Download - June 22, 2007 12:51 RPS rpb001 fm - National Institute of ... 22, 2007 12:51 RPS rpb001_fm vi CONTENTS Chapter 6 Phytoplankton diversity, biomass, and production S. G. Prabhu Matondkar,

June 22, 2007 12:51 RPS rpb001_fm

The Mandovi and ZuariEstuaries

Editors

Satish R. Shetye, M. Dileep Kumar, and D. Shankar

National Institute of Oceanography

Dona Paula, Goa, India

June 22, 2007 12:51 RPS rpb001_fm

Copyright © 2007 by National Institute of Oceanography, Goa, India.

ISBN 978–81–7525–872–3

This book is unpriced. Print copies available on request.Contact: Director, National Institute of Oceanography, Dona Paula,

Goa 403 004, India.

Cover picture: False colour satellite image of the Mandovi and Zuariestuaries constructed by assembling four scenes.

(Prepared by C. Pradnya Vishwas.)

Typeset by Research Publishing Services, Chennai 600 035, Indiaemail: [email protected]

Printed by Lotus Printers, Bangalore 560 044, Indiaemail: [email protected]

June 22, 2007 12:51 RPS rpb001_fm

Contents

Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Prologue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 1 The environment that conditions the Mandoviand Zuari estuariesS. R. Shetye, D. Shankar, S. Neetu, K. Suprit,G. S. Michael, and P. Chandramohan . . . . . . . . . . . . . . . . . . . . 3

Chapter 2 Tides and sea-level variabilityS. R. Shetye, I. Suresh, and D. Sundar . . . . . . . . . . . . . . . . . . . 29

Chapter 3 Numerical modelsA. S. Unnikrishnan and N. T. Manoj . . . . . . . . . . . . . . . . . . . . 39

Chapter 4 Mixing and intrusion of saltS. R. Shetye, G. S. Michael, and C. Pradnya Vishwas . . . . . . . . . 49

Chapter 5 Variability of nitrate and phosphateS. Sardessai and D. Sundar . . . . . . . . . . . . . . . . . . . . . . . . . . 59

v

June 22, 2007 12:51 RPS rpb001_fm

vi CONTENTS

Chapter 6 Phytoplankton diversity, biomass, and productionS. G. Prabhu Matondkar, Helga do R. Gomes, SushmaG. Parab, Suraksha Pednekar, and Joaquim I. Goes . . . . . . . . . . 67

Chapter 7 ZooplanktonR. Jyothibabu and N. V. Madhu . . . . . . . . . . . . . . . . . . . . . . . 83

Chapter 8 Benthic macrofaunaZ. A. Ansari, S. Sivadas, and B. S. Ingole . . . . . . . . . . . . . . . . 91

Chapter 9 Distribution of iron and manganeseAnalia M. Mesquita and Sujata Kaisary . . . . . . . . . . . . . . . . . 99

Chapter 10 Distribution of tributyltin (TBT) in the Mandovi estuaryNarayan B. Bhosle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Chapter 11 Sewage-pollution indicator bacteriaN. Ramaiah, V. Rodrigues, E. Alvares, C. Rodrigues,R. Baksh, S. Jayan, and C. Mohandass . . . . . . . . . . . . . . . . . . 115

Chapter 12 The khaznam of GoaS. N. de Sousa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Epilogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

June 22, 2007 12:51 RPS rpb001_fm

This book is dedicated to our colleague

Dr. Mahesh D. Zingde

on his 60th birthdayin appreciation of his service

to the study of the Indian estuarine and coastal regionsand his endeavours towards controlling its pollution

and ensuring its sustenance.

vii

June 22, 2007 12:51 RPS rpb001_fm

Acknowledgements

Every chapter in this book has benefited from support provided by governmentfunding agencies. Most notable amongst these has been the support received fromthe Department of Ocean Development (DOD), Government of India, New Delhi,which is now a part of the Ministry of Earth Sciences (MoES). In 1993, DODsupported a study that, for the first time, permitted measurement of tides at overa dozen locations throughout the Mandovi and Zuari estuaries. In 2002–2003, theIntegrated Coastal Monitoring and Management (ICMAM) Directorate of DODfunded a programme of observations that has provided critically important inputsto studies described in this book. In both the 1993 and the 2002–2003 programme,Dr. B. R. Subramanian, ICMAM Program Director, took special interest in helpingthe projects. We are grateful for his support. Dr. E. Desa, then Director of NIO,took special care to ensure that the observations planned under the project werecarried out without a hitch. We are thankful for his support and encouragement.

Prof. Karl Banse of the University of Washington read all the chapters in the shorttime available to him. His critical comments have helped improve the clarity ofseveral chapters. We are grateful to him.

