characterization of the ripening of carambola … · years, little is known about the physiology of...

5

Upload: others

Post on 17-Jun-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: CHARACTERIZATION OF THE RIPENING OF CARAMBOLA … · years, little is known about the physiology of carambola ripening. Although changes in color, soluble sugars and acids have been

Table 2. Changes in standard curve correlation coefficient (r) through

time.

Time(min) r

Table 3. Activity estimates from assaying known amounts of OPPME;

r = 0.99986 for 0.009, 0.012 and 0.015 units.

0

5

10

15

20

25

30

-0.997149

-0.996521

-0.994151

-0.990756

-0.988862

-0.984617

-0.980215

rimetric estimates were in close agreement, giving a value

of 288.6 ± 11.21 units • ml-1 (n = 16). Colorimetric assay

of 0.006, 0.009, 0.012 and 0.015 units of OPPME indicated

that the microplate assay was accurate between 0.009 and

0.015 units of enzyme (Table 3). Higher concentrations

could be assayed but were limited by the rapidity of the

change in optical density. The activity estimate obtained by

assaying 0.006 units was only one-half of the predicted

value (0.0029 units).

The commercial non-pasteurized orange juice con

tained 0.0133 ± 0.0005 [x equivalents acid released -min-1

• jxl"1 (n= 18) while the hand-expressed juice contained only

0.001 ± 0.00003 fx equivalents acid released -mirr1 ixh1

(n=60). These values are equal to 13.3 units -ml-' and 1.0

unit -ml"1 respectively.

Conclusions

The colorimetric assay for PME activity introduced by

Hagerman and Austin (1986) has been adapted successfully

to a kinetic microplate reader. Results presented here dem

onstrate that the method is accurate and applicable to a

commercial PME as well as a commercial, non-pasteurized

and a hand-expressed orange juice. The sensitivity of the

assay is slightly greater than originally reported by Hager

man and Austin (1986). As little as 0.009 units of PME

could be measured accurately. The assay requires very small

volumes of substrate and enzyme. The microplate format

allows for extensive replication of individual samples, mak-

Added Activity

(units)

Estimated Activity

(units)

0.006

0.009

0.012

0.015

0.00294 ± 0.00022 (n = 16)

0.00898 ± 0.00037 (n = 32)

0.01205 ± 0.00056 (n = 32)

0.01495 ± 0.00042 (n = 16)

ing statistical treatment possible. Finally, adaptation of the

colorimetric PME assay to a kinetic microplate reader saves

a tremendous amount of time. After becoming familiar

with the procedure, 24 samples could be assayed, with each

sample replicated three times, in less than one hour. Addi

tionally, computer software is available to operate the

microplate reader, collect and manipulate data and present

final estimates of enzyme activity.

Literature Cited

1. Baldwin, E. A. and R. Pressey. 1989. Pectic enzymes in pectolyase.

Separation, characterization, and induction of ethylene in fruits. Plant

Physiol. 90:191-196.

2. Fox, J. D. and J. F. Robyt. 1991. Miniaturization of three carbohydrate

analyses using a microsample plate reader. Anal. Biochem. 195:93-96.

3. Hagerman, A. E. and P. J. Austin. 1986. Continuous spectrophotomet-

ric assay for plant pectin methyl esterase. J. Agric. Food Chem. 34:440-

444.

4. Koch, J. L. and D. J. Nevins. 1990. The tomato fruit cell wall. II.

Polyuronide metabolism in a nonsoftening tomato mutant. Plant

Physiol. 92:642.

5. Plantner, J. J. 1991. A microassay for proteolytic activity. Anal.

Biochem. 195:129-131.

6. Rombouts, F. M. and W. Pilnik. 1978. Enzymes in fruit and vegetable

juice technology. Process Biochemistry 13:9-14.

7. Tucker, G. A., N. G. Robertson and D. Grierson. 1982. Purification

and changes in activities of tomato pectinesterase isoenzymes. J. Sci.

Food Agric. 33:396-400.

8. Versteeg, C, F. M. Rombouts, C. H. Spaansen, and W. Pilnik. 1980.

Thermostability and orange juice cloud destabilizing properties of

multiple pectinesterases from orange. J. Food Sci. 45:969-971, 998.

9. Wenzel, F. W., E. L. Moore, A. H. Rouse, and C. D. Atkins. 1951.

Gelation and clarification in concentrated citrus juices. I. Introduction

and present status. Food Technol. 5:454-457.

