upgrading waste for feeds and food. proceedings of previous easter schools in agricultural science

Proceedings of Previous Easter Schools in Agricultural Science, published by Butterworths, London

•SOIL Z O O L O G Y Edited by D . K. McL. Kevan (1955) *THE G R O W T H OF L E A V E S Edited by F. L. Milthorpe (1956) • C O N T R O L OF T H E P L A N T E N V I R O N M E N T Edited by J. P. Hudson (1957) • N U T R I T I O N OF T H E L E G U M E S Edited by E. G. Hallsworth (1958) -• T H E M E A S U R E M E N T OF G R A S S L A N D P R O D U C T I V I T Y Edited by J. D . Ivins

(1959) • D I G E S T I V E P H Y S I O L O G Y A N D NUTRITION OF T H E R U M I N A N T Edited by D .

Lewis (1960) • N U T R I T I O N OF PIGS A N D P O U L T R Y Edited by J. T. Morgan and D . Lewis (1961) •ANTIBIOTICS IN A G R I C U L T U R E Edited by M. Woodbine (1962) • T H E G R O W T H OF T H E P O T A T O Edited by J. D . Ivins and F. L. Milthorpe (1963) • E X P E R I M E N T A L P E D O L O G Y Edited by E. G. Hallsworth and D . V. Crawford (1964) • T H E G R O W T H OF C E R E A L S A N D GRASSES Edited by F. L. Milthorpe and J. D .

Ivins (1965) • R E P R O D U C T I O N IN T H E F E M A L E M A M M A L Edited by G. E. Lamming and E. C.

Amoroso (1967) • G R O W T H A N D D E V E L O P M E N T OF M A M M A L S Edited by G. A Lodge and G. E.

Lamming (1968) • R O O T G R O W T H Edited by W. J. Whittington (1968) • P R O T E I N S A S H U M A N F O O D Edited by R. A . Lawrie (1970) • L A C T A T I O N Edited by I. R. Falconer (1971) •PIG P R O D U C T I O N Edited by D . J. A . Cole (1972) • S E E D E C O L O G Y Edited by W. Heydecker (1973)

H E A T LOSS F R O M A N I M A L S A N D MAN: ASSESSMENT A N D C O N T R O L Edited by J. L. Monteith and L. E. Mount (1974)

• M E A T Edited by D . J. A . Cole and R. A . Lawrie (1975) •PRINCIPLES OF CATTLE P R O D U C T I O N Edited by Henry Swan and W. H. Broster

(1976) • L I G H T A N D P L A N T D E V E L O P M E N T Edited by H. Smith (1976)

P L A N T PROTEINS Edited by G. Norton (1977) ANTIBIOTICS A N D ANTIBIOSIS IN A G R I C U L T U R E Edited by M. Woodbine (1977) C O N T R O L OF O V U L A T I O N Edited by D . B. Crighton, N. B. Haynes, G. R. Foxcroft and G. E. Lamming (1978) P O L Y S A C C H A R I D E S IN F O O D Edited by J. M. V. Blanshard and J. R. Mitchell (1979) S E E D P R O D U C T I O N Edited by P. D . Hebblethwaite (1980) PROTEIN DEPOSITION IN A N I M A L S Edited by P. J. Buttery and D . B. Lindsay (1981) P H Y S I O L O G I C A L PROCESSES LIMITING PLANT P R O D U C T I V I T Y Edited by C. Johnson (1981) E N V I R O N M E N T A L ASPECTS OF H O U S I N G FOR A N I M A L P R O D U C T I O N Edited by J. A . Clark (1981) EFFECTS OF G A S E O U S A I R POLLUTION IN A G R I C U L T U R E A N D H O R T I C U L T U R E Edited by M.H. Unsworth and D.P . Ormrod (1982) C H E M I C A L M A N I P U L A T I O N OF CROP G R O W T H A N D D E V E L O P M E N T Edited by J. S. McLaren (1982) C O N T R O L OF PIG R E P R O D U C T I O N Edited by D.J .A. Cole and G.R. Foxcroft (1982)

S H E E P P R O D U C T I O N Edited by W. Haresign (1983)

• The titles are now out of print but are available in microfiche editions

upgrading Waste for Feeds and Food D.A. LED WARD, MSc ,PhD ,FiFST A.J. TAYLOR, BSC, PhD R.A. LAWRIE, B S C , Pho, DSC , SCD, F R S E , F R S C , F I F S T

University of Nottingham School of Agriculture

BUTTERWORTHS London Boston Durban Singapore Sydney Toronto Wellington

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, without the written permission of the copyright holder, application for which should be addressed to the Publishers. Such written permission must also be obtained before any part of this publication is stored in a retrieval system of any nature.

This book is sold subject to the Standard Conditions of Sale of Net Books and may not be re-sold in the UK below the net price given by the Publishers in their current price list.

First published 1983

© The several contributors named in the list of contents 1983

British Library Cataloguing in Publication Data

Upgrading waste for feeds and food. 1. Waste products as feed—Congresses L Ledward, D . A . Π. Taylor, A.J. in. Lawrie, R.A. 338.1'6 SF95

ISBN 0 - 4 0 8 - 1 0 8 3 7 - 1

Library of Congress Cataloging in Publication Data Main entry under title:

Upgrading waste for feeds and food.

Based on the 36th Easter School in Agricultural Science.

Bibliography: p. Includes index. 1. Agricultural wastes—Congresses. 2. Waste

products as feed—Congresses. 3. Food industry and trade—Congresses. I. Ledward, D . A . II. Taylor, A.J. (Andrew John), 1951- . III. Lawrie, R . A . (Ralston Andrew) IV. Easter School in Agricultural Science (36th : 1982? : University of Nottingham) TP995.A1U63 1983 664'.096 83-7548

Typeset by Scribe Design Ltd, GiUingham, Kent Printed and bound by Robert Hartnoll Ltd, Bodmin, Cornwall

PREFACE

It is now many decades since the world was alerted to the possibihty that the number of human beings might increase beyond the capacity of the available nutrients to feed them.

There have been, and there wih continue to be, vigorous and successful attempts by agriculturalists to produce more food, but this is not the only approach to the problem. Large quantities of nutrients are wasted after they have been produced because they are unpalatable or because they have been improperly stored. Moreover, insofar as such waste contributes to environmental pollution, it is doubly undesirable.

It was the purpose of the 36th Easter School in Agricultural Science to consider how currently wasted or underutilized nutrients could be recovered and upgraded in order to make available more food, either directly or through animal intermediaries; and to assess what progress had already been made in seeking a solution to this problem.

The various chapters in this volume are the contributions made at the School by invited experts. The editors hope that readers will find in this volume the breadth and depth of coverage necessary to appreciate this field of scientific endeavour, which is increasingly important and of concern to all.

ACKNOWLEDGEMENTS

The editors are glad to take this opportunity of acknowledging the expertize and efforts of all those who contributed papers at the Easter School.

They are also indebted to the following gentlemen who kindly acted as session chairmen: Sir David Cuthbertson, CBE, formerly Honorary President of the British Nutrition Foundation Ltd; Dr H. Egan, Government Chemist, 1970-81; Professor R.F. Curtis, Director, ARC Food Research Institute, Norwich; Dr W.J.F. Cuthbertson, OBE, Consultant, London; Dr R.B. Hughes, Technical Director, C T . Harris (Calne) Ltd; and Professor A.E. Bender, Professor of Nutrition, Queen Elizabeth College, University of London.

The University of Nottingham wishes to express its gratitude for the generous financial contributions of the following organizations. These assisted in meeting the costs of bringing overseas speakers to the School.

Albright & Wilson Ltd Alginate Industries Ltd Batchelors Foods Ltd The British Council The British Petroleum Company Ltd Imperial Foods Ltd Pedigree Petfoods Ltd Pork Farms Ltd Purina Protein Application Ltd Seymour, Arthur H. & Son Ltd Shell Research Ltd United Biscuits (UK) Ltd Walkers Crisps Ltd

In conclusion, the editors wish to thank warmly all those members of staff and students at the School of Agriculture who gave their time in the interests of the Symposium. The help of Mrs D.M. Borrows, Mrs B.E. Dodd, Mrs D. Treeby, Mr G. Millwater, Mr P. Glover and Mr J. Rosillo was particularly appreciated.

VI

1

Introduction

At no time in human history have food shortages been as widespread and affected as many people as they do today. An estimated five hundred milHon people in the world are malnourished (NAS, 1977)—and the food supply problem will become more severe as the world population rapidly grows from the present level of nearly five thousand million to 12 to 16 thousand million by 2150 (UN, 1973). We know humans must have an adequate amount of food and that these foods must contain the many nutrients essential to sustain life. The basic problem then, is how to provide such a food supply in the face of increasing populations and diminishing resources needed to produce this food.

Therefore, in planning for the coming decades we need to consider not only the present conditions affecting food production, but the many constraints that may impede our achieving these goals in the future. Thus the interplay among population growth, energy resources, land availability, water supplies and use of biological wastes needs to be examined. Only when these interrelationships are clearly understood will we be able to make viable plans for the future.

World population

For 99% of the time that humans have inhabited the earth, the world population numbered less than eight miUion (Coale, 1974), and the total population of North America numbered less than 200000. Now every day more than 200000 humans are added to our rapidly growing numbers so it is no wonder the human population is projected to increase to 6.5 thousand million by the turn of the century. Numerous studies hke that of the National Academy of Sciences pessimistically state there is no feasible way to stop the explosive increase of the world population short of some catastrophe (NAS, 1971). To provide food to feed the rapidly growing numbers of humans during the next 25 years will require a doubling of world food supply.

Probably one of the most important factors responsible for the population explosion has been the escalating use of fossil energy {Figure 1.1).

WORLD OUTLOOK FOR FOOD

DAVID PIMENTEL College of Agriculture and Life Sciences, Cornell University, USA and MARCIA PIMENTEL Division of Nutritional Sciences, College of Human Ecology, Cornell University, USA

4 World outlook for food

1 6 0 0 1 7 0 0 1 8 0 0 1 9 0 0 2 0 0 0 2 1 0 0 2 2 0 0 2 3 0 0 2 4 0 0 Y e a r s AD

Figure 1.1 Estimated world population numbers ( ) from 1600 to 1975 and projected numbers ( ) (?????) to the year 2250. Estimated fossil fuel consumption ( — ) from 1650 to 1975 and projected ( ) to the year 2250 (after Pimentel et al., 1975)

Basically, increased food production and more effective control of human diseases, have contributed most to the alarming growth of world population (NAS, 1971). Of the two, evidence suggests that reducing death rates through effective public health programs has contributed the most to increased population growth (Freedman and Berelson, 1974). For example, in Mauritius, eradication of malaria-carrying mosquitoes by using DDT, a fossil-based pesticide*, produced a dramatic reduction in death rates (PEP, 1955; UN, 1957-1971). In just one year, death rates fell from 27 to 15 per 1000 over a period of five years.

Then, because fertility rates did not decrease, an explosive increase in population has occurred.

Events in recent history document similar occurrences where medical technology and availabihty of medical supplies have significantly reduced

*To produce and apply 1 kg of D D T uses about 8 € of oil; 1 kg of D D T provides effective control for several months in about 70 small homes.

From 1800 to the early 1970s, fossil energy has been ample in supply and low in cost. As a result, industries have flourished; agriculture has become more productive through mechanization, but more dependent on pesticides and fertilizers; human disease control operations have been more successful; and unfortunately military armaments have become more deadly.

David Pimentel and Marcia Pimentel 5

Energy resources for agriculture

Energy use in agricultural production has evolved and changed over the thousands of years humans have cultivated the earth. As human numbers increased, many regions could no longer support the primitive hunting-gathering economy and a shift was made to a more permanent type of agriculture (Boserup, 1965). 'Slash and burn' or 'cut and burn' agriculture (i.e. cutting trees and \)rush and burning them on site) was the first agricultural technology used. Because this practice killed weeds and added nutrients to the soil, crop production was satisfactory for two to three years. Then soil nutrients became depleted and about 20 years had to elapse before the forests regrew and soil nutrients were renewed.

Cut and burn crop technology required an ax and hoe and much manpower. For example, Lewis (1951) who investigated 'slash and burn' corn culture in Mexico, reported that a total of 1144 h of labor was required to raise a hectare (ha) of corn {Table 1.1). Other than human energy, the only other inputs were the ax, hoe and seeds. This corn yield of

Table 1.1 E N E R G Y INPUTS IN C O R N P R O D U C T I O N IN MEXICO USING O N L Y M A N P O W E R (PIMENTEL A N D PIMENTEL, 1979)

Inputs Quantity/ha kJ/ha kcal/ha

Labor 1144 h 2462690 589160 A x + hoe 69260 kJ 69260 16570 Seeds 10.4 kg 153020 36608

Total 2684970 642338 Corn 1944 kg 28847020 6901200 kJ output/kJ input

1944 kg 10.74

1944 kg/ha provided about 28.8 miUion kilojoules (kJ) (6.9 million kcal) of food.

Gradually humans have augmented their own power with other sources of energy, first animals, then wood and coal. But it wasn't until the twentieth century that fossil fuel became the dominant fuel, especially in the industrialized nations. Now, in these countries, fossil energy powers crop and livestock production and is as vital an agricultural resource as land and water.

death rates (Corsa and Oakley, 1971). Based on experience, the inevitable conclusion is that it is relatively easy to reduce death rates, but birth rates are difficult to curtail because they are dependent on multidimensional factors and deeply rooted social customs. Consequently, our efforts must be focused not only on population control, but must be redoubled to find ways to augment a nutritious food supply. The latter aim is the focus of this discussion.


Table 1.2 E N E R G Y INPUTS PER H E C T A R E FOR C O R N P R O D U C T I O N IN T H E U S A (PIMENTEL A N D B U R G E S S , 1980)

Quantity/ha kJ X Κ

Inputs Labor 8.05 h Machinery 55 kg 4.14 Gasoline 26.96 € 1.14 Diesel 78.45 € 3.74 Liquefied petroleum gas 34.54 € 1.11 Electricity 3L62 kWh 0.38 Nitrogen 135 kg 8.30 Phosphorus 65.04 kg 0.82 Potassium 95.32 kg 0.64 Lime 354.35 kg 0.47 Seeds 23.79 kg 2.49 Insecticides 2.47 kg 0.90 Herbicides 5.14 kg 2.15 Transportation 186.04 kg 0.20 Total 26.48

Output Total yield 8000 kg 117.04 kJ output/kJ input 4.42

Yearly about 1500 £ of oil are expended to produce, process, distribute and prepare the food for each American. Collectively this represents about 17% of the total energy used in the USA each year (Pimentel and Pimentel, 1979). Agricultural production uses only about 6% of total energy, while food processing, packaging, transport, storage and home preparation together use the remaining 11%.

For example, to raise 1 ha of corn, a typical grain crop, in the USA, approximately 600 € of gasoline equivalents are required and this is equivalent to an expenditure of about 1 € of gasoline per 9 kg of corn produced. Or put another way, about 4 kJ of corn are produced for each kJ of fossil energy expended {Table 1.2). For corn, approximately one-third of the fossil energy is used to make fertihzers and another one-third is used to power the various farm machines. For most grain production in the USA only 0.25 to 0.5 kJ of fossil energy are expended per kJ of food produced.

Of course, manpower is still used, but it is a relatively small input. Under present mechanized systems, only about 8 h of on-farm labor are required to produce 1 ha of corn compared with producing corn by hand, which requires about 1200 h of labor. This is more than a 100-fold difference {Tables LI and i.2).

Although fossil energy is expended in many phases of food production, the major uses of energy in actual crop production are for the fuel to run farm machinery and for the manufacture of fertilizers and pesticides {Table 1.2). Both pesticides and nitrogen fertilizers are produced directly from fossil energy. Pesticides are made primarily from petroleum, while nitrogen fertilizer is made from natural gas.


Industrial

Transport

Food s y Stenn

Connnnercial

Residential

E N E R G Y U S E IN F O O D S Y S T E M

Processing and

packaging

Distribution and

preparation

Agriculture

Figure 1.2 Percentage of total energy used in the U S economy and the proportion expended specifically for agricultural production, processing and packaging, and distribution and preparation

Producing other food crops, however, is not as energy efficient as grain production. For example, in apple and orange production, 2-3 kJ of fossil energy are expended per kJ of food produced (Pimentel, 1980) and in culturing vegetables 1-5 kJ of energy are expended per food kJ produced.

Although fruits and vegetables require larger energy inputs per food kJ than grain, neither are as energy-expensive as producing animal protein. From 10 to 90 kJ of fossil energy are expended to produce 1 kJ of animal protein (Pimentel et ai, 1980). Animal protein products are significantly more energy-expensive than plant protein because forage and grains have first to be grown, then consumed by animals, who in turn are used as human food. The forage and feed that maintain the breeding herd are additional energy costs. At present in the USA about 90% of the grain produced is cycled through livestock to produce the milk, eggs and meat that consumers prefer (Pimentel et aL, 1980).

Yet many of these grains are entirely suitable for human food. Thus an important consideration for future planning would be to use the grains directly as food and thereby decrease energy expenditure.

Fertilizer is responsible for costly energy inputs in modern agriculture and therefore ways to reduce this energy expenditure, while adding necessary nutrients to farm lands, need to be developed. One way that

T O T A L E N E R G Y U S E


Land resources

One obvious way to increase the food supply is to bring more land into production. Although energy expenditure is involved in this choice and may be a constraint, let us consider what our land resources really are. Worldwide, the arable land now available per person is estimated to be 0.35 ha (USDA, 1978). Increasing the world population to a projected high of 12 to 16 thousand million by the year 2150 (UN, 1973) would reduce the arable land per capita to slightly more than 0.1 ha. This estimate assumes no more agricultural land will be lost either because of soil erosion or because of population pressure for housing and highways. Based on the rate of land degradation presently occurring in the USA, a more reahstic estimate would be that less than 0.1 ha per capita will be available by the year 2150.

Unfortunately, worldwide environmental degradation of land is even more severe than in the USA, mainly because of high rates of erosion and to a lesser extent because of population pressure for highways and housing (Eckholm, 1976). The soil erosion problem in the developing countries of the world is estimated to be nearly twice that of the USA (Ingraham, 1975) and can be expected to intensify as the need for food increases. Even now, more marginal land, especially that with steep slopes, is being cultivated for crops; forests are rapidly being removed for fuel (Eckholm, 1976); and deserts continue to advance on good land, partly because of overgrazing

holds great promise is the more effective use of manure, often considered an agricultural waste. Although 90% of livestock manure produced is applied to US agricultural land, a large portion of this goes to waste. Conservative estimates are that more than 50% of the nutrients, particularly valuable nitrogen, are lost throughout the year because Hvestock manure is spread directly from the barn and feedlot onto the land (Muck, 1982). Then, too, during the eight-month fallow season typical of the temperate zones, rain and snowfall leach much of the nitrogen from the manure and wash it into adjacent streams and lakes. This constitutes a loss of nutrients from the soil, and in addition, the leached chemicals cause serious water pollution problems. More effective systems of fertilizer application would reduce the amount of fossil-based fertilizer needed and thereby cut energy expenditures used to make commercial fertilizers and thus cut the cost to the farmer.

Thus far discussion has centered on energy expenditure in actual production of food but energy, especially fossil energy, is expended in many other segments of the food system.

Transportation, storage and preservation as well as home preparation all use a share of energy expended for food production {Figure 1.2). In the future when food supplies must be increased, energy expenditure used to produce them will also escalate. Meanwhile fossil energy supplies will diminish and continue to be more costly than at present. Our choice will be to find and use fuels other than these fossil-based ones and also to find ways to make agricultural productivity and the entire food system more energy efficient.


The land-energy tradeoff

A fundamental interdependence exists between energy and land resources. For example, the yield of corn from 2 ha of land, with energy inputs of about 9.2 million kJ/ha is about 2500 kg/ha (Pimentel and Pimentel, 1979). To achieve the same corn yield (5000 kg) on 1 ha of land, the energy inputs would have to be increased to about 27.2 million kJ (Pimentel, 1980). Thus, to reduce the land area by half yet maintain total corn yield, about three times more fossil energy must be expended.

Certainly the loss of arable land to urbanization and erosion will continue to affect food production adversely. As mentioned, some marginal land will be put into production, but at the cost of increased fossil energy expenditure for fuel, fertilizers and pesticides. For desert areas the additional resources will be water, plus fossil fuel to pump and distribute it. Thus both nonrenewable energy and water supplies are major constraints to cultivating marginal lands.

Project these multiple problems to the future world situation, where doubling of food production on current land resources in less than 25 years would require about a threefold increase in energy for agriculture (RSAS, 1974). Indeed, it is unHkely that our known fossil energy resources are adequate to meet this need. This is because developing countries already use more than 60% of their energy, including wood, for food production (RSAS, 1974) and worldwide about 25% of all energy, including wood, is spent in the food system (Pimentel, 1974). Thus, the outlook for doubling food production on existing arable land is not promising, especially when balanced against diminishing energy supplies of both fossil and fuel wood.

Water resources

Energy, land and water are all essential to food production and ultimately to human life itself. Of these, water is the most vital. Humans must consume more than 1 € of water per day for survival (Pimentel et al., 1982). Crops require and transpire massive amounts of water. For instance, a corn crop that produces 6000 kg/ha of grain will take up and transpire about 9 million ( of water per ha during the growing season (Penman, 1970). To pump and apply this much irrigation water would require about 900 ( of oil.

In the USA about 7200 €/day of water per person are withdrawn from rivers, lakes, and underground aquifers to maintain the high standard of living. Some of this is used directly by individuals and industry, but almost

(Kassas, 1970; Ormerod, 1976). All these situations foster intense erosion, which accelerate the deterioration of the soil.

Although in the future some marginal land undoubtedly will be put into production, the inevitable conclusion is that available land resources in the world are far from unlimited, and that increased food production will have to be carried out primarily on the land we already are using.


72% is used to supply food and energy needs (Murray and Reeves, 1977; USWRC, 1979). Specifically, the production of about 2 kg of food consumed per person each day requires about 2700 i of pumped water.

Irrigated crop production requires large quantities of water. For example, to produce 1 kg of food and fiber under irrigation in California requires: 1420 t for corn; 1920 € for sugar (sugar beets); 4670 € for rice; 17200 t for cotton (Ritschard and Tsao, 1978). Water needed to produce 1 kg of grain-fed meat is calculated to be from 4200 to 8300 t when the water input for irrigated grain is included.

Not only is water an essential resource in irrigated crop production, but also large quantities of energy are required to pump and distribute it to the crop. For example, the production of 1 ha of corn with irrigation water pumped from a depth of only 30 m requires about three times more energy than producing the same quantity of corn under normal rainfed conditions (Pimentel and Pimentel, 1979; Pimentel and Burgess, 1980).

Although at present the amount of water withdrawn per capita on a global basis is less than one-third of US per capita use, the growth in world population can be expected to double water demand by the year 2000 (CEQ, 1980). In addition, increased agricultural production is projected to consume more water than at present, and will probably increase to about 64% of total water withdrawn. This is alarming, especially when we know that even now some 80 countries, which account for nearly 40% of the world population, are experiencing serious drought (Kovda, Rozanov and Onishenko, 1978).

Although the volume of water resources is certainly crucial, its location will influence how effectively it can be used. For instance, water in some mountainous areas of the western USA is relatively inaccessible for irrigation use on land suitable for crop production. Furthermore, the competition for water will extend beyond national boundaries to conflicts between nations that share common water supplies. Because one-third of the major river basins in the world are shared by three or more countries (CEQ, 1980), it is likely that disputes will arise as need for water grows.

Food waste and losses

Lost or wasted food amounts to lost food energy and nutrients and represents no return for energy invested in production. Worldwide too much of our food crops is being lost directly to pest attack, simply wasted, or not used because potentially valuable waste products are not recycled into usable food products.

The magnitude of these losses has become critical in view of the need to augment future food production. Indeed it is timely that the concern of this book is focused on waste, ways to reduce food losses and ways to convert waste into usable food.

At present about half of the world food is lost or wasted. Estimates are that worldwide pre-harvest losses to pests, primarily insects, pathogens and weeds, range as high as 35% (Pimentel, 1976). After harvest, an additional 10-20% of food is lost to fungi, bacteria, insects and rodents. Safer, more effective pest control and more effective storage will undoubtedly help curtail these losses.


Conclusion

The need for increased food supplies and the magnitude of the need will depend on the status of the human population and the standard of living and type of diet desired by that population. Whether the anticipated need can be met will be determined largely by water supphes, the amount of arable land available for crop production as well as supplies of reasonably priced energy. Other interacting factors affecting the human food supply will continue to be stringent pest controls, reduction in table wastes, and improved use of waste products that have the potential either for improving crop production or being used directly as food.

Vast as it seems, the earth ecosystem is not inexhaustible. Many of its most precious resources are finite and irreplaceable. And many of these resources are being used unwisely in the agricultural production systems of today.

Science and technology will help overcome some of the food and other crises facing humankind as numbers rapidly increase, but the obvious solution is effective, organized population control. Clearly if humans do not control their numbers, nature will.

References

B O S E R U P , E. (1965). Conditions of Agricultural Growth. 124 pp. Aldine-Atherton, Chicago

CEO. (1980). The Global 2000 Report to the President. Council on Environmental Quality and the Department of State. Vol. 2. US Government Printing Office, Washington, DC

C O A L E , A.J. (1974). Scientific American, 231(3), 40

In addition in the USA, and probably in most industrial countries of Europe, about 14% of the edible food that reaches the table is wasted simply by being discarded in the garbage (Harrison, Rathje and Hughes, 1975).

Also responsible for some waste have been the stringent 'cosmetic standards' set for many foods in the USA. Presently, apples, oranges, broccoli and lettuce, for example, must be almost blemish-free. Raising the cosmetic standards has increased food waste because previously acceptable foods are no longer allowed in the marketplace.

This, in part, is documented by the fact that despite a more than tenfold increase in the use of insecticides in the USA during the last 35 years, insect losses have not decHned. Instead losses to insects have nearly doubled from about 7% in 1945 to 13% to date (Pimentel et aL, 1978). Estimates are that 10-20% ofnhe insecticide presently appHed to US crops is used just to maintain the high 'cosmetic standards' set for US food (Pimentel et ai, 1977). There are, in addition, clear public health and environmental hazards associated with the increased use of insecticides on food.


C O R S A , L. and O A K L E Y D . (1971). Consequences of population growth for services in less developed countries—an initial appraisal. In Rapid Population Growth. Vol. 11. pp. 368-402. Research Papers, National Academy of Sciences. Johns Hopkins Press, Baltimore, MD

E C K H O L M , E.P. (1976). Losing Ground. Environmental Stress and World Food Prospects. Norton, New York

F R E E D M A N , R. and B E R E L S O N B. (1974). Scientific American, 231(3), 30 H A R R I S O N , G.G. , R A T H J E , W.L. and H U G H E S , W.W. (1975). / . Nutr. Educ., 7,

13 I N G R A H A M , E.W. (1975). A query into the quarter century. On the interrela

tionships of food, people, environment, land and climate. Wright-Ingraham Institute, Colorado Springs, CO

K A S S A S , M. (1970). Desertification versus potential for recovery in circum-Saharan territories. In Arid Lands in Transition, pp. 123-142. American Association for the Advancement of Science, Washington, DC

K O V D A , V . A . , R O Z A N O V , B.G. and O N I S H E N K O , S.K. (1978). On probability of droughts and secondary salinization of world soils. In Arid Land Irrigation in Developing Countries, pp. 237-238. Ed. by E.B. Worthing-ton, Pergamon, London

LEWIS, O. (1951). Life in a Mexican Village: Tepoztlan Restudied. University of Illinois Press, Urbana

M U C K , R .E . (1982). Personal communication. Dept. of Agricultural Engineering, Cornell University, Ithaca, NY

M U R R A Y , R.C. and R E E V E S , E .B . (1977). Estimated use of water in the United States in 1975. US Geol. Surv. Circ. 765

N A S (1971). Rapid Population Growth. Vols. 1 and 2. Johns Hopkins University Press, Baltimore, MD

N A S (1977). Supporting Papers: World Food and Nutrition Study. Vol. II. National Academy of Sciences, Washington, DC

O R M E R O D , W.E. (1976). Science, 191, 815 P E N M A N , H.L. (1970). Scientific American, 223(3), 99 PEP (1955). World Population and Resources. Political and Economic

Planning, London PIMENTEL, D . (1974). Environ. Biol. Rept. 74, Cornell University, Ithaca,

NY PIMENTEL, D . (1976). Bull. Entomol. Soc. Am., 22, 20 PIMENTEL, D . (ed.) (1980). Handbook of Energy Utilization in Agriculture.

CRC Press, Boca Raton, FL PIMENTEL, D . and PIMENTEL, M. (1979). Food, Energy and Society. Edward

Arnold, London PIMENTEL, D . and B U R G E S S , M. (1980). Energy inputs in corn production,

pp. 67-84. In Handbook of Energy Utilization in Agriculture. Ed. by D. Pimentel. CRC Press, Boca Raton, FL

PIMENTEL, D . , DRITSCHILO, W., K R Ü M M E L , J. and K U T Z M A N , J. (1975). Science, 190, 754

PIMENTEL, D . , T E R H U N E , E . , DRITSCHILO, W., G A L L A H A N , D . , K I N N E R , N . , N A F U S , D . , P E T E R S O N , R., Z A R E H , N . , MISITI, J. and H A B E R - S C H A I M , O. (1977). Bioscience, 27, 178

PIMENTEL, D . , K R Ü M M E L , J., G A L L A H A N , D . , H O U G H , J. MERRILL, Α . , S C H R E I N E R , I., VITTUM, P., KOZIOL, F. , BACK, E . , Y E N , D . and F I A N C E , S. (1978) Bioscience, 28, 772, 778


PIMENTEL, D . , O L T E N A C U , P .A. , NESHEIM, M.C. , K R Ü M M E L , J., A L L E N , M.S.

and CHICK, S. (1980). Science, 207, 843 PIMENTEL, D . , FAST, S., C H A O , W.L. , S T U A R T , E . , DINTZIS, J., E I N B E N D E R , G.,

SCHLAPPI, W., A N D O W , D . and B R O D E R I C K Κ. (1982). Bioscience, 32, 861 R I T S C H A R D , R.L. and T S A O K. (1978). Energy and water use in irrigated

agriculture during drought conditions. Lawrence Berkeley Laboratory, University of California, Berkeley

RSAS (1975). Energy uses. Royal Swedish Academy of Sciences. Energy Conference, Aspenasgarden, Oct. 27-31. Manuscript.

U N (1957-71). Statistical Yearbooks. Statistical Office of the United Nations, Department of Economic and Social Affairs, New York

U N (1973). World Population Prospects as Assessed in 1968. Population Studies, No. 53. United Nations, New York

U S D A (1978). Agricultural Statistics 1978. US Government Printing Office, Washington, DC

U S W R C (1979). The Nation's Water Resources, 1975-2000. Vols. 1-4. Second National Water Assessment. United States Water Resources Council. US Government Printing Office, Washington, DC

2

Introduction

Food wastage occurs at all points along the food chain as a result of agricultural practices, of food manufacture and in final consumption of food by people. It is not the intention here to review all wastage in the agriculture/food complex but to concentrate on sources of food waste as they occur after the primary stage of processing agricultural raw materials, i.e. post-slaughter in the case of animals and beyond the farm-gate in the case of crops. Significant food loss, as opposed to agricultural waste production, does occur, of course, before commodities leave the farm and such sources which are significant will be mentioned briefly. Much of the information which follows was obtained by a Food Waste Survey Unit set up in 1976 by the Ministry of Agriculture, Fisheries and Food (MAFF) 'to collect and collate data on waste as it arises from the point at which food commodities enter into food processing, distribution and consumption, and for reviewing ways in which waste may be reduced or may be recycled within the food chain or otherwise usefully employed'.

Wastage losses before food processing

Precise information on total losses of agricultural raw materials before processing is limited. In the European context statistics on losses are published by the EEC Commission (Eurostats) and a summary of figures published recently for a few agricultural commodities are shown in Table 2.1. In these statistics loss is defined as any production which cannot be sold for its market value. Each EEC country interprets this definition in its own way so that the figures probably present an order of loss rather than a precise estimate. In a total EEC production of some 35 million tonnes of wheat 500000 tonnes is considered to be lost before processing. For potatoes the losses are higher, in the order of 4% of production. Some 10% of fresh vegetables and fruit produced in the EEC are wasted before marketing and in the case of apples this amounts to about 500000 tonnes.

In considering losses or wastage from plant crop production it is important to distinguish between 'crop residues' and 'crop wastage'. The former refers to that part of the plant which is removed with the crop but

15

SOURCES OF FOOD WASTE—UK AND EUROPEAN ASPECTS

A. TOLAN Ministry of Agriculture, Fisheries and Food, London, UK

16

EEC UK Production Loss Production Loss

(Ό00 tonnes) (Ό00 tonnes)

Wheat 35395 538 9027 34 Fresh vegetables 29016 2793 3287 39 Fresh fruit (excluding citrus) 14401 1377 575 — Apples 5857 493 334 — Potatoes 33784 1652 7237 —

Table 2.2 UK F R U I T A N D V E G E T A B L E P R O D U C T I O N A N D E S T I M A T E D W A S T E ARISINGS (1970-71) ( A D E R A N D PLASKETT, 1975)

Salable output

Crop wastage

Crop residue

(Ό00 tonnes)

Industrial process waste

Vegetables Potatoes 5882 1083 149 380 Cabbages, savoys, broccoli 602 74 544 — Carrots 469 71 314 62 Peas 314 12 2329 65 Cauliflowers 279 12 299 — Brussels sprouts 231 37 698 2 Lettuces 136 13 34 — Turnips, swedes 127 10 80 — Onions (dry bulb) 111 9 11 — Beetroot 111 4 63 19 Tomatoes 106 — — — Runner and french beans 100 5 175 10 Navy beans — — — 48 Fruit Apples 512 91 — 3 Plums 72 10 — 1 Pears 69 11 — — Strawberries 44 1 — 2 Blackcurrants 20 1 — — Raspberries 18 1 — 1

Table 2.3 T O T A L S L A U G H T E R I N G W A S T E FOR I N D I V I D U A L COUNTRIES OF T H E E E C IN 1973 (WEIERS A N D FISCHER, 1978)

('000 tonnes)

FRG 2035.0 France 1932.5 Italy 1333.8 Netherlands 461.5 Belgium 443.5 Luxembourg 9.3 U K 1409.6 Ireland 281.7 Denmark 418.8

Table 2.1 SUPPLY A N D LOSS OF SOME A G R I C U L T U R A L COMMODITIES IN T H E EEC A N D UK IN 1980 ( E U R O S T A T S , 1981)

Α. ΤοΙαη 17

Waste material (tonnes)

Protein content (tonnes)

Blood 1 0 0 0 0 0 1 7 0 0 0

Head meat (sheep) 3 0 0 0 7 0 0

Lungs 2 5 0 0 0 3 6 0 0

Spleen 6 0 0 0 1 1 0 0

Oesophagus 2 0 0 0 3 4 0

Brain 1 4 0 0 - 4 7 0 0 1 5 0 - 5 1 0

Cattle stomach 2 3 5 0 0 2 6 4 0

Sheep stomach 9 0 0 0 1 5 0 0

Pig stomach 3 0 0 0 - 4 0 0 0 4 0 0 - 6 0 0

Pancreas 6 0 0 - 8 0 0 — Large intestine 7 0 0 0 0 — Small intestine 1 5 0 0 0 — Feet 3 0 0 0 0 - 4 0 0 0 0 —

The total amount of proteinaceous material available was estimated to be about 380000 tonnes of which about 300000 tonnes could possibly be upgraded for human consumption. At an average protein content of about 16% this is equivalent to about 45000 tonnes of protein. Any up-graded waste would of course need to comply with food regulations if used as food.

plays no part in its utilization. Thus crop residues have no value in edible food terms but may form suitable substrates for further recycling or upgrading. Crop wastage refers to the unusable or surplus crop as exemplified by the figures shown above. A breakdown of the various waste arising from fresh vegetable and fruit production in the UK is given in Table 2.2, For all vegetables with the exception of potatoes, crop residues are higher and in many cases much higher than crop wastage.

On-farm storage can result in substantial wastage of some vegetables. For example in the UK loss of carrots averages 30%, potatoes 8-10%, onions 11% and winter cabbage and cauliflowers 25-45%.

The largest and most important source of wastage occurring at the primary stage of food production is from the slaughter of animals for meat. Figures published by the EEC Commission for slaughterhouse waste in member states are given in Table 2.3.

The figures include the contents of animal organs arising from slaughter in abattoirs but do not include edible by-products such as heart, liver, tongue, etc. which are marketed with the meat. They also include animals dying on their way to slaughterhouses and meat condemned by official inspection as unfit for human consumption. Some wastes are already upgraded for human use, for example, small intestines are used in the manufacture of sausage casings but most is disposed of as low value products such as fertilizer, pet-food, animal feed. A detailed survey of waste arising in UK abattoirs has been published (Richards, 1978). In this survey the amounts of high-grade protein materials going to pet-food, fertilizer, rendering and which could be upgraded if a suitable process were available was estimated as shown in Table 2.4.

Table 2.4 Q U A N T I T Y O F A B A T T O I R M A T E R I A L S A T P R E S E N T D I S P O S E D O F

A S L O W V A L U E P R O D U C T S W H I C H C O U L D B E U P G R A D E D ( R I C H A R D S , 1 9 7 8 )

18 Sources of food waste - UK and European aspects

Estimating food wastage

More than 30 years ago it was observed (FAO, 1952) that in developed countries a considerable gap always existed between estimated per capita supplies at the retail level and calorie requirements as estimated by a recognized system. The reasons for this gap were not fully understood. Various attempts have been made over the years to reconcile the figures which account for the gap (e.g. Baines and Holhngsworth, 1961). In the UK it has been estimated (Roy, 1976) that on a weight basis only 79% of the food supplies entering the food chain from farm-gate or dockside are usefully consumed. The largest source of food wastage was in the home

primary supply level (CLE • a lcohol )

L o s s e s in process ing distribution and retail

Gap I

Retail level

National Food Survey

9 .54

Gap Π

L o s s e s in home

13.01

Total l o s s e s ) 0 .96

approx.

Total

12.05 approx.

Total l o s s e s 2 .85

approx.

Intake level

9 .20 (daily intake ]

Losses in processing distribution retail and catering preparation

2.51 Other approx.

Losses of alcohol , s w e e t s , soft drinks, e t c . and catering p la te l o s s e s

Figure 2.1 Food energy: supply, purchase and intake data (All data in Μ J per head per day, 1976). *Other' includes alcohol, sweets, soft drinks and catering

So far as fish is concerned it has been estimated that some 275 000 tonnes of fish and fish offal are processed each year in the UK to yield about 50000 tonnes of protein meal (MAFF, 1982). Total landings of fish and shellfish amounted to about 1.13 million tonnes in 1980 (MAFF, 1981).

Α. ΤοΙαη 19

Waste in food processing

Physical waste in food processing would normally be regarded as that part of the food leaving the farm-gate or port which is not actually used. There are however two important distinctions to be drawn between types of physical waste. The first is between waste which has actual or potential market value and that which has not; and the second is between avoidable and unavoidable waste. Using these two broad distinctions physical waste can be categorized as follows:

(1) avoidable waste that has no value; (2) avoidable waste that has actual or potential value; (3) unavoidable waste that has no value; (4) unavoidable waste that has actual or potential value.

The distinction between avoidable and unavoidable waste is not absolute but can be regarded as reasonably fixed during any short-term period. The levels of unavoidable waste depend on the current state of technology, i.e. any physical waste which arises from industrial and distributive processes when the best-known commercial methods and technologies are used will be unavoidable until such time as even better methods are developed and

directly by the consumer, but the loss in the industrial sector was still of some magnitude.

An attempt has been made by MAFF to obtain more precise information on the points at which wastage occurs in the national food supply by carrying out an 'energy audit' of the various stages in the food chain from the primary supply level to the point of consumption. This analysis is presented schematically in Figure 2.1. Using figures from the Consumption Level Estimates (CLE) (MAFF, 1977b) it was estimated that supplies of food energy available amounted to 12.30 MJ (2940 kcal) per head per day. A further 0.7 MJ (170 kcal) was supplied by alcohoHc drinks giving an estimate of total food energy available of about 13.01 MJ (3110 kcal) per head per day. The average energy content of food purchased and consumed at household level obtained from the National Food Survey (MAFF, 1977a) amounted to 9.54 MJ (2280 kcal) per head per day. Purchases of alcoholic drinks, soft drinks, sweets, snacks and ice-cream and meals eaten out, which are not included in the National Food Survey, were 0.71 MJ (170 kcal), 0.17 MJ (40 kcal), 0.59 MJ (140 kcal), 0.17 MJ (40 kcal) and 0.92 MJ (220 kcal) respectively. The most recent estimates of actual energy intake of the UK population is 9.20 Μ J (2200 kcal) per head per day (DHSS, 1979). Thus there is a crude gap of 3.81 MJ (910 kcal) between the recommended food requirement and that available from total food supplies. This crude gap may be divided into two parts, i.e. Gap I representing losses between the primary supply level and the retail level and Gap II losses between the retail level and actual consumption of food. On this basis Gap II at about 2.85 Μ J (680 kcal) is considerably larger than Gap I at about 0.96 MJ (230 kcal) and the majority of waste would seem to occur in the home.


F L O U R A N D B A K E R Y P R O D U C T S

During the production of bread and flour confectionery products in the UK, small losses of food materials occur at several stages of processing. Each loss in itself is not sufficiently large to make itself amenable to up-grading but the overall loss is such that any reduction will yield economic benefits for the manufacturer. Losses during grain transport are less than 0.5%. Storage losses in mills due to pest infestation are unlikely to be greater than 1%. Prior to miUing approximately 2% (by weight) of the grain is removed during the 'gleaning' process and sold off as wheat feed at approximately the same price as the original wheat. In flour mills, mould and other infestation damage of flour is negligible and flour rejected by bakeries is reblended. 'Flour mill sweepings' are used for animal feed and less than 1% of flour is lost in flour mills. Bakery wastage due to dust loss, mis-shapen dough, mechanical breakdown, etc. account for less than 1% of in-factory loss. The major losses of the order of 2-3% occur at the packaging stage. Bread which is unsold at the retail level and returned to bakeries as 'stale returns' probably represents a maximum of 3% total bread production. The amount of waste bread sold locally to pig farmers

introduced. At any particular time, therefore, there will be in each sector of the industry some absolute level of unavoidable waste, the rest being, in principle, avoidable.

There are no accurate figures for total food processing waste either for the UK or Europe, but approximate estimates may be obtained by application of wastage factors to food supply and population statistics. The UK has a population of about 56 million people each consuming on average about 0.94 kg of 'soüd' food per day giving a total quantity of food eaten each year in the UK of about 19.2 million tonnes. About 70% of this total amount is processed by the food manufacturing industries. By equating calories 'lost' in food processing of about 7.3% (Gap I) with actual waste, the amount of food wasted in food processing may be estimated crudely to be about 980000 tonnes annually. Using a similar procedure another estimate of wastage in the UK food processing industry was 840000 tonnes (Crawford and Whitman, 1980). In that study some 550000 tonnes was accounted for by direct measurement of wastage in the major food manufacturing sectors; however not all sectors were included. By applying the same crude weighting factors to European statistics it is possible to estimate a total wastage in the EEC food processing industries of the order of 4.5 million tonnes annually.

In terms of up-grading food processing wastes for food or feed the possibiHties vary from sector to sector because sources of waste differ markedly in each sector. For example in grain and flour processing wastage occurs in small, mostly unquantifiable, amounts along the length of the processing chain whereas in milk processing wastes arise in large and measurable quantities. It will be apparent that in the latter case the possibilities for further use of wastage are considerably greater than in the former case. It is therefore worth considering in further detail the amounts of and ways in which waste arises in the major food processing sectors.

A, ΤοΙαη 21

D A I R Y P R O D U C T S

Large quantities of waste occur in the dairying sector. In 1980 some 780 million litres of skim-milk and about 800 million litres of whey were wasted in the UK. These wastes usually arise at the major dairies in large and relatively constant volumes and since they are already food-grade materials they are readily amenable to further processing. Dairy product wastes have become a major area for technical developments. If dairy waste production for Europe is as great as in the UK then the potential source of raw material in this sector for upgrading is large.

FATS A N D OILS

It has been estimated that the total fat lost during the manufacture of the main food product groups in the UK amounts to some 35000 tonnes annually (Crawford and Whitman, 1980). The largest wastage occurs in the manufacture of biscuits, chocolate confectionery, sausages and pies, crisps and snacks and in fish-finger production. Losses for each of these sectors was in the region of 2000-7000 tonnes per annum. Measures are taken to recover waste fat in food manufacture because it is a relatively expensive ingredient and being insoluble is fairly readily recoverable from the waste stream. Fat losses therefore take the form of small but steady 'leakage', often of an unavoidable or unrecoverable nature, during processing operations. Where a frying process is used during the production process fat losses are higher. Of total fat lost in all sectors about 13000 tonnes are recovered from unsalable food products and some 1400 tonnes are sent for re-refining by the frying industries. A quantity of fat is collected from catering outlets but this has been subjected to high temperatures for prolonged periods and might not be considered suitable for further use in food production. A considerable quantity of fat and oil is disposed of for animal feed in the form of spoiled food, and much of this will not have been used for high temperature frying. Using actual recovery and waste figures for the various manufacturing sectors it was estimated that some 16000 of the 35000 tonnes of total waste could not be accounted for.

S U G A R A N D C H O C O L A T E C O N F E C T I O N E R Y

The total amount of sugar lost, as a proportion of the national production of food containing sugar, is higher than in the case of fat. Sugar lost in food manufacture in the UK is estimated at about 131000 tonnes per annum (Crawford and Whitman, 1980). One reason for this is that sugar is a

from large bakeries is less than 2% of production. Approximately 20% of total UK flour production is used for the manufacture of cake and biscuits. Much of the material lost during flour confectionery manufacture is recovered. The main losses of finished product occur at the packaging and distribution stages. It is estimated that overall wastage in this sector is not more than 230000 tonnes of flour equivalent per annum.


C A N N I N G A N D F R E E Z I N G — F R U I T A N D V E G E T A B L E S

The main sources of wastage arise from unit operations of washing, size grading, trimming, cutting, sorting, pitting and peeling processes. Average losses in preparing vegetable and fruit for canning or freezing in the UK are shown in Table 2.5.

Higher wastage rates occur in vegetable processing than in fruit canning because of the large amounts of pods, peel, or outer leaves which are

Table 2,5 E S T I M A T E D W A S T E A N D P R O D U C T I O N O F C A N N E D A N D F R O Z E N

V E G E T A B L E S I N T H E U K ( 1 9 7 7 ) ( C H I P P I N G C A M P D E N F O O D P R E S E R V A T I O N

R E S E A R C H S T A T I O N , 1 9 8 2 )

Estimated Canned Frozen waste production production (%) ( Ό 0 0 tonnes

net can contents) ( Ό 0 0 tonnes)

Vegetables 3 1 . 7 Beans, runner and french 2 0 1 7 . 0 3 1 . 7

Beans, broad 8 2 (pod weight) 8 . 7 4 . 2

Beetroot 4 0 5 . 0 — Carrots 4 5 9 5 . 1 — Peas, fresh 8 7 1 . 3 123.8 Peas, processed 1 8 1 7 1 . 0 — Potatoes 3 0 2 0 . 8 1 3 7 . 9

Beans in tomato and other sauces — 3 5 4 . 3 — Fruit ' Gooseberries 10 2 . 2

Plums, damsons, etc. 8 8 . 0

Rhubarb 2 5 1 4 . 4

Strawberries 1 8 8 . 8

Other berries and currants 1 0 3 . 3

Table 2.6 C A N N E D V E G E T A B L E P R O D U C T I O N IN T H E E E C IN 1 9 7 1 ( INSTITUT N A T I O N A L D E R E C H E R C H E CHIMIQUE A P P L I Q U E E , 1 9 7 7 )

FDR Belgium France Netherlands Italy Belgium ( Ό 0 0 tonnes)

Peas 2 7 5 1 2 5 1 2 0 6 0

Beans, kidney and string 6 7 4 2 2 5 6 6 8 1 8

Other vegetables 1 2 2 7 8 3 2 6 2 3 3 8

All vegetables 2 1 6 1 7 1 8 3 3 1 1 1 1 1 6

Mushrooms 0 . 5 1 . 5 7 7 . 1 2 7 2

Tomato puree 1 . 2 — 2 7 . 1 — 1 4 0

Peeled tomatoes 0 . 5 — 2 8 . 7 — 3 9 0

cheaper ingredient than fat and is relatively difficult to recover from a dilute waste stream which may also contain other contaminants. The sugar confectionery and chocolate confectionery sectors sustain the greatest losses. Of total waste losses actual, accountable waste was estimated to be about 40%, recoverable waste to be about 23% leaving an unaccountable loss of 37%. It was suggested that a saving of half the raw material loss would double the profit in a factory operating on a 5% margin on turnover.

Α. ΤοΙαη 23

Production ( Ό 0 0 tonnes/annum)

Effluent discharge ( Ό 0 0 tonnes/annum of suspended soHds and organic matter)

France 3 1 1 9 1 7 6

F D R 2 0 3 7 1 2 3

Italy 1 1 7 2 6 0

Netherlands 6 9 6 4 1

Belgium/Luxembourg 6 1 6 4 0

U K 8 7 4 4 5

Denmark 3 1 5 1 7

Ireland 1 6 1 1 0

removed. Much of this waste will be inedible being used mainly for animal feed; however there is a potential for upgrading since the waste arises in large volumes at a few sources. Canned vegetable production in various EEC countries is shown in Table 2,6.

The canning of vegetables and fruit also produces large amounts of aqueous effluent water containing various amounts of useful wastes depending upon the operation (Holdsworth, 1971).

S T A R C H

The starch industries of Europe produce significant amounts of waste much of which goes unutilized at the present time, largely because it arises in dilute effluent streams. The two main producers of potato starch in 1974 were the Netherlands and France producing 549000 and 140000 tonnes per annum respectively. In the Netherlands the waste is discharged into lagoons for treatment and in France is disposed of by irrigation. Starch is produced in the UK mainly from maize and from the 726000 tonnes per annum used in 1973 some 2100 tonnes were discharged as waste. Some 760000 tonnes of maize and 8000 of wheat per annum were processed to starch in France in 1973-74 and about 5500 tonnes per annum of waste were produced in the operations. Grain starch output in the FDR was 450000 tonnes and waste arising was 4200 tonnes per annum (Institut National de Recherche Chimique Appliquee, 1977).

S U G A R

In the manufacture of sugar from sugar beet substantial amounts of solid waste are produced, amounting to about 870 kg/tonne of beet processed. About 60% of the beet remains as spent pulp which is used for animal feed. The remainder of the solid waste is lime waste used mainly as a fertilizer and molasses which has several uses. Other losses occur in various effluent streams produced during sugar manufacture. For 1973 the EEC Commission has published figures for the amounts of sugar produced together with the amounts of waste in effluent streams associated with this production. These are shown in Table 2,7.

Table 2.7 B E E T S U G A R P R O D U C T I O N A N D A S S O C I A T E D W A S T E E F F L U E N T

I N T H E E E C ( I N S T I T U T N A T I O N A L D E R E C H E R C H E C H I M I Q U E A P P L I Q U E E ,

1 9 7 7 )


Category Food Total Kitchen Service Customer input waste waste waste waste

(MJ/person/meal)

Schools 3.4 0.22 0.003 0.06 0.167 Place of work 4.3 0.44 0.036 0.27 0.138 Restaurants, hotels, public houses 8.0 1.24 0.559 0.13 0.562 Cafes, snack bars 2.7 0.13 0.037 0.01 0.082 Hospitals 2.8 0.82 0.099 0.30 0.424 Welfare services 2.2 0.37 0.003 0.23 0.136

wastage rates in the various catering sectors studied. The highest percentage waste was in hospital catering where some 30% of food was wasted. In absolute terms the greatest amount of wastage occurred in commercial restaurants, hotels and public houses where 1.24 Μ J (300 kcal) was wasted per person per meal. These units also had the greatest food inputs averaging some 8.0 MJ (1910 kcal) per person per meal which is equivalent to 2.5 average meals. Cafes and snack bars produced the least amounts of waste. These also had the lowest food inputs, restricted menus and a high usage of convenience food. School meals produced lower than average amounts of waste and schools which provided a cafeteria-type meal wasted less food than did schools providing a 'traditional' school meal. About half of the food wasted in all the sectors was customer waste and the remaining food waste was divided approximately equally between kitchen waste and service waste. Factors associated with high food waste in catering establishments (Colhson and Banks, 1982) are high food input, large menu size, high fat content, poor control and low usage of convenience foods.

B R E W I N G

Of the ingredient by-products resuhing from the process of brewing beer, spent grains are by far the greatest in quantity and value. Other byproducts are barley screenings, malt culms and kiln dust, which are produced in more limited quantities. Fresh hops used in the brewing process also yield waste which is sold or used as an organic fertilizer. Considerable quantities of spent yeast are produced and most of this now goes for the manufacture of yeast extract. The UK brewing industry uses about 649000 tonnes of malt barley and adjuncts yearly and this would yield 743400 tonnes of spent wet grains or approximately 150000 tonnes of dried grains (Garscadden, 1973). The amounts of malted barley used in 1973 for production of malt extract and vinegar, in distilleries, and for export were 65000, 492000 and 49000 tonnes respectively.

Food waste in catering

In the UK about one meal in five is now eaten outside the home and studies have recently been carried out to determine whether the catering sector is a significant source of food waste. A survey of food waste in 39 catering estabhshments in the UK showed that on average 11% of the potentially edible food was discarded (Banks, 1981). Table 2.8 summarizes

Table 2.8 E D I B L E W A S T E F O R D I F F E R E N T C A T E G O R I E S O F C A T E R I N G I N

T H E U K ( B A N K S , 1981)

Α. ΤοΙαη 25

Summer Winter Summer Winter (% of total net weight) (% of energy content)

Meat 12.1 13.9 16.7 20.7 Fat 7.5 6.1 21.7 19.7 Potato 5.6 9.6 2.3 4.5 Cereals 33.2 24.7 39.4 30.6 Milk 12.6 6.1 3.5 2.1 Other 29.1 39.8 16.4 22.4

The survey also showed that larger families wasted more food in total but less in proportion to expected energy intakes and highest wastage in terms of energy occurred in the smallest household units. With the continuing reduction in the size of the household unit in the UK such findings may have implications for the food industry in terms of the sizes in

Domestic food waste

The largest source of food wastage in statistical terms (Gap II) appears to be that which occurs between the point of purchase and consumption. An attempt can be made to quantify this in actual terms in two ways, firstly by using national waste production and composition figures and secondly by surveying actual food waste in the home. Using the first approach some 18.0 million tonnes of domestic refuse are produced annually in the UK. Of this about 20% by weight is comprised of vegetable and putrescible matter (Barber, 1982) most of which will be food waste. In Scandinavia food waste comprises 35-40% of domestic refuse on a wet weight basis (Heie and Minsaas, 1981). A rough estimate of domestic food wastes produced annually in the UK would be 3.6 million tonnes. This figure will be on the low side because of waste food fed to pets, use of waste disposal units, compost heaps, etc. About 60% of domestic food waste is inedible waste such as peel, leaves, bones, eggshells, inedible trimmings, etc.

In order to obtain more detailed information on food wastage in the home a survey was organized by MAFF to investigate inedible waste in a random sample of 1000 households in Britain (Wenlock et al, 1980). Overall the total quantity of food discarded was higher in summer than in winter averaging 738 g or 9.3 MJ per household per week in summer and 591 g or 7.1 MJ per household per week in winter. Considerable quantities of otherwise edible food were also given to pets and birds, accounting, on average, for a further 2.4 MJ and 3.0 MJ per household per week in summer and winter respectively. The energy content of all food wasted in the home therefore averaged 11.7 MJ per household per week in summer and 10.1 MJ per household per week in winter; this is equivalent to 0.6 MJ and 0.5 MJ per person per day respectively. As Table 2.9 shov/s, bread and cereal products were wasted in the greatest amounts, and in energy terms averaged about 35% of the total. In terms of wet weight, wasted milk became more important especially in summer. Meat and fat were also important sources of wasted energy.

Table 2.9 C O M P O S I T I O N O F H O U S E H O L D F O O D

W A S T A G E I N B R I T A I N ( W E N L O C K ETAL., 1980)


Conclusions

In statistical terms the largest source of food waste appears to occur at the household level. Such waste is not readily amenable to upgrading or recycling and furthermore, in practical terms much of it cannot even be accounted for. Other sources of waste which occur before the stage of primary processing or during the process of food manufacture may be more readily recovered, recycled or upgraded. Often the major constraint upon the reuse of food waste is however non-technical and may be attributed to the structure of the industry, unit size dispersion and profitability. For example many animal tissues now rejected for human food could be used but for public prejudice, lack of capital investment, small abbattoirs, old-fashioned operations, and the inability to harness research and technical advance commercially. Firms in the food industries will evaluate the costs and benefits of various methods of recovery and disposal of waste products, and for the most part the chosen method of dealing with waste will be the most economically efficient from the standpoint of the individual firm. The enormity of the subject and the lack of relevant, particularly economic, information suggests a selective approach to further research in this area rather than a broad frontal attack on the problem.

Acknowledgements

I am indebted to my colleagues Mr D.D. Singer, Miss G.A. Smart and Dr E. Brewster, who were responsible for the Ministry's Food Waste Survey Unit and were instrumental in obtaining much of the information in this paper.

References

A D E R , G. and PLASKETT, L.G. (1975). Food Process. Ind., 14 B A I N E S , A.H.J , and HOLLINGSWORTH, D.F. (1961). Proc. Symp. Family

Living Studies, p. 120. International Labour Office (ILO), Geneva B A N K S , G. (1981). An Investigation into Food Utilisation in the UK Catering

Industry with Particular Reference to Food Waste. M. Phil. Thesis, Huddersfield Polytechnic, UK

which it packages food. A significant inverse relationship was found between the number of children and amount of food wasted. A single child was associated with the highest additional energy wastage and each successive child tended to add less waste to the household than did the previous one. Wastage also tended to increase with increasing income. In overall terms the amounts of wastage found in the survey accounted for less than a quarter of the 'calorie' gap between total food supplies and food thought to be eaten. This type of study is considered to underestimate food wastage which leaves in doubt again the estimate of actual total wastage at the domestic level.

Α. ΤοΙαη 27

B A R B E R , C. (1982). Notes on Water Research, No. 31, p.4. Water Research Centre, UK

CHIPPING C A M P D E N F O O D P R E S E R V A T I O N R E S E A R C H ASSOCIATION (1982) Personal communication

COLLISON, R. and B A N K S , G. (1982). Caterer & Hotel-keeper, May 13, p.79 C R A W F O R D , A . G . and W H I T M A N , W.E. (1980). The Utilisation and Disposal

of Sugar and Fat in British Food Manufacture. Special Project p. 942 and 948. British Food Manufacturing Industries Research Association, Leatherhead, UK

DHSS (1979). Report on Health and Social Subjects No. 15. Recommended Daily Amounts of Food Energy and Nutrients for Groups of People in the United Kingdom. HMSO, London

E U R O S T A T S (1981). Crop Production, Various issues F A O (1952). Second World Food Survey. Rome. p. 12 G A R S C A D D E N , B .A . (1973). Brewer, 59, 612 H E I E , A. and M I N S A A S , J. (1981). Recycling of Domestic Food Waste. In

Household Waste Management in Europe—Economics and Techniques. Eds. A. V. Bridgwater and K. Lidgren. Van Nostrand Reinhold Co. Ltd

H O L D S W O R T H , S.D. (1971). Proc. Int. Congress Industrial Wastewater. Stockholm, Sweden

INSTITUT N A T I O N A L D E R E C H E R C H E CHIMIQUE A P P L I Q U E E (1977). Pollution by the Food Processing Industries in the EEC A Report Prepared for the Directorate-General for Industrial and Technological Affairs and for the Environment and Consumer Protection Service of the Commission of the European Communities, Graham & Trotman

MAFF (1977a). Household Food Consumption and Expenditure: 1976. Annual Report of the National Food Survey Committee. HMSO, London

M A F F (1977b). Estimates of Food Supplies Moving into Consumption in the UK. Food Facts No. 7

M A F F (1981). Sea Fisheries Statistical Tables 1980. HMSO, London M A F F (1982). Personal communication R I C H A R D S , S.P. (1978). A Survey of the Utilisation of By-Products in British

Abattoirs. Special Project No. Ρ 755. British Food Manufacturing Industries Research Association, Leatherhead, UK

R O Y , R. (1976). Wastage in the UK Food System. Earth Resources Research Ltd, London

WEIERS, w. and FISCHER, R. (1978). The Disposal and Utilisation of Abattoir Waste in the European Communities. A Report Prepared for the Environment and Consumer Protection Service of the Commission of the European Communities. Graham & Trotman

WENLOCK, R.W., BUSS, D . H . , D E R R Y , B.J. and D I X O N , E.J. (1980). Br. J. Nutr., 43, 53

RECOVERY AND UTILIZATION OF PROTEIN FROM SLAUGHTERHOUSE EFFLUENTS BY CHEMICAL PRECIPITATION

R.N. COOPER, J.M. RUSSELL Meat Industry Research Institute of New Zealand (MIRINZ), Hamilton, New Zealand and J.L. ADAM Ruakura Agricultural Research Station, Hamilton, New Zealand

Introduction

Effluents produced during the slaughter of animals and processing of meat and by-products are characterized by high concentrations of organic nitrogen and fat. Only a proportion of this material is readily removable by conventional means such as sedimentation or screening resulting in an effluent high in soluble and colloidal organic nitrogen and fat together with a substantial oxygen demand. The characteristics of typical New Zealand slaughterhouse effluents after primary treatment are shown in Table 3.1.

Table 3.1 shows clearly the high proportion of soluble nitrogen typically found in slaughterhouse effluents.

Table 3.1 C H A R A C T E R I S T I C S O F N E W Z E A L A N D S L A U G H T E R H O U S E

EFFLUENTS

Range (g/m^)

Biological Oxygen Demand (BOD)-, 700-1800 Chemical Oxygen Demand (COD) 1200-3000 CODf* 700-1800 Total Kjeldahl Nitrogen (TKN) 70-200 TKNf* 55-160 Ammonia-Nitrogen (NH3-N) 5 -50 Total fat 100-900

*Subscript f refers to filtered through Whatman GF/C filter paper

These effluents are commonly treated by anaerobic/aerobic lagoons and oxidation ponds or by a high rate aerobic process such as activated sludge or trickling filters. The net result of these treatments is that the organic nitrogen is incorporated into cell biomass or degraded to ammonia.

Partial purification of proteinaceous wastes can be effected by protein precipitation following pH adjustment and/or dosing with coagulants such as aluminium or iron salts. Recent developments (Hallmark et al., 1978; Anon., 1981) have shown that a potentially valuable protein-containing material can be recovered from such effluents; the financial return obtained from this material offsetting to some extent the cost of treatment.

31

32 Recovery and utilization of protein from slaughterhouse effluents

Properties of proteins

A C I D - B A S E R E A C T I O N S

Proteins in solution carry a net charge, the nature of which may be positive or negative, depending on the pH of the solution {Figure 3.1). At a particular pH value, the isoelectric point (pi), the net charge is zero. The net charge on the protein is positive at pH values below the isoelectric point and negative at pH values above.

G O G H m \ C O O H C O O " NH3 C O O " C O G " N H 2 COÖ

' G H " ^ A 1- U G H " J - L

T : — \ r Ί Γ 1—r~ T NH3 C G ' O H NH3 N H ; C G O ' NH3 N H 2 C O G " N H 2

Net charge +3 Net charge 0 Net charge -3

Figure 3.1 Schematic representation of the effect of pH on protein charge

Proteins also behave in solution as colloidal particles and are stabilized by their surface charge. At the isoelectric point, attractive forces (e.g. Van der Waals) may lead to agglomeration and flocculation. However some proteins such as serum albumin and haemoglobin do not flocculate at the isoelectric point and cannot therefore be removed from solution by simple adjustment of pH.

These phenomena are made use of in processes in which slaughterhouse effluents are acidified to the pH corresponding to minimum solubiUty of the proteins. The effect of acidification on the soluble organic carbon of slaughterhouse effluents is shown in Figure 3.2 and it can be seen that maximum precipitation of organic material occurs over the pH range 4-5.

The relationship between pH and soluble organic carbon for a slaughterhouse effluent will not depend on the properties of any particular proteins but rather on the solubility relationships of the mixture of proteins over the pH range.

The incomplete removal of organic carbon from solution is due to proteins which do not precipitate and to proteins which are not isoelectric at the pH of net minimum solubility of the organic components of the effluent.

An important disadvantage of isoelectric pH adjustment processes is that blood proteins, which can be a major contributor to the soluble organic nitrogen of a slaughterhouse effluent, are not precipitated.

A number of chemical treatment processes have been reported in the literature (Stephenson, 1978; Cooper and Denmead, 1979; Hopwood, 1980). Differences in treatment efficiencies and the composition of the precipitated soHds observed for the different processes can be explained in terms of the mechanism involved. These mechanisms are related to the properties of proteins in solution and their reactions with anions and cations.

33

450

400

Ε ^ 350 i o u 300 Ε o σ> o ^ 250 - Q _D O

2 0 0

150

100 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

PH

Figure 3.2 Effect of pH on soluble organic carbon for three slaughterhouse effluents

2 5 0 r

200

5 150

o lOOh

5 0 h

o 0 g / m ^ Al*^

Δ 2 0 g / m 3 Ar3

+ 6 0 g / m 3 Al*^

_L - L - L 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

pH

Figure 3.3 The effect of increasing aluminium ion doses on the soluble organic carbon of slaughterhouse effluent


P R O T E I N - C A T I O N REACTIONS

The charged surface of colloidal molecules attracts ions of the opposite charge, resulting in the concentration of the counter ion being higher near the surface of the colloidal molecules than in the bulk solution. This phenomenon is known as the charged double layer. Adding an electrolyte to the solution reduces the range over which repulsive forces act by compressing the double layer and may promote destabilization and agglomeration of colloids. With negatively charged protein colloids, cations are responsible for the compression of the double layer and solutions of ferric and aluminium salts are often used.

Russell and Cooper (1981) reported that treating a slaughterhouse effluent with 40 g/m^ of aluminium (added as Al2(S04)3.18H20) and reducing the pH to 5.0 produced results similar to those obtained by acidification to the average isoelectric point. Increasing the aluminium dose {Figure 3.3) increased the range of minimum solubility. The material produced by this process is unlikely to have any value due to the high aluminium content of the precipitated soHds.

Cations can also form complexes with proteins by chemical reaction with the protein groups resulting in flocculation. They may also form bridges between protein colloids and polyelectrolytes resulting in enhanced agglomeration and flocculation.

P R O T E I N - A N I O N REACTIONS

Anions interact with positively charged protein colloids, the interactions being of a similar type to those described for cations. Anionic species reported include polyphosphates (Spinelli and Koury, 1970; Cooper and Denmead, 1979), lignosulphonates (Hopwood, 1980) and glucose trisul-phate (Jorgensen, 1971). Only the interactions of sodium hexametaphosphate and lignosulphonates with proteins are considered here.

Precipitation of proteins

S O D I U M H E X A M E T A P H O S P H A T E

SpineUi and Koury (1970) and Cooper and Denmead (1979) report the use of sodium hexametaphosphate (NaHMP) at pH 3.5 for the treatment of fish processing and slaughterhouse effluents respectively.

The effect of varying doses of NaHMP on the solubihty of haemoglobin and fibrinogen are shown in Figures 3.4 and 3.5 respectively. Haemoglobin has a pi of approximately 6.8 but does not precipitate under acid conditions. The addition of sufficient NaHMP and the adjustment of the pH of the solution results in almost complete precipitation of the haemoglobin. Increasing NaHMP doses lower the pH at which complete precipitation occurs.

Fibrinogen is normally precipitated at its isoelectric point (pH 6.0). Increasing sodium hexametaphosphate doses result in a shift of the

35

5 0 0 t -

^ AGG Ε

350

-e 3G0 Ö υ

2 5 0 | -c ? 200

In 15Gh

1G0

5 0 -

0 1

pH

o 0 g / m 3 NaHMP

Δ 70 g / m 3 NaHMP

+ 1 5 0 g / m 3 NaHMP

X 300 g /m3 NaHMP

Figure 3.4 The effect of increasing sodium hexametaphosphate (NaHMP) doses on the solubility of haemoglobin

500

A5G

Í AOGh

^ 350

Ξ 2 5 0 h

° 2G0l-

- Q

Ξ 150

1GGh

5 0 h

o 0 g / m 3 NaHMP Δ 15 g /m3 NaHMP • 100 g/m3 NaHMP X 300 g /m3 NaHMP

2 3 5 6 7 8 pH

Figure 3.5 The effect of increasing sodium hexametaphosphate (NaHMP) doses on the solubility of fibrinogen


isoelectric point to lower pH values. The NaHMP therefore interacts with proteins that are normally precipitated and it*is likely that for a given pH the appropriate NaHMP dose is determined by the amount of protein present in solution. The observed shift in the isoelectric point with respect to NaHMP dose can be explained as follows. At the isoelectric point the protein colloid has net zero charge. Addition of NaHMP results in the HMP anions neutralizing the positive charges resulting in a net negative charge on the protein-HMP complex. Further addition of hydrogen ions is now required to restore the protein to an isoelectric condition.

Examination of Figure 3.4 illustrates that at a given pH it is possible to overdose with NaHMP resulting in non-isoelectric conditions.

Optimum dose rates of NaHMP and percentage precipitation of several proteins are shown in Table 3.2. The exact amount of NaHMP required to precipitate proteins at a stated pH will depend on the number and availability of positively charged groups in the protein colloid. Denmead and Cooper (1975) report a dose of 42 g/m^ of protein. If the protein is assumed to be 50% carbon, then this represents 0.21 g NaHMP per g of organic carbon which is in good agreement with the values given in Table 3.2.

Table 3.2 R E M O V A L OF PROTEIN B Y NaHMP A T pH 3.5

Protein Optimum dose rate (g NaHMP/g total carbon removed)

Removal (%)

Serum albumin 0.17 97 Fibrinogen 0.15 89 Haemoglobin 0.26 99 γ-globulin 0.19 98 Dilute blood 0.23 98

3 0 0 -

2 8 0 -

'E 260 -

CT c 2 A 0 -o

Si

S 2 2 0

S 2 0 0

180

-δ 160 CO

UO

120

10

O 0 g / m 3 NaHMP

Δ 3 0 g / m 3 NaHMP

3.5 ^.0 5.0 pH

5.5 6 .0 6.5 7.0

Figure 3.6 The effect of sodium hexametaphosphate (NaHMP) on the soluble organic carbon of a slaughterhouse effluent

R, Ν. Cooper, J. Μ. Russell and J. L. Adam 37

L I G N O S U L P H O N A T E S

The precipitation of protein from waste waters using purified Hgnosulphon-ates (a by-product of the wood pulping industry) has been reported by Jorgensen (1968) and Hopwood and Rosen (1972). Lignosulphonates exist in solution at pH values above 1 as large anions and effect precipitation by

1000

Ε 800

c 700

g 600

o 120 g / m 3 CaLSA Δ 300 g /m3 CaLSA + 6 0 0 g / m 3 CaLSA X 9 0 0 g /m3 CaLSA

Figure 3.7 The effect of increasing calcium Ugnosulphonate (CaLSA) doses on the solubility of haemoglobin

charge neutralization. Precipitation by lignosulphonates shows many features in common with NaHMP. For example they precipitate haemoglobin and the pH at which minimum protein solubility occurs and the degree of precipitation is dependent on the dose of lignosulphonate. This is illustrated in Figure 3.7.

PRECIPITATION B Y T W O - S T A G E pH A D J U S T M E N T

In New Zealand two-stage pH adjustment procedures have been developed which can produce an effluent superior to that produced by acidification alone but without the cost of specific precipitants such as

Figure 3.6 compares the effect on a slaughterhouse effluent of NaHMP treatment at pH 3.5 with acid addition alone. A significant reduction in soluble organic carbon is produced by the NaHMP treatment. This is attributed to the almost complete removal of blood protein and those proteins which are not precipitated at the average isoelectric point of the effluent.


350

300

'E 250

c SI

S 200 u Ξ Ö tyi 150 o

-Q 100

50

0. I

J L

3 ii 5 6 7 8 9 pH

Figure 3.8 The influence of the addition of 0 .1% blood on soluble organic carbon in the double pH adjustment process

In the case with added blood. Figure 3.8 shows that additional soluble organic carbon is precipitated on raising the pH from 2.8 and a second isoelectric minimum is observed between pH 6 and 7. Thus a minimum amount of blood has to be present in slaughterhouse effluents if two-stage pH adjustment is to produce an effluent superior to that obtained by simple pH adjustment to the isoelectric point.

These observations can be explained in terms of the known properties of haemoglobin, a major blood protein, which does not normally precipitate at its isoelectric point (pH 6.8). If haemoglobin is acidified to pH 3 and then neutralized to between pH 6.5 and 7.0 flocculation and precipitation of protein is observed. This is due to the splitting of the haemoglobin into haem and globin units at pH 3 (Fanelh, Antonini and Caputo, 1964), the globin being precipitated at pH 6.5-7.0 (FanelH and Antonini, 1958; Tybor, Dill and Landmann, 1973). This is illustrated in Figure 3.9.

In practice it has been found necessary to raise the pH to between 8 and 9 to obtain good floe formation with anionic poly electrolytes. The effect of

NaHMP or lignosulphonates. The essential feature of these processes is to lower the effluent pH to 3 with sulphuric acid and then to raise the pH with an alkali (calcium or sodium hydroxide).

Our studies have shown that two-stage pH processes are most effective for effluents which contain high concentrations of blood (Cooper, Russell and Adam, 1982). Figure 3.8 shows the soluble organic carbon-pH relationships for an effluent with and without 0.1% added blood. It can be seen in the case of the effluent without added blood, that on readjustment to pH values in excess of 6.5 the amount of residual soluble organic carbon is greater than that at the isoelectric point of the effluent.

o Effluent . 0.1 7 . blood Δ Effluent

39

7 0 0 r -

600h-

E 5 0 0 h -

AOO

c

I 300

ω 2 0 0 k -

l o o h -

2.5 3 .0 3 .5 pH

4 .0 4 . 5 5.0

Figure 3.9 The effect on soluble organic carbon of acidifying a 0.12% haemoglobin solution to the indicated pH value and neutralizing to pH7 with N a O H

o ΝαΟΗ

Δ CaíOH )2

8.5 9 .0 9.5 10.0

Figure 3.10 The effect of Ca(OH )2 and N a O H as neutralizing agents in the double pH adjustment procedure


o 0 g /m^ superphosphate Δ 200g /rT i3 superphosphate

Figure 3.11 The effect of supeφhosphate addition on the soluble organic carbon of a slaughterhouse effluent in the double pH adjustment procedure

Figure 3.11 compares the effect of acidification of an effluent to pH 3 with and without superphosphate additions. Superphosphate results in a slight reduction in soluble organic carbon which may be due to occlusion of stable colloidal material during the formation and precipitation of hydroxy-apatite (Caio(P04)6(OH)2).

Process comparisons

A summary of chemical treatment processes reported in the literature and those developed at MIRINZ (Meat Industry Research Institute of New

*Fospur International Ltd, Derby, England

different neutralizing agents on residual soluble organic carbon is shown in Figure 3.10, Calcium and sodium hydroxides both produce minimum soluble residual organic carbon at pH 6-7. On raising the pH with sodium hydroxide appreciable resolubilization occurs at values higher than 7. When calcium hydroxide is used as the neutralizing agent the amount of resolubilization occurring between pH 7 and 9 is much reduced.

An alternative two-stage pH adjustment process, involving the addition of 200 g/m^ of ground agricultural superphosphate to the effluent followed by acidification to pH 3 and subsequent raising of the pH to 9 with calcium hydroxide, has been reported (Stephenson, 1978). The precipitated soHds are removed by sedimentation after flocculation with an anionic polyelec-trolyte (Decapol A33*). The final pH of 9 was dictated by the polyelectro-lyte used.

41

Table 3.3 SUMMARY OF CHEMICAL TREATMENT

Process name Operating Chemicals Solids Solids dewatering conditions separation

(1) MIRINZ isoelectric pH 4.0-4.5 dilute H2SO4 dissolved air lime condition, heat coagulate, 95-105 °C precipitation Magnafloc LT28^ flotation (daf)

(2) Aminodan*' pH 4.0-4.5 dilute H2SO4 daf lime condition, heat coagulate, 95-105 °C + polyelectrolyte

(3) MIRINZ HMP pH3.5 dilute H2SO4 + HMP daf lime condition, heat coagulate, 95-105 °C (4) Alwatech*^ pH3.0 dilute H2SO4 daf lime condition, heat coagulate, 95-105 °C

+ lignosulphonate (5) Superfloc"* pH 3.0-9.0 dilute H2SO4 sedimentation heat coagulate only, 95-105°C

supeφhosphate, Hme, + Decapol A33

(6) MIRINZ Alum pH 5.0-5.5 dilute H2SO4 daf lime condition, heat coagulate, 95-105 °C 40mg/€ Al Magnafloc E24^

(7) MIRINZ two stage pH 3.0-8.5 dilute H2SO4 lime daf heat coagulate only pH adjustment Magnafloc 156^

(8) MIRINZ two stage pH 3.0-7.0 dilute H2SO4 Hme heat coagulate only pH adjustment 4 mg/€ Al"^ Mk II Magnafloc 156^

^Allied Colloids, Bradford, England. ^Aminodan A/S, Skagen, Denmark. "Alwatech (UK) Ltd, High Wycombe, England. '^Development Finance Corp., Wellington, New Zealand.


Settled effluent

COD*^ 2 2 4 0 9 5 0 8 9 0 7 5 0 5 8 0

CODf^ 1 4 4 0 8 8 0 7 5 0 6 7 0 5 0 0

TKN^ 1 6 5 1 0 0 9 5 7 0 6 0

TKNf^ 1 3 0 9 0 8 0 6 5 5 0

N H v N ^ 1 0 1 0 1 0 1 0 1 0

Total fat 2 5 0 5 5 2 0 3 5 3 0

Soluble organic carbon 4 5 0 2 4 5 2 1 5 1 4 5 1 5 5

""Calcium lignosulphonate was used in this study ^See Table 3.1 for key to abbreviations.

It can be seen that isoelectric pH adjustment shows the poorest purification with the best results being obtained from those processes using specific protein precipitants. From a knowledge of the mechanisms occurring in the processes not listed in the above comparison it is possible to predict their performance.

The MIRINZ alum process would produce results similar to that obtained by acidification to the isoelectric point. The MIRINZ two-stage processes would produce results similar to those obtained by the Superfloc process. The performance of the two-stage pH processes relative to isoelectric precipitation will depend on the amount of blood present in the effluent.

Recovered solids

The principal objective of the treatment processes outlined is to purify protein-containing effluents and at the same time recover a solid whose composition makes it of value as a source of protein. There are two principal factors which influence the composition of such solids, namely the characteristics of the influent to the process and, secondly, the process itself.

I N F L U E N C E OF W A S T E W A T E R CHARACTERISTICS

Chemical treatment processes not only precipitate protein and destabilize colloidal material but also break fat emulsions. Thus the composition of an

Zealand) is shown in Table 3.3. Process numbers 2, 4, 5 and 7 have been operated at full scale and 1, 3 and 8 are pilot scale.

Comparisons of the effectiveness of these processes are difficult since tests were carried out on widely different effluents. Simulations of four of the above processes were therefore carried out at bench scale. The results are shown in Table 3.4. Sedimentation of the precipitated flocculated solids was used as the separation technique. Table 3.4 T R E A T M E N T OF S L A U G H T E R H O U S E E F F L U E N T BY C H E M I C A L M E T H O D S

Influent Isoelectric Superfloc Alwatech^ MIRINZ ρ Η 4.5 process HM Ρ

R. Ν. Cooper, J. Μ. Russell and J. L. Adam 43

23

22

21

20

18 J I I L 20 22 24 26 28 30 32

Fat in the influent (V. volatile so l ids )

Figure 3.12 The influence of influent fat on the recovered solids composition

A similar relationship will exist relating the proportion of precipitable organic nitrogen in the influent to the percentage protein in the dry recovered solids.

A potential disadvantage of the double pH adjustment processes is the high ash content of the recovered solids. Figure 3.13 illustrates how the ash content of the recovered solids is related to the chemical oxygen demand (COD) of the untreated effluent. The results were obtained at one particular slaughterhouse and it is likely that each slaughterhouse will exhibit its own unique relationship. For this particular slaughterhouse a 25% drop in COD from 2000 to 1500 g/m^ results in a 54% increase in the ash content of the recovered solids.

I N F L U E N C E OF T H E PROCESS

The processes used for the purification of protein containing effluents will influence the composition of the recovered solids. Those processes which remove more organic nitrogen per unit volume of influent will produce

influent to such a process will dictate the composition of the recovered solids.

For example, an influent containing high levels of total fat will result in recovered solids which contain higher levels of fat than would be the case for a low fat influent. This is illustrated in Figure 3.12

2 4 r -


3 0 r

2 5 h

20h

•σ "o 15h

o ιομ

O MIRINZ t w o - s t a g e pH Δ MIRINZ t w o - s t a g e Mkfl + MIRINZ HMP X MIRINZ i soe lect r ic

500 1000 1500 Influent COD

2000 2500

Figure 3.13 The relationship between influent C O D and % ash in the recovered solids

two-Stage pH processes produce solids which can be dewatered satisfactorily after heating to 95-105 °C. The solids produced by the other processes require conditioning with calcium hydroxide prior to heating if satisfactory dewatering is to be obtained. Such conditioning will result in an increase in ash of the recovered solids. For example, Cooper and Denmead (1979) report a calcium hydroxide addition rate of 3 kg/m^ of wet recovered solids to produce satisfactory dewatering. If all the calcium hydroxide was retained in the solids the ash content would increase to approximately 13%. This is unlikely, as approximately one-third of the calcium hydroxide will be lost in the liquid phase separated during mechanical dewatering resulting in a final solids ash of approximately 11%.

Utilization of recovered solids

Chemical treatment processes are expensive to build and operate. If the recovered solids have no value the costs of disposal of the solids produced (10 tonnes/day dry weight for a large New Zealand slaughterhouse) could be prohibitive and alternative treatment technologies more attractive.

solids containing a higher percentage of protein. This is illustrated in Figure 3.13 in which the composition of the solids produced by four different processes are shown.

The solids recovered by MIRINZ isoelectric and hexametaphosphate processes cannot be compared directly with those from the MIRINZ two-stage process with and without aluminium for the following reason. Table 3.3 reports details of the solids handling procedures which have to be adopted to ensure adequate mechanical de watering prior to drying. All the


E V A L U A T I O N O F R E C O V E R E D S O L I D S F R O M T H E S U P E R F L O C P R O C E S S A S A

F E E D I N G M E A L S U P P L E M E N T

The prime objective in building the plant was to recover sufficient dried solids to enable extensive feeding trial evaluation of these solids (known as Apex meal) to be undertaken.

The trials were designed to determine the effects of different dietary inclusion levels of Apex meal with and without added lysine on the digestibility and growth performance of pigs between 20 and 90 kg live weight. Details of the experimental design are presented in Appendix 3.1 (p.47).

Apex meal was found to contain more ash than typical New Zealand meat and bone meals but only about 55% of the crude protein. In general their amino acid profiles were quite similar, but levels of methionine and isoleucine were somewhat lower. The apparent digestible energy (ADE) of the Apex meal was 8.9 MJ/kg air dry sohds. The composition and the essential amino acid profile of Apex meal is shown in Table 3.5.