In addition to the support from DOD and MoES, individual chapters have benefitedfrom assistance received from the Ministry of Shipping, Government of India, NewDelhi, and Public Works Department, Government of Goa, Panaji, Goa. K. Supritand N. T. Manoj acknowledge the Research Fellowships received from the Councilof Scientific and Industrial Research, New Delhi. India Meteorological Department(IMD) and Central Water Commission (CWC) provided the rainfall and river runoffdata. We acknowledge the role of GRASS GIS and Ferret in this study. Most of thefigures were drafted using the Generic Mapping Tools (GMT).

Many of our colleagues at NIO helped in many aspects related to putting together ofthis book. In particular, we are grateful to Christabelle Fernandes, Teja Dhawasker,

ix

June 22, 2007 12:51 RPS rpb001_fm

x ACKNOWLEDGEMENTS

Lavinia Gonsalves, Nasreen Shaikh, L. B. Naik, S. T. Yeshwant, Mr. Blasco Fernan-des, Sanjay Kankaonkar, Nilesh Parsekar, Karen Lobo, Xavita Vaz, Nisha Pires,Imran Mirza, and Marlene D’Cunha. Most of the figures in this book were draftedby G. S. Michael and K. Suprit, with support from D. Sundar and from Arun Mahaleand his colleagues (including S. Akerkar and K. G. Chitari) in the DTP section ofNIO. A book of this nature, with several contributing authors, goes through severalrevisions. The expertise and patience of Tresa Fernandes, who incorporated themany changes that were made to the original draft, is gratefully acknowledged. Wealso acknowledge the contribution of Dr. M. D. Rajagopal towards the logistics ofpreparing this book.

This book has enjoyed support from those working outside of NIO too. The chapteron khaznam has benefited from the help given by the staff of the Central Library(Rare Books Section), Panaji, and by Mr. Damodar Phadte, Chairman, CorlimKhazan Tenants’ Association, Corlim, Goa. The final copy-editing was done byHema Wesley of the Indian Academy of Sciences, Bangalore. She also designedthe layout of the book, and oversaw its production. We are indeed grateful.

Editors

June 22, 2007 12:51 RPS rpb001_fm

Map A: Map of the Mandovi and Zuari estuaries based on publications from theSurvey of India (1967), the Naval Hydrographic Office, India, and the Admiralty,U.K. The depth contours (m) are with respect to mean sea level which is 1.3mabove Spring Low Water at Mormugao (see Map C). To avoid clutter, the depthcontours are shown only near the mouths of the Mandovi and the Zuari. “T” ina circle shows location up to which tidal influence is felt in the rivers Kushavati,Uguem, Guloli, Khandepar, and Mandovi. Land elevation is shown in colour, thekey for which is given above the figure.

xi

June 22, 2007 12:51 RPS rpb001_fm

73°48'E 74°00'E 74°12'E

15°12'N

15°24'N

15°36'N

TT

T

T

12

3

4

5 6

7

8

9

10

11

12

13

MZ

M7 M6M5

MN

M4M3

M2 M1

Z5Z4

Z3

ZN

Z2 Z1TT

T

T

0 5 km

Map B: The locations where water level observations were carried out areshown with a star. The name of the location can be determined by the num-ber in a circle adjoining the star: (1) Verem; (2) Britona; (3) Akkada; (4) Valpoi;(5) Usgao; (6) Ganjem; (7) Dona Paula; (8) Cortalim; (9) Borim; (10) Sanvordem;(11) Sanguem; (12) Banastarim; and (13) Madkai. A solid circle with a station num-ber next to it shows a station where biogeochemical variables were measured. Thesestations are marked as: M1, M2, M3, MN, M5, M6, M7, and MZ in the Mandovi;M4 in the Cumbarjua Canal (see Map A); Z1, Z2, ZN, Z3, Z4, and Z5 in the Zuari;and, MZ near the mouths of the two estuaries. “T” in a circle shows location upto which tidal influence is felt in the rivers Kushavati, Uguem, Guloli, Khandepar,and Mandovi.

xii

June 22, 2007 12:51 RPS rpb001_fm

73°48'E 74°00'E 74°12'E

15°12'N

15°24'N

15°36'N

TT

T

T

5

4

13

11

10

1

2

3

6

7

8

9

15

14

12

0 5 km

Map C: The locations where sea level and salinity were measured during April andAugust 1993 are shown with a star. The names of locations can be found from thenumber inside a circle next to the star: (1) Mormugao; (2) Cortalim; (3) Loutulim;(4) Sanvordem; (5) Sanguem; (6) Aguada; (7) Penha de Franca; (8) Sarmanas;(9) Volvoi; (10) Sonarbaag; (11) Ganjem; (12) Banastarim; (13) Khandepar;(14) Sirsai; and (15) Mhapsa. “T” in a circle shows location up to which tidal influ-ence is felt in the rivers Kushavati, Uguem, Guloli, Kandepar, and Mandovi.

xiii

June 13, 2007 18:51 RPS rpb001_prologue

Prologue

The International Indian Ocean Expedition (IIOE) of the 1960s was the impetusfor oceanographic research in India. Indian veterans of IIOE found a new hometo continue their research at the National Institute of Oceanography (NIO), whichwas founded on 1 January 1966 in newly liberated Goa.