Proc. Fla. State HorL Soc. 104:104-108. 1991.

CHARACTERIZATION OF THE RIPENING OF

CARAMBOLA (AVERRHOA CARAMBOLA L) FRUIT

Elizabeth J. Mitcham and Roy E. McDonald

Agricultural Research Service

U.S. Department of Agriculture

2120 Camden Road

Orlando, FL 32803

Additional index words: cell wall, firmness, color, ethylene,

respiration, pectin.

Acknowledgements: The authors would like to thank Craig Campbell

of J.R. Brooks & Son for hand-selecting and donating the fruit used in

this study. We also acknowledge the excellent technical assistance of Ms.

Heather Tucker.

Abstract. Carambola fruit were harvested at four stages of

ripeness: dark green (DG), light green, color break (CB) and

ripe. Additional ripe fruit were stored at 21C until overripe

(OR). The ratio of CIE color a/b increased during ripening in

versely to the decrease in fruit firmness. Respiration and

ethylene production of carambola fruit of different ripeness

stages suggested a possible climacteric pattern. However,

daily monitoring of individual fruit respiration and ethylene

production provided inconclusive evidence as to the climac-

teric/nonclimacteric nature of carambola fruit. The cell walls

of DG carambola fruit were comprised mainly of cellulose

(60%) and hemicellulose (27%), with pectin polymers account

ing for only 13%. There was an increase in the proportion of

less tightly bound chelator-soluble pectin and a decrease in

104 Proc. Fla. State HorL Soc. 104: 1991.

Page 2: CHARACTERIZATION OF THE RIPENING OF CARAMBOLA … · years, little is known about the physiology of carambola ripening. Although changes in color, soluble sugars and acids have been

covalently-bound pectin during ripening. The amount of total

cell wall uronic acid decreased at the OR stage. The amount

of hemicellulose decreased beginning at the CB stage and the

proportion of cellulose increased throughout the ripening pro

cess.

Although carambolas have been commercially grown in

Asia for many years, the industry in South Florida is rela

tively new. While production has increased rapidly in recent

years, little is known about the physiology of carambola

ripening. Although changes in color, soluble sugars and

acids have been documented during fruit ripening

(Campbell et al., 1989; Vines and Grierson, 1966), there is

little or no information available on the textural changes

that occur during ripening of carambola fruit. The soften

ing process is of great commercial importance because the

postharvest shelf life of the fruit is limited by increased

softness which leads to an increase in susceptibility to

mechanical damage and disease. Softening of fleshy fruits

is mainly due to a modification of cell wall structure. Cell

wall structure and modification thereof have not been inves

tigated in the carambola fruit.

Also, uncertainty exists about the classification of caram

bola fruit as climacteric or nonclimacteric. Vines and Grier

son (1966) observed a respiratory climacteric in 'Golden

Star' carambola and classified carambola as climacteric.

However, Oslund and Davenport (1983) found no evidence

that 'Golden Star' carambola was climacteric. Similarly, Lam

and Wan (1983) found no evidence for a climacteric pattern

in carambola (cv. BIO), and exposure to Ethrel, a treatment

which releases ethylene gas, failed to induce ripening (Lam

and Wan, 1983). It was recently reported that treatment

with 100 ppm ethylene can promote carambola ripening

(Sargent and Brecht, 1990). This type of response to

ethylene treatment is normally associated with climacteric

fruit. It has been suggested that carambola is a climacteric

fruit and that the climacteric occurs very late in the ripening

process.

The objective of this investigation was to increase our

knowledge of the physiology of carambola ripening by

exploring changes in fruit texture and cell wall composition

during ripening, and additionally, to search for evidence

of a climacteric pattern in carambola.

Materials and Methods

Fruit material. Carambola (Averrhoa carambola L., cv.

Arkin) fruit were harvested in Homestead, Fla. at four

different stages of development as determined subjectively

by color development: 1) dark green (DG), no yellow color;

2) light green (LG), green with pale yellow between ribs;

3) color break (CB), mostly yellow with some green remain

ing, orange color between ribs; and 4) ripe (R), yellow to

light orange. Fruit were transported to Orlando, gently

dipped in 0.015% sodium hypochlorite (30C, pH 7.7),

rinsed with deionized water (dH2O) and dried. Fruit were

stored at 21C, 90% RH overnight. Eight to 12 fruit of good

condition representing DG, LG, CB, and R stages were

selected for respiration, ethylene production, color, firm

ness, and cell wall analysis. Additional R fruit were stored

(1 to 3 weeks) until overripe (OR, dark orange color), at

which time they were also sampled.