Table 3.5 C O M P O S I T I O N A N D E S S E N T I A L

A M I N O A C I D P R O F I L E O F A P E X M E A L

( % A I R D R Y W E I G H T )

Dry matter 98.5 Crude protein ( N x 6.25) 28.5 Diethylether extract 24.7 Crude fibre 17.4 Ash 24.4 Calcium 3.55 Phosphorus 1.35 Lysine 2.05 Histidine 0.92 Threonine 1.30 Valine 1.76 Methionine 0.58 Isoleucine 0.75 Leucine 2.69 Tyrosine 0.87 Phenylalanine 1.62 Tryptophan 0.13

The four diets for the growth trials were formulated on an isocaloric and isonitrogenous basis using graded levels of Apex meal to contain 10 g/kg of lysine (Table 3.6).

The results from the growth performance trials in which the diets were offered once daily to the pigs as a dry meal are shown in Table 3.7. Preliminary trials had shown that when the proportion of Apex meal exceeded 20% of the diet of pigs, feed refusals occurred. No problems were encountered in the growth performance trials in which Apex meal did not exceed 15% by weight of the diet.

In order to establish the nutritive quality of the material produced by one of these processes a prototype Superfloc plant (100 m^/h) was built in 1980 at the Wairoa works of Waitaki NZ Refrigerating Ltd.

46

Table 3.6 D I E T C O M P O S I T I O N A N D S E L E C T E D N U T R I E N T C O N T E N T S O F

T R I A L D I E T S (g/kg A I R D R Y W E I G H T )

Diet (Apex meal % by weight)

Ingredient 0 5 10 15 Maize meal 62.64 60.16 57.27 53.99 Barley meal 20.00 20.00 20.00 20.00 Apex meal 0.00 5.00 10.00 15.00 Blood meal 3.14 2.24 1.98 2.33 Fish offal meal 6.00 6.00 6.00 6.00 Meat bone meal 6.59 5.16 3.46 1.50 Salt 0.40 0.40 0.40 0.40 Lime 1.03 0.84 0.69 0.58 Vitamin and trace mineral supplement^ 0.20 0.20 0.20 0.20 A D E (MJ/kg) 14.2 13.6 13.5 13.6 Crude protein Lysine^

175.3 170.3 171.4 174.4 Crude protein Lysine^ 10.4 9.6 9.8 10.4 Methionine 4.0 3.0 3.3 3.6 Threonine 6.5 6.0 6.7 6.7 Isoleucine 5.6 5.0 5.1 5.1 Tryptophan'' 1.7 1.6 1.6 1.6

^Containing per kg supplement: vitamin A , 10 million l U ; vitamin D , 1 million l U ; vitamin E , 25000 l U ; vitamin B 2 , 8.8 g; vitamin B Ö , 3.4 g; vitamin B i , 3 g; vitamin B12 , 30 mg; niacin, 28 g; calcium pantothenate, 22 g; biotin, 100 mg; Mn, 30 g; Co, 1.6 g; Zn, 100 g; Fe, 80 g; I, 2 g; S e , 2 0 0 mg. ^Diets not containing added L-lysine monohydrochloride. Calculated .

Table 3.7 G R O W T H , C A R C A S E CHARACTERISTICS A N D A P P A R E N T DIGESTIBILITIES WITH PIG F E D DIETS CONTAINING D I F F E R E N T P R O P O R T I O N S OF A P E X M E A L

Diet SE Significance (Apex meal % by weight) of linear''

5 10 15 diff

Period 20-50 kg Growth rate (g/day) 650 646 642 611 8.2 xxx Feed conversion ratio 2.33 2.36 2.38 2.49 0.03 XXX

Period 50-90 kg Growth rate (g/day) 913 892 872 829 14.4 XXX

Feed conversion ratio 2.93 2.97 3.09 3.24 0.06 XXX

Period 20-90 kg Growth rate (g/day) 111 766 756 719 10.1 XXX

Feed conversion ratio l.dl 2.71 2.78 2.92 0.05 XXX

Killing out percentage 80.2 79.4 80.0 79.5 0.44 ns Hot carcase (kg) 72.3 71.9 72.2 71.7 0.40 ns Fat depth P2 (mm) 19.9 19.7 19.8 18.5 0.74 ns Lean body gain (g/day) 381 370 372 358 6.71 X

Fat tissue gain (g/day) 183 174 172 158 3.65 XXX

Fat (%) 30.6 30.1 29.9 28.8 0.64 ns Apparent digestibility coefficients of Nitrogen 0.81 0.82 0.81 0.79 0.011 ns Gross energy 0.85 0.82 0.80 0.79 0.009 XXX

Dry matter 0.85 0.83 0.81 0.80 0.009 XXX

^ns, not significant X , significant at the 5% level X X , significant at the 1% level X X X , significant at the 0 .1% level

R. Ν. Cooper, J, Μ. Russell and J. L. Adam 47

Cole (1980) 100 50 — 60 50 18 Diet formulation 100 — 35 65 55 17

0% Apex meal 100 — 38 62 54 16 15% Apex meal 100 — 35 64 49 15 0% Apex meal + lysine 100 — 34 53 48 15

15% Apex meal + lysine 100 — 30 57 44 13

Calcu lated

Diets for the present trial were formulated using the concept of an ideal protein (Cole, 1979), but with reference to the amino acids given in Table 3.8. Analysis of the diets showed that relative to lysine actual proportions of amino acids were lower than intended and for diets with added lysine were lower than the recommendations of Cole (1980) {Table 3.8). In the present trial the addition of L-lysine did not affect liveweight gains or tissue growth and neither was any interaction found between levels of Apex meal and lysine. Thus comparing amino acid ratios for extremes of the Apex meal diets with recommendations of Cole (1980) it can be postulated that levels of isoleucine and tryptophan were implicated in the depressed growth with increasing levels of dietary Apex meal.

In summary. Apex meal is a useful addition to the range of feed materials available for pigs. It contains less protein and apparent digestible energy than meat and bone meals and more ash. Apex meal can be used at up to 10% by weight in the diet of growing pigs.

Appendix 3.1 F E E D I N G TRIALS E X P E R I M E N T A L DESIGN

Growth trial

The trial was designed as a 4 x 2 x 2 factorial with the factors being the level of dietary Apex meal, added lysine level and sex.

Table 3.7 shows that increasing the proportion of Apex meal in the diet resulted in decreases in daily gain and corresponding increases in the feed conversion ratio. The depressions in live weight growth occurred at all stages of growth. In contrast, Pearson and Smith (1982), in evaluating a meal produced by an isoelectric precipitation process, only found significant linear growth depressions at live weights in excess of 45 kg. The depressions in liveweight growth in the present trials were also reflected in corresponding linear depressions in both lean body and fat growth rates. These linear decreases for tissue growth were probably not due to differences in ADE intake. Increasing levels of Apex meal in these trials, and of the meal used in the trials of Pearson and Smith (1982), resulted in a hnear decrease in the apparent digestibility of gross energy which was not reflected in dietary ADE levels {Table 3.7). This may be a reflection of protein quality.

Table 3.8 D I E T A R Y A M I N O A C I D B A L A N C E R E L A T I V E TO LYSINE (=100)

Lysine Methionine + Methionine Threonine Isoleucine Tryptophan^ cystine


Digestibility trial

Two groups of eight entires weighing 22 kg were allocated at random within groups, but in equal numbers to each of the four diets containing graded levels of Apex meal without added L-lysine {Table 3.6).

Pigs were housed in individual metabolism crates. Following an acclimatization on period and a further four-day period at a constant feed intake for all pigs, a full faeces collection was done over five days.

Data were subjected to analysis of variance with treatment sums of squares being positioned into components based on single degrees of freedom to assess linear effects.

References

A N O N (1981). Meat Processing, June, 40 A D A M , J.L. , D U G A N Z I C H , D .M. and H A R G R E A V E S , K. (1982). N.Z. Soc.

Anim. Prod., (in press) C O L E , D .J .A . (1979). In Recent Advances in Animal Nutrition, pp.55-72.

Ed. by W. Haresign and D. Lewis. Butterworths, London C O L E , D .J .A . (1980). Pigs News & Information, 3, 201 C O O P E R , R.N. and D E N M E A D , C P . (1979). / . Wat. Poll. Contr. Fed., 51, 1017

Twenty-four entires and the same number of gilts, weighing about 19 kg, were allocated at random within sexes to the eight dietary treatments. Pigs were then allocated at random to individual feeding for the duration of the trial covering the interval 20-90 kg live weight. A further three pigs of both sexes, with the same background were killed at 20 kg live weight.

Four isocaloric and isonitrogenous diets were formulated with graded levels of Apex meal to contain lOg/kg of lysine. The minimum levels of methionine, threonine, isoleucine and tryptophan were set at 35, 65, 55 and 17% respectively in these latter diets. To half of each original batch of diet was added 0.17% by weight of L-lysine monohydrochloride and this reduced the proportions of other dietary amino acids relative to lysine.

Diets were offered to pigs once daily as a dry meal and water was available ad libitum from drinking nipples. Daily feed offered was based on live weight (120 g/W^^^) and was adjusted weekly after weighing.

On reaching their final Hve weight of either 20 or 90 kg pigs were stunned and slaughtered. The P2 fat depth was marked (Kempster, Jones and Cuthbertson, 1979), the carcase backed down to expose both sides of the dorsal spinous processes and P2 measured by fat depth indicator (Adam, Duganzich and Hargreaves, 1982). The P2 fat depth is the fat thickness over the longissimus dorsi muscle lateral to the dorsal mid-line at the position of the head of the last rib.

After chilling for 24 h at 3°C heads and feet were removed and the carcases sawn into two sides through the middle of the vertebral column. The right-hand sides were cut into fat (rind, subcutaneous and peritoneal fat) and lean (the remainder).


C O O P E R , R . N . , RUSSELL, J.M. and A D A M , J.L. (1982). Proceedings 37th Industrial Waste Conference, Purdue University (in press)

D E N M E A D , C.F. and COOPER, R.N. (1975). Meat Industry Research Institute of New Zealand (inc.) Publ. No. 448

FANELLI , R . A . and A N T O N I N I , E. (1958). Biochim. Biophys. Acta, 30, 608 FANELLI , R . A . , A N T O N I N I , E. and C A P U T O , A. (1964). Adv. Prot. Chem. 19,

74 H A L L M A R K , D . E . , W A R D , J.L., ISAKSEN, H.C. and A D A M S , W. (1978). Pro

ceedings 9th National Symposium on Food Processing Wastes, Denver, Colorado, EPA-600/2-78-188, 288

H O P W O O D , A .P . (1980). Wat. Pollut. Control, 79, 225 H O P W O O D , A .P . and R O S E N , G . D . (1972). Process. Biochem., 7, 15 J 0 R G E N S E N , S.E. (1968). Vatten, 24, 332 J 0 R G E N S E N , S.E. (1971). Vatten, 27, 58 KEMPSTER, A.J . , JONES, D .W. and C U T H B E R T S O N , A. (1979). Meat Sci., 3, 109 P E A R S O N , G. and SMITH, W.C. (1982). N.Z. J. Exptl Agric, 10, 119 RUSSELL, J.M. and COOPER, R.N. (1981). Environ. Tech. Lett., 2, 537 SPINELLI, J. and K O U R Y , B. (1970). / . Agr. Food Chem., 18, 284 S T E P H E N S O N , P. (1978). Proceedings 9th National Symposium on Food

Processing Wastes, Denver, Colorado, EPA-600/2-78-188, 306 T Y B O R , P.T. , DILL, C.W. and L A N D M A N N , W.A. (1973). / . Fd Sci., 38, 4

ULTRAFILTRATION IN THE RECOVERY OF FOOD WASTE

J.K. WALTERS and K.L. ELLIOTT Department of Chemical Engineering, University of Nottingham, UK

Introduction

When two solutions of different concentrations are separated from each other by a semipermeable membrane, there is a tendency for the solvent to flow through the membrane from the dilute solution to the more concentrated solution. This phenomenon is known as osmosis and its first recorded observation is attributed to the Abbé Nollet in 1748. If allowed to reach equilibrium the passage of water will cease when osmosis is balanced by the pressure difference across the membrane. If pure solvent is on one side of the membrane, this pressure difference is known as the osmotic pressure of the solution on the other side of the membrane, as shown in Figure 4,1. The application of a pressure difference ΔΡ greater than the osmotic pressure ΔΠ will cause solvent to flow in the reverse direction.

Osmotic pressure

Solvent Solution

^

Membrane Figure 4.1 Schematic diagram of osmosis

51

52 Ultrafiltration in the recovery of food waste

^ , 1 · % % · — _ _ Concentrate

Feed ^ · φ # ·

Permeate

Figure 4.2 Schematic diagram of ultrafiltration: % large molecules · small molecules

overlap between ultrafiltration and reverse osmosis but the distinction can be made between the two processes at a pore size of about 10"^ to 10"^ μm. The upper limit for ultrafiltration is around 10"̂ to 1 μm and above that size the membrane process is known as micro filtration. The separation of even larger sized materials is done by conventional filtration. Both of these latter processes remove suspended solid particles whereas ultrafiltration and reverse osmosis are used to separate dissolved solutes. The essential feature of ultrafiltration and reverse osmosis is the application of pressure to separate the components of a solution. Conventional separations are based on the application of heat to effect a phase change, by for example, evaporation. Such processes are energy-intensive compared with membrane separations. Figure 4.3 shows the size characteristics of membrane processes and is based on Porter (1977) and Porter and Michaels (1971).

This is the basis of reverse osmosis as a process for recovering solvent from a solution. The pores in the membrane are sufficiently small to prevent the passage of even small inorganic ions and only the even smaller solvent molecules can pass through. Consequently the flux (J) of solvent through the membrane is small, and a large area of membrane is required for a reasonable throughput. The membrane is characterized by the pure water permeability constant A so that

J = Α ( Δ Ρ - Δ Π ) (1)

Membrane manufacturers can now produce membranes of a specified pore size with a high degree of accuracy. Membranes with larger pores will allow the passage of inorganic ions and molecules of low molecular weight through the membrane with the solvent while retaining molecules of high molecular weight. The process is then known as ultrafiltration. The osmotic pressure of solutions of molecules of high molecular weight is quite small, because it depends on the molar concentration rather than the mass concentration, and so the osmotic pressure difference across the membrane in ultrafiltration is usually ignored in comparison with the applied pressure drop. The solvent flux is given by

J = AP/(R^ + Rs) (2)

where R^ is the resistance of the membrane and R̂ is the resistance characterized by the upstream conditions of flow. The solvent flux is larger in ultrafiltration than reverse osmosis and the applied pressure difference is smaller. Ultrafiltration is shown schematically in Figure 4.2. There is some

/. κ. Walters and Κ. L. Elliott 53

Oí C C 13

sie CO > < DO Q

Microfiltration

Ultrafiltration

Conventional f i l tration

Reverse o s m o s i s

ΙΟ Ι 0"̂ 10" ,-2 10 10- 1

Size (pm)

Figure 4.3 Size characteristics of membrane processes

10^ 10^

Michaels (1981) has followed this with a further review and Cooper (1980) and Turbak (1981) have edited very useful symposium books.

Types of membrane

The earliest membranes developed by Loeb and Sourirajan in the late 1950s were of cellulose acetate and this is still widely used today. However, many synthetic polymers have since been developed, such as aromatic Polyamides, poly acrylic acid/poly vinyl chloride copolymers and sulphon-ated polysulphones. These can withstand higher temperatures and have better mechanical strength and resistance to chemical attack. Many of these membranes are made by precipitation of the polymer from a polymer/solvent casting solution by a non-solvent which is usually water. Cellulose acetate membranes are made from a solution of cellulose acetate in acetone. The acetone is allowed to evaporate for a few minutes. The cellulose acetate film is then precipitated in cold water and annealed in hot water. Incorporation of a water soluble filler which is leached out in the annealing stage provides the porous structure. The membrane is not uniform throughout its thickness. Under the electron microscope it is seen to be asymmetric consisting of a thin surface layer, about 0.25 μm thick which contains very small pores, bound to the bulk of the film which may be 100 μm thick and which contains much larger pores as shown in Figure 4.4. It is the surface layer that is the essential feature of the membrane for reverse osmosis and ultrafiltration applications. Permeate passes through the surface layer first and then through the bulky porous mass of the membrane. If flow were in the other direction the surface layer would simply be knocked off and very little separation would be achieved. It is important therefore that the membranes be installed the correct way round. Permeate fluxes are generally in the range 5-50 €/(m^ h) for reverse osmosis and 40 to 400 €/(m^h) for ultrafiltration.


Spongy porous ma te r i a l -

v ioo 'PM P ° ^ ^ ^ ' " 0 , 1 -0 .4PM

Smooth Figure 4.4 Cross-section of membrane

Another process for producing pores in membranes is the 'track-etch' technique. The membrane is first made in a completely non-porous state and then subjected to a collimated beam of radiation. The radiation passes through the membrane perpendicular to the surface and causes ionization of the polymeric material along its path. The damaged polymer regions are subsequently dissolved away to leave an array of pores, which are relatively uniform in diameter and orientation. They are, however, quite large—about 0.5 μm in diameter—and are more suited to microfiltration than ultrafiltration or reverse osmosis. If the technique can be developed to produce pores of finer diameters, it would provide a great advance in the separations that could be achieved by uhrafiltration and reverse osmosis.

The pores in the active surface layer have a distribution of pore sizes so that a few large molecules may still find their way through the larger pores and appear in the permeate. However the manufacturers are now able to

S i z e (jjm )

Figure 4.5 Typical cut-off curves for proteins (MW 50000 and MW 300000): Cp permeate concentration, Cp feed concentration

Rough -1 • - — •

0.25jüm surface layer-pore size <0.01jum _L- — — - —

y. κ. Walters and Κ. L. Elliott 55

Membrane configurations

Because the permeate flux through the membrane is small a large area of membrane is required to obtain a reasonable throughput. Pressure differences across the membrane are significant so sufficient mechanical support must be provided. Adequate passage for the flow of feed, concentrate and permeate is necessary and it must be possible to dismantle and reassemble the module easily for cleaning. This is particularly important in applications in the food industry where hygiene is paramount. Many modules with no 'dead space' within them may be cleaned in situ and this is preferable, where daily cleaning is necessary as in many food processing applications, for reasons of reliability and economy. There are four main types of construction, all of which are commercially available: flat sheet, tubular, spiral wound and hollow fibre, and each finds favour with different manufacturers.

FLAT S H E E T

Membrane sheets are stacked together with appropriate spacers and feed supplied to alternate compartments. The permeate is removed from the

Feed

1 1 i J Permeate

Concentrate Figure 4.6 Schematic diagram of flat sheet membrane configuration

produce membranes with fairly narrow pore size distributions so that this is minimized. Rather than use pore size as such, manufacturers often characterize membranes by the molecular weight of a globular protein that will be 90% retained by the membrane (Porter, 1979). Control of pore size during manufacture is such that molecular weight differences of 10000-20000 can be specified, and typical cut-off curves are shown in Figure 4.5 for nominal 50000 and 300000 molecular weight proteins. A historical outline of membrane structure and production has been given by Porter (1977). Solubility diagrams for sulphonated polysulphones are given by Friedrich et al. (1981).


T U B U L A R

The membrane is mounted on the inside of a strong porous tube which is contained in a tubular sleeve. The permeate passes through the tube and is collected in the sleeve. Usually many tubes are arranged in a common sleeve so that the module is similar to a conventional shell-and-tube heat

Feed

J i L

I Concentrate

Permeate

Figure 4.7 Schematic diagram of tubular membrane configuration

exchanger as in Figure 4.7. Such an arrangement is very easily cleaned and individual tubes can be easily replaced. The disadvantage is the large module volume per unit area of membrane.

SPIRAL W O U N D

This is essentially a flat sheet unit with a long narrow membrane which is then rolled up, with suitable spacers to leave passages for liquid flow, so that it is rather like a 'swiss roll' as in Figure 4.8. It is very compact with a low hold-up per unit area of membrane, but is very susceptible to plugging and very difficult to clean. However recent developments described by Bailey and Skelton (1981) suggest that these difficulties are being overcome.

H O L L O W FIBRE

Small diameter, thick walled tubes of membrane material are manufactured by a number of companies. They are known as hollow fibres and have an internal diameter to wall thickness ratio of from 3 to 5. DuPont have a reverse osmosis fibre of internal diameter 0.04 mm while Romicon

intermediate compartments as shown in Figure 4.6. The advantage is a low module volume per unit area of membrane but care has to be taken to prevent the presence of 'dead space' or regions of low liquid velocity, which would make cleaning difficult.

Feed

J, Κ, Walters and K. L. Elliott 57

Membrane

Spacer

Figure 4.8 Schematic diagram of spiral wound membrane configuration

have ultrafiltration fibres of two sizes: 0.5 mm and 1.1 mm internal diameter. Photomicrographs of the cross-section of the hollow fibre show that it is anisotropic with a very tight thin skin on the inside about 0.1 μm thick with a spongy outer structure with much larger pores. More detailed descriptions are given by Blatt (1972), Breslau et al, (1975), Porter (1975 and 1979) and Breslau et al. (1980). It is clear that the basic structure is similar to the asymmetric cellulose acetate membranes described earlier. Because the hollow fibres are so small, many millions are put together in a bundle in one sleeve and the permeate seeps through the fibre walls into the sleeve. The concentrate passes through the inside of the fibres and

Table 4.1 M E M B R A N E C O N F I G U R A T I O N S FOR F O O D PROCESSING

Manufacturer Configuration Application^

Abcor^ Tubular R O , U F Amicon'' Flat plate UF, MF

Hollow fibre U F DDS'^ Flat plate R O , U F Dorr Oliver^ Flat leaf UF DuPont^ Hollow fibre RO Gelman^ Flat plate MF Millipore^ Flat plate UF, MF Nuclepore' Flat plate MF Paterson Candy InternationaP Tubular R O , U F Romicon*^ Hollow fibre RO Sartorius' Flat plate MF Schleicher Schuir Flat plate UF

^RO-reverse osmosis, UF-ultrafiltration, MF-microfiltration. ^Abcor Environmental Systems Ltd, Surbiton, Surrey, UK. ''Amicon Corporation, Lexington, Massachusetts, U S A . ^De Danske Sukkerfabrikke, 1001 Kobenhavn, Denmark. ^Dorr-Oliver Inc., Stamford, Connecticut, U S A . ^Du Pont Inc., Wilmington, Delaware, U S A . ^Gelman Instrument Co . , Ann Arbor, Michigan, U S A . ^Millipore Corporatin, Bedford, Massachusetts, USA. 'Nuclepore Corporation, Pleasanton, CaHfornia, USA. Paterson Candy International Ltd, Laverstoke Mill, Whitchurch, Hampshire, UK. •'Romicon Inc., Woburn, Massachusetts, U S A . 'Sartorius Membranfilter GmbH, Gottingen, W. Germany •"Schleicher & Schüll GmbH, Einbeck, W. Germany

Permeate

Concentrate


Membrane polarization

Concentration polarization in membrane processes is inevitably associated with the rejection of solutes by the membrane. Bulk flow of the concentrated solution is towards the membrane; solvent and some solutes pass through the membrane; higher molecular weight species are rejected. Consequently the concentrafion. Cm, of these rejected species builds up adjacent to the membrane, until diffusion away from the membrane balances the bulk flow towards the membrane. The concentration profile of solute through the membrane is thus as shown in Figure 4.9.

Concentration polarization

Gel polarization

Concentration

Bulk flow

P3Sition

k-Membrane

Flux(J)

Figure 4.9 Polarization on the upstream side of the membrane: C ß bulk concentration, C M concentration adjacent to membrane

The flux increases with increase of pressure and so does C^. This causes an increase in solute concentration in the permeate. Solutes rejected in ultrafiltration can attain values of Cm adjacent to the membrane of the order of ten times the bulk concentration in the feed. In fact the solute may precipitate in the form of a sHme or gel and provide added resistance to flow. Increase in pressure then only thickens this gel layer and results in a

comes out at the opposite end to that to which the feed is suppHed. The flux rates are small in hollow fibre units but a large area of membrane is packed into a small volume so the overall permeate throughput is of a reasonable value. Hollow fibres modules are difficult to clean and a very efficient filter is usually required as a pretreatment unit.

Further comparisons of the four major types of membrane configuration can be found in Bailey (1977), Crits (1976/77), Harper (1980), Fell and Payne (1980) and Crossley (1980). Table 4.1 summarizes configuration by manufacturer.


Concentrate

Ultrafiltration

nnodule

P e r m e a t e

Figure 4.10 Recirculation around an ultrafiltration unit

reduced flux through the membrane. This is known as gel polarization. Most practical ultrafiltration systems operate near maximum flux conditions and so are at incipient gel polarization. Polarization effects can be reduced by having a large cross-flow of solution along the upstream surface of the membrane. To achieve this a recirculation of solution around the ultrafiltration unit as shown in Figure 4.10 is required. Further theoretical approaches to polarization may be found in Erikson et al. (1978), Fane et al. (1981), Papenfuss et al. (1978) and Probstein et al. (1978).

Membrane fouling

It is difficult to distinguish between polarization effects and fouling. Both reduce the flux by increasing the resistance across the membrane. However, polarization is to a large extent reversible or avoidable by suitable choice of operating conditions, whereas fouUng is not. Eykamp (1978) attributes fouling in whey ultrafiltration to fat, casein and metabolites from micro-organisms such as lactobacilli. Calcium salts may also be deposited. Fouling can only be removed by thorough cleansing with peptizing agents or solubilizing agents for the inorganic salts. FouUng is pardcularly important in food processing and ultrafiltration units must be cleaned regularly on a daily basis both to maintain the flux and for hygienic reasons. Non-cellulose acetate membranes are generally more robust and easier to clean. Membrane replacement is a large item in the operadng costs of ultrafiltration units and membrane damage occurs largely during the cleaning operations, rather than during actual ultrafiltrafion.

The dairy industry

In the UK alone, milk at 15.8 miUion tonnes a year has the third largest annual tonnage after water (supply and sewerage) and crude steel (Dun-nill, 1981). It is not surprising therefore to find the majority of applications


W H E Y PROCESSING

Whey is the fluid portion of milk obtained after the coagulation of casein in the manufacture of cheese. The compositions of wheys vary but they generally contain about 6.5% wt TS made up of 5% lactose, 1% protein and 0.5% ash. Prior to the use of ultrafiltration whey was

(1) used as a pig food, (2) evaporated and spray dried to give whole whey dried solids, (3) evaporated and the lactose recovered by crystallization, or (4) demineralized, evaporated and dried at low temperature to give a

lactose/protein product for use in baby foods.

Reverse osmosis may be used to replace the initial evaporation stages in (2) and a combination of reverse osmosis and ultrafiltration may be used to separate the protein and lactose as in (3). The protein concentrate is obtained first and the permeate from the ultrafiltration unit is treated in the reverse osmosis unit to obtain a lactose concentrate. The permeate from that unit is an aqueous waste comparatively low in biochemical oxygen demand (BOD) and may be sent to the sewer or possibly recycled. A block diagram is shown in Figure 4.11. The economics of the process depend on the satisfactory use of the lactose containing permeate from the ultrafiltration stage. Coton (1979) has discussed this in detail and Figure 4.12 summarizes the possible uses to which the permeate may be put.

Three modes of operation of the ultrafiltration unit are currently employed. In the batch mode whey is pumped from a feed storage tank through the unit and back to the tank. Cross-flow velocities sufficient to prevent gel polarization are used and the storage tank is depleted by the

of membrane processes in the food industry in dairy processing. Reverse osmosis is used mainly for concentrating wheys. Operation at room temperature is usual and an upper Hmit of 24% total soHds (TS) is imposed to prevent lactose crystallization. Full milk may also be concentrated by reverse osmosis to 30% TS which may replace evaporated milk or be used in ice-cream, yoghurt or cheese manufacture. It is reported to give a better texture and taste than the use of powdered milk. It may also be used as a first stage in the preparation of powdered milk although concentration from 30 to 50% TS where spray drying can begin, must still be done by evaporation. Ultrafiltration is being successfully used to prepare milk concentrates for making into soft cheese, yoghurt and ymer. Whey may be ultrafiltered to produce a protein concentrate and a lactose permeate. The proteins are not denatured to any great extent because ultrafiltration takes place at room temperature. The advantages of membrane processes are the higher yield and higher nutritional value of the product, a smaller plant, the production of a sweet permeate and the ease of continuous processing. Glover et al (1978) have a particularly good review while Bansal (1977) and Horton (1974) put dairy processing into perspective in the spectrum of membrane usage and Pepper (1977 and 1978) and Klinkowski (1978) give costs of whey treatment plants.

/. κ, Walters and Κ. L. Elliott 61

Whey

6.5% TS

5%U3dose r/o protein as % Ash

Concentrate

^15% Protein >20P/o TS

Permeate

1 Lactose 1

6% TS Reverse

o s m o s i s

Concentrate

>10% Lactose >20P/o TS

Permeate

(water)

for reuse

Figure 4,11 Block diagram for ultrafiltration and reverse osmosis in whey processing

WHEY PERMEATE

.^Pig feed

^ Concentration

drying (singly or with other materials) crystallization to cattle lick reaction with urea ANIMAL USE

» Fermentation

reaction to ammonium lactate biomass

anaerobic - methane for fuel INDUSTRIAL

alcohol

other organics (e.g. lactic acid)

.Lactose extraction

. Hydrolysis

galactose/glucose syrup sugar confectionery brewing syrup

DIRECT HUMAN USE

Figure 4.12 Uses of whey permeate (based on Coton, 1979)

passage of permeate through the membrane. The continuously fed batch mode operates in a similar fashion except that fresh raw whey is continuously fed to the feed tank. Both these modes are similar to that shown in Figure 4.10. The third mode of operation is cascade operation where concentrate from the first stage becomes feed for the second stage. It is common to have two stages but there is no reason why three or four or more stages should not be used. A three-stage system is shown in Figure


concentrate

Pernneate 1 Pernneate 2

Figure 4.13 Three-stage ultrafiltration system

P e r m e a t e 3

These properties have important uses in the sugar and flour confectionery industries. It is stable in acid conditions and can be used in soft drinks. These functional properties are discussed more fully by De Wit (1975). Whey protein also has a high nutritional value and may be used in yoghurt, tortillas and macaroni.

Donnelly (1971) was one of the first to consider membrane processes in the treatment of dairy wastes and since then many papers have appeared. Coton (1974) discussed the fractionation of whey by ultrafiltration, De-laney, Donnelly and Bender (1974) looked at product quahty, Horton (1974) reported on three large industrial plants using tubular membrane units and Bakel et al. (1975) discussed hollow fibre ultrafiltration of cottage cheese whey. Richert (1975) considered various milk manufacturing processes and put ultrafiltration into perspective and Bodzek and Kominek (1979) show block diagrams with protein and lactose concentrations for combined ultrafiltration, reverse osmosis and electrodialysis. Muller and Harper (1979) have discussed the effect of whey pretreatment on the operating characteristics of membrane units and conclude that ultraflhra-tion performance is improved by prior clarification and pH control and reverse osmosis by demineralization and pH control.

C H E E S E M A N U F A C T U R E

A liquid product with the same composition as a cheese can be obtained by ultrafiltration of milk under suitable condidons (Maubois and Mocquot, 1975). This is known as 'pre-cheese'. The whey proteins are retained along with the casein giving an improved yield and reducing the quantity of whey for disposal. The 'pre-cheese' is inoculated with a starting material, rennetted, incubated and fermented. Variadon of pH and the use of whole milk or skim milk allow different types of cheese to be made. Soft cheese manufacture is well developed and in commercial production. Figure 4.14 shows the differences between conventional cheesemaking and the method using ultrafiltration. Production of mozzarella has been reported by Covacevich and Kosikowski (1978), blue cheese by Jolly and Kosikowski

4.13. The ultrafiltration membrane allows lactose, lactic acid, salt and water components to pass through, but retains and concentrates the higher molecular weight proteins (mostly lactoglobulin and lactalbumin). Uhrafil-tered whey concentrate is a valuable material. It is soluble, has not been denatured and has good organoleptic properties. It will coagulate and form complexes with other proteins on heating and form gels and stable foams.


Mi lk p r e t r e a t m e n t

R e n n e t i n g

Curd and w h e y

Moulding

U l t ra f i l t ra t ion

W h e y w i thou t proteins

W h e y d ra inage

P r e - c h e e s e ( concen t ra te )

R e n n e t i n g moulding

Coagulat ion

Cheese

Figure 4.14 Cheesemaking—conventional and using ultrafiltration

(1975), Camembert by Ernstrom, Sutherland and Jameson (1980) and Maubois (1980). Feta, ricotta, ymer and camembert ultrafiltration systems are marketed by Alfa Laval* who use Romicon hollow fibre membrane units. Production of cheddar cheese is difficult (Ernstrom, Sutherland and Jameson, 1980; Covacevich and Kosikowski, 1978; Sood and Kosikowski 1979), although acceptable processed cheese can be made by blending with conventional cheddar.

W H O L E MILK

Prior to the use of ultrafiltration for the concentration of milk, three milk products were widely accepted,

(1) evaporated milk—some water removed, (2) condensed milk—some water removed and some sugar added, and (3) dried milk—all the water removed.

Reverse osmosis can be used to remove some of the water and hence produce a product similar to evaporated milk but with the protein still in undenatured form. Some evaporation is still required after reverse osmosis and before spray drying for dried milk. With appropriate selection of

*Alfa-Laval Co. Ltd, Great West Road, Brentford, Middlesex, UK.


Vegetal extracts

The protein that can be extracted from soy bean, cotton seed, peanuts and sesame seed may be used as a protein supplement for human foods. A large number of food companies now use vegetable protein in their products and a number of formulations using this as the sole source of protein are available. Cheryan (1980) quotes a growth rate of 10% per annum. Undesirable components are also present in the form of oligosaccharides, phytic acid and trypsin inhibitors and these must be removed. This can be done by ultrafiltration because such molecules are smaller than the desired proteins so it is possible to select a membrane and process conditions for this separation.

The recovery of starch and gluten from wheat flour is widely practised. The Martin process and the Batter process are in common use and in both large quantities of water are required to separate the starch and gluten. Losses of over 10% of the flour entering the plant are lost in the effluent stream, and the protein loss is an even greater percentage (Fane and Fell, 1977). Disposal is a problem because of the high BOD. Ultrafiltration is an effective treatment and will recover the protein and leave a permeate free

Flour

Lime

Extraction Extract Extraction

Residue

y' Ultrafiltration Concentrate Dryer

Permeate

Reverse Concentrate Dryer

Permeate

Figure 4.15 Block diagram of soybean whey processing

membrane, ultrafiltration may be used to produce special dietary milks that are free of salt, lactose and sugars, and these are possible at no extra cost. Yan, Hill and Amundson (1979) have discussed the production of yoghurt and Rangarajan et al. (1981) have reported protein recoveries of 92% on both whole and skim milk. Poulsen (1978) discussed the feasibility of using ultrafiltration for standardizing the protein content of milk but agreement on a standard or even its desirabihty has not been achieved.


S O Y B E A N W H E Y

Soybean flour is hydrolysed in alkaline suspension to produce a whey containing the soy protein which is separated from the solid residue in a centrifuge or by conventional filtration. The whey extract is then fed to an ultrafiltration unit for recovery of the protein concentrate, which may be spray dried, and a permeate containing sugars. These may be recovered in a reverse osmosis unit and the permeate from that may be recycled to the hydrolysis stage as in Figure 4.15 (Lawhon et al. 1977, 1979 and 1980). Jackson et al. (1974) show detailed flowsheets and suggest a three-stage ultrafiltration system and a ten-module series/parallel reverse osmosis unit. Pressures in the ultrafiltration system are typically 0.1-0.7 Mpascals (MPa) (1-7 bar) and in the reverse osmosis unit 4-10 MPa (40-100 bar).

W H E A T S T A R C H EFFLUENTS

The effluents from wheat starch are broadly similar to those from soy whey or cheese whey. All contain proteins, sugars and inorganics. For a factory processing 400 tonnes/week a three-stage ultrafiltration unit concentrating the effluent to 16% wt sohds would require 1000 m^ membrane area in the first stage and 400 m^ in each of the second and third stages. The flux would be 20 €/(m^ h) in the first stage and 10 €/(m^ h) in the second stage. Fane and Fell reported that the ultrafiltrate concentrate, which was cloudy and brown in colour, could be spray-dried to produce a stable yellow powder. This was used in a series of baking tests and shown to be an acceptable substitute for gluten at up to 7.5% subsdtution. As such its market value would be some five times greater than the value based on its protein content. The permeate contains sugars and low molecular weight solutes and may be suitable for recycling if the wheat flour processing can tolerate higher concentrations of sugars, or it could be disposed of to the sewers. An alternative is the use of reverse osmosis for further processing to produce a sugar concentrate and a water permeate.

P O T A T O PROCESSING E F F L U E N T

The conventional way of treating potato processing effluents consists of three stages and has been described by Holladay (1976). The wastes contain high levels of starch and protein, have a BOD of 1500 g/m^ and ferment easily. The treatment comprises screening, sedimentation, activated sludge aerafion and final settling to give a final effluent BOD of 20 g/m^ or less. All useful materials are lost, and there is clearly the opportunity to use membrane processes for protein and water recovery. A plant to handle 4300 m^/day (180 m^/h) of potato effluent water has been

of micro-organisms which may be recycled. A similar problem occurs with potato processing effluents and a large reverse osmosis plant handling 4300 m^/day has been buih in Holland (Crossley, 1981).


built for Avebe in Holland by Paterson Candy International*. It is described by Crossley and by Pepper and Orchard (1981). The flowsheet is shown in Figure 4.16. A reverse osmosis system is used to recover half of the flow (90 vo?lh) as permeate with a relatively low chemical oxygen demand (COD) of 400 g/m^ which is recycled and an equal flow of concentrate containing 8% TS. The protein in the concentrate is coagulated by steam injection and then separated from the liquid and dried to

Potatoes Grinding and

separation

Starch

IBO t/h J

Grinding and

separation fac tory

90m7h Waste potato VMoter 180frfh

4 % TS

R e v e r s e / ^ Cor̂ oentrate

Dsmosis

protein separation

Evaporator Dryer

60% TS potato'protein

Figure 4.16 Diagram of potato processing effluent (based on Pepper and Orchard, 1981)

give the potato protein. The liquid from the separator is concentrated by evaporation to 60% TS. The reverse osmosis line consists of six units operating in parallel with five on Une and one being cleaned at any one time. Each unit contains 1030 m^ membrane area in three reverse osmosis stages in series, similar to that in Figure 4.12, giving a mean flux of 17 €/(m^ h). There are six main reciprocating feed pumps driven by a 55 kW motor and 18 high pressure recirculation pumps each driven by a 45 kW motor.

Other applications

Membrane processes may be used for the separation of solutes from each other or from the solvent by the application of pressure alone, without the need to supply latent heat for phase change. The following sections highlight some of the wide range of separations that merit further investigation.

* Paterson Candy International Ltd, Laverstock Mill, Whitchurch, Hampshire, UK.

J. κ. Walters and Κ. L. Elliott 67

P A U N C H C O N T E N T M A T E R I A L

Paunch content material consists of partially digested grass with a large amount of water (90-95%) and has a high BOD (cattle, 50000 g/m^ sheep 30000 g/m^). Conventionally it is conveyed by water to a screen and then to the drain. Fernando (1980) reports that the liquid portion contains 60% to 80% of the protein which in turn is 2.5% of the original contents, and so is suitable for concentration by ultrafiltration. The fibrous material must be

F E R M E N T A T I O N P R O D U C T S

The use of ultrafiltration in the fermentation industries is still largely at the development stage, but Drioli (1980) has reported its use in controlling the polyphenol content of must in white wine production, without change of sugar content. Beaton (1980) has described the concentration of enzymes and their purification using diafiltration in which water is added to the circulating concentrate at a rate equal to the permeate flow. Such operation increases the purity of the concentrate at the expense of permeate dilution. He also describes fermentation broth clarification by ultrafiltration.

E N Z Y M E A N D PROTEIN PURIFICATION

Ribonuclease, peroxidase and serum albumin have been used by Hatch and Price (1978) to characterize the retentivity of ultrafiltration membranes. They showed that a two-stage system could be used to separate a mixture of peroxidase (MW 40000) and serum albumin (MW 68000) with minimal denaturation of the enzymes. Staged cascade systems have also been discussed theoretically by Tutunjian and Reti (1978). Balfanz and Hirsch (1981) have concentrated a-amylase, while Adamski-Medda, Nguyen and Dellacherie (1981) have developed a technique for chelating a very high molecular weight molecule with a selected protein so that uhrafiltration may be used in a very specific way. They were able to separate trypsin (MW 24000) from a-chymotrypsin (MW 22500) using p-aminobenzamidine linked with dextran (MW 2 x 10^) as a chelating agent for trypsin. Upon ultrafiltration the trypsin was retained and the a-chymotrypsin passed through to the permeate.

VITAMINS A N D VIRUSES

Increase of concentration of vitamin Β12 from 20 g/m^ to 10"̂ g/m*̂ by ultrafiltration has been reported by Rangarajan, Desai and Rao (1981) in a laboratory scale flat plate unit. Virus concentration using hollow fibre membranes has been discussed by Belfort, Rotem-Borensztajn and Katzenelson (1978) and Rotem-Borensztajn, Katzenelson and Belfort (1979) and shown to be a useful and quick technique, although the hydrodynamics of the system are important.


A N I M A L B L O O D

Blood contains about 10% of the protein in an animal and is a valuable by-product of the meat industry. Raw blood contains about 20% TS and ultrafiltration is attractive as a means of concentrating it, without damage, to about 30% TS. Fernando (1981) reports experience with hollow fibre ultrafiltration units and indicates favourable economics compared with vacuum evaporation.

F R U I T JUICES

In concentrating fruit juices it is necessary to retain the sugars which have quite low molecular weights. Reverse osmosis is therefore more applicable than ultrafiltration and this is discussed by Merson, Paredes and Hosaka (1980). A fundamental thermodynamic approach which may be used for predicting the solute transport parameter for many low molecular weight compounds was presented by Matsuura and Sourirajan (1978) and they have applied it to the concentration of fruit juices.

Economics

Cost data is expressed in a number of different ways by different authors and detailed comparisons are difficult to make. Pepper (1978) gives costs of a 10 m % whey plant operated at one-third, one-half and two-thirds water removal but he omits labour costs. Crossley gives costs of an 18 m^/h cheese whey plant including labour, but also including steam costs for reheating the chilled and pasteurized whey. Their figures are given in Tables 4,2 and 4,3, For a plant removing half the water, Pepper's figures may be adjusted upwards to account for labour to give a cost of £0.5/m^ permeate and Crossley's adjusted downwards to give about £l/m^ permeate. For a larger potato processing effluent unit Pepper and Orchard (1981) give the figures shown in Table 4,4. Again labour costs are omitted and if included would increase the cost to about £0.8/m^ permeate. American costs are given by Beaton (1980) for the ultrafiltration of fermentation broths and are compared with his costs for rotary vacuum filtration in Table 4.5. To obtain a comparison we should ignore the depreciation term when his costs become $2.44/m^ permeate (about £1.3/m^ permeate). Crossley notes that the annual savings of the reverse osmosis unit over the original evaporator system exceeded the capital cost of the membrane unit. The future of ultrafiltration and reverse osmosis systems in the food

removed first by filtration or in a centrifuge and may be utilized as a fuel. The ultrafiltration concentrate can be utilized as pig food in Uquid form. It contains 6.2% TS of which about one-third is protein. The permeate BOD is lowered to 1500 glw? and represents a very significant reduction over the full paunch content material. Fernando's costings indicate a halving of the operating costs by using ultrafiltration.

J. κ. Walters and Κ. L. Elliott 69

Water removal: 33% 50% 66% Operating costs £/m^ £/m^ £/m^

membrane replacement 0.13 0.15 0.21 electrical power 0.06 0.04 0.05 maintenance, spares 0.02 0.02 0.03 chemicals 0.10 0.12 0.16

Total 0.31 0.33 0.45 Capital cost (£) 40000 65000 100000

Table 4.3 A N N U A L COSTS OF A N 18m'/h (FEED) C H E E S E W H E Y P L A N T (CROSSLEY, 1981) B A S E D ON 9 m % P E R M E A T E

Original evaporator plant Reverse osmosis plant

Steam £173000 £24000 Cleaning £ 2500 £ 9500 Labour £ 9300 £ 9300 Electricity £ 5200 £ 7000 Membranes — £15000 Total £ 1 9 0 ()()() £65000 Cost/m^ filtrate £4.20 £1.43

Table 4.4 O P E R A T I N G COSTS OF A 180m^/h ( F E E D ) P O T A T O E F F L U E N T PLANT (PEPPER A N D O R C H A R D , 1981)

Membrane replacement 0.18 Electrical power 0.26 Cleaning, chemicals 0.16 Total (£/m^ permeate) 0.60

Table 4.5 C O S T C O M P A R I S O N B E T W E E N U L T R A F I L T R A T I O N ( U F ) A N D

R O T A R Y V A C U U M F I L T R A T I O N ( R V F ) F O R F E R M E N T A T I O N B R O T H S

( B E A T O N , 1980)

Operating costs ($/m^ filtrate) R V F U F

filter aid 5.30 — membrane replacement — 1.50 energy 0 . 0 8 0.18 labour 0.26 0.37 maintenance 0.06 0.11 cleaning chemicals — 0.18 depreciation 0 . 1 9 1.10

Total 5.89 3.44 Capital costs ($/m^/day) 470 2760

industry would seem to be assured with good return on investment and with operating costs of about £l/m^ permeate.

Acknov^ledgements

The groundwork for this chapter was presented by one of us (KLE) as a dissertation in part-fulfilment of the penultimate year of the Chemical Engineering course at the University of Nottingham.

Table 4.2 ECONOMICS O F A lOm^/h ( F E E D ) W H E Y P L A N T (PEPPER, 1978)


References

A D A M S K I - M E D D A , D . , N G U Y E N , O.T. and D E L L A C H E R I E , E. (1981). / . Memb. ScL, 9, (3) 337

B A I L E Y , P.A. (1977). Filtration and Separation, May/June, p. 213 B A I L E Y , P.A. and SKELTON, R. (1981). Filtration and Separation, May/June,

207 B A K E L , J.T. , MORRIS , H . A . , RICHERT, S.J. and MORR, C.V. (1975). / . Dairy

Sci., 58, (12), 1794 B A L F A N Z , F. and HIRSCH, S. (1981). Chemische Technik, 33, (11) 569 B A N S A L , I.K. (1977). A.I.Ch.E. Symp. Ser., 73, No. 166, 144 B E A T O N , N.C. (1980). In Cooper, A.R, (Ed.), below p. 373 B E L F O R T , G., R O T E M - B O R E N S Z T A J N , Y. and K A T Z E N E L S O N , Ε. (1978).

Frogr. Water Tech., 10, (1/2), 357 B L A T T , W.F. (1972). International Laboratory, Nov./Dec, p. 14 B O D Z E K , M. and KOMINEK, O. (1979). Environment Protection Engg., 5, (3),

285 B R E S L A U , B . R . N , A G R A N A T , E . A . , TESTA, A.J. , MESSUIGER, S. and CROSS,

R . A . (1975). Chem. Engg Prog., 71, (12), 74 B R E S L A U , B .R. , TESTA, A.J . , MILNES, B . A . , and MEDJANIS , G. (1980). In

Cooper, A.R. (Ed.), below, p. 109 C H E R Y A N , M. (1980). In Cooper, A.R. (Ed.), below, p. 343 C O O P E R , A .R . (Ed.) (1980). Ultrafiltration Membranes and Applications.

Plenum Press, Washington C O T O N , S.G. (1974). / . Soc. Dairy Tech., 27, (3), 121 C O T O N , S.G. (1979). Address to International Dairy Federation on T h e

Utilization of Permeates from the Ultrafiltration of Whey and Skim Milk'

C O V A C E V I C H , H.R. and KOSIKOWKSI, F.V. (1978). / . Dairy Sci., 61, 701 CRITS, C.J. (1976/77). Ind. Wat. Engng., 13, (No. 6), 20 CROSSLEY, I .A. (1980). Chartered Mechanical Engr, Jan. p. 33 CROSSLEY, I .A. (1981). I. Mech. E. Conf. C6/81 Future Developments in

Process Plant Technology, p. 41 D E L A N E Y , R . A . M . , D O N N E L L Y , J.K. and B E N D E R , C D . (1974). Dairy Indus

tries, Nov., p. 426 D E W I T , J.N. and D E B O E R , R. (1975). Netherlands Milk Dairy J., 29, 198 D O N N E L L Y , J.K. (1971). Food and Farm Research, Sept./Oct., 2(5) 113 DRIOLI , E. (1980). In Cooper, A.R. (Ed.), above, p.291 D U N N I L L , P. (1981). Chemistry & Industry, 4 Apr., p. 204 ERIKSSON, G., ERIKSSON, P., HALLSTRÖM, Β. and WIMMERSTEDT, R. (1978).

Desalination, 27, 81 E R N S T R O M , C A . , S U T H E R L A N D , B.J. and JAMESON, G.W. (1980). / . Dairy

Sci., 63, 228 E Y K A M P , w. (1978). A.I.Ch.E. Symp. Ser., 74, (No. 172) 233 F A N E , A . G . and FELL, C.J.D. (1977). A.I.Ch.E. Symp. Ser., 73, (No. 163),

198 F A N E , A . G . , FELL, C.J.D. and W A T E R S , A . G . (1981). / . Membr. ScL 9, (No. 3),

245 FELL, C.J.D. and P A Y N E , M . D . (1980). Chem. Engg in Australia, 6, (1), 26 F E R N A N D O , T. (1980). Process Biochem., 15, (3), 7


F E R N A N D O , T. (1981). Biotechnology and Bioengineering, 23, 19 F R I E D R I C H , C , D R I A N C O U R T , Α . , N O E L , C. and M O N N E R I E , L. (1981), De

salination, 36, (No. 1), 39 G L O V E R , F .A . , S K U D D E R , P.J., S T O T H A R T , P.H. and E V A N S , E.W. (1978). 7.

Dairy Res,, 45, 291 H A R P E R , W.J. (1980). In Cooper, A.R. (Ed.), above, p.321 H A T C H , R.T. and PRICE, J .D. (1978). A.I.Ch.E. Symp. Ser., 74, (No. 172),

226 H O L L A D A Y , D . G . (1976). Process Biochem., 11, 14 H O R T O N , B.S. (1974). Proc. IV Int. Cong. Food Sci. & Technol., IV, 332 H O R T O N , B.S. (1974). In Advances in Preconcentration and Dehydration of

Foods, Ed. by Spicer, A. Applied Science Publishers Ltd, London JACKSON, G., STAWIARSKI, M.M., WILHELM, E.T. , GOLDSMITH, R.L. and

E Y K A M P , W. (1974). A.I.Ch.E. Symp. Ser., 70, (No. 136), 514 JOLLY, R.C. and KOSIKOWKSI, F.V. (1975). / . Dairy Sci., 58, (9), 1272 KLINKOWSKI, P.R. (1978). Chem. Engg., May 8, p. 165 L A W H O N , J.T. , H E N S L E Y , D . W . , MIZUKOSHI, M. and MULSOW, D . (1979). J.

Food Sci., 44, (1), 213 L A W H O N , J.T. , M A N A K , L.J. and L U S A S , E.W. (1980). In Cooper, A.R. (Ed.),

above, p. 685 L A W H O N , J.T. , MULSOW, D . , CATER, C M . and MATTIL, K.F. (1977). / . Fd Sci.,

42, (2), 389 M A T S U U R A , T. and S O U R I R A J A N , S. (1978). A.I.Ch.E. Symp. Ser., 74, (No.

172), 196 M A U B O I S , J.L. (1980). In Cooper, A.R. (Ed.), above, p. 305 M A U B O I S , J.L. and M O C Q U O T , G. (1975). / . Dairy Sci., 58, 1001 M E R S O N , R.L. , P A R E D E S , G. and H O S A K A , D . B . (1980). In Cooper, A.R.

(Ed.), above, p.405 MICH A E L S, A .S . (1981). Chem. Tech. Jan. p. 36 M U L L E R , L.L. and H A R P E R , W.J. (1979). / . Agric. Food Chem., 27, (4), 662 P A P E N F U S S , H . D . , GROSS, J.F. and S A N C H E Z - R U I Z , F. (1978). A.I.Ch.E.

Symp. Ser., 74, (No. 172), 218 PEPPER, D . (1977). Chemistry and Industry, 20, 834 PEPPER, D . (1978). /. Chem. E. Symp. Ser., No. 54, p. 247 PEPPER, D . (1978). The Chemical Engineer, No. 339, 916 PEPPER, D . and O R C H A R D , A.C.J. (1981). Paterson Candy International Ltd.,

TPRO 41-1 PORTER, M . C (1979). In Handbook of Separation Techniques for Chemical

Engineers, Ed. by Schweitzer, P.A. Section 2-1, p. 2-3. McGraw Hill, New York

P O R T E R , M . C (1975). Chem. Engg Prog., 71, (12), 55 P O R T E R , M.C. (1977). A.I.Ch.E. Symp. Ser., 73, (No. 171), 83 PORTER, M . C and MICHAELS, A .S . (1971). Chem. Tech., Pt. 1, Jan. p. 56; Pt.

2 Apr. p. 248; Pt. 3 July p. 440; Pt. 4 Oct. p. 633; Pt. 5 Jan. 1972 p. 56 P O U L S E N , P.R. (1978). J. Dairy Sci., 61, 807 PROBSTEIN, R.F. , S H E N , J.s. and L E U N G , W.F. (1978). Desalination, 24, (No.

1/2/3), 1 R A N G A R A J A N , R., D E S A I , N.V. and R A O , A . V . (1981). Indian J. Tech., 19,

(1), 1


R A N G A R A J A N , R., D E S A I , N .V. , D E V M U R A R I , C.V. , R A O , A . V . and P A R E K H , J.V. (1981). Indian J. Tech., 19, (No. 2), 43

RICHERT, S.H. (1975). / . Dairy Sei., 58, (7), 985 R O T E M - B O R E N S Z T A J N , Y. , K A T Z E N E L S O N , Ε. and BELFORT, G. (1979). / .

Env. Engg Div., ASCE, 105, (EE2), 401 S O O D , V.K. and KOSIKOWSKI, F.V. (1979). / . Dairy Sei., 62, 1713 T U R B A K , A .F . (Ed.) (1981). Synthetic Membranes, Vol. II Hyper- and

Ultrafiltration Uses, ACS Symp. Ser. 154, Washington T U T U N J I A N , R.S. and RETZ, A .R . (1978). A.I.Ch.E. Symp. Ser., 74, (No.

178), 210 Y A N , S.H., HILL, C G . Jnr. and A M U N D S O N , C.H. (1979). / . Dairy Sci., 62,

23

5

Introduction

Fat extraction is not new and in particular rendering has been with us for a very long time. However, the whole business of fat extraction including rendering, mechanical and solvent extraction, to produce fats and protein-rich meals for use in livestock and poultry feeding is under continual review. This is because the capital cost of equipment, the running cost, the cost of energy, the availabihty of materials and the value of the final products all have to be balanced.

The world's supplies of fats and oils were forecast to be 66 million tonnes (t) for 1981-82 during the USDA's Outlook 1982 meeting in Washington in November, 1982 (Anon., 1982). The total includes 59.4 million t of production plus stocks of 6.6 million t. Of the 59.4 million t of production the animal fats contribution will be 14.8 miUion t made up of 4.9 million t of butter (fat content) and 9.9 million t of lard, tallow and grease. Recovery of fat from poultry to give a protein rich meal continues to expand and methods used to recover fat from these sources are discussed in this review. However, in spite of this vast amount of fats and oils now available there may well be a shortage in the future and, in addition, experts predict there will continue to be a shortage of meat, eggs, fish and dairy products. Although there are improved agricultural and farming methods the fact is that the production cannot be stepped up to cope with the anticipated protein demand, particularly in developing countries where protein consumption averages will remain a problem. Some answers to this problem may be found in industrial food production techniques. Fermentation which results in the production of single cell proteins (SCPs) is an emerging industry which permits the rapid production of micro-organisms rich in proteins. Fat extraction from these systems is important and thus current work in a 115 t per day plant in the USSR to extract lipids from biomass will also be described in this review.

Whether talking about fat extraction from oilseeds, animals or SCP there is a basic similarity in the processes used. It is in the preparation of materials prior to extraction that the greatest differences in the unit operations involved are seen. In this review the methods of extracting fat from meat and bone tissues (greaves), poultry wastes and SCPs are discussed.

FAT EXTRACTION

J. DAVIE Simon-Rosedowns Ltd, Hull, UK

73

74 Fat extraction

Stages of extraction

The following section deals with the extraction of inedible fat/grease from animal and poultry sources. The extraction of lipids from SCP is dealt with in a later section.

The raw materials from which animal fat is extracted include fallen stock, butcher waste, reject carcases, bones, slaughter house material, chicken and turkey waste, etc. The finished products are fat and meat and bone meal.

R a w mater ia l

i

Solvent extraction

plant Solvent extracted fat I

Mea l Cake Fa t

Figure 5.1 Flow diagram outlining the stages of extraction involved in the production of animal fat and meat and bone meal

In order to reduce protein degradation, minimize the increase in free fatty acids and limit the problem of smell as much as possible the raw material should be received in as fresh a state as possible and processed rapidly. The various stages involved in the extraction procedure are outlined in Figure 5.1, although for several of the stages shown alternatives may well exist and be used in certain circumstances. The detailed stages are outHned below.

P R E B R E A K I N G

For subsequent handling it is important to break the material into pieces less than about 2 cm^. The main factors to be considered in the design of prebreakers are safety and ruggedness. Irrespective of the design of the prebreaker the products sent to the next stage of the process vary widely in consistency and this variation must be catered for by the machines which carry out these subsequent operations.

/. Davie 75

R E N D E R I N G / C O O K I N G

The cooking process is carried out to release fat from the material and thus produce a high protein meal of good storage stability. In addition it does serve to retain the protein quality and to sterilize the meat by destroying micro-organisms whilst effectively removing moisture.

An efficient cooker is designed to give maximum heat transfer to all the material in the cooker and to give a high rate of evaporation to enable the moisture content to be reduced from 65 to 10% in as short a process time as possible. Atmospheric cooking is normally sufficient for sterilization but where the regulations in a particular country require the material to be subject to temperatures in excess of 100 °C, pressure cooking must be carried out. The most common cooker, certainly in the UK, is the horizontal, cylindrical type shown in Figure 5.2 where a heat transfer

fct. \

Figure 5.2 One of the SD range of new generation cookers. (Simon-Douglas Ltd, Hempstead Lane, Gloucester)

surface is provided by both the shell and the agitator located within this shell. This cooker has the flexibility, when processing poultry material, to yield a product of high protein digestibility even when feathers are present since the conditions can be set to satisfactorily hydrolyse the feathers whilst minimizing the amount of amino acid degradation and thus loss of protein quality.

Where rendering/cooking is a batch operation it is necessary to have more than one unit and/or storage bins if continuous running of the downstream plant is to be achieved.

76 Fat extraction

Fat separation

This step is achieved by percolation followed by mechanical extraction and, where necessary, solvent extraction.

P E R C O L A T I O N S T A G E

When handling animal greaves the free fat present after cooking can be run off into a tank fitted with a perforated plate or wedge wire screen bottom or onto a conveyor with a trough that has been similarly modified. After draining, the fat content of the animal greaves is usually in the region of 25-35%. With poultry the fat content at this stage can be as high as 45% which can give rise to problems in the subsequent mechanical extraction stage.

M E C H A N I C A L E X T R A C T I O N

The percentage of fat which can be tolerated in the extracted cake or meal will determine whether this stage yields the final product or whether it is a preUminary operation to remove some of the easily-won fat prior to a subsequent solvent extraction stage.

Solely mechanical means can remove a significant amount of the fat in the greaves but will invariably leave between 8% and 10% remaining in the resulting cake. The machine used is a high pressure continuous screw press {Figure 5,3) which receives a metered supply of hot greaves via a small heated cooker/conditioner whose purpose is to 'top up' the temperature of the drained greaves and make any small adjustments to moisture content as required for efficient pressing.

The screw press is basically a long wormscrew revolving inside a 'slotted' barrel; the metered feed of hot greaves is conveyed along and squeezed

The cooking cycle commences with the loading of a predetermined amount of broken material through a charging door in the top of the vessel. It continues with the agitator revolving whilst controlled steam on the shell jacket and agitator shaft enable the necessary cooking temperature for the specified period to be achieved. An aspiration system is used as required to exhaust steamy vapours and reduce the moisture content of the charge. On completion of the cooking period a discharge door is opened in the end of the vessel and most of the loose fat released by this operation flows out into the drainage pan. The agitator shaft is then run in the reverse direction bringing into play angled paddles on the back of the stirrer arms to discharge the cooked greaves and the remaining loose fat into the drainage pan.

Other methods of cooking used in the industry include continuous rendering (Burton, 1979), wet rendering (Blogg, 1976) and a process where the raw material is mixed with Uquid fat and then dried under vacuum to induce rendering (Granofsky, 1978).

/. Davie 77

Figure 5.3 Simon-Rosedowns Mark 3M Screw Press designed to process up to 30 t of meat and bone greaves every 24 h

progressively through a combination of diminishing wormflight pitch and increasing boss diameter. The fat released by this progressive volume reduction passes through the slots out of the barrel and is collected, the solid residue being discharged at the end of the wormshaft as 8-10% fat cake which is then cooled and ground.

If the market demands a very low-fat product containing 3% fat or less this can only be achieved by solvent extraction. In this case mechanical extraction is used as a preparation stage to take the fat content of the drained greaves down to 16-20% to ease the duty of the solvent plant. This

Figure 5.4 Simon-Rosedowns Low Pressure Greaves Press designed to remove the easily won tallow from up to 75 t of meat and bone greaves every 24 h

78 Fat extraction

S O L V E N T E X T R A C T I O N

If required greaves from the low pressure screw press or from the basket centrifuge can be passed to a solvent extraction plant to further reduce the fat content, to as low as 1% if desired. The type of plant normally used for this operation is shown in Figure 5.5.

Although several solvents have been used in the past to defat the feedstock, including trichloroethylene, hexane is now used almost exclu-

Figure 5.5 Continuous Rotary Solvent Extraction Plant designed by Simon-Rosedowns Ltd with a capacity to handle 150 t of meat and bone greaves every 24 h

preparation task traditionally has been performed by basket centrifuges, particularly in smaller plants where spun greaves from the centrifuges are sold for solvent extraction elsewhere. However, because of safety problems and the inherent batchwise operation of centrifuges a marked preference is now being shown towards the use of a light continuous screw press to perform this operation. A low pressure press which performs this task is shown in Figure 5.4.

The fat from the drain tanks, screw presses or basket centrifuges is usually clarified using a decanter centrifuge.

Assuming that the feedstock has been allowed to cool after rendering the high pressure screw press {Figure 5.3) requires about 37 kJ of energy and about 28 kg of steam to process 1 t of material to a fat content of 8-10%. However, the low pressure press used before solvent extraction {Figure 5.4) only uses about 3 kJ of energy and 25 kg of steam whilst partially defatting 1 t of material to a content of 16-20% fat. Obviously if the feedstock has not been allowed to cool after rendering then the steam consumption figures will be lower.

y. Davie 79

Greaves Solvent(hexane)

Extractor

Extracted meal

Desolventizer Solvent

Vent Solvent recovery

Solvent

Miscella (tallow/ solvent)

To

atmosphere

Vent

Distillation system

Solvent condensers

Solvent

Vent

Condensate separation

Meal I

Fat Water to dram

Figure 5.6 Flow diagram of a typical solvent extraction plant

sively to perform this task. In this process the fat is normally removed in a percolation type extractor by a commercial hexane boiling in a specified temperature range. A typical simplified flow diagram for this process is shown in Figure 5.6.

In the usual solvent extraction process the material is fed into the extractor via a screw conveyor equipped with counter-weighted discharge gates which minimize the back-flow of solvent vapours from the extractor and reduce the ingress of air with the feed material. A pneumatically operated slide gate is located in the feed system to prevent solvent vapours from escaping to the preparation area when the plant is not in operation.

Several types of extractor are marketed but for large plants the most efficient and trouble-free has been shown to be the rotary extractor. This extractor consists of a vertical cylindrical rotor divided into cells, rotating continuously within a vapour-tight shell. Each cell is fitted with a perforated discharge door at its base which is covered with a screen to support the material. At the discharge point, the door swings downwards, permitting the extracted material to fall into the discharge hopper. The bottom of the extractor shell is divided by radial baffles into compartments which comprise extraction stages and segregate the various strengths of miscellas (tallow/solvent mixes). Clean solvent liquid is pumped to the extractor and enters through manifolds above the cells. After percolating through the partially extracted material, the solvent falls into one of the stage tanks and is then pumped from stage to stage in a direction countercurrent to the rotation of the rotor, discharging over a fine mesh screen into a rich miscella buffer tank. This type of extractor has several advantages and some of its more important features are:

(1) Its simple drive gives many years of trouble-free operation and its fractional horse power motor and gearbox give infinitely variable speed.

80 Fat extraction

(2) There are no rubbing surfaces between the meal and the extractor wall.

(3) The extractor is compact and has an economic diameter to height ratio which can be varied to suit the characteristics of the material to be processed.

(4) Flooding of the material in each cell can be observed and controlled. (5) The bed of extracting material is sufficiently stable to act as a filter bed

for the solvent or miscella passing through it and this, combined with the effect of the self-cleaning tent screen, results in an exceptionally clear miscella being fed to the evaporation system.

(6) The quick opening cell doors and controlled door impact keep the draining screens clean. In addition these screens can be replaced easily when necessary.

(7) The stage tanks and miscella tanks are so designed that sufficient capacity is available to take the entire Uquid content of the extractor without hazard in the event of a shut-down resulting from a power failure or some other cause.

(8) A self-closing seal is provided on the extractor to limit the excess pressure inside the machine to a maximum of 2.26 kPascals (kPa) (23 cm water gauge). If the pressure exceeds this figure, the seal is blown off to atmosphere by the escaping vapours. It is automatically replaced when the pressure again falls below 2.26 kPa.

(9) Under normal circumstances the pressure in the extractor is controlled slightly below atmospheric pressure by a pneumatically operated control valve.

The type of machine used for removing the solvent from the defatted meal leaving the extractor is selected to give the required product characteristics. In greaves extraction plants the horizontal type of desol-ventizer, or Schnecken is normally used (Becker, 1971).

The miscella distillation system is designed to minimize contact time between the oil and heated surfaces. In the usual, simple two-stage evaporation system miscella is pumped from the miscella buffer tank to a steam heated rising film evaporator via a small steam heater which raises the temperature of the miscella to its boiling point. The contact time between miscella and the steam heated tubes of the rising film evaporation is only about 1 s, during which time though the oil concentration of the miscella reaches about 95%. Final traces of solvent are removed from the oil in a disc and doughnut type stripping column operating under a vacuum. A heat economizer in the form of a vertical cylindrical vessel employs the heat in the evaporator exit vapours to maintain the temperature of the clean solvent fed to the extractor. In addition it helps to smooth operating fluctuations in the evaporation system. The rate of flow of miscella to the evaporation system is set manually and is adjusted to balance the feed of miscella from the extractor. From this point the distillation system operates automatically, delivering desolventized dry oil to the oil cooler or storage vessel. Vapours from both the desolventizer and evaporator are condensed in water cooled vertical shell and tube condensers.

It is inevitable that a certain amount of air will enter the system with the

/. Davie 81

feed material and through minor leaks in the vacuum vessel. It is therefore necessary to provide a system for recovering the solvent from the air leaving the plant. Air is normally recovered by drawing all the vapours in the common vent system through some type of absorption unit before discharging to the atmosphere. The type of unit most favoured in modern plants employs a non-hygroscopic mineral oil of edible grade to absorb the solvent. Care must be taken though in the design of these systems when extracting greaves because of the corrosive nature of the vapours. Pressure in the vent system is maintained sUghtly below atmospheric by a fan of sparkproof design and is controlled automatically by pneumatically operated valves. Correct operation of the absorption system ensures that the air leaving the plant has a solvent content well below the lower explosive Umit.

The cost of utilities per tonne of material processed will depend largely on the system adopted, the nature of the produce required and the economizer systems incorporated in the plant. Typically though, steam consumption in the extraction plant itself will be approximately 300-320 kg/t of material processed; cooling water consumption, assuming water at a maximum temperature of 30 °C, will be approximately 15 m^/t and power consumption will be approximately 20 kJ/t.

All these modern fat extraction techniques are aimed at ensuring safety, operating economy, close control of product characteristics and ease of operation. Cost obviously is of paramount importance and everyone is well aware of the costs of energy, plant, maintenance and running processes. Of increasing importance though is the need to make allowances in design to ensure clean air and water. Suppliers and operators have had in recent years to develop systems for control that are within the Factory Act and that are also acceptable to both Public Health and Environmental Control Authorities. In addition, the AlkaHne Inspector has to be satisfied.

Single cell protein

'Single cell protein comes of age' said Dr Martin Sherwood in the New Scientist in 1974. In this article he reported that for nearly 20 years, petrochemical companies have been developing processes for growing protein on hydrocarbons, or simple hydrocarbon derivatives. Several processes are now on stream or are almost ready to go on stream on a large scale. Although the whole outlook has recently received a setback with the decision of British Petroleum (BP) to abandon its programme, the situation is reviewed in the article with particular reference to a plant installed in the USSR.

From the European viewpoint, methane is an ideal substrate, as it is an indigenous raw material (natural gas) and it is a raw material that can be produced from renewable resources. Shell is a company still working with methane while Imperial Chemical Industries (ICI) has switched to using methanol as the substrate. In other areas of the world agricultural by-products such as bagasses, molasses and whey and agricultural and urban wastes such as manures and cellulose are more readily available and Ukely to be used. Worldwide, however, the current production of SCP is considerably less than 1 million t per annum (Barlet, Holve and Meriel, 1978).

82 Fat extraction

Sample % Water % Oil extracted Sample on a dry weight basis

Original powder 9.0 8.6 Powder cooked without moisture addition — 7.7 Powder cooked with moisture addition 2 L 4 9.9 Powder cooked with moisture addition and flaked 18.9 13.3 Powder cooked with moisture addition and pelleted 20.6 14.7 Powder cooked with moisture addition, pelleted and flaked 19.4 18.8

The Russians now have in operation a plant capable of processing 115 t/day of biomass.

At Hull University research into the use of mould as a source of both protein and edible oil is currently being undertaken (Routledge, 1982). Simon-Rosedowns Ltd are developing means of extracting the oil from these moulds that are grown on sugar substrates. Moulds are being grown at Hull under conditions where they have ample carbohydrate food supply but hmited phosphorus or nitrogen supplies. Under these conditions the mould grows at a steady rate, synthesizing edible oil, but does not require either elaborate cooHng arrangements to limit temperature rise or highly energy intensive aeration systems to maintain oxygen supphes. It is hoped

In 1972 the Russians, who were preparing biomass from yeast cells grown on a substrate of mineral oil, petroleum gas or a waste carbohydrate, wished to develop a process to extract the lipids in the cells they were producing from these impure hydrocarbon substrates. A suitable process was developed by Simon-Rosedowns Ltd and was patented in 1977 (Alexander, Foster and Farmery, 1977). This process is discussed in the following section. The initial product yeast from the fermentation is a mass of unicellular organisms and when harvested is obtained as a fine powder. For extraction of the contained oil, the walls of the yeast cells must be first ruptured, so that when a solvent for the oil is percolated through the mass of yeast cells, the oil makes intimate contact with the solvent. In an attempt to simultaneously cause cell breakdown and to bring the powder into a form suitable for percolation by the solvent the powder was formed into flakes. The actual procedure involved conditioning the harvested powder, at say 9% moisture, by the addition of steam and water to yield a product of 20% moisture at 70-80 °C. This product is then fed to a pelleting machine and converted into pellets of approximately 5 mm diameter. The pellets are then fed to flaking rolls where the roll gap is set at 0.075-0.2 mm. The resulting flakes are well formed but, being wet, have low mechanical strength. Before further treatment they are therefore dried to a moisture content of 8%.

These flakes are then solvent extracted with n-hexane in a continuous solvent extraction plant similar but not identical to the one described in Figure 5.6, The importance of rupturing the yeast cells by flaking is shown in Table 5.1. In each of the samples the yeast after treatment was extracted for 7 h in a standard Soxhlet extractor.

Table 5.1 EFFECT OF FLAKING, PELLETING A N D MOISTURE A D D I T I O N ON T H E E X T R A C T I O N OF OIL FROM SCP

J. Davie 83

Discussion

Fat extraction plays a vital role in supplying the world's food needs. That part of it which can be considered as upgrading waste for feeds and foods is not insignificant. In this article an attempt has been made to give a broadbased general view of how fat extraction plays its role in the modern animal and poultry rendering industry and has also highlighted some of the various options open to the processor. Beyond doubt there is pressure on everyone in the rendering and related industries to balance the costs of running their business against the demands of the environmental lobby but there is evidence that they (the processors) can adjust.

In this review some information on a plant designed to extract lipids from some novel lipid-containing potential feed and food (biomass) obtained from a waste paraffin feedstock has also been discussed.

For readers requiring more information about potential edible and non-edible fats and oils from new sources a recently published monograph is available (Pryde, Princen and Mukherjee, 1982). This monograph also includes the utilization, or potential utilization, of by-products as valuable raw materials.

References

A L E X A N D E R , D . G . , FOSTER, A. and F A R M E R Y , D.W. (1977). British Patent No. 1466853.

A N O N . (1982a). JAOCS, 59, 7A A N O N . (1982b). Newswatch, (March 1st), 3 B A R L E T , Α . , H O L V E , W.A. and MERIEL, J. (1978). Food Engg Int., ll(Nov),

45 B E C K E R , K W . (1971). JAOCS, 48, 299. B L O G G , M.J. (1976). Process Biochem, l l(Dec), 9 B U R T O N , F. (1979). Render, Oct. G R A N O F S K Y , D . B . (1978). The National Provisioner, (April 15), 8 P R Y D E , E . H . , PRINCEN, L.H. and M U K H E R J E E , K.D. (1982). New Sources of

Fats and Oils, AOCS Monograph 9. American Oil Chemists Soc. Champaign, USA

R O U T L E D G E , C. (1982). Personal communication S H E R W O O D , M. (1974). New Scientist, 64, 634

ultimately that by the correct choice of materials and conditions to be able to produce edible oils to suit. The oil extracted from the moulds currently being studied are rich in γ-linolenic acid thus they may have potential use in the treatment of multiple sclerosis.

To conclude this section the creation of biomass (biotechnology) is really in its infancy and the real story of microbial protein may have only just begun. Recently it has been announced that the German Democratic Republic have developed a protein, trade name Fermosin, jointly with Soviet scientists (Anon., 1982b).

PRECIPITATION AND RECOVERY OF WHEY PROTEIN WITH CARBOXYMETHYL CELLULOSE AND PREPARATION OF A SOLUBLE COMPLEX BY AMMONIA ADSORPTION

P.M.T. HANSEN AND R. BALACHANDRAN* Department of Food Science and Nutrition, Ohio State University, USA

Introduction

The recovery of protein in waste fluids from food processing plants is a difficult problem especially when the protein concentration is low and the fluid contains relatively high concentrations of non-protein constituents. An example of the problem is represented by the efforts to reclaim protein in whey from cheese and casein manufacturing in which the protein content is 0.5% or less while other soHds constitute as much as 6-7%, principally in the form of lactose, milk salts and lactic acid. Considerable advances have been recorded over the last decade for whey protein recovery through a number of processes, including ultrafiltration (Kosikowski, 1979). These processes have largely been directed towards the treatment of sweet whey from cheese manufacture (pH 6.0-6.5). Recovery of protein from acid whey (pH 4.5), resulting from cottage cheese or caseinate production, presents additional problems arising from the presence of mineral or organic acids, which limit the acceptability of the recovered protein product, unless, by special efforts, the acid constituents have been eliminated. In this respect, complexing of proteins by anionic polyelectro-lytes including polysaccharides (Hansen, Hidalgo and Gould, 1971), polyphosphates (Jones et al., 1972; Hidalgo, Kruseman and Bohren, 1973), and polyacryHc acid (Sternberg, Chiang and Eberts, 1976), is an alternative approach for protein reclamation which would appear to be particularly suited for acid whey systems since the complex formation requires a low pH environment to allow the isoelectric precipitation to occur.

A number of polyelectrolytes are already in use for the treatment of sewage, the clarification of industrial wastewater and the purification of contaminated streams, rivers and lakes (Gutcho, 1972). Treatment of water with polyelectrolytes for purification is primarily designed for removal of suspended or dispersed solids and not for the purpose of reclaiming the constituents for further use. However, it should be recognized that the availability of food grade polyelectrolytes, including a variety of food gums, offers the possibility of applying the principles of complex formation for reclamation and upgrading of proteins from food waste fluids (Ledward, 1979; Stainsby, 1980).

*On deputation from Division of Dairy Technology, N D R I , India

85

86 Precipitation and recovery of whey protein

Early work by Smith et al. (1962) demonstrated nearly complete protein recovery from soybean whey by the use of edible food gums and detergents under controlled conditions, with a corresponding reduction of 8-18% in biological oxidation demand (BOD). A similar approach, using carboxy-methyl cellulose (CMC), has been used in our laboratory for the recovery of proteins from sweet and acid whey (Hansen, Hidalgo and Gould, 1971; Hansen and Crauer, 1971). A patent for the continuous fractionation of whey by CMC treatment has been issued to the DeLaval Separator Co. (Crauer, 1976).

Interactions between proteins and anionic polysaccharides are pH-dependent and involve the formation of complexes at a pH usually below the isoelectric point of the protein. The complexes are least soluble when the proteins and the polyanions are matched in predetermined amounts and the pH is adjusted to form isoelectric aggregates. Excess of polyanion in the mixture causes formation of soluble complexes, whereas high salt concentration interferes with complex formation (Hidalgo and Hansen, 1969; 1971). In a series of papers, Hill and Zadow (1974; 1978 a, b) and Zadow and Hill (1975; 1978 a,b) have described the mechanism of interaction between whey protein and carboxymethyl cellulose and its derivatives. These investigators reported that the efficiency of precipitation was increased with increasing degrees of substitution of CMC, and more importantly, that the tolerance of the induced complex towards high salt concentration was also increased. Thus, at a degree of substitution of CMC of 1.4, i.e. with, on average, 1.4 of the three available sites per monomer substituted, a 90% efficiency in protein recovery was recorded, without resorting to dilution of the whey system. In the USA, specific regulations prevent the use in food of carboxymethyl cellulose with a degree of substitution greater than 0.95 (Code of Federal Regulations, 1980). However, it is worthwhile to note that the Australian authors have observed that derivatives of CMC modified with substituents of increasing hydrophobicity were also increasingly effective as protein precipitants. Thus, selection from these various types may make it possible to meet the requirements of effectiveness within the limitations of the current food regulations.

Isolated whey protein complexes typically contain 70-80% water, and on a dry weight basis, approximately 65% protein and 35% of hydrocolloid. Although in some cases it may be possible and desirable to separate the complexes into their constituent protein and precipitant (Hidalgo and Hansen, 1971; Sternberg, Chiang and Eberts, 1976), separation may not be warranted if it were possible to utilize the intact complex for edible or technical purposes. Complexes of a nutritious protein with an edible food stabilizer, having specific functional properties, would be of interest to the food industry, for example for ice-cream stabilization. Such complexes may also have application in biomedicine in treatment of ulcer patients and for other special dietary needs.

While whey protein reclamation, using CMC as a precipitant, offers advantages with respect to low cost for investment in equipment, there have been up to now certain limitations to this approach. These relate to

(1) lack of salt tolerance of food grade CMC for complex formation which necessitates dilution of the whey system to counter this effect, and

p. Μ. Τ. Hansen and R. Balachandran 87

(2) problems in spray drying of the neutralized complex because of the extremely high viscosity of the product due to the presence of CMC.

In this chapter it is not intended to review the whole of protein recovery by charged polysaccharides because, as indicated earlier, several such reviews are already available. Rather the development of a scheme by which the acidification step is accomplished by the use of a cationic exchange resin in order to minimize the effect of ionic strength on directly acidified whey will be discussed. Furthermore, a method by which the isolated complex is dried conventionally or by freeze drying and subsequently converted to a soluble product by exposure to anhydrous ammonia is reviewed (Girdhar and Hansen, 1974; Holsinger, Hafez and Hansen, 1977).

Materials and methods

Cottage cheese whey (pH 4.4) was obtained from a local dairy plant at the day of manufacture and kept refrigerated until use. Carboxymethyl cellulose was obtained from Hercules, Inc., Wilmington, Delaware. Four types of CMC were used differing in their degree of substitution (DS): 9M31F with a DS of 0.9 and of medium viscosity; 7LF with a DS of 0.7 and of low viscosity; 4M6F with a DS of 0.4 and of medium viscosity; and 4H1F with a DS of 0.4 and of high viscosity.

For protein precipitation the whey was warmed to ambient temperature (23 °C) and mixed with a cationic wet resin (Amberlite G-120, 100-200 mesh, H+) in sufficient quantities (approximately 12% wet weight) to reduce the pH to 3.2. The whey was then decanted and the resin reclaimed for further use. Solutions of CMC at 0.25% concentration were prepared in tap water and added to the acidified whey in predetermined quantities to yield maximum precipitation. Following the addition of CMC, the protein-CMC complex was permitted to settle by gravity and the supernatant decanted and discarded. The complexed protein was resuspended in water and recovered by gravity sedimentation. The resuspension/sedimentation procedure was repeated for a total of three times. The wet sediment was dried by freeze drying and the dry material treated with anhydrous ammonia gas according to Holsinger, Hafez and Hansen, (1977) as follows: the complex was placed in a vacuum chamber and a vacuum was drawn by aspirator to approximately 98 kPa (29 inHg). Dry ammonia gas was admitted to reduce the vacuum to approximately 17-34 kPa (5-10 inHg) which was maintained for approximately 30 min. The chamber was then degassed and flushed with air to remove excess ammonia and the product was retrieved.

The nitrogen content of dry complexes, receiving different treatments, was determined by the Dumas method using a Coleman Nitrogen Analyzer, Model 29. The total nitrogen content was converted to protein using the factor 6.38.

Results and discussion

A pilot plant process capable of handling 1 m^ (1000 €) of whey was previously developed in our laboratory for reclamation of protein by utilizing carboxymethyl cellulose and other hydrocolloids (Hansen and


CMC type Mixing ratio^ Yielct" Protein in dry complex^ (g/1) (%)

9M31F 0.6 5.00 69.0 7LF 0.7 6.03 67.4 4HIF 0.8 4.67 54.4 4MGF LO 5.07 61.1

^ml of 0.25% CMC per ml of whey (pH 3.2) required for maximum precipitation. of dry complex per Í of whey

''means of duplicate determination, Standard Deviation (means) 1.60%

increasing DS value, with type 9M31F with a DS of 0.9 showing the highest value. However, viscosity grade is apparently also an important factor, at least for gravity sedimentation, as shown by the improved yield and protein content for the medium viscosity grade, 4MGF, over the high viscosity grade, 4K1F, both with a DS of 0.4. The total protein recovery (3.9 g/€) was highest for type 7LF (DS 0.7 and of low viscosity), possibly because of the viscosity factor.

Figure 6.1 illustrates a possible approach that should lead to an improvement of the industrial process. Acidification with a cationic resin is readily achieved and controlled and results in an overall reduction in the electrolyte content, and thus in ionic strength, as indicated by the specific conductivity values listed in Table 6.2. The need for a regeneration cycle to reactivate the resin has been indicated in the diagram.