The main campus of NIO is located on a promontory between the mouths of twoestuaries, the Mandovi and the Zuari. Researchers at NIO soon began to studythe estuaries. In 1977–1978, the Government of Goa funded a study of the twoestuaries, often described as lifelines of the state because of their extensive usefor fisheries, agriculture, transportation, dumping of waste, etc. The two estuarieshave since continued to attract scientific curiosity over the years, thanks in largemeasure to support from national and local funding agencies. As a consequence,our understanding of the Mandovi and Zuari estuaries has grown, making themalmost certainly, the best studied estuaries of India (Qasim 2003).

Being a national institute dedicated to the study of ocean sciences, NIO owes it tothe country to disseminate knowledge generated by its scientists for the bettermentof society. It also bears the responsibility of creating awareness among our peopleabout the environmental consequences of human activity, and inspiring a newgeneration of researchers to undertake research in marine systems. To encourageresearch on other estuaries of India, often home to large human settlements, it wasfelt that a summary of what is known about the Mandovi and Zuari estuaries wouldbe most helpful. It was also felt that the summary be written in a manner that wouldappeal to both professionals and amateurs, especially environmental managers andthose with special concerns for the environment. This book has been put together tofulfill these and other needs, particularly in the light of the immense environmentalpressures our country will face as its economy continues to grow.

1

June 13, 2007 18:51 RPS rpb001_prologue

2 PROLOGUE

The first chapter of the book describes the environment in which the Mandovi andthe Zuari exist. Covered in this chapter are the geometry of the terrain over whichthe estuaries flow, the rainfall and runoff they experience, the winds, etc. Chapters 2,3, and 4 provide an overview of the basic physics of the two estuaries: nature oftides, efforts made so far to simulate them in numerical models, and characteristicsof stratification and mixing. These are followed by chapters that examine the fun-damentals of biology and chemistry of the estuaries. A distinct characteristic of theestuaries is the annual cycle that they exhibit. Chapter 5 describes the annual cycleof nutrient concentration, chapter 6 of phytoplankton, chapter 7 of zooplankton,and chapter 8 of benthic macrofauna. A motivation for estuarine research aroundthe world stems from the desire to keep them healthy and free of pollutants andhuman interference. Chapters 9, 10, and 11 describe three pollutants that are ofspecial relevance to the Mandovi and the Zuari. Chapter 9 describes the observedconcentrations of iron and manganese, whose source lies in the extensive iron andmanganese mines of Goa. Chapter 10 discusses the distribution of TBT, which iswidely used in building of ships and in their maintenance. The channels of theMandovi and the Zuari have been used extensively in the past for transportingpeople and goods. Over the last five decades they have been used to transport ironand manganese ore. The estuaries have also been used to discharge sewage, mostof it raw. Chapter 11 discusses the distribution of bacteria that are linked to sewagedisposal. Chapter 12 discusses characteristics of wetlands found along the banks ofthe two estuaries. Known locally as khaznam, these wetlands have sustained a specialform of agriculture and fishery for almost ten thousand years. The chapter describeshow khaznam work and their relationship with the estuaries. The book concludeswith a chapter that discusses future directions for research on the two estuaries.

It is our fond hope that the contents of this book will encourage further studies ofthe Mandovi and Zuari, and of our country’s several other estuaries.

We dedicate this book to our colleague, Dr. Mahesh Zingde, on his 60th birthday.Studies on the estuarine and coastal regions of India have been the hallmark of hisprofessional career. Dr. Zingde’s contributions have greatly improved our knowl-edge of the structure and functioning of the estuaries in the states of Maharashtraand Gujarat. For several years, Dr. Zingde has served as consultant to state and cen-tral governments and to the industry. In this capacity, he has been of immense helpto the industry and to several diverse institutions, helping them to minimize pollu-tion and to ensure the health of riverine and coastal ecosystems. As he turns sixty,we express our gratitude to Dr. Zingde for his contributions and for his mentorshipover the years.

Satish R. Shetye, M. Dileep Kumar, and D. Shankar(Editors)

June 2007NIO, Dona Paula, Goa, India.