Respiration and ethylene production. Eight single fruit or

pairs of small fruit at each stage of ripeness were sealed in

Proc. Fla. State Hort. Soc. 104: 1991.

930 ml jars at 21C. After 15 min, 1 ml gas samples were

analyzed for CO2 using a Hewlett Packard 5880 gas

chromatograph equipped with a Porapak Q column and a

thermal conductivity detector (TCD). Nitrogen served as

the carrier gas at a flow of 30 ml-mirr1. Only five measure

ments of CO2 production were obtained from OR fruit as

fruit showing any sign of decay were not used. The same

fruit were then sealed in 930 ml jars at 21C with 30 g of

soda lime (sodium hydroxide with calcium oxide or hydro

xide) to absorb CO2. After 22 to 24 hr, 1 ml gas samples

were analyzed for ethylene using a Hewlett Packard 5880

gas chromatograph equipped with an alumina column and

a flame ionization detector (FID). Nitrogen served as the

carrier gas at a flow rate of 30 ml-min-1.

An additional 20 fruit of various stages of ripeness from

LG to R were monitored daily for CO2 production as de

scribed above. A separate group of 15 fruit were monitored

every second day over a 20-day period for ethylene produc

tion as described above.

Color. After gas measurements, color measurements

(CIE L*a*b) were taken on the same 8 to 12 fruit at each

ripeness stage. A Minolta CR200 Colorimeter fitted with

an aperture plate (11 mm diameter) was used to measure

external color. The mean of three measurements per fruit,

on the side of three wings, were averaged for fruit of each

ripeness stage.

Firmness. Carambola fruit were supported by a triangu

lar platform with the flat side of one wing perpendicular

to a round, convex probe (8 mm diameter) attached to an

Instron Universal Testing Instrument. The Instron was

fitted with a 100 kg load cell and operated at a crosshead

speed of 25 mm-min-'. Resistance to compression (3 mm)

was measured on three wings per fruit.

Cell wall extraction. Fruit were peeled, and seeds and

endocarp removed. The mesocarp from two to three fruit

per stage were combined, frozen in liquid nitrogen, and

stored at —80C for subsequent cell wall extraction. Later,

tissue was thawed, homogenized 3 min with a Polytron in

3 volumes of 80% ethanol, treated in a cell disruption bomb

to break remaining whole cells and rinsed through Mirac-

loth with 2 volumes 20 mM Hepes-NaOH (pH 6.9). The

residue was stirred in 2 volumes phenol:acetic acid:dH2O

(2:1:1, w/v/v) for 20 min, and successively rinsed in 3 vol

umes chloroform:methanol (1:1, v/v) and 3 volumes

acetone. Cell walls were dried in a vacuum oven at 40C

over phosphorous pentoxide.

Cell wall uronide determination. Cell walls (10 mg) were

incubated on ice in 2 ml concentrated sulfuric acid on a

gyratory shaker. Two 500 |xl aliquots of dH2O (4C) were

added slowly and the solution incubated on a shaker until

dissolved (6 hr). Uronic acid concentration was estimated

by the carbazole method (Dische, 1947).

Cell wall fractionation. Cell walls (150 mg) were fraction

ated into chelator-soluble pectin (CSP), covalently-bound

pectin (CBP), a hemicellulosic fraction (HF), and a cellulosic

fraction (CF) as described previously (Mitcham et al., 1989),

except that the HF was extracted using a single incubation

in 8 N KOH for 3 hr. Carbohydrate concentration of cell

wall fractions was determined by the phenol-sulfuric acid

method (Dubois et al., 1956). Uronide concentration was

determined using the carbazole method (Dische, 1947).

Cellulose determination. Cellulose content of crude cell

walls was estimated according to the procedure of Updeg-

raff (1969). Cell walls (10 mg) were incubated in teflon-cap-

105

Page 3: CHARACTERIZATION OF THE RIPENING OF CARAMBOLA … · years, little is known about the physiology of carambola ripening. Although changes in color, soluble sugars and acids have been

ped tubes with 1 ml 2 N tritluoroacetic acid (TFA) for 1

hr at 121C to hydrolyze noncellulosic neutral sugars. The

TFA solution was removed and the residue rinsed twice

with dH2O and incubated at 30C for 1 hr with 1 ml 78%

sulfuric acid. The volume was brought to 25 ml with dH2O

and incubated at 121C for 1 hr. Cellulose content was esti

mated using the Anthrone method for total hexose (Spiro,

1966) with glucose as the standard.