The complex recovered by centrifugation in the solids discharging separator may, if desired, receive a first wash prior to sohds ejection. The purpose of this wash is to remove lactose and other soluble constituents, by diverting rinse water to the separator in a controlled manner. The curd, which is normally well compacted, can be subjected to additional washing in a tank and subsequently collected by centrifugation or gravity sedimentation.

The proposed ammonia treatment of the dry complex was adapted from the process used by Girdhar and Hansen (1974) for the treatment of dry, isoelectric casein by a column arrangement operating at a positive pressure. The conversion, however, is more conveniently accomplished in a vacuum tank under slightly reduced pressure, an approach which is also

Crauer, 1971). While the pilot plant experiments demonstrated the feasibility of the process, using in most instances CMC with a DS of 0.70, it became evident that improvements would be needed to decrease the ionic strength of the mixture so as to exert better control over the complex formation and, more particularly, to circumvent the need for spray drying. This was desirable because of the extraordinary difficulties encountered in processing the highly viscous material resulting from neutralization of the acidic complex.

Since Hill and Zadow (1974) reported improved efficiency in protein precipitation using CMC of high DS value, it was decided to examine the performance of a number of available food grade CMC types. The results in Table 6.1 confirm that the protein content of the complex increases with

Table 6.1 Y I E L D A N D PROTEIN C O N T E N T OF COMPLEXES O B T A I N E D B Y D I F F E R E N T TYPES OF CMC

89

V a c u u m ^

Figure 6 . 1 Flow diagram for, manufacture of soluble CMC/whey protein complex

Table 6.2 S P E C I F I C C O N D U C T I V I T Y O F A C I D W H E Y B E F O R E A N D A F T E R

T R E A T M E N T

Whey pH Specific conductance ( 1 0 " ' m h o s / c m )

Untreated 4 . 4 4 5 2

Acidified, IN HCl 3 . 2 6 0 7

Cation exchanged 3 . 2 3 2 6


Figure 6.2 Change in pH of ammonia-treated CMC/whey protein complex during degassing in stream of ambient air and by prolonged vacuum treatment (pH of 5% solution)

rapidly lost when the product is exposed to a stream of ambient air. In contrast, ammonia is retained under simple vacuum exposure, even after 24 h. This suggests that removal and loss of ammonia occurs through contact of the treated complex with moist air.

The values in Table 6.3 show that ammoniation followed by degassing to pH 7.0 resuhed in a 7-11% increase in nitrogen, corresponding to a total retention of approximately 1.2% ammonia.

Ammoniation of complexed whey protein, dried by conventional means, would be expected to be an energy conserving process compared to alkali conversion in the wet state followed by spray drying. The question of food safety and nutritive value of proteins treated in this manner has been addressed by Holsinger, Hafez and Hansen (1977) who found no signifi-

Table 6.3 N I T R O G E N C O N T E N T O F W H E Y A N D W H E Y P R O T E I N / C M C

C O M P L E X E S B E F O R E A N D A F T E R A M M O N I A T R E A T M E N T

Product Nitrogen'' Ammonia Untreated Ammoniated uptake (%) (%) (%)

Freeze dried whey 1.68 — — Complexes

9M31F 10.81 11.54 0.89 7LF 10.57 11.76 1.45 4H1F 8.52 9.48 1.17 4M6F 9.58 10.60 1.24

""Means of duplicate determination, standard deviation (means) 0.15%

better suited for larger scale operation. The treatment results in the adsorption of ammonia to the dry matrix and apparently also in ammonium salt formation with carboxyl groups on the protein and the CMC (Girdhar and Hansen, 1974).

The treated product exhibits a pungent odor of ammonia which may be partly eliminated by degassing. The effect of degassing on the pH of the complex is shown in Figure 6.2. It may be observed that the ammonia is

p. Μ. Γ. Hansen and R. Balachandran 91

Acknowledgement

Supported in part by Public Health Service Research Grant FD-00117 through the Office of Research and Training Grants, Food and Drug Administration, Washington, DC.

References

C O D E OF F E D E R A L R E G U L A T I O N S (1980), Title 21 Ch. i . Part 182,1745 C R A U E R , L.s. (1976). Can. Patent 995,971 G I R D H A R , B.P. and H A N S E N , P.M.T. (1974). J. Fd Sci., 39, 1237 G U T C H O , S. (1972). Waste Treatment with Poly electrolytes, Noyes Data

Corporation, Park Ridge, New Jersey, USA H A N S E N , P.M.T. , H I D A L G O , J. and G O U L D , L A . (1971). J. Dairy Sci., 54, 830 H A N S E N , P.M.T. and BLACK, D . H . (1972). / . Fd Sci., 37, 452 H A N S E N , P.M.T. and C R A U E R , L.S. (1971). / . Dairy Sci., 54, 756 (abstract) H I D A L G O , J. and H A N S E N , P.M.T. (1969). / . Agr. Fd Chem., 17, 1089 H I D A L G O , J. and H A N S E N , P.M.T. (1971). / . Dairy Sci., 54, 1270 H I D A L G O , J., D R U S E M A N , J. and B O H R E N , H . U . (1973). / . Dairy ScL , 56, 988 HILL, R . D . and Z A D O W , J.G. (1974). J. Dairy Res., 41, 373 HILL, R . D . and Z A D O W , J.G. (1978a). Aust, /. Dairy TechnoL, 33, 97 HILL, R . D . and Z A D O W , J.G. (1978b). N.Z. J. Dairy Sci., TechnoL, 13, 61 H O L S I N G E R , V . H . , H A F E Z , Y. and H A N S E N , P.M.T. (1977). / . Agr. Fd Chem.,

25, 1109 JONES, S.B. , K A L A N , E .B . , JONES, T.C. and H A Z E L , J.F. (1972). Agr. Fd Chem.,

20, 229 KOSIKOWSKI, F.V. (1979). J. Dairy Sci., 62, 1149 L E D W A R D , D . A . (1979). Proc. 27th Easter School Agrie. Sei., Univ.

Nottingham, 205 M O R R , e .V., SWE NSON, P.E. and RICHTER, R.L. (1973). / . Fd Sci., 38, 324 SMITH, A .K. , N A S H , A . M . , E L D R I D G E , A . C . and WOLF, W.J. (1962). / . Agr.

Food Chem., 10, 302 STAINSBY, G. (1980). Food Chem., 6, 3 S T E R N B E R G , M., C H I A N G , J.P. and E B E R T S , N.J. (1976). / . Dairy Sci., 59, 1042 Z A D O W , J.G. and HILL R . D . (1975). / . Dairy Res., 42, 267 Z A D O W , J.G. and HILL, R . D . , (1978a). J. Dairy Res., 45, 85 Z A D O W , J.G. and HILL, R . D . (1978b). N.Z. J. Dairy TechnoL, 13, 162

cant differences in growth of weanling rats fed high levels of ammoniated casein.

The market potential for complexed whey protein has not been explored to any large extent. In the early stages of development of the pilot plant process it was observed that the complexes provided adequate stability for milk fat emulsions over a wide range of pH and calcium ion concentrations (Hansen, Hidalgo and Gould, 1971). Complexes of whey protein and CMC as well as other hydrocolloids may have application as whipping agents and foam stabilizers (Hansen and Black, 1972; Morr, Swenson and Richter, 1973).

THE MECHANICAL RECOVERY OF MEAT—A NEW LOOK AT THETECHNOLOGY

P.B.NEWMAN Meat Research Institute, Bristol, UK

Introduction

In recent years, as sources of protein for human food have become more expensive and less plentiful, the major animal-protein consuming countries have sought alternative resources of both animal (e.g. krill protein, Spillmann, 1980) and vegetable proteins (e.g. alfalfa. Knuckles et al., 1972) and have investigated methods for more efficient use of existing protein foods. Since meat is one of the most expensive proteins, maximal efficiency in its use is of prime importance.

The last decade has seen new food ingredients from meat and its by-products, suitable for human consumption, become available to the food processor. As the properties of blood plasma protein, bone protein and mechanically recovered meat, have become better understood, they have found their place alongside other relatively new but now well-established protein sources such as soya and other bean proteins.

The recovery of protein from meat by-products is not new. Solubilization of muscle proteins for analytical purposes is well documented (e.g. Keller and Block, 1960). However, for reasons such as high cost or protein denaturation, these processes have not been developed as methods of commerical production. Jelen et al. (1978) and Jelen, Earle and Edward-son (1979) have pubhshed details of alkahne and acid solubilization techniques which retain many of the important functional properties of the muscle proteins. Young and Lawrie (1975) and Swingler and Lawrie (1977) have processed animal protein sources such as lung and stomach, unsuitable in their native form, into edible food. Henning (1974) has utilized the residual recovered meat protein fraction which has a high collagen content, while Jobhng (1978) and others have looked at ways of extracting protein from bones and using it in processed foods.

In a recent review. Field (1980) estimated that there is a potential 289000 tonnes of recoverable pork and beef meat from current US livestock production. In the Netherlands, between 2500 and 5000 tonnes of mechanically recovered pork is produced (Bijker et al., 1979). Field (1981) concluded that worldwide 2.3 million tonnes of recoverable red meat could be added to the food supply; to which could be added an estimated 1 million tonnes of recoverable poultry meat and 2 million tonnes of recoverable fish meat. Newman (1981) estimated that in Britain

93

94 The mechanical recovery of meat - a new look at the technology

Separation of meat from bone

Since the interest of the meat trade was awakened to the potential of waste meat recovery from bones by the Japanese and their earliest fish meat collectors, 'Gyonikusaishuki' {see Figure 7.1), the technology has improved tremendously, although the mode of action of these first fish presses is still the basis of much of today's machinery.

Mechanical recovery however, is not the only approach by the trade in finding effective methods of separating meat from bones. Figure 7.2 illustrates the variety of methods used. Because mechanical methods seriously damage texture, the last few years have seen an increase in research into alternative processes.

Eastern European countries have been particularly interested in pneumatic, cryogenic and other physical methods of separation. Ilyukhin, Kiselev and Shebel'kov (1981) describe a method of separation which, after freezing and crushing, subjects a continuous centrifugally-fed thin layer of the crushed meat-bone mix to an air stream at - 2 0 °C. The meat, connective tissue and bone particles were separated by their different densities and resistance to movement within the air stream. A similar method of separation is suggested by Baier et al. (1980). Limonov, Snitsar and Mamedov (1978) outhne a low frequency vibration method, whilst Shuvaev, Afanas'ev and Yu (1978) have demonstrated a combined soaking and air jet technique which they claim removes much larger pieces of native meat from bones than other methods. A similar method of removal after partial heating to 40°C is outlined in a recent patent (St. Clair Fisheries, 1980). Bombarding bones with high velocity frozen meat particles is suggested by Roth (1980). High velocity water and air jets (13.8-55.2 MPa, i.e. 2000-30001b/in^) are used by Gordon (1980), whilst the recovery of meat by abrasion with the meat-bone mixture submerged in a solution of alkaU salts and saccharides is described in a recent patent (Bibun Co. Ltd, 1980). The method of Herrmann and Nitzsche (1981) combines liquid comminution and two different deboning machines, the separation resulting in a high yield, good quahty meat product.

Despite the increased interest in other methods, in attempts to improve texture, mechanical methods of meat recovery are still the most widely

alone the monetary value of this protein source was in the order of £9 million per annum which represents 1-1.5% of total edible meat production.

An earlier publication on mechanically recovered meat (MRM) (Newman, 1981) contains a more detailed study of the mechanics of red meat recovery and should be read in conjunction with this paper which concentrates on changes in technology, legislation and improvements in our understanding of the processing of bone, the presence of bone marrow, lipid oxidation, bacteriological considerations and the inclusion of MRM in processed products. Further background information can be found in Field (1981) on mechanically recovered red meat with particular reference to the USA, and Froning (1981) who discusses mechanical recovery of fish and poultry meat.

95

ON

C

(3

96 Separation of meat from bone

Chemical ^ Physical

I I I Hydraulic | Acoustic Thermal | Magneto Inorganic Biochemical Electro- Optical Mechanical Pneumatic Hvdro-

P^^y^'^^l I I dynamics 0 Grinding

^ Alkaline Screw pressure

1 Neutral Enzymatic Ultrasound Percussion Cooking

Ö Acid Impaction

¿ Mincing

Granular

·—rartial—• aggregation

Tota] •

<υ ^ Thermal S High Laser Sieving expansion Plasma

Filtration frequency beam Centri'fugation ^ cutting nitration separation separation , ^ Cutting ^

0 Flotation i

- High pressure Density "^.^aring High pressure

1 Centrifugation water/ f luid . shredding a.r/gas jets

S- J^^^ ^ Flaking

Scrapmg

Milling

Protein solution ' ' ' Separated meat and bone 1 and suspension

Figure 7.2 Methods used to separate meat from bone

p. Β. Newman 97

Figure 7.3 Baader meat separator, model 694

convenient size. UnUke the sieve machines which are continuous, the press-type are batch producers. Automatic hoppers deliver a predetermined amount into a thick-walled cylinder. A hydraulically-powered piston compacts the mix under 10-25 MPa (100-250 atm), the meat and fat *flow' out through screens or along microgrooves, whilst bone is retained and then ejected, {Figure 7.5). (A typical meat recovery layout is shown in Figure 7.6). Although other mechanical methods are being developed commercially, these two types of mechanical separation are still monopolizing the market.

*Paoli International, Rockford, Illinois. tHydrau, B V , Holland. iProtecon, B V , OSS, Holland.

used commercially. The sieve-type (e.g. Paoh and Baader*) requires the bones to be pre-ground in a bone grinder. The bone material is forced into the nip between belt and drum {Figure 7.3) or between microgrooves {Figure 7.4). The softer meat and fat pass through while the harder bone and connective tissue are retained. In the press-type (e.g. Hydrant and Protecont) machines the bones need not be ground, merely broken to a


Figure 7.4 Paoli red meat deboner, model 21

Recent publications by Nordischer Maschinenbau (1979), Holding (1979) and Protecon BY (1980) show that there has been little fundamental change in the technology, merely refinement. The machine described by Machinefabriek Amersfoort BV in their patent (1980) controls the pressure applied to the bones by monitoring the calcium content of the resultant meat slurry.

Processing and storage problems associated with MRM

B O N E PROCESSING

Bones for processing must be subjected from the outset to the same hygienic control as meat. In the UK there are no restrictions on the bones that can be used except those prohibited by the Offals in Meat Products' Order (1952). In practice, however, processing of marrow bones results in

p. Β. Newman 99

Figure 7.5 Protecon press-type meat separator

reduced stability (Froning and Johnson, 1973), and high fat content in the final product (Moerck and Ball, 1973) and this, together with the limited amount of meat adhering to them after hand-boning, makes them commercially unattractive.

Detectable bone or grittiness in the final product markedly reduces panel acceptability scores (Chant et al., 1977) and recent improvements to machines, particularly the press-type have concentrated on an improved separation of bone from the fat/lean paste. Protecon machines, for example, have three screens with 1.6 mm holes, slightly offset to retain large bone particles. When properly adjusted these machines will not pass organoleptically detectable bone. Nevertheless, Mawson and CoUinson (1974), Murphy and Engel (1979) and others have shown that under normal plant operating conditions, machines frequently exceed either maximum acceptable calcium levels, i.e. 1%, {see Table 7.1), or organoleptically detectable bone particle size, i.e. 0.46mm.

Opinion differs as to the best way of measuring bone content whether it be by bone particle measurement (Geräts and Terbijhe, 1975), calcium content in the meat or ash (Hujnakova, 1976), or by bone sieving (Ramsey and Rossi, 1976). Bijker, Geräts and Fransen, (1978) and Bijker et al. (1979) have developed a method for the determination of hard bone residues using KOH digestion and light microscopy. Although accurate, correction factors are necessary because of differences in desiccation and evaporation rate of bone from different species. It is also time consuming and unsuitable for routine use.

100

Figure 7.6 Commercial meat recovery layout using a beehive deboner

p. Β. Newman 101

Machine type A Β C D

Bone source pork back 0.45 0.15 0.46 0.73 pork neck 0.40 0.15 0.23 0.55 pork ham L65 0.16 1.52 pork picnic L81 0.17 1.44 beef chine/neck 0.04 0.24 0.40

Sources: Goldstrand (1975); Anon. (1976). Machine A is continuous action separating ground bones through a strainer under pressure. Machines Β and C are batch action using crushed bones forced against a stationary strainer. Machine D is as A but using a stationary strainer.

The level of calcium in boneless meat is usually less than 0.01%. Because mechanical recovery inevitably results in an increase, measurement of the calcium content has been used to estimate the nutritional quahty and composition in the recovered meat. Germs and Stennenberg (1978) compared the results obtained from two different chemical methods with calcium determinations by atomic absorption spectrophotometry and found significant differences between chemical methods. The results were considered sufficiently accurate for in-plant measurement. However, because organoleptically detectable bone reduces consumer acceptance, it may be useful, for quality control purposes to monitor both bone particle size and total calcium levels.

M A R R O W B O N E S

Physical and chemical properties of marrow bones differ from those of compact bones, major problems arise from the amount and degree of unsaturation in the fat, release of haem pigments and the accumulation of possible harmful constituents in the bone marrow.

The amount of lipid in compact bone such as ribs and vertebrae is very small, usually less than 0.1% (Leach, 1958), and the degree of unsaturation in red meat compact bone is corresponding small. However, there is an increased level of polyunsaturated fats, particularly phospholipids, in red meat marrow bones (Moerck and Ball, 1973), with pork marrow bones containing most unsaturated fat, beef the least. Overall, red meats are considerably lower in unsaturated fats than poultry, which in turn, contains less than fish. Jantawat and Dawson (1980) showed there were significant differences in the lipid, especially phospholipid composition of different types of poultry meat. This has an important bearing on storage and stability since the rate of polyunsaturated fatty acid oxidation is exponential with the number of unsaturated bonds (Igene, 1979).

Haem pigments composition (measured by free iron) is normally below 0.01% in hand-boned meat (Eastoe, 1961). In MRM this level is much higher due to the presence of marrow haem-released pigments. The consequences are threefold: First, the colour is 25-30% darker than

Table 7.1 M A C H I N E P E R F O R M A N C E R E L A T E D TO P E R C E N T A G E C A L C I U M C O N T E N T OF D E B O N E D M E A T S


Figure 7.7 meat

0 -20 -10 0 1̂0 *20 .30

S t o r a g e t e m p e r a t u r e CO

Haem protein and lipid oxidation rate constants in mechanically deboned turkey

Hultin, McDonald and Kelleher (1979) showed that although the chemistry of lipid oxidation in fish differs from that in other meats, the presence of iron is as essential.

Janky and Froning (1975), as well as Takama and Mizushima (1978), have suggested that protein denaturation and breakdown may result from the haemoglobin as well as myoglobin undergoing ion binding by reacting either with metal ions solubilized by contact of the meat with the sides of the vessel or by combination with the enhanced levels of calcium ions released from bones. Jankiewiez and Pisula (1979), showed in model sausage systems the addition of bone marrow accelerated the rate of oxidative fat changes; pig marrow having a much greater effect than beef. Also the stability decreased as the initial ratio of unsaturation to haemin content increased.

The third restriction on the use of marrow bones is in their ability to accumulate a number of constituents harmful either by direct action or a concentration effect. Lead and heavy metals (Schroeder and Tipton, 1968), barium and strontium, especially strontium 90 (Sowden and Stitch, 1957), are undesirable since they are accumulated irreversibly. The level of barium, though, remains relatively stable although its role in metabolism is

hand-boned meat (Goldstrand, 1975) and while this may improve the final colour of a number of processed products it severely limits the amount of MRM that can be used in products such as the British sausage.

Increased intake of iron in the diet could be beneficial (Food and Nutrition Board, 1980). However, work by Mahoney etal. (1980) indicates that the bioavailability of iron from meat may be depressed by the action of mechanical recovery. Secondly, there is an interactive catalytic effect between haem pigments and unsaturated lipids, both of which undergo oxidation {Figure 7.7), even at temperatures below freezing (Janky and Froning, 1975).

p. Β. Newman 103

LIPID O X I D A T I O N

The high fat content (often 20-30%) of MRM compared to hand-boned meat {Table 7.2) severely limits its storage life. The intereaction between haem protein and lipid has already been discussed, but many other physical factors such as light, storage temperature and gaseous medium can influence the rate of lipid oxidation (Janky and Froning, 1975; Uebersax, Dawson and Uebersax, 1978). Particle size was also shown to be an important factor (Schnell et al, 1973).

Several workers have investigated the use of carbon dioxide 'snow' chilling, liquid nitrogen and blast freezing and the effect of different atmospheres with freezing medium, to improve storage of mechanically recovered meats. Jantawat (1979) concluded that samples of poultry MRM could be stored for up to four months at -18°C when vacuum-packed or packed in carbon dioxide or nitrogen and that freezing before packing was more beneficial than just chilling and packing.

Igene et al (1980) showed that all the lipid components were involved in loss of keeping quality with the phospholipid fraction most responsible, while the involvement of the neutral lipids was dependent upon the degree

unclear. Recent evidence {see Report of Select Panel, USDA, 1977) suggests that in a normal diet these elements are unlikely to be found in harmful concentrations. However, where MRM may form a much greater part of the diet, or where much more rapid uptake may occur, as in young children, the effects are less predictable.

Higher levels of fluorine and calcium have been discussed elsewhere (Newman, 1981), and, but for a few exceptions mainly relating to calcium sensitivity and osteoporosis, they can generally be considered beneficial. Indeed, in the UK, calcium supplementation of foodstuffs such as flour has been considered desirable for many years and is at present legally required.

Accumulation of antibiotics in most meat animals is unhkely to cause any problems since levels are generally low (Honikel, 1976). However, where animals are grown by intensive feeding methods, e.g. pigs, veal and, in particular poultry, much higher levels can be demonstrated (Waltersdorf and Schmidt, 1976). Recent work by Honikel et al (1978) has shown that tetracycHnes are especially resistant to breakdown below 60 °C. Again, in a normal diet their overall effect in MRM is unhkely to be a health hazard. (Within the UK there is no specific legislation on the regulation of antibiotic residues, therapeutic or growth-promoting, in fresh meat but it is hkely that this aspect will be covered in the new regulations.)

Although only limited data are available (Arasu, 1980), the elevated levels of nucleic acids found in MRM compared with hand-boned meats are still well below the levels concluded by the US Protein Advisory Group (1970) to be safe.

The problems which marrow bones can cause if used in mechanical meat recovery make it essential to measure the quantity of marrow and marrow bones used. Field etal. (1980) have demonstrated the relationship between increased marrow and pigment concentration suggesting that this could be a useful measure of bone marrow.


Table 7.2 C O M P O S I T I O N O F M E C H A N I C A L L Y R E C O V E R E D M E A T

( A V E R A G E A N A L Y S E S )

Bone source Percentage of total composition Total fat Crude protein Ash Calcium

Pork: ham 39.0 (27.9) 10.2 (15.7) 4.0 (0.5) 1.4 (0.03) picnic 42.3 (22.3) 9.1 (19.1) 3.7 (0.7) 1.2 (0.04) loin 29.5 (23.6) 14.0 (16.7) 1.8 (0.7) 0.4 (0.04) neck 27.3^ 14.7« 1.5« 0.4« Boston butt 26.0 (12.8) 13.5 (19.2) 2.7 (0.9) 0.7 (0.08) ribs 23.0^ 15.5« 1.2« 0.2« mixed bones 3 1 . Γ 12.5« 1.4« 0.3«

Veal: shoulder 7.6 (3.0) 12.8 (20.2) 5.4 (0.9) 1.8 (0.04) frame 6.8 (5.6) 17.5 (18.8) 2.7 (0.9) 0.7 (0.05) backs 5.8 (3.7) 16.0 (18.7) 2.2 (1.1) 0.5 (0.04)

Cow: rib 23.2^ 16.4« 1.2« 0.2« rib, plate 31.8 (31.6) 12.9 (14.1) 4.5 (0.8) 1.5 (0.01) rump 41.9 (11.8) 10.0 (17.6) 4.3 (0.8) 1.5 (0.08) short loin 33.4 (22.5) 11.6 (16.4) 4.3 (1.0) 1.5 (0.01) neck (choice) 13.7 (9.0) 17.2 (19.3) 3.4 (1.0) 1.1 (0.06) plate (choice) 32.7 (27.9) 11.4 (13.2) 4.3 (0.5) 1.5 (0.05)

Mutton: carcase 19.7 (19.1) 19.1 (19.6) 1.4 (1.0) 0.2 (0.02) breasts 36.5 (38.1) 15.0 (15.5) 1.2 (1.0) 0.1 (0.02)

Chicken: back/necks 22.0 12.3 spent layers 22.3 14.1 broiler neck 17.6 (16.1) 12.4 (12.0) 0.7 (0.5) 0.04(0.01)

Turkey: frames 13.6 12.8

Fish: English sole 2.6 13.4 4.9 true cod 2.1 15.6 4.3 ling cod 4.6 17.2 3.5 rockfish 7.3 16.2 6.5

Sources: Ackroyd, 1979; Crawford etaL, 1972; Field etal., 1974; Field etai, 1976; Froning, 1970; Satterlee et al., 1971; Froning and Johnson, 1973; Gründen et ai, 1972. «Denotes data obtained from press-type machines; other data obtained from sieve-type machines. Figures in brackets relate to data from hand-boned material (where available).

of lipid unsaturation and the length of time in frozen storage. Topol'nik and Ratushnyi (1980) and Younathan, Marjan and Arshad (1980) showed in red meats a decrease in both phospholipid and triglyceride lipid during storage with a corresponding increase in their decomposition products. Igene (1979) concluded that the breakdown products from the phospholipid fraction were primarily responsible for off-flavours in stored processed meats and meat products. Kunsman, Field and Kazantzis (1978) concluded there was no significant difference in the manner or rate of lipid breakdown in MRM or ground beef.

The overall conclusions of Jurdi, Mast and MacNeil, (1980); Uebersax, Dawson and Uebersax (1978) and Kraft et al. (1979) were that cryogenic freezing prior to conventional frozen storage (-20°C) was significantly more effective in prolonging the storage life of MRM than blast freezing or

p. Β. Newman 105

MICROBIOLOGICAL C O N S I D E R A T I O N S

The effect of mechanical recovery on the bacterial populations of the product and the measures available to reduce numbers has been reviewed by Field (1981), Froning (1981) and Newman (1981) amongst others. More recent work has examined the possible interactions between lipid oxidation and microbial spoilage in the meat and the development of alternative methods for reducing the bacterial population during storage. Branen (1979) discussed the possibility of extending shelf-life by inhibiting the interaction between oxidative rancidity and microbial spoilage. Raccach and Baker (1979) showed inoculation of lactic acid bacteria into mechanically recovered poultry meat inhibited the growth of the normal psychro-trophic bacterial population and retarded spoilage. They noted no adverse effects on the protein, colour, water-holding and emulsifying capacities of the meat. Dubois, Beaumier and Charbonneau (1979), found similar results. Williams etal. (1980) showed that freezing ground beef with solid carbon dioxide at —70°C before subsequent storage at normal freezer temperatures (-20°C) significantly reduced the resident bacterial population. Kraft et al. (1979) confirming the effectiveness of cryogenic freezing, found that there was a change in the flora as pseudomonads survived cryogenic freezing but not conventional mechanical freezing.

Some natural herbs and spices used in the formulation of processed products using mechanically recovered meat may be acting as bacter-iocides. MacNeil et al. (1973) stated that the herb, rosemary, was effective in extending shelf life, although Farbood, MacNeil and Ostovar (1976) were unable to show the same effect. However, recent work by Shelef, Naglik and Bogen (1980) showed that both Gram-positive and Gram-negative bacteria were susceptible to a number of herbs particularly sage and rosemary. At 0.3% they were inhibitory but bactericidal at 0.5%;

chilling, and any form of vacuum packaging with or without an inert gas atmosphere improved the storage life of MRM. Kraft et ai (1979) also demonstrated a greater survival of micro-organisms in products with increasing fat content.

The search for antioxidants suitable for inclusion in processed products containing MRM has continued. Hazelkamp et al (1980) and Ockerman et al. (1981) showed that the addition of sodium ascorbate or a mixture of ascorbyl palmitate and a-tocopherol significantly increased the storage life and acceptance scores of mechanically recovered pork products. Arneth (1981) observed similar antioxidant properties with the addition of citrate to sausage products containing MRM. Younathan, Marjan and Arshad (1980), demonstrated that a number of additives including onion juice and peel, egg-plant and sweet potato extracts all greatly decreased the rate of deterioration in MRM containing poultry products. Hydroxy anisóle, α-tocopherol and lecithin were effective in reducing lipid oxidation in fish protein (Takama and Mizushima, 1978; Avdeev et al, 1977).

It has also been suggested that supplementation of poultry feed with salts of a-tocopherol subsequently improves the keeping qualities of that mechanically recovered meat (Uebersax, Dawson and Uebersax, 1978).


Functional properies of MRM and their effect on products

Functional properties of mechanically recovered meat differ from those of hand-boned meat and when used in products, can affect their texture and possibly consumer acceptance.

Colour, texture and functional characteristics such as water holding, water binding and emulsifying capacity of mechanically recovered meat have been studied in both the raw state and in processed products. Both sieve and press-type mechanical deboners introduce a considerable amount of air into the recovered meat paste and as a consequence the myoglobin is fully oxygenated. While this may be an advantage if the meat is processed immediately, it can be the cause of rapid deterioration by haem-lipid oxidation and rapid oxidation of the pigment to metmyoglobin. Recent work on chilling and cryogenic freezing has shown it can also influence the colour of the MRM. Uebersax, Dawson and Uebersax (1978) obtained a paler colour in MRM after carbon dioxide freezing and MacNeil and Mast (1980) found a similar effect with liquid nitrogen freezing. Mast, Jurdi and MacNeil (1979), and Ockerman et al. (1981), showed that with mechanical freezing there was little initial change in colour but with increasing storage time, the colour darkened.

Bone marrow content, pH and water-holding capacity (WHC) of MRM are all interrelated. Increase in pH results in increased protein extractabil-ity (Anderson and Gillett, 1974). Recent experimental work at the Meat Research Institute (MRI) has shown a large increase in protein solubility of MRM compared with hand-boned meat. Unlike hand-boned meat though this solubility is not increased by subsequent freezing. The water-holding capacity of MRM is substantially higher than conventionally boned meat but a large depression of WHC occurs if MRM is frozen before further processing. Orr and Wogar (1979) showed similar effects for poultry meat suggesting that the composition of the material had a greater effect than the meat recovery machine used. Sebranek et al. (1979), demonstrated a lowering of WHC for beef used for patties when blast rather than cryogenically frozen.

Froning (1970) in his investigations with recovered poultry meat suggested that water and salt-soluble proteins could play a significant role in the improved emulsion capacity and stability of mechanically recovered meats and their products. Neelakatan and Froning (1971) subsequently showed that the elevated pH of mechanically recovered poultry meat (MRPM) substantially increased the stabilities of actinomyosin, myosin and myofibrils, although the stability of the sarcoplasmic proteins fell markedly.

MRM is already extensively used in processed food products (Table 7.3)

allspice was the least effective against the bacterial population. Hitokoto et al. (1980) showed allspice as well as aniseed and clove were capable of inhibiting fungal growth including yeasts.

Therefore there is still considerable scope for improving the overall storage stability of MRM and its products, whether by reducing bacterial growth, inhibiting lipid oxidation, or by effective storage conditions.

p. Β. Newman 107

Beef patties Canned corned beef Fresh pork sausage Fresh beef sausage Breakfast sausage Whole hog sausage Smoked pork sausage Frankfurters Bologna Fish sausage Fish cakes, sticks, etc. Stuffings Braunschweiger Liver sausage Luncheon meat Meat loaf Scrapple Bockwurst Chilli con carne Chow mein Chop suey Pizza Potted meat Devilled meat Ham, tongue spreads Fish patties Fish spreads Salami Meatballs

Sources: U S D A , 1976; Martin, 1976

but its lack of certain textural characteristics limits or even prevents its inclusion in some products. Structured protein fibre (SPF), a texturized soya protein product, has been used to improve textural properties. Lyon (1980) found that inclusion of 15% SPF significantly increased springiness, cohesiveness and chewiness of MRM and with the addition of salts and seasoning decreased cooking losses. Lyon, Lyon and Townsend (1978) also showed that if the MRM was flaked-cut there was a further improvement over conventional mixing of ingredients.

Despite the apparent limitations of MRM, it is generally agreed that inclusion of limited quantities (up to 20%) in processed products has httle effect on their properties and acceptance scores.

Nutritional and health implications

The composition of mechanically recovered meat differs appreciably from hand-boned meat, containing higher levels of fat, ash and bone (in the form of bone dust) and with less moisture and connective tissue. Protein content is determined by the source of the bone, e.g. pork, ham and picnic bones yield low protein, while beef neck bones yield higher protein (Goldstrand, 1975). MRM may also contain higher levels of other ingredients such as haem pigment, free iron and ascorbic acid. Despite this variability the nutritional content of MRM is more than satisfactory. In the USA though, there has been consumer disquiet over the allegedly poorer nutritional status of MRM.

The amino acid complement should be adequate, this has been shown to be the case for mechanically recovered fish by Meinke, Choy and Mattil (1975), poultry (Essary and Ritchey, 1968) and red meat (Field, 1976). The practical measurement of nutritional efficiency, the protein efficiency ratio (PER), has been extensively measured for MRM. Macneil, Mast and Leach (1979) found that the PER of mechanically recovered poultry meat treated with an antioxidant was significantly higher than the standard casein value of 2.5. Without antioxidant the PER was lower, possibly due

Table 7.3 M E A T P R O D U C T S I N W H I C H M E C H A N I C A L L Y R E C O V E R E D M E A T

H A S B E E N S U C C E S S F U L L Y I N C O R P O R A T E D


Legislation

The development of legislation on mechanically recovered meat in Europe has been summarized elsewhere (Newman, 1981). The peculiar situation that arose in the USA as a result of public consumer alarm as a consequence of inaccurate press reporting has been fully outlined by Field (1981).

In the UK, provisional regulations have defined MRM as fresh meat and have not required any declaration of its inclusion in products. However, consumer pressure and food standards officers are now demanding that MRM should be declared when included in meat products. MRM has not been included in any specific category regarding its handling or storage, although individual countries have specific regulations on the holding, storage and use of bones, much of the technology within the EEC is controlled by codes of practice rather than by statutory regulations. Consequently, within the EEC, mechanically recovered poultry and red

to protein alteration by rancidity, and varied widely. The source of the meat evidently has an effect; Babji, Froning and Satterlee (1980) found that turkey meat mechanically recovered from raw turkey frames had a PER of 2.59-2.75, while that from broiler backs, necks and fowl carcases had a much greater range of 2.34-2.94.

Attempts have been made to correlate the PER of mechanically recovered meat with certain amino acid components of the protein. Lee et al (1978) showed collagen content was highly correlated with essential amino acid content and consequently suggested that quantitative measurement of collagen can be used as a rapid estimation of protein quality. Field, Chang and Kruggel (1979) found similar results with mathematical projections using the content of leucine and proline, (the original concept postulated by Alsmeyer, Cunningham and Happich, 1974), or hydroxypro-line and isoleucine. Dvorak (1978) outhned a method for catagorizing meat products according to two nutritional characteristics, the energy (calorific content) and the composition of amino acids within the protein content of the product.

Two other components in MRM are important from the nutritional standpoint, bone and marrow. The former has already been discussed (Newman, 1981). The latter has been more difficult to quantify as no suitable methods have been available to measure it. Although the presence of marrow does introduce some protein, it also results in a substantial increase in the overall fat content and the higher the levels of MRM the lower the nutritional value of the material. The presence of marrow results in MRM having a higher pH than hand-boned meat. Field and Arasu (1981) proposed estimating bone marrow by pH measurement having shown good correlation. However, variation between factories due to different machines, different operations and the variability in the source of material necessitates the establishment of average values for individual factories. Marrow may also be estimated by the determination of haem pigment in the recovered meat, although again parameters must be established because of variability in the source material.

p. Β. Newman 109

Discussion and conclusions

Since the concept of mechanical recovery of meat was introduced in the late 1940s, the technology and understanding of the material have improved considerably but some basic problems remain.

The mechanism of recovery ensures that MRM is basically a structureless matrix of meat proteins and fat. It has lost many of the native physical and textural properties of conventionally boned meats, and other problems have been introduced, such as high fat content with increased incidence of

meats have been included in the provisional regulations of their respective fresh meat directives and are to be treated as fresh meat (Proposed Meat Directives, poultry and red meat, 1982).

The only bones excluded from processing by the EEC directives are head bones and pig tails. In the UK, prohibition is at present controlled by the Offals in Meat Products' Orders (1952) which precludes the use of brain, feet and spinal cord in mechanical meat recovery operations. The report of the Food Standards Committee on Meat Products (1980) does not suggest any change in these restrictions, although in the UK tails are an acceptable source for processing. Likewise, the present WHO proposals do not place MRM in a separate category nor do they define it as different from conventional boned meat (Alinorm 81/16, 11th Session of Codex, 1981).

In Europe, there are no restrictions at present on the types of machine for separating meat from bones. In the USA all machines used in the meat industry must be approved by the Animal and Plant Health Inspection Service (APHIS), a US Department of Agriculture regulatory body. Within Europe, the press-type separator is most widely used. In the UK, the Food Standards Committee have stated that while the press-type of machine is acceptable, the sieve type, which is used extensively in the USA, could create problems particularly as bones must be crushed before separation resulting in a rise in temperature.

Within the UK and Europe, there is little overall restriction on the use of MRM, although there is still some concern about the possible imphcations of calling MRM meat and its widespread use in meat products. The situation in the USA has been much more confused because of pressures from the very active consumer lobby. After two years of legal proceedings, and with considerable evidence from all interested parties, a Select Panel Report (1977) decided there was no danger in using MRM in food products. It made a number of recommendations, including maximum permitted level of fat and the maximum percentage of MRM in formulated products, labelling and protein quahty requirements. It also recommended that, until such information became available, MRM should not be used in baby and junior foods. Labelling was required to state levels of calcium and fluorine in the product for the safety of people suffering from calcium hypersensitivity and renal problems. The recent change in administration in the USA has led to a shift in attitudes concerning MRM. Although there is still concern in the consumer lobbies, the problem has received much less publicity and attempts are being made to reduce the restrictions on its use.


References

A C K R O Y D , H.B. (1979). MLC Seminar on Recovering Meat from Bones, p.21. MLC, Bletchley, UK

A L S M E Y E R , R .H. , C U N N I N G H A M , A . E . and HAPPICH, M.L. (1974). Food Technol, 28(7), 30

A N D E R S O N , J.R. and GILLETT, T. (1974). 7. Fd ScL, 39, 1143 A N O N . (1976). Ind. Rpton Evaluation of Protecon Deboner. Protecon UK

Ltd, Bedford, UK A R A S U , P. (1980). M.Sc. thesis. University of Wyoming, USA A R N E T H , w. (1981). Fleischerei, 32(6), 456, 463

rancidity and reduced storage life. If MRM could be recovered in a more native state with properties nearer those of hand-boned meat many of these problems would disappear.

Problems of microbial contamination can be overcome in two ways. Firstly, by maintaining the same temperature and hygiene conditions for the raw material, i.e. the bones, and the recovered meat as for fresh meat. Secondly, by ensuring that machines used for meat recovery are constructed in a sound manner, that metal contamination is Hmited to a minimum and that commercial practice allows for regular and effective cleaning.

Careful selection of the raw material and effective use of machinery can reduce two potential hazards. The elimination of the use of marrow bones (a common practice with UK MRM manufacturers), will increase the nutritive content of the recovered meat by reducing the level of added unsaturated marrow fat and haem pigment. Limiting the level of marrow fat will also improve keeping qualities. The quality of the product will be improved by reducing bone particle and fat content with machines which do not require extensive bone crushing before use and by reducing the operating pressure.

There is no doubt that with the increasing world shortage of animal protein, maximizing the yield from carcases and improvement in its use will be of paramount importance. In Europe in the mid-1970s, the existence of surplus meat in frozen storage and its availability at fairly low cost tended to reduce the rate of advance in MRM technology. Now the economics of mechanical meat recovery are very evident. The material is readily available at prices well below the cheapest meat cut. Knowledge of the material has enabled the food technologist to develop new and novel ways of using it in processed products, with little hindrance from legislation.

In the end, however, consumer acceptance will determine the extent to which the manufacturer can incorporate MRM in his products. Grittiness, lack of texture and poor colour are inherent problems. These, together with natural resistance towards new products, are identified as major causes for consumer rejection. Nevertheless, providing the trade ensures that the consumer is fully aware of what he or she is buying and is assured of its wholesomeness, there is little doubt that mechanically recovered meat can play an increasingly important role in satisfying the world need for good quality animal protein.

p. Β. Newman 111

AVDEEV, I.L, AKULIN, V.N., NAUMOV, V.V., ROMASHINA, N.A. and SYSKIN, G.A. (1977). Issledovaniya po Tekjnologii Rybnukh Produktov, 7, 3

BABJI, A.S., FRONING, G.W. and SATTERLEE, L.D. (1980). J. Fd Sci., 45, 441 BAIER, Α . , BAUM, W., HAACK, E. and ZOHRER, P. (1980). Fleisch, 34(6), 111 BARNETT, H.J., NELSON, R.W., HUNTER, P.J., BAUER, S. and GRONINGER, H.

(1971). Fisheries Bulletin, 69, 430 BAUM, W., ERNST, H. and ZOHRER, P. (1980). Fleisch, 34(4), 71 BIBUN CO. LTD (1980). Fish Processing UK Pat. No. 2 037 568 BIJKER, P.G.H., GERÄTS, G.E. and FRANSEN, Τ. (1978). Tijdschrift VOOr

Diergeneeskunde, 103, 583 BIJKER, P.G.H., GERÄTS, G.E., LOGTESTIJN, J.G. VAN, KOOLMEES, P.A. and

FRANSEN, Τ. (1979). Proc. 25th Europ. Meeting Meat Res. Workers, Utrecht, Netherlands

BIJKER, P.G., SCHÖLTEN, J.I., FRANSEN, Τ. and KOOLMES, P. (1980). Jijdschr. Diergeneeskd., 105, 428

BRAEGER, H. (1979). Fish meat recovery US Pat. 4 151 629 BRANEN, A.L. (1979). Proc. 31st Ann. Recip. Meat Conf., 156 CHANT, J.L., DAY, L., FIELD, R.A., KRUGGEL, W.G. and CHANG, Y.O. (1977).

J. FdScL, 42, 306 CODEX ALIMENTARIUS COMMISSION (1982). Draft Provisions on Code of

Practice for processed meat and poultry products. FAO/WHO, Rome CRAWFORD, D.L., LAW, D.K. and BABBITT, J.K. (1972). / . Fd Sci., 37(4), 553 DUBOIS, G., BEAUMIER, H. and CHARBONNEAU, R. (1979). / . Fd Sci., 44,

1649 DVORAK, z . (1978). Zpavodaj Masneho Prumyslu, 1, 46 EASTOE, J.E. (1961). In Biochemists Handbook, Ed. C. Long, p. 175 (Van

Nostrand-Reinhold, Princeton, New Jersey) ESSARY, E.O. and RITCHEY, S.J. (1968). Poultry Sci., 47, 1625 EUROPEAN ECONOMIC COMMUNITY (1982). Draft Provisions regarding the

production, storage and composition of mechanically recovered meat and poultry meat. EEC, Brussels

FARBOOD, M.I., MACNEIL, J.H. and OSTOVAR, K. (1976). / . Milk Food Technol, 39(10), 675

FIELD, R.A. (1975). Proc. 21st Europe. Meeting Meat Res. Workers, p. 100 FIELD, R.A. (1976). Food Technol, 30(9), 34 FIELD, R.A. (1981). In Advances in Food Research, Vol. 27, Ed. C O .

Chichester, Academic Press, New York FIELD, R.A. and ARASU, P. (1981). / . Fd Sci., 46(5), 1622 FIELD, R.A., CHANG, Y. and KRUGGEL, W.G. (1979). / . Fd Sci., 44(3), 690,

695 FIELD, R.A., RILEY, M.L., MELLO, F.C., CORBRIDGE, M.H. and KOTULA, A.W.

(1974). J. Anim. Sci, 39, 493 FIELD, R.A., KRUGGEZ, W.G. and RILEY, M.L. (1976). / . Anim. Scl, 43(4),

762 FIELD, R.A., SANCHEZ, L.R., KUNSMAN, J.E. and KRUGGEL, W. (1980). / . Fd

Scl, 45(5), 1109 FOOD AND NUTRITION BOARD (1980). Toward Healthful Diets. National

Academy of Science, Washington DC FRONING, G.W. (1970). PouU. Scl, 49, 6 FRONING, G.W. (1970). Poultry Scl, 49, 1625


F R O N I N G , G . w . (1981). In Advances in Food Research, Vol. 27. Ed. by C O . Chichester. Academic Press, New York

F R O N I N G , G . W . and J O H N S O N , F . (1973). / . Fd Sci., 38, 279 G E R M S , A . C . and S T E N N E N B E R G , H . (1978). Food Chem., 3(3), 213 G E R Ä T S , G . E . and T E R B I J H E , R . J . (1975). Archiv fur Lebensmittelhygene,

26(6), 227 G O L D S T R A N D , R . E . (1975). Proc. Recip. Meat Conf, 28, 104 G O R D O N , M . (1980). Meat Deboning US Pat. No. 4 217 679 G R Ü N D E N , L . P . , M A C N E I L , J . M . and D I M I C K , P . S . (1972). / . FdScL, 37(2), 249 H A Z E L K A M P , G . P . V A N D E M , H O U B E N , J . H . and K R O L , B . (1980). Tijdschrift

voor Diergeneeskunde, 105(21), 900 H E N I N G , W . (1974). Feinkost Wirtschaft, 11(2), 39 H E R R M A N N , Η . and N I T S C H E , W . (1981). Mechanised preparation of commi

nuted meat, GDR Patent 147 452 H I T O K O T O , H . , M O R O Z U M I , S . , W A U K E , T . , S A K A I , S . and K U R A T A , H . (1980).

Appl. Environ. Microbiol, 39(4), 818 H O L D I N G , M . (1979). Apparatus and process for separating meat scraps

from bones. UK Pat. No. 2 016 258 H O N I K E L , K . O . (1976). Fleischwirtsch., 56(5), 722 H O N I K E L , K . O . , S C H M I D T , U . , W O L T E R S D O R F , W . and L E I S T N E R , L . (1978). / .

Assoc. Official Anal Chem., 61(5), 1222 H U J N A K O V A , M . (1976). Zpravodaj Masneho Prumyslu, 3/4, 54 H U L T I N , H . O . , M C D O N A L D , R . E . and K E L L E H E R , S . D . (1979). Am. Chem. Soc.

Abstr., 178(1), AGFDl I G E N E , J . O . (1979). Role of meat lipids and meat pigments in development of

rancidity and warmed-over flavor in frozen and cooked meat. Dissert. Abs. International, B40(6), 215pp. Michigan State University, Michigan, USA

I G E N E , J . O . , P E A R S O N , A . M . , D U G A N , L . R . and P R I C E , J . F . (1980). Food Chem., 5(4), 263

I L Y U K H I N , v., K I S E L E V , I . I . and S H E B E L ' K O V , V . V . (1981). Myasn. Ind. SSSR, 6, 43

J A N K I E W I E Z , M . and P I S U L A , A . (1979). Zeszyty Naukowe Szkoly Glownes Gospodarstwa, 13, 67-77

J A N K Y , D . M . and F R O N I N G , G . W . (1975). Poultry Sci 54, 1378 J A N T A W A T , P . (1979). Composition and stability of lipids from mechanically

processed poultry meats. Dissert. Abst. Int., B40(4), 150pp Michigan State University, Michigan, USA

J A N T A W A T , P . and D A W S O N , L . E . (1980). Poultry Scl, 59(5), 1043 J E L E N , P . , E A R L E , M . and E D W A R D S O N , W . (1979). / . Fd Scl, 44(2), 327 J E L E N , P . , R I C H E R T , S . , E A R L E , M . , O L D F I E L D , S . , E D W A R D S O N , W . and

H A M I L T O N , R . G . (1978). Food Technol N.Z., 13(8), 3 J O B L I N G , A . J . (1978). In Recovering meat from bones, MLC symposium.

West Bromwich J U R D I , D . , M A S T , M . G . and M A C N E I L , J . H . (1980). J. Fd Scl, 45(3), 641 K E L L E R , S . and B L O C K , R . J . (1960). In Laboratory Manual of Analytical

Method of Protein Chemistry, I, Academic Press, London K N U C K L E S , B . E . , B I C K O F F , E . M . and K O H L E R , G . O . (1972). Ag. Fd Chem.,

20, 1055 K R A F T , A . A . , R E D D Y , K . V . , S E B R A N E K , J . G . , R U S T , R . E . and H O T C H K I S S , D . K .

(1979). / . FdScl, 44(2), 350

p. Β. Newman 113

K U N S M A N , J .E . , FIELD, R . A . and K A Z A N T Z I S , D . (1978). / . Fd Sci., 43(5), 1375

L E A C H , A . A . (1958). Biochemistry / . , 69, 38 LEE, Y . B . , ELLIOTT, J.G., R I C K A N S R U D , D . A . and H A G B E R G , E.C. (1978). / .

FdSci., 43(5), 1359 LIMNOV, G .E . , SNITSAR, A.I . and M A M E D O V , I.I. (1978). In ProC. 24st Europ.

Meeting Meat Res. Workers, H11:1-H11:5 L Y O N , C.E. (1980). Poultry Sci., 59(5), 1031 L Y O N , B .G. , L Y O N , C.E. and T O W N S E N D , W.E. (1978). / . Fd Sci., 43(6), 1656 M A C H I N E F A B R I E K A M E R S F O O R T B V (1980). A press for separating meat

from bones. British Pat. No. 1 563 750 M A H O N E Y , A . W . , TSO, T .B. , M c L A U G H L I N , K. and H E N D R I C K S , D . (1980). Fed.

Prog., 39(3,11), 1042 M A C N E I L , J .H. , DIMICK, P.S. and MAST, M.G. (1973). / . Fd Sci., 38(8), 1077 M A C N E I L , J.H. and MAST, M.G. (1980). / . Fd Sci., 45(3), 645 M A C N E I L , J .H. , MAST, M.G. and L E A C H , R.M. (1979). / . Fd Sci., 44(5), 1291 M A R T I N , R .E . (1976). 2nd Tech. Seminar on Mechanical Recovery and

Utilization of Fish Flesh. Nat. Fish. Inst., Cambridge, Mass. M A S T , M.G. , J U R D I , D . and MACNEIL, J.H. (1979). / . Fd Sci., 44(2), 346 M A W S O N , R.F. and COLLINSON, B.R. (1974). Studies on the mechanical

recovery of meat. MIRINZ pubhcation 391 M E I N K E , w.w., C H O Y , A . A . and MATTIL, K.F. (1975). In 35th Ann. Meeting

Inst. Food Technologists, USA MINISTRY OF A G R I C U L T U R E , FISHERIES A N D F O O D (1952). Offals in Meat

Products Order. HMSO, London MINISTRY OF A G R I C U L T U R E , FISHERIES A N D F O O D (1980). Food Standards

Committee Report on Meat Products. HMSO, London M O E R C K , K.E. and B A L L , H.R. Jnr (1973). / . Fd Sci., 38, 978 M U R P H Y , E.w. and E N G E L , R .E . (1979). In Nutrition in transition, 413,

USDA, Washington DC N E E L A K A N T A N , S. and F R O N I N G , G.W. (1971). / . Fd Sci., 36(4), 612 N E W M A N , P.B. (1981). Meat Sci., 5(3), 171 N O R D I S C H E R M A S C H I N E N B A N (1979). US Pat. No. 4 151 629 O C K E R M A N , H.W. , H O U B E N , J.H., KROL, B . , PLIMPTON, R.F. and S C H A D , M.

(1981). / . FdScL, 46(1), 220 O R R , H.L. and W O G A R , W.G. (1979). Poultry Sci., 58(3), 577 P R O T E C O N BV (1980). Meat-bone separation. UK Pat. No. 1 559 931 PROTEIN A D V I S O R Y G R O U P (1970). Single cell protein. Guidehne No. 4.

FAO/WHO/UNICEF, Rome R A C C A C H , M. and B A K E R , R.C. (1979). Poultry Sci., 58(1), 144 R A M S E Y , J .D. and ROSSI, G. D E L . (1976). Detection of bone in meat. US Pat.

No. 3 995 164 R O T H , E.N. (1980). Deboning of meat. US Pat. No. 4 186 216 ST. CLAIR FISHERIES (1980). Fish meat recovery. UK Pat. No. 1 563 662 S A T E R L E E , R . D . , F R O N I N G , G.W. and J A N K Y , D .M. (1971). J. Fd Sci., 36(7),

978 SCHNELL, P.G. , N A T H , K.R., D A R F L E R , J.M., V A D E H R A , D .V . (1973). Poultry

Sci., 52(4), 1361 S C H R O E D E R , H . A . and TIPTON, LH. (1968). Arch. Environ. Hlth, 17, 959 S E B R A N E K , J .G. , S A N G , P.N. , TOPEL, D . G . and RUST, R .E . (1979). J. Fd Sci.,

48(5), 1101


SHELEF, L .A . , NAGLIK, O.A. and B O G E N , D.W. (1980). / . Fd Sci., 45(4), 1042 S H U V A E V , I .A. , A F A N A S ' E V , T. and Y U , I (1978). Method of separating

residual meat from broken bone. USSR Pat. 587 918 S O L B E R G , M. (1968). Proc. Meat Ind. Res. Conf. AMIF, Arlington,

Vancouver S O W D E N , E.M. and STITCH, S.R. (1957). Biochem. / . , 67, 100 SPILLMAN, P. (1980). The recovery of meat from small crustaceans, especial

ly Antartic Krill. German Democratic Republic Pat. No. 140 698 SWINGLER, G.R. and L A W R I E , R .A . (1977). Meat Sci., 1, 159 T A K A M A , K. and MIZUSHIMA, Y. (1978). Int. Congress Food Sci. Technol,

Kyoto, Japan 200 TOPOL'NIK, V .G. and R A T U S H N Y I , A .S . (1980). Izvestiya Vysshikh Ucheb-

nykh Zavendenii Pishchevaya Tekhnologiya, 6, 55 U E B E R S A X , K.L. , D A W S O N , L.E. and U E B E R S A X , M.A. (1978). Poultry Sci.,

57(3), 670 U S D E P A R T M E N T O F A G R I C U L T U R E (1976). Fed. Reg., 17535 and 17560 US D E P A R T M E N T O F A G R I C U L T U R E (1978). Fed. Reg., 43, 26416 WILLIAMS, R.R. , WEHR. Η.Μ. , STROUP, J.R., PARK, Μ. and P O I N D E X T E R , B . E .

(1980). I. FdSci., 45(4), 757 W O L T E R S D O R F , W. and SCHMIDT, U. (1976). In Rückstände in Fleisch und

Fleischerzeugnissen, pp. 60-72. Boppard, West Germany Y O U N A T H A N , M.T. , M A R J A N , Z.M. and A R S H A D , F.B. (1980). J. Fd Sci.,

45(2), 274 Y O U N G , R.H. and L A W R I E , R .A . (1975). J. Fd Technol., 10, 459

8

Introduction

Wet fodder preserved and stored in a silo is called a silage. This is the traditional way of preserving fresh grass for animal feed. Under the anaerobic conditions which soon develop in such a silage, bacteria naturally present in the fodder crop will convert sugars into lactic acid, which causes the pH to drop sufficiently to prevent spoilage. This preservation process, acidification by natural fermentation of sugars present in the raw material, is not always rehable. The process may 'run wild' and result in considerable deterioration, in particular if the raw material is rich in protein and wet with rain when ensiled. To ensure proper preservation of such crops it is nowadays most common to add acid, a practice which was initiated by Virtanen in the 1920s. He suggested the use of mineral acids in sufficient quantities to lower the pH of the silage to about 4. Formic acid has almost completely replaced the mineral acid method of Virtanen, mainly because the organic acid has a preservation action in addition to that due to low pH. Less organic acid was therefore needed and formic acid was considered less hazardous to handle than the mineral acids.

Rapid acidification by acid addition causes an instantaneous stop in the respiratory processes in the plants, preventing loss of organic carbon from the raw material, notably sugars. Acid tolerant bacteria surviving in the acidified grass silage will slowly convert sugars into lactic acid and thus keep the pH low. Residual sugars in the grass will also contribute to preservation by repressing the production of deaminating enzymes in the bacteria and consequently prevent ammonia formation from amino acids.

The microbial processes in ensiled grass are complex. Initially the grass is loaded with mainly obhgate aerobic bacteria, facultative anaerobes and yeasts. The aerobic species are Sarcina, Micrococcus, Flavobacterium, Alcaligenes and Pseudomonas (Kroulik, Burkey and Wiseman, 1955; Gibson et al, 1958; Pedersen, 1976), while the facultative anaerobic organisms belong to the coli-aerogenes group (Kroulik, Burkey and Wiseman, 1955; Pedersen, 1976; StirHng and Whittenbury, 1963). Among this variety of micro-organisms, the lactic acid bacteria occur only in very low numbers, 0-100/g fresh weight (Kroulik, Burkey and Wiseman, 1955; Keddie, 1951; Nilsson and Nilsson, 1956). Although Clostridia are not part of the flora normally found on grass they are introduced into the silage by contamination from soil.

117

SILAGE PRODUCTION—THEORY AND PRACTICE

J . R A A Institute of Fisheries, University of Troms0, Norway and A. GILDBERG and T. STR0M Institute of Fishery Technology Research (FTFI), Troms0, Norway

118 Silage production - theory and practice

Ensiling of fish by lactic acid bacterial fermentation

WHY FISH SPOIL SO QUICKLY

Fish spoil more quickly than carcases of warm blooded animals for a number of reasons. First, fish tissues become less acid post mortem in contrast to mammalian tissues. Secondly trimethylamine oxide, a specific component found in marine organisms, stimulates anaerobic growth of spoilage bacteria. Trimethylamine oxide is reduced by the spoilage bacteria to trimethylamine, which is responsible for the ammoniacal off-flavour of spoiled fish. Since both fish and warm blooded animals contain little free sugar, which is the preferred source of energy for growth of most micro-organisms involved in spoilage, degradation of amino acids and formation of ammonia start soon after microbial attack post mortem. In the

In ensiled grass with no added acid, the obligate aerobic flora will vanish quickly, yeasts after a few days and the facultative anaerobes more slowly (Pedersen, 1976). There is a simultaneous rapid increase in the number of lactic acid bacteria, which after a week become the predominant bacterial group, being present in numbers of about lO'̂ Vg wet silage (Pedersen, 1976). Streptococcus sp. dominates among the lactic acid bacteria during early phases of natural fermentation, later Lactobacillus sp. takes over (Kroulik, Burkey and Wiseman, 1955; Gibson etai, 1958; Pedersen, 1976; Stirling and Whittenbury, 1963; Nilsson and Nilsson, 1956; Allen and Harrison, 1936; Allen etal., 1937). In a mature grass silage Pediococcus sp. is a characteristic inhabitant (Pedersen, 1976). Lactic acid bacteria develop in the same succession in a grass silage preserved by adding formic acid, but the time span is longer. In the acid preserved silage however, the facultative anaerobes steadily increase in number to sometimes surpass the lactic acid bacteria (Pedersen, 1976), and the obligate aerobes survive for a longer time, probably because oxygen will remain longer in a silage where the respiratory processes in the grass are arrested by acid.

Lack of success in preserving forage grass by natural fermentation may be due to low levels of fermentable sugars in the raw material and high buffering capacity. In such cases the addition of sugars may stimulate the lactic acid bacterial fermentation and produce proper silages. Molasses (McDonald and Purves, 1956; Anderson and Jackson, 1970), potato and sugar beet (Barnett, 1954) have been used as sugar sources with acceptable results. Since the plants initially have a low number of lactic acid bacteria, it is logical to combine sugar addition with starter cultures of such bacteria (McDonald and Purves, 1956). Alternatively, the effect of adding cellu-lase, amylase and starter cultures should be more systematically studied. From a theoretical point of view it seems most feasible, however, to combine sugar addition and chemical acidulation in the initial stage of grass ensilage, and ensure lactic acid fermentation by adding a starter culture together with agents which favour growth of these bacteria. This is the strategy followed when applying natural fermentation for preserving flsh and other raw materials which have a high buffering capacity, are low in carbohydrate and rich in protein.

J. Raa, Α. Gildberg and Γ. Str0m 119

LACTIC A C I D B A C T E R I A IN FISH

Lactic acid bacteria are natural inhabitants of fish (Schr0der et al, 1980; Kn0chel, 1981), but they are present in low numbers, lO^-lO'̂ /g (Kn0chel, 1981). The fish contains, however, only small amounts of free sugar, which is the essential substrate for growth of such bacteria. They will therefore soon vanish in a dead fish and the typical spoilage microflora degrading amino acids will take over. If a fermentable sugar is added, however, the lactic acid bacteria become predominant and cause the pH to fall below 4 and thus preserve the fish.

ENSILING

Due to the high buffering capacity of a flsh (protein, minerals), large quantities of acid are required to produce a pH below 4 and consequently considerable amounts of fermentable sugar must be added to obtain a stable silage with a pH around 4. For example 20 kg of a dry mixture of malt and oatmeal were required for 100 kg of fresh herring (Nilson and Rydin, 1963) or more than 10% molasses (Roa, 1965; Kompiang, Yushadi and Creswell, 1980). Malt was essential for the preservation with the oatmeal additive, because the amylolytic enzymes present in the malt convert the starch to glucose which can be fermented by the lactic acid bacteria. These bacteria are unable to produce such enzymes themselves. In such a silage the free glucose will repress the production of deaminating enzymes by spoilage microbes in the raw material and in this manner suppress ammonia production and prevent pH increase. Glucose will simultaneously be continuously converted to acid, which lowers the pH in the silage and finally renders it resistant to spoilage. Both spoilage bacteria and lactic acid bacteria will contribute to the initial acid production, because the conditions are anaerobic and sugars are available, but growth of the lactic acid bacteria will be favoured as the silage becomes more acid. If so much sugar is added that pH falls to below 4, lactobacilli will become the predominant organism present. Coliforms, enterococci, typhoid bacteria and even spores of Clostridium botulinum are destroyed in such a silage (James and Nair, 1977; Durairaj et al, 1976; Wirahadikusumah, 1968). The low pH is probably the primary reason why they are killed, but other factors may also be involved, for example enzymes in the fish raw material and antibiotic substances produced by lactic acid bacteria (Schr0der et al, 1980; Lindgren and Clevstr0m, 1978a; Lindgren and Clevstr0m, 1978b).

presence of a fermentable sugar, however, amino acid degradation is suppressed in most spoilage bacteria. To preserve fish or animal waste products by fermentation it is therefore essential to add a sugar source, preferably together with a starter culture of proper lactic acid bacteria, which by rapid conversion of the sugar to acid preserves the whole mass.


N U T R I T I O N A L V A L U E

Fermented fish silage maintains a good nutritional value during long-term storage (Nilson and Rydin, 1963; Kompiang, Yushadi and Creswell, 1980), and rancidity is prevented even during drying (Wirahadikusumah, 1968). Lactic acid may be the active antioxidant, but unpublished observations by the authors suggest that other antioxidants generated in the silage are probably also involved. The fermentation method is suitable for preserving waste fish products and since it does not require expensive equipment or special chemicals, it can easily be adapted to the conditions in small fishing villages where periodic surpluses of waste could be stored as a local animal feed. There is a particular demand for such simple technologies in many tropical countries where fish landings are scattered and the losses are high due to rapid spoilage under non-chilled conditions. It has been demonstrated that tropical waste fish can be preserved by lactic acid bacterial fermentation and stored as a silage for at least a month at 30 °C without loss of nutritional value (Raa, 1981; Raa and Gildberg, 1982). In fact fermented fish silage has been shown to have a significantly better nutritional value for chickens than fish silage preserved by acid (Kompiang, Arifudin and Raa, 1980). It has also been shown that uptake and incorporation of calcium into the eggshell is facilitated by fermented fish silage (Wirahadikusumah, 1969). Despite the good nutritional value, ammonia formation in fermented silage is considerably higher than in a corresponding acid preserved silage (Raa and Gildberg, 1982). This shows that repression of deamination by glucose is not absolute, but essential amino acids are hardly degraded since the nutritional value was unaffected. Since this ammonia formation occurs early during the fermentation process and before the pH has dropped much, it may be avoided by combined acidulation and sugar addition. By this same strategy another problem or risk-factor associated with fish silage fermentation may also be avoided, namely gas production due to hetero-fermentative bacteria which are active at a pH above 5.5. Ensuring lactic acid bacterial fermentation by adding, after acidulation, proper starter cultures and selective agents like sodium nitrite is also to be recommended (Raa, 1981).

FISH F E E D

At the moment there is an increasing interest to ensile fish waste products and low value fish for use as a feed for sea farmed fish; in Norway this is mainly salmon and trout. Fish silage preserved by adding acid (formic acid, hydrochloric acid, sulphuric acid or combinations of these) is acceptable to the salmonids and the nutritional value is as good as that of the fresh raw material (Anstreng, 1982). It is a significant problem however, that the fish silage becomes so liquefied that large volumes of dry binder meals are needed to produce a moist pellet of acceptable consistency; 60-70% of the dry matter in such a moist feed may derive from the added dry meal. A fermented fish silage is better in this respect, because the solubilization may be reduced considerably. Currently a process which combines low preservation costs and a reduced need for binder meal to produce an acceptable moist pellet for fish is being developed at our Institutes.

J. Raa, Α. Gildberg and Τ. Str0m 121

Acid silage—preservation by adding acid

Addition of acid is an alternative to fermentation for preserving waste fish. The main advantage is that the volume of additives can be reduced compared with the fermentation method and the success rate can be well predicted. Acid preserved fish silage has been produced commercially since about 1950 in Denmark, where the annual production over the last years has been about 60000 tonnes. There is also a silage industry in Poland. In Norway silage is produced commercially from fish guts and other waste products from the fishing industry, and silage plants with equipment for de-oihng are in operation.

P R E S E R V A T I V E ACTION OF ACIDS

If inorganic acids are used, fish cannot be fully preserved as a silage, unless the pH is lowered to 2, or below (Edin, 1940). If organic acids are used, however, the silage is stable at pH 3.5-4.0 with formic acid (Tatterson, 1976) and pH 4.5 with propionic (Gildberg and Raa, 1977). Such moderately acid fish silages can be incorporated in animal rations and fed without neutralization.

The antimicrobial activity of weak organic acids is associated with the undissociated molecule. The non-charged molecules can pass into the cell, whereas cell membranes are in general impermeable to the anion of the weak acid, and to protons. Inside the cell, where pH is close to neutral, the weak acid will immediately dissociate and both the anion and the proton become trapped in the cytoplasm. The internal pH of the microbial cell therefore gradually falls, and the anion accumulates. Both these factors contribute to the antimicrobial effect of weak organic acids. The anion may, however, be metabolized by the cell, which in this way counteracts its toxicity. This probably happens with acetate, which has a considerably lower antimicrobial activity than propionate and formate. This theory of the mode of action of weak organic acids, implies that their antimicrobial activity should increase with falling pH, according to their acid/base titration curves. This has in fact been demonstrated, using the carcinogenic fungus Aspergillus flavus as the test organism (Str0m et al., 1979). Propionic acid, which has a pKa of 4.9, inhibited this fungus at pH values below 5, whereas formic acid, having a pKa of 3.7, inhibited below pH 4. In a medium lacking organic acids, the same fungus grew fast even at pH 2.

Production of fermented fish silage is a simple and recommendable practice for producing moist feed from local waste for domestic animals in the vicinity. But if lactic acid bacterial fermentation is to be adopted for preserving industrial fish for meal and oil production, or whole fish for human consumption, the methods have to be refined considerably, as discussed in a recent paper (Raa, 1981). For certain raw materials with variable lipid content, such as fish guts with variable liver content, it may be more practical to use the chemical ensihng method described in the following section.


O R G A N I C OR M I N E R A L ACID?

Before mineral acid silage can be fed to animals it must be neutralized, for example by addition of 2-5 kg of chalk/100 kg of silage (Petersen, 1953). However, mineral acids alone are no longer widely used as ensiling acids because the handling of such acids is very hazardous, the neutralization of the acid is a laborious procedure with practical problems, the high salt level which results from neutralization is undesirable in nutrition and because corrosion problems are more severe at the low pH of a stable mineral acid silage. Although organic acids are more expensive than mineral acids, the difference in price is partly counterbalanced by the higher efficacy of the organic acids, and it is not necessary to neutralize a silage produced with organic acids. To reduce the price of preservation of fish with high buffering capacity, a mixture of inorganic and organic acids can be used; the cheap inorganic acid will lower the pH so that the organic acid becomes antimicrobial. Such experiments have been carried out mainly with formic acid in mixture with sulphuric (Olsson, 1942; Disney, Tatterson and Olley, 1977). Approximately 3% of a 3:1 (v/v) mixture of sulphuric and formic acid preserved fish offal equally well as 4% (v/w) of pure sulphuric acid. Sulphuric acid can be replaced by hydrochloric acid (Disney and Hoffman, 1976; Disney, Tatterson and Olley, 1977) or phosphoric acid (Jensen and Schmidtsdorff, 1977).

Because of the higher ash content of tropical by-catch fish, more acid is needed for preservation; 2.5% formic acid, or a mixture of formic and propionic acid being the minimum concentration which ensures preservation of fresh tropical by-catch fish (Kompiang, Arifudin and Raa, 1980). For practical purposes, when it is necessary to have a high safety margin, it is advisable to use at least 3% of such acids. This represents a significant cost, and it is accordingly important for the viability of the silage process to find the most economic combination of acids for such raw materials.

MINCING OR CHOPPING?

To produce a fish silage it is common to mince the fish and mix the mince with the acid. This may be a significant practical problem if the fish is fresh because the muscle components become rubber-like when exposed to acid. The mince therefore tends to form closed pockets where the acids do not enter quickly enough to prevent spoilage. If the raw material is not fresh then this is not a major problem. However, the solution is not to let the fish

It is moreover in accordance with the theory suggested here that a stable silage can be obtained at a higher pH with propionic (Gildberg and Raa, 1977) than with formic acid (Tatterson, 1976). The quantity of inorganic acid required to lower the pH to 2 in a fish homogenate depends on the concentration of protein and ash (minerals) in the raw material, but varies within the limits 9 and 4 € of 14 Ν inorganic acid for the most bony fish low in oil and the most oily fish respectively. This corresponds to 6.3 kg (3.4 €) and 2.8 kg (1.5 €) of concentrated sulphuric acid/100 kg fish.


LIQUEFACTION (AUTOLYSIS)

A silage gradually liquefies due to the activity of tissue degrading enzymes which are naturally present in the fish, mainly in the digestive organs (Tatterson and Windsor, 1974; Backhoff, 1976; Gildberg and Raa, 1979; Raa and Gildberg, 1976; Hjelmeland, 1978; Hjelmeland and Raa, 1980; Hjelmeland and Raa, 1982). This self-digestion, or autolysis, is an advantage if oil is to be removed from the silage. Oil removal usually improves the feed quality of fish silage (Raa and Gildberg, 1982).

The rate of autolysis is determined by the activity of hydrolytic enzymes in the raw material, the physiological condition of the fish at the time it was caught, the pH, the temperature and the nature of the preservative acids. Since protein is the major structural component in fish tissues, it is mainly the proteases which are responsible for autolysis. Most active at acid conditions are pepsin-type stomach proteases, but certain lysosomal enzymes Hke cathepsin D, which is the major muscle protease, may also be of some importance. Collagen degrading enzymes are probably present in fish tissue (Yoshinaka, Sato and Ikeda, 1978; Dugal and Raa, 1978), but the importance of these enzymes in autolysis of fish silage is doubtful. The denaturation temperature of collagen decreases with increasing acidity (Hayashi and Nagai, 1973), and fish collagen is certainly denatured and rendered susceptible to degradation by ordinary proteases under normal ensiling conditions.

About 80% of the protein in fish silage usually becomes solubilized after one week at temperatures around 23-30 °C (Tatterson and Windsor, 1974; Backhoff, 1976). However, the yields vary significantly if different fractions of the fish are ensiled separately (Backhoff, 1976). Usually viscera gives the highest yield, muscle the lowest. Protein solubilization may be reduced by lipid in the fish (Sheikh and Shah, 1974; Raa and Gildberg, 1976).

Even after prolonged autolysis there is always an undigested fraction which is resistant to further enzymatic degradation. Such a fraction also remains when fish tissues are solubilized by commercial enzymes (Freeman

spoil before the acids are added, but rather to cut it or chop it into pieces before the acids are added. This will ensure good preservation because the acid will sterilize the fish surfaces and the guts which carry the spoilage microbes. In practice it may be advisable, however, to mix the newly chopped, or minced, fish with the liquid silage from the container/storage tank to make the mixture pumpable. This serves also to obtain an efficient mixing of fish and acid. Attempts have been made to ensile waste meat from Arctic seals. Being a warm blooded animal, its tissues contract much more upon exposure to acid than is the case with fresh fish. In fact the acidified mince became so rubbery that it could not be pumped at all. The problem can be circumvented nevertheless by chopping and mixing with already liquefied silage before more acid is added. Such problems are not encountered with fish tissues with low muscle content, such as guts, and with small fish which are mixed whole with the preservative acids (Kompiang, Arifudin and Raa, 1980).


STABILITY OF A M I N O ACIDS

Amino acids are rather stable in fish silage. Only 1.3% of the amino nitrogen was released as ammonia in a silage of by-catch fish after three weeks at tropical temperatures (Kompiang, Arifudin and Raa, 1980). Nevertheless, this might imply a significant reduction of the nutritive value of the silage, if the ammonia derived from essential amino acids. Free tryptophan decomposes in an acid silage (Backhoff, 1976; Kompiang, Arifudin and Raa, 1980) and there are reports claiming that methionine (Atkinson, Lamprecht and Misplon, 1974) and histidine (Disney et aL, 1978) are also unstable.

After storage for 40 days at 30 °C about 30% of the tryptophan was lost in silages of cod and herring preserved with formic acid at pH slightly below 4 (Backhoff, 1976). At lower storage temperatures the rate of degradation was significantly lower. Whether enzymes are involved in the degradation of tryptophan is not known, but there are observations in favour of this suggestion; for example, tryptophan was shown to be stable in the heat-treated aqueous phase of a silage of cod or saithe viscera (Str0m and Eggum, 1981).

It has been reported that methionine was the growth limiting amino acid in fish silage (Jensen and Schmidtsdorff, 1977), but this was probably due to a low level of this amino acid in the raw material. There are no reports to disagree with the observation that methionine is stable in an acid fish silage (Wignall and Tatterson, 1976; Str0m and Eggum, 1981).

Histidine may be the limiting amino acid in fish silage (Disney et al., 1978), particularly if the silage is prepared from partly spoiled fish, since this amino acid is quickly degraded by spoilage bacteria at neutral pH.

T H I A M I N A S E

Vitamins are usually added when an animal diet is prepared. Since most fish contain an enzyme which degrades thiamine (vitamin B^) there is a certain risk that feeding silage rations may cause a deficiency in this vitamin, and poor growth on silage diets has been attributed to vitamin Bj deficiency (Disney et al., 1978; Disney and Hoffman, 1976). Heating the silage is a means of avoiding this risk, a treatment which also safeguards against the spreading of possible acid resistant virus and makes oil removal easier.

N U T R I T I O N A L V A L U E

The nutritional value of fish silage incorporated in animal rations has been shown to be good (Hillyer et al., 1976; Anstreng, 1982; Cameron, 1962; Disney, Parr and Morgan, 1978), or only slightly inferior to fish meal

and Hoogland, 1956; Hale, 1969; Tarky, Agarwala and Pigot, 1973; Möhr, 1980).

y. Raa, Λ. Gildberg and Τ. Str0m 125

ENSILING OF FISH VISCERA

Fish viscera is a waste problem where whole fish is landed for filleting. Earlier in Norway it was dumped in close proximity to the factories, but still some 20000-25000 tonnes, or about 50% of the total amount of viscera from cod, saithe and haddock are being wasted An industrial silage process for producing animal feed and oil is now gradually being introduced to make full use of this resource.

The fish viscera offal which comes out from a fish filleting line has very variable chemical composition, depending on whether liver is removed (cod) or not (saithe, haddock). Even small variations in the amount of liver in the offal will affect the chemical composition greatly, because liver has a high dry weight (60-70%) compared with the other digestive organs (15-20%) and a fat content of 70% of dry weight. Offal with one-third its weight of liver will, for example, contain 55% fat and 40% protein, on a dry weight basis. The guts without liver will by comparison contain 80% protein and 10% fat (dry weights). Such variations in chemical composition could not be accepted in a standardized feed product, which seemed the most realistic outlet for ensiled viscera. The strategy for utilizing fish guts became, therefore, to let the tissues autolyse so that floating lipid could be removed from a liquid phase of dissolved protein by centrifuging.

I N D U S T R I A L PROCESSING OF FISH VISCERA

The industrial process includes:

(1) ensiling, (2) autolysis and lipid separation, (3) concentration of the aqueous phase by evaporation, and (4) preparation of feed.

A mixture of propionic and formic acid has been used as the ensiling acid at a concentration of 1.5-2%, which results in a pH of about 4.4 after 24 h. A 1:1 mixture of the two acids has been used, but sufficient protection against moulds may be obtained with less propionic acid, a modification which improves the economy of the process. The acids are added automatically to the fish viscera before mincing and controlled by the pumped

(Fonge, 1976; Whittemore and Taylor, 1976). In rations for poultry, however, a negative effect of feeding a high proportion of silage has been demonstrated (Disney et al., 1978; Disney and Hoffman, 1976), due to the lipids (Kompiang, Arifudin and Raa, 1980). Nevertheless, fish silage is a realistic alternative to fish meal in utilizing fish waste, surplus fish and low value fish, particularly in tropical countries (Disney, 1979). Salmon grows well on silage diets, but does not tolerate propionic acid among the preservative acids, and differs in this respect from trout (Austreng, 1982). Freshwater carp also accept and grow well on silage based feeds (Djajase-waka and Djajadiredja, 1980).


S t o r a g e t a n k s for S I l ä g e

A u t o l y s i s t a n k s w i t h s t i r r i ng ν

Steann heat ing

Decanter

λ Λ Λ

S t e a m heat ing

[Heat exchänger |

S to rage tanks for protein concent ra te and oi I

Sepa ra to r

Figure 8.1 Flow chart of fish gut silage processing plant

volume of viscera. The preserved mixture is pumped into silage storage tanks. This ensiling unit can be installed on board vessels or in each factory. For further processing this silage is transported to a larger processing plant (capacity 1000-2000 tonnes per annum). This plant is equipped with storage tanks for the silage, autolysis tanks (10 m^), decanter centrifuge to remove sludge, lipid separator to remove the oil, and storage tanks for the oil and the deoiled silage. Autolysis occurs for three days at 30-35 °C before the autolysate is heated to 95 °C to facilitate oil separation from sludge and dissolved matter. This process yields an aqueous phase, containing dissolved protein, peptides and amino acids, with 0.1-0.3% lipid (wet weight). The sludge phase may constitute from 0.5 to 10% of the silage volume, depending on the quantity of bone residues in the raw material. The oil is sold as such to the oil refining industry. The aqueous phase is concentrated by evaporation to a dry weight of about 45%. The pH will increase to 4.9-5.0 after evaporation, but the concentrate is still resistant to spoilage and can be stored for a long time in tanks. The energy spent on heating to 95 °C is collected by heat exchangers and used for water evaporation during concentration, so that almost no extra energy has to be used for removing water. The flow chart of the commercial plant for processing ensiled gut is shown in Figure 8.1. The same technology can be used to process entrails from slaughterhouses.

There are two limitations to the use ot this amino acid/protein concentrate at high inclusion levels in rations for ruminants and chicken, respectively. Firstly, due to the high content of free amino acids, rumen microbes cause significant deamination, resulting in slightly lower growth and milk yield in cows fed on a silage diet than on a diet with an equal

/. Raa, Λ. Gildberg and Τ. Str0m 127

ECONOMICS

The process is close to economic balance (Str0m et al, 1979), but it is dependent on several factors such as access to raw materials and continuity in processing. A calculation based on full capacity production of 4000 tonnes of silage (raw material) per annum and continual operation of the plant is shown in Table 8.1. This indicates that a protein price of 5.85 N.kr*/kg 100% protein (~ 0.50 £/kg) must be obtained in order to run the plant in economic balance. Such a protein price is reaHstic at the present time. Thus a pollution problem, with all its indeterminate costs, can be solved on a commercial basis. Possible future developments of the process might be to combine it with the production of a few high priced components, such as cholesterol derivatives from the oil, and enzymes from the

Table 8.1 P R O D U C T I O N ECONOMICS OF OIL A N D D E O I L E D SILAGE FROM FISH VISCE RA (AFTER STORMO, INSTITUTE OF FISHERY T E C H N O L O G Y R E S E A R C H , T R O M S 0 , N O R W A Y )

Raw material (guts) per annum Protein produced: 450000 kg Oil produced: 400000 kg Investments Facory (buildings) and office including taxes Equipment/mounting/starting up

Interest and depreciation—buildings Interest and depreciation—equipment Costs (1) Capital costs

Buildings L2 million N.kr 0.117 = 14000 Equipment 4.8 million N.kr 0.163 = 782000

(2) Maintenance (2%) (3) Salaries and social expenses (three persons) (4) Energy (1 million kW @ 0.15 N.kr/kWh (5) Insurance (6) Various (7) Transport of raw material (0.10 N.kr/kg) (8) Transport of product (0.20 N.kr/kg) (9) Purchase of preserved raw material (300 N.kr/tonne)

Income Oil (2 N.kr/kg) Protein (X N.kr/kg x 450000 kg)

Balance is achieved if X = 5.85 N.kr/kg pure protein

4 million kg

N.kr* 7200000 4888000

12088000

10%/20 years 10%/10 years

N.kr

922000 96000

360000 150000 25000

100000 400000 180000

1200000 Total

Total

3433000

800000 2633000 3433000

*11 N.kr = £1.00 at July 1982 rates.

quantity of protein. Secondly, due to the low level of sulphur amino acids in the aqueous phase, a high level of silage in the feed may result in a sulphur deficiency for chickens in the moulting phase.


Ensiling of shrimp waste

T H E W A S T E PROBLEM

About 15000 tonnes of waste (shells/heads) from Norwegian shrimp processing factories are dumped every year. This waste contains about 1500 tonnes of protein and approximately 1000 kg of the pigment astaxan-thin, worth altogether 10-15 million N.kr. In addition to this there is the possible commercial value of chitin for chitosan production, a product with many potential outlets in industry, biotechnology and medicine (Muz-zarelh, 1977). A minor proportion of the shrimp processing waste is frozen in blocks and used in moist feeds for cultured salmonids as a source of natural pigment, but most of it is dumped, at a significant expense for the factories. Despite the fact that astaxanthin present in shrimp is the natural pigment of salmonids, another carotenoid, canthaxanthin, has become the predominant pigment source in commercial diets; first because it produces nearly the same coloration of muscle as does astaxanthin (Schmidt and Baker, 1969; Ugletveit, 1974; Torrissen, 1978), and secondly because this pigment is available as a dry powder which can be incorporated into any fish diet at a controllable level. Shrimp waste, on the other hand, spoils quickly, the level of astaxanthin is variable, it is bulky and not available to fish farms far from shrimp processing plants, and is difficult to incorporate in feeds.