June 13, 2007 19:40 RPS rpb001ch01

1The environment thatconditions the Mandoviand Zuari estuaries

S. R. Shetye, D. Shankar, S. Neetu, K. Suprit,G. S. Michael, and P. Chandramohan∗National Institute of Oceanography, Dona Paula, Goa 403 004, India.∗Indomer Coastal Hydraulics (P) Ltd., Vadapalani, Chennai 600 026, India.

1.1 INTRODUCTION

The Deccan plateau, which makes up the greater part of the Indian Peninsula,extends approximately from 10◦N to 20◦N. As noted by Wadia (1975), the plateauis a central tableland, with an average elevation of 600m above sea level, and issurrounded on all sides by hill ranges. To the west of the plateau and close tothe western coastline of the peninsula are the Sahyadris, or Western Ghats. Theyextend southward from about 20◦N to the Malabar Coast, where they merge intothe Nilgiri Mountains (figure 1.1). South of the Palghat Gap, the Sahyadris extendthrough the Anaimalai Hills to the extreme south of the Indian Peninsula. TheSahyadris are called “ghats” because they are steep and terraced hills that runparallel to the coastline. The mean height of the hills is about 900m, and they aremade of horizontally bedded lavas that on weathering have taken a characteristic“landing stair” shape (Wadia 1975).

3

June 13, 2007 19:40 RPS rpb001ch01

4 S. R. SHETYE ET AL.

Figure 1.1 Relief of the Indian subcontinent and its surroundings.

Between the west coast of India and the start of the rise in elevation due to theSahyadris, there often occurs a coastal plain interspersed by small hills and othertopographic features. As depicted in the schematic in figure 1.2, the typical cross-shore extent of the plain is a few tens of kilometres. With average annual rainfallin excess of 2m, the coastal plain and the western slopes of the Sahyadris form oneof the two rainiest areas of India, the other being the mountain slopes in northeasternIndia. Most of the rain occurs during the four months, June–September, of theIndian summer monsoon. Runoff due to the rainfall on the ridge and western slopesof the Sahyadris is carried westward down the slopes by a number of streams,rivulets, rivers, and their tributaries. The runoff is high during the four months ofthe monsoon, and is particularly high when the monsoon is in an active phase.After withdrawal of the monsoon, the runoff decreases rapidly, and by Novemberreaches negligible levels.

June 13, 2007 19:40 RPS rpb001ch01

THE ENVIRONMENT THAT CONDITIONS THE MANDOVI AND ZUARI 5

SahyadrisCoastalPlain

10 km

Arabian Sea

Ele

vatio

n (m

)

800

400

SeaLevel

Figure 1.2 A schematic of the typical variation in topography along the cross-shore direc-tion on the west coast of India. Many estuarine channels that experience tides meander onthe coastal plain and join the Arabian Sea. At the upstream end of these channels are riversthat arise in the Sahyadris.

The coastal plain often has estuarine networks with a main meandering channel anda number of smaller channels that are connected to it. The channels are usually ofthe converging kind, i.e., the width of a typical channel decreases from mouth tohead. The estuarine network forms an important conduit to carry to the ArabianSea the runoff that comes down the slopes of the Sahyadris, and that which arisesfrom the rainfall on the plain. Hence, the freshwater content of the estuarine watersis high during the monsoon. During this season, the water level in the upstream partof the estuarine channels is controlled by the amount of runoff in the channel. Thewater level at the downstream end of the estuarine channel is controlled throughoutthe year by the astronomical tide at the coast. After withdrawal of the monsoon,the flow imposed by the tide at the mouth of the channel becomes the sole drivingmechanism for transport in the estuarine network. Hence, oceanographic processesin these networks differ significantly between the wet monsoon season, when runoffis high, and the dry season, when runoff is negligible and the tide dominates circu-lation and mixing in the estuaries. Such estuaries are often referred to as monsoonalestuaries.

Such monsoonal estuaries occur along the entire west coast of India. They areused for dumping of domestic and industrial waste, fishing, transport of goodsand people, and recreation. The banks of the estuarine channels have traditionallybeen preferred locations for human settlement and setting up of industries. Withrapid increase in population and in industrialization during the last few decades,the estuarine channels have come under increasing stress due to anthropogenicactivities. It is suspected that these activities could be causing irreversible changesin the chemistry and biology of the estuaries. As a result, issues related to the healthof the estuarine networks and their ability to absorb additional impacts has gainedimportance. Addressing these issues requires understanding of how these estuaries

June 13, 2007 19:40 RPS rpb001ch01


work, i.e., how the physics, chemistry, biology, and geology have in the past workedtogether to keep the estuary healthy, and how anthropogenic influence could bedisturbing the pristine system.