Results and Discussion

Color and firmness. CIE color "a" and "b" values both

increased during ripening of carambola fruit (data not

shown); however, the ratio of a/b provided the most linear

relationship to ripeness stages which had been selected

based on subjective color analysis (Fig. 1). Resistance to

compression (firmness) decreased linearly as ripening prog

ressed (Fig. 1). The decrease in firmness was inverse to the

increase in a/b color indicating that determination of ripe

ness stage based on fruit color is very accurate.

Respiration and ethylene production. The rate of CO2 pro

duction of fruit picked at different ripeness stages (Fig. 2)

was 30 ml-kg-'-h"1 at the DG and LG stages, decreased to

20 ml-kg-'-hr1 at the CB stage, then increased to the original

rate by the OR stage. The pattern of CO2 production indi

cated a possible preclimacteric minimum at the CB stage

with a peak in respiration at the OR stage. However, there

is no evidence for a decrease in CO2 production after the

peak. Ethylene production from the same fruit began to

increase at the CB stage and increased sharply at the OR

stage (Fig. 2). The results for ethylene production were

similar to those of Oslund and Davenport (1983).

To search for additional evidence of a climacteric pat

tern in carambola fruit, CO2 and ethylene production was

monitored over time (Fig. 3) in fruit harvested at different

stages of ripeness. No consistent pattern in respiration was

observed in fruit harvested at the CB stage (Fig. 3A) or any

other ripeness stage (data not shown), and we did not ob

serve a peak in CO2 production like that reported by Vines

and Grierson (1966). As fruit became OR, decay often in

terfered with respiration measurements. The presence of

decay was associated with an increase in respiration, which

may have been due to a wound response or to respiration

of the decay organisms.

Ethylene production, however, followed a climacteric

pattern with a sharp peak in production at approximately

-2

-4

-6

DG LG CB R

Stage of Ripeness

OR

Fig. 2. CO2 and ethylene production of carambola fruit at different

stages of ripeness. Each value represents the mean production rate of 5

to 8 individual fruit or pairs of fruit at each stage of ripeness. Vertical

bars represent SE.

the R stage (Fig. 3B). All fifteen fruit monitored underwent

a peak in ethylene production, but the timing and mag

nitude varied with the ripeness stage at harvest. Two CB

fruit underwent an immediate peak in ethylene production

while the ethylene peak of R fruit occurred several days

later and was lower in magnitude (Fig. 3B). It is possible

that R fruit underwent peak ethylene production prior to

harvest. Two LG fruit underwent a peak in ethylene pro-

60

50 g-

40 il zi O -i'

30 § «

20 2

10

20

DG LG CB R

Stage of Ripeness

Days

Fig. 1. Color and firmness of carambola fruit at different stages of

ripeness. Each value is the mean of 3 measurements per fruit and 8 to

12 fruit per stage. Vertical bars represent SE.

106

Fig. 3. A) Daily CO2 production of 3 individual CB carambola fruit.

B) Ethylene production of individual LG, CB, and R carambola fruit measured every second day. Arrows designate the point when decay was

visible.

Proc. Fla. State Hort. Soc. 104: 1991.

Page 4: CHARACTERIZATION OF THE RIPENING OF CARAMBOLA … · years, little is known about the physiology of carambola ripening. Although changes in color, soluble sugars and acids have been

duction at the CB to R stage. Most fruit exhibited signs of

decay 2 to 4 days after the peak in ethylene production.

The decay organisms did not appear to contribute to ethylene production since the production rate continued

to decrease after decay began. However, ethylene produc

tion by the carambola fruit may have been a pathogen-in

duced, wound response that occurred before decay became visibly noticeable. Alternatively, normal production of

ethylene by the fruit may have stimulated decay develop ment.

Our data regarding the climacteric nature of carambola

are inconclusive. However, as explained by Biale (1960),

placing a fruit in the nonclimacteric category calls for a

more tentative decision than listing one as climacteric be

cause positive evidence can be experimentally obtained

much more readily than convincing negative proof.