Ensiling acid

Shrimp processing

w a s t e / / Α^^^^φφ^

///

Product

Water vvuiei Figure 8.2 Ensiling and processing of shrimp waste

autolysate. It might be of interest also to remove bitter peptides from the aqueous phase and make a sauce-like product (cf. Raa and Gildberg, 1982).


SEMI-MOIST SILAGE PRODUCTION

These problems can now be circumvented by an ensihng process which produces a semimoist silage in which astaxanthin is stable and which is resistant to spoilage {Figure 8.2). Water is drained off the shrimp processing waste in a drum sieve (a), the waste transported via a screw (b) to a screwpress (c) which squeezes out water and produces a ground shrimp meal with a dry weight of about 58%; of this dry weight 50% is protein, 33% ash and the remainder mainly chitin. The dewatered shrimp shells are transported to an ensiling unit (d) where acids are added. The final product may be packed semimoist and stored in sacks or containers and either used on the fish farms as an additive to moist feed for salmonids, at a level of 5%, or transported to feed compounding factories to be incorporated in dry rations.

STABILITY OF PIGMENT AND CHOICE OF ACID

It has been demonstrated that astaxanthin is stable in an acid silage of shrimp waste and that the digestion of the pigment by rainbow trout was improved by ensiling, to about 70% as compared to 45% in the corresponding fresh or dried material (Torrissen et al., 1981/82). The shrimp waste, being rich in minerals, has a high buffering capacity and much acid is needed to lower the pH sufficiently to ensure preservation. For example, about 18 kg (10 €) of concentrated sulphuric acid is needed to lower the pH to 2.5 (after seven days) in 100 kg of screw-pressed shrimp waste. A silage produced by adding concentrated sulphuric acid will absorb the free water. The astaxanthin is not stable, however, in such a mixture, probably due to oxidative breakdown. By modifying the ensiling acid by including phosphoric acid, propionic acid and antioxidants, both lipid and water soluble, the pigment will be stable for months even at room temperature with access to air (unpubhshed data). The rationale behind using sulphuric and phosphoric acid is to precipitate the calcium which will be released by acid from the carbonate structure of the shells and thus prevent water seepage from the ensiled waste. Use of hydrochloric acid, on the other hand, results in the volume of solids decreasing to about half due to release of calcium chloride and protein, and evaporation of carbon dioxide. Almost all the pigment remains, however, in the solid fraction after ensiling in hydrochloric acid. This ensiling method is therefore a means of producing a more concentrated pigment source in an almost pure chitin. If chitosan production from chitin becomes an alternative method for utilizing shrimp processing waste, the pigment can be extracted directly in the feed oil (Raa and Hansen, patent pending) before the remaining chitin is subjected to chitosan processing.

ECONOMICS

The value of screwpressed shrimp waste preserved by acid is 2-2.50 N.kr/kg wet weight, calculated on the basis of market prices for protein


References

A L L E N , L.A. and H A R R I S O N , J. (1936). Ann. Appl, Biol., 231, 546 A L L E N , L.A. , H A R R I S O N , J., WATSON, S J . and F E R G U S O N , W.S. (1937). J.

Agric. ScL, 27, 271 A N D E R S O N , B.K. and JACKSON, N. (1970). / . Sci. Fd Agric, 21, 235 A N O N . (1981). Fish Farmer, 4, 4 A T K I N S O N , Α . , L A M P R E C H T , Ε. and MISPLON, D . (1974). In 28th Ann. Rep. of

the Director, Fishing Industry Res. Inst., Cape Town, 44 A U S T R E N G , E. (1982). Aktuelt fra Statens fagtjeneste i landbruket (Hus-

dyrförsöksmötet 1982), Norwegian Agricultural University, 525 B A C K H O F F , H.P. (1976). / . Fd Technol., 11, 353 B A R N E T T , A.J .G. (1954). Silage fermentation. 208 pp. Butterworths, Lon

don C A M E R O N , C.D.T. (1962). Can. J. Anim. Sci., 42, 41 DISNEY, J.G. (1979). Prospects for fish silage in Malaysia, Sri Lanka,

Bangladesh and the Philippines. IPFC Occasional Paper. FAO Regional Office, Bangkok, Thailand

D I S N E Y , J.G. and H O F F M A N , A. (1976). In Proc. Torry Research Station Symp. on Fish Silage, Torry Research Station, Aberdeen

D I S N E Y , J.G., H O F F M A N , Α . , OLLEY, J., CLUCAS, I.J., B A R R A N C O , A. and FRANCIS, B.J. (1978). Trop. Sci., 20,(2), 129

D I S N E Y , J.G. , P A R R , W.H. and M O R G A N , D.J. (1978). Proc. IPFC Symposium on Fish Utilization, Technology and Marketing in the IPFC Region, Manila, Philippines, 543-553

D I S N E Y , J.G. , T A T T E R S O N , I.N. and OLLEY, J. (1977), In Proc. Conf. on the Handling, Processing and Marketing of Tropical Fish, p. 231. Tropical Products Institute, London

D J A J A S E W A K A , H. and D J A J A D I R E D J A , R. (1980). In Proc. IPFC Workshop on Fish Silage, FAO Fish. Rep. No. 230:74

D U G A L , B. and R A A , J. (1978). IRCS Medical Sci., 6, 546 D U R A I R A J , S., S A N T H A N A R A J , T.. Md. S U L T A N , K.M., D O R A I R A J A H , K . A . P . A .

(1976). In Proc. Symp. Fish Processing Industry in India, Mysore, India E D I N , H. (1940). Nord. lordbr. Forsk., 22, 142 F O N G E , J. (1976). Pig Farming, May, 43 F R E E M A N , H.C. and H O O G L A N D , P.L. (1956). / . Fish. Res. Can., 13, 869 G I B S O N , T., STIRLING, A . C . , K E D D I E , R.M. and R O S E N B E R G E R , R.F. (1958). / .

Gen. Microbiol., 19, 112 G I L D B E R G , A. and R A A , J. (1977). J. Sci. Fd Agric, 28, 647 G I L D B E R G , A. and R A A , J. (1979). Comp. Biochem. Physiol, 63B, 309 H A L E , M.B. (1969). Food Technol, 23, 107 H A Y A S H I , T. and N A G A I , Y. (1973). / . Biochem., 73, 999

and astaxanthin. The preservatives account for about 10% of this value, leaving room for depreciation of the necessary investment in this process, which has become even more attractive in the light of the proposed ban on canthaxanthin in fish feeds (Anon., 1981).

J, Raa, Α. Gildberg and Τ. Str0m 131

H I L L Y E R , G . M . , P E E R S , D . G . , M O R R I S O N , R . , P A R R Y , D . A . and W O O D S , M . P .

(1976). Proc. Torry Res. Station Symposium on Fish Silage, Aberdeen, UK

H J E L M E L A N D , K . (1978). Buksprenging hos lodde. Thesis, Institute of Biology and Geology. 115 pp. University of Troms0, Norway

H J E L M E L A N D , K . and R A A , J . (1980). In Advances in Fish Science and Technology, Ed. by J.J. Connell, p. 456. Fishing News Books Ltd, Farnham, Surrey

H J E L M E L A N D , K . and R A A , J . (1982). Comp. Biochem. Physiol., 71B, 557 J A M E S , M . A . and N A I R , M . R . (1977). Proc. Conf. Handling, Processing and

Marketing of Tropical Fish. p. 273. TPI, London J E N S E N , J . and S C H M I D T S D O R F F , w. (1977). In Proc. lAFMM Symp. on the

Production and Use of Fish Meal, Szczecin, Poland K E D D I E , R . (1951). Proc. Soc. Appl. BacterioL, 14, 157 KN0CHEL, S . (1981). Mikrobiell fermentering af fisk ved hjelp av naturligt

forekommende laktobaciller. Hovedoppgave, Fiskeriministeriets For-S0gslaboratorium, K0venhavn Univ.

K O M P I A N G , L P . , A R I F U D I N , R . and R A A , J . (1980). In Advances in Fish Science and Technology, Ed. by J.J. Connell, pp. 349-352. Fishing News Books Ltd, Farnham, Surrey

K O M P I A N G , L P . , Y U S H A D I A N D C R E S W E L L , D . C . (1980). In Proc. IPFC Workshop on Fish Silage, FAO Fish. Rep. No.230,38

K R O U L I K , J . T . , B U R K E Y , L . A . and W I S E M A N , H . G . (1955). J. Dairy Sci., 38, 256

L I N D G R E N , S . and CLEVSTR0M, G . (1978a). Swedish J. Agrie. Res., 8, 61 L I N D G R E N , S . and CLEVSTR0M, G . (1978b). Swedish J. Agric. Res., 8, 67 L I S A C , H . (1961). In Proc. GFCM, 6, 111 M C D O N A L D , P . and P U R V E S , D . (1956). J. Sci. Fd Agrie. 3, 189 M O H R , V . (1980). Process Biochemistry, 15(6), 18 M U Z Z A R E L L I , R . A . A . (1977). Chitin. Pergamon, Oxford N I L S O N , R . and R Y D I N , C . (1963). Acta Chem. Scand., 17, 174 N I L S S O N , G . and N I L S S O N , P . E . (1956). Arch. Mikrobiol., 24, 412 O L S S O N , N . (1942). Lantbrukshögskolan Husdjurförsöksanstalten. Sweden,

Rep. No. 7:55 pp. P E D E R S E N , S . (1976). Lic. techn. thesis, Norwegian Institute of Technolo

gy, University of Trondheim (English summary) 220 pp. P E T E R S E N , H . (1953). FAO Fish. Bull., 6(1), 18 R A A , J . (1981). Plenary lecture in GIAM VI, Global impacts of Applied

Microbiology, Lagos, Nigeria. Academic Press Congress Proceedings, Emejuaive, Ogúnbi, Sanni, ed.), 3

R A A , J . and G I L D B E R G , A . (1976). / . Fd Technol., 11, 619 R A A , J . and G I L D B E R G , A . (1982). CRC-Critical Reviews Series in Food

Science and Nutrition, 16,(4), 383 R A A , J . and H A N S O N , K . Norwegian patent application 803582 A23K R O A , P . D . (1965). Fish. News Int., 4(3), 283 SCHR0DER, K . , C L A U S E N , E . , S A N D B E R G , A . M . and R A A , J . (1980). Advances

in Fish Science Technology, Proc. Torry Jubileum Conference, Torry Research Station, Aberdeen, UK, 1979, 480

S C H M I D T , P . J . and B A K E R , E . G . (1969). / . Fish. Res. Board Can., 28, 357 S H E I K H , A - S . and S H A H , F . H . (1974). Pak. J. Sci. Ind. Res., 17, 136


STIRLING, A . C . and W H I T T E N B U R Y , R. (1963). / . Appl. Bacteriol., 26, 86 S T R 0 M , T. and E G G U M , B .O. (1981). J. ScL Fd Agric, 32, 115 S T R 0 M , T., G I L D B E R G , Α . , STORMO, Β. and R A A , J. (1979). Proc. Torry

Jubileum Conf. Torry Res. Station, Aberdeen, UK T A R K Y , W., A G A R W A L A , O.P. and PIGOTT, G.M. (1973). / . Fd Sci., 38, 917 T A T T E R S O N , I.N. (1976). Proc. Torry Res. Station Symposium on Fish

Silage, Aberdeen T A T T E R S O N , I.N. and WINDSOR, M.L. (1974). / . Sci. Fd Agric, 25, 369 T O R R I S S E N , O.J. (1978). Hovedfagsoppgave i ernaeringsbiologi til mate-

matisk naturvitenskapelig embetseksamen, Universitetet i Bergen, 134 pp.

TORRISSEN, O. , T I D E M A N N , E . , H A N S E N , F. and R A A , J. (1981/1982). Aquaculture, 26, 77

U G L E T V E I T , S. (1974). Fisken Havet, 9, 31 WESSELS, J .p.Η. and L A B U S C H A N G E , S. (1974). In 28th Ann. Rep. of the

Director, p. 48. Fishing Industry Res. Inst., Cape Town W H I T T E M O R E , C T . and T A Y L O R , A . G . (1976). J. Sci. Fd Agric, 27, 239 W I G N A L L , J. and T A T T E R S O N , I. (1976). Process Biochem., 11,(10), 17 W I R A H A D I K U S U M A H , S. (1968). Lantbrukshögsk. Annlr., Sweden, 34, 551 W I R A H A D I K U S U M A H , S. (1969). Lantbrukshögsk. Annlr., Sweden, 35, 823 Y O S H I N A K A , R., S A T O , M. and I K E D A , s. (1978). Bull. Jap. Soc. Sci. Fish.,

44, 639

9

Introduction

An overview of the apphcation of enzymes in upgrading food by-products was given at the Institute of Food Science and Technology Annual Symposium in July 1981 Food Technology in Europe (Fullbrook, 1981). The main objective was to put forward a few concepts and ideas which might prove of value in stimulating food manufacturers to consider enzymes when dealing with by-product problems. The points made were a consideration of why the secondary products of the food processing industry should be re-processed and an estimation and characterization of the available wastes within the UK. This was supplemented with a brief consideration of enzymes as suitable processing agents and their commercial availability, and concluded with two examples of how enzymes are used to create higher value products from food waste materials. In particular, the production of protein hydrolysates from oil seeds was detailed as an example of possible future interest. Certainly the increasing costs of raw materials, energy and effluent disposal, combined with the pressure for new products within the food industry, the availability of food grade industrial enzymes and the improvement of separation techniques, make the recovery of higher-value components worth considering.

However, what is possible in the laboratory and what is practicable on the factory floor are often on opposite sides of reality. In this chapter, therefore, a different approach is taken to that pursued previously, and a serious attempt is made to be completely objective in assessing the large-scale use of enzymes in the modification of food waste.

By now, almost all interested parties are aware of the potential of the new biotechnology and most are apprehensively awaiting some viable and down-to-earth examples of its exploitation. In an area so ripe with potential, it is all too easy to make out a case for investigating interesting possibihties, and to pose stimulating questions for research—but this is not necessarily going to lead to realistic solutions for what are essentially technological problems. But it is not only the universities and institutions which present perhaps an over-enthusiastic case for biotechnical solutions to these problems: the commercial enzyme manufacturers, constanfly

*Present address: Imperial Biotechnology Ltd, 8 Princes Gardens, London, SW7 I N A .

133

THE USE OF ENZYMES

P.D.FULLBROOK* National College of Food Technology, University of Reading, UK

134 The use of enzymes

The attitude of the food industry

The UK Food Research Association recently sent out a biotechnology questionnaire to all its members and to various related trade associations (Holmes and Jarvis, 1981). As only 24% were completed and returned, one might—perhaps ungenerously—conclude that this reflects a less than enthusiastic attitude of the food industry in general to biotechnology and an apparent disinterest in the potential for improving its efficiency by upgrading its waste products. Nevertheless, of the 103 returned questionnaires, 22% of the companies (i.e. 35 organizations) stated that during 1980 they were actively involved in upgrading their food wastes or in fermenting them to produce a combustible fuel (methane or ethanol). A further 61% said that they would be interested. An indicated 24% said that they were involved in the upgrading of dairy by-products such as skim milk or whey. In fact, it is probably the dairy and confectionery branches of the food industry who seem most committed to reprocessing of waste products—possibly because they produce the highest value by-products or are operating on the narrowest margins. A wide variety of uses for microorganisms and enzymes to process or upgrade waste products was indicated, across the whole spectrum of food materials processed by the various industry sectors.

Availability of food waste materials

One of the most valuable parts of the report was a quantification and value estimation of available wastes, and Table 9.1 is derived from some of the data collected. This would represent the substrate availability in consideration of enzyme reprocessing and gives a useful input for a consideration of process economics. Although incomplete, Table 9.1 indicates that approximately 37 X 10^ tonne per annum of waste are produced on 137 locations. Roughly four-fifths of this are currently considered to be of such low value

seeking new markets, also claim to have the capability to produce any enzyme for which a need exists or a market can be developed. Undoubtedly they have the capability—but at what cost? This is perhaps the crux of the whole matter, and we should therefore be more cautious in applying essentially high technology to medium to low value by-products, processed at intermediate through-put rates.

In order to validate this statement, this chapter considers three points: the attitude of the food industry to the possibilities of upgrading byproducts using biotechnological methods or agents; the availability of food wastes and enzymes; and finally some current problems associated with the industrial use of enzymes in the processing of by-products. Whilst it is often relatively easy to show some promising technological advance at laboratory scale, many academic colleagues fail to recognize that in the industrial situation technical success, but economic failure, is still failure. Economic risks involved in any piece of development work should always be considered along with, and not after, any technical development—a point sadly illustrated by several of the chapters in this Easter School volume.

p. D. Fullbrook 135

Table 9.1 A V A I L A B I L I T Y OF SOME WASTES IN T H E UK ( D A T A FROM H O L M E S A N D JARVIS, 1981)

Type of waste No. of Total wt (tonne Wtsold Sales value Disposal cost sites per annum) (%) (£/tonne) (£/tonne)

Liquid, total 46 30 X 10^ 0.003 47.3 0.2 effluent 37 29 X 10^ 0 — 0.2 valuable X 10^

Whey 1 140 0 0 — Sugar and glucose 2 80 0.04 175 2.5 Non-milk 1 2 0 0 0 Oil 1 0.4 100 75 — Fat 1 0.005 100 130 — Solids, total 91 6.7 X 10^ 95 23 1.0

Carbohydrate X 10^ Sugar 9 1.9 79 165 3.4 Fruit/veg 9 40 58 1.4 — Spent grains 6 171 100 15.9 — Dough waste 4 4.7 100 102.5 — Cake waste 2 2.5 100 39.2 — Draff 1 2.4 100 80.0 — Protein and fat Offal 5 80.3 99 8.0 8.1 Rind fat 3 0.4 100 48.5 — Fat 3 0.6 100 416.5 — Fat and meat sludge 2 5.2 3 12.0 5.6 Fish shells and bones 3 1.8 100 39.2 — Protein and fat 2 1.3 100 234.9 — Vegetable protein 2 1.25 60 6.0 6.0 Miscellaneous Yeast 3 5.6 100 160.2 — Cocoa shell 5 7.5 100 38.4 — Inorganic Spent catalyst 1 0.5 100 150 — Spent earth 1 3.3 59 35.7 10.9

as to be considered as effluent which accrues an average disposal cost of 20p/tonne (costing in total approx. £6 million). Of the remaining one-fifth considered worth processing, only 15% is liquid material (the majority of which is sugary material), only a minute proportion of which is utilized. Most of the easily separable wastes, including water insolubles such as fats, are recovered and sold. Caution should be used in interpreting these data as they do not take into account the large volumes of cheese whey (>600000 tonnes) which are available for processing. Of the solids processed, roughly 300000 tonnes were of predominantly carbohydrate nature and 100000 tonnes essentially higher value protein and fat. Only 80 out of 1000 tonnes of material had a sale value greater than £100/tonne which represented about 1% of the total. Not surprisingly, the highest proportion of material sold was that which was either the easiest to separate, or of the highest potential value.

The possibilities for improving the value of waste materials must therefore concentrate on:

(1) the large volume of low solids liquid waste material, currently considered as effluents;


The availability and use of enzymes generally

The availability of industrial grade enzymes has been very much in balance with demand, mainly because the initiative for creating the demand has been that of the enzyme producing companies themselves in developing the new enzyme processes. This has meant that the price of an enzyme for a particular appHcation for which it has been developed, has also been reasonably within the control of the enzyme producers—a point which has great significance, and which is discussed later.

Table 9.2 SOME I M P O R T A N T CHARACTERISTICS OF I N D U S T R I A L E N Z Y M E S

E N Z Y M E S A R E • true catalysts • (glyco) proteins

A N D T H E R E F O R E • highly specific (also stereospecific) • highly active substances • low cost • available in virtually unümited quantities • non-polluting • biodegradable

O P E R A T E A T • low temperature (10-110°C) • low pressure (760 mm) • neutral conditions (pH 4 - 1 0 , low salt)

U S U A L L Y IN • aqueous phase • homogeneous phase (?)

A F F E C T E D B Y • activators/inhibitors • 'extreme' conditions • metal ions/chelating agents

A C T ON • 'natural' biological materials and organics A N D T H E R E F O R E • easily controllable

Some of the most important characteristics of industrial enzymes are given in Table 9.2. Because of these general properties, enzymes would seem technically to be the ideal agents for processing food wastes, especially when it is remembered that industrially used enzymes are relatively crude products, as indicated by Table 9.3. However, even though their chemical composition may appear crude, their catalytic action under

Table 9.3 C O M P O S I T I O N O F T Y P I C A L I N D U S T R I A L E N Z Y M E P R O D U C T

Component Dry solids (%)

Proteins and amino acids 10-15 Active enzyme protein 2 - 5 Complex carbohydrates 5 -12 Sugars 2 - 4 0 Inorganic salts 3 - 4 0 Preservatives/stabilizers 0 - 0 . 3

(2) the small portion of sohd waste currently considered as not worthy of recovery or sale;

(3) producing products of higher value from the currently recovered materials.

p. D. Fullbrook 137

Table 9.4 TYPICAL COSTINGS FOR I N D U S T R I A L E N Z Y M E PROCESSING ( D A T A F R O M G O D F R E Y , 1982)

Process area

Starch glucose syrup fructose syrup

Natural extracts

Proteins

Fermentation

Detergents

soluble functional flavour

brewery distillery alcohol

Textile desizing Leather production

Enzymes Approximate Unit process involved cost{i) (tonnes)

amylases 11 starch amylases 14 starch isomerase complex 60 source

carbohydrates material acidic 120 neutral 180 protein alkahne 70 source amylases 9 raw material glucanases 7 'as is' proteinases 10 alkaline 40 product

proteinases amylases 3 size proteinases 4 skin

correct conditions of storage and use is guaranteed and their microbiological and toxicological specifications are of a consistent and high standard, often well within pharmaceutical limits. The UK currently consumes about 2000 tonnes per annum of industrial enzymes mainly imported from other EEC countries, having a combined value of £6.5 million. This represents only about 4% of the world market for these products. If we were to express these amounts of industrial grade products in terms of the amount of pure active protein, the current UK usage would only be about 46 tonnes, 95% of which would be proteinase and amylase produced by Bacillus spp. and amyloglucosidase produced by Aspergillus niger. These relatively low amounts of material are of course due to the high catalytic nature of enzymes, since they are only required in small amounts when acting under ideal conditions. Even when used as crude industrial products, they are only used in small doses, typically in the region of 0.01-1.00% w/w of substrate acted upon.

As imphed above, the costing basis for industrial enzymes is related more to the added value of the primary products produced by their mediation and the R & D costs involved in the development of the enzyme products, than the actual production cost of the enzyme itself. Some typical process costings for several typical industrial enzymes used in the processing of food products and natural biological materials of commercial importance are given in Table 9.4, This gives a clear indication of the orders of added processing costs incurred when using enzymes to produce or treat primary products. They range from the very low enzyme processing costs involved in the non-food process through the intermediate costs associated with the conversion of carbohydrate materials, to the relatively higher costs in the emerging protein modification processes. A notable exception is the intermediate cost of alkaline proteinases used as ingredients of biological detergents—but this is due to higher formulation costs, as these enzymes are sold as solid granulated products rather than stabilized Hquids.


Problems in the use of industrial enzymes

In addition to the above point, the use of enzymes in treatment of by-products highlights some of the problems associated with the current range of industrial enzymes.

A point not generally appreciated is that currently marketed industrial enzymes are of low catalytic performance in comparison with the majority of enzymes. This is due to two fundamental reasons. Firstly, most enzymes operating in industrial processes act upon substrates which are just not their natural substrates. Their use exploits the broad specificity of the enzyme and often the catalytic efficiency of the enzyme, when acting on its industrial substrate, is low. This is illustrated in Table 9.5 for the industrial

Table 9.5 N O N - I D E A L I T Y O F I N D U S T R I A L L Y - U S E D E N Z Y M E S F O R T H E I R

C O M M E R C I A L L Y - U S E D S U B S T R A T E S ( D A T A F R O M N I E L S E N , 1980)

Parameter value Substrate D-glucose D-xylose D-ribose L-arabinose

Kn, (M) 0.17 0.01 0.23 0.15 V ^ 3 x ( M / g / m i n ) 0.09 0.1 0.07 0.005 P E ratio 0.5 10 0.3 0.03

enzyme glucose isomer ase (EC: 5.3.1.5). The kinetic constants and Vmax (related to the enzyme turnover number) for a variety of substrates is shown, and clearly the efficiency of the enzyme as measured by the PE ratio = Vmax/Km, is low for the commercial substrate, glucose in comparison with the natural substrate xylose—by a factor of 20 in fact.

Secondly, since most of the more established apphcations of industrial enzymes were developed to supplement, supersede or replace chemical processes—which themselves were originally designed to run at elevated temperatures (50-140°C) in order to improve solubility, viscosity, etc., of natural products and allow faster reaction rates—the majority of industrially used enzyme systems operated today have been specifically selected and designed to be primarily thermostable rather than super-efficient. One might hope that the next generation of industrial enzymes will be designed to be both efficient at low temperatures and stable under operating

Even though we now recognize over 3000 different enzymes, to date only about 20 or so are available as products suitable and cheap enough for industrial processing. Of these, most are, however, of the type suitable for degrading complex biological polymers—the type of compounds in fact we would expect to find in food wastes—being of broad specificity and able to hydrolyse a wide variety and mixture of complex substrates. In general, only those products for which there already exists a finite primary industrial use are available for the secondary activity of treating by-product wastes. As the form and specification to which they are manufactured are to a large extent specifically tailored to their primary use, they are often only available in a higher quality grade than is really necessary for treatment of secondary or by-products—with obvious economic penalties.

p. D. Fullbrook 139

conditions, in order to minimize rising energy costs, and hence lower the accrued costs in secondary processing.

Another problem with the use of enzymes in processing by-products concerns their robustness. As is evident from Table 9.2, enzymes are typically sensitive to both inhibitory substances and extremes of pH, temperature and salt concentration. Hence removal of potential inhibitory substances from effluents or by-products is often mandatory, and reaction conditions will require careful monitoring. These factors will combine to have a bearing on the overall treatment cost, which will ultimately determine the viability of any enzyme/by-product process.

The amount of enzyme required for treating a given material to a defined degree of conversion is dependent on the interaction of a number of factors which are Hsted in Table 9.6. As well as the chemical nature of the substrate, and the general chemical and physical environment, the kinetic constants and characteristics of the enzyme have to be taken into account. Physical conditions and reaction times will need to be tailored in order to obtain maximum enzyme productivity—whilst if the by-product is to be used as a food ingredient—the residual enzyme activity in the product will have to be minimal, or preferably zero.

Table 9.6 F A C T O R S AFFECTING A M O U N T OF E N Z Y M E R E Q U I R E D

Factors Parameters Example

Substrate Concentration SolubiHty Physical condition Particle size

Denaturation Physical conditions Temperature

pH Environment Ionic strength, salts

potential inhibitors Enzyme characteristics StabiHty

Activity ' productivity

^ p H , temp: optima range kinetic constants K„,,k3 (V_)

Time available for reaction

Clearly then, for a viable process, the added-value obtained when using an enzyme to process a by-product must be significantly greater than the extra processing cost incurred. These will be directly related to the concentration of enzyme required multiplied by its unit cost plus energy and depreciation costs of any extra equipment required. Typical enzyme processing costs are up to £l/m-^ digester capacity for effluent treatment or £7/tonne substrate for upgrading food waste materials (Godfrey, 1982). These costings equate to the lower end of the typical processing costs summarized in Table 9.4.

In conclusion therefore, the reality is that when considering the use of enzymes for upgrading by-products, caution should be exercised. Consideration should be given concurrently to both the technical and economic feasibihty, bearing in mind the points outlined above. Whilst enzymes definitely have the potential for upgrading food by-products, the practical reality is often at variance with this. Because of currently operated recovery practices and outlets; the need for the rechannelling of often


References

F U L L B R O O K , P. (1981). IFST Pwc., 14,(4) G O D F R E Y , A . (1982). Market Opportunities in Biotechnology. Conference,

University of Sheffield H O L M E S , A . W . and JARVIS, B. (1981). Application of biotechnology to the

food industry—an appraisal. FRA Publication, Leatherhead, UK NIELSEN, M. (1980). In Chemical Engineering in a Changing World.

Elsevier Scientific, Amsterdam

already hard-pressed R&D resources and the current high demand for industrial enzymes for primary processing, the use of enzymes for secondary processing of materials is less attractive and often uneconomical—at least for the present.

10

Introduction

The need for an ever increasing volume of food for an expanding population and for a nutritional balance for the present and future populations forces the world to consider novel sources, and to re-evaluate conventional sources of food and feed. If a large fraction of the world's population continues to be malnourished and hungry, it will be a politically unstable world. Therefore, it is in the best interest of all nations to try to provide an adequate diet for all people. In the absence of effective population control, an adequate diet for all people will continue to be an elusive goal. While the yields of conventional agriculture have increased dramatically in this century, the loss of agricultural land to urban sprawl, to erosion, to depletion and to salinization, limit its potential for further growth. Therefore, it is essential that new sources of food and feed be found or developed.

Microbial growth on agricultural and food processing wastes offers the potential for additional foods and feeds of a wide variety. Because the doubhng times are hours or days instead of months or years, microbes can produce many times as much protein per unit area as slower growing plants and animals. It is possible to use not only waste biomass but also high carbohydrate, low protein crops for microbial conversion into foods and feeds. The phosphate and nitrogen fertilizers can be added to the microbial fermentation tanks instead of the fields thereby reducing run-off, soil loss and stream pollution.

In any discussion of foods and feed, it is necessary to distinguish between need and demand. Need is a physiological or nutritional term. Demand is a commercial term of the marketplace. Demand is created by buyers with funds or credit for purchase at the marketplace. It is well established that hundreds of milHons of people throughout the world are undernourished or malnourished and they are undernourished because they cannot create a demand in the world food market. Microbial conversion of wastes offers the opportunity to convert a large fraction of their nutritional need into an effective demand by helping the undernourished in the Third World to supply their own food and feed through labor intensive, low technology, low capital processes.

The problems related to introduction of a new food or feed are of three

THE USE OF MICROBIOLOGICAL AGENTS IN UPGRADING WASTE FOR FEED AND FOOD

W.D. BELLAMY Food Science Department, Cornell University, USA.

141

142 The use of microbiological agents in upgrading waste

Protein source Biological value''

Digestibility^ Net protein utilization''

Egg 100 97 97 Milk 93 97 90 Oatmeal 79 60 47 Yeast

C . utilis sulfite waste liquor 32-48 85-88 27-42

Bacteria Bacillus megaterium

whole 62 56 35 broken 70 67 47

Fungi Fusarium sp. 70-75 — —

Algae Chlorella 54 65 35 Spirulena maxima 72 84 60

Reproduced by permission of the publishers a n - 1 · 1 1 Retained Ν ^Biological value = - r r — τ τ τ x 100 ^ Absorbed Ν b T ^ - . u i - . Absorbed Ν "Digestibihty^ Total Food Ν ''Net protein utilized = Biological value x Digestibility

content of cells is directly related to the rate of growth, therefore, micro-organisms contain a higher fraction of nucleic acids than most plant and animal tissues. Although about 90% of the purines are recycled, the remaining 10% are oxidized through xanthine to uric acid. Because man has lost the enzyme uric acid oxidase that oxidizes relatively insoluble uric acid to soluble allantoin, consumption of foods with a high nucleic acid content leads to high levels of plasma uric acid. Free uric acid and ureate

Table 10.2 N U C L E I C A C I D C O N T E N T O F F O O D S

Source % Protein NAIlOOg protein

Cereal flour 9 -16 1-2 Lean beef 45 2 Liver 77 4 Mushrooms 13-53 1-4 Yeasts 40-60 3 - 6 Spirulina 64-70 2 - 4 Bacteria 40 -70 8-25

kinds: technical, economic and cultural. A product must pass all three tests to be a success. There are numerous examples of products that were technical and economic successes but failed to get consumer acceptance and, therefore, failed. Single cell protein (SCP) produced from hydrocarbons in Japan is a good example where the consumers refused to purchase either hydrocarbon-grown SCPs or chickens grown on the hydrocarbon-grown SCPs (Rockwell, 1976). Many SCPs are comparable with the better grains in biological value, digestibility and net protein utilization {Table 10.1). They contain, however, high levels of nucleic acids. The nucleic acid

Table 10.1 N U T R I T I V E V A L U E O F S C P ( F R O M C H E N A N D P E P P L E R , 1978)

W. D. Bellamy 143

(1) aseptic pure-culture, and (2) non-sterile mixed cuhure fermentations.

INDUSTRIAL FERMENTATION PRODUCTS

Non-s t er i l e environment controlled

fermentations

Pure culture asept i c

fermentations

W i n e s

C h e e s e

M i s o

S C P

Vinegar

Sauerkraut

Si l ä g e

Compost

A Vaccines

Ant ibiot ics

H o r m o n e s

Amino a c i d s

E n z y m e s

S C P

Λ Figure 10.1 Relative value of products produced by pure-culture aseptic and by non-sterile mixed culture fermentations. Commercial products of low value cannot be produced by aseptic fermentations

Figure 10.1 presents examples of each type and the relative value of the products. The products of aseptic pure-culture fermentations include antibiotics, hormones, amino acids, enzymes, food additives, etc. Products of non-sterile fermentations include wine, cheese, sauerkraut, silage and compost. In general, the products of aseptic fermentation are worth $/g, while those of the second process are worth $/kg or less. There are obvious exceptions where special wines, exotic cheeses and truffles are worth more than food additives and some amino acids.

The products of the sterile process are valuable enough to support gleaming stainless steel fermentors, sophisticated monitoring and controlling instrumentation. It is a high technology, capital intensive, skilled labor operation. Only in times of economic upheaval such as World Wars I and II has it been profitable to produce animal feed by this method. The non-sterile mixed culture operation, on the other hand, is low capital, low technology and labor intensive. It includes processes such as cheesemak-ing, wine and beer fermentations, sauerkraut and pickle preservation; processes that were developed empirically long before microbes were known to exist.

salts may precipitate in the kidney forming kidney stones, or in cartilaginous tissues producing the disease, gout. Chnical studies have shown that the safe intake of nucleic acids for a healthy adult is about 2 g/day (Scrimshaw, 1975). Enzymatic and chemical methods have been proposed for removal of nucleic acid from SCP (Sinskey and Tannenbaum, 1975; Shetty and Kinsella, 1978; Otero and Cabello, 1980, Otero et al, 1982) {Table 10.2).

There are two types of microbial conversions that need to be considered:


Pretreatment

When plant tissue is the substrate, pretreatment to rupture the hgnocellu-lose complex is necessary in all pure culture or enzymic degradations. Without pretreatment, microbial digestion of fibrous wastes would be too slow for economical operation of pure culture fermentors. It is not necessary to completely remove the lignin but the physical barriers to cellulase must be removed. Pre treatments include hot alkali, hot dilute acid, cold concentrated acid, mechanical grinding, steam explosion and hot organic solvent extraction. The lignin degrading enzymes in fungi and actinomycetes require an aerobic environment and weeks or months for significant degradation (Crawford, 1981).

Aseptic monocultures

Whey and fruit juices constitute unique food processing wastes because they are completely edible and, therefore, products from these wastes do not meet with consumer resistance. The advent of commercial ultrafiltration, reverse osmosis and electrodialysis have produced a revolution in the milk processing industry {see Walters, Chapter 4). These processes may find application in other waste processing. By removal of the albumins and globulins as well as the casein from whey, it is possible to increase the cheese yield by 10-20%. Whey contains 55% of the nutrients and 85-90% of the original volume of milk. Fluid whey contains 4.8-4.9% lactose and 0.75-0.8%^ protein.

Whey, the liquid serum residue from cheesemaking, is fermented by lactose utilizing yeasts such as Kluyveromyces fragilis to SCP and ethanol in both the USA and Europe (Kosikowski, 1979), or by Saccharomyces sp. after hydrolysis of the lactose to monosaccharides by immobilized lactase. At this time, alcohol is the primary product while SCP is a secondary but essential product. Economic operation depends upon a ready market for the high protein concentrate as SCP as well as for the alcohol (Rajagopalan and Kosikowski, 1982).

Pure cultures of fungi such as Trichoderma reesei and Chaetomium cellulolyticum have been proposed for pure culture production of SCP from straw and corn stover (Moo-Young, MacDonald and Ling, 1981).

It now appears that pure cultures may be used in two areas of waste recycling in addition to those already mentioned. The production of microbial enzymes for waste processing will probably require pure cultures grown under sterile conditions. Enzymes such as cellulase, amylase, protease, pectinase, and hemicellulase all could find application in waste conversion to fermentable sugars either with or without sterilization by heat or filtration (Chahal, Swan and Moo-Young, 1977; Cantneros et α/., 1982; Fullbrook, Chapter 9). The second appHcation involves production of specific proteins or other valuable biologicals by the unique organisms produced by genetic engineering or recombinant DNA techniques. This application will be discussed later.

While the estimated economics for the previous described processes look favorable, no commercial plants have been constructed in the USA or

W. D. Bellamy 145

6 8 10 Tinne(h)

Batch cultivation of C utilis stover hydrolysate at 30 °C

CMI -233 I I on corn and pH 4.5

Figure 10.2 Utilization of hexoses and pentoses in hot acid pretreated corn stover by pure culture of Candida utilis (Gonzales-Valder and Moo-Young, 1980, reproduced by permission of the publishers)

Fähnrich and Irrgong (1981) reported that the yield of cellulolytic enzymes was suppressed by substrate concentration of 2% compared with 1%. Selection of mutants without substrate inhibition may be necessary for any commercial appHcation of SCP production by submerged fermentation.

A soHd state (28-30% dry matter) sterile fermentation with a dual pure culture of Chaetomium cellulolyticum or Trichoderma lignorum and Candida lipolytica has been reported (Viestures et al, 1981).

Canada. Chahal, Swan and Moo-Young (1977) described a process for fermenting wheat straw with Chaetomium cellulolyticum to produce SCP with up to 43% protein. The microbial utilization of straw for SCP was reviewed by Hah (1978). He concluded 'Economic and technological conditions in our modern society make current uses of straw impractical...'. Kargi and Shuler (1981) described a mixed yeast/bacteria process for conversion of poultry waste into high protein feedstuff. A uricase producing strain of Candida utilis was grown with the natural flora of poultry manure. Molasses was used as the carbon source. The product had a lysine and methionine content comparable with other feedstuffs.

The advantages of the mixed pure cultures are reported to be more rapid substrate utilization and removal of product inhibition. Strains of yeast such as Candida utilis can utilize pentoses from hemicellulose as well as hexoses from cellulose {Figure 10.2).


Non-sterile mixed cultures

The non-sterile mixed culture process finds extensive apphcation today and, I believe, will find greater application in the future for reasons that I will mention later. The non-sterile process depends upon selecting the proper conditions of temperature, pH, nutrients, and humidity that will favor the growth of the desired organisms and will inhibit many of the unwanted organisms. In some cases, a heavy inoculum of the desired culture will outgrow undesirable organisms that may be present.

The microbial process that is most applicable and needs the least amount of new technology through research and development is mushroom cultivation on food processing and agricultural wastes. Mushroom cultivation on wastes is a labor-intensive process that requires little capital and only a minimum knowledge of microbiology (Gray, 1973; Kurtzman, 1979; NRC, 1981). The common mushroom, Agaricus bisporus, does not utilize cellulose nor lignin and depends upon a pretreatment by composting of the cellulosic wastes. Other species utilize cellulose, lignin, or both. Table 10.3

Table 10.3 M U S H R O O M CULTIVATION CONDITIONS ( K U R T Z M A N , 1979)

Species Temperature (°C) Level of Waste substrate Species environmental

Spawn Fruiting control running required''

Agaricus bisporus (common 20-27 10-20 + 4- + + Composted horse mushroom) manure or rice

straw Agaricus bitoquis 25-30 20-25 + + + + Composted horse

manure or rice straw Auricularia sp. (ear 20-35 20 -30 + + + Sawdust-rice

mushroom) straw Coprius fimetarius 20-40 20 -40 Straw Flammulina velutipes 18-25 3 - 8 + + + Sawdust-rice

(winter mushroom) bran Lentinus edodes 20-30 12-20 Logs or sawdust-

(shiitake mushroom) rice bran Pholiota nameko (nameko 24 -26 5-15 + + + Logs or sawdust-

mushroom) rice bran Pleurotus ostreatus 20-27 10-20 Straw, paper

(oyster mushroom) sawdust-straw Stropharia rugosoannulata 25-28 10-20 Straw, paper

sawdust-straw Tremella fuciformis 20-25 20-27 + + + Logs or sawdust-

(white jelly mushroom) rice bran Volvariella volvacea 35-40 30-35 Straw, cotton

(straw mushroom) wastes

^+ + + + = greatest; + = least.

Hsts some of the more commonly grown mushrooms, their environmental requirements and their substrates. The paddy straw mushroom, Volvariella valvocea, and the oyster mushroom, Pleurotus ostreatus, can utilize up to 70% of rice or wheat straw. The remainder has been reported to have a feed value equivalent to hay (NRC, 1981). While growth on sohd wastes is a labor-intensive operation, growth of selected mushrooms on liquid

W. D. Bellamy 147

wastes in aerated fermentation tanks is possible and offers a more labor-efficient process but requires more stringent control of contaminants. Many species of ascomycetes and basideomycetes have now been grown in submerged aerated cultures on such diverse products as asparagus juice, corn steep Uquor, citrus fruit press water and orange juice. Many do not produce the characteristic mushroom odor and flavor nor the morphology of sohd grown mushrooms (Gray, 1973).

As mentioned, the non-sterile mixed culture processes constitute the major paths for waste recycling now and in the foreseeable future. These processes can be divided into two types. In the first, mixed culture microbes are grown on waste and harvested for food or feed. In the second process, the harvesting is performed by fish, crayfish or other aquatic animals. There is an extensive literature related to the use of mixed cultures of algae and bacteria to process domestic sewage, animal wastes and food processing wastes (NRC, 1981). The algae can be harvested for animal feed with potential yields as high as 82000 kg of protein/ha per annum. There is a concomitant reduction in biological oxygen demand (BOD) and chemical oxygen demand (COD) of the wastewater. Most of the algae studied are micro algae and must be harvested by centrifugation, fine filtering or autoprecipitation. Oswald (1978) and associates (Eisenberg et al,, 1981; NRC, 1981) reported on a process of autoflocculation in algae cultures as a means of preconcentrating the algae suspension and have achieved better than 90% algae recovery. The high yield of algae protein and the reduction in water pollutants make this a very attractive process depending upon environment, temperature and sunshine.

Spirulina, a multicellular Cyanophyceae, is large enough to be harvested by filtering with cloth and has been used as human and animal feed in Chad and Mexico from historic times (Clement, 1975; NRC, 1981).

The septic mixed culture system that is most used throughout the world today is aquaculture; polyculture fish ponds, crustaceans in flooded rice fields, oysters and fish in waste treatment ponds. All these systems are finding increased use both in developed and developing countries and they all depend upon a complex mixed culture of micro-organisms for their success (Figure 10.3).

Roo ted m a c r o p h y t e s

Benth ic and aquatic protozoa and inver tebra tes

Benth ic Clams a lgae

Figure 10.3 Microbiotic and macrobiotic relationships in aquaculture. Flow of energy — ; Flow of manure - > ; A , Big Head carp; B, Silver carp; C, Gross carp; E, Common carp; F, Black carp. (After N R C , 1981)


Future prospects

An area of research with non-sterile cultures that have received little attention is the growth of thermophilic actinomycetes under low moisture-high solids conditions (Bellamy, 1974; 1979; NRC, 1979). Under aerobic conditions, at a temperature of 55-65 °C, water content of 40-60% and pH 7.0-8.0, the thermophihc actinomycetes grow so rapidly that other microorganisms are virtually excluded. The moisture content is too low for rapid growth of bacteria and the fungi grow much more slowly than the thermophilic actinomycetes.

The control of temperature 55-65 °C and of moisture (25-50% solids) can be by sophisticated mechanical and electronic equipment or by labor-intensive manual methods. An operator can learn to judge the temperature by touch and the moisture content by physical consistency. Because of the wide tolerance of these organisms to temperature and moisture, the system is well suited for use as a labor-intensive, low technology process. Only the pH has to be controlled within the range of 7.0-8.5 by the addition of alkah, ammonia or buffer.

The actinomycetes have not been found to produce mycotoxins as do many of the fungi, therefore, consumption does not present the hazard that the consumption of uncontrolled fungi does. Actinomycetes are, however, a major cause of farmers' lung disease. Farmers' lung is an allergic disease from inhalation of spores from moldy hay or straw; it is not an infectious disease (Blyth, 1973). Cellulolytic, thermophilic actinomycetes can be grown on biomass under low capital, low technology conditions by controlHng the temperature, pH and moisture within the above described range. Two modifications will greatly improve the feeding quality of the product. Selection of asporogenous strains would greatly reduce the hazard of spore inhalation and subsequent allergic symptoms.

Actinomycetes have been found to be a major cause of the earthy odor in soils and the musty odor of damp earth. Geosmine, 1,10-dimethyl-trans-9-decalol, has been identified as the source of the earthy odor in some actinomycetes cultures (Gerber and Lechevalier, 1965; Gerber, 1971). Mucidone, C 1 2 H 1 8 O 2 , imparts a musty odor to water and soils (Dougherty, Campbell and Morris, 1966). Selection of thermophilic and cellulolytic

These systems represent complete biomass,recycle systems. There is little need for pretreatment of the biomass; the waste from chickens, pigs, or humans along with added straw or other fibrous or wood plants are converted directly into a human food—fish or crayfish or other crustaceans (NRC, 1981). The proper mix of fish has been found important. The fish feed on the waste straw and on the micro-organisms that are digesting the cellulose and lignin in the straw. In such a system the fiber particles may pass through the chain several times. On each pass, a fraction of the cellulose and lignin is degraded. While these are effective in conversion of plant biomass and wastes into human food, they can also be sources of human and animal diseases. More thorough knowledge of the biochemistry, the microbiology and the entire biotic interaction is necessary for maximum yield and maximum safety from these ancient processes.

W. D. Bellamy 149

strains or mutants that do not produce these unpleasant odors will be necessary.

The application of genetic engineering or recombinant DNA technology to waste recycling could have dramatic effects on pure-culture applications but not on non-sterile mixed cultures. It seems probable that the engineered or modified micro-organisms will not be able to compete in the natural environment. Therefore, it will be necessary to grow them under sterile conditions where their unique properties can be utilized. Several examples will illustrate the types of improvements that may find widespread apphcation.

The availability of low cost, highly active cellulases could find application in both human food and animal feeds. By increasing the available carbohydrates, many feeds that are now limited to ruminants could be digested by monogastric animals such as pigs and chickens. For human consumption, the digestibility and nutritive value of many fruits and vegetables could be improved by cellulase enzyme(s) pretreatment. There are several laboratories working on cellulase enhancement. The microorganisms reported are Trichoderma reesei, Cellulomonas sp., Clostridium thermocellum and Thermomonospora sp. There are at least three enzymes involved in cellulose degradation in all the micro-organisms studied in detail. Trichoderma reesei (formerly Γ. veredi) cellulases are the most thoroughly studied. Endoglucanase (carboxymethyl cellulase) [EC 3.2.1.4] is a 1,4 ß-glucan glucano hydrolase which causes random breaks in extended cellulose fibers. Exoglucanase (avicellase) [EC 3.2.1.91] 1,4 ß-glucan cellobiohydrolase, removes cellobiose from the non-reducing end of cellulose fibers. Cellobiase (ß-glucosidase) [EC 3.2.1.21] hydrolyzes the dimer, cellobiose to glucose.

In Thermomonospora, the exo- and endocellulase enzymes are quite heat stable while the ß-glucosidase is relatively unstable. In a cell free system, ß-glucosidase is the limiting enzyme because it is unstable and is less abundant in normally grown cultures {Figure 10.4) (Hagerdal, Ferchak and Pye, 1978; Ferchak, Hagerdal and Pye, 1980). A team at the University of British Columbia reported on cloning of the cellulose gene from Cellulomonas fimi in E. coli (Whittle etal.,l9%l). They reported that the enzyme(s) is not excreted during growth but could be recovered after cell rupture.

In addition to studies on cellulase cloning, studies are in progress to clone the cellobiase gene into ethanol producing micro-organisms such as Zymomonas mobilis (Dally, Stokes and Eveleigh, 1982). It is not apparent that an alcohol tolerant thermophilic micro-organism can be developed. None have been found in nature so far and the problems of protein stabihty at 55-70 °C in the presence of 7-12% ethanol may be insurmountable.

Lignin degradation appears to require mixed function oxidases as well as NADH or NADPH. It is not clear how much one could expect to increase the rate of hgnin digestion by cloning multiple copies of these enzymes. The rate may be limited by available energy or available sites on the surface of the lignocellulose complex.

The role of microbiological agents in the complex non-sterile ponds is understood in a general way. It is difficult to suggest methods for improvements of these systems until the functions of the individual agents


100

80 h

A 60

CH 40

20

• β glucosidase pH 6.6 55"C

I- Δ—Δ Avicel lase pH 6.6 eO '̂C

O—O CMCase pH 6.0-7.3 60°C

5 10 15 20 25 30 Time (h)

Figure 10.4 Stability of crude enzyme preparations from Thermomonospora (After Hager-dal et ai, 1980, reproduced with permission of the publishers)

have been defined. A thorough study of these systems should provide new knowledge of basic understanding of the microbiology involved. Better control of human and animal pathogens as well as increased yields should result.

Summary It appears that application of pure culture fermentation to waste processing will depend upon products in addition to SCP for animal feed. Genetic engineering will help in designing microbes to produce these specialty products. Controlled non-sterile fermentations will continue to be the most economical method for processing most wastes. Significant improvement in yields of foods and feed can be expected as the role of microbial agents becomes better understood and exploited in these systems.

References

B E L L A M Y , w .D. (1974). Biotech, Bioeng., 16, 869 B E L L A M Y , Ψ,Ό. (1979). Am. Soc. Micro. News, 45(6), 326

W. D. Bellamy 151

BLYTH, W. (1973). In Actinomycetales, Ed. by G. Sykes and F. Skinner, p. 261. Academic Press, NY

C A N T N E R O S , I., G O N Z O L E G , R., R O N C O S , A. and A S E R G I , J.A. (1982). Biotech. Lett., 4(1), 51

C H E N , S.L. and PEPPLER, H.J. (1978). In Developments in Industrial Microbiology, 19, 79. Ed. by L.A. Underkofler. SIM Publ., Washington, DC

C H A H A L , D . S . , S W A N , J.E. and M O O - Y O U N G , M. (1977). In Developments in Industrial Microbiology, 18, 433. Ed. by L. Underkofler. SIM Publ., Washington, DC

CLEMENT, G. (1975). In Single-Cell Protein II. Ed. by S.R. Tannenbaum and D.I.C. Wang. MIT Press, Cambridge, Mass.

C R A W F O R D , D.L . (1981). In Third Symposium on Biotechnology in Energy Production and Conversion, p. 275. Ed. by C D . Scott. InterScience, NY

D A L L Y , E .L . , STOKES, H.w. and E V E L E I G H , D . E . (1982). Biotech. Lett., 4(2), 91

D A U G H E R T Y , J . D . , CAMPBELL, R . D . and MORRIS , R.L. (1966). Science, 152, 1372

E I S E N B E R G , D . M . , K O U P M A N , B . , B E N E M A N N , S.R. and O S W A L D , W.J. (1981). In Third Symposium on Biotechnology in Energy Production and Conservation. Ed. by C D . Scott. John Wiley & Sons, NY

F Ä H N R I C H , P. and I R R G O N G , K. (1981). Biotech. Lett., 3(5), 201 F E R C H A K , J .D . , H A G E R D A L , B. and PYE, E.K. (1980). Biotech. Bioeng., 22,

1527 G O N Z A L E S - V A L D E R , Α . and M O O - Y O U N G , Μ. (1980). Biotech. Lett., 22, 1515 G O N Z A L E S - V A L D E R , A . and M O O - Y O U N G , M. (1981). Biotech. Lett., 3(3),

149 G E R B E R , N.N. and L E C H E V A L I E R , H .A . (1965). Appl. Microbiol., 13, 935 G E R B E R , N.N. (1971). Tetrahedron Lett., 2971 G R A Y , W . D . (1970). The Use of Fungi as Food and in Food Processing. I

CRC Press, Cleveland, Ohio G R A Y , W . D . (1973). ibid. II. H A G E R D A L , B. , F E R C H A K , J .D. and PYE, E.K. (1980). Biotech. Bioeng., 22,

1515 H A H , Y.w. (1978). In Advances in Appl. Microbiol., 23, 119. Ed. by D.

Perlman. Academic Press, NY K A R G I , E. and S H U L E R , M.L. (1981). Biotech. Lett., 3(8), 409 KOSIKOWSKI, F.v. (1978). Cheese and Fermented Milk Foods. Edwards

Brothers, Inc., Ann Arbor, Mich. KOSIKOWSKI, F.v. (1979). J. Dairy Sci., 62(7), 1149 K U R T Z M A N , R.H. , Jr. (1979). In Annual Report on Fermentation Processes,

3, Ed. by D. Perlman. Academic Press, NY M O O - Y O U N G , M., M a c D O N A L D , D . G . and LING, A. (1981). Biotech. Lett., 3(7),

154 N A T I O N A L R E S E A R C H COUNCIL US (1979). Microbial Process: Promising

technologies for developing countries. NTIS Accession No. 80-144-696. National Academy Press, Washington, DC

N A T I O N A L R E S E A R C H COUNCIL US (1981). Food, fuel, and fertilizer from organic wastes. NTIS. National Academy Press, Washington, DC

NOZINIC, R. and D R A Z I C , M. (1982). Biotech. Lett., 4(2), 109


O S W A L D , W J . , LEE, E.W. , A D A M , B. and Y A O , K.H. (1978). WHO Chronicle, 32, 348

O T E R O , M.A. and C A B E L L O , A . (1980). Biotech. Lett., 2(9), 379 O T E R O , M . A . , G O N Z A L E Z , A . C . , B U E N O , G.E. and G A R C I A - R E V I L L A , J.L.

(1982). Biotech. Lett., 4(3), 149 R A J A G O P A L A N , K. and KOSIKOWSKI, F.V. (1982). Industrial & Eng. Chem.

Product R & D, 21(1), 82 R O C K W E L L , P.J. (1976). Single-Cell Proteins from Cellulose and Hydrocar

bons. Noyes Data Corp., Parkridge, NJ SCRIMSHAW, N.s. (1975). In Single-Cell Protein II, p. Ed. by S.R. Tannen

baum and D.I.C. Wang. MIT Press, Cambridge, Mass. SHETTY, K.J. and KINSELLA, J.E. (1978). Biotech. Bioeng., 21, 329 SINSKY, A.J. and T A N N E N B A U M , S.R. (1975). In Single-Cell Protein II. p. 158.

Ed. by S.R. Tannenbaum and D.I.C. Wang. MIT Press, Cambridge, Mass.

V I E S T U R E S , U . E . , APSITE, A . F . , L A U K O V I C S , J.J., OSE, V .P . , B E K E R S , M.J. and T E N G E R D Y , R.P. (1981). Biotech. Bioeng. Symposium, 11, 359

WHITTLE, D.J . , KILBORN, D . G . , W A R R E N , R.J. and MILLER, R .C . , Jr. (1982). Gene, 17, 139

11

Introduction

The very large amounts of organic wastes generated by the food industry constitute a double problem; firstly, they must be disposed of without causing pollution, secondly, considerable amounts of energy-rich resources are lost during their disposal. The problems are compHcated even more because many such wastes have noxious odours and do not decompose readily.

Any system whereby food wastes could be converted into a material of value for disposal on agricultural land or elsewhere, or for growth of plants, would have considerable economic potential. If, during the process, a form of animal protein suitable for animal feed could be produced, the economic attractiveness would be even greater.

A typical waste of this kind is that produced by the processed potato industry. The national potato crop in the UK is currently 6700000 tonnes, of which 475000 tonnes are used for potato crisps, 179000 tonnes are dehydrated and 545000 tonnes are frozen as chips, a total of about 1200000 tonnes. The waste from these processed potatoes is mainly in the form of solid potato peelings and large quantities of sludge. The solid portion is extremely malodorous and does not decompose readily, and the liquid can contaminate waterways. Potato waste disposal can be costly to the processor; for instance, if it is passed into rivers the water authority has to be paid to monitor the effluent so as to avoid the waters becoming over polluted. Alternatively, it is sprayed on to land at the processor's expense. The solids can be sold as pig food at about £l/tonne, but the cost of transport is prohibitive if pig farms are not close to the processing factory. Potato production in the UK is increasing annually (Figure 11.1) so the quantities of waste to be disposed of in the future are likely to increase accordingly.

The production of waste during the processing of potatoes is summarized in Figure 11.2.

Approximately 210000 tonnes of solid potato waste are produced per annum. This represents the raw material that can be broken down by earthworm activity into a much more useful soil additive, at the same time producing valuable protein for animal feed.

PRODUCTION OF EARTHWORM PROTEIN FOR ANIMAL FEED FROM POTATO WASTE

C A . EDWARDS Rothamsted Experimental Station, Harpenden, UK

153

154

AOOO CD C

g 3000 ^

2000

1000

74 75 76 77 78 79 60 81 82 83 8A 85 86 Y e a r

Figure 11.1 Potato production for human consumption in the U K

P e e l i n g l o s s 1 5 - 2 0 %

( M e a n 17.5°/o)

P o t a t o s o l i d s 210 000

P o t a t o e s 1 200 000

P e e l e d p o t a t o e s 990 000

P o t a t o s o l i d s 227 000

P o t a t o s l i c e s o r c h i p s 199 0 0 0

M o i s t u r e c o n t e n t 77 °/o

S l i c i n g l o s s 12.5°/o

S l u d g e ( 2 - 5 % s o l i d s )

28 000

Figure 11.2 Production of waste from processed potatoes (tonnes)

7 0 0 0

6000 i

5000

C. A, Edwards 155

Fraction Percentage composition

Protein 6 0 - 7 0 % Fat 7 -10% Carbohydrate 8 -20% Minerals 2 - 3 % Gross energy, kJ/kg 16750

It has a better essential amino acid spectrum than meat meal or fish meal and is rich in lysine, methionine and cysteine (Sabine, 1978; Edwards, 1983), large quantities of long-chain fatty acids that non-ruminant animals cannot synthesize, a good mineral content and many useful vitamins, particularly niacin and vitamin Β12 (Edwards, 1983). When assessed for animal nutritive value on a computer program it was found to have a greater monetary value than fish meal, meat meal or soya bean meal.

All the wastes investigated had a better structure after working by worms, an improved water-holding capacity and more available mineral nutrients (Edwards, 1982).

Quite early in the research programme, the potential of potato waste solids for growing worms was investigated and found to be excellent. This chapter aims to assess this potential further and give some idea of the economic feasibihty of growing worms on potato wastes.

Growth of Eisenia foetida in potato solids

In preliminary experiments, it was found that potato solids provided an acceptable habitat and food source for E. foetida and needed no preparation or modification. Their moisture content of about 77% is almost identical with that of E. foetida, so the worm is under no osmotic stress. Its open texture allows air to percolate so that it becomes aerobic with little ammonia or mineral salts, all conditions favourable to the growth of E. foetida. This differs from many animal wastes which are often unacceptable to the worm initially and need manipulation or pretreatment before worms

Eisenia foetida (Savigny), the tiger or brandhng worm is a common inhabitant of compost heaps and is reared commercially for fish bait. It has been used in the USA to break down activated sewage sludge. Its biology is understood better than that of most other species of earthworm (Watanabe and Tsukamoto, 1976; Hartenstein, Neuhauser and Kaplan, 1979).

A programme of research at Rothamsted Experimental Station was begun in 1980 with the aim of using E. foetida and other species of earthworms to break down various kinds of animal wastes (pig, cattle, horse, duck, chicken, turkey) and other organic wastes such as potato, paper, spent mushroom compost and brewery wastes. This project also aimed to use the earthworms produced as feed for fish, poultry and pigs.

E. foetida is an ideal food for such animals. Its overall composition is given in Table 11.1.

Table 11.1 O V E R A L L A N A L Y S I S OF E. FOETIDA ( D R Y M A T T E R )

156 Production of earthworm protein for animal feed from potato waste

STOCKING R A T E S OF WORMS

The growth of E. foetida at different stocking rates and temperatures was investigated in 125 ml crystallizing dishes. Into each dish, 40 g (w/w) of potato sohds was placed and 1, 2, 4, 8 or 16 worms were inoculated with two replicates for each stocking level. The worms inoculated were young hatchlings weighing approximately 0.05 g each. Similar batches of dishes were kept at 15, 20, 25 and 30°C. At regular intervals, the dishes were emptied out and the worms washed and weighed. If worms were dead, then equivalent numbers of similarly-sized live worms were added to maintain the stocking rate. Such deaths were rare except at the highest temperatures. The experiment continued until all worms began to lose weight, indicating that the nutritional value of the waste was beginning to decrease.

The resuhs of this experiment are summarized in Figures 11.3-11.8, in terms of average weight per worm and of total weight of worms in each crystallizing dish. Clearly, the worms increased in weight individually more rapidly when they were fewer in numbers. However, ttíe overall productivity, in terms of total weight of worms in the whole 40 g of waste, was greatest at the highest stocking rates {Figures 11.3, 11.5 and 11.7) used in the experiment. It is feasible that even higher stocking rates are possible.

0.6h

0.4h

c

s 0.2h

0 4 I ' ' ' Ί M a r c h A p r i l M a y

Figure 11.3 Potato waste. Average weight per worm at 15°C

0.6h

I c σ d) -Σ

0 .4h

0.2h

Ί ' ' ' Ί M a r c h A p r i l M a y


can be grown in them. If the waste is kept covered, there is Httle change in moisture content.

157

χ 16

M a r c h A p r i l M a y


3.0h

2.0h

o -Μ

Ρ 1 . 0

' • ' Μ ' ' ' Ί M a r c h A p r i l M a y

Figure 11.6 Potato waste. Total weight of worms at 15°C

x 1 6

M a r c h A p r i l M a y


Ί ' ' ' Ί M a r c h A p r i l M a y


158

^ ^ Figure 11.9 Effects of temperature March 'April' ' ' ' May ' ' ' growth of E. foetida in potato waste

3 . 0 Η

20 X

25 X

1.0

Figure 11.10 Effects of temperature on March' "Afiril' 'May' growth of E. foetida in potato waste

C. Α. Edwards 159

Temperature Conversion percentage (°C) (waste/worm weights)

15 4.50 20 4.82 25 4.94 30 erratic^

^There was variable mortality at this temperature which made an exact conversion ratio difficult to calculate.

of worm and waste are almost identical, on the wet weights. The maximum conversion ratios were all obtained at the highest stocking rate and were as shown in Table 11.2.

These calculations were made on a limited range of stocking rates and on a relatively small laboratory scale. In our experience, larger scale breeding systems often produce comparatively higher conversion ratios. Moreover, manipulation of the wastes by additives such as straw and wood shavings or

E F F E C T O F T E M P E R A T U R E

When the growth of individual worms {Figure 11.9) and of total biomass, at the maximum stocking rate of 16 worms in 40 g of waste {Figure 11.10) was compared at 15, 20, 25 and 30°C it was clear that the ambient temperature had a considerable effect on the rate of growth of the worms. Individual worms put on weight much more rapidly at 20 °C than at 15 °C and there was a further increase in rate of weight gain at 25 °C. However, at 30 °C there was considerable mortality; moreover the rate of increase in weight was less than at 25 °C. It would seem to be impracticable to grow worms in potato waste at temperatures higher than 25 °C. There was only a relatively small increase in rate of growth from 20 to 25 °C, and in view of the cost of maintaining growing systems at the higher temperature, it seems that the optimum temperature should be about 20 °C.

T I M E T O R E A C H M A T U R I T Y

E. foetida took 61 days to reach maturity at 25 °C, 65 days at 20 °C and 74 days at 15 °C; this is a little longer than on most animal wastes. Cocoon production was not monitored in detail, but seemed very prolific. From studies on other wastes, we found E. foetida to be capable of producing about 20 young per week, which gives it a very considerable potential for increase and rapid colonization of new waste.

C O N V E R S I O N R A T I O S

From the stocking rate experiment it was possible to calculate basic conversion rates of waste to earthworm protein. These could be made on the basis of dry weights of waste and worms or, since the moisture contents

Table 11.2 C O N V E R S I O N R A T I O S F O R W A S T E

T O E A R T H W O R M T I S S U E


Changes in potato waste caused by earthworm activity

E. foetida is believed to obtain most of its nutrition from micro-organisms that grow on organic wastes. These are mainly bacteria, but also include fungi and protozoa. Coincidentally, the worm fragments the waste with the aid of a grinding gizzard as the material passes through its intestine. This increases the surface area of the waste and further microbial inoculation occurs, so that the worm faeces contain considerably more microorganisms than the ingested waste (Bater, 1982). In this way the worm and micro-organisms interact with each other to break down the waste progressively. During this process, the mineral nutrient content and form changes (Tables 1L3 and 11.4).

Table 11.3 C H E M I C A L A N A L Y S E S OF W O R M - W O R K E D P O T A T O SOLID W A S T E

% Dry matter Waste Ρ Κ Ca Na Mg

With no worms 0 1 2 3 Í 905 0 7 5 8 0033 0.204 With worms 0.126 2.049 1.318 0.044 0.334

Table 11.4 C H A N G E S IN T H E FORM OF N I T R O G E N IN W O R M - W O R K E D P O T A T O SOLID W A S T E

Waste Sample Nitrogen in solution Total nitrogen no. (dry matter)

NH4 ( p p m ) NO3 (ppm) N H 4 (ppm) NO3 (ppm)

With no a 108 112 4392 4554 worms b 108 124 4392 5042 With a 9 520 320 18504 worms b 9 515 320 18326

The changes in mineral nutrients were not large but, in general, they were higher in the worm-worked potato waste with the exception of potassium. The most striking difference was in the form of nitrogen. Almost all of the nitrogen in the unworked waste was in the ammonium form whereas in the worm-worked waste it was nearly all as nitrate.

Methods of processing potato wastes with worms

The requirements of the worm is that the waste in which it lives is aerobic; this precludes depths of waste much greater than 30 cm. However, once the worms have processed such a depth of waste it is quite feasible that further layers of waste can be added progressively to a depth of 1 m or more. The worms will move gradually to each successive layer processing them as they go.

by microbial inoculation can increase the productivity of wastes and hence the conversion ratios.

C Α. Edwards 161

Temperature (°C)

1:800 Ratio: worm/waste (wet weight)

1:400 1:200 1:100 1:50 1:25

15 71 62 49 39 30 18 20 50 46 38 32 24 16 25 42 37 32 28 21 14

even more rapid in practice, when done on a larger scale. Probably the most convenient method would be to do the processing at the maximum stocking rates of 20 °C giving a breakdown time of two to three weeks. In practical terms, there are several ways of inoculating with worms. If containers were large, some worked waste carrying worms could be left at the bottom of the container before addition of a fresh batch of waste.

Separation of the worms from the worked waste has been achieved with a special separator designed at the National Institute for Agricultural Engineering. This consists of a series of rotating screens which not only separate waste and worms but also size-grade the worms.

Economics of using worms to break down potato waste Approximately 210000 tonnes of potato solid waste is produced annually in the UK. With a 5% waste to worm conversion ratio this could produce

Such systems of processing could be either in batch systems, i.e. in boxes or other containers which could be stacked on shelves, racks or even free-standing one upon the other. Alternatively, production could be in beds on the floor with bottom drainage but with a fine mesh base to prevent loss of worms. In either method it is preferable to keep a covering over the box or bed to minimize loss of water. The worm needs a temperature between 15-25 °C and in colder climates some form of heating to a relatively constant temperature may be necessary.

The time the worms take to process the waste can be manipulated by the number and age of worms added to the waste. If cocoons are used to inoculate waste the process will be slow with about three weeks needed for the cocoons to hatch and seven to eight weeks for the worms to reach sexual maturity and begin a new generation.

If the aim is merely to process the potato waste, the number of worms added is not important, it is more relevant to know the live weight of worms added, i.e. a large number of small worms are equally effective in processing a unit of waste as a smaller number of large worms with the same biomass. However, to obtain maximum harvestable worm tissue it is best to inoculate with at least half-grown worms because the rate of increase of weight of such worms with time is much greater than that of small worms.

It is possible to calculate the stocking rates for specific rates of breakdown of potato waste {Table 11.5) from graphs calculated from the laboratory data.

From larger scale breeding studies in potato waste, it seems that these figures tend to the conservative side and break down of waste could be

Table 11.5 T I M E S T O P R O C E S S W A S T E A T D I F F E R E N T S T O C K I N G R A T E S

( D A Y S )


Acknowledgements

Very considerable thanks are due to Dr M. Kirkman of Walkers Crisps for his help and advice with the project, and Miss Barbara Jones for her work on the stocking rates.

References

B A T E R , J. (1982). The effect of earthworm activity on microbial populations in organic waste materials. Thesis, Hatfield Polytechnic

E D W A R D S , C.A. (1982). Report of the Rothamsted Experimental Station, Part I, 103

E D W A R D S , C.A. (1983). Aquaculture (in press) H A R T E N S T E I N , R., N E U H A U S E R , E.F. and K A P L A N , D.L . (1979). Oecologia,

43, 329 S A B I N E , J. (1978). In Utilization of soil organisms in sludge management,

Ed. by R. Hartenstein, p. 122. SUNY, Syracuse, NY W A T A N A B E , H. and T S U K A M O T O , J. (1976). Revue d' Ecologie et Biologie du

Sol., 13(1), 141

potentially 10500 tonnes of worms which when dried, would give approximately 2500 tonnes of dried worm meal protein. Computer nutrition analyses have shown that worm protein is worth between £200 and £400/tonne, giving a value between £500000 and £1000000.

Additionally, the waste is turned into a material which has value in horticulture and might be worth up to £20/tonne. Assuming a loss of 20% in weight during the worm processing, 210000 tonnes of potato waste would produce 168000 tonnes of useful worked material and this could be worth:

160000 tonnes @ £20/tonne = £3200000

Clearly, this utilization of potato sohds could be an attractive commercial proposition.

12

Introduction

Although it is their content of essential amino acids which makes proteins obhgatory components of human diets, it is for the organoleptic qualities they possess, or which they confer upon other food systems, that consumers enjoy eating them. Of these quahties, texture is second only to flavour in determining satisfaction. It encompasses a large number of sensations detected by the palate and many parameters which can be determined objectively (Szczesniak and Torgesen, 1965).

The manipulation of cereal flours in making bread is an ancient example of an empirical attempt to improve the original form of a protein source in the interests of human consumers. Systematic attempts to add desirable texture to recovered or otherwise derived protein have been made for only 100 years or so (Gerrard, 1886). A patent was granted to Kellogg for the fabrication of organoleptically desirable products from wheat gluten and casein in 1907.

Since World War II, there has been greatly increased interest in the texturizing of proteins after their extraction from various sources. This has arisen against a background of heightened social awareness, of a believed worldwide lack of protein and of an appreciation that protein is synthesized much more efficiently within the Vegetable Kingdom, and by unicellular organisms, than within the Animal Kingdom. There is also increasing concern that much of the protein which is produced is wasted or underutilized because of its initially unaesthetic form. Thus, meat-hke analogues and extenders, have predominated among the forms which have been fabricated.

Texturization of recovered proteins has been principally achieved by one or other of three broad procedures—fibre spinning, thermoplastic extrusion or heat gelation. These will be considered in some detail.

In this chapter, however, the extensive use of recovered proteins in untexturized form for the benefits which their functional properties confer on food systems—foaming, water-binding, emulsifying—will not be reviewed.

163

TEXTURIZATION OF RECOVERED PROTEINS

R.A. LAWRIE and D.A. LEDWARD Department of Applied Biochemistry and Food Science, University of Nottingham, UK

164 Texturization of recovered proteins

Fibre-spinning

For hundreds of miUions of years spiders and silk worms have instinctively exploited the properties of proteins in forming long fibres. Man's efforts to do so have been very recent—and it is not surprising that these should have been made firstly in the textile industry. A patent was granted to Millar in 1898 for the production of artificial silk by forcing concentrated solutions of the proteins of egg albumen and blood serum, together with alkah or alkaline earth phosphates, into air, when the resultant fine stream of protein dried as fibres as they passed over rollers. It was not until after World War II, however, that such procedures were employed on a large scale to upgrade proteins from relatively inexpensive, originally indigestible or unaesthetic sources, by fabricating them into meat-hke products of superior consumer appeal.

In what is now regarded as standard technique, Boyer (1954) extracted proteins from defatted plant matrices (such as soya) using food grade alkali at pH 10-11, concentrated them to about 15% and held them at 40-50 °C until the viscosity of the 'spinning dope' had reached predetermined values, as the proteins unfolded from globular configurations into randomly coiled polypeptides. The dope was then forced through a metal spinneret (consisting of thousands of apertures of approximately 0.2 mm diameter) into a coagulating bath (containing salt and acid at pH 3-4.5), when the protein precipitated as long, thin filaments. In a continuous operation, the latter were gathered together longitudinally, over godet wheels, as a 'tow' of fibres. Since the wheels rotated at a faster rate than the filaments emerged from the spinneret, the latter were stretched and the polypeptide chains ahgned. If desired, fibres could then be passed through further baths in which they were bound together by added fat, egg albumen or polysaccharides such as carrageenan and associated with flavour and colour.

The texture and appearance of fibres produced by the Boyer process can be altered by controlling the viscosity of the dope, the rate of flow of dope, the rate of removal of the precipitated fibres from the bath, the tension applied to the fibres and by the temperature, pH, concentration and nature of the bath constituents (Horan, 1974). Typically, the fibres contain 50-70% moisture and, on a dry weight basis, 60% protein, 20% fat, 17% carbohydrate and 3% ash (Smith and Circle, 1972).

When subjected to agents capable of disrupting the ionic, disulphide, hydrogen and hydrophobic bonds which determine their native secondary and tertiary structure, the abihty of proteins to unfold varies considerably, as does the speed with which they reaggregate thereafter. It is evident that differences in the overall amino acid composition, and in the sequence of amino acids along the polypeptide, affect their behaviour. Thus, those polypeptides which have relatively high contents of free hydroxyl and amino groups will tend to form hydrogen bonds readily and the latter determine the secondary and tertiary configurations the proteins will assume (Nagano, 1974).

For effective fibre spinning it has been suggested that molecules should be at least 100 nm in length and have a molecular weight (MW) of 10000-50000 daltons. They should have no bulky side chains but have

R.A. Lawrie and D.A. Ledward 165

abundant polar groups and cystine residues (Lundgren, 1945,1949). Below 10000 daltons the fibres which form are weak and above 50000 daltons difficulties due to high viscosity occur. The degree of alignment determines intermolecular binding and thus fibre strength. During spinning shear forces disentangle the polypeptides, and flow through the spinneret enhances parallel alignment, thus favouring association into crystalline regions.

The Boyer process has been applied to the proteins extracted from many plant sources—soya, casein, cotton-seed, safflower, sesame, field beans, and wheat gluten (Horan, 1974); but it has also been used in efforts to upgrade protein recovered from unaesthetic animal sources, such as offal, when it has become evident that the source determines the procedural details which must be followed (Young and Lawrie, 1974b, 1975a). Thus, with blood plasma (partially freeze-dried to a protein concentration of 11%), the dope tends to gel and must be stabilized at a viscosity suitable for spinning (~ 250 poise, P) by first adding NaOH until the ratio of NaOH/protein is 1:10, allowing the mix to stand for 15 min and then reducing the pH to 11 by adding acetic acid (Young and Lawrie, 1974a,b). The dope can then be pumped (at a pressure of 150-200 kPa) into a bath containing 20% NaCl in IM acetic acid. Alternatively, if handling circumstances permit, the dope can be spun before it has had time to gel (Swingler and Lawrie, 1977). The protein fibres thus formed contain about 17% protein and 73% moisture. Their high ash content can be reduced to acceptable levels by water-washing (Swingler and Lawrie, 1977).

On the other hand, the behaviour of proteins extracted by alkaU from lung and intestinal tissues differs insofar as the exponential rise in viscosity (and gelation) does not develop (Young and Lawrie, 1975a). Fibres spun from the proteins isolated from lung, stomach and rumen also differ from those of blood plasma in being less elastic and more brittle. They all contain collagen, the amount of which depends on the pH, duration and temperature of extraction (Swingler and Lawrie, 1979). Boyer (1954) reported that, unless fibres were stretched during their formation, they were weak and lacked desirable organoleptic properties. However, Young and Lawrie (1975b) found that a higher velocity of the godet wheels (a greater take-away speed), which caused more stretching of the fibres, lowered the shear resistance significantly, this effect being particularly marked when the speed of pumping into the bath was low. Electron micrographs confirmed these impressions. Fibres of higher shear strength (lower take-away speed) showed a clearly defined structure in cross-section, whereas weak fibres revealed a much more random orientation of proteins.

Whereas fibres spun from plasma proteins had a relafively well-defined structure, those from lung protein showed Uttle specific orientation and those from stomach protein were intermediate in structural organization. These features were reflected in the shear strengths of the respective fibres {Table 12.1). Such findings are not unexpected. There are, of course, considerable differences in the pattern of proteins originally present in the various offal sources (Young and Lawrie, 1975c). Moreover, the solubilization by alkali, required with soHd offal, clearly altered the nature of the proteins, since a component of MW 70 000 daltons which was present in the


1975b)

Source Pump speed Take-up reel speed Mean shear resistance (rpm) (cm/min) (relative values)

Lung 2 245 0.22 2 580 0.20

10 245 0.12 10 580 0.13

Plasma 2 245 0.69 2 580 0.44

10 245 0.20 10 580 0.28

Stomach 2 245 0.30 2 580 0.27

10 245 0.13 10 580 0.15

protein dope, was not incorporated in the fibres which precipitated on acidification. Again, in respect of amino acid composition, plasma proteins contain noticeably higher percentages of lysine and threonine than those of other offal (Young and Lawrie, 1975a) and this may be significant in relation to the relative ease with which three-dimensional structures can be formed from them.

In spun fibres from all sources a considerable quantity of a protein component of MW around 130000 daltons was detected by gel electrophoresis. Evidently it was formed in the spinning process since it was absent from the corresponding protein isolates (Young and Lawrie, 1975c). Because the electrophoretic conditions employed incorporated sodium dodecyl sulphate and ß-mercaptoethanol, and thus any protein-protein associations involving electrostatic, hydrogen, hydrophobic or disulphide bonds should have been broken, some type of covalent linkage must arise in the spinning process. Certainly when alkaline conditions are extreme, racemization of amino acids, and both hydrolysis and synthesis of cross-links between polypeptide chains, take place (DeGrott and Slump, 1967).

The formation of unusual derivatives of proteins by alkali has been widely reported (Bohak, 1964; Asquith, Booth and Skinner, 1969). Thus, it is evident that when the temperature of alkahne extraction is relatively high (—60 °C) and the time is prolonged (~8 h), appreciable quantities of lysinoalanine are formed from lysine and cysteine (Swingler and Lawrie, 1979).

Because of the possibility of lysinoalanine formation (whatever its nutritional significance) when employing alkah, alternative means for protein extraction have been sought. Thus Tombs (1975) used a process whereby proteins, isolated from various sources, were extracted in water with added sodium chloride. In this medium the proteins formed a mesophase (a colloidal suspension in equilibrium with a true solution) and, when extruded into water, set as filaments. The process, however, requires undenatured proteins and a low level of insoluble material. Sodium dodecyl sulphate is another well-known solubilizing agent which can

Table 12.1 S H E A R S T R E N G T H FOR FIBRES SPUN FROM PROTEINS I S O L A T E D FROM V A R I O U S OFFAL S O U R C E S (AFTER Y O U N G A N D L A W R I E ,


Effect of lipid

Lipid-protein interactions evidently influence the nature of extracted proteins even before controlled texturization is attempted. Thus, the percentage of residual lipid (and its chemical nature) in protein isolates, as determined by the origin of the raw material (e.g. rumen, small intestine, lung), the mode of protein extraction employed (e.g. alkah at pH 10-10.5, sodium dodecyl sulphate (SDS)-FeCl3-acetone) and the solvent used for lipid extraction (when this is subsequently carried out), affect the degree of folding of the protein chains.

The amount of lipid extracted from protein isolates increases with increasing dielectric constant of the solvent. The level of residual lipid affects both affinity constant and monolayer values of the proteins {Table 12.2: Areas, 1981) and the capacity of extracted proteins to recover

Table 12.2 L I P I D C O N T E N T S , M O N O L A Y E R V A L U E S A N D A F F I N I T Y

C O N S T A N T S F O R P R O T E I N P R E P A R A T I O N S B E F O R E A N D A F T E R L I P I D

E X T R A C T I O N ( A R E A S , 1981)

Source of protein Lipid Before After and (%) lipid extraction lipid extraction mode of extraction mode of extraction

(mo) (C) (mo) (C)

R u m e n - S D S 11.35 0.054 15.33 0.055 8.23 Rumen-alkal i pH 10 19.19 0.057 13.39 0.062 8.25 L u n g - S D S 2.62 0.059 17.56 0.058 9.30 Lung-alkali pH 10 6.32 0.064 16.57 0.071 8.10

mo = monolayer value (g H20 ) / g protein C = affinity constant (dimensionless)

partially after thermal treatment decreases as the percentage of residual lipid falls (Areas, 1981). Again the functionality of alkali-extracted proteins, when subsequently precipitated at pH 5 (the region of the isoelectric point), is superior to that of those precipitated at pH 4 or pH 6 (Mittal, 1981).

This surprising finding also can be related to the greater residual lipid content of the former which, by lowering the degree of protein folding, would limit intramolecular interactions, leaving free more functional groups, despite the electrically neutral state of the proteins in this pH range.

It has also been demonstrated that admixture of charged polysaccharides in the spinning dope, enhances fibre-spinning abihty (Giddey, 1960 ; Imeson, Ledward and Mitchell, 1979). For coagulation a pH range of 1-3 is required when the polysaccharide is carrageenan (which has strongly acidic

extract protein even more effectively than alkali (Putnam, 1948; Gault and Lawrie, 1980). In this case, however it is necessary to remove the detergent from the extracted protein. This can be done effectively using precipitation by F e C l 3 but the recovered protein thereafter has a high content of residual iron (Elhson, Gault and Lawrie, 1980).


CaCh Warner-Bratzler^ Moisture Protein Alginate Ash (%) shear strength (%) (%) (%) (%)

(N/cm^)

5 4 2 . 2 ± 7 . 4 89.15 4.43 5.14 1.28 4 7 . 1 ± 1 . 4 93.92 2.61 2.69 0.78 3 9 . 0 ± 1 . 3 95.41 1.75 2.18 0.66 1.5 9 . 2 ± 1 . 0 95.19 2.99 1.16 0.66

^Mean ± standard deviation of eight measurements from two tows

sulphate groups) but a bath of calcium chloride solution is required for alginates and pect ates. The composition and rheological properties of such fibres can be varied by adjustment of any one of several parameters including the spinning dope composition, the pH and ionic strength of the coagulating bath, and the chemical composition of the polysaccharide (Imeson, Ledward and Mitchell, 1979).