The Mandovi and Zuari estuaries in the state of Goa, central west coast of India, aretypical of the west coast estuaries and are also the best studied of the estuaries alongthe coast. This chapter describes the environment that is external to the estuaries, butinfluences the working of the estuaries. We begin in section 1.2 with an overview ofthe geometry of the channels of the estuaries. Section 1.3 examines runoff in the estu-aries. Winds observed at the mouths of the estuaries are discussed in section 1.4, andwaves near the mouths of the two estuaries are described in section 1.5. The chapterconcludes with a discussion on special features that characterize the estuaries.

1.2 GEOMETRY OF THE MANDOVI AND ZUARI ESTUARINE NETWORK

Map A shows the estuarine network of the Mandovi and the Zuari. The map alsoshows the location up to which tides reach in the two estuaries. The Mandovi iswider, the width being approximately 4km at the Aguada Bay. The 4 km longstretch of the bay is, on average, marginally deeper than the rest of the estuarinechannel, the average depth in the bay being about 5m. The Sinquerim River joinsthe bay. The 6 km long stretch immediately upstream of the bay is, on average,750m wide and 5m deep, i.e., the channel narrows considerably from the bay tothis stretch of the estuary. The Mhapsa River joins the Mandovi at the upstream endof this stretch. Farther upstream, Diwar Island, which is approximately 11 km long,bifurcates the Mandovi into two channels. Before rejoining at the upstream endof the island, the two channels lead into an extensive network of narrow channelsin a marshy area. The Cumbarjua canal joins the Mandovi about 4 km upstreamof Diwar Island. The 30 km stretch of the main channel of the Mandovi, from theeastern edge of Diwar Island to Ganjem, gets progressively narrower and shallowerin the upstream direction. Rivers Dicholi, Valvat, Kudnem, and Khandepar join theMandovi along this stretch. The Khandepar is the largest of the four rivers and isfed by the Dudhsagar River (not shown in Map A). At the upstream end of theMandovi, near Ganjem, lies the Mhadei River, the major supplier of runoff to theMandovi during the monsoon. A smaller river, Ragda, also joins the Mandovi nearGanjem. The Mhapsa River, which joins the Mandovi at Penha de France, has at itsupstream end the rivers Asnoda and Moide. The Mhapsa River and its tributariesare joined by a large network of small rivulets that often flow through marshy areas.

The 10 km stretch upstream from the mouth of the Zuari, known as the MormugaoBay, is approximately 5 km wide and 5m deep (Map A). At the upstream end ofthe bay, the channel narrows to a width less than 1km. The 30 km long channelfrom Cortalim to Sanvordem narrows progressively. It is less than 50m wide atSanvordem. The Kushavati River (also known as Paroda) joins the Zuari 4 kmdownstream of Sanvordem. The 10 km long stretch of the main channel from

June 13, 2007 19:40 RPS rpb001ch01


15

10

5

0

5D

epth

(m

)

0 10 20 30 40 50

(a) Mandovi

15

10

5

0

5

Dep

th (

m)

0 10 20 30 40 50

(b) Zuari

0.00

0.01

0.02

0.03

0.04

Are

a (k

m2 )

0 10 20 30 40 50Distance from the mouth (km)

(c)MandoviZuari

Figure 1.3 Depth (m) with respect to mean sea level along a line in the middle of the mainchannel of (a) Mandovi and (b) Zuari. The horizontal scale gives distance (km) from themouth of the two estuaries. The depth was computed by adding 1m to the depths given incharts published by the Ministry of Shipping and Transport, Government of India. Thesecharts are available only for those parts of the estuarine network used by iron-ore barges.Mean sea level at Mormugao is 1.3m above Spring Low Water. At the locations wheredepth is zero or positive, mudflats bifurcate the main channel and the flow is along the sidesof the mudflats. (c) Cross-sectional area (km2), with respect to mean sea level, of the mainchannels of the Mandovi and Zuari. (After Shetye et al. 1995.)

June 13, 2007 19:40 RPS rpb001ch01


Sanvordem to Sanguem is also known as the Sanguem River. At Sanguem, tworivers —- Uguem and Guloli —- join to form the Sanguem River. During the dryseason, the effect of tides is felt at Sanguem, and possibly as far as 4 km upstreamin the Uguem River and 1 km in the Guloli. In the Kushavati, the tidal influenceis present up to about 8 km upstream of the point where it joins the Zuari. Duringthe wet season the flow at Sanguem is dominated by river runoff.

The depth of the main channels of the Mandovi and Zuari varies considerably withlocation. This is brought out in figure 1.3, which shows the variation of the depthalong a line in the middle of the main channels of the Mandovi and Zuari. The cross-sectional areas of the channels of the two estuaries drop rapidly in the upstreamdirection (figure 1.3c) primarily owing to decrease in channel width. Such channelshave often been described as ‘strongly-convergent’ (Friedrichs and Aubrey 1994).The convergence has important implications for dynamics of tides in the estuarinechannels.