Cell wall composition. The carbohydrate portion of DG

carambola cell walls was comprised largely of cellulose

(60%) and hemicellulose (27%) polymers with pectin polym

ers comprising only 13%. In unripe fruit, a greater percent

age of the pectin was covalently-bound pectin (CBP) which

is more tightly bound within the cell wall (Fig. 4A). The

proportion of loosely associated chelator-soluble pectin

(CSP), which can be solubilized from the cell wall with cal

cium chelators, was lower than CBP in the unripe fruit (Fig.

4 A). However, in OR fruit, the increase in CSP and decrease

in CBP resulted in equal amounts of the two types of pectin

polymers. In addition to a shift in the type of pectin polym

ers present during ripening, there was also a slight decrease

in cell wall uronides, the major component of pectin, after

the R stage (Fig. 4B). The modification of pectin polymers

from CBP to CSP began after the LG stage along with the

decrease in fruit firmness; however, there was no net loss of uronic acid from the cell wall until after the R stage.

The changes in carambola pectin polymers during ri

pening are similar to those found in strawberry fruit. The

amount of cell wall uronide decreases late in ripening (Knee

et al., 1977), and a shift to CSP also occurs during ripening

of strawberry fruit (Woodward, 1972). Woodward pro posed that a change in the cell wall po/ysaccharides takes

place in such a way that the uronide polymers become

rearranged to allow more plasticity in the walls of the fruit.

Neal (1965) suggested that chelation of calcium ions, which

are thought to form crosslinks between carboxyl groups of

polyuronide chains, may lead to softening of the tissue.

In addition to the pectin changes, there was also a de

crease in the amount of hemicellulosic material which began

at the CB stage (Fig. 5A). The amount of cellulosic material

increased linearly throughout ripening (Fig. 5B). The in

crease in the proportion of cellulose may be due to a loss

of other cell wall polymers.

It is interesting that the amount of all four types of cell

wall polymers (mg/150 mg cell wall) increased from the DG

stage to the CB stage (Figs. 4 & 5). These data indicate a

decrease in another cell wall component, perhaps protein.

Conclusions

The development of carambola fruit CIE a*/b* color

was inverse to the decrease in fruit firmness indicating color

is an accurate means by which to determine fruit ripeness.

Carambola cell walls were comprised mainly of cellulose

DG LG CB R

Stage of Ripeness

Fig. 4. A) Galacturonic acid equivalents of chelator-soluble pectin (CSP)

and covalently-bound pectin (CBP) extracted from 150 mg cell wall at

different ripeness stages. B) Total galacturonic acid equivalents extracted

from 10 mg cell wall.

Proc. Fla. State Hort. Soc. 104: 1991.

LG CB R

Stage of Ripeness

Fig. 5. Glucose equivalents of hemicellulosic (A) and cellulosic (B) polymers extracted from 150 mg cell wall at different stages of ripeness.

107

Page 5: CHARACTERIZATION OF THE RIPENING OF CARAMBOLA … · years, little is known about the physiology of carambola ripening. Although changes in color, soluble sugars and acids have been

and hemicellulose with only 13% pectin; however, the pro

portion of less tightly bound chelator-soluble pectin in

creased with ripening.

Literature Cited

1. Biale, J. B. 1960. Respiration of fruits. In: Encyclopedia of Plant

Physiology, Vol. XII/2, J. Friend and M.J.C. Rhodes (eds.), p. 536-

592, Springer-Verlag, Berlin.

2. Campbell, C. A., D. J. Huber, and K. E. Koch. 1989. Postharvest

changes in sugars, acids, and color of carambola fruit at various

temperatures. HortScience 24:472-475.

3. Dische, Z. 1947. A new specific color reaction of hexuronic acids. J.

Biol. Chem. 167:189-198.

4. Dubois, M., K. A. Gilles, J. K. Hamilton, P. A. Rebers and F. Smith.

1956. Colorimetric method for determination of sugars and related

substances. Anal. Chem. 28:350-356.

5. Knee, M., J. A. Sargent, and D. J. Osborne. 1977. Cell wall metabolism

in developing strawberry fruits. J. Exp. Bot. 28:377-396.

6. Lam, P. F. and C. K. Wan. 1983. Climacteric nature of the carambola

(Averrhoa carambola L.) fruit. Pertanika 6(3):44-47.

7. Mitcham, E. J., K. C. Gross, and T. J. Ng. 1989. Tomato fruit cell

wall synthesis during development and senescence. In vivo radiolabel-

ing of wall fractions using [14C] sucrose. Plant Physiol. 89:477-481.