Imeson, Ledward and Mitchell (1979) found that, when a solution of blood plasma (6%) and sodium alginate (2%) was extruded into unbuffered coagulating baths of calcium chloride (pH'-'8), there was a rapid increase in the shear strength of the fibre bundles with increasing salt concentration up to 3% calcium chloride. Above this level the strength of the fibres was independent of the calcium concentration. When dopes were extruded into 5% calcium chloride, fibre strength was independent of pH in the range 4-8. Below pH 4, however, the fibre bundles rapidly decreased in strength exhibiting a minimum value at pH 3.5. Decreasing the pH still further increased fibre strength once more. In all cases the fibres contained about 5-6% alginate and, above pH 3, 5-6% protein. Below pH 3 the protein content increased significantly being about 15% at pH 2. The rheological properties of the spun alginate-blood plasma fibres were found to vary in a complex manner with the guluronic acid block content and the molecular weight (viscosity) of the alginate (Imeson, Mitchell and Ledward, 1980).

It has been suggested that, at neutral pH values, the protein is merely trapped within the calcium alginate filaments since it can be easily washed out from the fibre (Imeson, Ledward and Mitchell, 1979). At lower pH values, however, the carboxylate groups of the alginate will tend to exist in the undissociated form (since the pK values of these groups are ~ 4 ; Haug, 1961). Consequently, extrusion into a bath of about this pH would be expected to involve the formation of fibres containing a high proportion of alginic acid filaments, which are of lower strength. As the pH is reduced still further, however, the precipitating conditions become similar to those employed in conventional protein fibre production and thus the acid denatured proteins will precipitate irrespective of the presence of the alginate. The protein fibres could then coexist with the calcium alginate and alginic acid filaments to give bundles of increased strength. Thus, at pH 2, there is almost complete recovery of both the alginate and protein (Imeson, Ledward and Mitchell, 1979). In addition, at low pH, there is a distinct possibility of electrostatic protein-polysaccharide interactions

Table 12.3 T H E EFFECT O F C O A G U L A T I N G B A T H CaCls C O N C E N T R A T I O N A T pH 3.5 O N T H E PROPERTIES OF FIBRES P R E P A R E D F R O M P L A S M A ( 6 . 0 % ) - A L G I N A T E (2.0%) MIXTURES (SMITH, 1982)


Soaking solution pH % protein loss

Distilled water 3.70 36.8 1 % S D S 3.75 34.0 1 % ß-mercaptoethanol 3.80 39.6 5% NaCl 3.80 47.2 2 % CaCl2 3.80 42.7

"Protein was the soluble hydrolysate from beef lung obtained after 60 min hydrolysis with 180°C

The co-spinning of protein with polysaccharides has also been applied to milk proteins. The texturization of casein as cheese, is of course, an ancient process but casein is a suitable protein for spinning, forming fibres when a neutral casein dope is spun into a coagulating bath containing lactic acid and salt. Casein fibres are unstable on rehydration, however, unless the dope also contains certain polysaccharides such as sodium alginate which co-precipitate with the casein when a calcium salt is present in the acid bath (Burgess, 1980).

An alternative procedure for stabilizing casein fibre is to exploit directly the sensitivity of casein to precipitation by calcium ions by spinning the dope into a solution of hot calcium salts. Calcium-coagulated casein fibres are more robust than those coagulated by acid.

As well as allowing an extensive range of textures to be generated, the incorporation of charged polysaccharides (especially alginate), into the spinning dope does have some further advantages. The dope is stable and of a suitable viscosity for spinning without the need to adjust the pH to

occurring (Imeson, Ledward and Mitchell, 1977; Imeson et al., 1978) and, although these may be insignificant at high ionic strengths, they may help to explain the increased strength (and protein content) of fibres produced by spinning plasma-alginate mixtures into baths containing 1.5% calcium chloride at pH 3.5 compared to spinning into baths containing 3% or 4% calcium chloride at this pH (Table 12.3).

Further evidence, indicating that protein-polysaccharide interactions may be of importance in alginate-protein fibres produced at pH 3-4, has been obtained by Knight (1981). Hydrolysates, prepared by enzymic treatment of beef lung, were mixed with alginate, calcium orthophosphate (CaHP04) and glucono-6-lactone. As the lactone hydrolysed, the pH decreased, calcium ions went into solution and a gel formed, the resultant pH being in the range 3.6-4.5. It was found that molecular size (degree of hydrolysis) had no significant effect on the amount of protein retained in the gel. Washing the fibres with solutions of sodium dodecyl sulphate or of ß-mercaptoethanol removed about the same amount of protein as a distilled water wash (Table 12.4). However, immersion in sodium chloride or calcium chloride caused a significant increase in the amount of protein leached out of the gel (Table 12.4) and a possible explanation of this effect is that the high ionic strength disrupts electrostatic linkages between the negatively charged polysaccharide and the positively charged protein.

Table 12.4 EFFECT O F V A R I O U S S O A K I N G SOLUTIONS O N T H E P E R C E N T PROTEIN" L E A C H E D O U T OF A L G I N A T E - C A L C I U M PROTEIN GELS A T pH 3 .7 -3 .8 (AFTER KNIGHT, 1981)


Extrusion processing

Of the techniques currently employed to texturize proteins, extrusion processing is the most popular as, in comparison with spinning, it requires less equipment and less sophisticated technology (Kinsella, 1978). This process, which was initially applied about 50 years ago to the production of shaped pasta products and ready-to-eat breakfast cereals (Rossen and Miller, 1973) was subsequently used to texturize vegetable proteins (Atkinson, 1970). It is now a widely practised technology, accounting for a significant fraction of fabricated proteinaceous foods (Harper, 1979). Since several comprehensive reviews have been written on the principles of, and equipment available for, the extrusion processing of proteinaceous material (Harper, 1979; 1981), these aspects will not be dealt with in detail in this chapter.

extreme values, the fibres so produced are of low ( -1%) ash content and, in addition, the protein used does not necessarily need to possess good functional properties and thus it may be a possible means of texturizing protein of low functionality (including those of hydrolysates). A possible disadvantage of using the alginate spinning system is that the dope must be of a reasonably low calcium content, otherwise gelUng may take place prior to spinning. Thus, whey uhrafiltrates may be spun into fibres using the alginate system but the protein must first be extensively dialysed to decrease the calcium content to a suitable level (Smith, 1982).

Although fibres produced by the alginate spinning process would need further development before they were acceptable as meat analogues, as meat extenders their value is already established. Rusig (1979) found that substitution of 40% of the meat protein of sausages by plasma caused a marked decrease in the acceptability of the cooked product. However, if this level was added in the form of alginate-plasma fibres no significant difference was detected between sausages containing no plasma and those deriving 40% of their protein from plasma.

As Millar showed in 1898, fibres can be produced from a protein dope by dry-spinning. Dry-spinning a casein dope into hot air improves the heat stability of the fibres formed in comparison with other spinning procedures (Burgess, 1980). An alternative dry spinning process (Visser et α/., 1980) involves preparing an aqueous mixture (pH 5.0-6.6) of casein and a heat-set able protein (such as soy protein) containing, per g of casein, 0.1 mmol calcium ions and 0.04 mmol orthophosphate. The mixture is spun at temperature below the gelHng temperature (40-70 °C), into a gaseous medium and the fibres dried.

Apart from the development of processes which yield a continuous flow of uniform, spun fibres, such fibres have been subjected to many subsequent texturizing operations. For example a patent was granted (Hartman, 1967) for the manufacture of a product which involved heat-binding alternating layers of red or white coloured soya protein fibres (together with added flavour and oil). The product could then be sUced and smoked to simulate bacon.


In extrusion processing the proteinaceous material (usually defatted soy flour) is placed in a conditioning chamber where it is moistened with steam or water to a moisture content of 15-40%. Additives such as salt, polysaccharides, colorants and flavours may also be added at this stage. The mixed homogeneous ingredients are fed through a feeder/hopper into the hollow barrel of the extruder where a tapered screw with whorled ridges forces the material towards the exit orifice of the barrel. The temperature of the barrel may be controlled. The heights of the ridges on the screw decrease towards the exit and this, together with the decreasing clearance between the flights and the inner barrel surface, causes high shearing as the material is moved along. As the clearance diminishes, the temperature and internal pressure also rise (to 120-175 °C and 2.8-4.2MPa respectively) converting the ingredients into a plastic viscous state in the metering section of the extruder. Under these conditions starchy components gelatinize, proteins denature and the tractile components are restructured and/or aligned. The shearing action of the rotating flights tends to align the denaturing proteins into parallel sheaths. The usual residence time in the extruder is 30-60 s after which time the molten mass is further aligned within the die prior to being squirted out into the atmosphere. With the instantaneous release of the high pressure, the superheated water within the structured protein 'flashes' off leaving an expanded porous structure. This evaporation causes rapid cooling and consequent thermosetting of the product to yield a puffed, fibrous structure which may be cut into strips or chunks or broken down into powder or granules. The moisture content is about 20% initially but it is usually reduced to about 8% prior to packaging.

Although textured protein products have been primarily made from soya; other sources such as cotton-seed, peanut, corn, wheat, sesame and yeast proteins have also been successfully employed. More recently admixtures of soya grits with several defatted, extracted offal proteins have been successfully extruded (Mittal, 1981) as also has defatted, dehydrated pork rind (Fox, 1981). Due to slipping (increased lubrication), textured thermoplastically extruded products can normally only be manufactured from protein sources containing 1% or less of fat.

The texture, density, chewiness, rehy drat ability, fat absorpdon and colour of extruded products can be influenced not only by the nature of the ingredients but also by moisture content, temperature profile within the extruder, pressure generated, shear rates (screw speed), residence time in the extruder, the type and configuration of the extruder, shape and size of the die, post die coohng and post-extrusion treatments. Many of these factors are interrelated and the precise effect of different variables in determining the texture of thermoplastically extruded products will also depend on the idiosyncracies of the individual machine. The importance of various parameters can thus best be illustrated by referring to data obtained from a given extruder. Mittal (1981) employed a Brabender 20 DN laboratory extruder to texturize soya grits, and mixtures of the latter with protein from abattoir offal. In all cases it was found that with a constant screw speed of 250 rpm and exit die diameter fixed at 4 mm, a feed moisture of 37.5% (dry sohds basis) and an extruder temperature (second barrel and die sections) of 170-180 °C yielded products which had


(1) Feed moisture^ {% dry solids basis)

35.0 1.00 1.80 3.2 22 13 37.5 1.45 2.60 7.0 47 27 40.0 1.10 1.70 6.9 43 25 42.5 0.70 2.30 3.8 24 15

(2) Barrel and die^ Temperature

150 °C 1.20 1.50 5.9 48 17 160 °C 1.30 2.60 8.1 50 20 170 °C 1.70 4.00 10.2 72 38 180 °C 1.68 3.50 7.8 47 27 190 °C 1.62 3.10 6.9 32 17

^@ 190°C barrel/die temperature ''@ 37.5% dry solids basis moisture in feed ^Chewiness = hardness x cohesiveness x elasticity, derived from profile analysis on Instron Universal Texturometer

the best laminar structure (as assessed by expansion ratio and hydratability) and the firmest texture (as assessed by shear-force, hardness and chewiness) {Table 12.5).

The texture of thermoplastic extruded products depends on a combination of starch gelatinization and cross-hnking of denatured proteins. The successful application of the process to abattoir offal, therefore, was found to depend on the presence of a carrier such as soya grits. In addition, the hpid content of the offal needed to be decreased. Thus, protein extracted by sodium dodecyl sulphate (SDS), a procedure which involves acetone treatment, from lung or small intestine had little residual fat and could be extruded directly but untreated offal, or protein isolated therefrom by alkali, required to be extracted by acetone as there still remained about 4% lipid in untreated small intestine and in protein isolated from the latter by alkah.

It was found (Mittal, 1981) that defatted, alkali-extracted offal protein could be incorporated in mixes with soya grits up to 65% and still produce a highly expanded, texturized product. At incorporation levels above 35%, however, products containing porcine or bovine lung protein, whilst having a good external appearance, contained internal channels. Products containing SDS-extracted proteins possessed httle or no texture, the extrusion process seeming only to shape the mix and not to introduce fibrous structure.

Typical data for the incorporation of untreated bovine small intestine, and of the alkali- and SDS-extracted proteins from this source, at the 35% level in a mix with soya grits, are given in Table 12.6. Products containing alkali-extracted protein expanded considerably more than those containing either untreated offal or SDS-extracted protein. Moreover the peak

Table 12.5 RELATIONSHIP B E T W E E N (1) F E E D MOISTURE A N D (2) B A R R E L / D I E T E M P E R A T U R E A N D V A R I O U S T E X T U f t A L P A R A M E T E R S OF S O Y A GRITS E X T R U D E D IN A B R A B E N D E R 20 D N A T SCREW SP EED 250 rpm A N D 4 mm D I E D I A M E T E R (AFTER MITTAL, 1981)

Expansion ratio Hydratability Shear force Hardness Chewiness^ (dia. product:) (g HzO/g sample) (N) (N) (dia. die)


Offal Barrel/ Expansion ratio Hydrat ability Shear force Hardness Chewiness^ die (dia. product: (gH^O/g (N) (N) temp. dia. die) sample)

Untreated 150 °C 1.10 0.95 7.6 50 13 160 °C 1.15 1.10 8.1 57 17 170 °C 1.22 1.50 8.2 58 21 180 °C 1.35 1.95 8.8 57 24 190 °C 1.18 1.70 8.6 50 18

Alkali- 150 °C 1.25 — 3.4 22 14 extracted 160 °C 1.27 3.15 3.8 21 11 protein 170 °C 1.51 3.10 3.8 20 9

180 °C 1.14 2.60 3.4 12 8 190 °C 1.13 2.20 3.0 12 6

SDS- 150 °C 1.18 1.15 11.5 108 44 extracted 160 °C — — — — — protein 170 °C 1.19 1.10 13.6 128 55

180 °C 1.15 1.15 16.2 84 40 190 °C — — — — —

"@ 37.5% dry solids basis moisture in feed Chewiness = hardness x cohesiveness x elasticity, derived from profile analysis on Instron

Universal Texturometer

expansion occurred at a barrel/die temperature of 170 °C with protein preparations extracted by either alkah or SDS but at a temperature of 180 °C with the untreated offal.

In respect of water-absorbing ability, products containing alkah-extracted protein were also superior, the maximum benefit again being when the extrusion temperature was 170 °C. The water-absorption ability of products containing SDS-extracted proteins was very poor at all extrusion temperatures.

Products containing alkali-extracted proteins required least force to shear, were least hard and had the lowest values for chewiness whereas those containing SDS-extracted proteins had the highest values for these parameters. Moreover, although products containing either SDS-extracted proteins or untreated offal showed maxima for these parameters with an extrusion temperature of 170-180 °C, those containing alkali-extracted proteins exhibited a continuous decrease in hardness and chewiness as the extrusion temperature rose from 150-190 °C.

The increase in shear strength with increasing processing temperature observed by Mittal {Table 12.6) in mixtures of soya grits with untreated offal protein and soya grits with SDS-extracted protein, agrees with previously reported work (Cummings, Stanley and de Man, 1972; de Man, 1976; Maurice, Burgess and Stanley, 1976) on soya grits alone. However the tensile strength of the fibres from soya grits alone was maximal after processing at 150-170 °C. If the hardness and chewiness parameters measured by Mittal (1981) are related to the tensile properties, then the behaviour of both the soya/untreated offal and soya/SDS extracted offal mixtures fits in with this pattern {Table 12.5). These results suggest that.

Table 12.6 RELATIONSHIP B E T W E E N B A R R E L / D I E T E M P E R A T U R E A N D V A R I O U S T E X T U R A L P A R A M E T E R S OF PREPARATIONS" C O N T A I N I N G 35% B A S I C OFFAL (SMALL INTESTINE)/65% S O Y A GRITS E X T R U D E D IN A B R A B E N D E R 20 D N A T SCREW SPEED 250 rpm A N D 4 mm D I A M E T E R (AFTER MITTAL, 1981)


with increasing temperature from 150-190 °C, increased orientation and fibre formation occurs leading to increased structural integrity (shear force) but that at the higher temperatures (>170°C), excess Assuring takes place leading to loss of tensile strength (and hardness and chewiness). For reasons that are not obvious, the alkali-extracted proteins drastically modify this behaviour {Table 12.6), so that maxiumu structural integrity is developed at a barrel/die temperature of 160-170 °C.

Although studies similar to those outlined above have been described for several protein sources little is known of the physical and chemical changes taking place in these concentrated systems at high temperatures and pressures.

Aguilera, Kosikowski and Hood (1976), using electron microscopy, studied the physical changes taking place in soya grits moistened to 25% and extruded in a Wenger (R) X-5 extruder with a barrel temperature of 145 °C and a die temperature of 120 °C. Their results show that mixing occurs during most of the length of the screw until the elevated temperature at the end is reached. Here the disrupted protein cells were seen to orientate into fibres and, following pressure release at the die, new surfaces were created and fibres formed.

With respect to the chemical changes taking place during extrusion processing, most workers have concentrated on those involving the proteins and have attributed the characteristics of the extruded fibres to the breaking of existing protein-protein linkages and the formation of new interprotein bonds. The resultant fibre is believed to consist of a protein matrix within which the carbohydrate and other non-protein material is embedded (Harper, 1981). Thermal denaturation of the proteins is obviously of paramount importance in the texturization process as this results in extensive unfolding of the native structure, allowing the protein to align in the shear field until reactive sites come into juxtaposition to form stable intermolecular bonds. The number and type of such bonds, at least to some extent, govern the physical characteristics of the fibre. This explanation of the texturization process will obviously explain why admixtures of proteins, e.g. soya and offal, extrude differently to soya alone. It is not unexpected that proteins from the same source, which have been extracted under differing denaturing conditions (SDS and alkali), should exert differing effects on the extrusion behaviour of the mix {Table 12.6).

Under the conditions of extrusion, it is likely that hydrogen, ionic and disulphide bonds will break, as also may a few of the more labile (non-disulphide) covalent bonds. Subsequently, at the elevated temperatures, hydrophobic interactions may become significant and a few specific covalent linkages form. On cooling, further additional Unkages may form. There is still some controversy about the nature of the crucial protein-protein bonds, formed during extrusion and cooling, which maintain the structure of the extrudate on rehydration in hot water. Evidence in the Hterature suggests that disulphide (Jenkins, 1970) and isopeptide linkages (formed by reaction between a carboxylic acid group and amino residue : Burgess and Stanley, 1976) are important and recent studies on the solubility of soya extrudates in a range of solvents lend qualified support to the role of these linkages in stabilizing the textured material {Table 12.7).

It is seen from Table 12.7 that, in comparison with distilled water.


Solvent Percent solubilization (Smith, 1982) (Burgess and Stanley, 1976)

Cold water (20 °C) 12.8 8.3 0.5M carbonate pH 10.6 18.8 23.6 0.1 Μ ß-mercaptoethanol 34.9 — Hot water (100 °C) 52.7 — O.IMSDS (pH 7) 81.8 13.9

explanation of this apparent discrepancy is that Burgess and Stanley subjected their protein mixes to excessive heating. They quote a product temperature of 178 °C (compared with the 160 °C used by Smith, 1982) and a slow screw speed (70 rpm compared to 250 rpm) which should lead to a longer residence time in the extruder. Under these conditions sufficient energy may be supphed to the system to permit extensive isopeptide bond formation. Obviously if this is so, the extrudate will remain largely insoluble even when all hydrophobic interactions are destroyed by SDS treatment.

Thus it may well be that the conditions of extrusion govern the number and types of linkages formed and these directly affect the textural hydration properties of the material.

In extrusion processing the chemistry is further complicated because the protein usually represents only about 50% of the ingredients used (the rest being carbohydrate, fibre and ash) and recent results suggest that the role of the polysaccharides is not passive (Smith, Mitchell and Ledward, 1982). In Table 12.8 the effects of including 1% of different polysaccharides on the extrusion of soya grits (moisture content 38% on dry solids basis) using a Brabender (Model 20 DN) extruder with barrel and die temperatures of 170 °C and screw speed of 250 rpm, are shown. It is seen from Table 12.8 that, even at the 1% level, alginate significantly lowers the expansion ratio, dough temperature and torque in comparison with both the control and

Table 12.8 EFFECT OF T H E INCLUSION O F C H A R G E D P O L Y S A C C H A R I D E S (1%) O N T H E E X T R U S I O N OF S O Y A GRITS (SMITH, 1982)

Control Alginate Carrageenan Polygalacturonic Carboxymethyl acid cellulose

Expansion ratio 1.44 1.14 1.54 1.26 1.28 Internal temp (°C) 163 156 163 163 162 Torque (Nm) 14.4 8.8 15.9 16.2 13.7

increased solubilization occurs in carbonate solutions of pH 10.6 (a pH at which isopeptide bonds will rupture), and in solutions of ß-mercaptoethanol (a reagent capable of breaking disulphide linkages). However, the results of Smith (1982) suggest that hydrogen bonds (which are broken in hot water and SDS) and hydrophobic bonds (which are broken by SDS) are primarily responsible for insolubilization of extruded proteins whilst the results of Burgess and Stanley (1976) suggest that isopeptide bonds are primarily responsible {Table 12.7). A possible

Table 12.7 P E R C E N T S O L U B I L I T Y O F T H E P R O T E I N O F E X T R U D E D S O Y A

G R I T S


Control Alginate Polygalacturonic acid

Moisture content (% dry solids basis) 42 38 42 Barrel temperature (°C) 190 215 190 Die temperature (°C) 130 140 130 Expansion ratio 2.10 2.14 2.08 Peak shear force (N) 251 230 216

Recent work has shown that alginates of high mannuronic/guluronic acid ratios are most effective in modifying the extrusion behaviour of soya grits {Table 12.10). Smith, Mitchell and Ledward (1982) have outhned mechanisms by which the alginate may modify the normal extrusion behaviour of the soya. Whatever the mechanism by which the alginate acts, the observations that only low levels cause marked differences and that these differences are dependent on the composition of the polysaccharide, may explain the reported variations in extrusion behaviour of apparently similar batches of soya grits, i.e. the differences may be due to differences in the carbohydrate component rather than the protein.

Table 12.10 D O U G H T E M P E R A T U R E , T O R Q U E A N D E X P A N S I O N R A T I O OF S O Y A E X T R U D A T E S CONTAINING 1% A L G I N A T E OF DIFFERING M A N N U R O N I C (M) TO G U L U R O N I C (G) A C I D RATIOS (SMITH, 1982)

Approx. M:G ratio Temp Torque Expansion (°C) (Nm) ratio

Control system (no added alginate) 160 14.5 6.7 0 . 7 : 1 153 13.5 6.3 0 .7 : 1 153 12.5 5.6 1.7: 1 149 14.5 4.9 1.9: 1 147 13.0 4.9 2 . 2 : 1 149 14.0 5.0

An advantage of extrusion processing is that, in many instances, the steam distillation occurring at the discharge of the extruder is an effective deodorizer, thus naturally occurring off-flavours are eliminated. For example, Mittal (1981) found that, following extrusion, highly odorous soya/offal protein mixtures were invariably bland in flavour and free of any objectionable odour. Because flavours are so volatile it has, to date, proved impossible to incorporate flavours into protein mixes prior to extrusion and obtain sufficient retention following processing. For this reason it is usual to flavour the textured material (following extrusion and coohng) but this requires unwieldy and costly external application.

other polysaccharides. The rheological properties of the doughs were also very different (Smith, Mitchell and Ledward, 1982). Using a Rosenbrock direct search procedure (Rosenbrock, 1960) Smith, Mitchell and Ledward (1982) were able to optimize the extrusion conditions for the soya alginate feed to give a product that was similar, with regard to both texture and expansion ratio, to the soya grits alone when these were extruded under optimum conditions {Table 12.9). However, as seen in Table 12.9, the optimum conditions were very different.

Table 12.9 OPTIMUM CONDITIONS F O U N D FOR E X T R U D A T E S WITH A N D W I T H O U T A D D I T I O N OF 1% P O L Y S A C C H A R I D E IN T H E MIX (SMITH, 1982)


Texturization by gel formation

Under appropriate conditions concentrated solutions of several proteins will form gels of suitable textures for use as foodstuffs. A well established, and commercially successful, apphcation of this principle is in the preparation of several meat products whereby mechanical and/or chemical treatment of meat pieces causes the muscle proteins to exude to the meat surfaces where, following heat processing, they effectively set and bind the chunks together into an appropriately shaped product. Myosin is the principal protein involved in binding (Turner, Mackenzie and Macfarlane, 1979). If large pieces of meat (~250g) are involved, vigorous mechanical working will cause sufficient exudation to cement the meat pieces together and, ahhough the binding at this stage is relatively weak, subsequent heat processing or cooking yields a high quahty product which maintains its integrity. This type of process has also been used for fish products. Thus Sugino (1979) described a system for moulding and mashing seasoned crab meat pieces whilst heating to yield a gelled product which could be subsequently shredded to resemble crab meat.

For smaller meat pieces it is necessary to tumble them in salt/phosphate solutions to extract sufficient salt soluble myosin to form, after pressing, a cohesive mass which, following cooking, is very similar in appearance, texture and sliceability to a single large piece of meat of the type from which the pieces were derived. To avoid the use of salt/phosphate solutions several workers have suggested that selected proteinaceous substances may be used as binders. Crude myosin, extracted from low grade meats, is a very effective binding (gelling) agent and may be very useful in reforming high quality meat products (Turner, Mackenzie and Macfarlane, 1979). Other proteins capable of forming gels on heat treatment may also be of use in reforming small meat pieces into large, integrated cuts and wheat gluten (Carlin, 1963) certainly appears to be capable of performing such a function. Siegel, Church and Schmidt (1979) showed that, of a whole range of non-meat proteins in the presence of 8% salt plus 2% phosphate, only wheat gluten, and to some extent egg white, gave improved binding to that observed with salt and phosphate alone. In the absence of salt and phosphate only wheat gluten, blood plasma, and to some extent isolated soya protein, gave any measurable binding.

As well as being used solely as a meat binder, several patents have been taken out, which utilize the gelling ability of gluten to create meat or fish-like analogues. For example, the Nippon Seifun Co. (1980) have a patent for the production of a meat-like product from gluten. In this process a paste of gluten and inactive yeast is mixed with seasoning, fats, emulsifiers and such like, and the mixture heated to create a textured product.

As well as wheat gluten, other non-meat proteins have been used to bind protein fibres to yield meat-like textures. Nisshin Mills, KK (1981) have recently patented a method involving the impregnation of the fibres with a binding solution of heat denatured soya bean protein and gelatin. Heat treatment of the impregnated fibres gives rise to binding in such a manner that the final product has the chewiness of meat.

An interesting gelling system that does not require heat to yield


Nutritional and microbiological aspects

Although secondary to the purpose of this presentation, it is desirable to mention the concomitant effects which texturization of recovered proteins have on the nutritive value and microbiological status of the products.

It might well have been supposed that the nutritive value of texturized products would be the same as that of the proteins from which they had been prepared. In certain circumstances, however, their biological value has proved superior to that of the parent material. In the case of fibres prepared from soya protein, such findings suggest that the spinning process may exclude or inactivate various of the antinutritional factors (e.g. goitrogens, haemolysins) which are naturally associated with this source (Kinsella, 1978). Again, fibres spun from various offal proteins were utilized better in rat feeding trials than the proteins from which they had been prepared (Swingler, Neale and Lawrie, 1978), suggesting the removal of substances responsible for the inhibition of digestive enzymes during

meat-like textures is that described by Tolstoguzov et al. (1978). In this system an anisotropic gel is formed during the breakdown of a two-phase system containing casein and a charged polysaccharide (alginate or gum arable). The gel produced has a fibrous consistency and high moulding capacity. The properties of the gel can be controlled by varying the potassium salt concentration.

Although several systems have been described involving the addition of functional protein to waste or low grade protein to yield textured products, most success appears to have been achieved utilizing the functionality of the recovered or waste protein itself.

One potentially successful product is a textured fish protein concentrate (Marinbeef) which has been developed in Japan (Suzuki, 1981). In a typical process the washed meat of the fish is mixed with 1-2% salt and the pH adjusted to 7.4-7.8 by the addition of sodium bicarbonate. At this salt concentration not all the muscle proteins are solubilized and thus those that are extracted serve to bind the remaining fibres together into a viscous paste. The paste is extruded as long, spaghetti-Uke strands into cooled ethanol (5-10 °C). After the ethanol has again been removed by centrifuga-tion the residue is dried to less than 10% moisture by hot air at 30-45 °C. The texture of the product (Marinbeef) can be adjusted by varying the salt concentration, the number of ethanol treatments and the length and temperature of such treatments.

The product formed by the above process has a good meat-like texture, good rehydration properties and excellent nutritional quality. Sensory evaluation showed that, when used as a partial meat replacement in hamburgers, meat loaves, meat balls and sauces, it was quite acceptable and was preferred to products containing meat flavoured soya bean extenders (Suzuki, 1981).

The possible disadvantages of Marinbeef production are that about 2100 kJ of energy are necessary to produce 1 kg of product and large amounts of ethanol are required. (Most of the ethanol can be recovered by appropriate distillation procedures.)


Conclusions

Although it may be argued that there is no real protein shortage in the world today, what is beyond dispute is that much of the available protein is in a form that is neither suitable nor acceptable for human consumption. To become acceptable the protein must be manipulated into a suitable

their preparation. Similarly, in extrusion processes, the high temperature/ short time treatment in the extruder barrel is generally sufficient to inactivate any antinutritional features present in the input material. Thus, in respect of the trypsin inhibitors present in soya protein. Harper (1981) believes that extrusion temperatures in excess of 138 °C inactivate 55-70% of these factors. Various enzymes detrimental to food quality are also denatured and inactivated during extrusion. Thus 90% of hpoxidase is destroyed (Harper, 1981).

Clearly, since extrusion processing involves the application of high temperature, it might be expected that vitamins would be inactivated and amino acids made unavailable. Because of the very short time of exposure, however, such damage is minimal. For example, there is little loss of riboflavin or vitamin A (Harper, 1981).

Numerous studies have also been performed on the quality of several proteins and protein mixtures, subjected to elevated temperatures similar to those attained during extrusion cooking. Not unnaturally, most of the reported work has been carried out on soya based products (Thompson, Wolf and Reineccius, 1976; Jokinen, Reineccius and Thompson, 1976; Wolf, Thompson and Reineccius, 1978), where the main reaction of interest is the loss of available lysine through non-enzymic browning reactions between the ε-amino group of the lysine and reducing sugars. The initial loss of available lysine follows first-order kinetics but because of the short time exposure at elevated temperatures the studies on the model systems (Thompson, Wolf and Reineccius, 1976; Wolf, Thompson and Reineccius, 1978) indicate that only minimal loss of protein quahty, due to browning reactions would occur during extrusion processing. The small amount of data on extruded products suggest that this is so (Mustakas et al., 1970; deMuelenaere and Buzzard, 1969).

Just as enzymes are extensively inactivated during normal extrusion processing, so the extrusion process also effectively destroys those microorganisms associated with food poisoning. Thus deMuelenaere and Buzzard (1969) found that Coliforms, Staphylococcus and Salmonella were completely destroyed during normal extrusion processing and that the total aerobic plate count was greatly reduced. The microbiological quality of extruded products is thus generally believed to be excellent (Harper, 1981).

In respect of fibre-spinning operations, the protein fibres formed are considerably more resistant to microbial attack than the parent material. The processes involved in isolation (salt and alkali) greatly reduce the level of psychrotrophs and mesophils (Swingler, Naylor and Lawrie, 1979) and subsequent fibre formation (including acid precipitation) reduces them still further. Such fibres have been observed to be stable for three years at 0°C.


References

A R E A S , J .A .G. (1981). Unpublished data A G U I L E R A , J.M., KOSIKOWSKI, F.W. and H O O D , L.F. (1976). / . Fd Sci., 41,

1209 A S Q U I T H , R.S . , B O O T H , A.K. and SKINNER, J .D. (1969). Biochim. biophys.

Acta, 181, 164 A T K I N S O N , w . T . (1970). US Patent No. 3488770 B O H A K , z . (1964). / . bioL Chem., 239, 2878 B O Y E R , R . A . (1954). US Patent No. 2560621 B U R G E S S , K.J. (1980). Cited by Burgess, K.J. and Coton, G. (1982), in

Proc. Kellogg Found. Intl. Sympos. Food Proteins, Applied Science, London (in press)

B U R G E S S , L . D . and S T A N L E Y , D .W. (1976). Can. Inst. Food Sci. Technol., 9, 228

C A R L I N , G.T. (1963). US Patent No. 3100710 C U M M I N G S , D . B . , S T A N L E Y , D.W. and de M A N , J.M. (1972). Can. Inst. Food

Sci. Technol., 5, 124 D E G R O T T , A .P . and SLUMP, P. (1967). / . Nutr., 98, 45 D E M A N , J.M. (1976). Cereal Foods World, 21, 10 D E M U E L E N A E R E , H.J.H. and B U Z Z A R D , J.L. (1969). Food Technol., 23, 345 ELLISON, N .S . , G A U L T , N.F.S. and L A W R I E , R .A . (1980). Meat ScL, 4, 77 F O X , D . (1981). BSc Hons. Dissertation, University of Nottingham G A U L T , N.F.S. and L A W R I E , R .A . (1980). Meat Sci., 4, 167 G E R R A R D , M.P.E. (1886). French Patent No. 178367 G I D D E Y , C. (1960), u s Patent No. 2952542 H A R P E R , J.M. (1979). CRC Crit. Rev. Food Sci. Nutr., 11, 1551 H A R P E R , J.M. (1981). Extrusion of Foods Vols. I & II. CRC Press Inc.,

Boca Raton, FL H A R T M A N , N . E . (1967). US Patent No. 3320070 H A U G , A . (1961). Acta Chem. Scand., 15, 950

form. In this respect cooking extruders have many advantages because of their capabihty to cook, minimize nutrient loss, form, mix, texturize and shape under conditions of high production and low cost (Harper, 1981). However, the technology of extrusion processing is very much in its infancy and both the Food Industry and the consumer will only gain the maximum benefit from this process when the complex physical and chemical changes undergone by the multicomponent mixtures in the extruder are more fully understood.

It is probably fair to say that the principles involved in fibre spinning are better understood than those involved in extrusion processing, but even here a better understanding of the factors affecting the functional properties of protein would lead to improvements in the established technology. Fibre spinning is economically less attractive than extrusion processing but it does possess the advantages of being a more easily controlled procedure and, especially if polysaccharides are also included in the mix, it is capable of generating a more diverse range of textures than is at present possible by extrusion processing.

R.A, Lawrie and D.A. Ledward 181

H O R A N , F.E. (1974). In New Protein Foods, Vol. lA, Ed. by A.M. Altschul, p.367. Academic Press, New York

IMESON, A . P . , L E D W A R D , D . A . and MITCHELL, J R . (1977). / . Sci. Fd Agric., 27, 621

IMESON, A . P . , L E D W A R D , D . A . and MITCHELL, J.R. (1979). Meat Sci., 3, 287 IMESON, A . P . , MITCHELL, J.R. and L E D W A R D , D . A . (1980). / . Fd Technol.,

15, 319 IMESON, A . P . , W A T S O N , P.R., MITCHELL, J.R. and L E D W A R D , D . A . (1978). / .

Fd Technol., 13, 329 JENKINS, S.L. (1970). u s Patent No. 3496858 JOKINEN, J .E . , REINECCIUS, G.A. and T H O M P S O N , D .R. (1976). / . Fd Sci., 41,

816 K E L L O G G , J.H. (1907). US Patent No. 869371 KINSELLA, J.E. (1978). CRC Crit. Rev. Fd Sci. Nutr., 10, 147 KNIGHT, S.E. (1981). BSc Hons. Dissertation, University of Nottingham L U N D G R E N , H.P. (1945). Textile Res. J., 15, 535 L U N D G R E N , H.P. (1949). Adv. Prot. Chem., 5, 305 M A U R I C E , T.J., B U R G E S S , L .D. and S T A N L E Y , D .W. (1976). Can. Inst. Fd Sci.

Technol., 9, 173 MILLAR, A . (1898). Br. Patent No. 6700 MITTAL, p. (1981). PhD Dissertation, University of Nottingham M U S T A K A S , G.C. , A L B R E C H T , W.J., B O O K W A L T E R , G.W., M c G H E E , J .E . ,

KWOLEK, W.F. and GRIFFIN, E.L. (1970). Food Technol., 24, 1290 N A G A N O , K. (1974). / . Mol. Biol., 84, 337 NIPPON SEIFUN CO. (1980). Japanese Patent No. 5516615 NISSHIN MILLS, KK (1981). Japanese Pat. 561 7062 P U T N A M , F.W. (1948). Adv. Prot. Chem., 4, 79 R O S E N B R O O K , H.H. (1960). Computer J. (Oct), 175 R O S S E N , J.L. and MILLER, R.C. (1973). Food TechnoL, Chicago, 27, 46 RUSIG, O. (1979). Meat Sci., 3, 295 SIEGEL, D . G . , C H U R C H , K.E. and SCHMIDT, G.R. (1979). / . Fd Sci., 44, 1276 SMITH, A.K. and CIRCLE, S.J. (1972). Soyabeans : Chemistry & Technology,

1, 368. AVI, Westport, Conn. SMITH, J. (1982). Unpublished data SMITH, J., MITCHELL, J.R. and L E D W A R D , D . A . (1982). Prog. in Fd and Nutr.

Sci., 6, 139. Pergamon, Oxford S U G I N O , Y. (1979). US Patent No. 4158065 S U Z U K I , T. (1981). Fish and Krill Protein, p. 148. Applied Sci., Lond. SWINGLER, G.R. and L A W R I E , R .A . (1977). Meat Sci., 1, 161 SWINGLER, G.R. and L A W R I E , R .A . (1979). Meat Sci., 3, 63 SWINGLER, G.R. , N E A L E , R.J. and L A W R I E , R .A . (1978). Meat Sci., 2, 31 SWINGLER, G.R. , N A Y L O R , P.E. and L A W R I E , R . A . (1979). Meat Sci., 3, 83 SZCZESNIAK, A .S . and T O R G E S O N , K.w. (1965). Adv. Food Res., 14, 33 T H O M P S O N , D . R . , WOLF, J.C. and REINECCIUS, G.A. (1976). Trans. ASAE,

19, 989 T O L S T O G U Z O V , U . B . , D I A N O V A , V .T . , MZHELSKIV, A.I . and Z H V A N K O , Y U , N.

(1978). Myasn. Ind., USSR, 6, 22 T O M B S , M.P. (1975). In Meat, Ed. by D.J.A. Cole and R.A. Lawrie, p.525.

Butterworths, London


T U R N E R , R . H . , M A C K E N Z I E , I. and M A C F A R L A N E , J.J. (1979). / . Fd Sci., 44, 1443

VISSER, J., OOSTHOEK, R .H. , G R O E N E W E G G , J.W. and DIJKSTRU, S. (1980). US Patent No. 4208436

WOLF, J .C. , T H O M P S O N , D.R. and REINECCIUS, G.A. (1978). / . Fd Sci., 43, 1486

Y O U N G , R.H. and L A W R I E , R .A . (1974a). / . Fd Technol., 9, 171 Y O U N G , R.H. A N D L A W R I E , R .A . (1974b). Proc. IV ML Congr. Fd Sci,

Techn., IV, 273 Y O U N G , R.H. and L A W R I E , R .A . (1975a). J. Fd Technol., 10, 453 Y O U N G , R.H. and L A W R I E , R .A . (1975b). / . Fd TechnoL, 10, 465 Y O U N G , R.H. and L A W R I E , R . A . (1975c). / . Fd Technol., 10, 528

13

Introduction

C O N V E N T I O N A L PROCESSING OF B O N E

Bone comprises about 12% of the Hve weight of a large meat animal and about 8% of a pig. Since roughly only half of the live weight is eaten in the form of meat and edible offals, bone constitutes 16-24% of the non-edible wastes. About one-sixth of the weight of raw bone is protein (dry basis).

A certain amount of bone is sold to the consumer in bone-in joints and cuts (of sheep and pig particularly) and is not available for further processing. Some bone is used for the manufacture of bone-stock for food use and some is ground to an emulsion which is incorporated into pet-foods. Manufacture of bone glue has ceased in the UK but there is still a production of ossein gelatin from indigenous bone. However, much the largest fraction of the bone arising from abattoirs and boning plants is collected by the animal by-products processing industry and rendered into bone tallow and meat-and-bone meal. Direct UK statistics are hard to come by but very approximately:

200 000 tonnes per annum are rendered 220 000 tonnes per annum are used for gelatin manufacture

30000 tonnes per annum are used in soup-stock, soups, pet-foods, etc. Total 250 000 tonnes per annum with a total dry protein content of —40 000 tonnes per annum.

The profitability of rendering has always been cyclical since the selling prices of the two products, protein and fat, are determined largely by world commodity prices for protein feeding meal and oils and fats (whether of vegetable or animal origin). In recent years, the rendering industry has been further burdened by the sharp rise in fuel costs and the additional capital and operating expenses of meeting ever tighter pollution control requirements of local communities.

The outlets for the various conventional products made from bone can be grouped as follows in descending order of product value:

(1) food and pharmaceutical products/ingredients (edible fats, gelatins, edible bone phosphate, soup-stock);

183

CONVERSION OF BONE TO EDIBLE PRODUCTS

A. JOBLING and C.A. JOBLING Lensfield Products Ltd, Bedford, UK

184 Conversion of bone to edible products

The Lensfíeld processes

GENERAL

Three new processes have been developed {Figure 13.1) which enable fresh bone to be fractionated into its constituents (protein, fat, calcium phosphate) in forms which are suitable for use as food ingredients. All material is first size reduced and defatted by a centrifugal washing process. There follow two alternative processes. In the 'cooking' process, the defatted bone is pressure-cooked and hydrolysed collagen extracted leaving a residue of bone phosphate. In the alternative 'acid' process, the defatted

Defatting process

Acid process

Cooking process

Dicalcium phosphate

Edible bone collagen

Edible bone " phosphate

^ Soluble bone ' protein

Figure 13.1 Lensfieid processes for fractionation of fresh raw bone

(2) pet-food ingredients (greaves, bone meal, bone emulsions); (3) general industrial products (technical gelatins, bone glues, tallows,

bone ash); (4) agricultural and horticultural products (meat-and-bone meal, steamed

bone meal, tallow, dicalcium phosphate).

The last category is by far the largest in volume but the lowest in unit value. The technological task is therefore to upgrade as much as possible of this material into higher categories and thereby increase the value added to the meat animal.

Λ. Jobling and CA. Jobling 185

Pork Beef (%) (%)

Moisture 43 32 Nitrogen 3.3 3.3 Protein (as Ν x 6.25) 20.6 20.6 Fat 12.4 15.2 Ash 21.4 29.0

Typical compositions of beef and pork bone are shown in Table 13.1. Ordinary fresh bones which have not been through a mechanical recovery system can be accepted into the process but the quality of certain of the products, particularly bone phosphate and soluble bone protein, is adversely affected for some purposes by the presence of meat residues. This disadvantage can be corrected by building a meat recovery step into the first (defatting) process at a convenient point.

D E F A T T I N G P R O C E S S

Water-defatting of bone was developed originally by Chayen and Ashworth (1953) and there are now several commercial processes available based on the general principle of crushing, washing and centrifuging bone.

bone is steeped in cold dilute hydrochloric acid which dissolves the bone mineral and leaves behind native bone collagen. Dicalcium phosphate is recovered as a by-product of lime treatment of the spent acid.

R A W M A T E R I A L

A key element in the development of these processes has been the commercial availability of fresh bone which has been processed through a mechanical recovery system (such as the Protecon, Paoli, Beehive, Hyd-rau, etc.). These systems permit bone which has been through a normal butchering operation to be stripped of a further 20-35% (depending on raw material and machine setting) of a paste of fat, water, and meat fibres which can be used in comminuted meat products. The other product of this mechanical recovery process is crushed clean bone with very little adhering meat. This is an excellent and consistent starting material for the processes to be described.

For optimum product quality, the bone has to be handled as a food product; chilled or frozen if necessary and transported and stored under refrigeration. Mainly beef and pork bones are available. These are processed separately for both technical and commercial reasons. Mechanical recovery is also extensively practised on poultry carcases and chicken bone has been processed experimentally on the Lensfield system with no problems.

The concentration of slaughtering into larger units and the establishment of large central boning plants are making available regular bulk supplies of fresh bone of consistent quality and uniform origin. This consistency affords a good opportunity for fine tuning of the subsequent processes.

Table 13.1 C O M P O S I T I O N O F B E E F A N D P O R K B O N E


C O O K I N G PROCESS

The defatted bone is loaded into the basket of a new type of pressurized slow-speed basket centrifugal cooker which has been developed in association with J Lildal A/s of Randers, Denmark (Olds and Jobhng, 1978).

When the basket is in position in the closed centrifuge, the first effect of rotation is to shape the bone mass into a hollow cyHnder. Superheated water is then circulated down the axis of rotation, through the bed of bone and out through the base of the centrifuge to the recirculating pump and pressurized reservoir. This system gives a fast and very efficient leaching of hydrolysed protein from the bone which also acts as its own filter bed, retaining any traces of fat and suspended solids to give a high-clarity extract. In the usual way, a series of extracts is taken, the later more dilute washings being recirculated on to the next batch. This sequencing can be controlled either manually or by a microprocessor. Total extraction time is 3-4 h.

In addition to the self-clarifying action referred to above, the advantages of this new extraction system are the completeness of the extraction throughout the mass of bone and the capability to produce higher concentration extracts than the conventional gravity system because of the more complete drainage that can be obtained under the centrifugal action.

The protein hydrolysate is normally recovered by evaporation to a suitable feed concentration and spray-drying to a stable powder.

The residue in the centrifuge basket is dried on a band dryer, milled and gristed as an edible bone phosphate. Typical yields on raw bone are 13% bone protein at 5% moisture and 28% bone phosphate at 2% moisture.

These are generally applied to the pretreatment of bone leading to manufacture of bone glue or ossein gelatin.

The requirements of the present work were that the defatting stage should be capable of:

(1) producing a low residual fat-in-bone content of the order of 2% (dry basis);

(2) being operated to food hygiene standards.

The bone is crushed in a two-stage operation to a maximum particle dimension of 7-10 mm. General experience is that a two-stage size reduction gives a narrower particle size distribution, which is desirable.

The bone is then slurried in hot water in a countercurrent system which removes much of the fat, washed again, and pumped to a decanter centrifuge which separates solids (bone) from fat and water. The fat/water mixture is then separated in a conventional disc centrifuge and the fat is polished, antioxidant added, and the fat pumped to store. The aqueous effluent is either recycled or passed to the effluent treatment plant.

Recovered fat is of grade 1 quahty or better with a yield of 10-14% on raw bone depending on animal species and whether or not the bone has been through a meat recovery system.

Α. Jobling and CA. Jobling 187

Products—properties and uses

E D I B L E B O N E C O L L A G E N

The primary product is a granular powder, each particle of which retains the size and shape of the bone particle from which it has been produced. It has a sponge-like structure, the pores having been formed by removal of fat and dissolution of bone mineral. For commercial use, the granular product is milled to a powder (95% through 1 mm aperture).

Bone collagens from different animal species (mainly beef, pork or poultry) are not readily distinguishable. There is little residual taste or

A C I D PROCESS

In the acid process (England et al., 1980, 1981), the cold, undried, defatted bone is introduced into cold dilute hydrochloric acid which dissolves the mineral (mainly hydroxyapatite) component of the bone leaving behind the collagen in suspension. The protein particles retain the form of the original bone pieces from which they have been formed and have a porous structure.

In practice, the demineralization is carried out as a countercurrent operation in three stages with subsequent washing of the protein and adjustment to pH 4-5. The protein is dewatered in a decanter centrifuge, dried on a band dryer and milled.

The spent acid is neutralized with a hme slurry which precipitates dicalcium phosphate. This is recovered on a vacuum filter and dried and is a standard ingredient in compound animal feeds as a source of calcium and phosphorus. Dicalcium phosphate yield is around 25% of raw bone.

In its chemistry, this process is, of course, identical to that used for making ossein as an intermediate in the production of bone (ossein) gelatins. The differences in the Lensfieid process are:

(1) the use of fresh bone and the operation of the defatting and deminer-alizing processes to food hygiene standards so that the bone collagen (ossein) is edible. In conventional gelatin manufacture, ossein is not an edible product but further purification takes place in the subsequent gelatin extraction stages;

(2) the use of a continuous countercurrent process for contacting acid and bone in which both the acid and the bone are moved. In conventional ossein manufacture, the dilute acid is percolated through static beds of bone.

A collagen product which has been demineralized to a residual ash content of around 5% is found to be generally satisfactory for food use. At this, the yield of dry collagen (5% moisture) from raw bone is about 15% although this figure can vary with animal species, age and type of bone. The main losses of collagen are as finely suspended solids in the spent acid and as soluble (hydrolysed) protein if the conditions of demineralizing (time, temperature and acid concentration) are too severe.


Moisture 6% max. Fat 5% max. Ash 5% max. Protein (N x 6.25) 90% min. Chloride (as NaCl) 3 % max. pH (10% suspension) 4 -5 Water absorption 300-400% at room temp. Total microbial count 1000/g max. E. coli Absent from 1 g sample Salmonellae Absent from 10 g sample

odour but the coUagen appears to be able to take on and even enhance meat flavours of other products with which it is mixed. A typical composition is shown in Table 13.2.

It is useful but perhaps not surprising that the dilute acid treatment has a powerful sterilizing action and routine production of bone collagen with a low bacterial content is not difficult. The main use of edible bone collagen is as a secondary source of animal protein in almost any comminuted meat product—for example, sausages, burgers, luncheon meats, pies, etc. The natural sponge-like structure has good absorbency of water and fat. The latter property is particularly useful in reducing free fat in such products which were originally developed as a means of making more palatable the fattier scraps and less attractive cuts of meat.

Most recently, interest has been developing fast in the combination of collagen proteins with blood proteins. Such blends can be used in meat products to give technical, nutritional and economic advantages.

The use of blood in traditional meat products has been limited by the dark colour and strong flavour which are generally developed. Hence it was restricted to a few special products such as the English black pudding. Such outlets consume only a very small proportion of the total amount of slaughterhouse blood arising.

A colourless plasma fraction can be separated by centrifugation but early commercial plasma preparations were not always of sufficiently high quality to gain wide acceptance. Recently, plasma protein production has been streamlined so that some very good quality material is now commercially available in chilled hquid, frozen, or low-temperature spray-dried form. If the blood is taken cleanly, chilled immediately and processed quickly, plasma products with little taste or odour and very good bacteriological quality can be produced.

Such plasma can be added to almost any comminuted meat product but the high water-binding and heat-gelling power of plasma protein is such as to produce a fairly tightly bound meat product, which may or may not be what is wanted.

Bone collagen on the other hand tends to loosen the texture of any meat product to which it is added and make it generally more crumbly and less chewy. Thus a blend of blood plasma and bone collagen can be used to obtain independent control of consistency.

There are other advantages of the collagen-plasma blend. If undried plasma is used in a meat product it is sometimes a problem to accommo-

Table 13.2 TYPICAL COMPOSITION A N D PROPERTIES OF E D I B L E B O N E C O L L A G E N


S O L U B L E B O N E P R O T E I N

Soluble bone protein is a bone collagen hydrolysate. As manufactured at present, the molecular weight is around 20000, i.e. too low for the protein to be gel-forming in cold aqueous solution. By adjustment of the cooking schedule, a geUing protein extract can be obtained with a Bloom gel strength of about 100. A typical composition is indicated in Table 13.3.

Table 13.3 T Y P I C A L C O M P O S I T I O N O F S O L U B L E B O N E P R O T E I N

Moisture 5% max. Nitrogen 15 .5-17% Ash 2% max. Chloride (as NaCl) 1% max. p H ( l % solution) 5 - 7 Total microbial count 5000/g E. coli Absent from 0.1 g sample

Chemically, soluble bone protein is similar to a bone gelatin hydrolysate with the difference that, whereas the gelatins are prepared by acid or alkaline hydrolysis of bone collagen (ossein), the soluble bone protein has been made by direct neutral hydrolysis of bone without the use of any chemicals at any stage. This may be a commercially important point if the product is intended to be used in, for example, health foods, baby foods or toiletry preparations.

The closest commercial analogue is probably bone-stock or soup-stock which is the concentrated (about 66% protein) salt-preserved extract obtained by pressure cooking of fresh bones and meat scraps. Soluble bone protein is a refined form of soup-stock with a lower residual fat content and no added salt.

Soluble bone protein is used as an ingredient in soups, sauces and gravies and for protein supplementation of meat products. It is also of interest as an ingredient in reaction flavour processes to produce new ranges of improved meat flavours.

E D I B L E B O N E P H O S P H A T E

Edible bone phosphate consists mainly of the mineral hydroxyapatite with some carbonato-apatite and fluoro-apatite. It also contains residues of unextracted protein and fat. Typical compositions are given in Table 13.4.

date the high water content of the plasma in the formulation. One possible solution is to absorb the hquid plasma on to bone collagen powder. One part of collagen powder will absorb naturally about four parts of liquid plasma to form a paste of similar consistency to that of comminuted meat. The collagen/plasma combination has a better balanced essential amino acid composition than that of collagen alone. Nutritional aspects of both are discussed more fully in the section on collagen in nutrition.


Table 13.4 T Y P I C A L C O M P O S I T I O N S O F E D I B L E B O N E P H O S P H A T E S ( T Y P E S A A N D Β A R E A R B I T R A R Y C L A S S I F I C A T I O N S B A S E D O N A S H C O N T E N T )

Type Λ Type Β

Moisture 2% max. 2% max. Fat 2% max. 2% max. Ash (650 °C) 84% min 8 1 - 8 4 % Calcium (CaO) 45% min 43 .5 -45% Phosphorus (P2O5) 34% min 3 3 - 3 4 % Fluorine 700 mg/kg max. Heavy metals (as Pb) 30 mg/kg max. Zinc (Zn) 150 mg/kg max. Total count 500/g max. E. coli Absent in 10 j ι sample S a l m o n e l l e Absent in 50 j ι sample

The other major use of bone phosphate is as a free-flow additive in powdered food products such as salt, sugar and powdered fruit drink concentrates. For the best effect, a very finely powdered grade is used at around 0.5-2% addition.

The adult body contains about 1200 g of calcium and 600-900 g of phosphorus. Most of this is in the bones and teeth. The skeleton provides a reservoir of calcium to maintain the level of blood calcium, which is regulated by the endocrine hormones. It has been estimated that some 700 mg of calcium are exchanged in the adult skeleton each day. The level of blood calcium is also important in relation to blood clotting and the maintenance of normal muscle activity.

Phosphorus of course occurs in nucleic acids, the phospholipids of nervous tissue, all cell membranes and many substances involved in the essential intermediate metabolic pathways.

Recommended daily intakes for calcium range from about 500 mg/day for adults to 1400 mg/day for adolescents and women in late pregnancy or in lactation. A phosphorus-deficient diet is rarely encountered.

Bone phosphate also contains some useful trace elements, particularly fluorine (as fluoroapatite) and a surprisingly high content (150mg/kg) of zinc, which is not readily obtained from many other foods. The US recommended daily intake for zinc is 5 mg/day for one year old children rising to 15 mg for adults and 25 mg for lactating women. Zinc is present in carbonic anhydrase in blood red cells and its nutritional importance is being increasingly recognized. Recently, low birth-weight in some children has been correlated with maternal zinc deficiency.

The basic porous microstructure is not destroyed in the extraction process and is Ukely to assist assimilation? Edible bone phosphate is a permitted food additive in the UK.

The main use of edible bone phosphate is as a natural source of calcium and phosphorus in the correct ratio in health foods, pharmaceutical and veterinary products. Many of these products are made in tablet form and it is a further advantage that certain grists of bone phosphate powder can be tabletted directly without the use of additives.


M I X T U R E S OF C O L L A G E N A N D B L O O D P L A S M A PROTEIN

As indicated previously, there is considerable practical interest in the potential for using collagen in combination with blood plasma protein.

Collagen in nutrition

M I X T U R E S OF C O L L A G E N PROTEINS WITH M E A T MUSCLE PROTEINS

A characteristic of collagen proteins is a relatively high content of hydroxyproline—indeed collagen is usually determined analytically on the basis of hydroxyproline x 7.1 rather than nitrogen x 6.25. In its content of essential amino acids (those which cannot be synthesized by the body and have to be regularly ingested in food) collagen lacks tryptophan almost completely and is low in methionine, cystine, and tyrosine. These deficiencies are responsible for the traditional view of collagen as a protein of low nutritional status.

However it is increasingly recognized that the study of the nutritional quality of single proteins is somewhat unreal and this is particularly so for collagen which is almost always eaten in association with muscle (meat) protein in meat cuts and meat products.

The digestibility and biological value of meat mixed with various amounts of collagen (as pigskin) and given various heat treatments have recently been measured by Laser-Reuterswärd et al. (1982) at the Swedish Meat Research Institute at Kävlinge. Digestibility of the collagen comminuted to ~2 mm was high (>95%) in agreement with earher work and unaffected by heat treatment.

A direct relationship was found between net protein utilization (NPU) (y) and collagen content (x) as % crude protein according to the equation

y = 82.8 - 0.6 χ

with a correlation coefficient of 0.99. These results are in close agreement with work by Bender and Zia (1976) who determined NPU values experimentally and found NPU = 69 for shin beef containing 23.6% collagen and NPU = 82 for fillet beef with 2.5% collagen. Values calculated from the equation above are 69 and 81 respectively.

For weaning foods, the UN Protein Advisory Group recommends a minimum NPU value of 60 and preferably 65. Using the above equation, these values correspond to collagen contents in the collagen/muscle protein system of 38 and 30% respectively. The lower figure is close to the 28.5% collagen limit (corresponding to a minimum protein efficiency ratio (PER) value of 2.5) recommended by Yu Bang Lee et al (1978) for interim legislation by the US Department of Agriculture.

Nutritional requirements of adults are generally less exacting and for the collagen/lean meat system, this is borne out by the work of Kofranyi and Jekat (1969). In trials with adult human subjects, they found that up to 50% of beef muscle protein could be replaced by the collagenous protein gelatin without loss of nutritional value.


Amino acid Collagen (C) Blood plasma (B) Muscle (beef) 1C:1B 2C:1B

Tryptophan 0 1.7 1.1 0.9 0.6 Phenylalanine 2.4 5.7 4.0 4.1 3.5 Lysine 4.0 8.3 8.4 6.2 5.4 Threonine 2.3 7.1 4.0 4.7 3.9 Methionine 0.7 1.3 2.3 1.0 0.9 Leucine 3.7 10.1 8.4 6.9 5.8 Isoleucine 1.9 3.4 5.1 2.7 2.4 Valine 2.5 7.4 5.7 5.0 4.1

Although it has been shown in the preceding section that the essential amino acid deficiencies of collagen do not in practice detract significantly from its nutritional value in collagen/muscle protein diets, nevertheless a further improvement in amino acid profile can be obtained by introducing blood plasma protein as a third component, as Table 13.5 shows.

B O N E PROTEINS A N D F O O D LEGISLATION

In the foregoing outline of the uses to which bone proteins can be put, no account has been taken of any legislative constraints that there may be on their use, particularly in meat products.

There is general acceptance that bone proteins can now be produced in forms which are wholesome and suitable for use as food ingredients. It is also clear that they are not novel proteins requiring to be subjected to special screening procedures—they are merely new forms of collagen which has always been present in a meat-containing diet.

The difficulty arises when it is required to know whether these proteins come inside or outside the various national and international legal definitions of meat and meat products, a matter of considerable commercial importance. Because meat technology has now advanced ahead of legislation, there is a group of new, wholesome, nutritious, commercially-available animal products whose existence was not envisaged when much of the current legislation was drawn up. This group includes the bone proteins, edible blood plasma and red-cell fractions, various animal protein hydrolysates, processed pigskin and others. In addition to their protein content, all of these have various technical advantages in the formulation of meat products.

Because, by their nature, the new products are very close to or identical with components of meat as it is generally understood, their detection in the presence of meat by chemical methods is at best very difficult and unsuited for routine control. The most that a routine analysis is hkely to show is a disproportion in the content of, say, collagen or plasma protein, which can do no more than suggest that one of the new products may be present, given the inherently wide compositional range of the meat itself.

Thus we arrive at the current situation in which meat products manufactured under the laws of one country are exported to other countries with different legislation.

Table 13.5 ESSENTIAL A M I N O A C I D COMPOSITIONS OF SOME A N I M A L P R O T E I N P R O D U C T S A N D M I X T U R E S ( D A T A FROM L A W R I E , 1971 A N D O L S O N , 1970)


References

B E N D E R , A . E . and ZIA, M. (1976). / . Fd TechnoL, 11, 495 C H A Y E N , LH. and A S H W O R T H , D.R. (1953). / . Appl. Chem., 3, 529 E N G L A N D , R. , B O W C O T T , J .E.L. , O L D S , J.S. and JOBLING, A . (1980). British

patent appl. 80 32267 E N G L A N D , R., BOWCOTT, J .E.L. , O L D S , J.S. and JOBLING, A . (1981). European

patent appl. 81 304549.9 K O F R A N Y I , E . and JEKAT, F. (1969), Hoppe-Seyler's Z. Physiol. Chem., 350,

1405 L A S E R - R E U T E R S W A R D , Α . , ASP, N-G. , BJÖRCK, I. and R U D É R U S , H. (1982). / .

Fd Technol., 17, 115 L A W R I E , R .A . (1971). Proc. Inst. Fd Sci. Techn., 4(111), (No 2), 190 O L D S , J.S. and JOBLING, A . (1978). Danish patent 143584B. O L S O N , F.C. (1970). Proc. Meat Ind. Res. Conf., 23 Y U B A N G L E E , ELLIOTT, J.G. , R I C K A N S R U D , D . A . and H A G B E R G , E.G. (1978).

/. FdScL, 43, 1359

The same new (animal) material incorporated into the same meat product in three countries may be regarded quite differently by each as follows:

(1) in country A, it is classified as meat and contributes towards the meat content of the total product (for which there may be a legal minimum);

(2) in country B, it is classified as an additional (non-meat) food ingredient, which must be declared in the ingredients list;

(3) in country C, its use in meat products is not permitted even though it is identical with and analytically indistinguishable from a component or components of meat.

The exporter has a duty to ensure compliance with the legal requirements of the importing country but, as indicated above, product analysis alone is increasingly unable to provide the required degree of control. This is because current legislation is framed in terms of what is allowed to be put into the meat product rather than of the end properties, nutritional or otherwise of that product.

There is therefore a growing need for the rationalization and harmonization of laws relating to international trade in this field. A good starting point might be to ensure that in future there is no conflict between any new national legislation and already estabhshed international legislation such as that set out in the various EEC directives on meat and meat products.

14

Introduction

The potential for using upgraded waste materials as ingredients of meat products is considerable, but it is influenced by a number of factors.

Firstly, and most importantly, meat and meat products are staple foods and central to the eating practices of many Western countries. In Britain they account for over a quarter of all domestic expenditure on food and contribute a similar proportion of protein and iron in the diet (National Food Survey Committee, 1982). The market is large but the consumer is quality-conscious and cost-conscious and has attitudes which are rooted in tradition.

Secondly, meat has great flexibility as a raw material for processing and is used for the manufacture of a diversity of products for consumption at every meal of the day, at any season and on any occasion. Their diversity is due partly to the nature of different types of meat and partly to their compatibihty with other foods as components of meals.

Thirdly, the butcher and the processor have traditions of frugality in their approach to the use of their expensive raw material and this is reflected in the technology which they have built up over the years to use high value and low value parts of the carcase to best advantage. In Britain the technology includes the use of cereals and other cheaper foods in association with meat to make products with a wide range of price and eating characteristics (Hannan, 1975).

In summary, therefore, there are opportunities for using upgraded waste materials in a number of ways, but the requirements of the consumer are stringent and there are many practical Hmitations. The types of waste materials that can be considered have been discussed by various workers (Young, 1980; Lawrie, 1981; Ranken, 1982).

Types of meat product

The demand for particular types of meat on the retail market is normally related to the size of individual pieces of meat and their ease of preparation in the domestic kitchen. Large tender muscles from the hindquarter of young animals are readily salable and command high prices, while tougher

UPGRADED WASTES IN MEAT AND MEAT PRODUCTS

R.S. HANNAN Meat and Livestock Commission, Bletchley, UK

197

198 Upgraded wastes in meat and meat products

Meat types Processes

Whole meat—bone in \ Cutting

—boneless 1 Addition of non-meat ingredients Large pieces | Heating (including canning) Small pieces / ^ Smoking

Coarse comminute I Curing Fine comminute ] Drying Reformed meat ' pH change

Freezing

In individual products—meat may be of different species —combinations of meat types ^ —combinations of processes

may be used

The overall market, therefore, includes products made from pieces of meat of all sizes and eating characteristics, processed in many ways, as summarized in Table 14J. The methods of manufacture are such that non-meat materials can be incorporated in all types of product and it is against this background that we must view the use of upgraded wastes. They can be considered in three broad categories and can be used in association with meat in varying combinations and proportions:

(1) materials which contribute to the meat content of products; (2) optional non-meat ingredients; (3) functional ingredients usually used at lower levels of addition.

or more complicated tissues from other cuts are correspondingly lower in value. The processor uses both types of meat in his product range but an important part of his skill lies in 'adding value' to the materials of lower value, including trimmings and residuals from the manufacture of high value products which would otherwise have limited value for human consumption. His first operation is usually to divide the meat into small pieces before use; and manufacture of comminuted products has always played an important part in the overall efficiency and economy of operation of the industry.

The scope of the processor was restricted in the past by limitations of the available technology and by a need for shelf stability at ambient temperatures. Modern methods of distribution have removed many of the latter limitations and modern technology has provided new methods of modifying the eating quality of products. A wider range of products can be made from comminuted meat using new methods of comminution which produce pieces of meat in the form of flakes and these can be moulded into 'reformed meats' with many of the eating properties of whole meat (Shaw, 1974; Ranken, 1982a; Briedenstein, 1982). With whole-meat products, such as ham, we have improved methods of injecting curing solutions and binding pieces of meat together with dispersions of meat proteins produced by gentle working of the meat after injection in tumblers or massagers. These dispersions form irreversible gels on heating and their understanding and use is an important feature of modern technology.

Table 14.1 N A T U R E OF C O M M O N E R M E A T P R O D U C T S

R.S. Hannan 199

Non-meat ingredients

Non-meat ingredients can broadly be classified as of animal origin or non-animal origin. Materials of animal origin can be further divided into offals and edible by-products and there is a tradition of restriction on their use in human food, stemming partly from their uncertain microbiological

Materials which count as 'meat'

The practice in many countries is to categorize meat products by specifying the permissible methods of manufacture, together with screening tests which confirm that the finished product has been made properly. In Britain we approach the problem differently and place the emphasis on including a known amount of meat in the product, the term 'meat' being defined in regulations. Current Meat Product Regulations specify different minimum requirements for the meat content of different products while proposed new regulations (Ministry of Agriculture, Fisheries and Food, 1981) will, if enacted, require declaration of meat content on labels of all meat products; statutory minima will only be retained for certain traditional products such as sausages and meat pies. The definition of 'meat' in the new proposals is summarized in Table 14.2.

Table 14.2 M A T E R I A L S WHICH C O N T R I B U T E TO M E A T C O N T E N T OF M E A T P R O D U C T S A S IN P R O P O S E D UK M E A T P R O D U C T R E G U L A T I O N S (MINISTRY OF A G R I C U L T U R E , FISHERIES A N D F O O D , 1981)

Flesh, including fat,

and

Skin, rind, gristle and sinew in amounts naturally associated with the flesh used

and

Specified offals^ mammahan species: diaphragm, head, heart, kidney, liver, pancreas, tail, thymus, tongue avian species: gizzard, heart, liver, neck

^May be used in any uncooked or cooked product.

The main category of meat is 'flesh, including fat', and this clearly covers all conventional carcase meat. The lower part of the table also hsts certain types of offal which would contribute to meat content and would not be restricted in their use. Skin, rind and gristle are 'meat' when used in amounts naturally associated with the flesh used and are commonly used in comminuted meat products at appropriate levels to add firmness of texture and shrink resistance on cooking (Poulanne and Ruusunen, 1981). Pig rind is usually separated from the flesh and cooked before use.

The behaviour of the various types of meat on processing varies widely according to the nature of the end product. Those which are likely to be graded as wastes are more limited in usefulness and tend to be used in comminuted products. Basic characteristics such as lean/fat ratio, pH, connective tissue content and water binding properties can be determined and used as a basis for least cost formulations (Newman, 1983).


RESTRICTED OFFALS

In Britain the use of certain types of offals in meat products has been restricted (Anon., 1953) for many years. The relevant regulations have been reviewed twice (Food Standards Committee, 1972; 1980) and the most recent recommendations are summarized in Table 14.3. Use of the

Table 14.3 OFFALS R E S T R I C T E D T O C O O K E D M E A T P R O D U C T S A N D N O T

C O U N T I N G T O W A R D S M E A T C O N T E N T A S IN P R O P O S E D UK M E A T P R O D U C T R E G U L A T I O N S (MINISTRY OF A G R I C U L T U R E , FISHERIES A N D F O O D , 1981)

Brains Lungs Spleen Feet Oesophagus Stomach Intestine—large Rectum Testicles Intestine—small Spinal cord Udder

hsted offals would only be allowed in cooked meat products and they would not count towards 'meat content'. In general these materials are underutilized for human consumption but they could conceivably find increased usage in certain types of cooked meat product if new regulations relax the requirement for minimum amounts of 'meat' (as in Table 14.1) to be present, subject to appropriate declaration.

The processing quahty of many of these products has been reviewed by Richards (1982). In general they contain less muscle protein than carcase meat and their functional properties are correspondingly different.

O T H E R OFFALS A N D B Y - P R O D U C T S

Tables 14.2 and 14.3 do not name all parts of the carcase which may be used for human consumption. Blood is a typical example. It is debatable whether this is offal or edible by-product but it is certainly used in black puddings or equivalent blood sausages of other countries. It would not be regarded as 'meat' under the proposed new British regulations, but as it is not listed in Table 14.3 there would be no restriction on its use in uncooked products.

The position on the continent of Europe is quite different; blood is usually regarded as 'meat' in all respects and finds wide use in meat products. Hygienic methods of collection in the slaughterhouse have been

status and partly from reservations of the typical consumer. A highly relevant factor in considering their use in present-day products is the improving technology of the primary meat and meat processing industries particularly in standards of hygiene and use of refrigeration. Residual meats, offals, and by-products from the modern industry are hkely to be cleaner and in a more suitable condition for further processing than their traditional counterparts.

The position regarding the use of individual offals and by-products tends to differ from country to country but there are broad similarities of approach.

R.S. Hannan 201

Derivatives of meat and animal products

SIMPLE D E R I V A T I V E S

Materials which have been derived from 'meat' by heat treatment or simple extraction procedures are, by common usage, usually regarded as still being 'meat'. Newer methods of extraction include separation of hard tissues from soft, as in mechanical desinewing of meat or mechanical recovery of meat from bones (Newman, 1983), and partial rendering of adipose tissue into usable fatty and non-fatty portions. Similar methods can also be used with offals and there is current interest in applying them to spleens to remove capsular and internal connective tissue (Bittel et al., 1981).

Simple physical separation in a centrifuge produces blood plasma which may be used directly in meat products or after freezing or drying. The red cell material can be extracted further with acid and acetone (Tybor, Dill and Landmann, 1973) or treated with proteolytic enzymes (Drepper, Drepper and Ludwig-Busch, 1981) to produce globin, which is similar in many ways to plasma protein except that it is nutritionally deficient in isoleucine. Both materials have heat-gelling and emulsifying properties broadly comparable with egg white (Hickson et al., 1982) and can be used for similar purposes.

Bone is not normally used as a food ingredient but bone marrow finds specialized food uses while aqueous extracts of bone are commonly used in soups and gravies. Traditional gelatine is an extracted derivative of high specification and Jobling (1983) has described the preparation of other bone derivatives, soluble bone protein and decalcified bone (ossein). This, clearly, is a further underutilized source of food grade proteins. Their legal status in Britain is similar to that of blood and its derivatives in that they can be used in all products, whether cooked or not, but do not contribute to meat content.

H E A V I L Y P R O C E S S E D B Y - P R O D U C T S

A further broad category covers materials derived from meat, offals and animal by-products by processes which are sufficiently drastic to change their nature completely. Treatment of intestines by cleaning, stripping and dry salting to produce sausage casings is a traditional example; similar collagenous materials can also be prepared from skins as by-products of leather manufacture. Gault and Lawrie (1980) have studied the use of alkali and detergent to extract the proteins of lung, stomach and intestines and Lawrie and Ledward (1983) have reviewed their conversion to fibrous

developed and this material is one of the most readily available for upgrading (Wismer-Pedersen, 1979; Ranken, 1980). Its strong colour tends to make its direct use self-limiting but this can be overcome to some extent by prior emulsification with milk protein and fat. It can also be decolorized by enzyme treatment (Quagha and Massacci, 1982).


Functional ingredients

The area of greatest technological interest is in using materials which have useful functional properties in their own right or which enhance the functional properties of meat itself. A number of materials of this type, including sah, polyphosphate, curing agents and milk proteins, are commonly used in meat products in relatively small amounts to modify colour, flavour or texture or act as containers during processing and distribution.

C O L O U R

The colour of meat and meat products is one of their most important attributes. It is particularly important in self-service displays where colours must be uniform, stable and able to withstand strong lighting. In Britain it is accepted practice to add small amounts of red colour to certain types of meat product and blood would seem to be a potential source of suitable colour. In practice, however, added haem pigments often prove to be insufficiently stable. Combination with materials such as carbon monoxide increases their stability but no derivative has yet found wide acceptance in the industry.

A further aspect of meat colour is important in the present context. Fresh meat products can discolour rapidly as a result of uncontrolled curing reactions if nitrite or nitrate are present. Even if these materials are not added deliberately they can be introduced as natural components of drinking water and vegetable products or by direct drying of ingredients in

materials. The possibilities of enzyme or chemical modification of all types of material of animal origin are extensive and have been reviewed by Brekke and Eisele (1981). Treatment of beef heart myofibrils with acetic anhydride and other acylating agents, for example, improves their functional properties by moving the isoelectric point (Eisele and Brekke, 1981). The fact that 'natural' sausage casings are regularly used on uncooked sausages suggests that heavily processed offals and other materials of this type are not to be restricted to use with cooked meat products. The operative criteria in considering their use, therefore, must be the cost of processing, the need for the end product, its safety and its market value.

It should be noted that in developing the use of a processed by-product it may often be necessary to break a chain of established practice whereby a starting material is left at ambient temperature to deteriorate into an offensive waste before recovery is attempted.

It is not proposed in this chapter to discuss the use of waste materials from vegetable sources as major ingredients of meat products. Soya derivatives, and cereals, for example, have long-standing uses but it is far from certain that these are waste materials in the same sense as the materials discussed so far. The use of micro-organisms grown on waste materials is a further possibility but there have been few accounts of practical application in meat products.

R.S. Hannan 203

flue gases. Any upgraded waste materials intended for use in uncured meat products must be free from criticism on this count.

F L A V O U R

The study of meat flavour is highly specialized and knowledge is fragmentary even with the most basic meat systems. Meat extracts have longstanding uses in soups and gravies and extracts of bone can be used in a similar manner. The protein hydrolysate industry also uses animal materials, including bone, as one of its starting materials. Newer understanding of flavour chemistry and use of newer enzyme preparations can be expected to contribute to further developments in this field.

Again there is a negative aspect of meat flavour. It is essential that meat and meat products should be free from unattractive or unfamihar flavours since these tend to be interpreted as danger signals. Any proposed new ingredient must be scrutinized closely from this point of view; it must be free from off flavours and must not produce them in the finished product within the expected total shelf-life. Memories of pronounced flavours can be long and an unfavourable reputation gained in the early days of the development of a product can persist for many years.

T E X T U R E

There is much current interest in the use of materials which influence the physical nature of meat products, including their response to cooking and texture when eaten. The textural characteristics of meat and meat products are themselves due largely to the nature of the various meat proteins. The insoluble proteins of connective tissue provide fibrous texture, while the soluble proteins provide binding and emulsifying properties, especially the salt soluble proteins of the myofibrillar system, which are able to form viscous solutions and gel irreversibly on heating. The gelling action can be used to bind pieces of meat, trap fat and help to control shrinkage during cooking. In cooked sausages such as frankfurters it forms a continuous structure throughout the product, in reformed meats it binds individual pieces of meat together and in cooked whole products for slicing it helps to control texture and losses on cooking (Schmidt, Mawson and Siegel, 1981). Its action appears to be due mainly to cross-linking of myosin molecules (Macfarlane, Schmidt and Turner, 1977; Samejima, Ishioroshi and Yasin, 1981). A certain amount of salt must be present for full development of these effects and the physical nature of many meat products depends on the use of salt to modify the relative contributions of the soluble and insoluble meat proteins. Further modification may be produced by addition of other ingredients from non-meat sources, particularly protein-containing materials.

I N S O L U B L E PROTEINS

Added insoluble proteins may contribute preformed fibrous texture or sponge-like proteinaceous matrix. The fibrous materials described by


S O L U B L E PROTEINS

Blood proteins are an example of a soluble protein with useful functional properties; they gel on heating and are also able to act as foaming and emulsifying agents. The proteins of milk and soya can be used in a similar manner and whey proteins, prepared by ultrafiltration of whey, are currently attracting interest (Evans and Gordon, 1980; Aarlbersberg, 1981; Evans and Gordon, 1982). As pointed out by Schut (1980) the solubility of whey proteins is relatively unaffected by concentrations of salt up to 5M and they are well suited to incorporation in meat curing brines which are commonly in the range 2-4M.

The relative gel strength of the various protein additives is most important but gelation in meat systems is not simple. Meat proteins in salt solution tend to gel in two stages: a loose gel forms between 50 and 60 °C and this becomes firm at 65-70 °C in the temperature range where shrinkage of the meat and cooking losses begin. Non-meat proteins, on the other hand, tend to gel at temperatures in excess of 70 °C and the behaviour of mixed systems is usually not quite the same as that of the all meat system. Interactions between individual proteins in meat systems have been described by Deng, Toledo and Lillard (1981) while proteins which merely increase viscosity, such as gelatine, can reduce cooking losses and improve eating quality, especially with products which are eaten cold.

The practical technologist, therefore, tends to pin his faith on direct testing of functional proteins in realistic processes and products, with only a nominal reliance on theoretical considerations. For example Terrell et al. (1982) have proposed a simple comparative test using a solution of protein to stick together two meat surfaces, an adaptation of the method originally used by Macfarlane, Schmidt and Turner (1977) with muscle proteins. Saterlee (1981) has compared the use of numerous proteins of current interest.

F A T EMULSIONS

Added proteins are also finding increasing use in the manufacture of preformed 'fat emulsions'. In a typical process for British meat products five parts of water and one part of protein are mixed in a bowl chopper and five parts of fatty tissue added to form a solid fat-containing material which can be used as an ingredient of many types of comminuted products

Lawrie and Ledward (1983) are of the former type and decalcified bone (Jobhng, 1983) and dried rind of the latter. When added to comminuted products they can enhance fibrous texture and firmness of bite of soft products and add juiciness by loose binding of meat juices. They can also assist in the control of shrinkage on cooking. Wheat gluten is an available insoluble protein from vegetable sources which can contribute texture of meat products. Mixing an insoluble dry material of this type with meat can present problems of aggregation but Hand, Crenwelge and Terrell (1981) have reported its successful use in reformed beef.

R.S. Hannan 205

U S E S O F C O L L A G E N

Collagenous materials can be used for functional purposes of a different kind. Sausage casings with highly specialized qualities have traditionally been made from various sections of the animal gut and these have been supplemented in recent years by casings made from regenerated collagen. The main source of collagen is the inner layer of cattle hide ('split') which is separated at an early stage of the leather making process and is httle used at present. It can readily be converted into a collagen dough by treatment with acid and then extruded in the form of a casing. Preformed casings are commercially available, or sausage meat and collagen dough can be co-extruded to make finished sausage in a single operation (Anon., 1981). The newer types of casing have greater reproducibility of physical dimensions (particularly diameter) and behaviour on cooking than the corresponding natural casings and are an excellent example of adding value to a low value waste.

Materials of this general type should also be applicable to the manufacture of edible packaging materials for products other than sausage, or edible labels which could overcome one of the persistent problems of the meat industry—marking or labelling of meat without the use of inks, chps or other materials which present 'foreign body' problems. Gelatine labels have already been developed in Australia for this purpose but are somewhat limited in application.

Social aspects of the use of upgraded wastes

The discussion so far has dealt primarily with technological and economic aspects of the use of upgraded wastes, but this is only part of the total picture. Food products must also be clean, wholesome and socially acceptable to the consumer. Many special requirements apply to meat and meat products.

S A F E T Y A N D NUTRITION

Detailed consideration must clearly be given to any proposal to change the nutritional status of meat and meat products in Western diets. It is generally accepted that the protein content of the typical British diet is more than adequate and that there is latitude within a mixed diet for a nominal substitution of meat by other foods. The possibihty of major substitution, however, or the introduction of totally new foods raise many problems and these have been considered at length by various committees (Food Standards Committee, 1975; 1980).

(Oliphant, 1982). Losses of fat from the finished product can be reduced considerably by this technique as compared with comminution of the same ingredients in a single processing operation.


L A B E L L I N G

Part of the overall social problem hes in communication between the manufacturer and consumer regarding the nature of foods on the market. The modern consumer is looking for more information and proposed new British Food Labelling Regulations (Anon., 1980), in line with the corresponding EEC Directive, make a number of provisions to that end. Table 14.4 summarizes the main requirements regarding the product

Table 14.4 SOME LABELLING R E Q U I R E M E N T S FOR N A M E S OF UK R E T A I L M E A T P R O D U C T S (LABELLING OF F O O D R E G U L A T I O N S , 1980 ( A N O N . , 1980) P R O P O S E D M E A T P R O D U C T R E G U L A T I O N S (MINISTRY OF A G R I C U L T U R E , FISHERIES A N D F O O D , 1981))

All products to carry:

Name prescribed by law (if any)

or

Customary name (not including trade name or fancy name)

or Name which is sufficiently precise

—to inform purchaser of true nature of product and —to enable it to be distinguished from products with which it could be confused and —if necessary includes a description of its use including indication of —treatment to which the product has been subjected, if omission could mislead the

purchaser

—presence of excess added water in cuts or joints of meat or cured meat^

^Details still under discussion.

names of meat and meat products. Similar principles apply to names used in ingredients lists except that the generic names 'meat', 'other meat' and 'offal' may also be used in appropriate cases. In general the names of a product must be a 'customary name' or an indication of its 'true nature' together with various items of supplementary information. This is relatively straightforward with well established products with widely understood identity but with newly developed products it can be fraught with difficulty. There are particular problems if the product is identified in the mind of the consumer with a low grade or offensive material.

Broader health aspects of the use of upgraded wastes have been reviewed by van der Wal (1983). A special problem with many such materials is that they are likely to contain more connective tissue than carcase meat and have a relatively lower content of essential amino acids. The nutritional implications have been discussed most recently by Laser-Reutersward et al. (1982); after mixing lean beef and pig skin in varying proportions and feeding the mixture to rats in nitrogen-balance studies they concluded that 'even if 25-40% of the protein in a meat product is collagen the protein nutritional value of that product would be adequate as the sole source of protein even for weaning children. In a mixed diet the practical levels of collagen in industrial meat products (15-30%) will be of negligible nutritional significance'.

R.S. Hannan 207

OTHER SOCIAL PROBLEMS

But even if we have a highly nutritious, properly labelled product selUng at a competitive price there may still be problems with a complex of other social factors. Strongly held prejudices can be reflected in social restrictions and arbitrary legislation no matter how extensive the technical evidence. Any proposal to make novel use of a waste material in meat products must take full account of the special image of meat and meat products in the mind of the typical consumer. Although the industry as a whole must be constantly alive to the need for efficiency and economy of working there must be limits to its readiness to embrace innovations which could be prejudicial to that image. Upgrading a waste material without upgrading its image might even be counter productive in attracting a stigma of 'adulteration'.