Each of the two estuarine channels can be divided into two parts if geometry aloneis considered. The first of these is distinctly wider and lies at the mouth. In thecase of the Mandovi, this part is known as the Aguada Bay; Mormugao Bay isthe corresponding part of the Zuari. One side of each bay is connected to thesea. The length and width of the two bays being comparable, the estuarine pro-cesses of mixing and advection here are expected to show more two-dimensionalbehaviour (processes along the length are as important as those perpendicular toit) than in the rest of the estuary, which in each of the two estuaries consists ofa 30–40 km long converging channel. The width of the channel is well below akilometre even at the widest end. Physical and other processes here are expectedto exhibit the distinctly one-dimensional character that is commonly observed inestuaries.

1.3 RUNOFF

Figure 1.4 elucidates the special feature of the west coast of India that was referredto earlier: the west coast is one of the two rainiest regions of the country. Thisspecial feature is a consequence of the presence of Sahyadris: this region of highrainfall is due to orography. The moisture-laden westerlies of the summer monsoonget lifted up as they cross the Sahyadris. This leads to their cooling, and conden-sation of moisture and rainfall follow. The rainfall on the slopes of the Sahyadrisleads to runoff in the rivers that join the Mandovi and Zuari estuaries. Virtuallyall of the freshwater influx into the two estuaries occurs during the summer mon-soon ( June–September), when most of the rainfall over the catchment area of thenetwork occurs. The nature of temporal variability of the freshwater discharge inthe main channels of the two rivers can be appreciated from the estimates of dis-charge (hydrographs) in the main rivers that bring freshwater to the two estuaries(figure 1.5). The monsoon season has distinct periods of high and low rainfall. These

June 13, 2007 19:40 RPS rpb001ch01


Figure 1.4 Geographical distribution of the normal monsoon ( June to September) rainfall(cm) over India. (After Rao 1976b.)

are often called active (higher than normal rainfall) and weak (lower than normal)phases of the monsoon. When a particularly active spell occurs, the entire catch-ment area of the Mandovi and Zuari receives copious rainfall, leading to high runoffin all the rivers. It is for this reason that some of the peaks seen in figure 1.5 occursimultaneously.

Rainfall over these basins, and hence runoff into the estuaries, being highly seasonal,the character of the estuary (salinity and nitrate, for example) varies markedly from

June 13, 2007 19:40 RPS rpb001ch01


0

200

400

600

800

Kushavati(d)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0

200

400

600

800

Sanguem(c)

0

200

400

600

800

Khandepar(b)

Dis

char

ge (

m3 s

−1)

0

200

400

600

800

Mandovi(a)

Figure 1.5 Daily discharge (m3s−1) hydrograph of (a) Mandovi River at Ganjem, (b) Khan-depar River, (c) Sanguem River, (d) Kushavati River for 1978. Note that the discharge datafor the rivers were not available for all days. The resulting ‘data gaps’ were removed to plotsmooth curves. (This is equivalent to using linearly interpolated values for filling missingdata or ‘data gaps’.)

the wet to the dry season. This difference makes it important to determine the spa-tial and temporal variability of the runoff received by the two estuaries. Empiricaldata on runoff, such as that shown in figure 1.5, are too sparse to construct a com-prehensive description of the runoff. Hence, we use a model to make this estimate.We begin by describing the variability of rainfall.

June 13, 2007 19:40 RPS rpb001ch01


Figure 1.6 The region of interest for rainfall interpolation and hydrological modelling. Thetopography of the region (along with 0 and 500m contours) as in the GLOBE (GLOBE2004) DEM (inset topography from ETOPO2 (ETOPO2 2007)). The Mandovi and Zuari(all rivers digitized from Survey of India maps) are the two major rivers of Goa (borderoverlaid on the map). The Mandovi drains into the Arabian Sea near Panaji. The Mandovibasin (black curve) has two discharge gauging stations, at Ganjem on the Mhadei andat Kulem on its major tributary, the Khandepar. The outer and inner rectangles denotethe interpolation and THMB model domains; rainfall interpolation was performed over alarger domain in order to include more rain gauges to map the variation. The region hastwo distinct topographical and climatic features: to the west lies a coastal plain with heavyrainfall (windward side), and to the east lies a plateau with less rainfall (leeward side). Thewhite jagged line is the ridge separating the windward and leeward sides of the Sahyadris.Stations on the windward (leeward) sides are marked in black (white). (After Suprit andShankar 2007.)

1.3.1 Rainfall variability

Rainfall over the Mandovi and Zuari basins (figure 1.6) varies considerably in bothtime and space. The seasonal cycle of rainfall, with a peak during the summermonsoon ( June–September), is obvious at the rain-gauge stations in these basins(figure 1.7). There is also considerable variation of seasonal (or monthly) rainfall inspace, with rainfall increasing as one proceeds from the coast towards the Sahyadris.