8. Neal, G. E. 1965. Changes occurring in the cell walls of strawberries

during ripening. J. Sci. Food Agric. 16:604-611.

9. Oslund, C. R. and T. L. Davenport. 1983. Ethylene and carbon

dioxide in ripening fruit of Averrhoa carambola. HortScience 18:229-

230.

10. Sargent, S. A. and J. K. Brecht. 1990. Ethylene pretreatment allows

early harvest of carambola. HortScience 25:1174.

11. Spiro, R. G. 1966. Analysis of sugars found in glycoproteins. In:

Methods of Enzymology, Vol. 8, E. F. Neufeld and V. Ginsburg,

(eds.), p. 4-6, Academic Press, NY.

12. Updegraff, D. M. 1969. Semi-micro determination of cellulose in

biological materials. Anal. Biochem. 32:420-424.

13. Vines, H. M. and W. Grierson. 1966. Handling and physiological

studies with the carambola. Proc. Fla. State Hort. Soc. 79:350-355.

14. Woodward, J. R. 1972. Physical and chemical changes in developing

strawberry fruit. J. Sci. Food Agric. 23:465-473.

Proc. Fla. State Hort. Soc. 104:108-111. 1991.

AN EXPERT SYSTEM FOR HYDROCOOLERS

Jean-Pierre Emond and Khe V. Chau

Agricultural Engineering Dept.

IF AS, University of Florida

Gainesville, FL 32611

Jeffrey K. Brecht

Vegetable Crops Dept.

IF AS, University of Florida

Gainesville, FL 32611

Additional index words, cantaloupe, celery, cucumber,

peaches, sweet corn, cooling.

Abstract. This paper describes an expert system that has been

developed to give the specifications required to cool produce

with a hydrocooler. Provided with a user friendly interface,

this expert system can perform a wide range of tasks such as:

hydrocooler design, precooling process optimization, quality

control, energy conservation and investigation of new applica

tions. In order to execute these tasks, the expert system needs

information provided by the user such as: kind of product,

loading capacity, entering and exit temperature of the product,

water temperature and residence time. Depending on the in

formation given, the expert system will use default values for

the missing information to compute the residence time, product

temperature, refrigeration capacity, water flow rate and the

maximum shelf life of the product if stored at the exit temper

ature. Assumptions made in this expert system are: cooling

efficiency of 50%, 3°F rise in temperature of the water going

through the hydrocooler, exit product temperature determined

by the 7/8 cooling time unless the chilling injury threshold

temperature of the product is higher. Further improvement

will make the software more versatile and applicable to a

wider variety of commodities.

The temperature of fresh fruits and vegetables is the

greatest determinant of the rate of deterioration by decay

Florida Agricultural Experiment Station Journal Series No. N-00536.

108

and senescence, and consequently, of the potential market

life (Kader et al., 1985). Produce temperature control is

critical from the moment of harvest when the process of

deterioration begins; delay in cooling the product can cause

loss of quality.

Several methods for rapid removal of heat from pro

duce are in commercial use (Ryall and Lipton, 1979). The

choice of a particular method depends mainly on the rate

of cooling desired, product surface to volume ratio, product

susceptibility to water damage, availability of equipment,

the value and the perishable nature of the product. It is

very important to select the optimal precooling system to

satisfy the commodity and the market structure needs. A

knowledge-based system for selecting precooling methods

was developed by Morey et al. (1988) to help producers

faced with a decision of selecting a precooling method. This

computer program (Morey et al., 1988) gives the users the

opportunity to specify the crop and information about their

operation, and uses this information to select possible cool

ing methods, which can be room cooling, forced-air cooling,

hydrocooling, vacuum cooling or package icing. However,

the above system does not provide-any pertinent informa

tion on the possible cooling methods selected.

Once a precooling method is selected, the service of an

expert to characterize the system is required. With their

knowledge, experts are able to determine the specification

needed to design and operate a precooling unit. This kind

of expertise can be costly and may not be readily available.

The same problem can be observed for owners of precool

ing units who want to change operating conditions of their

units or to use them for other crops. Only an expert can

provide this kind of information. In many cases, a joint

effort of a team of experts in postharvest physiology and

engineering would be useful. These teams cannot always

be available at the time the information is needed. For these

reasons, computer programs that simulate human experts

in this area are needed. By storing the knowledge of many

experts in its memory bank a computer program ("expert

system") can provide answers to questions from users. Com-

Proc. Fla. State Hort. Soc. 104: 1991.