References

AALBERSBERG, W.J. (1981). Pwc, Inst. Food Sci. Technol, 14, 172 ANON. (1953). Offals in Meat Products Order 1953. HMSO, London ANON. (1980). Food Labelling Regulations 1980. HMSO, London ANON. (1981). Meat Ind., 27, (6), 81 BITTEL, R.J., GRAHAM, P.J., YOUNG, R.W. and BOVARD, K.P. (1981). / . Fd

Sci., 46, 336 BREKKE, C.J. and EISELE, T.A. (1981). Food Technol., 35, (5) 231 BRIEDENSTEIN, B.C. (1982). Intermediate value beef products. National

Livestock & Meat Board, Chicago DENG, J.C., TOLEDO, R.T. and LILLARD, D.A. (1981). / . Fd Sci., 46, 1117 DREPPER, G., DREPPER, K. and LUDWIG-BUSCH, H. (1981). FleischwirtS., 61,

1393 EISELE, T.A. and BREKKE, C.J. (1981). J. Fd Sci.. 46, 1095 EVANS, E.w. (1982). In Developments in Food Proteins I. Ed. by Hudson,

B.J.F. p.31. Applied Sci. Pubhshers, London

It is, for example, difficult to see how the true nature of derivatives of blood and bone could be indicated without the use of the words 'blood' and 'bone' on labels, with probable emotive reactions by the British consumer. If, however, as already indicated, blood can reasonably be regarded as an offal the new Labelling Regulations would allow its use in ingredients lists of cooked products to be covered by the generic term 'offal'. The problem is even more complicated in international trade since blood is 'meat' on the continent of Europe and the term 'animal protein' is more or less synonymous with 'meat protein'. Retrospective analysis for true meat protein content in the presence of non-meat protein involves many difficulties, especially if the latter are of animal origin (Llewellyn, 1982) and it may well prove that methods based on the presence of 3 methyl-histidine as developed by Lawrie and colleagues (Hibbert and Lawrie, 1972; Poulter and Lawrie, 1980) will prove to be the only effective recourse.


E V A N S , M.T.A. and G O R D O N , J.F. (1980). In Applied Protein Chemistry, Ed. by Grant, R.A., p.31. Applied Sci. Publishers, London

F O O D S T A N D A R D S COMMITTEE (1972). Report on offals in meat products, HMSO, London

F O O D S T A N D A R D S COMMITTEE (1975). Report on novel protein foods, HMSO, London

F O O D S T A N D A R D S COMMITEE (1980). Report on meat products, HMSO, London

G A U L T , N.F.S. and L A W R I E , R .A . (1980). Meat ScL, 4, 167 H A N D , L.W., C R E N W E L G E , C.H. and TERRELL, R.N. (1981). / . Fd Sci,, 46,

1004 H A N N A N , R.S. (1975). In Meat, Ed. by Cole, D.J.A. and Lawrie, R.A.,

p.205. Butterworths, London H I B B E R T , I. and L A W R I E , R . A . (1972). / . Fd TechnoL, 7, 333 HICKSON, D . W . , DILL, C.W., M O R G A N , R.G. , S W E A T , V . E . , S U T E R , D . A . and

C A R P E N T E R , Z.L. (1982). J, Fd 5d. , 47, 783 JOBLING, A . and JOBLING, C A . (1983). In Upgrading Waste for Feeds and

Food, Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie, p. 183 Butterworths, London

L A S E R - R E U T E R S W Ä R D , Α . , ASP, N.-G. , BJÖRCK, I. and R U D É R U S , H. (1982). / . Fd Technol,, 17, 115

L A W R I E , R . A . (1981). Proc, Inst, Food Sci, Technol., 3, 118 L A W R I E , R .A . and L E D W A R D , D . (1983). In Upgrading Waste for Feeds and

Food. Eds, D.A. Ledward, A.J. Taylor and R.A. Lawrie, p. 163 Butterworths, London

L L E W E L L Y N , J.w. (1982). In Developments in Food Proteins—I. Ed. by Hudson, B.J.F., p. 171 Applied Sci. Publishers, London

M A C F A R L A N E , J.J., SCHMIDT, G.R. and T U R N E R , R.H. (1977). / . Fd Sci., 42, 1603

MINISTRY O F A G R I C U L T U R E , FISHERIES A N D F O O D (1981). Proposals for Meat Product Regulations

N A T I O N A L F O O D S U R V E Y COMMITTEE (1982). Household Food Consumption and Expenditure: 1980. HMSO, London

N E W M A N , D . (1982). Meat, 55, 9 N E W M A N , P.B. (1983). In Upgrading Waste for Feeds and Food. Ed. by

D.A. Ledward, A.J. Taylor and D.A. Lawrie, p. 93. Butterworths, London

O L I P H A N T , G. (1982). Personal communication P O U L A N N E , E. and R U U S U N E N , M. (1981). Meat Sci., 5, 371 P O U L T E R , N .H. and L A W R I E , R .A . (1980). Meat ScL, 4, 15, 21 Q U A G L I A , G.B. and MASSACCI, A. (1982). / . Sci. Food Agric, 33, 634 R A N K E N , M . D . (1980). In Applied Protein Chemistry, Ed. by Grant, R.A.

p. 169. Applied Sci. Publishers, London R A N K E N , M . D . (1982a). Report of the EEC Scientific Group Meeting,

Copenhagen 28-29 October, 1981 (In press) R A N K E N , M . D . (1982b). Reforming of Meat. In press R I C H A R D S , S.P. (1982). Investigations of the use of animal by-products. PhD

Thesis, Brunei University SAMEJIMA, K., ISHIOROSHI, M. and YASIN, T. (1981). / . Fd ScL, 46, 1412 S A T E R L E E , L .D . (1981). Food Technol., 35, (6), 53

R.S. Hannan 209

SCHMIDT, G.R. , M A W S O N , R.F. and SIEGEL, D . G . (1981). Food TechnoL, 35, (5), 235

S C H U T , J. (1980). The role of milk proteins in foods. Dutch Dairy Bureau, Rijswijk, The Netherlands

S H A W , R. (1974). Proc. 10th Anniv. Sympos. Inst. Food Sci. Technol., p.15 T E R R E L L , R .N . , C R E N W E L G E , C.H. , D U T S O N , T.R. and SMITH, G.C. (1982). / .

FdScL, 47, 711 T Y B O R , P.T. , DILL, C.W. and L A N D M A N N , W.A. (1973). / . Fd Sci., 38, 4 V A N D E R W A L , P. (1983). In Upgrading Waste for Feeds and Food. Ed. by

D.A. Ledward, A.J. Taylor and R.A. Lawrie, p.275. Butterworths, London

W I S M E R - P E D E R S O N , J. (1979). Food Technol, 33, (8), 76 Y O U N G , R.H. (1980). In Developments in Meat Science—I. Ed. by Lawrie,

R.A., p. 145. Applied Sci. Publishers, London

15

Introduction

Burgeoning world population growth has resulted in serious inequalities in the available food supply. Malnutrition is widespread; the most critical nutrition problem is protein-calorie malnutrition of young children and pregnant and lactating women, particularly in the so-called developing nations in which two-thirds of the world's population live (Hardin, 1979). The basic cause is a diet which is likely to be low in protein, provide insufficient energy and often be marginal in other nutrients.

High quality animal proteins are scarce and priced out of reach of the general population in many of the developing nations where food shortages are most acute. Therefore, vegetable proteins, particularly oilseed proteins, have become attractive alternative sources of low-cost protein foods. Among the oilseeds, soybean is the most promising because it contains high levels of essential amino acids and manufacturing techniques for the production of a wide variety of food products are already well developed (Lischenko, 1979).

Milk production is almost always inadequate to meet the nutritional needs of the population in developing countries. Expanding the available milk supply either by extending it with vegetable protein or by preparation of milk analogues and other refreshing drinks from vegetable protein or previously wasted animal protein such as that from cheese whey, provide acceptable and economical alternatives to expensive animal proteins.

Numerous protein sources have been evaluated in beverages. Beverages containing cheese whey or cheese whey protein concentrate, soybean, or groundnut proteins are available commercially in some parts of the world. Beverages with proteins from cotton-seed, sunflower, rape, sesame, leaf protein concentrate, single-cell protein, and flsh protein concentrate have been developed in the laboratory. In some cases they have undergone large-scale consumer trials; but, in general, they still remain in the developmental stage.

The purpose of this chapter is to describe some of the many research and development activities directed toward production of nutritious beverages from recovered proteins. *Agricultural Research Service, US Department of Agriculture

211

UNDERUTILIZED PROTEINS FOR BEVERAGES

V.H. HOLSINGER Eastern Regional Research Center"^, Philadelphia, Pennsylvania, USA

212 Underutilized proteins for beverages

Beverages with cheese whey and whey fractions

The economical salvage of cheese whey, a by-product of cheese manufacture, has become a serious problem to the dairy industry in the USA because of stringent antipollution regulations. The estimated US production of fluid sweet and acid wheys in 1981 was 18.95 thousand million kg, only 52.7% of which was utilized (Whey Products Institute, 1982). Although whey contains only about 6.5% solids, these sohds represent 50% of the nutrients of milk; small amounts of high quality protein equivalent to that of egg are present, but fluid whey is essentially a dilute solution of lactose {Table 15.1).

Table 15.1 C O M P O S I T I O N O F W H E Y S O L I D S ( A F T E R H O L S I N G E R , 1976)

Component Sweet whey Acid whey

Total protein 11.5% 11.4% Lactose 74.4% 66.8% Ash 7.4% 10.2% Lactic acid < 1 . 0 % 9.6% Fat 2.7% < 1 . 0 % pH 6.5 4.7

The use of cheese whey or its fractions as a base for the manufacture of beverages, both alcoholic and nonalcoholic, has been attempted for many years with some commercial success in Europe (Holsinger, Posati, and DeVilbiss, 1974). Perhaps the most successful whey drink is Rivella*, a sparkling crystal clear herbal infusion in deproteinized whey that appeared on the market in 1952 in Switzerland (Sush, 1956). This product is currently sold in most of western Europe, being promoted as a therapeutic tonic. In 1977, its production was absorbing 75000 i of whey per day (Anon., 1978a). Rivella resembles ginger ale in flavor and appearance; it has been pasteurized, and must be refrigerated after opening. Small amounts are being imported into the USA; 0.3 € sells for about $0.89.

Ύοο Hoo' beverage has occupied a niche in the soft drink market in the USA for many years. Originally sold as a chocolate-flavored, shelf-stable, sterilized drink, based on skim-milk, it has been reformulated so that its principal ingredient after water is sweet dairy whey. In taste and mouth feel, it more closely resembles chocolate milk than the carbonated soft drinks with which it competes.

Alcoholic beverages produced from whey include beer-like drinks that have been marketed with some success in Russia and a whey 'champagne' in Poland (Holsinger, Posati and DeVilbiss, 1974). Yang et al. (1977) developed a process for making wine from whey which underwent commercial trials in the USA. By adding dextrose to the whey, they could use a nonlactose fermenting organism to produce the alcohol. A recent re-evaluation of whey as an ingredient in beer was carried out on small commercial scale; results showed that beer meeting trade specifications for

*Reference to brand or firm name does not constitute endorsement by the US Department of Agriculture over others of a similar nature not mentioned

ν,Η. Holsinger 213

flavor and chemical composition could be brewed readily with whey permeate from an ultrafiltration operation (Reaves, 1980).

A conventional type nonalcohohc orange-flavored carbonated beverage fortified with 1.5% whey protein was test marketed in Brazil in 1971 (Anon., 1973a). Since the whey protein concentrate contained milk-fat, lactose, and whey salts, these were also present in the drink. Although this product resembled the carbonated soft drinks that dominate the US market, it did not advance beyond the test market stage.

Research leading to the development of beverages with whey components has generally taken three forms: fermented deproteinized alcoholic type drinks with less than 0.5% protein; nonalcoholic snack beverages, including carbonated soft drinks, fruit ades*, and drink powders with 0.5-1.0% protein; and imitation milks, dietary supplements and hquid breakfasts supplying 1.0-3.5% protein.

The most efficient means of producing a whey-based beverage is to use whole whey. Whey flavor and saltiness of whole whey have been problems in producfion of a successful beverage in the past. The 'whey taint', particularly that of acid whey, is most compatible with citrus flavors, especially orange. Several experimental citrus-flavored beverages have been developed from sweet and acid wheys for which high consumer acceptabihty is claimed (Holsinger, Posati and DeVilbiss, 1974). Acid whey powder has also been successfully used to fortify reconstituted frozen orange juice; the finished product had 2.5 times more protein than the original (Kosikowski, 1968).

More recently, it has been reported that acid cheese whey produced by the direct acidification process in cottage cheese manufacture does not have the characteristic off-flavor that is present in whey from cottage cheese made by the culture method; orange, lemon-lime and tomato-flavored beverages have been prepared and rated acceptable by a consumer panel (DeMott, 1975; DeMott, Helms, and Sanders, 1977).

The rapid improvement in commercial scale ultrafiltration and reverse osmosis technology has resulted in increased availability of a whole spectrum of whey fractions and modified wheys for ingredient use. Consequently, recent whey beverage research has been directed towards utilization of whey fractions as beverage bases. Much research time has been devoted to the production of various types of alcoholic beverages, including wines of 10-12.5% alcohol content (Kosikowski and Wzorek, 1977; Lang and Lang, 1979). The presence of protein is undesirable in this type of beverage; the whey is deproteinized by heat treatment or by membrane fractionation, and the whey permeate is used for the beverage base.

Whey protein concentrates produced by membrane fractionation or electrodialysis offer a good starting material for production of protein fortified beverages of improved nutritive value. An example of such a product is an orange juice beverage with 3% protein prepared from a whey protein concentrate containing 74% protein. The finished drink may be frozen or canned hot for refrigerated storage; it may also be processed further into a powder (Anon., 1978b).

*Non-carbonated fruit-flavored drinks


COTTAGE CHEESE WHEY

Clarif ication Ul t ra f i l t ra t ion

PRE-CONCENTRATE

Clori f icat ion Gel f i l t ra t ion

HIGH PROTEIN ELUATE

Iciar i f ication Evaporation Spray or stielt dry

LOW MOLECULAR WT FRACTION

DRY PRODUCT 81.4% P ro te in 10.2% Lactose 1.51% Asli

Figure 15.1 Isolation of undenatured whey protein concentrates from acid cheese whey

A large-scale process was developed to permit the isolation of undenatured proteins from cottage cheese whey {Figure 15.1) (Holsinger et al., 1975). The whey proteins were preconcentrated by ultrafiltration at pH 4.7; low molecular weight materials were removed from the concentrate by gel permeation on Sephadex G-25. The high protein eluate thus obtained was condensed under vacuum and spray dried. To produce a highly soluble material, centrifugal clarifiers were used to remove residual milkfat and insoluble material during the purification steps. The amount of protein lost under the experimental conditions used was about 50%. With this process, dehydrated isolates containing 75-90% protein were routinely prepared. Whey proteins are exceptionally rich in the essential amino acid lysine; isolates prepared by this procedure had over 90% of the total lysine in a

The magnitude of the soft drink market in the USA has attracted the interest of the whey protein processors as an outlet for their products. In 1981, 141.7 ( of soft drinks were consumed per capita, representing an estimated retail value of over $23 billion (Anon., 1982a). 'Ade'-type beverage powders, consumed at a rate of 18.4 ^/capita after reconstitution, had estimated sales worth more than $1.5 billion. Because the appeal of these drinks to young people is strong, and their average cost in July of 1982 was estimated to be $2.50/gallon compared to $1.83/gallon for fresh milk (Anon., 1982b), the soft drink companies have been under great pressure from nutritionists and consumers to improve the nutritive value of their drinks.

One approach to increasing the nutritive value of soft drinks is to fortify them with whey proteins without detectable change in flavor or appearance. This means that the proteins used must be completely soluble at the acid pHs characteristic of most soft drinks and yield a clear solution when dissolved. Researchers at the US Department of Agriculture demonstrated that proteins isolated from acid whey had the functional characteristics required for soft drink fortification (Holsinger et ai, 1973a).

V.H. Holsinger 215

Table 15.2 S O L U B I L I T Y O F C O T T A G E C H E E S E W H E Y P R O T E I N S

P A S T E U R I Z E D A T 73.9 ° C F O R 20 s B Y A H I G H - T E M P E R A T U R E S H O R T - T I M E P R O C E D U R E ( A F T E R H O L S I N G E R , 1980a)

Protein concentration % of solids

11.0 45.4 77.6

% of protein nitrogen insoluble at pH 5.1

Raw Pasteurized % Total solids pasteurized

1.8 9.2 8.6 10.6 24.0 26.1

6.44 12.5 3.90

Carbonated beverages prepared with conventional beverage ingredients and fortified with added whey protein at a level up to 1% by weight of the total weight of the beverage, maintained clarity, color, anf flavor during 203 days' storage at room temperature. Carefully prepared spray-dried whey protein concentrates could also be used to fortify 'ade'-type powdered products. The sucrose crystals in presweetened drink powders were ideal for dispersing whey protein concentrate powder by dry-blending; no problems with foaming on reconstitution were encountered. Sensory evaluations of both types of beverages showed that trained judges of dairy products could not detect the presence of whey proteins when either type of drink was fortified with 1% protein or less (Holsinger et al., 1973a).

Although empirical studies showed that soft drink fortification with whey protein concentrates was feasible, more fundamental studies were needed. Holsinger etal. (1973a) studied the solubility and stability of whey proteins under acid conditions to provide some of the additional data required.

When 1% protein solutions were heated for 6 h at 80 °C, solution clarity was not impaired, but some structural changes had occurred, since 37% of the protein present precipitated when the pH was shifted to 4.7. Increased stability against heat insolubilization under acid conditions was conferred by some of the soft drink ingredients. The addition of sucrose reduced protein insolubilization by one-half but sodium saccharin had no effect. The type of acidulant used in the beverage also influenced the rate of protein insolubilization; at pH 2.68, greater protection was conferred by phosphoric acid than by citric acid.

nutritionally available form as measured by chemical means, one indication that the quality of the dried products was excellent (Holsinger et al., 1973b).

Investigation of the causes of formation of insoluble protein during processing showed that the heat required for pasteurization of the high protein concentrate prior to evaporation was responsible {Table 15.2). High-temperature short-time pasteurization was sufficient to insolubilize 26.1% of the protein present in a protein concentrate with 77.6% protein. A similar amount of insolubilized material (24.0%) was measured in a concentrate with 45.4% protein; only 10.6% of the protein was insolubilized when raw whey was pasteurized (Holsinger, 1980).


Results showed that most soft drinks with pH below 4 could be successfully fortified with undenatured cheese whey proteins. Precipitation problems could be encountered with the protein fortification of drinks containing natural fruit juices high in tannins; some caramel colorings might also act as protein precipitants. However, with only slight formula modifications, a stable system should be attained (Holsinger et ai, 1973a; Holsinger, 1978).

A US patent, which covers a process for manufacture of a whey protein concentrate suitable for incorporation into beverages, has been granted to a major soft drink company. The process involves:

(1) passing the whey through a filter of diatomaceous earth; (2) subjecting the filtrate to ultrafiltration with water injection, and (3) passing the concentrate over a strongly acidic cation exchange resin to

produce a whey protein concentrate low in microbial count, low in mineral saUs, and having a pH in the range 2.7-3.6.

The whey is heated to 37 °C between steps (1) and (2) and cooled to about 5°C before step (3). The concentrate is then dried to yield an acid-soluble powder suitable for beverage use (Malaspina and Moretti, 1975). The technological problems involved in the fortification of soft drinks have been reviewed (Fenton-May, 1975).

Beverages with soybeans

The most important protein raw material for the manufacture of imitation milks is the soybean because of the nutritive value of the protein. In many parts of the world, the soybean is a staple food and a variety of beverages with soybean proteins have been developed. Some are designed as infant formulas, some for vegetarian diets, and some specifically for developing countries; problems of flavor, acceptability, cost, texture, and product stability have been encountered.

A comprehensive review of the literature describing research and development activities directed toward improved utilization of the soybean as a protein resource is beyond the scope of this chapter. Detailed information about the chemistry, technology, and nutritive value of soybean protein may be found in Smith and Circle (1978), and Wilcke, Hopkins and Waggle (1979).

The traditional Asian preparation of soy milk is a cold water extraction procedure of water-soaked and ground soybeans. The whole beans are soaked overnight, wet ground, slurried in water, filtered, and the extract boiled to inactivate the trypsin inhibitor that reduces the nutritive value of the raw bean (Smith and Circle, 1972). The soy milk may be used to prepare other foods such as tofu, yuba, or consumed directly.

The acceptance of soy beverages produced by the tradifional process has been limited in non-Oriental countries because of their characteristic flavor and odor (Hand et al., 1964). Soybeans ground in the presence of water possess a characteristic flavor that has been described as beany or painty and is generally undesirable. Wilkins, Mattick and Hand (1967) concluded that the undesirable flavor was the result of lipoxygenase enzyme activity.

V. Η. Holsinger 217

I Γ ™ τ 1 A l d e h y d e s K e t o n e s A l c o h o l s Others

0

Η Η II Η Η HC — C — C — C=-=C

Η Η Η Ethy l v iny l k e t o n e

Figure 15.2 Mechanism of lipoxygenase activity

The enzyme system is heat labile so heat treatment at an early processing stage inactivates lipoxygenase.

The most complete research study of a soybean beverage has been conducted at the University of Illinois. Nelson, Steinberg and Wei (1975, 1976) developed a process to prepare a soy beverage from whole soybeans, with or without hulls.

Whole soybeans were tenderized by blanching in boiling water, usually with prior soaking, while cotyledons were tenderized only by blanching. Usually sodium bicarbonate was present in the soak and blanch solutions. After blanching, the soy solids were milled with additional water in a hammermill, followed by heating the slurry to 93.3 °C, and homogenizing at 246.1 kg/cm^ (3500 psi). The resulting beverage base was diluted with water to the desired total solids level, sweeteners and flavoring were added, and the beverage pH was adjusted to about 7.2. The formulated

Lipoxygenase acts specifically on fatty acids that contain a cis-cis-penta-1,4-diene system, resulting in the production of cis-trans-diene hydroperoxides that have been isolated and identified (Lao, 1971). In an extensive investigation, Mattick and Hand (1969) isolated 80 volatile compounds claimed to be the result of lipoxygenase activity; 40 compounds were identified. The majority of the compounds were aldehydes, ketones, and alcohols; ethyl vinyl ketone was identified as a primary component of soybean off-flavor {Figure 15.2).

R R,

I Γ HC CH c i s - c i s p e n t a d i e n e s y s t e m

II " II HC — C — C H

Η L i p o x y g e n a s e

[Free r a d i c a l s ] • R

I HC

II " HC — CH c i s - t r a n s d i e n e h y d r o p e r o x i d e s

I " HC — C — R,


Figure 15.3 Shelf stable soy milk commercially produced in Japan. (Photograph by A . Rivenburgh)

* Reference to brand or firm name does not constitute endorsement by the US Department of Agriculture over others of a similar nature not mentioned.

beverage was heated to 82.2 °C, rehomogenized, and bottled. Blanching the soybeans before grinding inactivated lipoxygenase, so the beverage was free from lipoxygenase-induced off-flavors. The beverage had good mouth feel and suspension stability.

Kuntz (1977) carried out an extensive investigation of the mechanisms affecting organoleptic quality of Illinois soy beverage. Chalkiness in the finished beverage was related to the presence of suspended insoluble material; removal of the particulates reduced chalkiness intensity to the imperceptible. Greatest correlafion with chalkiness was found with that fraction of insoluble material retained by a 150 mesh screen. Astringency and cereal flavor were also reduced significantly by removal of the beverage insolubles; the effect of desludging on bitterness was not significant. Analysis showed that the bulk of the sludge was particulate fiber fragments. There has been some commercial interest in the Illinois process.

Lo (1971) patented a process for preparing a soy beverage from full fat flour. The flour was prepared from soybean cotyledons that were flaked, extrusion cooked, toasted, and ground to 270-300 mesh. The beverage was prepared by slurrying with water and adding carrageenan as stabilizer, heat treating to 82-88 °C, holding for 30 min, and homogenizing at 562.6 kg/cm^ (8000 psi). Some of the insoluble carbohydrate was removed by continuous centrifugation. Clarification was followed by homogenization at 175.8 kg/cm^ (2500 psi), bottling and sterilization.

Lo's process was the basis for the first successful commercial effort for the promofion of soy milk. The product is marketed in Hong Kong as a soft drink under the trade name 'Vitasoy'*; in 1974, it passed the leading carbonated beverage to become Hong Kong's top selling soft drink with sales of 150 million botfles per annum (Shurtleff, 1981).

'Vitasoy' is being imported into the USA from Hong Kong to satisfy the large Oriental market for a traditional soy milk. It sells for about $0.35 and is marketed as a shelf stable product in Tetra-pak* or Britpak* containers.

V.H. Holsinger 219

F l a s h cool to 9 0 ° C

H o m o g e n i z e 1 8 0 k g / c m 2

Chi l l

P a c k a g e a s e p t i c a l l y

Figure 15.4 Manufacturing scheme for shelf stable ultra-high-temperature processed soy milk

mesh, centrifuged, and the supernatant standardized to 12% total solids (3.2% fat), including added sweetener and flavoring. The soy milk is then sterilized at 140 °C for 1 min to inactivate the trypsin inhibitor, deodorized by flash cooling in a vacuum pan, homogenized single-stage at 180 kg/cm^, chilled and aseptically packaged. The aseptic packager handled 9000 *Reference to brand or firm name does not constitute endorsement by the U S Department of Agriculture over others of a similar nature not mentioned

'Vitasoy' as presently marketed contains water, soybean solids, malt, sugar, rapeseed oil, sodium bicarbonate and salt. A 243 ml serving provides 110 calories, 4 g of protein, 18 g of carbohydrate and 3 g of fat (Levitón, 1981).

Soy milk has become very popular in Japan in the past five years. In 1981, I visited the two-year old Okazaki Marusan Company factory where shelf stable ultra-high-temperature processed soy milk is manufactured. Both plain and flavored soy milks are being marketed in cans (for vending machines), retort pouches, and 200-ml Tetrabriks* {Figure 15.3). The product is made by a novel process based on the original research of Mustakas et al. (1971; 1972), whereby the dehuUed beans are steam injected in a tubular screw conveyer to inactivate lipoxygenase {Figure 15.4). The heated beans are ground with water, hammermilled to 100

W h o l e s o y b e a n s

D e h u l l

S t e a m i n j e c t U O ' C

W a t e r

Grind, 60°C, H a m m e r m i l l , 1 0 0 - n n e s h

C e n t r i f u g e

S o y m i l k

F o r m u l a t e

Heat U O ° C , 50 s


Beverage powders with cheese whey and soybeans

The limiting amino acids in soybean protein are methionine and cystine. Because cheese whey is particularly rich in lysine and contains significant amounts of methionine and cystine, their combination offers good potential for development of beverages of high-nutrition to cost ratio (Table 15.3).

Table 15.3 ESSENTIAL A M I N O ACIDS OF S O Y B E A N PROTEIN A N D SWEET C H E E S E W H E Y PROTEIN C O M P A R E D TO THE F A O R E F E R E N C E P A T T E R N

Amino acid FAO reference^ Defatted soybean meal^ (g amino acid/100 g protein)

Sweet whey''

Lys 5.5 6.9 8.7 His — 2.5 1.6 Thr 4.0 4.3 6.1 Cys + Met 3.5 3.2 3.5 Val 5.0 5.4 5.6 l ieu 4.0 5.1 6.0 Leu 7.0 7.7 9.8 Tyr + Phe 6.0 8.9 5.3 Trp LO 1.3 2.0

"After Anon. (1973b). ^After Rackise iö / . (1961). After Posati and Orr (1976).

Loewnstein and Paulraj (1972) demonstrated that a powder, made by co-precipitating and drying defatted soybean flour and whey protein to a final blend of three parts soy to one part whey protein, was clearly superior to soy protein alone in promoting the growth of rats.

Sasaki and Tsugo (1953) described the preparation of synthetic milk powder from whey arid soybeans by extraction of the beans with hot whey. This research led Guy, Vettel and Pallansch (1969) to develop a process for spray drying a mixture of full fat soy flour and fluid sweet cheese whey to yield a free flowing powder of good nutritive value suitable for beverage use. The powder contained 67% sweet whey solids and 33% full fat soybean flour. Another powder containing 55% sweet whey sohds, 28% soy flour and 17% corn oil was also prepared by this process.

200-ml units/h. The shelf-hfe of the product was only 60 days because of flavor deterioration; to maintain best flavor, the company recommends that the product be stored below 10 °C, even though sterile. In 1981, the soy milk was selling for 70 yen, about $0.30.

A program, carried out by the Government of Mexico and private industry, was established in the State of Chihuahua, Mexico, for the purpose of manufacturing high nutrition, low-cost foods to supply low-income groups (Del Valle et al., 1978). Two plants were built, one to make full fat soy flour and the other to convert the flour into a line of products, including soy milk. To promote consumption of the foods, the Government conducted a state-wide advertizing campaign with great success. It was planned to extend the program to other parts of Mexico in the future.


Ingredient Formulation (%)

Using only Using only full fat flour defatted flour

Sweet cheese whey sohds 41.3 41.3 Soybean flour 36.5 29.7 Soybean oil 12.2 19.0 42 DE^ corn syrup solids 9.0 9.0 Added vitamins and minerals 1.0 1.0

^DE = Dextrose equivalent.

homogenized double-stage at a pressure greater than 141 kg/cm^ for the first stage and 35.2 kg/cm^ for the second stage, pasteurized at 70.4 ®C for 25 s, condensed under vacuum, and spray-dried using the techniques developed by Guy, Vettel and Pallansch (1969). One per cent of a vitamin-mineral pre-mix was dry-blended into the powder to increase the nutritive value.

The proximate composition of WSDM showed that the powder was a high-protein high-fat, high-calorie product which, upon reconstitution, simulated milk in many ways {Table 15.5). The bulk of the protein was supplied by the soybean flour, whereas the carbohydrate came mainly from the lactose in the whey.

To make the powder, hquid sweet whey was combined with full fat soy flour and then pasteurized continuously at 77°C for 20s, homogenized in two stages with pressures of 387 and 38.7 kg/cm^, condensed to 40-50% total solids in vacuum, and conventionally spray-dried. The formulation of the product was flexible and allowed the addition of sweetener and flavorings so that a beverage containing 2.7% protein was obtainable. Homogenization reduced the amount of settling of the soybean solids in the reconstituted beverages, yielding a suspension more like milk. Concentration under vacuum and spray drying reduced the bean flavor of the soybean flour. The addition of citrus or cherry-vanilla flavors to the drink increased acceptability (Guy, Vettel and Pallansch, 1968).

In the late spring of 1973, a joint effort by the US Department of Agriculture and the US Agency for International Development was begun to develop a nutritious beverage powder mix specifically formulated as a dietary supplement for preschool children receiving inadequate protein. It was not intended to serve as the sole source of food. When reconstituted with water, the beverage was to supply all the nutrients of whole milk; it was also to be cheaper than nonfat dry milk to produce and ship overseas to meet Food-for-Peace commitments in developing countries (Holsinger et al, 1977b).

Processing parameters leading to the development of the new product were investigated in great detail, some of which are described below.

The new milk analogue, named Whey-Soy-Drink-Mix (WSDM), was formulated from sweet cheese whey, soy flour (full fat, PDI 35-45, or fully toasted defatted, NSI10-30), soybean oil, and 42 dextose equivalent (DE) corn syrup solids {Table 15.4). After formulation, the mixture was

Table 15.4 F O R M U L A T I O N OF W H E Y SOY DRINK MIX


Table 15.5 A P P R O X I M A T E COMPOSITION O F W H E Y - S O Y DRINK MIX

Component Percent

Moisture 2.7 Protein (total Ν x 6.25) 20.0 Fat 20.0 Ash 6.1 Fiber 1.2 Carbohydrate 50.0

WSDM was readily produced with conventional milk plant equipment by various procedures {Table 15.6). Soybean oil, soy flour and corn syrup solids were wet blended into fluid whey or whey precondensed to 16% total solids. After two-stage homogenization and flash pasteurization, water evaporation under vacuum was necessary to remove some of the beany flavor associated with the use of soy flour in the formulation. The low-heat process employing a single effect evaporator did not lend itself too readily to WSDM production in milk drying plants in the USA; the high-heat process, with a maximum temperature of 68 °C during vacuum evaporation permitted use of the multiple effect evaporators common in drying plants. Both evaporative processes yielded products with good flavor and good nutritional quality.

A 16 °C 'flash' in a vacuum chamber increased the total solids from 43-45%, reduced beany flavor, and met heat treatment parameters defined by the evaporative processes without affecting nutritional quality. 'Flashed' products were prepared with both the full fat soy flour and the

Table 15.6 A L T E R N A T I V E PROCESSES FOR WHEY-SOY DRINK MIX

Evaporative processes

Low-heat (single High-heat (double 'Flash' process effect evaporator) effect evaporator)

Whey (16% total solids) Whey (16% total solids) Whey (23% total solids) Soy flour Soy flour Soy flour Soybean oil Soybean oil Soybean oil Corn syrup solids Corn syrup solids Corn syrup solids Wet-blend in vat (31.5% TS) Wet-blend in vat (31.5% TS) Wet-blend in vat (42% TS) Heat to 43 °C Heat to 43 °C Heat to 43 °C Homogenize Homogenize Homogenize Two stages—175.8 kg/cm^ Two stages—175.8 kg/cm^ Two stages—175.8 kg/cm^

— 38.7 kg/cm' — 38.7 kg/cm^ — 38.7 kg/cm^ Pasteurize 7 9 ° C , 2 5 s Pasteurize 7 9 ° C , 2 5 s Pasteurize 7 9 ° C , 2 5 s Evaporate to 45% TS Evaporate to 45% TS Vacuumizer to 45% TS 60 °C maximum First effect 68 °C (0.93 kg/cm^ vacuum)

Second effect 49 °C In a t79°C Out at 46 °C

Spray dry Spray dry Spray dry 141 °C Inlet 141 °C Inlet 141 °C Inlet

82 °C Outlet 82 °C Outlet 82 °C Outlet Cool to 43 °C Cool to 43 °C Cool to 43 °C Dry blend Dry blend Dry blend Vitamin-mineral premix Vitamin-mineral premix Vitamin-mineral premix Pack for shipment Pack for shipment Pack for shipment


Table 15.7 EFFECT OF V A C U U M 'FLASH' D E O D O R I Z A T I O N ON N U T R I T I O N A L Q U A L I T Y OF W H E Y SOY DRINK MIX

Sample Standardized Net protein efficiency protein ratio^ utilization

Control evaporation process full fat flour formulation 'Flash' full fat flour formulation 'Flash' defatted flour formulation

2.48

2.45 2.25*^

81%

79% 80%

^ANRC casein = 2.50. •^Significantly different (P = 0.05).

Wet-blending of the vitamins and minerals into the fluid mixture before condensing and drying would reduce processing costs. To evaluate the effect of this technique on vitamin stability in the finished powder, samples containing full fat soy flour were prepared in which vitamins and minerals were either wet-blended into the fluid mixture or dry blended into the finished powder. Although vitamin A was one of the less labile vitamins in the vitamin premix used, its deterioration was monitored because vitamin A deficiency mostly affects young children, and the xerophthalmia and blindness it leads to are serious public health problems in many developing countries.

The effect of processing on vitamin A stability in WSDM in which both vitamins and minerals were added by wet-blending and in which only the vitamins were added by wet-blending was monitored (Table 15.8). This was done to determine if the mineral mix adversely affected vitamin stability during processing. During holding at 43 ®C, considerable losses occurred with the mixture containing only vitamins showing the greater

Table 15.8 EFFECT OF PROCESSING ON STABILITY OF VITAMIN A A D D E D B Y W E T - B L E N D I N G TO W H E Y SOY DRINK MIX F O R M U L A T E D FOR FULL F A T F L O U R

Sample Vitamins Vitamins and minerals wet-blended wet-blended

Concentration % Concentration % USunits/100 g Lost US P^ units/100 g Lost

Amount of vitamin A added 1925 — 1925 — Fluid mix, held at 43 °C 1041 46 1251 35 45% TS concentrate evaporative process 985 49 1114 42

Spray-dried powder 838 56 964 50

United States Pharmacopeia

defatted flour formulations and compared with a control prepared with full fat flour and condensed by vacuum evaporation. There was no significant difference in standardized protein efficiency ratio (PER) between samples prepared with full fat flour (Table 15.7). Although the 'flashed' sample with defatted soy flour had a significantly lower PER than the control, this value was still above the minimum PER of 1.9 required for WSDM (Holsinger et al., 1977). There was no difference in net protein utilization among the three samples.


σ> 2000 σ o (Λ

I 1500 ΟΙΟ

^ 1000

< C

Ε σ

5 0 0

Dry blend (control)

Wet blend

I 2 3 4 5 6 Storage time (months)

Figure 15.5 Stability of vitamin A in whey-soy drink mix stored at 37 °C

o 2000 o

i 1500 CL

r 1000

500

Ε ^ O

^ Dry blend (control )

o Wet blend

10 20 30 4 0 50 Peroxide value (mEq 02/kg fat)

60

Figure 15.6 Effect of peroxide development on vitamin A content of whey-soy drink mix stored at 37 °C

loss. Condensing from 31 to 45% total solids resulted in a 49% loss in the mixture containing only vitamins and a 42% loss in the vitamin-mineral containing mixture. Fifty per cent or more of the initial vitamin A content was lost after spray-drying, the sample containing only vitamins showing the greater loss.

An accelerated storage study was carried out to determine if wet blending affected the storage stability of vitamin A in the dry powder. Air-packed cans of the sample containing vitamins and minerals added by wet blending were stored at 37 °C for six months; vitamin A deterioration was compared to a control, stored under similar conditions, in which the vitamin-mineral premix was added by dry-blending {Figure 15.5). The results showed that the control had lost only 32% of its vitamin A content after six months of storage, whereas the sample in which the vitamins and minerals were wet-blended lost 80% of the vitamin A remaining after processing.

V.Η. Holsinger 225

Peroxide development was measured in the stored samples as another evaluation of oxidative stability. Peroxide development during storage was greater in the samples in which the vitamins and minerals were added wet {Figure 15.6). This may be due to the presence of iron in the mineral mix; the data suggest that the greater vitamin A deterioration during storage might be related to peroxide development in the wet-blended sample. Subsequently, to ensure that adequate levels of vitamin A and other vitamins were maintained in WSDM during storage, a 30% overage of the vitamin premix was routinely dry-blended into the finished powder.

Cottage cheese (acid) whey was evaluated as an ingredient in WSDM (Holsinger et al., 1977a). Samples prepared with neutralized acid whey or with a 1:1 mixture of acid to sweet whey showed decreased flavour score both initially and during storage. These samples contained over 1% lactate reported as lactic acid; decreased flavor score of commercially produced WSDM was correlated with the lactic acid content of the powder. If WSDM were manufactured with only small amounts of acid whey, storage stability as measured by flavor acceptability was impaired.

Substitution of either 26 to 29 DE corn syrup solids or 9 to 12 DE dextrin for 42 DE corn syrup sohds was also evaluated for effects on nutritional quality and flavor of WSDM (Holsinger et al, 1978b). Results showed that although substitution had no deleterious effects on protein quality and digestibihty during one year of storage at 37 °C, only the 26-29 DE corn syrup solids provided the required flavor qualities.

Rehydration properties of WSDM were studied in detail (Holsinger, 1980b). Response surface methodology was used to investigate optimal levels of three processing variables, homogenization pressure, total sohds of the mixture homogenized, and emulsifier level on dispersibihty, nitrogen solubility index (NSI), free fat content, particle size, and phase separation after reconstitution.

Emulsifier effects were shown to be undesirable. Emulsifier destabilized the fat emulsion of the whey-soy system during the drying process, resulting in reduced dispersibihty and increased free fat content of the powder, and increased phase separation and decreased NSI after reconstitution. The particle size was not the controlhng factor in phase separation; both phase separation and particle size were related to viscosity which was significantly influenced by all three processing variables. Phase separation was least in samples homogenized at high pressure and at high-total solids; these samples had the highest viscosities on reconstitution.

WSDM was one of the most thoroughly tested food products ever placed in the US food distribution program. Before the product was distributed abroad, acceptability studies were carried out in six developing countries representing a variety of cultures; approximately 5000 children participated in the feeding trials (Rodier et al, 1973). Their mothers were also given the beverage and asked for their opinion of the product as food for their children. In only one country. Sierra Leone, was the product found to be unacceptable; this was ascribed to the fact that the people there had had some experience with nonfat dry milk.

Over 10 million kg of WSDM were distributed abroad during the five year period 1974-78. The product, although a highly desirable dietary


Beverages from other protein sources

One of the most abundantly available protein rich foods in India and Africa is low-fat groundnut flour. Peanut protein is deficient in the essential amino acids lysine, methionine, threonine, and tryptophan; if combined with milk protein, the amino acid balance may be improved {Table 15.9). Extensive studies have been carried out, particularly in India with both milk analogues and 'toned' milks based on groundnut flour; guidelines have been issued for their manufacture (Anon., 1973).

Table 15.9 E S S E N T I A L A M I N O A C I D C O N T E N T O F P E A N U T F L O U R A N D ' M I L T O N E ' B E V E R A G E C O M P A R E D T O T H E F A O R E F E R E N C E P A T T E R N

Amino acid FAO reference'' (g amino acid/100

Peanut flour^ g protein)

' Μ iltone' protein"^

Lys 5.5 2.9 5.8 His — 1.8 Thr 4.0 2.6 4.0 Cys + Met 3.5 1.9 3.3 Val 5.0 3.8 5.6 l ieu 4.0 3.2 4.6 Leu 7.0 6.4 8.2 Tyr + Phe 6.0 8.4 4.7̂ ^ Trp 1.0 0.9 1.2

W t e r Anon. (1973b). Âfter Ayres, Branscomb and Rogers (1974). Âfter Chandrasekhara etai (1971). Ônly Phe reported.

'Toned' milk is a milk of animal origin, cow or buffalo, extended with vegetable protein, usually groundnut. 'Toned' milk is prepared by first extracting groundnut flour with dilute alkali and adjusting the extract to the isoelectric point to precipitate the protein {Figure 15.7). The protein isolate is slurried to pH 6.8 with water, buffer salts, and alkah; carbohydrate (usually sucrose), fat (a mixture of hydrogenated and refined vegetable oils), vitamins, and minerals are added. The mixture is blended with fresh animal milk, sterilized, homogenized and bottled (Chandrasekhara et al, 1971). The 'toned' milk contains 3.5% protein, between 4 and 8% carbohydrate, at least 2% fat and added vitamins and minerals. The pattern of essential amino acids of the 'Miltone' is very close to that of the FAO reference pattern {Table 15.9). Large-scale production of 'Miltone' is under way in India with the aid of UNICEF (Winkelmann, 1975). It is important to remember that any groundnut flour used for beverage production must be free of aflatoxin contamination.

Using the principles learned in the production of whey-soy drink mix, Holsinger et al (1978a) prepared and tested a prototype of a spray-dried whey-peanut powder to serve as a base for beverage type products. The formulation contained 50% sweet whey sohds and 24.6% defatted peanut

supplement, well accepted by the recipients, is no longer distributed because US Government stocks of nonfat dry milk have risen to unprecedented highs and WSDM is no longer competitive in price.

V.Η. Holsinger 227

G r o u n d n u t f lour

D i l u t e a l k a l i

Prote in e x t r a c t

A d j u s t to i s o e l e c t r i c point

P r e c i p i t a t e

W a t e r , a l k a l i , b u f f e r s a l t s

S l u r r y , pH 6 . 8

C a r b o h y d r a t e , f a t , v i t a m i n s , m i n e r a l s

Blend wi th f resh a n i m a l milk

I S t e r i l i z e and h o m o g e n i z e

B o t t l e

Figure 15.7 Manufacture of 'toned' milk containing groundnut protein flour among other ingredients. The standardized PER was 2.0 compared to casein at 2.5; nitrogen digestibility was 90%. The flavor quality was bland initially; however, decreased flavor scores coupled with increased peroxide values during storage were indicative of a stability problem still to be solved.

The only animal protein other than the milk proteins used in imitation milks is fish protein concentrate (FPC). FPC is being used to fortify a traditional rice beverage called 'chicha' in Venezuela (Kodaira, Rey and Luna, 1978). It has also been reported that commercial production of imitation milk from FPC was underway in Chile with assistance from the FAO (Winkelmann, 1975).

Many novel proteins have been isolated and studied for their nutritive value and functionality as food ingredients. Virtually all of them have been proposed as milk extenders, as a base for the production of imitation milks or as fortifiers for acidic beverages.

Cotton-seed is rarely used as an edible protein even though, among oilseeds, cotton-seed production of 22-24 tonnes annually is second only to soybean (Spadaro and Gardner, 1979). Development of edible protein products has been inhibited by the presence of pigment glands containing gossypol in the kernels. Gossypol is a highly reactive polyphenolic compound that is toxic to most monogastric animals in its metabolically active form. With the emergence of processes to remove intact glands without deleterious effect to the meal, the preparation of a variety of nutritious


protein concentrates became feasible. The storage proteins are acid soluble; a protein isolate, dried at pH 3.5, is white, and almost totally soluble at that pH. It has been used at levels up to 6% to fortify acidic beverages with protein (Spadaro and Gardner, 1979).

Because of their high content of lysine, methionine, and cystine, it has been claimed that rapeseed proteins have a higher nutritive value than any other vegetable protein (Ohlsan and Anjou, 1979). Unfortunately, they contain about 4% of toxic glucosinolates that must be removed during processing. Standardized PERs of detoxified rapeseed protein concentrates average 3.0-3.5, equivalent to cheese whey proteins or egg albumen. Because rapeseed is one of only two oilseeds to be grown successfully in all parts of the world, its potential as a protein resource is excellent, provided the toxicity problem can be overcome.

The commercial use of sunflower seed protein meal as a food depends on the development of low-chlorogenic acid cultivars and efficient processes for dehulling high oil cultivars (Sosulski, 1979). Chlorogenic acid is a phenolic compound that, when present, contributes a gray color to baked goods, and turns green under alkaline conditions.

Sunflower protein concentrates have been evaluated as protein sources for milk-hke beverages (Sosulski and Fleming, 1977). Unhke soy and peanut proteins, sunflower proteins are highly soluble in low or high concentrations of sodium or calcium chloride, which makes them more desirable for producing extended milk beverages. Although nitrogen solubility of a sunflower protein concentrate was significantly reduced by the aqueous diffusion method used to remove the color producing phenolic acids, over 80% of the protein could be resolubilized by a combination of heat treatment, mechanical agitation, and emulsifiers. A 1:1 blend of sunflower protein concentrate with milk gave a grayish-white or milky-white color and a slight cereal flavor compared to a tan-white or yellow-white color and beany taste of a 1:1 soy flavor milk blend. Chemical score of the sunflower-milk blend was 90.5 compared to 94.3 for cows' milk alone.

Sesame is probably the oldest crop grown for edible oil. Asia and Africa produce 90% of the world supply and most of the seed is consumed in the countries where grown (Johnson, Suleiman, and Lusas, 1979). Harvesting characteristics have precluded a successful crop in developed countries because of the tendency of the seed to shatter with mechanical harvesting. The development of indehiscent varieties has made the future more promising.

Sesame protein has a unique balance of amino acids. Although the PER is only 1.2-1.3 because of the low concentration of lysine, sesame protein contains from 3.8 to 5.5% sulfur amino acids and about 2% tryptophan, which means that the amino acid pattern is complementary to those of other oilseeds; for example a 1:1 blend of sesame with soy protein has the same nutritive value as casein. Unfortunately, sesame protein is insoluble at acid and neutral pHs, making it more desirable as a protein supplement in baked goods than as an ingredient in beverages (Johnson, Suleiman and Lusas, 1979).

Some nonoilseed pulse proteins have been evaluated in beverages. A chocolate-flavored beverage from whole chick-pea had acceptable sensory

V.Η. Holsinger 229

Conclusions

At the present time, only proteins from soybeans or groundnuts are widely used for beverage production on a commercial scale. Cheese whey protein has been used for beverage fortification in some isolated instances; its usefulness is expanding with improved technologies for whey fractionation. Of the numerous proteins recovered from a variety of sources, most are still in the research phase; many problems of availability, flavor, functionality and cost still remain to be overcome.

References

A N O N . (1973a). Industry Week, 111, (4), 42 A N O N . (1973b). Kept. Joint FAOIWHO Ad Hoc Expert Committee, FAO

Nutrition Meetings Report No. 52. (FAO, Rome) A N O N . (1973c) FAG Bulletin, Vol. HI, No. 1 p.l4 United Nations, NY A N O N . (1978a). Schweizerische Milchzeitung, 104, (57), 439 A N O N . (1978b). Food Engng Intl, 3, (11), 20 A N O N . (1982a) Beverage World, 101, 30

properties (Fernandez de Tonella, Taylor and Stull, 1981). Experimental lupine proteins have been evaluated in gruels in Chile (Cerletti and Duranti, 1979). Field pea and faba bean proteins are also in the early stages of development (Bramsnaes and Olsen, 1979). Functionality of these proteins for beverage fortification is not known.

Leaf protein concentrate has repeatedly been investigated as a protein source and assessment of its nutritional value has been a major preoccupation of many researchers. Carefully made leaf protein contains 60 to 65% true protein and 0.1 to 0.2% ß-carotene. Feeding experiments with rats, chickens, pigs and mice have shown the protein to be safe and nutritionally useful. Human trials with purified green leaf protein concentrate have shown it to have a favorable effect on the growth of malnourished infants when fed in a 50:50 mixture with milk. Children with kwashiorkor have also responded well. However, because of still unsolved problems with flavor and color, leaf protein concentrate has not found widespread appreciation as food for humans (Vinconneau, 1979; Kinsella, 1979; Price, 1979).

Cereals have also been investigated as protein sources for beverages. Oat protein concentrates have been used to fortify neutral or acidic beverages with up to 4% protein (Cluskey et al, 1976). The authors claim that milk analogues may also be easily prepared from this protein source. Wheat gluten has also been proposed for use in nutritious beverages, if rendered soluble by acidic or enzymatic treatments (Kahn, 1979).

Other proteins under development include those from safflower, okra, grapeseed, coconut, potato, hma bean, and single-cell protein. Much work remains to be done to determine if proteins from these sources have the necessary functional and nutritional characteristics to make them desirable proteins for beverage development.


A N O N . (1982b). News and Views Sept.-Oct. 3. Dairy Council, Inc., Southampton, PA

A Y R E S , J.L. , B R A N S C O M B , L.L., and R O G E R S , G.M. (1974). J. Am. Oil Chemists Soc., 51, 133

B R A M S N A E S , F. and OLSEN, H.S. (1979). / . Am. Oil Chemists Soc, 56, 450 CERLETTL P. and D U R A N T I , M. (1979). J. Am. Oil Chemists Soc, 56, 460 C H A N D R A S E K H A R A , M.R. , R A M A N N A , B.R. , J A G A N N A T H , K S . and R A M A -

N A T H A N , P.K. (1971). Food Technol., 25, 596 CLUSKEY, J . E . , W U , Y . V . , I N G L E T T , G . E . and WALL, J.S. (1976) / . Fd Sci., 41, 799 D E M O T T , B.J. (1975). J. Milk Fd Technol., 38, (11), 691 D E M O T T , B.J., HELMS, A . B . and S A N D E R S , O.G. (1977). / . Fd Protection, 40,

(8), 540 D E L V A L L E , F.R., C A M A C H O , Α . , A C O S T A , Η. and LUJAN, F.J. (1978). Abst.

5th Intl Congress Food Sci. Technol., Kyoto, Japan, September 17-22, p. 81

F E N T O N - M A Y , R. (1975). In Technology of Fortification of Foods, p. 100 National Academy of Sciences, Washington, DC

F E R N A N D E Z D E T O N E L L A , M.L. , T A Y L O R , R.R. and STULL, J.W. (1981). Cereal Foods World, 26, 528

G U Y , E.J., VETTEL, H.E. and PALLANSCH, M.J. (1968). / . Dairy Sci., 51, 932 G U Y , E.J. , VETTEL, H . E . and PALLANSCH, M.J. (1969). / . Dairy Sci., 52, 432 H A N D , D . B . , S T E I N K R A U S , K.H. , V A N B U R E N , J.P., H A C K L E R , L.R., EL R A W I ,

I. and PALLESEN, H.R. (1964). Food Technol., 18, 1963 H A R D I N , C M . (1979). / . Am. Oil Chemists Soc, 56, (3), 173 H O L S I N G E R , V.H. (1976). Manuf Confectioner, 56, (1), 25 H O L S I N G E R , V.H. (1978). In Nutritional Improvement of Food and Feed

Proteins, Ed. Mendel Friedman, p. 735 Plenum Press, New York H O L S I N G E R , V . H . (1980a). Proc Whey Products Conf, p. 16. Chicago, IL,

October 21-22, USDA, Agricultural Research Service, Eastern Regional Research Center, Philadelphia, PA

H O L S I N G E R , V.H. (1980b). A Study of the rehydration properties of a milk analogue containing soy products and cheese whey. PhD Thesis. Ohio State University, Columbus, Ohio

HOLSINGER, V . H . , POSATI, L.P. and DEVILBISS, E . D . (1974). / . Dairy Sci., 57, 849

H O L S I N G E R , V . H . , POSATI, L.P., DEVILBISS, E . D . and P A L L A N S C H , M.J. (1973a). Food Technol., 27, (2), 59

H O L S I N G E R , V . H . , POSATI, L.P., DEVILBISS, E . D . and P A L L A N S C H , M.J. (1973b)./. Dairy Sci., 56, 1498

HOLSINGER, V . H . , DEVILBISS, E . D . , M C D O N O U G H , F.E. , POSATI, L P . , VET-TEL, H . E . , BECKER, D . E . , T U R K O T , V.S. and PALLANSCH, M.J. (1975). Abs. 169th Nat. Meeting Am. Chem. Soc, Philadelphia, PA, April 6-11, No. 41, AGFD

H O L S I N G E R , V . H . , SUTTON, C.S., VETTEL, H . E . , A L L E N , C. and T A L L E Y , F.B. (1977a). / . Dairy Sci., 60, 1841

H O L S I N G E R , V . H . , SUTTON, C.S., VETTEL, H . E . , A L L E N , C , T A L L E Y , F .B. , and WOYCHIK, J.H. (1978a). Peanut Sci., 5, 97

H O L S I N G E R , V . H . , SUTTON, C.S., VETTEL, H.E . , E D M O N D S O N , L.F., CROWLEY, P.R., B E R N T S O N , B.L. and PALLANSC H, M.J. (1977b). ProC. 4th Intl. Congress Food Sci. Technol., Madrid, Spain, September 23-27, 1974, 5, 25

ν, Η. Holsinger 231

HOLSINGER, V . H . , WOMACK, Μ., SUTTON, C S . , VETTEL, Η . Ε . , A L L E N , C. and T A L L E Y , F.B. (1978b). / . Dairy ScL, 61, 1061

J O H N S O N , L.A. , S U L E I M A N , T.M., and L U S A S , E.W. (1979). / . Am. Oil Chemists Soc., 56, 463

KALIN, F. (1979). / . Am. Oil Chemists Soc, 56, 477 KINSELLA, J.E. (1979). / . Am. Oil Chemists Soc, 56, 471 K O D A I R A , M., REY, J. and L U N A , G. (1978). Abst. 5th Intl. Congress Food

ScL TechnoL, Kyoto, Japan, September 17-22, p. 71 KOSIKOWSKI, F.V. (1968). / . Dairy Sei., 51, 1299 KOSIKOWSKI, F.V. and W Z O R E K WIESLAW (1977). / . Dairy Sei., 60, 1982 K U N T Z , D . (1977) Processing factors affecting the organoleptic quality of

Illinois soy beverages. PhD Thesis, University of lUinois, Urbana, IL L A N G , F. and L A N G , Α. (1979). Milk Ind., 81, (11), 30, 34 L A O , T.B. (1971). A study of the chemical changes relating to flavor of

soybean extracts. PhD Thesis, University of lUinois, Urbana, IL LEVITÓN, R. (1981). Soyfoods, 1, (4), 16 LISCHENKO, V.F. (1979). / . Am. Oil Chemists Soc, 56, (3), 178 LO, K.S. (1971). Process for preparing a soybean beverage. US Patent No.

3563762 LOEWNSTEIN, M. and P A U L R A J , V.K. (1972). Food Product Development, 5,

(8), 56, 60 M A L A S P I N A , A. and MORETTI, R.H. (1975). Preparation of a whey protein

concentrate. US Patent No. 3896241 MATTICK, L.R. and H A N D , D . B . (1969). / . Agric Food Chem., 17, 15 M U S T A K A S , G.C. , A L B R E C H T , W.J. and B O O K W A L T E R , G.N. (1972). Produc

tion of vegetable protein beverage base. US Patent No. 3639129 M U S T A K A S , G.C. , A L B R E C H T , W.J., B O O K W A L T E R , G.N. , SOHNS, V . E . , and

GRIFFIN, E .L. , Jr. (1971). Food Technol., 25, 534, 540 N E L S O N , A. I . , STEINBERG, M P . and WEI, L.S. (1975). Soybean beverage base.

US Patent No. 3901978 N E L S O N , A . L , STEINBERG, M.P. and WEI, L.S. (1976). J. Fd ScL, 41, 57 O H L S O N , R. and A N J O U , K. (1979). / . Am. Oil Chemists Soc, 56, 431 PIRIE, N.w. (1979). / . Am. Oil Chemists Soc, 56, 472 POSATI, L.P. and ORR, M.L. (1976). Composition of foods: Dairy and egg

products: Raw-processed-prepared. Agricuhure Handbook No. 8-1, USDA, Washington, DC

RACKIS, J.J., A N D E R S O N , R.L. , S A S A M E , H.A. and SMITH, A.K. (1961). / . Agric. Food Chem., 9, 409

R E A V E S , W.J. (1980). Proc. Whey Products Conference, Chicago, IL, October 21-22, p. 122 US Department of Agricuhure, Agricultural Research Service, Eastern Regional Research Center, Philadelphia, PA

R O D I E R , W.I . , III, WETSEL. W . C , JACOBS, H.L. , G R A E B E R , R . C , MOSKOWITZ, H.R. , R E E D , T.J.E. and W A T E R M A N , D . (1973). The acceptability of whey-soy mix as a supplementary food for preschool children in developing countries. Technical Report 74-20PR, US Army Natick Laboratories, Natick, MA

SASAKI, R. and T S U G O , T. (1953). Proc 13th Intl. Dairy Congress, 4, 606. The Hague, Netherlands

SHURTLEFF, W. (1981). Soyfoods, 1, (4), 28


SMITH, A.K. and CIRCLE, S.J. (1972). In Soybeans: Chemistry and Technology, Vol. I. Proteins Eds. by Smith, A.K. and Circle, S.J. p. 339 Avi Publishing Co., Inc., Westport, CT

SMITH, A.K. and CIRCLE, S.J. (1978) Soybeans: Chemistry and Technology, Vol. I: Proteins. Rev. Ed. Avi Publishing Co., Inc., Westport, CT

SOSULSKI, F. (1979). / . Am. Oil Chemists Soc, 56, 438 SOSULSKI, F.W. and FLEMING, S.E. (1977). Canad. Inst. Food Sci. Technol.

J., 10, 229 S P A D A R O , J.J. and G A R D N E R , H.K., Jr. (1979). / . Am. Oil Chemists Soc. 56,

422 SUSLI, H. (1956). 14th Ind. Dairy Congress, 1, (II), p. 477. Rome, Italy V I N C O N N E A U , H.F. (1979). J. Am. Oil Chemists Soc, 56, 469 W H E Y P R O D U C T S INSTITUTE. (1982). Estimated US fluid whey and whey

solids production (by type) and resulting quantity of whey solids further processed. Chicago, IL

WILCKE, H.L. , HOPKINS, D .T . and W A G G L E , D . N . , Eds. (1979). Soy protein and human nutrition. Academic Press, New York

WILKINS, W.F. , MATTICK, L.R. and H A N D , D . B . (1967). Food Technol., 21, 1630

W I N K E L M A N N , F. (1975). Wld. Anim. Rev., 14, 31 Y A N G , H.Y. , B O D Y F E L T , F.W., B E R G G R E N , K.E. and L A R S O N , P.K. (1977).

Utilization of cheese whey for wine production. Environmental Protection Technol. Ser. EPA-600/2-77-106, Grant No. 803 301. USEPA-Industrial Environmental Research Laboratory, Cincinnati, Ohio

16

Introduction

As a consequence of the meat-eating habit of man the major role of the world's animal production is to provide human food. Inevitably some food which would have been used directly for human consumption becomes the food of these animals. Thus, a conflict can arise in the use of the world's resources. Clearly any scarce commodity needs to be used as efficiently as possible and each material needs to be diverted to its most appropriate use. Consequently, there are attractions in considering waste materials in animal feeding as they help to avoid the immediate competition for resources between man and animals. At the same time their value may be enhanced when otherwise they might have been of little value, no value or even a cost in terms of their disposal. The value of waste materials depends on their ability to meet the needs of the animal for particular productive processes. The materials themselves will be of value depending on their initial concentration of energy and nutrients, the presence of any non-nutritive influences, e.g. toxins and the influence of processing and preparation on the availability of their components.

The needs of the animal

Before considering the value of various waste materials it is important to consider how the needs of the animal might influence decisions on their utilization. A major item influencing utilization by the animal will be whether or not it is a ruminant. While it has been suggested that ruminants and non-ruminants may well fít into the same broad nutritional pattern when fed diets of similar nutrient density (Swan and Cole, 1975), the presence of a rumen allows the utilization of a wider and generally poorer range of feedstuffs. However, the fermentation of food in the reticulo-rumen with its digestion in the abomasum is associated with poorer conversion efficiency for ruminants compared with, for example, pigs. Nevertheless, this should not detract from the fact that the ruminant can utilize materials which are not suitable for, or only poorly utilized by, the non-ruminant. For example, dried grass has a metabolizable energy (ME) value of about 10 MJ/kg for cattle whereas a value of 6.48 MJ ME/kg was

THE UTILIZATION OF WASTE IN ANIMAL FEEDS

J. WISEMAN and D.J.A. COLE University of Nottingham School of Agriculture, Sutton Bonington, UK

233

234 The utilization of waste in animal feeds

E N E R G Y

In Britain dietary energy is usually described in terms of apparent digestible energy (DE) for pigs and apparent metabolizable energy (ME) for ruminants and poultry.

DE = Gross energy of food - gross energy of corresponding faeces

ME = DE - Gross energy of corresponding urine - gross energy of any combustible gases.

Net energy (NE) systems, which make allowance for energy losses through heat increment, are also used in some countries. Consideration of a wide range of energy systems has been given by Kielanowski (1972) for pigs. De Groóte (1974) for poultry and Niemann-Sorensen (1980) for ruminants.

An animal needs energy for maintenance and its productive processes. Generally the requirement for maintenance is taken as basal metabolic rate (based on the animal's metabolic body size) together with an increment for activity. The requirement for production depends on the amount of production, the energy concentration per unit product and the efficiency of conversion of dietary energy to product (e.g. live weight, milk, eggs, fetus). Growing animals, particularly pigs, have considerable ability to store excess energy as fat in the body and this may detract from carcase value.

PROTEIN

For many purposes crude protein (N x 6.25) has been taken as a measure of total protein status of a feedingstuff. In some circumstances digestible protein has been used and a whole range of other measures are available (e.g. net protein utilization, protein efficiency ratio, etc.) but not often used for farm livestock. All measures of protein quality assess the limitation imposed by the limiting amino acid(s) and amino acid composition has become a more common descriptor of the protein status of an ingredient. Measures of amino acid availability would add further precision but there is considerable debate about appropriate techniques.

The animal requires a supply of essential amino acids together with adequate nitrogen for the synthesis of non-essential amino acids. Requirements for maintenance are small, relative to those for production, in the high yielding animal. Thus, the animal product has considerable influence on requirements which are hkely to be similar, at tissue level, for ruminants and non-ruminants. Associated with rumen fermentation is the

reported for pigs (Morgan, Cole and Lewis, 1975a). Some materials, e.g. hay and sugar beet pulp, are not usual ingredients in the diets of non-ruminant animals.

The major role of the ingredients of animal diets is the provision of energy and protein. The adequacy of the mineral and vitamin content is largely assured by supplementation.

J. Wiseman and D.J.Α. Cole 235

COMPOSITION OF T H E DIET

While diets are described mainly in terms of their protein and energy content they may not necessarily behave according to their determined composition. For example, it has been shown that there have been considerable reductions in digestibility of non-ruminant diets with increases in crude fibre level. Some further reports with pigs have also shown reductions in crude protein digestibility with high levels of crude fibre (reviewed by Cole, 1973). It has been suggested that this may resuk from increases in metabolic faecal nitrogen or lack of penetration by digestive enzymes. Thus, the inclusion of ingredients high in fibre may reduce the overall nutritive value of a diet as a consequence of such interaction.

Conversely, the presence of fat may aid the digestibility of other ingredients such that the nutritive value of the diet is elevated. Such an effect has been reported by Mateos and Sell (1981) with poultry and was attributed to fat slowing the rate of passage of digesta, allowing gastric enzymes more time to act. Additionally a fat from one particular source (e.g. soya oil) may assist in the absorption of another (e.g. tallow) such that the combined nutritive value of a mixture of the two may be greater than could be predicted from their independent values (Sibbald, 1978).

Such interactions must be accounted for in diet formulation, in addition to meeting the animal's protein, energy and micronutrient requirements.

APPETITE

A number of animal production systems depend on allowances of feed. Under these circumstances it is important that the allowances prescribed are within the appetite limits of the animal. Thus, a knowledge of the animal's reaction to its food in terms of feed intake is important and will also be of value when the aim is to maximize feed intake (e.g. in the young fast growing animal).

Animals attempt to eat to meet their requirements for maintenance and production. The amount that they actually eat is modified by the food. For example, it has been shown that non-ruminants often eat more of a poor quality feed than a good quality feed in order to attain a particular energy intake. Such a compensatory mechanism is often not complete. Obviously, the animal cannot indefinitely make up for poor quality by eating more food as a bulk hmitation will be reached. Generally ruminants are fed diets which mean that they are in the area of bulk restriction rather than in the area of a compensatory mechanism through physiological control, although Swan and Cole (1975) have suggested that cattle and pigs may behave similarly when compared over the same range of energy concentration in the diet {Figure 16.1).

Thus, the nutritive value of a diet may be viewed in terms of its ability, when fed within appetite limits, to supply (in addition to micronutrients)

production of microbial cells which are rich in protein and subsequently available for digestion by the host animal.


U)

^ Ν

.Ξ ο 0; υ

1 9 0

> Ο;

1 7 0

| 5 150[-t σ 140 · -

Pfgs ^.--^

For p i g s 1 4 / 0 . 6 7 5 For c a t t l e kV^-^^

B e e f c a t t l e

^ i E 2 . 0 9

Ε

• . " " ' P i g s . . B e e t c a t t l e

3 . 3 3 .1 2 . 9 2 . 7 2 . 5 2 . 3 2 . 1 D i g e s t i b l e e n e r g y (Meal /kg a ir d r y f e e d )

Figure 16.1 Relationship between the voluntary intake of food dry matter, digestible energy and nutrient density of the diet in pigs and beef cattle (Swan and Cole, 1975)

sufficient quantities of amino acids and energy-yielding ingredients to allow the animal to realize its biological potential for whatever purpose it is kept (i.e. production of meat, milk or eggs, etc.).

Limitations to use

Feedingstuffs used in diet formulation should obviously contribute to the overall nutritive value, but there are a number of possible limitations to their use which need to be considered prior to their incorporation.

The detrimental effect of high levels of fibre on the utilization of other dietary ingredients has already been mentioned. In addition an appreciation is needed of how such levels may increase the bulk density of the diet such that the animal's physical capacity for food is reached. The presence of non-nutritive factors is relevant not only because they may reduce palatability due to taints or odours, but because they may precipitate digestive disorders, or even be toxic (Leiner, 1981). Additionally, such factors may be responsible for tainting the salable product; a good example being the sinapine content of rapeseed meal which has been implicated in producing fishy taints in brown eggs (Hobson-Frohock, Fenwick and Heaney, 1977). Non-nutritive factors in feedingstuffs may occasionally be present as a consequence of the previous history of the material rather than be naturally occurring. Thus, health hazards may result from the use of feedingstuffs of animal origin (e.g. excretory products and slaughterhouse wastes) which may be contaminated with micro-organisms such as Salmonella sp. (Williams, 1980) or with pharmaceutical residues (Battacharya

/. Wiseman and DJ. A. Cole 237

Utilization of wastes and by-products

Fundamental to the utilization of wastes and by-products is an appreciation that if any primary raw material is subdivided into a number of fractions during initial processing (i.e. at least one major portion, which is the objective of the process, together with one or more by-products) then the nutritive value of all the fractions cannot usually exceed that of the original unless they are subjected to further processing. An illustration of this

Table 16.1 NUTRITIVE V A L U E OF MILLING C O M P O N E N T S OF W H O L E W H E A T ( A S S U M I N G 80% E XTRACTION RATE)

Yield Protein (g/kg) (g/kg)

Fraction Contrib. to whole wheat

Fibre (g/kg)


Metabolizable energy (MJ/kg : pigs)


Flour 805^ 120^ 96.6 2^ 1.6 15.00^ 12.08 Fine wheat feed 125^ 143^ 17.9 84^ 10.5 9.28^ 1.16 Bran 70^ 124^ 8.7 I I P 7.8 8.19^ 0.57 Whole wheat

(1) Calculated 1000 123.2 from above 1000 123.2 19.9 13.81

(2) Determined'^ 117.7 17.6 13.33

For example, the protein content of flour is 120 g/kg and the contribution of flour to the overall protein content of wheat is 96.6 g/kg. "Kent, 1975 ^Feedstuffs, 1981 ''Wiseman, Cole and Lewis, 1982

and Taylor, 1975), although such problems as the latter are only likely to occur if adequate withdrawal periods prior to slaughter have not been used (Webb and Fontenot 1975). Similarly, heavy metal contamination of activated sludge may pose problems (Beszedits, 1981).

The potential value of a feedingstuff may be adversely affected by indirect factors such as moisture content. This applies to washings from various processes (e.g. during vegetable canning) and to many dairy by-products. For example, the high cost of either drying, or transporting hquid whey often precludes its economic use as a feedingstuff (Sching-oethe, 1976). Degeneration during storage is a serious problem, and the development of moulds, for example those associated specifically with Aspergillus flavus (producing aflatoxins), may have important repercussions on both the animal and human food industry (Arafa et al., 1979; Lindsay, 1981). Adverse storage conditions are also important in the development of, for example, rancidity in those feedingstuffs which are high in polyunsaturated fatty acids.

Finally, it is important that any feedingstuff that is to be incorporated into animal diets should be in regular supply and should be of moderately constant and fairly predictable nutritive value. Its utilization should of course be an economically viable proposition—a point particularly important with the development of least cost ration formulation in most modern feed mihs.


appears, with reference to wheat, in Table 16.1. Assuming an extraction rate of 80% during milhng, the resuUant respective proportions of flour, fine wheatfeed and bran are shown. Based upon published data for the protein, fibre and dietary energy content of these three fractions, it is possible to calculate the nutritive value of the original whole wheat. This compares favourably with an a^ t̂ual determined value, the difference being acceptable considering the various assumptions associated with the procedure and biological variability. Similar calculations could be performed for other primary raw materials. Thus, oilseed cakes, whilst being a very important source of protein to the livestock feeding industry, have much lower dietary energy values than the original seeds prior to oil extraction. Currently, there is a trend to utilize full fat oil seeds for animal feeding-stuffs as a source of both protein and energy (particularly soyabeans, Wiseman 1981); a situation which has arisen largely due to the relative economics of oil extraction, and the cost of alternative sources of energy-yielding ingredients.

This, then, explains a major limitation to the utilization specifically of the majority of wastes and by-products. By definition, they tend to arise from processes which are designed to remove those fractions that are of a higher value leaving those that, for non-ruminant animals at least, are of relatively poorer quality, and which usually require a considerable degree of supplementation when incorporated into complete diets.

However, it would be wrong to limit a discussion of wastes and by-products to those arising solely from the food or feed processing industries. Any commodity produced as a residue from any process, as long as it conforms to the criteria of what constitutes an animal feeding-stuff, has potential use. In fact the realization that some chemical by-products have such potential value can be of considerable economic value, particularly if they are troublesome pollutants, and a whole new technology has been developed to upgrade them.

It is possible to differentiate between those wastes and by-products that may be incorporated directly into animal diets, and those which, for a variety of reasons, normally undergo further processing to enhance their nutritive value. It should be borne in mind though that some potential feedingstuffs may be found in either category.

Wastes and by-products used directly as feedingstuffs

Before dealing with their upgrading it is worth considering that there are numerous wastes and by-products that can be utilized perfectly satisfactorily usually by ruminant livestock. Thus, the harvested portion of most crops is limited to, for example, the grain in cereals. This results in large quantities of plant material, from both arable and vegetable sources, which is of potential value (Francis, 1980), an example being those from brussels sprouts {Table 16.2).

The brewing industry generates large quantities of wastes, such as brewers' grains (Barber and Lonsdale, 1980; Table 16.3) and barley rootlets which are well known animal feedingstuffs. Liquid whey and skimmed milk are also traditional ingredients in animal feeds (Sching-oethe, 1976).

J. Wiseman and D.J.A. Cole 239

Table 16.2 C O M P O S I T I O N O F B R U S S E L S S P R O U T S W A S T E S ( D R Y M A T T E R

B A S I S ) ( F R A N C I S , 1980)

Stripped stems Foliage waste

Dry matter (g/kg) 209 157 Crude protein (g/kg) 168 158 Crude fibre (g/kg) 174 117 Metabolizable energy (MJ/kg) 11.8 11.3

Table 16.3 N U T R I T I O N A L V A L U E O F

B R E W E R S ' G R A I N S ( D R Y M A T T E R B A S I S ) ( B A R B E R A N D L O N S D A L E , 1980)

Dry matter (g/kg) 263 Crude protein (g/kg) 234 Crude fibre (g/kg) 176 Metabolizable energy ( M J / k g ) 11.20

Some non-food industries also provide by-products which have a possible nutritive value. Medium chain length fatty acids (from C12 to C20), for example, are receiving considerable attention as possible sources of energy in both ruminant and non-ruminant diets. Thus, the production of blends of free fatty acids, glycerides and glycerol to produce commodities to compete with more traditional fats may alter the emphasis of an industrial process whereby fatty acids could become the prime objective rather than a by-product.

Upgrading of wastes and by-products

Many wastes and by-products may require further processing before they can be considered as potential feedingstuffs, whether this be in order to improve their nutritive value (i.e. digestibihty) directly, to sterilize, detoxify or concentrate them, or, by biological means, to produce a nutritionahy utilizable commodity from substrates of little initial feeding value. Procedures have been described elsewhere in this book, and it is proposed here to consider specific examples of them in terms of the feeding value of the final product.

I M P R O V I N G D I G E S T I B I L I T Y

The improvement in digestibility following processing should be viewed in the knowledge that such changes are rarely accompanied by alterations in levels of any readily assayable chemical measurement, so that the prediction of any enhancement in nutritive value is difficult. There has been a considerable degree of interest in upgrading the vast amounts of crop residues produced annually for use as animal feeds. Cereal straws have been given most attention, but any cellulose wastes could be considered equally; ranging from those already utilized but which could possibly


2 4 6 NaOH ( k g / 1 0 0 k g DM)

Figure 16.2 The effect of sodium hydroxide concentration during processing on subsequent in vitro organic matter digestibihty (OMD) (Rexen et al., 1975)

benefit from upgrading (e.g. brewers' grains) to hitherto unused commodities such as sawdust and vegetation including heather (Greenhalgh, 1980). All these examples have one thing in common, in that, being fibrous, they provide a potential fermentable substrate for rumen micro-organisms, although it should not be forgotten that due to their usually low nitrogen content, supplementation is essential.

A major hmitation to the use of such commodities is that they are usually derived from mature plants and are invariably highly hgnified. Polysaccharides, such as cellulose and hemicellulose, complexed with lignin have a much lower rumen digestibility. Although physical processes, such as chopping, have been employed to enhance the nutritive value of straw (Swan and Clarke, 1974), it is the use of chemical treatments that have received most attention, in particular sodium hydroxide (due to its relatively low cost) although hydroxides of calcium and potassium have also been considered, as has ammonia.

The nutritive value of the resultant product is a consequence largely of the conditions prevailing during processing. Thus the concentration of sodium hydroxide can have a marked effect upon in vitro digestibility (Figure 16.2). However, only small changes in levels of dietary fibre (as determined by the acid detergent technique) or of crude hgnin accompany the improvement in nutritive value following alkali treatment (Kohler, Walker and Kuzmicky, 1979; Table 16.4). A second method of improving

/. Wiseman and DJ.Λ. Cole 241

Unheated 40 s steam, 5% H2O, 28 kg/cm control ( 0 % N a O H ) ( l % N a O H ) (4% NaOH)

Acid detergent fibre (g/kg) 506 488 494 498 Crude lignin (g/kg) 165 152 149 137 In vitro digestibility (%) 39.7 62.2 67.0 69.5

the potential nutritive value of crop residues is ensilage. This process is unlikely to improve digestibility directly but is a useful means of conserving residues, that may only be available for short periods of the year, and thus rendering them available for greater lengths of time. Sugar beet tops are frequently considered for ensilage (Francis, 1980) and although mean data for the nutritive value of both fresh and ensiled product are presented (Table 16.5) considerable variabihty is possible dependent upon the degree of soil contamination. Another part of the sugar beet crop considered for ensilage is the pulp left after sugar extraction. Frequently it is molassed although the source of molasses is not usually the beet—this being used as an industrial fermentation substrate (Barber and Lonsdale, 1980). While many wastes and by-products have potential nutritive value, economic forces may result in their use in other industries.

Table 16.5 S U G A R B E E T B Y - P R O D U C T S ( D R Y M A T T E R BASIS) (FRANCIS, 1980; B A R B E R A N D L O N S D A L E , 1980)

Tops Molassed pulp Fresh Ensiled Pressed Ensiled

Dry matter (%) 15.5 15.3 24.7 22.1 Crude protein (g/kg) 140 153 134 132 Crude fibre (g/kg) 85 119 141 141 In vitro organic matter digestibility (%)of the dry matter 50-65 49 -58 79.1 82.0

Metabolizable energy (MJ/kg) 9.6 8.8 12.1 12.9

A major limitation to the use of ensiled crop products is that, due to their high moisture content, they cannot be transported and have to be fed in situ. Thus their value is generally limited to mixed farming operations. Additionally, once the silage clamp is opened, there is considerable risk of aerobic fermentation causing the nutritive value of the product to fall. By-products from the fishing industry have also been considered for ensilage. These may either be in the form of offals from filleting plants, or whole fish which may be present as by-catch. The fish is minced prior to mixing with formic acid. The nutritive value of the resultant product is very much dependent upon the raw material ensiled, but DE for pigs has been estimated as 23.4 MJ/kg DM and 17.5 MJ/kg DM for oily and non-oily fish respectively (Green, Wiseman and Cole, 1981). Inclusion in diets is limited to approximately 10% on a dry matter basis, due to problems of palatabil-ity (Wiseman, Green and Cole, 1982). Organoleptic qualities of the pork meat may be adversely affected at higher levels although assessments have been very subjective. In some countries, the product could be satisfactorily dried in the sun.

Table 16.4 EFFECT OF PROCESSING CONDITIONS O N T H E S U B S E Q U E N T F E E D I N G V A L U E OF STRAW (KOHLER, W A L K E R A N D KUZMICKY, 1979)


H E A T I N G

The apphcation of heat is important for a number of reasons, including the concentration of hquid feedingstuffs, denaturing of non-nutritive factors and sterilizing of any micro-organisms that may be present. Additionally, heat may be necessary to obtain a desirable ingredient from a by-product, for example the rendering of fat from slaughterhouse waste. Large amounts of hquid wastes of potential feeding value are produced from a number of industrial processes within the food industry. Reference has already been made to the production of mah culms and brewers' grains from distilhng which may be fed directly to livestock without further processing.

However, there are also quantities of other by-products of low dry matter content, an example from the production of malt whisky being pot ale. This is a particularly troublesome pollutant with a very high biological oxygen demand. Its discharge into sewage or river systems is discouraged, and alternative outlets for its disposal have been investigated. In its dried form, it is a valuable feedingstuff for both ruminants and non-ruminants particularly when blended with draff to product distillers' dark grains {Table 16.6). However, it is considered (Pass, 1981) that production costs

Table 16.6 NUTRITIVE V A L U E OF DISTILLERS' B Y - P R O D U C T S (PASS, 1981; N O S C A , 1982)

Draff Pot ale Dried Dried Draff syrup distillers' distillers'

solubles grains

Dry matter (g/kg) 240 450 970 900 Crude protein (g/kg) 50 350 291 230 Digestible crude protein (g/kg) 39 270 235 161 Crude fibre (g/kg) 43 — — 101 Metabolizable energy (ruminants) (MJ/kg)

2.6 7.1 9.0 9.7

Digestible energy (pigs) (MJ/kg) — — 10.0 11.4

incurred during drying may render this process uneconomic in the future. The raw material (of 4% dry matter) still has value where it could be incorporated into liquid feed systems but costs of transport could be prohibitive. This, interestingly, has led to the suggestion that distilleries may need to consider housing livestock alongside the main plant to dispose of the large amounts of liquid wastes produced, a system that was used successfully during the early 1800s in the great London breweries.

Sterilization is of particular importance when animal wastes are used as feedingstuffs, although culinary wastes are also potential sources of infection. There are strict regulations governing the processing of both types of wastes (Watson, 1980).

The upgrading of animal offals in this fashion produces feedingstuffs of good quahty which have been (Batterham et al., 1980) referred to as meat meals (above 55% crude protein), meat and bone meals (between 40 and 55% crude protein) or bone meals (less than 40% crude protein). However, due to the varying proportions of different offals from various

J. Wiseman and DJ.A. Cole 243

( B A T T E R H A M ETAL., 1980)

Crude protein (g/kg) 438-567 Ether extract (g/kg) 78-139 Ash (g/kg) 224-345 Gross energy (MJ/kg) 14.18-18.22 Digestible energy (MJ/kg) 9.44-13.21 Crude protein digestibility (%) 73 .1-91 .0

Batterham et al. (1980) were able to predict energy digestibility of meat meals and meat and bone meals fairly accurately, but no significant relationship between protein digestibility and any chemical measurement could be detected. Both protein digestibility and subsequent availability of amino acids are influenced considerably by processing conditions, but in ways which are not fully understood. More information relating to this problem can only increase the efficiency of processing and utilization of these types of feedingstuffs.