June 13, 2007 19:40 RPS rpb001ch01


0

200

400

600

800

1000

1200

1400

1600

Kulem

Ganjem

Rai

nfal

l (cm

) D

isch

arge

(M

cum

)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0

50

100

150

200

250

300

350 Asoga

Gavali

Valpoi

Mhapsa

Panaji

Figure 1.7 Climatology of monthly rainfall (cm; averages computed over 1981–1997) at thefive rain-gauge stations in the Mandovi basin (Panaji, Mhapsa, Valpoi, Gavali, and Asoga)and the climatology of river runoff (Mcum; averages computed over 1981–1998) at thetwo runoff gauging stations in the Mandovi basin (Ganjem and Kulem). The vertical barsindicate the standard deviation of the monthly rainfall and runoff; the height of the bars isa measure of the inter-annual variability.

There exist only five rain gauges in the Mandovi basin, which has an area of approx-imately 1895 km2 (Suprit and Shankar 2007). This sparseness of rain gauges makes itdifficult to map the variability in space, the problem being most acute on the slopesof the Sahyadris, where the rainfall increases rapidly as against that encountered inthe flatter plains to the west, and on the leeward (eastern) side, it being difficult tocapture the rapid decline in rainfall as one crosses the ghats from west to east. Theannual rainfall varies from an average of 286 cm at Panaji on the coast to an averageof 661 cm at Gavali on the windward slope of the Sahyadris. At Valpoi, which lies atthe foothills of the Sahyadris, the average annual rainfall is 413 cm. On the leewardside, at Asoga, the average rainfall declines to 161 cm. The distance between Valpoiand Gavali and that between Gavali and Asoga is about 10 km (figure 1.6; also seethe schematic in figure 1.2), showing just how rapid the spatial variation in rainfall

June 13, 2007 19:40 RPS rpb001ch01


7348' E 7400' E 7412' E 7424' E

1512' N

1524' N

1536' N

1

2 34

5

50

100

150

200

250

300

350cm

Figure 1.8 Map of rainfall (cm) during July 1992. This rainfall map was created by interpo-lating the monthly rainfall separately for the windward and leeward sides and then mergingthe two maps. The ridge separating the windward and leeward sides is shown as the blackcurve. The grey curve is the Mandovi basin. The five rain-gauge stations in the basin (1–5) arethe ones for which daily climatology is plotted (top–bottom in rainfall panels) in figure 1.9.(After Suprit and Shankar 2007.)

owing to the Sahyadris is. Hence, given the paucity of data to resolve this largevariation in rainfall, Suprit and Shankar (2007) used a multivariate interpolationscheme that incorporated elevation as the third variable (in addition to the two hor-izontal co-ordinates) to map the rainfall separately for the windward and leewardsides. The ridge separating the windward and leeward sides was prescribed a priori.The two maps were then merged to produce a spatial map of the monthly rainfallover the region. One such map, for July 1992 (figure 1.8), shows that mapping tech-nique captures the gradual increase in rainfall as one moves inland from the coast,the increase becoming much more rapid as one climbs the Sahyadris; the rainfalldecreases rapidly on the lee side of the ghats. This picture of spatial variability ofmonthly rainfall is typical of the peak monsoon months of June–September, whenthe moisture-laden winds blow from the Arabian Sea towards the Indian west coast.

There is considerable inter-annual variability in annual rainfall (figure 1.7), thestandard deviation about the mean rainfall at Panaji (286 cm), Mhapsa (303 cm),Valpoi (413 cm), Gavali (661 cm), and Asoga (161 cm) being 45, 58, 56, 173, and50 cm respectively. This inter-annual variability in rainfall is reflected in the runoff.

There is also considerable variation in rainfall from day-to-day even during thepeak monsoon months of July and August. For example, during July 1992, the totalmonthly rainfall at Gavali was 271 cm, but ∼ 80% of this rainfall occurred over just11 days (figure 1.9). Rainfall during the summer monsoon consists of several bursts

June 13, 2007 19:40 RPS rpb001ch01


0

50

100

150

200Ganjem

Dis

char

ge

(Mcu

m d

ay-1

)

Jun Jul Aug Sep

01020304050

Kulem

0

100

200

300 Asoga

0

100

200

300 Gavali

0

100

200

300 Valpoi

Rai

n (

mm

day

-1) 0

100

200

300 Mhapsa

0

100

200

300 Panaji

Figure 1.9 Daily rainfall (black; mmday−1) at Panaji, Mhapsa, Valpoi, Gavali, and Asogaand daily runoff (black; Mcumday−1) at Ganjem and Kulem during June–September 1992.The horizontal line in the panels for runoff indicates the volume of the Mandovi channel.The lower line indicates the volume of the channel; alone; the upper line indicates thevolume of the entire estuary, including the Aguada Bay. Overlaid on the daily rainfall andrunoff are the 5-day (red) and 30-day (blue) moving averages.