Another major source of animal wastes of potential feeding value is excretory by-products, although due to their usually high levels of fibre and (in poultry wastes) non-protein nitrogen, they are best suited to use in ruminant diets. However, it has been claimed that inclusion of 5% dried poultry waste did not depress subsequent performance of broilers, and that although feed conversion was poorer at 15% inclusion, this could be offset by the price of the material (El Boushy and Vink, 1977). It is interesting to note that levels of non-protein nitrogen may render the calculation for crude protein (% nitrogen x 6.25) somewhat meaningless. Thus, in the

Table 16.8 E X C R E T O R Y WASTES ( D R Y MATTER BASIS) ( B A T T A C H A R Y A A N D T A Y L O R , 1975)

Dehydrated cage layer Broiler litter

Dry matter (%) 90.0 84.7 Crude protein (g/kg) 280 313 True protein (g/kg) 113 167 Crude fibre (g/kg) 127 168 Digestible energy (cattle) (MJ/kg) 7.85 10.21

sources together with changes in processing conditions, the finished products may be of variable quahty (Cooke and Pugh 1980). This presents a considerable problem in their utilization. Accurate diet formulation requires precise information relating to the nutritive value of component ingredients. Whilst it is possible to monitor changes in, for example, proximate analysis, variability in nutritive value (which can only be measured accurately through laborious animal metabolism trials) is more difficult to assess. The use of regression equations (where DE or ME is predicted from proximate analysis) has proved moderately successful for primary feedingstuffs (Morgan, Cole and Lewis, 1975b; Wiseman and Cole, 1980) but for processed feedingstuffs such as meat and bone meal where changes in nutritive value are not accompanied by concomitant changes in proximate analysis {Table 16.7), success has been limited. Thus,

Table 16.7 V A R I A B I L I T Y I N C H E M I C A L C O M P O S I T I O N O F M E A T A N D B O N E M E A L S ( D R Y M A T T E R B A S I S )


F E R M E N T A T I O N

There has been considerable interest in the use of micro-organisms to upgrade wastes and by-products and to provide a source of good quality protein for the animal feed industry. Many types of micro-organisms, including bacteria, fungi and algae, have been investigated. They are a particularly useful means of detoxifying troublesome pollutants (e.g. hquid discharges from food processing or blood from slaughterhouses) of high biological oxygen demand and having a high content of suspended solids, and of concentrating the protein. In addition, inorganic substrates may be used (such as effluent from pulp mills). In fact the number of substrates capable of being fermented by, for example, yeasts, is considerable (Vananuvat, 1977). A consequence of this is that the nutritive value of the final product may be variable, a point that is also relevant when considering both the type of micro-organism used and harvesting conditions.

References

A R A F A , A . S . , H A R M S , R .H. , MILES, R . D and B L O O M E R , R.T. (1979). Feed-stuffs, 51, No.38, pp. 37, 38, 52

B A R B E R , W.P. and L O N S D A L E , C R . (1980). In By-products and wastes in animal feeding. Ed. by 0rskov, E.R. p.61. British Society of Animal Production

B A T T A C H A R Y A , A . N . and T A Y L O R , J.C. (1975). / . Anim. 5d. , 5, 1438 B A T T E R H A M , E.S . , LEWIS, C.E. , L O W E , R.F. and McMILLAN, D.J. (1980).

Anim. Prod., 31, 273 B E S Z E D I T S , S. (1981). Feedstuffs, 53, No. 14, p.25 COLE, D.J .A. (1973). In Nutrition Conference for Feed Manufacturers : 7.

Ed. by Henry Swan and Dyfed Lewis, p.81. Butterworths, London C O O K E , B.C. and P U G H , M.L. (1980). In By-products and wastes in animal

feeding. Ed. by 0rskov, E.R., p. 79. British Society of Animal Production

D E G R O Ó T E , Α . (1974). In Energy Requirements of Poultry. Ed. by T.R. Morris and B.M. Freeman, p.113. British Poultry Science Ltd, Edinburgh

EL B O U S H Y , A .R . and VINK, F .W.A. (1977). Feedstuffs, 49, No. 51, p.24 FEEDSTUFFS (1981). 53, No.3, p.44 FRANCIS , G.H. (1980). In By-products and wastes in animal feeding. Ed. by

0rskov, E.R., p.33. British Society of Animal Production G R E E N , S., W I S E M A N , J. and COLE, D.J .A. (1981). Anim. Prod., 32, 356

examples given in Table 16.8 true protein levels are considerably lower than crude protein levels. Processing of such wastes is concerned primarily with sterilization—excreta usually contain potentially harmful organisms (bacteria, nematodes) that would pose a threat to any animal to which they may be subsequently fed, although levels are very much dependent upon the health status of the original animal.

J. Wiseman and D.J.A. Cole 245

G R E E N H A L G H , J .F .D. (1980). In By-products and Wastes in Animal Feeding. Ed. by 0rskov, E.R., p.25. British Society of Animal Production

H O B S O N - F R O H O C K , Α . , FENWICK, G.R. and H E A N E Y , R.K. (1977). Br. Poult. Sci., 18, 539

KENT, N.L. (1975). Technology of cereals. Pergamon Press, Oxford KIELANOWSKI, J. (1972). In Pig Production. Ed. by Cole, D.J.Α., p.183.

Butterworths, London K O H L E R , G.O. , W A L K E R , H.G. and KUZMICKY, D . D . (1979). Fed. Proc, 38,

1934 LEINER, I.E. (1981). Toxic Constituents of Plant Foodstuffs. Academic

Press, New York L I N D S A Y , D . G . (1981). In Recent Advances in Animal Nutrition—1981. Ed.

by Haresign, W., p. 153. Butterworths, London M A T E O S , S.G. and SELL, J.L. (1981). Poult. Sci., 60, 2114 M O R G A N , D.J. , COLE, D.J .A. and LEWIS, D . (1975a). / . agric. Sci., 84, 7 M O R G A N , D.J. , COLE, D.J .A. and LEWIS, D . (1975b). J. agric. 5d. , 84, 19 N I E M A N N - S O R E N S E N , A . (1980). In Energy and Protein Feeding Standards

Applied to the Rearing and Finishing of Beef Cattle. Ed. by Beranger, C. Ann. Zootech., 29, 17

N O S C A (1982). Pot Ale Syrup. North of Scotland College of Agriculture Leaflet No. 47

PASS, R.T. (1981). Brewers Guardian, 110,(9), 17 R E X E N , F., STIGSEN, P. and FRISS KRISTENSEN (1975). In Feed Energy

Sources for Livestock. Ed. by Henry Swan and Dyfed Lewis., p.65. Butterworths, London

S C H I N G O E T H E , D.J. (1976). Feedstuffs, 48, No.31, pp. 18, 19, 41 S I B B A L D , I.R. (1978). Poult. ScL, 57, 473 S W A N , H. and C L A R K E , V. (1974). In Nutrition Conference for Feed Manufac

turers: 8. Ed. by Swan, H. and Lewis, D., p.205. Butterworths, London S W A N , H. and COLE, D.J .A. (1975). In Meat. Ed. by Cole, D.J.A. and

Lawrie, R.A. p.71. Butterworths, London V A N A N U V A T , P. (1977). CRC Crit. Revs. Fd Sci. Nutr., 9, 325 W A T S O N , W.A. (1980). In By-products and Wastes in Animal Feeding. Ed.

by 0rskov, E.R., p. 19. British Society of Animal Production W E B B , K.E. and F O N T E N O T , J.P. (1975). / . Anim. Sci., 41, 1212 WILLIAMS, J.E. (1980). Wld Poult. Sci. J., 36, 97 W I S E M A N , J. (1981). The Use of Full Fat Soyabeans in Diets for Poultry and

pigs. Proc. Joint ASA/SFT/AFTAA, 1980 W I S E M A N , J. and COLE, D .J .A . (1980). In Recent Advances in Animal

Nutrition 1980. Ed. by Haresign, W., p.51. Butterworths, London W I S E M A N , J., COLE, D.J .A. and LEWIS, D . (1982). J. agric. Sci., 98, 89 W I S E M A N , J., G R E E N , S. and COLE, D.J .A. (1982). Anim. Prod., 34, 364

17

Introduction

In the 1930s the Californian Fruit Growers' Exchange described a process for treating sohd plant material under mild alkaline conditions (Wilson, 1938). The object of this process was to de-esterify the pectin present, thereby producing a crude insoluble material which was later called pectate pulp (Baier and Wilson, 1941). Sodium pectate* may then be extracted from the crude pectate pulp by boiling with an alkahne phosphate sequestrant and further purified by alcohol precipitation.

In addition to being an intermediate material for the product of sodium pectate, pectate pulp can be used directly in some applications without the need to extract the pectate. The crude pulp from citrus waste contains about 20-30% pectin and, on heating in the presence of phosphate sequestrants, the pectate is extracted and a dispersion of predominantly cellulose material is obtained in a pectate solution. For some uses the presence of the dispersed material is not a disadvantage and Baier and Wilson (1941) described the apphcation of pectate pulp dispersions in rubber latex creaming, steel quenching and in the production of antistick coatings. As far as the authors are aware, however, none of these applications achieved significant success.

If calcium ions are introduced into the dispersion prepared from pectate pulp it is possible to form a calcium pectate gel. Because of its potential as a very cheap gelling agent there has recently been a revival of interest in the use of pectate pulp in foods. It has been shown that unlike most forms of pectin, pectate pulp can be made to gel in canned neutral pH foods, particularly meat products (Buckley, Mitchell and Burrows, 1978). This application is potentially of major importance to the pet-food industry because, in several countries, canned products consisting of meat in a polysaccharide gel are a major sector of the pet-food market. Indeed this application represents one of the largest food uses of another gelling agent, carrageenan. The interest kindled in pectate pulp by the discovery of its geUing properties in heat processed foods led to an examination of potential applications in a range of other food systems. The use of pectate *The ACS nomenclature (Kertesz et al, 1944) suggests the term pectate should be reserved for a pectic substance containing a 'negligible' proportion of methoxyl groups. In this chapter the term will be applied rather more loosely to materials with a low degree of esterification.

247

CRUDE PECTÄTE GELLING AGENTS IN HEAT PROCESSED FOODS

J.R. MITCHELL and A.J. TAYLOR Department of Applied Biochemistry and Food Science, University of Nottingham, UK

248 Crude pectate gelling agents in heat processed foods

pulp as a stabilizer in frozen products, in reformed fruit and meat chunks, yoghurts, blancmange type deserts and dietetic jams has been described (Mitchell, Buckley and Burrows, 1978).

More recently work at Nottingham has been directed towards understanding the factors governing the gelation behaviour of pectate pulp in heat processed foods as well as determining if the process which was originally applied to orange waste can be used with other materials. In this chapter this work will be reviewed and the future potential of pectate pulp discussed.

Before considering the crude system and the use of wastes as a source of pectate pulp, the relevant properties of pectin and its behaviour in heat processed systems will be outlined. For more detailed information the reader is referred to the many reviews that have been written on pectin (e.g. Doesburg, 1965, 1973).

Properties of extracted pectins

PECTIN S T R U C T U R E A N D O C C U R R E N C E

Pectin is found in the primary cell wall of all green land plants. It is chemically heterogeneous and there is still much to be discovered about its structure. The uniform structural feature is a backbone of a-l,4-linked polygalacturonic acid {Figure 17.1). A number of neutral sugars have been identified in pectin and these include galactose, rhamnose, arabinose and xylose (Rombouts, 1972).

Figure 17.1 A partially esterified polygalacturonic acid chain

The pectin content of plant tissue is normally expressed in terms of the amount of galacturonic acid present. When expressed on a dry weight basis, values are generally in the range of 5-25% with the pectin content of citrus waste being close to the upper hmit of this range (Rouse, 1977).

Of the neutral sugars present, rhamnose is the most important in determining the conformation of the molecule because it occurs in the polygalacturonic acid backbone whereas the other neutral sugars are present as side chains. Rees and Wight (1971) have shown that the presence of rhamnose will cause kinks of approximately 90 degrees in the

J.R. Mitchell and Λ J. Taylor 249

G E L A T I O N OF PECTINS

The gelation behaviour of pectin depends upon its degree of esterification. At high degrees of esteriflcation (>50%) pectin will form gels at low pH in the presence of high concentrations of a humectant. In food systems sucrose is invariably used and the pectin-sucrose-acid gel provides the gelled texture in jams and preserves. This is still the major application for pectins. Like other polysaccharide gelling systems the structure of the cross-links that form on high methoxyl pectin gelation has been the subject of recent investigations (Walkinshaw and Arnott, 1981; Morris et al, 1980). Walkinshaw and Arnott suggest that each junction zone or cross-hnk contains three to ten polymer chains orientated in parallel 3' helices.

At degrees of de-esterification below about 50% pectins can form gels with divalent ions over a wide pH range. A low water activity is not required for the formation of this type of gel. In food applications calcium ions are invariably employed. Determination of the structure of the junction zone or cross-link is complicated by the fact that a conformational change appears to take place on drying (Morris et al, 1982) and therefore the results of X-ray fibre diffraction studies may not be relevant to the gelled state. It seems probable from the level of calcium required to promote gelation that the junction zone in pectate gels is similar to that in alginate gels and is of the egg-box type in which only two chains participate (Grant et al., 1973; Morris et al., 1982) though a junction zone involving a larger number of chains has also been suggested (Walkinshaw and Arnott, 1981).

The properties of low methoxyl pectin gels depend not only on the degree of esteriflcation but also on the pectin molecular weight and the

polygalacturonic acid backbone, though in the model for pectin in sycamore ceU walls proposed by Keegstra et al. (1973), this kink will be nullified by the presence of a second rhamnose residue with a single galacturonic acid residue lining the two rhamnose units.

It has recently been suggested that, in pectin from other sources, the rhamnose residues are regularly located along the polymer chain with about 25 galacturonic acid residues between each rhamnose unit (Powell et al., 1982).

An important structural feature of pectins which influences both gelation behaviour and thermal depolymerization, is the extent to which the galacturonic acid residues are esterified (Figure 17.1). The % degree of esteriflcation (DE) of pectins in plant tissue varies widely (Doesburg, 1965) but in the native material values in the range 60-80% are often encountered. Pectins are commercially available with a range of DEs. Low methoxyl pectins typically have DEs in the range 20-40% and are prepared by de-esterifying extracted pectins generally by the action of enzyme, alkali, acid or by the action of ammonia in alcohohc dispersion. The last two methods are those most often employed on an industrial scale. Treatment with ammonia not only de-esterifíes but also introduces amide groups into the molecule. These amide groups will obviously affect the gelling properties (Kim, Rao and Smit, 1978a, 1978b).


0 0 5 0 . 1 0 .15

M o l a r Cd^ cone

Figure 17.2 The dependence of the rigidity modulus (G) of alginate and pectate gels on the concentration of calcium ions. The modulus was evaluated as the reciprocal of the creep compliance response measured 10 s after the application of the stress. The source data is that of Mitchell and Blanshard (1976a, 1976b)

As the degree of esterification of pectin is reduced, the gels formed with calcium have similar properties to calcium alginate gels. For example, both calcium pectate and calcium alginate gels are thermostable. One difference between the two systems is that pectate will gel at lower calcium levels and the optimum gel rigidity for pectate is reached with a smaller increase in calcium above the level required to initiate gelation than is the case for alginate. This is illustrated in Figure 17.2 which displays the rigidity modulus of alginate and pectate gels as a function of the level of calcium present. The data are derived from results reported by Mitchell and Blanshard (1976a, 1976b).

T H E R M A L D E G R A D A T I O N OF PECTINS

Among the polysaccharides used to control food texture, pectin shows a unique pH dependence for the rate of thermal depolymerization. Whereas other polysaccharides degrade by hydrolysis, which occurs far more rapidly at acid pH, pectin also degrades by a ß-ehmination mechanism which

distribution of carboxyl groups along the chain (Kim, Rao and Smit, 1978b). This latter parameter has been discussed in a recent review (Taylor, 1982). The carboxyl group distribution is undoubtedly a function of the de-esterification method used. Enzyme de-esterified pectins have a blockwise distribution of free carboxyl groups, whereas chemically de-esterified materials have a more random distribution. Thus Powell et al (1982) reported a completely different relationship between gel strength and degree of esterification for alkali and enzyme de-esterified low methoxyl pectins. Other chemical features, such as the presence of amide groups and the neutral sugar content and distribution, will also have an effect. Disentanghng these various factors represents a formidable problem, which is made more difficult by the fact that the influence of calcium concentration and the calcium release mechanism on gel strength will also depend on the chemical nature of the pectin molecule.

J.R. Mitchell and A.J. Taylor 251

C O O M e

A l k a l i

A c i d

C O O M e

Figure 17.3 Mechanisms for pectin degradation. Treatment by alkah causes degradation by ß-elimination. Treatment by acid results in degradation by hydrolysis

requires hydroxide ions, and therefore takes place more rapidly at neutral and alkahne pHs (Albersheim, Neukom and Deuel, 1960). The two mechanisms for pectin degradation are shown in Figure 17.3. ß-ehmination requires at least one of the two carboxyl groups on the adjacent residues to be in the esterified form. The rate of degradation by ß-elimination increases rapidly with temperature with a Qio for the reaction of 3.6 (Albersheim, Neukom and Deuel, 1960). It therefore follows that if the degree of esterification is reduced the neutral and alkah stability of the pectin molecule to heat degradation is improved. This is clearly illustrated by the resuUs obtained by Pilnik and McDonald (1968) which are reproduced in Figure 17.4. The viscosity decrease obtained when pectin solutions are autoclaved for 10 min at 120 °C is displayed as a function of pH and DE.

PECTINS A N D A L G I N A T E S IN C A N N E D F O O D S

Mitchell, Sommerville and Speirs (1978) compared these gelling agents in a model canning system. The thickening ability of sodium alginate (gulur-onic/mannuronic ratio —0.8), a low methoxyl pectin (DE = 22%) and


8 0

6 0

α

4 0

2 0

p H

Figure 17.4 Viscosity decreases on autociaving pectins for 10 min at 120 °C as a function of D E and pH. Data of Pilnik and McDonald (1968). The ordinate shows the specific viscosity after autociaving, expressed as a % of the initial specific viscosity

sodium pectate (DE = 4%), after autociaving at approximately neutral pH in a rotating retort for 40 min at 121.1 °C, was measured. A calcium release system consisting of sodium tripolyphosphate and calcium sulphate dihydrate (CaS04.2H20) was included. Only in the case of the pectate system was it possible to obtain homogeneous viscous solutions on cooling after retorting. Alginates give either very thin solutions, at low levels of included calcium, or lumps of gel in a thin hquid, at high calcium levels. Low methoxyl pectin (DE —22%) always gave solutions with a very low viscosity. This latter observation was explained in terms of thermal degradation of the pectin, resulting in a molecular weight below the threshold required for gelation. The sodium tripolyphosphate (STPP) and calcium sulphate system released more calcium as the STPP was degraded on heating to lower molecular weight phosphates that are less effective as calcium-chelating agents. Therefore the level of available calcium increased slightly during autociaving. This increase was sufficient to gel the calcium sensitive pectate system {see Figure 17.2), but not the less sensitive alginate system. Increasing the CaS04.2H20 level caused pregelation in the alginate system and, on autociaving the thermostable gel broke into lumps which then synerized. It was not possible to formulate a calcium release system that resulted in a sufficient increase in available calcium to gel alginate at the end of the autociaving cycle.

1 0 0 r

J.R. Mitchell and A J. Taylor 253

T i m e ( m i n ]

Figure 17.5 Can centre temperatures for systems thickened with pectate and starch (Coiflo 67). Processing was carried out in a rotating retort at a retort temperature of 121 °C (Mitchell et aL, 1978). Pectate concentrations: 0 .1% (Δ), 0 .2% (V), 0 .3% ( • ) , 0 .4% ( • ) , 0.5% ( · ) and 0.6% (O). Starch concentrations: 1% ( • ) , 2% ( • ) , 3 % ( x ) and 4% (0)

and the non-Newtonian behaviour would also be expected to be different. It should be appreciated that highly cross-linked starches are available which would have given better heat penetration behaviour than the one chosen for this study (O'Dell, 1979). Although thermo-reversible geUing agents such as carrageenan would also be expected to show good heat penetration properties, ionic gelhng agents such as pectate and alginate have the important advantage that the gel will not meU on reheating.

If the pectate concentration is increased to levels approaching 1%, it is possible to obtain recognizable gels after autoclaving. This is illustrated in Figure 17.6 which shows three canned meat products containing 1% levels of low methoxyl pectin, alginate and pectate respectively.

The lack of success in obtaining thickened or gelled, canned neutral pH products with alginate, discussed above, is perhaps at variance with the process described in a patent by McDermott (1966). Sodium alginate was used as a partial replacement for starch in thickened canned products to improve heat penetration, reduce the starchy flavour and avoid problems due to starch retrogradation. This process, which relies on controlhng the formation of a calcium alginate gel through an autoclaving cycle, was used commercially in a range of canned Chinese Foods (Messina and Pape, 1966). One possible reason why it was apparently successful is that the product was maintained at a high temperature after the alginate and calcium source had been mixed and was not allowed to cool until after

A possible advantage of employing pectate as a thickener in canned goods, rather than starch, is that the former allows more rapid heat penetration because it only thickens, after autoclaving, when the can cools. This is illustrated by Figure 17.5 which compares the rate of heat penetration into gelatinized starch and pectate systems. After cooling the viscosity of 4% starch and 0.6% pectate were roughly comparable though the temperature dependence of viscosity was different for the two systems

254

Figure 17.6 Canned meat products containing low methoxyl pectin, sodium alginate and sodium pectate. Products contained by weight 49% minced beef, 49.25% water, 1% gelling agent, 0.5% sodium tripolyphosphate and 0.25% C a S 0 4 . 2 H 2 0 , processed in a 210 x 300 can for 48 min at a retort temperature of 126 °C

0 . 4

0 . 3

Ε

Q .

α

0 . 2

0 .1

• A l g i n a t e - c a l c i u m

• B S A - a l g i n a t e - c a l c i u m

• M y o g l o b i n - a l g i n a t e -

c a l c i u m

6 . 0 6 . 2 6 . 4 6 . 6 6 . 8 7 . 0 7 . 2 7 . 4 pH

Figure 17.7 Effect of pH on the apparent viscosity of alginate-calcium and alginate-calcium protein systems measured at a shear rate of 173/s (Hughes et al., 1980). System contained 1% alginate, 1% protein and 0.006 Μ CaCl2

J.R, Mitchell and A.J.Taylor 255

Crude pectate systems in heat processed foods

O R A N G E W A S T E

Production of pectate pulp

Orange waste is available in large quantities as a by-product of the fruit juice industry. It is used extensively as a cattle feed but it is normally dried before incorporation into the animals' rations. If the wet waste is first treated with Ca(OH)2 then dewatering by pressing is facihtated and hence drying costs are reduced. A detailed account of this process is given by Rebeck and Cook (1977).

Orange waste contains high levels of the enzyme pectin methylesterase (PME), which has a pH optimum around 7.5-8.0 (Pilnik and Rombouts, 1979), although at higher pHs de-esterification occurs due to a combination of chemical (alkali) and enzymatic mechanisms. The pectin-degrading enzyme, polygalacturonase (PG), has not been rehably reported in oranges (Mannheim and Siv, 1964), although it is present in many other forms of plant tissue.

It is probable that the effect of the Ca(OH)2 treatment is to activate PME, thus causing partial in situ pectin de-esterification, and this change in the form of the pectin explains why the orange waste becomes much easier to de water by pressing. Analysis of a range of dried citrus pulps used for cattle showed that the degree of esteriflcation of the constituent pectin was typically in the range 20-40% (Mitchell, unpublished results).

autoclaving. It is known that it is possible to formulate an alginate gelation system which will remain hquid when hot and only set on cooling (Anon., 1979), though once the gel is formed it remains thermostable.

In addition to the differences in calcium sensitivity of pectate compared with alginate, another possible reason for the failure to obtain thickened canned meat products with alginate {Figure 17.6) may be some specific interaction with meat which inhibits alginate gelation. Edhn and Rocks (1969) have reported that alginate is less effective as a thickener in canned meat products compared with non-meat systems and claimed this problem can be overcome if high levels of magnesium ions are incorporated in the product. Some evidence for an interaction between myoglobin and alginate has been reported recently by Hughes et al. (1980). If myoglobin is included in a sodium alginate solution, containing levels of calcium just above the amount required to initiate gelation, then there is a significant reduction in viscosity, the minimum being at pH 6.3 {Figure 77.7). No such effect is found with myoglobin-pectate or alginate-bovine serum albumin mixtures. This pH is relevant to the canned meat products of interest, though it is of course far from certain that an observation made on dilute protein solutions at room temperature will be applicable to canned meat systems. It was suggested that the viscosity decrease was due to an electrostatic interaction between the protein and the polysaccharide. Myoglobin has an isoelectric pH of 6.9 and will therefore be positively charged at pH 6.3.


O r a n g e w a s t e f r o m juice p r o d u c t i o n

Ca(OH)2-

O r a n g e w a s t e f r o m j u i c e product ion

N a 2 C 0 3 -t o a b o u t pH 9 . 0

P r e s s W a t e r w a s h

Dry P r e s s

C a t t l e f e e d D r y

Grind

P e c t a t e pu lp

Figure 17.8 Comparison of processes for producing pectate pulp and cattle feed from orange waste

The effect of sodium carbonate treatment on waste from Valencia oranges has been investigated in detail by Speirs (1979). A suspension of albedo (400 g minced wet peel plus 600 ml water) was treated with sodium carbonate. The pH of the system was monitored during treatment and the final degree of esterification of the pectin in the treated material was measured by assaying the methanol liberated on de-esterification using gas-solid chromatography (Gouch and Simpson, 1970).

The intrinsic viscosity of the pectin extracted from the treated material with sodium hexametaphosphate was also determined (Speirs, Blackwood and Mitchell, 1980). The sahent points obtained from Speirs' investigation are shown in the composite Figure 17.9. From this the following facts are clear:

The process for the manufacture of pulp described by Baier and Wilson (1941) involved treatment of citrus waste with sodium carbonate rather than calcium hydroxide, and also included a water wash prior to pressing to remove soluble sugars and excess sodium carbonate. After drying, the pectate pulp is ground to a fine powder. A comparison of the two processes is given in Figure 17.8. In the authors' experience Na2C03 treatment is as effective as Ca(OH)2 in promoting dewatering. An addition level of 20-30 parts of sodium carbonate to 1000 parts of wet citrus waste was suggested by Baier and Wilson (1941). A yield of 100-140 parts of dry pectate pulp was obtained.

J.R. Mitchell and A J. Taylor 257

6 0 120 T ime(nn in )

(c)

6 0

A Δ 4 0

Δ

Lü Q _

2 0

— O

0 10 2 0

10^

3

5

N a 2 C 0 3 concen t ra t i on ( g / 1 0 0 0 g )

2 5

(d)

15 3 0 Tennpe ro tu re ( ° C )

4 5

Figure 17.9 Sodium carbonate treatment of orange waste (Speirs, 1979). (a) Effect of different concentrations of N a 2 C 0 3 on the pH of orange suspensions during de-esterification. (Minced orange peel 400g peel, 1000/g water with I g ( · ) , 5 g ( O ) , lOg ( Δ ) or 20g ( • ) of N a 2 C 0 3 at 25 ° C ) . (b) Time course of de-esterification for blanched ( • ) and unblanched ( O ) orange peel. (Minced orange peel suspension (400g/1000g) treated with 10g of N a 2 C 0 3 at 2 5 ° C ) . (c) Effect of N a 2 C 0 3 concentration on the D E ( O ) and intrinsic viscosity ( Δ ) of orange peel suspensions. (Minced orange peel suspension (400g/1000g) was treated for 120 min at 25 ° C ) . (d) Effect of temperature on the final D E of orange peel suspensions. (Minced orange peel suspension (400g/1000g) was treated with 10 g N a 2 C 0 3 for 30 min)

(1) A t the levels of Na2C03 addition suggested by Baier and Wilson (1941) (expressed in terms of wt of Na2C03/wt of wet peel) the p H of the system is raised initially to about 9.5 and then drops rapidly as de-esterification proceeds.

(2) De-esterification does not take place if the peel has first been blanched. This implicates PME in the de-esterification process.

(3) There is no evidence for pectin degradation as de-esterification proceeds.

(4) The rate and/or final extent of de-esterification decreases with decreasing temperature during treatment.

(b)


Conditions for Junctional behaviour in canned foods

If the gelation behaviour in canned meat products of pectate pulp prepared from oranges is compared with extracted pectate then, at equal polygalacturonic acid levels, it is found that pectate pulp performs better (Buckley, Mitchell and Burrows, 1978). The reason for this is not entirely clear but may be due to the water-binding properties of the celluloses and hemicellu-loses in the crude material.

Product development work on meat systems suggested that there was a critical DE for the in situ de-esterified pectin of about 20 above which the gelation process would not work (Buckley, Mitchell and Burrows, 1978). In order to define this more precisely Speirs, Blackwood and Mitchell (1980) devised an aqueous system containing guar gum, the treated orange waste and the calcium release system. The purpose of including guar gum was to maintain the insoluble constituents (CaS04.2H20 and

3 0 r -

10 2 0 3 0 4 0 5 0 6 0 S h e a r r a t e ( s )

Figure 17.10 Flow curves measured at 25 °C for canned systems processed at 121.1 °C for 75 min in a rotating retort. Processed suspensions contained by weight 1.5% guar gum, 0.5% sodium tripolyphosphate, 0.5% CaS042H20 ( • ) plus 0.5% de-esterified orange albedo ( · ) or 0.5% untreated orange albedo ( A ) (Speirs et al, 1980)

(5) Pectin DEs below 20% can be achieved after treatment of only 15-20 min. This latter observation is not consistent with the original patent describing the process (Wilson, 1938), which imphes that a treatment time of 12 h is necessary for extensive de-esterification.

There is an indication that lower DEs can be obtained by N a 2 C 0 3 treatment than with the Ca(OH)2 used in the cattle feed process. This may be because the added Na"*" in the former case plays a role in activating PME. Sodium ions are known to be important in this respect (Roxova-Bankova and Markovic, 1976).

259

0 . 9 i -

0 . 8 h

o7\-

- 0 . 6 f -

8 0 . 5 1 -

- 0A\

I 0 . 3 1 -

<

0 . 2 -

0 . 1 -

0

- L - L - L J 5 10 15 2 0 2 5 3 0 3 5 4 0

D e g r e e of e s t e r i f i c a t i o n iVo)

0 . 9 r

0 . 8

0 . 7

0 . 6

' 0 . 5

0 . 4

0 . 3

' 0 . 2

0 . 1

/ (b)

0 1 2 3 4 5 6 7 8 I n t r i n s i c v i s c o s i t y ( d l / g )

Figure 17.11 Conditions for functional behaviour of orange waste (Speirs et al., 1980). (a) Relationship between apparent viscosity of processed solutions and pectin degree of esterification in treated orange albedo. Viscosity was measured at a shear rate of 20 s from the rising part of the flow curve. Processing conditions as described in the legend to Figure 17.10. Pectin intrinsic viscosity was in the range 5 .0 -7 .0 dl/g. (b) Relationship between apparent viscosity and pectin intrinsic viscosity. Measurement and processing conditions as for Figure 17.11a. Pectin degree of esterification was in the range 10-14%


N O N - O R A N G E SYSTEMS

Other citrus systems

The pectin content of other citrus wastes (grapefruit, lemon and lime) is at least as high as orange, however the PME activity is somewhat lower (Rouse, 1977; Evans and McHale, 1978). From the limited amount of work that has been done it appears possible to prepare crude, heat-stable pectate gelling agents by N a 2 C 0 3 treatment of lemon, grapefruit and lime peel. In the case of grapefruit and lemon, treatment with N a 2 C 0 3 (11 g/814 g wet peel) at room temperature for 18 h was used; and this appeared adequate to reduce the degree of esterification of the pectin to below 20% (Buckley, Mitchell and Burrows, 1978). King (1982) has recently treated underripe limes with N a 2 C 0 3 for 2 h at room temperature in a system identical to that described by Speirs (1979) (10 g N a 2 C 0 3 to a suspension containing 400 g minced lime peel and 600 g water). The resultant material functioned well in a canned meat product. This result was surprising since the pH achieved was only 7.0 immediately after N a 2 C 0 3 addition, in contrast to 8.5 to 9.0 obtained with the less acid orange system, and the final in situ pectin DE was measured as 54%. A blanched control had a final DE of 59% and showed completely different behaviour, with no gelling or even thickening behaviour in the canned meat product.

Non-citrus systems

Speirs (1979) subjected a number of other waste materials {Table 17.1) to sodium carbonate treatment and then evaluated their performance in the

peel) in suspension during the initial stages of heat processing in the rotating retort. The peel concentration employed was 0.5% w/v, which was below the level required to give a firm gel. The product could therefore be evaluated by viscosity measurements. Figure 17.10 displays typical flow curves obtained for systems containing guar gum alone and systems containing blanched and unblanched treated orange peel. It is apparent that the guar gum degrades during the severe heat treatment conditions employed and contributes very little to the final viscosity. Using this system to evaluate the performance of the treated orange waste suggested that, for the material to be effective, a degree of esterification below about 22%, and a pectin molecular weight expressed in terms of intrinsic viscosity of above 2 dl/g, were both required {Figure 17.11) (Speirs, Blackwood and Mitchell, 1980). The sharp cut-off found for both DE and instrinsic viscosity is somewhat surprising. If degradation during autoclaving occurs by ß-elimination at all the esterifled regions, then, after heat processing, the molecular size of the remaining pectate chains will equal the size of the original free carboxyl blocks within the molecule. Work is currently in hand to determine if this free carboxyl block size also shows a discontinuity at a degree of esterification of around 20%. To this end a method has recently been developed to measure the block size of the free carboxyl regions (Tuerena, Taylor and Mitchell, 1981, 1982).


Orange albedo 6 -9 'Green' tomato skins 4.7 'Red' tomato skins 1.2 Potato peel 0.9 Mango peel 2.5 Grape skins 0.8 Pineapple bran 0.4 Pea pods 1.4

decrease in molecular weight with increasing maturity is probably due to the action of polygalacturonase which is present in tomato skins at high levels (Sawamura, Knegt and Brunsma, 1978). With the exception of pea pods, none of the other systems listed in Table 17.1 showed any significant thickening in canned systems after sodium carbonate treatment. The failure of sugar beet residue, potato peelings, grape skins and pineapple bran is probably due to the low molecular weight of the pectin present. Mango peel pectin had an intrinsic viscosity above this critical level but it was impossible to reduce the degree of esteriflcation below 33% using the Na2C03 treatment, and it may be that this was the reason why no thickening effect was obtained (Speirs, 1979). Despite the relatively low intrinsic viscosity of the pectin present in pea pods, Speirs (1979) reported a significant thickening effect. When compared at equivalent galacturonic acid levels the apparent viscosity in the guar system was about 50% of that achieved with orange waste. Taylor and Pritchard (1982) investigated the potential of pea pods as food thickeners in more detail and confirmed that a functional material with thickening properties could be produced from this source. De-esterification was carried out by Na2C03 addition to pea pods to maintain the pH at 8.0 for 30 min. Treatment was carried out both at room temperature and 50 °C for 30 min on blanched and unblanched samples. The lowest in situ pectin degree of esteriflcation was achieved for the unblanched sample at 50 °C where a DE of 10 was obtained compared with about 60 in the native molecule. The greater effectiveness of treatment at 50 °C compared with ambient is not surprising since pectin

autoclaved system previously described. Of the materials investigated only green tomato skins showed the characteristic pH fall indicative of in situ de-esterification. This is reasonable since, of the materials investigated, this was the only one with a reported PME activity of the same order as oranges (Vas et al., 1967). After retorting, green tomato skins gave viscosities similar to orange waste when evaluated at equal galacturonic acid levels, whereas mature red tomato skins did not thicken. The degrees of esteriflcation of the two materials were 5.6 and 4.8 respectively (Speirs, Blackwood and Mitchell, 1980). It thus appears that the difference in performance could be explained by the low molecular weight of the mature tomato skins as evidenced by their low intrinsic viscosities, which are below the threshold level required for the orange system {Table 17.1). The

Table 17.1 INTRINSIC VISCOSITIES OF PECTATES E X T R A C T E D FROM SOME Na2CO, T R E A T E D P L A N T WASTES (SPEIRS, 1979)

Intrinsic viscosity (dl/g)


Potential of crude pectate gelling agents

As is apparent from Figure 17.8 crude pectate gelling agents can be prepared from orange waste by a process which is similar to that which is currently used to produce a material suitable for cattle feed. It should therefore be possible to produce a crude gelling agent at a price which is very much lower than extracted polysaccharides. Thus Baier and Wilson (1941) talked about producing a firm gel containing about 2% pectate pulp at a materials cost of the order of $l/tonne. That is to say a pectate pulp cost $50/tonne. It goes almost without saying that this would bear no resemblance to a current price, but nevertheless it does suggest that if in some applications a material of this type could be used to replace pectins, carageenans or alginates which now cost of the order of $500(>-10000/ tonne, then there would be considerable financial benefit to the user.

A range of applications for pectate pulp in food systems has been demonstrated (Mitchell et al., 1978a,b) and some of these have been listed in the introduction to this chapter. It should be appreciated that although products with a satisfactory texture may be produced there are a number of 'negatives' associated with crude pectate systems which may prevent their use and the most important of these are listed in Table 17.2.

Table 17.2 FACTORS LIMITING T H E U S E OF PECTATE PULP IN F O O D SYSTEMS

(1) Flavour (2) Colour (3) Opaque gel (4) Heating in the presence of a sequestrant required to solubilize functional pectate

component (5) Brittle gel texture (6) Careful control of calcium release system required for homogeneous texture (7) Higher levels required than for extracted gelling agents

With orange waste the flavour problem may be reduced to some extent by removing the flavedo (the orange part of the peel containing the orange oil) prior to treatment. This is sometimes done anyway in conventional processes since orange oil is a useful by-product. In the pet-food industry many of the problems listed in Table 17.2 are not important and it is within this industry that this material has recently been used successfully, although only to a hmited extent. For this reason this chapter has concentrated on the apphcation of pectate gelhng systems in canned meat products.

methylesterase is known to be a relatively heat stable enzyme which may be activated on heating (Vas et al, 1961 \ Taylor Brown and Downie, 1981). De-esterification also occurred in the treated blanched samples although to a lower extent presumably because of chemical de-esterification. On a dry weight basis the treated pea pods contained about 10% galacturonic acid and when the sample with a pectin in situ DE of 10 was incorporated at a level of 2% on total product weight in a canned meat system a gelled pack was obtained (Taylor and Pritchard, 1982).


References

A L B E R S H E I M , P., N E U K O M , H. and D E U E L , H. (1960). Arch. Biochem. Biophys., 90, 46

A N O N . (1979). Gel Formation with Alginates. Alginate Industries Ltd, London

B A I E R , W.E. and WILSON, C.W. (1941). Ind. Eng. Chem., 33, 287 B A R T O L O M E , L.G. and HOFF, J.E. (1972). / . Agr. Food Chem., 20, 266 B U C K L E Y , K., MITCHELL, J.R. and B U R R O W S , I.E. (1978). Food Product and

Method. UK Patent 1 508 993 D O E S B U R G , J.J. (1965). Pectic Substances in Fresh and Preserved Fruits and

Vegetables. Sprenger Inst. Inst, for Res. Storage and Processing of Hort. Prod. Commun. No. 25. Wageningen, The Netherlands

D O E S B U R G , J.J. (1973). In Phytochemistry Vol. 1, p.273. Ed. by C P . Miller, van Nostrand Reinhold Co., New York

E D L I N , R.L. and ROCKS, J.K. (1969). Canned Creamed Meat-Containing Food Products by Incorporation of a Water Soluble Alginate and a Soluble Magnesium Salt. US Patent 3 480 450

E V A N S , R. and M c H A L E , D . (1978). Phytochemistry, 17, 1073 G O U C H , T.A. and SIMPSON, C F . (1970). / . Chromat., 51, 129 G R A N T , G.T. , MORRIS , E.R. , REES, D . A . , SMITH, P.J.C. and T H O M , D . (1973).

FEBS Letters, 32, 195 H O O G Z A N D , C. and D O E S B U R G , J.J. (1961). Food Technol., 15, 160 H S U , C P . , D E S H P A N D E , S.N. and DESROSIER, N.W. (1965). / . Fd Sci., 30, 583 H U G H E S , L., L E D W A R D , D . A . , MITCHELL, J.R. and SUMMERLIN, C. (1980). / .

Texture Stud., 11, 247 IMESON, A . P . , W A T S O N , P.R., MITCHELL, J.R. and L E D W A R D , D . A . (1978). / .

Food Technol, 13, 329

Other apphcations both inside and outside the food industry are clearly a possibility. For example Imeson et al, (1978) showed that crude pectate systems could be used to precipitate protein from waste streams. Another idea of interest is the possibility of treating materials that are natural components of food systems with dilute alkali to de-esterify the pectin in situ so as to give the normal ingredient functional behaviour (for example, treated orange peel as a gelling agent for dietetic marmalade or treated peas and onions as components of soups). A related development is the use of native PME to de-esterify pectin in situ in fruit or vegetables that are to be canned. Reduction of the DE reduces thermal degradation by ß-elimination and calcium ions naturally present or added to the canning hquor form calcium pectates in the cell wall. This process has had beneficial effects on a variety of canned products e.g. tomatoes (Laconti and Kertesz, 1941; Hsu, Deshpande and Desrosier, 1965) snap bean (Van Buren et al, 1960) cauliflower (Hoogzand and Doesburg, 1961) carrots (Lee, Bourne and van Buren, 1979) potato (Bartolome and Hoff, 1972) and mung beans (Taylor, Brown and Downie, 1981).

It will be interesting to see whether the revival of interest in pectate pulp, started by the discovery of an application within the pet-food industry, will result in this material being used for other applications.


K E E G S T R A , K., T A L M A D G E , K.W., B A U E R , W . D . and A L B E R S H E I M , P. (1973). Plant Physiol, 51, 188

KERTESZ, Z.I . , B A K E R , G.L. , JOSEPH, G.H. , MOTTERN, H.H. and OLSEN, A . G . (1944). Chem, Eng. News, 22, 105

KIM, W.J., R A O , V . N . M . and SMIT, C.J.B. (1978a). / . Fd ScL, 43, 74 KIM, W.J. , R A O , V . N . M . and SMIT, C.J.B. (1978b). ibid., 43, 572 KING, K. (1982). PhD Thesis, Univ. Nottingham L A C O N T I , J .D . and KERTESZ, Z.I. (1941). Food Res., 6, 499 L E E , C.Y. , B O U R N E , M.C. and V A N B U R E N , J.P. (1979). / . Fd ScL, 44, 615 M c D E R M O T T , F.X. (1966). Process for Preparing a Canned Food Product by

the Addition Therto of a Water Soluble Alginate. US Patent 3 257 214 M A N N H E I M , C.H. and SIV, S. (1964). Biochem. J., 92, 324 MESSINA, P.T. and P A P E , D . (1966). Food Eng., 38, 48 MITCHELL, J.R. and B L A N S H A R D , J.M.V. (1976a). / . Texture Stud., 7, 219 MITCHELL, J.R. and B L A N S H A R D , J.M.V. (1976b) ibid., 6, 341 MITCHELL, J.R., B U C K L E Y , K. and B U R R O W S , I .E. (1978a). Food Binding

Agent. UK Patent 1 525 123 MITCHELL, J.R., SOMMERVILLE, A . and SPEIRS, C I . (1978b). / . Fd Technol,

13, 425 M O R R I S , E .R. , G I D L E Y , M.J., M U R R A Y , E.J., POWELL, D . A . and R E E S , D . A .

(1980). Int. J. Biol Macromol, 2, 326 M O R R I S , E .R. , POWELL, D . A . , GIDLEY, M.J. and REES, D . A . (1982). J. Mol.

Biol, 153, 507 O'DELL, J. (1979). In Polysaccharides in Food. Ed. by Blanshard, J.M.V.

and Mitchell, J.R., p. 171. Butterworths, London PILNIK, W. and M c D O N A L D , R . A . (1968). Gordian, 68, 531 PILNIK, W. and R O M B O U T S , P.M. (1979). In Polysaccharides in Food. Ed. by

Blanshard, J.M.V. and Mitchell, J.R., p. 109 Butterworths, London POWELL, D . A . , MORRIS , E.R. , GIDLEY, N.J. and REES, D . A . (1982). J. Mol

Biol, 153, 517 R E B E C K , H.M. and COOK, R.w. (1977). In Citrus Science and Technology

Vol. 2, Ed. by Nagy, S., Shaw, P.E. and Veldhuis, M.K., p.368 AVI, Connecticut, USA

R E E S , D . A . and WIGHT, A .W. (1971). / . Chem. Soc. B., 1366 R O X O V A - B A N K O V A , L. and M A R K O VIC, O. (1976). Adv. Carbohyd. Chem.

Biochem., 33, 323 R O M B O U T S , P.M. (1972). Verslagen Landbouwk. Orderzoek., No. 179 R O U S E , A . H . (1977). In Citrus Science and Technology, Vol. 1, Ed. by

Zagy, S., Shaw, P. and Veldhuis, M., p.llO AVI, Connecticut S A W A M U R A , Μ., K N E G T , Ε. and B R U N S M A , J. (1978). Plant Cell. Phys., 19,

1061 SPEIRS, C.I. (1979). Heat Stable Food Gelling Agents from Plant Waste. PhD

Thesis, Univ. Nottingham SPEIRS, C.I. , B L A C K W O O D , G.C. and MITCHELL, J.R. (1980). / . ScL Food

Agric, 31, 1287 T A Y L O R , A.J. (1982). Carboh. Polymers, 2, 9 T A Y L O R , A.J . , B R O W N , J.M. and D O W N I E , L.M. (1981). / . Scl Food Agric.,

32, 134 T A Y L O R , A.J. and P R I T C H A R D , S. (1982). / . ScL Food Agric, 33, 384


T U E R E N A , C.E. , T A Y L O R , A.J. and MITCHELL, J.R. (1981). / . Sci. Food Agri., 32, 847

T U E R E N A , C .E . , T A Y L O R , A.J. and MITCHELL, J.R. (1982). Carboh. Polymers 2, 193

V A N B U R E N , J.P. , M O Y E R , J . C , WILSON, D . E . , R O B I N S O N , W . B . and H A N D , D . B . (1960). Food Technol., 14, 233

V A S , Κ., N E D B A L E K , Μ. SCHEFFER, Η. and KOVACS-PROSZT, G. (1967). Fruchtsaft Ind., 12, 166

W A L K I N S H A W , M . D . and A R N O T T , S. (1961). / . Mol. Biol, 153, 1075 WILSON, C.W. (1938). Pectate and Method of Making Same. US Patent No.

2 132 065

18

Introduction

The primary objective of the pet-food industry is to provide adequate nutrition to the pet animal and to maintain its heahh at a minimum cost to the owner. This approach is quite different from that of the agricultural feeds industry where maximum feed conversion in minimum time at minimum cost is the prime concern. The pet-food industry aims at two customers: there is the owner who buys the product and the pet animal that consumes it. To enable the product to reach the latter, the desires of the former must be satisfied. Experience has shown that a large majority of pet owners extend their preferences or dishkes from the human food area into that of the pet-food area. Should the purchaser not like any aspect of the product, it will not be offered to the pet, regardless of its nutritional qualities.

The edible products manufactured for domestic pets cover a wide area, ranging from treats such as confectionery, through complementary foods like biscuits and mixers, canned meats or fish products to complete foods containing animal and vegetable materials presented as dry, semi-moist or canned foods. In western Europe the industry annually uses around a million tonnes of meat and fish-based raw materials plus several hundred thousand tonnes of cereals and vegetables at present. The demand continues to rise. It is understood that similar quantities are used in North America and a somewhat smaller amount in Australasia. It is beheved that significant quantities are used in Eastern Europe but actual figures are unavailable. The pet-food industry is a user of large amounts of human food wastes. In Great Britain the industry is represented by the Pet Food Manufacturers Association (PFMA). Most major countries have a similar body. The PFMA represents the UK at the Federation Européenne de rindustrie des Ahments pour Animaux Familier (FEDIAF, : European Pet Food Industry Federation), which in turn represents the industry at Brussels.

Wastes used as raw materials

There are many waste materials from the human food industry used as primary raw materials. Examples are: blood, fish and fish by-products

267

UTILIZATION OF FOOD WASTES AS RAW MATERIAL IN THE PET-FOOD INDUSTRY

N.V. ASHLEY Applied Research, Pedigree Petfood, Melton Mowbray, UK

268 Utilization of food wastes as raw material in the pet-food industry

(filleting wastes), dairy waste (whey), bone, fat (tallow), poultry mix, egg waste, fish meal, dried yeast (brewery waste), abattoir wastes (offal meats), greaves, whole fish, oilseed waste, bran and millers offals. The meat offals are mainly unwanted organs such as green tripe (stomach), melts (spleen), udders, pig stomachs, lungs. Meat unfit for human consumption is also used. All the above is processed such that it is rendered commercially sterile. This ensures the complete destruction of any pathogenic or potentially pathogenic organisms. There is a complex chain of availability operating within the industry. The pet-food industry takes, as raw materials, waste products which the human food industry cannot, or does not want to, use or which are used in small amounts. As new technologies and techniques of food processing develop the human food industry is able to use previously unusable wastes. Thus the pet-food industry is increasingly researching into ways of using novel or 'exotic' raw materials as well as yet other human food wastes. Some examples of novel areas investigated or of potential benefit are single cell protein (SCP), mycoprotein and β protein of gluten. In the production of SCP energy requirements are low, yet it is high in protein and may have the correct nutritional balance required. Aspects to overcome when using novel wastes or products are those such as safety, appearance, acceptable forms (will it have the functionahty required or the correct texture?) and, in the case of SCP, the nucleotide base ratio.

There are many waste products of the human food industry which can be used together with novel waste or by-products of other industries. The difficulty is in actually using them to produce a product which fulfils all the nutritional, purchaser acceptance and commercial aspects of the final product. Pet-food raw materials are very variable, displaying intra- and interbatch variabihty. The challenge is to produce a uniform product having all the desired characteristics.

Properties of wastes as raw materials

Many waste materials intended for use as pet-food raw materials are susceptible to microbiological spoilage. To prevent or reduce such spoilage it is transported chilled ((M°C) or deep frozen (< -10°C). A proportion of fresh material is used which is not chilled. Rapid transportation and in-plant handhng is required. With regard to cereals and flours, the low water activity of these materials is usually sufficient to render them microbiologically stable, providing they are not mishandled.

Since many of the materials used are classified by their producers as waste, it is not uncommom for such material to be contaminated with other quite undesirable matter which has to be removed. To avoid this an active vendor assurance scheme is necessary. Deep frozen meat blocks may have polythene wrappers trapped within folds. For example, when lungs are wrapped and then evacuated—the polythene is trapped by masses of collapsed alveolar tissue. Such wrapping material presents problems when blocks are stripped and broken prior to use. Filleting waste, derived from dock-side operations, may contain metallic foreign bodies such as fish hooks. The fish skips, however, often have the appearance of a receptacle

N.V.Ashley 269

to dump anything unwanted such as wood, paper, etc. Fortunately, programmes of education aimed at the producer with respect to minimizing unwanted pollution of such waste products has been most effective, as have alterations in the way meats are packed. Increased sophistication in sensor technology has allowed a very high level of detection and ehmination of metalhc and non-metallic foreign bodies. The most desirable way of foreign body elimination and the one that is pursued with vigour is elimination at the supplier's or manufacturer's end.

Written raw materials specifications are produced giving precise details of the quahties of raw materials that will be acceptable and the limits of variabihty that are acceptable. Thus for certain meats there will be specifications for the percentage of bone, fat, smooth muscle protein; these descriptions vary depending on the role of the raw material in the final product. Vendor assurance schemes operate with the supphers who carry out their own strict quality control procedures. For example, specific properties may be required of a particular raw material such as hydration characteristics, sheer tolerance, break strength, etc. Manufacturers using specified tests, sell to these targets and the product is assessed by the industry to check whether it conforms; so the supphers' performance is audited. Many of the raw materials used, especially abattoir wastes and other proteinaceous wastes, may be contaminated with bacteria, usually originating from the gut of the animals, skins and hide and from abattoir equipment. Incoming raw materials such as these are subject to sensory organoleptic evaluation by skilled operators and may be subject to automated bacteriological techniques. In this way they are assessed as to their overall biological quahty. Evaluation by these methods greatly reduces the risk of pre-process spoilage of product. Thus considerations of hygiene are of very great importance. Raw materials arrive as non-sterile batches and are converted into a usable state (unwrapped, broken up, thawed, etc.).

Cleanliness at this stage is important to prevent a reservoir of bacterial infection, with the attendant spoilage problems. The plants should be designed to minimize direct handling of raw materials. Unfit meats should be treated to sterilize them and so kill all pathogens and spoilage organisms that may be present. The operational design of such plants is important to facihtate high process speeds and effective sterilization (often using hydrostatic continuous sterilizers as weh as static retorting) is most important—indeed a good knowledge of the heating profiles and characteristics of any such canned products is required. The commercially sterile canned products must be separated from non-sterile cans and materials. To achieve an acceptable product the hygiene has to be of human factory standards. Even with vendor assurance specifications, many of the raw materials used (since they are mainly wastes) have a significant natural variabihty and the industry has to produce a uniform product from them. Recipes for various products have been constructed so as to ensure proper nutrition for the pet. Each recipe has variations, so that if one ingredient is in short supply, it may be properly replaced to ensure the appropriate physical, organoleptic and nutritional properties are maintained {Tables 18.1-18.6). Extensive market research has found the preferences and desires of both the purchaser and the pet while scientific research has

270

Table 18.1 N U T R I T I O N A L P A R A M E T E R V A R I A T I O N S I N R A W M A T E R I A L S O F A N I M A L O R I G I N

Lungs No. Typical analysis Observed min-max Max. standard samples (%) (%) deviation

( Beef 72 Moisture 900 < Av. 76 60-89 7.8

V Pork 80 Protein 970 14.8 7 .2-21.4 2.65 Fat 970 9 1.3-36 8 Ash 336 1.4 0 .4-2 .6 1.1

The analyses differ significantly according to the species origin of the materials. Beef lungs: average protein 16.6%

average fat 13.8% average moisture—a standard difference of 8% between beef and pig lungs.

Pig lungs: average protein 12.9% average fat 4 .1%

Significant differences also occur between untrimmed lungs (i.e. with or without trachea). Trimmed lungs: average fat 5 %

average protein 16.3% (beef) 12.9% (pig) Untrimmed lungs: average fat 13.8% (beef) 6 .1% (pig)

average protein 15% (beef) 14.5% (pig)

Table 18.2 V A R I A B I L I T Y OF M E A T

Meat No. samples Typical Observed Max. standard analysis (%) min-max (%) deviation

Moisture 1301 67.2 41-84 .4 12.60 Protein 1290 14.1 7 .3-23.6 4.55 Fat 1310 18.7 4 .7-55 .6 11.03 Ash 691 2.3 0 .3-13.4 2.72

The term 'meat' covers all fibrous meat such as ears, tongues, trimmings, cheeks, lips, etc. from one animal or from a mixture of animals or species except poultry.

Table 18.3 V A R I A B I L I T Y OF F A T FISH

Fat fish No. samples Typical Observed Max. standard Fat fish analysis (%) min-max {%) deviation

Moisture 156 69.2 53.1-84.5 4.82 Protein 130 14.4 7-19 1.60 Fat 159 12.9 3 .8-26.7 4.46 Ash 117 2.9 0 .5-5 .2 1.86

Variability is intimately linked to the physiological state of the fish, e.g. pre- or post-breeding season.

Table 18.4 VARIABILITY OF P O U L T R Y

Poultry No. samples Typical Observed Max. standard Poultry analysis (%) min-max (%) deviation

Moisture 1155 66.1 23.4-83.1 5.71 Protein 1135 15.8 6 .6-22.0 3.54 Fat 1164 15.5 4 .2-35.1 7.39 Ash 538 6.8 1.2-16.3 4.02

The term poultry refers to carcases, necks, wing tips, back, heads, feet of chicken, turkeys, geese and ducks.

N.V.Ashley 271

Table 18.5 V A R I A B I L I T Y OF M E A T A N D B O N E M E A L

Meat and No. samples Typical Observed Max. standard bone meal analysis (%) min- •max (%) deviation

Moisture 299 7.9 1.7--14.7 3.10 Protein 299 46.1 31.2--66.3 3.82 Fat 223 6.7 2.1--16.3 2.74 Ash

Continent 120 43.2 38.2--49.8 2.81 UK 174 28.6 12.3--40.8 4.32

Table 18.6 E X A M P L E OF A F O R M U L A T I O N C H A N G E FOR A PET-FOOD OF M O I S T U R E C O N T E N T N O T E X C E E D I N G 14%

Standard formulation A Formula change'' Β (%) (%)

Ingredients Dry greaves 20 10 Meat meal 5 5 Fish meal 5 15 Meat and bone meal 10 10 Maize 20 20 Oats 20 20 Wheat 20 20 Total TÖÖ 100

A typical formulation change brought about by a temporary shortage of dry greaves and a better supply of fish meal.

elucidated nutritional requirements for a healthy life of the pet animal. The use of hnear programming, enables these areas to be synthesized into a product fulfilling all the required criteria. This technique is capable of calculating the lowest cost formulation which will comply with the finished product specification. By means of this approach several companies have been able to keep increases in price below the retail price index even though pet-food products carry 15% VAT unlike human foodstuffs. Long-term planning with forecasting of raw material supplies ensures an adequate buffer stock of raw materials in appropriate storage systems. This logistic approach is essential to ensure the production of a uniform end product.

Nutritional status of pet-foods

The waste products of food industries used as pet-food raw materials have been found to possess differing acceptability to pet animals. With respect to dogs and cats, it has been shown that there is a 'league table' of palatability regarding offal meats. Extensive research has identified factors which explain the higher palatability of certain raw materials and also factors imparting a reduced animal acceptance. Blending of these raw materials in short- and long-term product development programmes enables highly palatable, cost effective recipes to be achieved. These raw materials and those used in the feeding of other types of pet animal (e.g. fish, cage birds, cavies, etc.) have been exhaustively analysed as to their nutritional status. Individual amino acids have been quantified in the


various types of protein sources used, vitamin and mineral assays performed, amounts of co-factors, essential oils, fats, carbohydrates (usable by the animal and that not assimilable) and calorie density of the finished product assessed and standardized. This information has been used in conjunction with published information to ensure that the pet-food satisfies necessary requirements of the animal. Pet-foods are either complete or complementary diets and such is specified on the packaging. A complete food is just that, a carefuUy balanced product supplying aU the pet animal's needs: physiological and nutritional in a palatable form. The complementary forms tend to be protein rich and are designed to be fed with a filler biscuit. This latter supplies essential minerals, oils and its mechanical properties 'exercise' the teeth. It has been shown that the nutritional requirements of cats and dogs are different. The cat is essentially a more carnivorous animal, deriving more energy from amino acid catabohsm (and hence requiring more animal based protein) than the dog. Cats also have requirements for specific amino acids that dogs do not. Thus a cat-food will adequately feed cats and dogs, whereas a dog-food is not necessarily satisfactory for long-term cat feeding. It is thus very important that the pet-food is fully adequate for the target species since in many cases the pet animal will eat prepared pet-food for all of its life. This tends not to be the case with human foods so such rigour is not required to them.

Many new food technologies have been applied to pet-foods. Analogue techniques and meat reforming technology have been pioneered by the pet-food industry, as has forced functionality of raw materials. Regardless of the technology used the final product must be shelf stable. It must be chemically stable. Certain chemical changes affect palatability negatively, and as such are undesirable. It must be stable microbiologically. This is achieved by heat sterilization (canned foods), a very low water activity (biscuit) or a relatively low heat treatment combined with a reduced water activity (IMF). Also it must contain only permitted food additives. Should such an additive be delisted, it is removed from use by the responsible pet-food manufacturer. Thus, if the cat- or dog-food is eaten by the inquisitive offspring of the owner, the child, or anyone eating it, will be none the worse for the experience. Certain areas of preservation open to the human food industry are not available to pet-food manufacturers even though EEC rules may allow it, because they affect the quality of the food detrimentally. An example is preservation with sulphur dioxide. This inactivates vitamin Bi and since the dog or cat may eat a product all its life, then sulphur dioxide and other techniques having similar effects could produce a nutritional deficiency syndrome.

Thus the waste materials of food industries used by the pet-food industry as major raw materials must impart good nutritional and acceptance effects on the product. The texture and palatability must appeal to the animal and the visual and organoleptic impact of the product on the purchaser be favourable. The same criteria apply when novel foodstuffs are introduced. Single cell protein is an area of interest. These materials are being examined minutely as to feeding quahty. Certain SCP sources are low in some essential amino acids and some have certain essential amino acids absent. Also SCP often may contain a relatively large proportion of nucleic acid. Safe intakes for man has been given as around 2 g/day. The purines

N.V.Ashley 273

Alga Amino 1 acid Val He Leu Lys Phe Met + Cys Tryp Thr

Uronema gigas^ 6.8 4.0 10.5 6.3 4.7 — — 4.0 Ulothrix sp.'' 2.6 0.60 1.4 1.5 3.4 — — 1.8 Spirulina maxima^ 6.5 6.03 8.02 4.6 4.97 1.77 1.4 4.56 Spirulina platensis'' 5.7 4.93 7.95 4.34 3.63 2.78 0.88 4.02 Chlorella pyrenoidosa^ 2.7 L70 1.20 2.4 2.1 0.6 0.4 1.9 Scenedesmus acutus^ 7.0 4.2 6.6 5.0 3.6 2.10 1.2 5.8

Val, Valine; He, Isoleucine; Leu, Leucine; Lys, Lysine; Phe, Phenylalanine; Met, Methionine; Cys, Cysteine; Tryp, Tryptophan; Thr, Threonine. "Priestly (1975) ^Clement, Giddey and Menzi (1967) ^Narasimha^ra/. (1982) ^Combs(1952) "Soeder(1970)

present are converted to uric acid in man, who does not possess the uricase enzyme which enables the oxidation of uric acid to easily excretable allantoin to occur. The dog however possess this enzyme and so may tolerate SCP sources humans can not. Table iS. 7 gives some information on the essential amino acid composition of algae considered as SCP sources (Combs, 1952; Narasimha et al, 1982). The advent of the compound feedingstuffs directive of the EEC is expected to be translated into new feedstuffs regulations in the UK and to be operative by the end of 1982. This means that agricultural animal feed and pet-food will be more tightly and completely legislated for than will human foods.

The future of waste material use by the pet-food industry

It would appear that the human food industry may in the future use meats and by-products that now go for pet-food, to a greater extent as technologies for doing so become more sophisticated. The pet-food industry will turn more to novel or 'exotic' areas to find other raw materials which meet or can be made to meet its requirements. Examples are new sources of protein such as biomass derived from biotechnology processes and industry. Other protein sources that may also be of interest are, pea and other leguminous plants, peanuts, rape seed, SCP produced specifically as a foodstuff or food supplement and protein from lower animals (e.g. krill). All these areas offer useful avenues of exploration.

It should be borne in mind that it often appears that where the pet-food industry is today, the human food industry is tomorrow.

Acknowledgements

My thanks to Mr Ρ A Cheney, Mr G Andrews and Mr Κ Buckley for much useful discussion on areas covered by this paper.

Table 18.7 ESSENTIAL A M I N O A C I D COMPOSITION OF V A R I O U S A L G A E C O N S I D E R E D A S S O U R C E S OF SINGLE CELL PROTEIN


References

C L E M E N T , G., G I D D E Y , C. and MENZI, R. (1967). / . Sci. Food Agric., 18, 497 C O M B S , C F . (1952). Science, NY, 116, 453 N A R A S I M H A , D . L . R . , V E N K A T A R A M A N , G.S. , D U G G A L , S.K. and E G G U M , B . O .

(1982). / . Sci. Food Agric, 33, 456 PRIESTLY, G. (1975). In Food from Waste. Ed. b y G.G. Birch, K.J. Parker

and J.T. Worgan, p. 114. Apphed Sci., London S O E D E R , C.J. (1970). Ber.dt. bot. Ges., 83, 607

19

Introduction

Nutritional and health implications of novel food from waste are often recognized at a rather late stage of its development. This is due to the fact that initially only analytical characteristics and physical process parameters are measured. These factors provide only a limited possibility for predicting nutritional and health aspects of the product under development. Biological tests with rats produce more indicative data. Animal species, however, react rather specifically with regard to acceptance, digestion, absorption and metabohsm of a product; and they are not equally susceptible to toxic factors. Thus when, ultimately, testing is performed with the animals for which the product is destined (the target species), unpleasant surprises may still arise. By then, usually, high development costs have already been incurred. Such a situation, tends to be embarrassing.

Much can be gained by an awareness of these risks. A cost effective product development programme should take them into account in advance. As a starting point a thorough chemical and physical characterization of the waste material and a definition of the objectives for the ultimate use of the upgraded product are mandatory. Incorporation of simple biological tests that are adequate in the light of the intended use of the product, are equally helpful for the early recognition of dead alleys and therefore for the avoidance of unnecessary costs.

In this chapter there is presented some experience of our Institute in studying nutritional and health implications of the use of novel feed sources, in the hope that it might contribute to the effective development of upgraded waste products of which so many promising examples have been described in this book.

Elements in waste processing relevant to nutritional and health implications

Figure 19.1 illustrates the main lines along which the possible upgrading of agricultural waste may proceed to animal or human food.

NUTRITIONAL AND HEALTH IMPLICATIONS

P. VAN DER WAL ILOB, Research Institute in Animal Nutrition and Toxicology, Wageningen, The Netherlands

277

278 Nutritional and health implications

T H E R A W M A T E R I A L S

Some nutritional and safety characteristics that may cause products to be categorized as waste materials and (more specifically) prevent their use as food, can easily be illustrated (Van der Wal, 1979a). A low content of protein and/or energy will limit their potential contribution to the supply of nutrients. Straw (with 3-4% protein) and many other lignocellulose-rich products are classic examples.

Low digestibility or bioavailability of these materials that are rich in lignocellulose, further hmits their utilization. Low acceptability or palata-bility are limiting factors for using manure as a feed, another of our abundant agricultural residues. Safety, of crucial importance for use as a feed, is inadequate or at least rather unpredictable when we consider solid urban waste or residues from sewage systems. Hazards for chemical and microbial contamination are obvious. Economic aspects place a material in the category of waste if it is widely dispersed in small amounts over a wide area or if it is heavily diluted. If the availability is limited to one season (as is often the case with agricultural residues), this adds to the logistic problems causing negative consequences for cost aspects.

If only small quantities of waste are available, the development costs are seldom justified. Large quantities facilitate a standardization of the feed product, which is a necessity for adequate feed quahty control.

T H E OBJECTIVES

When it is desired to upgrade a waste the above-mentioned constraints in the chemical, physical and organoleptical sense, must be alleviated while controlling the economics of the final product. Especially for the safety requirements that are to be met, there is a little room for manoeuvre.

Regulatory authorities, while protecting consumers and animals from heahh hazards, have erected barriers in many countries. These are sometimes hard to surmount. The international market for food and feed ingredients usually sets rather strict limits to the economics of upgraded wastes as well. Fortunately in the case of animal feed the economic prospects are fairly predictable. The price can then be calculated accurately in advance, when the content and the bioavailability (i.e. digestibility) of the nutrients is known.

The more expensive process development and research are, the more one is inclined to look direct for a market in the human food sector, where (usually) higher prices can be claimed than for animal feed. Organoleptical and psychological acceptabihty problems then arise. Consumers are usually slow in accepting novelties and tend to maintain standards for acceptability regardless or their state of nutrition. The attempts to introduce novel sources of nutrients in developing countries form a challenge that is seldom met successfully.

In most cases it is considered advisable to explore the potential for an upgraded waste product in the animal feed market first. In that field the nutritional, safety and price requirements to be met can be measured objectively. Furthermore, the nutritional and toxicological data that have

p. Van der Wal 279

T H E R O U T E S FROM WASTE TO F O O D

In Figure 19.1 three possibilities for waste processing are mentioned. They can be applied either alone or in combination with each other.

Chemical and physical processing have been dealt with in the course of this book in the form of mechanical meat processing, ultrafiltration, precipitation and texturization. For upgrading of wastes to animal feed, chemical treatment of straw with NaOH has improved digestibility (Jackson, 1978). More recently it has been shown that the addition of urea, under tropical conditions, leads to an improved digestibility; and involves fewer safety hazards (Verma et al, 1982). In general, the toxicity of chemicals used and lessened bioavailability of nutrients through heat treatment, are important points to be observed in this category.

O r g a n i c r e s i d u e s C e l l u l o s e r i c h S t a r c h y , s u g a r y Animal m a n u r e

C h e m i c a l and I physical p r o c e s s e s

M i c r o b i a l c o n v e r s i o n

A n i m a l c o n v e r s i o n

Mush- Meat Milk E g g s r o o m

H u m a n f o o d

Figure 19.1 From waste to food along various lines of processing

Microbial conversion may lead directly to products for human consumption, as is the case with mushrooms grown on straw. Microbial conversion of several wastes received considerably greater attention following the development of large scale single cell protein production processes on the basis of gas-oil, paraffins and methanol (Van der Wal, 1976). With these materials it has been demonstrated that microbial proteins can be used as a major protein source effectively and safely. Bioconversion processes with waste as a substrate are under development in various contexts (Anon.,

to be acquired for obtaining clearance as animal feed, provide a good starting point for the evaluation of the acceptability in the human market as a follow-up target.


1979; Bellamy, 1983). Potential pathogenicity and possible toxic production by the micro-organisms require consideration in such processes.

Conversion by animals, either directly or in combination with chemical or microbial pretreatment, forms the classic route along which wastes are converted into food. The system copes effectively with the problem of consumer acceptance. Furthermore the capacity of the ruminant in utilizing materials rich in lignocellulose offers a unique mechanism for converting what is by far the most abundant residue in the world, straw, into human food. In the case of the animal conversion process, nutritional problems are dealt with at the animal level. Safety problems may have to be faced in terms of potential transfer of toxic substances (heavy metals, pesticide residues) or bacterial contamination {Salmonellae spp.) from the waste to the animal products.

General aspects, having nutritional and health implications, which are inherent in all technological treatments in the above-mentioned processes, are the interrelated factors of scale, sophistication, standardization and quality control. The case may be illustrated by comparing as examples large (10000 tonnes per annum) well-controlled production units for single cell protein from straw on the one hand {Figure 19.2), with a small scale (500 tonnes per annum) mushroom production unit, based on straw and

Figure 19.2 Yeast production for feed on corn residues in Bulgaria, applying fermentation technology on an industrial scale

Figure 19.3 Mushroom production for food on rice straw in a developing country, applying appropriate technology

ρ. Van der Wal 281

Nutritional implications

Q U A N T I T A T I V E ASPECTS

In many countries wastes form a potential source for food that is hardly tapped. This is of enormous relevance for countries with a food production that does not meet their requirement and that cannot be improved by an increase of the production of the classic food crops.

The amount of cassava residues in Asia alone (30 million tonnes) illustrates the wide perspectives in the category of starchy and sugary residues. Even more fascinating is the potential of 650 million tonnes of straw that is produced annually in Asia. Of the same order of magnitude as the lignocellulose rich wastes is the amount of animal manure. The above-mentioned perspectives have been recognized worldwide, and discussed in conventions organized by the United Nations. Increased efforts to solve the related problems in a way relevant for developing countries, are justified (Muller, 1980).

Q U A L I T A T I V E N U T R I T I O N A L ASPECTS

Protein and energy are usually by far the two predominant nutritional elements of a food that determine its price and its place in the diet. Minerals and vitamins seldom play a significant role in the value of an upgraded waste. If psychological or palatability problems of a food do not interfere, the content of these constitutents form an adequate and objective base for price calculations.

Table 19.1 CHEMICAL COMPOSITION (%) OF MICROBIAL PROTEIN, G R O W N O N W A S T E S U B S T R A T E S , A N D OF S O Y B E A N OIL M E A L

Yeast Bacteria Fungi Algae Soybean oil meal

Dry matter 96 90 86 94 88 Ash 6 8 2 7 6 Organic matter 90 81 84 87 82 Crude protein (N x 6.25) 60 74 32 52 45 True protein (amino ac id-N x 6.25) 47 55 22 46 38 Crude fat 9 8 5 15 1 Crude fibre — — 28 11 6 Nitrogen free extract 20 — 20 12 30

making use of simple appropriate technology on the other hand {Figure 19.3). It is obvious that a guarantee of nutritional and toxicological adequacy form quite different problems in both cases. It is encouraging that household bioconversion has been successfully applied for many ages, in particular in South Asia. This circumstance, however, may not convince pubhc health authorities that extrapolation and upscaling of these processes is justified without a more systematic analysis of the health hazards (Shacklady, 1979).


Yeast Bacteria Fungi Algae Soybi

Lysine 7.0 5.5 4.8 4.6 6.2 Meth + cystine 2.9 3.1 2.5 3.2 2.9 Arginine 4.8 4.7 5.2 — 7.2 Histidine 2.0 1.9 2.0 — 2.5 Isoleucine 4.5 3.9 4.1 3.1 4.9 Leucine 7.0 6.3 6.4 7.0 7.6 Phenyl + tyro 7.9 6.2 8.1 6.0 8.4 Threonine 4.9 4.2 4.4 4.9 4.2 Tryptophan 1.4 0.8 1.4 1.7 1.3 Valine 5.4 4.8 5.6 4.7 5.0

Under favourable conditions products with a high nutritional value may result from the upgrading of waste. This is illustrated in Tables 19.1 and 19.2 (UNU, 1978). The composition of microbial proteins that can be grown on waste substrates, are here compared with a classic high quahty protein source, soya. Table 19.1 demonstrates, at the same time, those analyses which have to be carried out to characterize the nutritional potential of a feed. In Table 19.2 di similar comparison is made of amino acids.

The content of amino acids that are essential for growth, production and reproduction of men and animals, to a large extent determines the nutritional value of the protein fraction. Of course the values presented can be entirely different for upgraded waste resulting from other types of processing than microbial conversion. The nutritional parameters with which their nutritional value is measured, remain the same. It is worth emphasizing at this point that microbial conversion, chemical and physical processing may add significant amount of nutrients to the waste.

SPECIFIC R E L E V A N C E OF DIGESTIBILITY A N D BIOA V A ILA BILITY OF W A S T E P R O D U C T S

Although the above-mentioned analytical data provide the basis for the nutritional evaluation of upgraded waste, the part of these nutrients that can be utilized by the animal must be determined subsequently: the fraction of the nutrients that is digested. Digestibility is highly specific for species. Therefore this factor has to be determined with the animal for which the product is destined.

Table 19.3 DIGESTIBILITY COEFFICIENTS IN PIGS FOR MICROBIAL PROTEIN, G R O W N ON W A S T E , A N D OF S O Y B E A N OIL M E A L

Yeast Bacteria Fungi Algae Soybean oil meal

Organic matter 92 90 79 — 83 Crude protein 90 93 71 54 91 Crude fat 95 87 34 — 34 Crude fibre — — 99 — — Nitrogen free extract 94 — — — 94 Metab. energy (MJ/kg) 16.15 15.55 12.30 — 13.35

Table 19.2 A M I N O A C I D COMPOSITION (g/16 g N) OF MICROBIAL PROTEIN, G R O W N ON WASTE SUBSTRATES, A N D S O Y B E A N OIL M E A L

p. Van der Wal 283

Pigs Chickens

Organic matter 79 24 Crude protein 71 59 Crude fat 34 18 Crude fibre 99 6 Metab. energy (MJ/kg) 12.30 4.20

A striking example of specificity of the digestive capacity of species is given in Table 19.4 (van Weerden and Schutte, 1980). Digestibility coefficients in pigs and chicks for the same product (a fungus grown on waste products), show here a dissimilarity to a degree that is unusual. In this case the use of the product for poultry is very hmited. Such disadvantages should be recognized early in the development of a new product for the market.

The need for a timely biological evaluation of waste material with the target species, i.e. the pig, is again illustrated in Figure 19.4 (van Weerden, Slump and Huisman, 1980).

In the upper line of the graph the digestibility of the separate amino acids is presented for a mixture of feed constituents with a high digestibihty and a low crude fibre content. The separate figures stay close to the average, the digestibihty coefficient of total protein. In these products one

4°/o c r u d e f i b r e r a t i o n

14% c r u d e f i b r e r a t i o n

- - - Ε >, 0; > > >

Ε

Q . cn I/) • ^ ^ O cn CD Ι

Ο.

Figure 19.4 Apparent digestibility coefficients of separate amino acids with relatively low and high fibre contents in pigs

Table 19.3 shows the digestibihty coefficients and metabohzable energy values for pigs of the same products shown in Table 19.1 and 79.2. Not only is there a difference in the digestibility coefficients for the various products, but also for the constituents within each product. The above-mentioned discrepancies tend to increase in products with a relatively low digestibility and a high crude fibre content. These characteristics are frequently found in waste products.

Table 19.4 D I G E S T I B I L I T Y C O E F F I C I E N T S O F A F U N G A L P R O D U C T G R O W N O N A W A S T E S U B S T R A T E


Health implications

Rather than overemphasize the health hazards of upgraded wastes, the potential of these materials especially in many developing countries, should be recognized.

A second major element in the evaluation of wastes tends to be overlooked. It is the fact that a waste is a potential feed and that it is not substantially different from other feed in its health hazards. Hence the requirements to be met are not different either.

M I C R O B I A L C O N T A M I N A T I O N

This hazard, especiaUy from potentially pathogenic micro-organisms, may easily occur in waste materials that have been in close contact with humans, animals or their excreta.

Salmoneha is probably the most widely recognized threat. It has attracted increasing international attention recently. Although only a very small proportion of the Salmonellae spp. found in feed are pathogenic, it is accepted that a pathogenic strain may follow the route: waste ^ f eed -»animal -^man. Table 19.5 illustrates how worldwide the contamination of feed with Salmonellae is spread, even in relatively common feed constituents (Cornelisse, 1974). Salmoneha contamination is hkely to occur in waste of human and animal origin like urban sewage, solid urban waste and manure. While these materials are under investigation as

would therefore not have made major errors when determining total protein digestibility only, and by assuming the same figure for all amino acids.

The situation is dramatically different in the lower hne. Here the same figures are given for a mixture of products with a relatively low digestibility and a high crude fibre content, characteristics common for many waste products. The coefficients differ very markedly from the average. In the case of the very important amino acid, lysine, one would have seriously overestimated the nutritional value of such a product when using the average values (protein digestibihty). This leads to the conclusion that, until the chemical analysis and the digestibility, with target species, are determined as specified, a considerable degree of uncertainty about the applicability and the value of an upgraded waste in the intended fields must be accepted. Uncertainties remaining after the above paramount evaluation concern mainly acceptabihty. This aspect is partly covered in (or prior to) the digestibihty studies so that serious shortcomings are detected in time. Finally the bioavailabihty of digested nutrients may offer problems, in respect of amino acids. Lysine can be damaged relatively easily during processing, because of its susceptibility to excessive heat treatment, a not uncommon feature in the processing of waste.

The total pattern of the nutritional value has, of course, ultimately to be verified in comparative feeding trials before the intended use can be realized on a large scale.

p. Van der Wal 285

Meal/feed Origin Batches Batches examined positive

Soybean meal U S A 2 2 Soybean meal Brazil 4 3 Soybean meal (pellets) Brazil 46 16 Sunflower meal (pellets) Argentina 18 4 Groundnut meal (pellets) Argentina 2 — Tapioca meal (pellets) Thailand 1 1 Feed for turkeys (pellets) Netherlands 1 — Feed for piglets (pellets) Netherlands 1 1 Feed for mice Netherlands 1 1 Feed for dogs Netherlands 6 3 Feed for minks Netherlands 2 2

Total 84 33 (39%)

substrates for growing yeast and algae, it should be emphasized that a heat treatment of either substrate or end-product is necessary in order to prevent proliferation of Salmonellae. Relatively simple methods prove here to be effective (van der Wal, Frik and van der Schaaf, 1966). There are indications that a treatment with lower fatty acids may offer a viable alternative (Van der Wal, 1979b)

In modern biotechnology, in operations which make use of microorganisms, it is essential to ensure that the micro-organism used is not pathogenic or closely related to a pathogen. It is also important to make certain that mutation of the organisms should not occur easily. There is divergence of opinion as to how to define pathogenicity more precisely; and on what test the evaluation should be based. Concepts based on the intravenous injection of organisms in the test animal seem to widen the range of pathogens to a bewildering extent.

TOXIC S U B S T A N C E S

These may enter waste products from various sources. Endogenous toxic constituents occur in many plants and micro-organisms. Contamination from exogenous origin may lead to unacceptable levels of a wide range of toxic factors in waste. Finally, during processing, a number of factors may introduce health hazards.

As in the case of the nutritional evaluation, two options are open for assuring acceptable safety aspects: analytical data and biological testing. Where the hst of necessary analyses for a nutritional evaluation is based on the search for a limited hst of well defined constituents, the list of potentially toxic constituents is far longer and bound to be incomplete. Within the EEC, a base is provided in four guidelines of the Council of Ministers that contain information about unwanted and unacceptable materials in feed:

74/63: gives maxima for unwanted materials and products. It covers primarily the heavy metals, nitrite, aflatoxin and toxic materials of vegetable origin (EEC, 1974).

Table 19.5 S A L M O N E L L A C O N T A M I N A T I O N OF F E E D COMPONENTS A N D M I X E D F E E D


Systematic stepwise evaluation of nutritional and health implications of novel products

Testing of safety and nutritional value is time-consuming and expensive. Exploring dead alleys in technological research and development could be even more so. A well-balanced and cost effective planning and coordination of the two elements proves to be feasible. In the course of the years we gradually developed a system for a stepwise evaluation of novel feed ingredients (Van der Wal, 1979c). It is based on the earher experience with the evaluation of single cell proteins grown on gas-oil and paraffins, and on the design of the guidelines of the Protein Advisory Group for testing novel feed (Protein Advisory Group, 1974). The system makes use of a flexible approach to be tuned to the data acquired during development and aims at providing:

(1) Timely guidance during technological research and development for optimization of process conditions.

(2) Adequate experimental evidence for clearance, for use by regulatory authorities.

(3) Nutritional evidence enabling potential customers to evaluate the merits of the product.

All experiments are designed to provide a maximum of data for determing nutritional and health implications of the product.

INITIAL ASSESSMENT OF T O X I C O L O G I C A L A N D N U T R I T I O N A L ACCEPTABILITY

In this stage (phase 1) a basis of data is created for a decision on further technological development and a more elaborate evaluation of safety and nutritional value.

If the results of this assessment are encouraging, a preliminary decision can be made regarding nature and extent of the evaluation programme for the product in phase 2 and, eventually, in phase 3.

70/524: dealing with additives in animal nutritiop (EEC, 1970). 81/851: regulating veterinary drugs (EEC, 1981). 82/471: refers to products used in animal nutrition and covers mainly

products of microbial origin (EEC, 1982).

Although the composition of the product may be in agreement with these guidelines, however, the health hazards are not yet completely identified. When biological materials are involved, a complete chemical analysis of potentially hazardous constituents is virtually impossible because of the sheer number of the constituents in a living cell. As a result, evidence for safety of a biological material can only be obtained by adequate biological testing with relevant animals under controlled conditions that make possible transfer of experimental data to practical conditions.

p. Van der Wal 287

Micro-organism involved

Possible pathogenicity, potential toxin production and the genetic stability are among the parameters to be studied in order to determine the safety risk that might result from the micro-organisms concerned.

Process conditions

These will be evaluated regarding the physical treatments that might adversely affect the quahty of the product. The risks of residual chemical solvents, the degree of standardization that might be obtained and the risk of contamination of the product with undesired micro-organisms are evaluated.

Product

The chemical, physical and bacteriological characteristics and the degree of uniformity of the product will be determined to provide a base for nutritional and toxicological evaluation.

Degree of exposure

The degree of exposure of the animal and human consumers of the product, which will be encountered directly or indirectly, will be evaluated.

Short-term feeding trials

Short-term feeding trials with target species (and possibly rats) will be carried out to determine the acceptabihty of the product to the animal for which it is destined and the effect on the animals at a range of levels of inclusion in the diet.

Substrate

The chemical composition, the degree of standardization and the degree of novelty of the substrate will be determined in order to define the potentials for use and the safety risks involved {Table 19.6).

Table 19.6 INITIAL ASSESSMENT OF T O X I C O L O G I C A L A N D N U T R I T I O N A L A C C E P T A B I L I T Y

Substrate (analysis) Organism (pathogenicity and toxin prod.) Process (physical and chemical hazards) Product (chemical and bacteriological analysis) Exposure (degree and frequency) {In vitro toxicity testing)


C H R O N I C TOXICITY STUDIES

If in the stages 1 (initial safety and efficacy assessment) and 2 (sub-chronic evaluation) potential hazards cannot be excluded with certainty, classic chronic toxicity studies for carcinogenicity and teratology must be carried out on laboratory animals and target species; and the validity of transferring results between laboratory and target species identified. In the foreseeable future these risks may be determined increasingly in the initial safety assessment period by the application of in vitro toxicity studies which are under development in order to collect information on potential carcinogenicity at an early stage. Materials that conform to the requirements in tests carried out in accordance with internationally approved experimental procedures (preferably Good Laboratory Practice), are likely to be acceptable internationally (FDA, 1978; OECD, 1981). It is neither possible nor even desirable to be dogmatic about the extent to which toxicological or nutritional evaluations should be carried out. This decision must rest with the authorities in the countries concerned in the hght of prevaihng circumstances.

Summary

(1) In many countries the population will increasingly have to rely on upgraded wastes as a source of nutrients.

E V A L U A T I O N WITH L A B O R A T O R Y A N D T A R G E T SPECIES OF A N I M A L S

When the resuhs of the above-mentioned safety and efficacy assessments are satisfactory, a more extended evaluation programme is carried out, with rats and with at least one avian or mammalian species by which the product will be consumed. The use of target species is necessary in order to ensure applicability of the experimental evidence to the food production systems.

In the studies with the target species the experimental design regarding groups, feeding levels and parameters studied should be such that a maximum of both toxicological and efficacy data will be obtained. An optimum use of test substances, animals and time will thus be achieved leading to the most economical programme. After short-term acceptabihty and toxicity evaluation, the following may be undertaken.

(1) Digestibility and energy determinations are carried out with the species for which the product is destined.

(2) Subchronic evaluations are carried out with rats and with at least one food producing target species (poultry, pigs, ruminants).

(3) Reproduction studies are optional and depend on the intended application of the product. If breeding animals are involved, these studies might be considered necessary.

(4) Metabolism studies are optional as well. Where in the initial safety and efficacy assessment and/or in the early stages of the biological evaluation programme, special risks or potentially hazardous substances can be identified, specific metabolism studies may have to be carried out.

ρ. Van der Wal 289

References

A N O N . (1979). Proc, 5 th Intl, Conf. Global Impacts of Applied Microbiology (Bangkok)

B E L L A M Y , W . D . (1983). In Upgrading waste for feeds and foods, Eds D.A. Ledward, A.J. Taylor and R.A. Lawrie. p.141. Butterworths; London

CORNELISSE, J.L. (1974). The isolation of salmonellae from animal feed-stuffs of vegetable origin and from mixed feeds. PhD Thesis, Utrecht, The Netherlands

E E C (1970). Additives in animal nutrition. Guideline 70/524. Council Europ. Commun; Brussels

E E C (1974). Maxima of unwanted materials and products in animal feeds. Guideline 74/63. Council Europ. Commun; Brussels

E E C (1981). Regulations of the use of veterinary drugs. Guidehne 81/851 Council Europ. Commun; Brussels

E E C (1982). Products used in animal nutrition. Guidehne 82/471. Council Europ. Commun; Brussels

F D A (1978). Good Laboratory Practice (GLP) regulations. Department of Health, Education and Welfare. Fed. Reg, 43, no. 247, Food and Drug Administration, USA

JACKSON, M.G. (1978). FAO Anim, Prod, Hlth Pap, No. 10, FAO; Rome M U L L E R , Z .O. (1980). FAO Anim, Prod, Hlth Pap, No, 18. FAO; Rome O E C D (1981). Guidelines for national GLP inspection and study audits.

OECD-guidelines GLP/81.63 (Paris)

(2) The nutritional value of novel products from wastes should predominantly be based on its contents of energy and digestible amino acids.

(3) The nutritional characteristics have to be determined as extensively and as early as possible with species that form the target for the use of the product.

(4) Palatability and acceptability problems are such formidable barriers to the application of novel products in man, that use as an animal feed usually is considerably easier and cheaper to achieve as a first objective.

(5) The main health aspect of the use of waste is its tremendous and unique potential to improve the nutritional status in developing countries, which is an unrivalled contribution to the health of the population.

(6) Negative effects on health can be controUed in most cases by adequate scientific identification and subsequent elimination via appropriate processing. Costs form the main bottleneck here rather than insurmountable technological problems.

(7) During the research and development phase of new processes for upgrading of waste, biological checks on nutritional and safety characteristics can provide timely guidance to avoid marches into dead aUeys.

(8) For the evaluation of nutritional and safety characteristics that will lead to acceptance by producer, consumer and regulatory authorities, emphasis should be on methods that are internationally standardized, cost effective and tuned to the target species for the novel products.


PROTEIN A D V I S O R Y G R O U P (PAG) (1974). Guideline on nutritional and safety aspects of novel proteins for animal feeding. Guideline No. 15. United Nations; New York

S H A C K L A D Y , C.A. (1979). UNU Food and Nutrition Bull. Suppl. 2: Biocon-version of organic residues, UNU; Tokyo, p. 156

U N U (1978). Food and Nutr. Bull. Suppl. 2. Bioconversion of organic residues for rural communities. UNU; Tokyo

V E R M A , M.L. , A G A R W A L , J.S., JAISWAL, R.S. and SINGH, R. (1982). 3rd Ann. Sem. Maximum livestock production from minimum land, Joydelpur, Dacca, Bangladesh

W A L , p. V A N D E R (1976). Proc. 1st Intl. Sympos. Feed Composition, Animal Nutrition Requirements, and Computerization of Diets. Utah State University; Logan, Utah. p. 227

W A L , P. V A N D E R (1979a). UNU Food Nutr. Bull. Suppl. 2: Bioconversion of organic residues, UNU; Tokyo p. 3

W A L , P. V A N D E R (1979b). Wld Poult. Sci. J. 35 (2), 70 W A L , P. V A N D E R (1979c). COST Workshop Production and Feeding of

Single Cell Protein, Jülich; FRG W A L , P. V A N D E R , FRIK, J.E. and S C H A A F , A . V A N D E R (1966). Wld Congr.

Animal Feeding, Madrid p. 323 W E E R D E N , E.J. V A N , and SCHUTTE, J.B. (1980). Proc. COST Workshop

Production and Feeding of Single Cell Protein. Jülich; FRG. p. 129 W E E R D E N , E.J. V A N , SLUMP, P. and H U I S M A N , J. (1980). 3rd EAAP Sympos.

Protein Metabolism and Nutrition, Braunschweig, FRG p.207

20

ECONOMIC CONSIDERATIONS

P.N. WILSON andT.D.A. BRIGSTOCKE BOCM Silcock Ltd, Basingstoke, UK

Introduction

Wastes are resources out of place (Taiganides, 1979) which, for economic or social reasons, are not fully utilized. However there is an enormous difference between wastes generated by the food production process and industrial wastes. The former are natural by-products which can be recycled, whilst the latter on the whole undergo little degradation and dispersion by natural processes. This is an important point because wastes are inherent by-products of agricultural production {Figure 20.1). This is not necessarily the case with industrial processes and their resultant wastes.

N a t u r a l r e s o u r c e s A i r W a t e r S o i l S u n

M a n - n n a d e r e s o u r c e s

_ _ B i o n n a s s _ S o i l S u n

M a n - n n a d e r e s o u r c e s c h e m i c a l s

L a b o u r T e c h n o l o g y C a p i t a l

A g r i c u l t u r a l p r o d u c t i o n o p e r a t i o n s

F o o d W o o d F i b r e W a s t e s

A g r i c u l t u r a l p r o d u c t i o n o p e r a t i o n s

F o o d W o o d F i b r e W a s t e s

Figure 20.1 Inputs and outputs in agricultural production systems (Taiganides, 1979)

There are many forms of waste in agriculture, including the excessive use of inputs (such as agricultural chemicals), the acceptance of sub-optimal levels of production (as by neglecting disease-control measures with animals) and the inefficient use of labour and capital resources (Wilson, Brigstocke and Wilhams, 1980). However, the most obvious waste is the under-use of waste resources of potential value. In this sense, waste is defined as a material where the costs of utilizing it are greater than the returns from utilizing it (O'Callaghan, 1975). Thus, in many countries, disposal of wastes of all types is rapidly becoming a major national problem (Wilson and Brigstocke, 1977a) Indeed it is a measure of the concern that an international scientific journal was launched in 1979 {Agricultural Wastes) with the specific aim of concerning itself with all aspects of agricultural waste management.

However, the idea should be rejected that there is some sort of sociopolitical imperative to use these waste materials. If their use is not

291

292 Economic considerations

The scale of the problem

The Joint Consultative Organization (1976) Working Party on Farm Wastes reported that the subject of farm wastes should be regarded as having major national importance for two reasons: (1) the environmental need to control pollution and the consequent

importance of providing substantiating information on which legislation to control such hazards can be based; and

(2) the economic need to make the best use of potentially valuable materials.

A large number of conferences and symposia have been held to quantify the extent of wastes and by-products used in animal feeding {see British Society of Animal Production, 1980) and it is clear there are many raw materials which are currently discarded or under-utilized and yet which, after a certain amount of processing, can be upgraded to become acceptable feedstuffs for livestock, including pets or man.

Recycling as animal or pet-feed is basically a matter of economics; technology is rarely a limiting factor. Recycling for human food is more complex and introduces new criteria such as consumer acceptability. Table 20.1 shows the range and scale of these potential waste resources on UK farms, which varied from 187 million tonnes of animal manure per year, through 9 million tonnes of cereal straw, 6 million tonnes of sugar beet tops and 2.3 million tonnes of pea-vining wastes down to 30000 tonnes of lettuce trimmings.

Table 20.1 POTENTIAL S O U R C E S OF WASTE ON UK F A R M S ( E N V I R O N M E N T A L R E S O U R C E S LTD, 1976)

Source tonne x 10^ per annum

Animal manure 187 cattle 163 pigs 17 poultry 7

Straw 9 Sugar beet tops 6 Pea-vining residues 2.3 Sprout residues 0.7 Other vegetable wastes L6

Worldwide the magnitude of food waste lost is immense. About half the world food production is lost or wasted (Pimentel and Pimentel, 1983). Pre-harvest losses due to pests, primarily insects, pathogens and weeds may be as high as 35% (Pimentel, 1976). After harvest, an additional 10-20% of food is lost to fungi, bacteria, insects and rodents. Undoubtedly

economic, then waste utihzation must involve the misuse or economic abuse of some other resource within the total production system (Raymond, 1978).

P.N. Wilson and T.D.A. Brigstocke 293

A N D J O H N S O N , 1979)

Constituent % by weight

Paper 33 Vegetable and putrescible 20 Metal 7 Glass 11 Textile 8 Plastic 5 Screening under 2 cm 6 Unclassified 10 Total 100

materials can be separated, and some recycled or utilized (for example as low-grade fuels), the major disposal route (76%) is untreated disposal in land infilling, which on present economic indications will continue for many years (Barber, 1982). Edible domestic waste has a nutritional potential equivalent to a conventional concentrate diet when fed to sheep (Hasdai and Ben-Ghedaha, 1982). It is only the logistics of the problem that preclude its use in this way.

Recycling of farm wastes for animal feed

It is the contention of the authors {see Wilson and Brigstocke, 1977a; 1980; 1981) that livestock production, due to its relative biological inefficiency compared with crop production, will have to rely more on diets based on

more effective storage and post-harvest control could reduce these losses considerably. In addition, in the USA, some 14% of the edible food that reaches the table is wasted and is discarded as domestic refuse (Harrison, Rathje and Hughes, 1975). Tolan and Singer (1983) have estimated that food wasted in the UK from the first stages of processing to the point of consumption could be up to 25% of the total food supply on an energy basis, while Wenlock and Buss (1977) have shown that domestic waste of food in the home is about 6% of the amount of dry matter eaten.

These enormous wastages indicate that there is considerable scope for recovery and recycling of potentially valuable resources. The system is dynamic. Effective R & D can transform yesterday's waste resource into a new productive asset. Furthermore the economic environment is not static. For instance, changes in the relative costs of feedingstuffs mean that farmers will now use large quantities of straw during the winter which before the mid 1970s would have been burnt in the fields. High cereal prices mean that more non-tariff materials, including by-products, are now being included in least-cost rations for animal feeding (Raymond, 1980).

It must be clearly understood that only a fraction of the waste generated can be economically utilized. A striking example of this is the 18.6 milhon tonnes of domestic sohd wastes which are produced each year in the UK, consisting of a variety of waste products which are generahy rich in degradable organic materials {Table 20.2). Although many of these waste

Table 20.2 T Y P I C A L C O M P O S I T I O N O F S O L I D D O M E S T I C W A S T E S I N T H E U K ( F R O M B A R B E R , M A R I S


Crop 1974 Food gap (Ό00 tonnes)

1980 1985

Wheat 31235 40557 57724 Coarse grain 312 4553 9215 Rice 1995 4596 8013 Total 33543 49707 74953

In such cases it is debatable whether economic considerations are suitable criteria for evaluating recycling of farm waste. A major national or international social subsidy may be needed, similar to that implicit in a national health service. What is required is a systems-synthesis approach based on a simple but complete model incorporating both the utilization of currently under-utilized or unused materials and a demonstratable increase in the production of animals and their products (Wilson and Brigstocke, 1980).

This chapter will now deal with the various types of farm waste used for animal feed in turn.

A N I M A L E X C R E T A

Many by-products and wastes are produced under less controlled conditions than primary products and may present unexpected chemical and bacterial hazards, such as drug residues. This could have particular relevance to the future role of treated animal wastes as feed for livestock. The US Department of Health, Education and Welfare (1977) set out their attitudes to Recycled Animal Waste and have adopted a 'wait and see' attitude. They state that the current level of R & D on the subject, although indicating good animal performance on recycled excreta, does not remove all possible hazards. Taylor and Geyer (1979) have noted that the US Food and Drug Administration does not sanction the use of poultry manure or litter as a feedstuff for animals because these materials could contain either drugs and antibiotics or their metabohtes, or disease organisms which could be transmitted in feed to other animals and hence eventually to man. It is probable that the EEC, and hence the UK, will tend to follow this American stance and adopt a similar cautious approach.

In addition to the legal constraints, the inclusion of some of these materials in feeds could prove a very contentious social issue and might well lead to adverse consumer reaction on the grounds of unwholesome-

forage and by-products rather than on large quantities of cereals which are suitable for direct human consumption. This is particularly relevant to developing countries where large sectors of the population do not receive their minimal energy requirements (i.e. they are starving to varying degrees). World Bank figures, quoted by Coombs and Parker (1979), estimate a growing food-grain gap reaching around 75 million tonnes in developing countries by 1985 {Table 20.3).

Table 20.3 PROJECTED F O O D G A P IN T H E D E V E L O P I N G C O U N T R I E S IN 1980 A N D 1985 (FROM C O O M B S A N D P A R K E R , 1979)

p. iV. Wilson andT.D.A. Brigstocke 295

V E G E T A B L E WASTES

The main problem with vegetable wastes either unprocessed or processed is whether these materials will continue to be under-utilized, not because of insufficiency of nutritional information but because the securing of the main crop will take even greater priority over secondary operations as labour and machinery become relatively more expensive (Raymond, 1978; Wilson and Brigstocke, 1977a). In addition many of these wastes are produced on units far removed from the hvestock which might consume them. Consequently if there is to be greater utilization of vegetable wastes then a change of direction would be needed away from specialization to more self-sufficient production systems where the by-products of one process become the raw materials for others. A move to integrated systems has much to commend it in biological terms but it would cause fundamental changes in the way rural development has occurred in developed countries over past decades.

Considerable efforts have been made by the animal feed compound trade to find alternative raw materials to human food crops for incorporation in compound diets. Whilst these efforts have not been entirely unrewarding, as illustrated by the success of upgrading of cereal straw, there is no immediate prospect of using large quantities of cheap byproducts which could be directly incorporated into animal feeds as substitutes for traditional raw materials (Wilson, Brigstocke and Cuthbert, 1981). In many cases, the cost of preparing the by-product in a form suitable for transport over long distances, and the additional cost of de-toxification and/or up-grading the material to release its potential

ness of the resultant food end-product (Wilson and Brigstocke 1977a). It might also cause the major wholesalers/retailers to stipulate 'zero excreta inclusion' clauses in their buying contracts with livestock producers. The point is generally missed that large amounts of excreta are invariably ingested by farm animals under normal conditions.

Once the UK Protein Processing Order, which is designed to control the processing of all animal proteins intended for animal feed (e.g. offal, meat and bone, feathers, blood and manure), becomes fully effective then UK government legislation might encourage the use of properly processed poultry wastes to ruminants. This would certainly lower the cost of animal feed and hence of human food. Additionally any such removal of manure from poultry farms would meet with approval from the environmentalists by getting rid of a potential problem and turning a major 'pollutant' into a valuable extra source of certain animal nutrients (Wilson, Brigstocke and Wilhams, 1980).

Some time has been spent on this topic as it illustrates the dilemma which affects the opportunities for recycling certain contentious waste products. The economic considerations of the operation may be favourable but it is hkely that subjective, sociological arguments will carry more weight than objective scientific principles. It is probable that society and not science or economics will determine the extent and use of recycled animal wastes.


nutritive value, have been found to exceed its economic value relative to conventional raw materials (Robb, 1976). There may be more opportunity for upgrading wastes on-farm, where high technology and expensive capital equipment is not involved. The quality of the end-product may be less but the costs will be lower.

The upgrading of cereal straw by chemical treatment provides a good example of the problem of quantifying the cost:benefit ratio. If the sum is calculated in terms of the energy contained in the untreated straw (5800 MJ/tonne) compared to the energy value of the treated product (9000 MJ/tonne) then the energetic benefit is 9000-5800 MJ/tonne = 3200 MJ/tonne (dry matter basis). However, the energy involved in the chemical and physical process of upgrading amounts to 5085 MJ/tonne, so for every tonne of straw upgraded there is an energy deficit of 5085 - 3200 = 1885 MJ/tonne. On this basis the process is not energetically efficient. However if the straw is not so treated it is likely to be either burnt, ploughed in or used for bedding, in which case the feeding value of the energy contained in the long straw is zero. If the injection of 5085 MJ/tonne in the treatment process enables 9000 MJ/tonne to be recovered in the treated straw, then the energetic balance is favourable at 9000 - 5085 = 3915 MJ/tonne.

In other words, 'energetically' this process would appear at first sight to be inefficient, but this comparison is confounded by the difference between the physical and nutritional energy parameters which, although quoted in the same units (in MJ), are clearly different. Perhaps a more reahstic method of describing the situation is to state that, although the bahng + storing + processing energy costs of straw treatment are 5085 MJ/tonne, a material which was previously wasted (by burning without any use being made of the heat of combustion released thereby) has been converted into a useful ruminant feed valued at 9000 MJ/tonne of ME (on a dry matter basis).

This argument would appear valid since the technique of chemical processing will not alter the existing pattern of use for untreated long cereal straw. It merely utilizes some of the 9 million tonnes which are otherwise burnt in the fields and wasted.

In economic, as distinct from energetic, terms the valid comparison is the cost of untreated straw plus the cost of treatment compared with other types of energy feed. In practice the main competing feeds would be other medium energy fibrous feeds, such as hay and sugar beet pulp.

Coursey (1972), in his review on the losses inherent in tropical horticultural production, calculated that some 30 million tonnes of vegetable food per year may be lost. If this estimate is reliable then, as Preston and Wilhs (1970) postulated, there is a good case for siting intensive livestock units near to centralized vegetable processing plants. There is therefore much to commend the system proposed by Preston (1975) where tropical animal production systems could be developed based upon the use of available by-products to balance forage shortages during critical periods.

G R E E N CROP F R A C T I O N A T I O N

Another indirect form of waste is the sub-optimal use made by ruminants of grass and legumes. Most grasses have a crude protein content in the dry

p.Ν. Wilson and T.D.A. Brigstocke 297

Utilization of wastes by the compound feed industry

Some industrial wastes, such as cereal by-products, are already extensively used in cattle and sow rations at around 30% of the total formulation. These by-products consist of wheatfeed, other mihing offals, rice bran, dried grains, grain screenings, etc. Barber and Lonsdale (1980) noted that from five industries in the UK—brewing, distilling, milling, sugar extraction, potato processing—at least 2.7 x 10^ MJ of ME and 4 x 10 tonnes of crude protein are available annuaUy to livestock farming as by-products. Other potentially useful raw materials, such as slaughterhouse wastes, are at present not being used to any extent. The need for more economical feeding, and the risk of pollution problems if some current methods of residue disposal continue, are likely to combine to encourage the greater use of a number of such resources. It may well be that, as Wiseman and Cole (1983) noted, this type of animal waste was merely regarded as troublesome until its potential nutritive value was established.

Since the great majority of compound feed manufacturers use least-cost formulation, changes in ingredient prices have a pronounced effect on a composition of compound feeds. The demand for any raw material by the animal feed manufacturer is determined by four factors:

(1) the price of the ingredient in relation to its nutritional specification; (2) the price of acceptable substitutes for the ingredient;

matter varying between 17 and 26, yet growing cattle only require 15% CP in the total diet, older beef animals 12% and dairy cows about 14% (Jones, 1976).

By expressing the juice from leafy materials, and then coagulating and drying the protein contained in the juice, it is possible to produce two useful feeds instead of one. The first is a dry, leaf protein concentrate which can be used as a high protein material for inclusion in pig, poultry or calf diets. The second is the residual, fibrous material of moderate protein content which can either be fed fresh to ruminants or ensiled for winter feeding at a later date (Connell and Houseman, 1976).

The chief economic difficulty with leaf protein extraction, or the less complex process of grass drying, is that the factory can only be used for four to five months each year. This is very expensive in overheads. The possibility of combining the chemically treated straw industrial process with other operations, such as grass drying or leaf fractionation, has therefore much to commend it with fixed capital being shared and the plant utilized throughout the year (Wilson and Brigstocke, 1977b). From a practical point of view, this type of shared operation would appear attractive. However, the experience of BOCM Silcock in running such a shared pilot plant is that the extremely high running costs, particularly power and fuel for drying, result in a final sale price of the two green crop feeds becoming prohibitively expensive. For this reason a promising technological development was abandoned on economic grounds. The BOCM Silcock plants are now used solely for the production of chemically treated straw.


(3) the upper and lower inclusion limits of the ingredient fixed by the compounder;

(4) other considerations, such as handling and processing characteristics (Crabtree, 1972; Wilson, 1980).

It therefore follows that any raw material has two values. One is its 'intrinsic value' dependent on its chemical composition, nutritional value, etc., whilst the other is its 'market value', which is determined by the competitiveness with other raw materials at a given time in a given formulation. The 'true market value' of a material, such as chemically treated straw, is dependent upon competition vis a vis all other raw materials competing for 'space' in the formula or diet (Wilson, Brigstocke and Cuthbert, 1981). Nevertheless in any stable market, a given material will always bear some relationship to the current price of a few 'major' alternatives (Wilson and Brigstocke, 1977a).

Recycling wastes for pet-foods

The wide range of products produced by the pet-food industry are also candidates for waste utilization. The vegetable raw materials used would be similar to those found in the animal feed market but the animal-based raw materials are largely those which the human food industry cannot or does not wish to use, such as stomachs, spleens, intestines and filleting waste from the fish industry (Ashley, 1983). The success and skill with which the pet-food industry produces and markets an organoleptically acceptable product to pet owners is an indication of the potential, much of which still remains to be exploited in other areas.

Recycling wastes for human food

Food from waste sounds a good idea, but the psychological problems of persuading people to actually eat such a commodity are tremendous. Any company must therefore consider the marketing problems involved, even if it can establish that food from waste is safe and wholesome. This problem is compounded by the usual technological, legislative and economic constraints. Nevertheless most major food companies now have substantial and continuing interests in minimizing their waste production and also in recovering and re-using such wastes as do exist.

There is therefore a need to apply 'quality' concepts to wastes. Everything else being equal, a waste of high quality will be utilized more effectively than one of low quality. This may well have marked consequences with respect to the efficiency and economics of a particular utilization procedure (Pearce, 1979). This is an extremely important general point concerning the upgrading of wastes. If they are only improved to low-or medium-grade materials, then their use is likely to be limited. It is only if the upgrading process is sufficiently successful to transform wastes into high-grade products that it is likely to be economically viable.

The chapter will now deal with the main types of waste used for human food.


A N I M A L WASTES

The recovery and utihzation of waste meat is becoming increasingly attractive in economic terms. Newman (1983) has estimated that over 2 million tonnes of red meat adhering to the bone is at present wasted, with an estimated monetary value in excess of £9 million per annum. Indeed the recovery of meat from bones is now in extensive commercial use. The most intractable problem has been changing the attitudes of the trade, the enforcement authorities and the public, by convincing them that correctly processed meat from bones is a useful nutritional food and not a waste material. The economics of the mechanical recovery of meat are very attractive as the cost of meat in real monetary terms is consistently rising.

There have also been considerable advances in the conversion of bone to edible products (Jobling, 1983). Jobhng notes two important points:

(1) the key to success is the regrading of the bone as a raw material in its own right and not a waste product;

(2) the economics of the process are highly favourable. This enables the conversion of 1 tonne of raw bones (worth currently (June, 1982) £60/tonne as a waste product for rendering into bone meal and tallow) into food ingredients (worth £350-£450/tonne).

The utilization of animal by-products is one possible method by which the meat industry can improve their very poor operating margins. The adage 'we use every bit of the pig except the squeak' is not a correct comment on current practice for, although most by-products are not thrown away, they are not turned into products of high added value.

One of the most interesting developments in the field of the meat product manufacturer is the technique of product reforming (Ashby, 1981). For example, low quahty meat can be converted into products more similar to top-grade cuts by breaking down the meat into its components of muscle, fat and gristle and then recombining the muscle and fat. The main objectives are greater cost effectiveness and greater palatability, but the technology also offers the potential for controlling the nutritional characteristics of the resuhant products.

Another interesting example is the use of hide splits, or other collagenous slaughter-house wastes, to make sausage casings (Davidson, 1972). The traditional material is the small intestine of the pig which has obvious disadvantages. It has to be thoroughly cleaned and is of varying length and diameter. It is discontinuous and cannot be used in a continuous extrusion process. Clearly a physically and chemically homogeneous collagen sheath of unhmited length has great production advantages, and emotively it is perceived to be more 'wholesome' than the pig intestine which it replaces.

The technology of meat recovery and the better utilization of animal by-products is already well established. The problem remains that consumers and legislators have many reservations about the use of waste materials, however well processed, in meat products and there are corresponding legislative, social and even religious barriers to this practice. The economics are highly attractive; the acceptance by the consumer may not be so favourable.


V E G E T A B L E WASTES

Singer (1980) has argued that the outstanding areas for greater utihzation of waste in the food industry are in the sectors of sugar utihzation, vegetable processing and distribution and catering, with recycling of these waste materials for feeding direct to humans. The problem, as Tolan (1983) has noted, is that at one extreme food is wasted if it is edible but not eaten whilst at the other it is wasted only if there would have been an economical return to be gained by using it for food. Between these two extremes, there are many social, technical, nutritional and economic factors to be considered.

There are a number of interesting technological developments currently being evaluated, such as the use of enzymes for the modification of food waste into products of higher added value (Fullbrook, 1983). Bellamy (1983) believes that genetic engineering and recombinant DNA will increasingly contribute to waste processing in the future. However, single cell protein production from hydrocarbons in Japan is a good example of technical and economic success but social, cultural (and hence marketing) failure. This is an important point for although in the animal feed industry low level inclusions of by-products can be achieved without farmer reaction, with the food industry more progress with the use of microbiological agents is hampered, not through lack of technological expertize but because of lack of social acceptabihty.

Although a great deal of research, money and effort have been devoted to textured vegetable protein and single cell protein, they have not yet had the market success that was expected and it is perhaps worth dwelhng on why this has occurred. The Food Standards Committee, in its Report on 'Novel Protein Foods', has made the point that since the average UK diet is not deficient in protein, any significant use of alternative sources of protein is likely to be on the basis of comparative costs. In other words, consumers are not seeking to increase their daily protein intake—in fact they have reduced it in the last two decades. Thus any new protein food can only gain market entry by displacing some other food. Apart from cost, the only other reason why one form of protein might displace another will be because of any specially desirable properties it may possess. However, as already seen, such new feeds can be readily acceptable into the animal or pet-food markets.

It is the lower cost of vegetable proteins and their technical attributes that have enabled them to gain an entry into certain sectors of the food industry. Thus vegetable protein has been used as a meat extender in the catering sector, while in food manufacturing vegetable proteins are now being used instead of the more traditional caseinates in sausages. However, the most sophisticated meat analogues have not been an unqualified commercial success since their cost/quahty ratio has not yet met with the fuU approval of consumers.

The use of previously wasted by-products containing valuable proteins has considerable merit, particularly in developing countries where milk and other high-quality protein sources may be scarce. Milk analogues prepared with plant protein, such as soybean, have much to commend them and indeed 'soymilks' are commerciahy successful in Hong Kong and


Effluent disposal

Many valuable raw materials are being discharged into the sewage system at nominal cost. In the past decade Water Authorities, due largely to the Control of Pollution Act in the UK, have accepted responsibility for the supervision and supply of water and disposal of sewage into rivers and streams. Disposal of waste into the sea is still relatively uncontrolled and a genuine cause for concern in many so-called developed countries. The financial penalties for unlawful discharge into water courses have resulted in new endeavours by the food industry to recover potentially valuable raw materials. Unfortunately in many situations, although the recovery of these materials may be financially desirable, it is not yet an economic proposition. As Lyon (1978) has noted in discussing a particular problem with confectionery effluent, systems such as evaporation, ultrafiltration, reverse osmosis and fermentation were all ruled out on the basis of either excessive operating or capital costs. It was cheaper to throw away potentially valuable raw materials rather than to recover them with the technology that was available.

A similar problem exists with wastewater from the meat, poultry and fish processing industries. The quantity of protein lost in the UK by these mdustries has been estimated at about 25000 tonnes per annum (Hop-wood, 1978). It all ot this material could be recovered as a meal containing 50% protein, it would have a current annual value to the animal feed industry of £7.5 milhon. The recovery of protein from wastewater as a usable commodity requires the satisfaction of a number of criteria (Hop-wood, 1978). Many of these points are relevant in all areas of waste management.

(1) Any process to recover materials from effluents must be designed to operate on a substrate of constantly varying composition containing many substances in dilute suspension and solution.

(2) The process should be capable of operating in a normal food factory and not require highly specialized skills.

(3) The chemicals used as precipitants must not render the recovered material unusable.

(4) The finished product must be readily salable and not require further expensive processing.

(5) When viewed in the long term the process must be more economical than conventional methods of waste disposal.

Brazil where they compete with soft drinks (Holsinger, 1983). Again it is a problem of consumer acceptabihty rather than one of economics. It is however worth noting that a milk analogue developed by the USDA in 1974 for use as a substitute for non-fat dry milk in developing countries failed because of a continuing world glut of milk powder. Various new techniques for whey fractionation, such as ultrafiltration and electrodialysis, have been developed for cheese whey disposal. However utilization remains a great problem with at least 50% still being wasted (Holsinger, 1983).


Year GNP (£ million)

Total public expenditure on water, sewerage and sewage disposal and refuse disposal (£ million) % of GNP

1950 11747 71 0.6 1960 22880 169 0.7 1970 44045 484 1.1

Table 20.5 UK PUBLIC E X P E N D I T U R E ON E N V I R O N M E N T A L C O N T R O L (1974) (FROM I S A A C , 1978)

Capital Operating (£ million) (£ million)

Water supply 145 326 Sewerage and sewage disposal 346 454 Refuse disposal 40 312

Total 531 1623 1092

GNP: £94095 million; % GNP: 1.7

The costs of waste disposal by traditional means are very high and continue to rise, both in absolute terms and also when expressed as a percentage of GNP (see Tables 20.4 and 20.5). Thus in 1975 the capital and operating costs were in excess of £1.6 billion, representing 1.7% of GNP. The incentive to reduce the cost by attracting revenue from recovered products is therefore very great.

Conclusion

A large number of processes described in this Easter School and at similar gatherings (see Process Biochemistry, 1979) are extremely ingenious and scientifically sound but often no costibenefit calculations have been presented. In the majority of cases economics will govern whether or not new technological developments will be exploited. This is especially so in the case of the animal feed industry but less so with the food industry where emotive considerations influencing consumer acceptability may predominate. Furthermore, legal constraints are hkely to be more strict where human food is concerned compared to the laws governing animal and pet-food manufacture.

Due to economic pressures individual farms will become more intensive with commensurately larger quantities of farm waste being produced. At the same time the costs of transportation and/or drying these wastes will increase. On the other hand agricultural pollution is becoming increasingly unacceptable and legal constraints will force farmers to dispose of their wastes more effectively in the future than in the past {see Guiver, 1982).

It is interesting to note that the most integrated agricultural waste management techniques are to be found in developing countries such as Thailand where the disposal of wastes is used in such a way and at such a

Table 20.4 υκ P U B L I C E X P E N D I T U R E O N E N V I R O N M E N T A L C O N T R O L ( F R O M I S A A C , 1978)


References

A S H B Y , A. (1981). Our Changing Diet—Lessons for Farmers. Paper presented at the Agricultural Education Association's Winter Conf., Imperial College, London. December 1981

A S H L E Y , N.V. (1983). In Upgrading Wastes for Feeds and Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie, p. 267. Butterworths, London

B A R B E R , C. (1982). Domestic waste and leachate. Notes on Water Research. No. 31. May 1982

rate that they are recycled into new resource materials by natural means. Indeed Black (1971) has pointed out that modern methods of agriculture, which may appear much more productive than primitive systems, are probably very similar in terms of the efficiency with which the total energy resources are used.

The situation past the farm gate either in the food or meat industries is considerably more developed. It is an open question whether this has occurred due to the awareness of the need for maximum recycling or because of the threat of excessive charges for illegal pollution of water courses or on landfill sites. Nevertheless the progress in recovery techniques has been impressive and in many situations the economics of the recovery process have been substantial.

The greatest problem facing the food producer in any new upgrading development is to gain the confidence and acceptance of the consumer. Upgrading a waste material without upgrading its image might even prove to be counter-productive by attracting a stigma of 'adulteration' (Hannan, 1983).

In spite of these constraints there will be a general trend towards greater utilization of waste resources, particularly for recycling as animal feed, either directly or after some form of processing treatment. The capital costs of these developments will be very great. If all the capital required has to come from the farmer producers or manufacturers then the pace of waste utilization will be slow. It is reassuring to note that the NRDC provided financial assistance to the various processes outlined by Jobling (1983) and Ross (1981). It is to be hoped that such assistance from governmental or parastatal bodies will continue to be forthcoming for if society at large, which is demanding the removal of these forms of pollution, accepts part or most of the cost on a national basis then the move towards recycling could take place very rapidly.

Another major problem is one of scale. It is usually uneconomic for farmers to process their own on-farm wastes on a small scale and factory-based industrial processes may be necessary. Unfortunately scal-ing-up pilot plants into full scale industrial processes can be fraught with difficulties and economies of scale may not always be forthcoming. At the end of the day in developed countries it is not so much what is socially or environmentally desirable but what is economically viable which will determine whether or not new process technologies will be put into practice, and whether or not they are farm or factory based.


B A R B E R , C , MARIS , P.J. and JOHNSON, R.G. (1979). Behaviour of wastes in landfill sites: study of the leaching of a selected industrial waste in large scale test cells. Edmonton, North London. WLR Tech. Note No. 69. Department of the Environment

B A R B E R , V^.P. and L O N S D A L E , C R . (1980). In By-products and Wastes in Animal Feeding. Ed. by E.R. 0rskov, p. 61. Occ. Publ. No. 3. British Society of Animal Production

B E L L A M Y , W . D . (1983). In Upgrading Waste for Feeds and Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie, p. 141. Butterworths, London

BLACK, J.N. (1971). Ann. Appl. Biol. 67, 272 BRITISH SOCIETY OF A N I M A L P R O D U C T I O N (1980). By-products and Wastes

in Animal Feeding. Occ. Publ. No. 3. British Society of Animal Production

C O N N E L L , J.C. and H O U S E M A N , R .A. (1976). In Green Crop Fractionation. Ed. by R.J. Wilkins, p. 57. Occ. Symp. No. 9. BGS/BSAP

C O O M B S , J. and P A R K E R , K.J. (1979). Proc ISES Conf Biomass-Future Development p.69

C O U R S E Y , D . G . (1972). In Post-harvest Deterioration of Raw Natural Products. Ed by A.H. Walters, E.H. Hueck-van der Pias, Vol. 2, p. 464. Applied Science Publ., London

C R A B T R E E , J.R. (1972). The composition of compound feedstuffs under UK and EEC conditions. Tech. Rep. No. 11. Grassl. Res. Inst. Hurley

D A V I D S O N , G. (1972). Collagen Casings. The British Fresh Sausage-Present Day Manufacturing Techniques. BFMIRA Symposium Proc. No. 11. March, 1972. London

E N V I R O N M E N T A L R E S O U R C E S LTD (1976). Study for the Commission of the EEC (unpubhshed)

F U L L B R O O K , P .D . (1983). In Upgrading Waste for Feeds and Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie. p. 133. Butterworths, London

G U I V E R , K. (1982). Publ. Hlth Engin., 10(1), 23 H A N N A N , R.S. (1983). In Upgrading Waste for Feeds and Food. Ed. by

D.A. Ledward, A.J. Taylor and R.A. Lawrie, p.197. Butterworths, London

H A R R I S O N , G.G. , R A T H J E , W.L. and H U G H E S , W W . (1975). J. Nutr. Educ., 7, 13

H A S D A I , A . and B E N - G H E D A L I A , D . W(1982). / . Dairy. Sci., 65, 65 H O L S I N G E R , V.H. (1983). In Upgrading Waste for Feeds and Food. Ed. by

D.A. Ledward, A.J. Taylor and R.A. Lawrie. p.211. Butterworths, London

H O P W O O D , A.P . (1978). Recovery of protein from food industry effluents. Paper presented at a Food for Thought conference organized by the National Industrial Materials Recovery Association, November 1978

I S A A C , P.C.G. (1978). Chem. Ind., 15 July 1978, p. 497 JOBLING, A . and JOBLING, C.A. (1983). In Upgrading Waste for Feeds and

Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie, p.183. Butterworths, London

JOINT C O N S U L T A T I V E O R G A N I Z A T I O N FOR R E S E A R C H A N D D E V E L O P M E N T IN A G R I C U L T U R E A N D F O O D (JCO) (1976). Report of the Farm Waste Working Party


JONES, A . S . (1976). In Green Crop Fractionation. Ed. by R.J. Wilkins, p . l . Occ. Symp. No. 9. BGS/BSAP

L Y O N , J . C M . (1978). Economics of Effluent Recovery. Paper presented at a Food for Thought conference organized by the National Industrial Materials Recovery Association, November 1978

N E W M A N , P.B. (1983). In Upgrading Waste for Feeds and Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie. p. 93. Butterworths, London

O ' C A L L A G H A N , J.R. (1975). J. Roy. Soc. Arts, 123, 139 P E A R C E , G.R. (1979). Agric. Wastes, 1, 223 PIMENTEL, D . (1976). Bull. Entomol. Soc. Am., 22, 20 PIMENTEL, D . and PIMENTEL, M. (1983). In Upgrading Waste for Feeds and

Food. Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie. p.3. Butterworths, London

P R E S T O N , T.R. (1975). In Proc. Seminar on Potential to Increase Beef Production in Tropical America, p. 149. Centro Internacional de Agri-cuhura Tropical, Colombia

P R E S T O N , T.R. and WILLIS, M.B. (1970). Intensive beef production. Pergamon Press, Oxford

PROCESS BIOCHEMISTRY (1979). Effluent Treatment in the Biochemical Industries. Process Biochemistry's Third International Conference, November 1979. Wheatland Journals, Watford

R A Y M O N D , W.F. (1978). Biologist, 24, 80 R A Y M O N D , W.F. (1980). In By-products and Wastes in Animal Feeding. Ed.

by E.R. 0rskov, p.3. Occ. Publ. No. 3. British Society of Animal Production

R O B B , J. (1976). In Feed Energy Sources for Livestock. Ed. by Η. Swan and D. Lewis, p. 13. Butterworths, London

ROSS, A . (1981). Food Manuf. October 1981, p. 61 SINGER, D . D . (1980). In By-Products and Wastes in Animal Feeding. Ed. by

E.R. 0rskov, p. 71. Occ. Publ. No. 3. British Society of Animal Production

T A I G A N I D E S , E.P. (1979). Agric. Wastes, 1, (1), 1 T A Y L O R , J.C. and G E Y E R , R .E . (1979). / . Anim. Sci., 48, 218 T O L A N , A . (1983). In Upgrading Waste for Feeds and Food. Ed. by D.A.

Ledward, A.J. Taylor and R.A. Lawrie. p. 15. Butterworths, London U S D E P A R T M E N T OF H E A L T H , E D U C A T I O N A N D W E L F A R E . F O O D A N D

D R U G A D M I N I S T R A T I O N (1977). Fed. Reg., 42, 248 WENLOCK, R.W. and BUSS, D . H . (1977). / . Human Nutr., 31, 405 WILSON, P.N. (1980). In Feeding Strategy for the High Yielding Dairy Cow.

Ed. by W.H. Broster and H. Swan, p. 374. EAAP Publ. No. 25. Granada, London

WILSON, P.N. and BRIGSTOCKE, T. (1977a). Agric. Prog., 52, 49 WILSON, P.N. and BRIGSTOCKE, T. (1977b). Process Biochem., 12(7), 17 WILSON, P.N. and BRIGSTOCKE, T . D . A . (1980). In Vegetable Productivity. Ed.

by C.R.W. Spedding, p.50. Symp. Inst. Biol. No. 25. MacMillan Publishers Ltd, London

WILSON, P.N. and BRIGSTOCKE, T . D . A (1981). In Biological Husbandry: A scientific approach to organic farming. Ed. by B. Stonehouse, p. 301 Butterworths, London


WILSON, P .N. , BRIGSTOCKE, T . D . A . and WILLIAMS, D .R. (1980). Process Biochem,, 50 (7), 36, 48

WILSON, P .N. , BRIGSTOCKE, T . D . A . and C U T H B E R T , N.H. (1981). Anim, Feed Sei, Technol,, 6, 1

W I S E M A N , J. and COLE, D .J .A . (1983). In Upgrading of Waste for Feeds and Food, Ed. by D.A. Ledward, A.J. Taylor and R.A. Lawrie. p. 233. Butterworths, London

LIST OF PARTICIPANTS

Areas, Mr J. Universidade de Sao Paulo, Brazil.

Ashley, Dr N.V. Pedigree Petfoods, Melton Mowbray, Leics.

Autio , Dr K. Technial Research Centre of Finland, Espoo, Finland.

Bellamy Professor W . D . Dept. Food Science, Cornell University, U S A .

Bender, Professor A . E . Queen Elizabeth College, University of London.

Berrington, Mr D . Dept. Applied Biochemistry & Food Science, University of Nottingham.

Blanchard, Mr J.M.V. Dept. Applied Biochemistry & Food Science, University of Nottingham.

Brady, Miss A . M . Dept. Applied Biochemistry & Food Science, University of Nottingham.

Brigstocke, M r T . BOCM, Basingstoke, Hants.

Butters, Miss L. Dept. AppHed Biochemistry & Food Science, University of Nottingham.

Cole, Dr D.J .A. Dept. Agriculture & Horticulture, University of Nottingham.

Cooper, Dr R.N. Meat Industry Research Institute, New Zealand.

Craven, Mr M.R. Dept. Applied Biochemistry & Food Science, University of Nottingham.

Crispin, Mr D.J. Brooke Bond Group pic, Croydon.

Crittenden, Mr M. Henry Telfer Ltd, Northampton.

Curtis, Professor R.F. Food Research Institute, Norwich.

Cuthbertson, Sir D . Dept. CHnical Biochemistry, The Royal Infirmary, Glasgow.

Cuthbertson, Dr W.J.F. 4 Coppermill Lane, Harefield, Middlesex.

Cyril, Mr H.W. Dept. Applied Biochemistry & Food Science, University of Nottingham.

Davie , Mr J. Simon-Rosedowns Ltd, Cannon Street, Hull.

Deeley , M s S . M . Butterworths (Pubhshers), Sevenoaks, Kent.

D o d o o , M r E . K . Dept. Veterinary Services, Accra, Ghana.

Downie , Mr H. Alginate Industries Ltd, Tadworth, Surrey.

D u r b y e , D r H . Effem GmbH, Verden, Germany.

307

308 List of participants

Edwards, Dr C A .

E g a n , D r H .

Elliott, Miss K.L.

Fevrier, Dr C.

Foscro f t ,MrP.D.

F u l l b r o o k , D r P . D .

Hall, Mr G.

Halsa l l ,MrI .

Hannan, Dr R.S.

Hansen, Mr P.I .E.

Hansen, Dr P.M.T.

Hector, Mr D . A .

Higgins, Dr J.G.

Holsinger, D r V . H .

Hopwood, Mr A . P .

Horrocks, Mr J.K.

Howson, Mr S.J.

Hughes, D r R . B .

Imeson, Mr A .P .

Jacobsberg, Dr F.R.

Jinap-Selamat, Mrs

Jobling, Dr A .

Jobling, Mr C.A.

King, Miss K.

K l y h n , M r F . A .

Kooi, Mr Eng-Teong

Kunstelj, Mr J.

Ladegaard, Mr T.

Lakin, Dr A . L .

Lawrie, Professor R .A .

Ledward, Dr D . A .

Lindner, Dr P.

MacFadyen, Miss L.G.

Mackie, Dr I.

McAuley, MrB.J .

Rothamstead Experimental Station, Harpenden.

Laboratory of the Government Chemist, London.

Dept. Chemical Engineering, University of Nottingham.

I N R A , Saint-Cilles, L'Hermitage, France.

Prosper de Mulder Ltd, Doncaster.

National College of Food Technology, Weybridge,

Dept. Applied Biochemistry & Food Science, University of Nottingham.

Shell Centre, London.

Meat & Livestock Commission, Milton Keynes.

Danish Meat Research Institute, Roskilde, Denmark.

Dept. Food Science & Nutrition, Ohio State University, U S A .

Tropical Products Institute, London.

Dornay Foods, Kings Lynn, Norfolk.

U S Dept. of Agriculture, Philadelphia, U S A .

Alwatech UK Ltd, High Wycombe, Bucks.

Walkers Crisps, Leicester.

Food Research Institute, Norwich.

C & T H a r r i s , C a l n e , Wilts.

Alginate Industries Ltd, Tadworth, Surrey.

Dalgety Spillers Ltd, Research & Technology Centre, Cambridge.

Food Science & Technology, Universiti Pertanian, Malaysia.

Lensfieid Products, Flitwick, Bedford.

Lensfieid Products, FUtwick, Bedford.


Alwatech UK Ltd, High Wycombe, Bucks.

Kuala Lumpur, Malaysia.

Nordreco A B , Bjuv, Sweden.

Aminodan, Skaaden, Denmark.

Dept. Food Science, University of Reading.

Dept . Applied Biochemistry & Food Science, University of Nottingham.


Division of Food Technology, Volcani Center, Israel.

Dalgety-Spillers Ltd, Research & Technology Centre, Cambridge.

Torry Research Station, Aberdeen.

Purina Protein, Application Laboratory, Baldock.

List of participants 309

Milbourne, Miss K.

Mitchell, Dr J.R.

Murphy, Mr D .

Newman, Dr P.B.

Nielsen, Mrs J.

Norton, Dr G.

0 'Conne l l ,MrP.P .

Oliphant, Mr G.

Pass, Mr R.T.

Phipps, M r T .

Pimentel, Professor D .

Raa, Professor J.

Sawyer, Mr R.

Smith, Mr J.

Tableros, Mr M.A.

Taylor, Dr A.J.

Tolan, Dr A .

Trugo, Mr L.G.

Tucker, D r G . A .

Tuerena, Mr C.E.

Tuominen, Mr R.

Verma, Mr M.M.

Van der Wal, Dr P.

Walker, Mr G.J.

Walters, D r K .

Wanke, Mr W.

Whittle, Dr K.J.

Wignal l ,MrJ.

Wilson, Professor, P.N.

Wiseman, Dr J.

Wood, Mr J.R.

W o o l e y , M r I . J .



Food Research Institute, Norwich.

Meat Research Institute, Langford, Bristol.

Ministry of Fisheries, Lyngby, Denmark.


National Board for Science & Technology, Dublin, Eire.

Meat & Livestock Commission, Milton Keynes.

Pentlands Scotch Whisky Research Institute, Edinburgh.

Henry Telfers Ltd, Northampton.

College of Agricultural & Life Sciences, Cornell University, U S A .

Institute of Fishery, University of Troms0, Norway.

Laboratory of the Government Chemist, London.


Food Technology, University of Hohenheim, Germany.


Ministry of Agriculture Fisheries & Food, London.

Dept. Food Science, University of Reading.



Finnish Meat Research Centre, Hameenlinna, Finland.


ILOB, Wageningen, Netherlands.

Ministry of Agriculture Fisheries & Food, Tettenhall, Wolverhampton.

Dept. Chemical Engineering, University of Nottingham.

Deutsche Gelatin-Fabriken Stoess GmbH, Eberbach, Germany.

Torry Research Station, Aberdeen.

Humber Laboratory, Hull.

BOCM, Basingstoke.

Dept. Agriculture & Horticulture, University of Nottingham.

Sun Valley Poultry Ltd, Hereford.

Simon-Rosedowns Ltd, Hull.

INDEX

Abattoir waste, improved utilization for feedingstuff, 297,

299 Acceptability,

incorporation of animal excreta a difficulty, 295

increased by conversion of waste via animals, 280

limiting factor in waste utilization, 278, 284

of milk analogues, 301 of novel products, initial assessment, 287

Acetic anhydride, enhances functional properties of meat, 202

Acids, in fish silage, 121-123 Ac id-base reactions, 32 Actinomycetes, potential for waste

conversion, 148 Activated sludge, treatment of effluent, 31,

65 Active surface layer of membranes, 54 Additions, permitted in pet-foods, 272 Anatoxin,

in mouldy feedingstuff, 237 required absence from groundnut flour,

226 toxic contaminant, 285

Agaricus bisporus (common mushroom), 146 Agglomeration, of proteins, 32, 34 Agricultural wastes, types of, 291 Alcoholic beverages, from cheese whey, 212 Alginate gels,

in canned foods, 251-255 rigidity affected by calcium ions, 250

Alkah, and improved digestibility of lignified

plant tissues, 240 inspectorate, 81 mild, and de-esterifícation of plant

pectins, 247 Allergy to fungi, limits use as food, 148 Amino acids,

available, 234, 243

Amino acids {cont.) differences in requirements for cats and

dogs, 272 in animal feeds, 235 in peanut protein, 226 in single cell protein, 273, 282 in soybean protein, 211 in whey protein, 214 limiting in feedingstuff, 234 limiting in soybean protein, 220 stability in flsh silage, 124

Aminodan, 41 Ammonia adsorpflon, 85-91 Analogue techniques, developed by pet-food

industry, 272 Anionic polysaccharide, 38, 85 Animal,

fats, 73 feed, 17, 23 health inspection service, 109

Anflbiotics, in meat, 103 Antioxidants, in mechanically recovered

meat, 105, 107 Apex meal, 4 5 - 4 8 Appeflte, animal, importance of in feedstuff

allowance, 235 Aseptic filling, of soy milk proteins, 219 Astaxanthin, from shrimp silage, 128 Avicellulase, production by

micro-organisms, 149-150

Bacterial see Micro-organisms Basket centrifuges, in rendering, 78 ß-ehminadon, in degradaflon of pectins, 251 ß-glucosidase, production by

micro-organisms, 149-150 Beverages,

alcoholic, from cheese whey, 212, 213 carbonated, fortified by whey protein, 215 from chick pea, 228 from cotton seed, 227-228

311

312 Index

Beverages (cont.) milk from soybean and cheese whey,

220-226 from peanut (groundnut), 226-227 protein content, 213 from soybeans, 216 from under-utilized proteins, 211-229

Bioavailability, of iron in meat, 102 of iron in mechanically recovered meat,

102 of nutrients, in waste utilization, 279

Biological oxygen demand, 31, 42, 60, 64, 6 5 , 6 7 , 8 6

Biological testing, for safety of novel products, 286-288

Biological value of proteins from waste, 283-284

Biomass, 83 recycHng in ponds, 148

Biotechnology, need to monitor micro-organisms, 285

Birth rates, 5 Black puddings, use of blood, 200 Blanching,

inhibits de-esterification of orange waste, 257 ,260

of soy solids, in beverage manufacture, 217

B l o o d , 3 2 , 3 8 , 6 8 functional properties, 201 globin,201 in meat products, 200 in pet-foods, 267 plasma, 201 plasma and bone collagen, 188 proteins, 32, 93

Blue cheese, manufacture, 62 Bone ,

acid process for collagen extraction, 187 collagen and blood plasma, 188 collagen composition and properties, 188 collagen in admixtures improves biological

value,191 collagen, nutritive value, 191 cooking process, 186 composition of, 185 content in mechanically recovered meat,

99 ,101 conversion to edible products, 183-193 defatting of, 185, 186 edible collagen from, 187 edible phosphate composition, 190 edible phosphate from, 189 fractionation into constituents, 184 legislative problems, 192 Lensfield processes, 184-190 marrow as ingredient, 201 marrow in mechanically recovered meat,

94, 98, 107, 108 minerals in, 190

Bone (cont.) processing, 98 product specifications, 187 production of and amounts, 183 protein, 93 separation, 94 -97 soluble protein and composition, 189 specification of, for conversion, 185 utilization of, 183, 187 yields of products, 186

Bread, 20, 25 Brewers' grains, as feedingstuff, 238-239 ,

297 Brewing, 24

syrup, 61 Bulk density of feedingstuff, limitation to

use, 236 Butter supplies, 73

Candida utilis, conversion of poultry waste by ,145

Canned foods, use of pectins and alginates, 250-255

Calcium, content of mechanically recovered meat,

98, 101, 103 ,108 ,110 hydroxide in treatment of orange waste,

255 ions and rigidity of alginate and pectate

gels, 250 pectate, gelling agent, 247 release and viscosity of autoclaved

products, 252 sensitivity of alginate and pectate, 255 supplementation, 103

Calorie gap, 25 Camembert, manufacture, 63 Carageenan, as gelling agent, 247 Carbon monoxide, enhances meat colour,

202 Carbonated beverages, fortified by whey

protein, 215 Carboxymethylcellulose,

as protein precipitant, 85-91 degree of substitution, 86 nitrogen content of whey complex, 90 types of, 86 -88 viscosity of, 87 whey complex, 83-91

Carboxyl groups and de-esterification of pectin, 250

Carcinogenicity, as factor in assessing novel products, 288

Casein, isoelectric, 88 Cassava, residues as potential food, 281 Cationic resin, in protein recovery, 88 Cellulase, production by micro-organisms,

149-150 Cellulose, water binding factor in pectate

pulp, 258

Index 313

Cereals, replacement by recycled waste in feedingstuff, 294

Chaetomium cellulolyticum, straw utilization by ,145

Chalkiness, disadvantage of soy beverages, 218

Charged double layer, 34 Cheese processing, 60, 6 2 - 6 3 , 69, 85 Cheese whey,

acid, reduced off flavour, 213 composition of solids, 212 salvage for beverages, 212-216

Chemical precipitation, 3 1 - 4 4 , 85 -91 by aluminium, 31, 33, 34 by iron, 31, 34

Chemical oxygen demand, 31 , 41 , 43, 66, 86 Chickens, digestibility of proteins from

waste, 283 Chlorogenic acid, and discoloration of

sunflower products, 228 Chocolate confectionery, 21 Chronic toxicity, in assessment of novel

products, 288 Citrus systems, as pectin source, 260 Collagen,

from bones, 187-188 biological value, 191 nutritive value, 191

in meat products, 206 in sausage casings, 205, 299 stability of, in fish silage, 123

Colour, of blood, restricts use, 201 of meat products, 202 of pectate pulp, limits use, 262

Comminution, offish silage, 122 new methods for meat products, 198

Composition, of excretory waste, 243 of meat and bone meals, 243 of meat, variabihty as pet-food raw

material, 270 of microbial protein, 281 of soUd wastes in UK, 293 of sugarbeet by-products, 241 of whey solids, 212 of whey soy drink milk, 222

Contamination, microbial, of feedingstuff, 284-285 of pet-food raw material, 268-269

Continuous rotary solvent extraction plant, 78

Cooker, for rendering, 74, 75 Cooking, shrinkage, control by insoluble

protein, 204 Cosmetic standards for foods, 11 Cost,

of crude pectate gelling agent, 262 of environmental control, 302 limiting factor in waste utilization, 278

Cost {cont.) of pet-foods, control, 271 relative, of upgrading or discarding waste,

295 Cost benefit analysis,

essential in all waste recovery systems, 302 in straw upgrading, 296

Cost formulation, feedingstuffs, 237 meat products, 199 waste utilization, 297

Cottage cheese, whey in beverages, 225 Cottonseed protein, in beverage

fortification, 227 Crop,

production, 5 residues, 15 wastage, 15

Death rates, 5 Deboning, 9 4 - 9 7

fish, 94 -95 press type, 97 sieve type, 97

Degree of esterification, in alkaline treatment of orange waste, 257 critical level for gelation, 258, 260 differences between use of Ca(OH)2 and

N a 2 ( C 0 3 ) , 2 5 8 effect of pectinmethylesterase, 261-262 in stabilizing protein drinks, 225

Desolventizer, 79 -81 disc and doughnut type, 80, 81 horizontal type, 80

Derivatives, of meat products, in drastic processing, 201-202

Developing countries, potential for waste utilization, 281, 302

Development costs, factor Umiting waste utilization, 278

Diet , animal, composition, 235 Dietetic jams, pectin as stabilizer, 248, 263 Digestibihty,

low, due to lignocellulose, 278 of waste, 282-284

improved by alkah, 240 a major problem, 291

Digestible energy, 234-235 Disc and doughnut stripping column, 80 Distilleries, liquid waste as feedingstuff, 242 D N A , recombinant in waste utilization, 300 Domestic waste, 293 Dumas method for nitrogen, 87

Economics of waste utilization, 291-303 Edible food gums, to recover protein, 85 -86 Effluent,

disposal or utilization, 301 economics of protein recovery, 301-302

314 Index

Egg box structure, in pectate, 249 Eisenia foetida,

chemical composition of, 155 conversion ratios, 159 economics of process, 161 effect of temperature, 158-159 effect on nitrogen in potato waste, 160 effect on potato waste, 160 growth in potato solids, 155 in potato wastes, 154-162 methods for processing potato waste, 160 separation from waste, 162 stocking rates, 156-157 time to reach maturity, 159

Electrodialysis, 62 in whey protein recovery, 213

Emulsifying properties, of b lood,201 of proteins, 204

Energy, audit, 19 cost of, in rendering, 73 fossil, 3, 4, 5, 6, 8, 9 from feedstuffs, definition of types, 234 in animal products production, 7 in desolventizing, 81 in fat extraction, 7 in grain production, 6 sources, 5 use in food systems, 7 values contrasted with economics, 296

Enforcement, inhibits waste utilization, 299 Ensilage see Silage Ensilation aids, 118 Environmental control authorities, 81 Enzymes,

added value, 139 availability, 134, 136 composition of industrial, 136 hydrolytic, 138 in fish silage, 123 in food processing, 133, 136 inactivation by process conditions, 139 kinetics, 138 modification of waste, 202, 300 potential of, 133, 134 problems in industrial use, 138 processing costs, 137,139 survey of use in food industry, 134 usage in UK, 137 use of, in upgrading food by-products, 133

Erosion of soils, 8 Esterification, degree of, in pectin structure,

249 Ethyl vinyl ketone, an off-flavour in

soybean,217 European Economic Community,

guidelines, 285-286 Excretory by-products, as animal

feedingstuffs, 243-294 Extraction, of meat, newer methods, 201

Extrusion, alginates and extrudate properties, 176 chemical changes during, 174 conditions for texturization, 171 deodorization by, 176 effect of lipid on extrudate, 172 effect of moisture and temperature on

extrudate, 172 effect on microbial contamination, 179 effect on nutritional value, 179 effect of polysaccharides, 175 isopeptide links, 175 of proteins, 170-176 of soya/offal mixtures, 172 proteinaceous materials used in, 171 relationship between conditions and

extrudate properties, 171

Factory Act , 81 Farmers' lung, caused by actinomycetes, 148 Fat,

depth (P2) ,48 dietary, and increased digestibility, 235 emulsification by proteins, 204 emulsions, 42 rendering, 73 -83

Fat extraction, 73 -83 by hexane, 7 8 - 8 0 , 8 2 by solvent, 78 -81 from yeast, 81 -83 mechanical, 76 percolation stage, 76, 79 tallow/solvent mixes, 7 9 - 8 0

Feedingstuffs, industrial use of wastes, 297-298 limitations in use, 236-237 protein status, 234 recycling of waste, 293-295 waste utilization, 233-234

Feeding trials, with apex meal, 4 7 - 4 8 with whey/CMC complex, 91

Fermentable sugars, in silage, 118 Fermentation,

by micro-organisms, for animal feedstuffs, 244

feed production by, 143 rumen, and waste utilization, 233-234

Fermosin, 83 Fertilizers, 4 , 6 , 7, 8, 9 , 1 7 Fibre, dietary and reduced digestibility, 235,

283 Fibre spinning, 164-170

Boyer process, 164 casein in, 169,170 effect of lipid in, 167 effect of polysaccharides in, 167,168 electrophoresis of proteins in, 166 lysinoalanine formation during, 166 offal proteins in, 165

Index 315

Fibre spinning (cont.) parameters for successful product, 164 plant proteins used in, 165 resistance to microbial attack, 179 stability of alginate/protein fibres, 169 strength of fibres, 166 structure of fibres, 165 utilization in food, 170 with alginate/protein mixtures, 168

Fibrinogen, 34 Fish,

deboners, 94 protein concentrate, 227 waste,

acid ensilation for feedingstuffs, 241 avoidable contamination, 268-269 in pet-foods, 267, 270 variability, 270

Fish silage, 118-130 amino acids, stability of, 124 ash content of, 122 autolysis of, 123 binders in, 120 by fermentation, 118 collagen in, 123 comminution, 122 economics of production, 127 enzymes in production of, 119, 123 for chickens, 120 for fish, 120 inorganic acids in, 121,122 lactic acid bacteria in, 119 liquid recycling, 123 methionine, stabiUty of, 124 neutralization of, 122 nutritive value of, 120 ,124 ,126 oil separation from, 126 organic acids in, 121, 122 preservation by acids, 121 process for, 119 production by acidification, 121 shrimps see Shrimp silage thiaminase in, 124 tryptophan, stabiHty of, 124 viscera processing, 125

Flavour, citrus, in whey beverages, 213 of leaf protein, 229 of meat products, 203 problem with orange waste, 262 o f s o y b e a n , 2 1 6 - 2 1 8 , 221

Flocculation of proteins, 32, 34, 38 Flour milling, 20 Foam stabilizers, 91 Food gap, projected, in developing

countries, 294 Food production, by micro-organisms, 141 Food waste, 10

availability, 134 avoidable, 19 canning and freezing, 22

Food waste (cont.) causes of, 10 ,11 conversion by micro-organisms, 141 domestic, 18, 25 estimating of, 18 from dairy products, 20, 85 from fish production, 18 from fruit and vegetable production, 15,

17 from potato production, 15 from potato processing see Potato waste from poultry, 73 from wheat production, 15 in brewing, 24 in catering, 24 in fats and oils, 20, 21 in flour products, 20 in fruits and vegetables, 22 in processing, 19, 20 in starch production, 23 in sugar products, 21 sources of, 15 ,134 ultrafiltration of, 51 unavoidable, 19, 20 value ,135

Formic acid, in silage, 117 Free fatty acids, in rendered fat, 74 Full fat soy flour, 218 Functional ingredients of meat products, 202 Functionality, of gels in canned foods,

258-260

Gel formation, by protein, 177 meat products, 178 myosin in, 177 reformed meat and, 177 texturization by, 177

Gelatin, from bone, 201 GeUing,

by crude pectate in heat processed foods, 247-263

of meat proteins, 203-204 of non-meat proteins, 204 of pectin, and degree of esteriflcation,

249-250 temperatures, 204

Glucose/galactose syrups, 61 Glucose trisulphate, 34 Glucosinolates, associated with rape seed

protein, 228 Gluten production, 64 Gossypol, removal from groundnut protein,

227 Grass,

dried as animal feedingstuff, 233 economic restraints, 297

Greaves, 73, 76, 78 hot, 76

316 Index

Green crops, under-utilized by ruminants, 297

Guar gum, use in pectate systems, 258 Gyonikusaishuki, 94

Haemoglobin, 32, 34, 38 Haem pigments,

and lipid oxidation, 103-104 in mechanically recovered meat, 101-103

Health hazard, in feedingstuffs of animal origin, 236

Heat, gelling, property of blood, 201 importance in upgrading waste, 242 penetration, of canned products, 253 processing, loss of nutrients, 279 processing, pectates as gelhng agents,

247-263 Herbs in mechanically recovered meat, 105 Hexane for fat extraction, 7 8 - 8 0 , 81 HydrocoUoids in protein recovery, 85-91 Hydroxyapatite, 40

Ice cream stabilization, 86 Insecticides, 11 Ion exchange, in concentration of whey

proteins, 216 Isoelectric precipitation, 32, 34, 37

Jams, and degree of pectin de-esterification, 249 dietetic, pectate pulp as stabilizer, 248, 263

Krill protein, 93

Labelling Regulations ( U K ) , 206 definition of meat, 207

Labels, edible, 205 Lactalbumin, 62 Lactic acid bacteria, in silage, 117 Lactoglobulin, 62 Lactose,

crystallization of, 60 extraction of, 6 0 - 6 2 in whey, 85

Lagoons, 31 Leaf protein,

flavour problems, 229 juice as source, 297

Legal constraints, encouragement of waste utilization, 302

Lensfield processes, for conversion of bone, 184-190

Lignification, impediment to plant utilization, 240, 278

Linolenic acid, from mould, 83

Lipoxygenase, mechanism of action, 217 in soybean products, 216, 219

Lignosulphonates, 34, 37, 38, 41 Lipase,

and protein digestibility, 284 in soy protein, 220 in whey protein, 214

Lipid oxidation, haem catalysis, 103,104 in mechanically recovered meat, 94 ,102 ,

103 Lungs, contamination by plastic wrappers,

268 Lysine,

and protein digestibihty, 284 in soy protein, 220 in whey protein, 214

Magnafloc, 41 Magnesium, aid in canning, 255 Maintenance, requirements for, in animals,

234 Maize starch, 23 Malted barley, 24 Management of waste, more efficient in

developing countries, 302 Manure, 8, 81

for single cell production, 81 Marrow bones, 98 -110

haem pigments in, 101 heavy metal content of, 102 stability on storage, 107 unsaturated fat content, 101

Meals, meat, from upgraded animal offals, 242

Meat, defined by Meat Product Regulations, 199 mechanically recovered, 201, 299 products,

consumer prejudice, 207 contributing materials, 199 and flavour, 203 inhibitor of gel thickening in canning,

255 modified by enzymes, 202 and pectate gels, 247-248 texture, 203 tumbhng and quality, 198 types available, 197, 198 from upgraded waste, 197-207 wastewater from industry, 301

protein determination, 207 reformed, added value, 299 reformed, technological development in

pet-foods, 272 separators, 94 -97 , 106 unfit for human consumption, in

pet-foods, 268

Index 317

Meat and bone meal, 45, 47, 73 -83 Mechanical dewatering, 44 Mechanically recovered meat, 93-113

amount of, 93 -94 antioxidants in, 105,107 ash content of, 104,107 bone content, 9 9 , 1 0 1 , 1 0 7 , 1 0 8 , 1 0 9 bones suitable for further processing, 185 calcium content, 98, 9 9 , 1 0 3 , 1 0 4 , 109 colour of, 98 ,106 composition of, 101,104 cryogenically prepared, 94 fat content, 9 9 , 1 0 3 , 1 0 4 , 1 0 7 fat stabiHty, 101-106 from beef, 93 from fish, 93, 94 from pork, 93 from poultry, 93, 94 functional properties of, 106-107 health implications, 107-108 herbs in, 105 iron content, 101,107 legislative aspects, 108-109 metal contamination, 110 method of manufacture, 94 -97 microbiological aspects, 105-106 nucleic acid content, 103 nutritional value, 107-108 packaging of, 105 pH, 106,108 products incorporating, 106 pneumatic preparation, 94 storage of, 98, 105

Membranes, cellulose acetate, 53 -57 cheese whey recovery, fractionation, 213 configurations, 5 5 - 5 8 flat sheet, 55, 57 fouling of, 59 hollow fibre, 56, 57, 58, 67 polarization of, 58 polyacrylic acid/polyvinylchloride, 53 Polyamides, 53 spiral wound, 56 sulphonated polysulphones, 53, 55 surface active layers, 54 tubular, 56, 57 types of, 53 -56

Metabolism studies, in assessing novel products, 286

Metabolizable energy, 234 Methionine,

in cheese whey and soy proteins, 220 in peanut protein, 226 in rapeseed protein, 228 stability in fish silage, 124

Microbial conversion, in upgrading waste for food, 244, 277

protein, amino acid composition of, 280 chemical composition of, 279

Microbial conversion (cont.) produced in rumen fermentation, 235

spoilage, avoidance by hygiene, 269 in waste for pet-food manufacture, 268

Microfiltration, 52, 53, 54 Micro-organisms,

aseptic monocultures, 144 cellulolytic, 145, 149 food production by, 141 mixed cultures, 146,147 plant waste degradation by, 144 polyculture, 147 pure and mixed cultures, 143 single cell protein from straw, 145 Spirulina, 147 straw as substrate, 144 utilization of pentoses, 145 waste conversion, 141 whey as substrate, 144

Milk, analogues, potential development, 300 condensed, 63 dried, 63 evaporated, 63 imitation from whey, 213 processing, 6 3 - 6 4 s o y , 2 1 8 - 2 2 0 'toned', from peanut protein, 226-227

MIRINZ, 40, 41 Miscella (tallow/solvent mix), 7 9 - 8 0 Moisture content, and feedingstuffs cost, 237 Mould,

feedingstuffs degradation, 237 source of protein, 82 source of oil, 82

Mozzarella cheese, 62 Muscle proteins, 93

acid soluble, 93 alkaline soluble, 93 stability in mechanically recovered meats,

105-110 Mushrooms,

conversion of waste by, 146, 280-281 submerged cultures of, in wastes, 147

Mycoprotein, use in pet-foods, 268 Myoglobin, interaction with gelling agents,

255 Myosin,

contribution to texture of meat products, 203

role in gel formation, 177

Net energy, 234 Non meat materials,

classifled, 199 incorporated into meat products, 198

Non-ruminants, preference for energy in feedingstuffs, 235

318 Index

Novel sources, difficulty of acceptance, 278 economic conditions for acceptance, 300 in pet-foods, 268 systematic evaluation of safety, 286-288

Nucleic acids, adverse effects on humans, 143

Nucleotides, in foods, 268, 273 Nutritional and health implications of waste

utilization, 277-289 Nutritional requirements, differences

between cats and dogs, 272 Nutritional states, of pet-foods, 270-273 Nutritive value,

of animal diet, 235 of brewers' grain, 241 of collagen, 206 of distillery by-products, 242 of peanut protein, 226 of rapeseed protein, 228 of sesame protein, 228 of whey/soy beverages, 233-235

Offals, contribution to meat content, 199 extrusion of, 173 in Meat Products Order, 98 ,109 rancidity inhibits utilization, 243 restricted to cooked products, 200

Oilseed, fullfat, for feedingstuff, 238 loss of nutrients in oil cake, 238

Orange waste, factors limiting use, 262 source of pectate pulp, 255-266

Organic acids, antimicrobial action in fish silage, 121

Organoleptic characteristics, important in waste utilization, 278

Osmosis, 51 reverse, 52, 53, 54, 60, 62, 65

Osmotic pressure, 51

Packaging, edible from collagen, 205 Palatability, of pet-foods, 271 Pea pods, as pectate source, 261-262 Peanut protein, in beverage manufacture,

226-227 Pectate,

crude, potential as gelling agent, 247-263 pulp, from citrus waste, 247, 255-260

Pectin, methylesterase, 261-263 structure and occurrence, 248-249 thermal degradation, 250-251 viscosity and degree of esterification, 252

Peroxide, index of stability in fortified beverages, 225

Percolation extractor, 79

PermeabiHty constant of membranes, 52 Pesticides, 4, 6, 9 Pet-foods, 17, 25

food waste as raw material, 267-273 nutritive status, 271-273 range of products, 267 in recycled waste, 298 use of pectate pulp, 263 variability of raw material a problem, 268,

271 wholesomeness for humans, 272

pH, changes in alkaline waste extraction, 257 and gelling of pectate in canning, 247 importance of soft drink fortification, 216 in viscosity of calcium gel systems, 254

Pharmaceutical residues, in feedingstuffs, 236

Phosphoric acid, lessens protein insolubility, 215

Pigs, digestibility of protein from waste, 283-284

Plasma, blood, as ingredient, 201 Pneumatic separation of meat and bone, 94 Pollution, control, 263, 292, 295 Polyelectrolytes, 85

in food recovery, 85 Polyphosphate, 34

adjunct in autoclaved products, 252 Polysaccharides,

as protein precipitants, 85 -prote in complexes, 8 5 - 9 0

Population, increase in, 5, 10 world, 3 , 4, 8

Potato processing, effluent from, 65, 68 Potato starch, 23 Potato waste, 153-162

amount produced, 153 processing by worms: see Eisenia foetida,

154-162 source of waste, 154

Poultry, manure as feedingstuff, 243, 294-295 variability as pet-food material, 270 waste, conversion into feedingstuff by

Candida utilis, 145 Precipitation, problems with proteins in

beverages, 216 Pre- and post-harvest waste, 292-293 Pressure cookers,

batch operation, 75 horizontal, cylindrical type, 75

Protein, blood, functional additive, 204 -calorie malnutrition, 211 concentrates, from acid cheese whey, 214 extrusion of see Extrusion fibre spinning see Fibre spinning gel formation with see Gel formation hydrolysates, 203

Index 319

Protein {cont.) ideal, 48 insoluble, contribution to texture, 203 insolubilized by heat, 215 loss, in wastewater, 301 organoleptic properties of, 163 solubility relations of, 32 status of feedingstuffs, 234 texturization of, 163-180 vegetable, low cost food, 211 whey, in beverages, 213

Protein-anion reactions, 34, 85 Protein-cation reactions, 34, 85 Protein degradation in rendering, 74 Protein precipitation, 31

by aluminium, 3 3 - 3 4 by hexametaphosphate, 3 4 - 3 6 , 4 4 by iron, 31, 34 by lignosulphates, 34, 37 by pH, 33, 37, 39, 44, 48

Psychology, and consumer acceptance, 298 Public Health Authorities, 81 Purines, relative importance to

human/animal consumers, 273

Rancidity in feedingstuffs, 237 Rapeseed protein, as potential source, 228 Recovered solids, 42, 43

ash in, 43, 44 fat in, 44 utilization of, 44

Recycling of wastes, for animal feedstuffs, 293-297 for human food, 298-301

Reformed meats, 177 Refrigeration, limits spoilage of raw

materials, 268 Regulations, national, as factors limiting

utilization, 278 Rehydration, of whey beverage powders,

225 Rendering of fat, 73 -83

continuous, 76 wet, 76

Reproduction, as indicator of safety of novel foods, 288

Residual meat, better utilization, 200 Resources,

energy, 3 land, 3, 6, 8, 9 livestock, 6 water, 3 , 6, 9

Restricted offal, recommended list, 200 Reverse osmosis, 5 2 , 5 3 , 54, 60, 62, 65

in recovery of cheese whey, 213 Rhamnose, structural component of pectin,

248-249 Rind, dried, contributes to texture, 204 Rind pig, cooked separately, 199 Rising film evaporator, 80

Rotary extractor, 79 advantages of, 7 9 - 8 0

Rotary vacuum filtration, 69 Ruminants, fed recycled excreta, 294

underutilization of green crops, 296-297 Rumination and utilization of waste,

2 3 3 - 2 3 4 , 2 8 0

Safety of novel products, systematic evaluation, 286-288

Salmonellae spp., in feedingstuffs, 236, 280, 284-285

Sausage casings, from collagen, 201, 205, 299 Scale of operations, economic factors in

waste recovery, 303 Schnecken desolventizer, 80 School meals, 24 SCP see Single cell protein Screening, 31, 65 Screw press,

low pressure, 74, 77, 78 high pressure, 74, 76, 78 Simon-Rosedowns mark 3M, 37, 77

Sedimentation, 3 1 , 4 1 , 6 5 Sesame, potential use, 228 Shear stress of canned products containing

citrus waste, 258, 261 Short-term feeding trials, in product

evaluation, 287 Shrimp silage, 128

economics, 129 pigment stabihty, 129 production, 129

Silage, 117 enhances digestibihty of feedingstuffs, 241 fish see Fish silage microbial flora, 118

Single cell protein, 73, 8 1 - 8 3 amino acid composition, 273 microbial conversion of waste, 279 -281 ,

300 nucleic acids in, 142-143 nutritive value, 142 in pet-foods, 268, 273 production on,

agricultural by-products, 81 methane, 81 methanol, 81 straw, 145

Spirulina, 147 Skin,

contribution to meat content, 199 source of collagen, 299

Slaughterhouse effluent, 3 1 - 4 9 Slaughterhouse wastes, 16, 17, 31, 74 Snack meals, 24 Sodium alginate, in canned foods, 251-255 Sodium carbonate, in treatment of citrus

waste, 256-258

320 Index

Sodium hexametaphosphate in protein recovery, 34, 35, 36, 38, 41

Soft drinks, nutritional enhancement by whey, 214 ,215

Soil erosion, 8 Solids discharging separator, 88 Solvent extraction, 73 Soy milk, 218-226 Soy oil, enhances fat absorption, 235 Soybeans,

in beverage manufacture, 216 chalkiness a disadvantage, 218 off-flavour, 216 -218 ,221

Soyflour, full fat for beverages, 218 Species, difference in digestibility of waste

protein, 283-284 Spirulina, as human and animal feed, 147 Spoilage, preprocess, limitation in pet-food

industry, 269 Spray drying, of soy and whey beverages,

220-221 Starch, 23

production, 64 Sterilization,

of feedingstuffs by heat, 242 of waste for use in pet-foods, 268, 272

Straw, cereal, upgraded for feedingstuff, 239 conversion by bacteria, 280-281 conversion by mushrooms, 280-281 increased utilization in feedingstuffs, 293,

295-296 limitations due to lignocellulose, 278, 280 potential for exploitation, 281 upgrading, cost-benefit analysis, 296

Structured protein fibres, in meat products, 107 from soy, 107

Sugar beet residues, ensilage for feedingstuffs, 241

Sugar confectionery, 61 Sugar waste, potential for ufilization, 300 Sulphur dioxide, preservation of pet-foods,

272 Sunflower protein, in beverages, 228 Superfloc in protein recovery, 41, 45 Superphosphate, 40 ,41

Taint, 'fishy', from rapeseed feedingstuff, 236 'whey', 213

Tallow/solvent mixes, 7 9 - 8 0 Target species, use in assessing toxicological

value, 283-284 , 288 Temperature,

degradation of pecfins, 250-257 elevated, and pecfinmethylesterase,

261-262 gelling of proteins, 204 HTST pasteurization of whey, 215 U H T sterilization of soy milk, 219

Teratology, in assessing novel foods, 288 Texture, of meat products, 203 Texturization,

antinutritional factors and, 178 by extrusion, 170-176 by fibre spinning, 164-170 by gel formation, 177-180 microbiological aspects, 178 nutrifional aspects, 178 of proteins, 163

Tomato skins, a source of pectate, 261 Toxins,

E E C guidelines, 285-286 a factor in waste utilization, 279 in feedingstuffs, 236 in vegetable proteins, 228

Transport, and economics of waste utilization, 295-296

Trickling filter, 31 Trimethylamine, in fish silage, 118 TumbUng, and improved meat products, 198

Ultrafiltration, 5 1 - 7 2 economics of, 6 8 - 6 9 in the dairy industry, 5 9 - 6 4 , 69 in enzyme and protein purificafion, 67 in trypsin purification, 67 in whey utilization, 213, 216 of blood, 68 of fermentation products, 67 of fruit juices, 68 of paunch materials, 67 of potato processing effluent, 65, 68, 69 of vegetable extracts, 6 4 - 6 6 of vitamins and viruses, 67 o fwhey , 5 9 - 6 9 , 85

Uric acid oxidase, missing in humans, 142 Uricase, and utilization of purine-rich foods,

273

Vacuum, and reducfion of beany flavour from soya,

222 and retention of nutritive value of soya,

223-225 Value, added to meat products, 198 Value, nutritive and 'market' of waste, 298 Variability, of raw materials, problem in

manufacture, 268, 271 Vegetable protein, 93 Vendor assurance scheme, value in pet-food

industry, 268-269 Vinegar, 24 Viscera, ensilation of fish, 125 Viscosity,

of autoclaved pectins, 252 of calcium gel systems, 254 of canned products containing orange

waste, 259

Index 321

Viscosity {cont.) comparative, of alkali treated waste, 261

Vitamin A , stability in whey/soy beverages, 223-224

Waste, abattoir, need for improved utilization,

299 conversion by micro-organisms, 279 digestibility of proteins, 282-284 direct use as feedstuffs, 238-239 disposal, a major problem, 291 economic need for better use, 292 ,298 liquid, as feedingstuff, 238 limits to utilization, 278 non-citrus as pectin source, 260-262 nutritional and health implications,

277-289 in pet-food industry, 269 rancidity a problem, 268-271 routes to food, 279-281 recychng, an economic problem, 292 sources on UK farms, 292 upgrading for feedingstuffs, 239-244 upgrading on farms, 296 vegetable, integration with animal

production, 295 Water binding, benefits of pectate pulp, 258 Water consumption, 9 Wheat gluten, as texture additive, 204

Wheat mining, nutritive value of components , 237

Whey, acid, 8 5 , 8 8

protein isolation procedure, 214 reduction of off-flavours, 213

cottage cheese, 87 liquid, as feedingstuff, 238 in pet-foods, 268 processing, 6 0 - 6 2 , 69 protein,

composition, 86 functional additive, 204 functional properties of, 62 insolubilized at high temperature, 215 lysine content high, 214 recovery, 85 -91 spray dried, 214 types of beverage, 213

soybean,65 uhrafiltration of, 5 9 - 6 2 , 85

Whey/soy beverages, 221-226 Whipping agents, 91 Whisky, malt, use of pot ale for

feedingstuffs, 242 Worms, in potato wastes see Eisenia foetida

Yeast, 24, 82 extraction of fat from, 82

Yoghurt manufacture, 60 Ymer manufacture, 60, 63

upgrading waste for feeds and food. proceedings of previous easter schools in agricultural science

Documents