June 13, 2007 19:40 RPS rpb001ch01


during active spells of the monsoon, there being weak spells or even monsoonbreaks in between them (Rao 1976b). This high-frequency variability in rainfall,evident at all rain-gauge stations, is reflected in high-frequency variability in therunoff (figure 1.9).

1.3.2 The observed temporal variability of runoff in the Mandovi

River runoff is gauged regularly at two locations in the Mandovi basin. One ofthem, Ganjem, is located on the Mhadei River about 50 km from the mouth of theMandovi; the other gauging station is at Kulem, which is located on the Khandepar,the biggest of the Mandovi’s tributaries (figure 1.6). The average annual runoff atGanjem and Kulem are 3400Mcum and 502Mcum, the standard deviation aboutthe mean being 648Mcum and 91Mcum respectively. (1Mcum is 1 million cubicmetres.)

The mean monthly runoff at the gauging stations integrates over a few rainfall burststhat result in a huge volume of freshwater being flushed through the estuarine chan-nels within a few days. That the monsoon rainfall is characterized by bursts impliesthat there is considerable high-frequency variability in runoff; in other words, a largefraction of the peak runoff during July–August can be expected to occur within afew days. During 1992, for example, the daily runoff at Ganjem had a peak of168Mcum on 26 July (figure 1.9).

Over the rest of the year, both before and after the summer-monsoon rains, therunoff into the Mandovi is negligible. The total average runoff during November–May is 117Mcum, much less than the 150Mcum peaks in daily runoff (figure 1.9)during the summer monsoon.

1.3.3 Spatial variability of runoff

The gauging stations at Ganjem on the Mhadei and at Kulem on the Khandeparprovide information on the amount of runoff in two of the key rivers in the Mandovibasin. The catchment area of the river at Ganjem is 872 km2, which is 46% of theMandovi basin (Suprit and Shankar 2007). The catchment area of the Khandeparat Kulem is 97 km2, implying that the effect of rainfall over the remaining 49%of the Mandovi basin is not measured. In particular, the data available do not tellus what the runoff is at Panaji at the mouth of the Mandovi, where the estuarinechannel broadens into the Aguada Bay. Hence, it is necessary to use a numericalmodel to make an estimate of this runoff. Shankar et al. (2004) assembled a frame-work for modelling river discharge under the sparse-data conditions prevalent inIndia. At the core of this framework was a hydrological model called THMB (Ter-restrial Hydrologic Modelling with Biogeochemistry; Coe 2000), which, given therainfall and evapo-transpiration at each 1 km × 1 km grid cell in the basin, routes

June 13, 2007 19:40 RPS rpb001ch01


Figure 1.10 Schematic representation of THMB. The figure shows a THMB grid cell andthe fluxes into and out of the three reservoirs (surface runoff, sub-surface runoff, and riverwater) associated with each cell. The dashed line indicates one THMB grid cell. (P–E)denotes the runoff, the difference between rainfall and evapo-transpiration; α determinesthe fraction of runoff (surface runoff RS) that goes into the surface water reservoir (WS),TS being the time scale over which water flows out of the reservoir; RD is the sub-surfacerunoff; WD denotes the sub-surface water reservoir, TD being the time scale over whichwater flows out of this reservoir. Water flows from both surface and sub-surface reservoirsinto the river-water reservoir (WR), from which water flows out of the cell to a downstreamcell over a timescale TR.

∑Fin is the total inflow from all upstream cells into the river water

reservoir.

the local grid runoff through a river channel to the sea. (This model was earliercalled HYDRA: Hydrological Routing Algorithm.) In other words, THMB needsat each grid point the surface runoff, the sub-surface runoff, and the flow into itfrom upstream cells. Each cell, in turn, feeds into one downstream cell, therebycreating an interlocking network of cells that together constitute the model basin.(See figure 1.10 for a schematic representation of THMB.) The flow out of the cellcorresponding to the gauging station is the model runoff. It is to create the rainfallforcing for THMB that Suprit and Shankar (2007) had to map the spatial vari-ability of rainfall. Their simulated runoff matches the observed runoff at Ganjem.Hence, since the model provides an estimate of the runoff at each grid cell withinthe basin, it becomes a viable source of information on the spatial variability ofrunoff.

Note that the model river channel terminates at the point where the height fallsbelow mean sea level in the GLOBE Digital Elevation Model (DEM) used forrouting the river. For the Mandovi, this point where the river ends is in the vicinityof Panaji. The rest of the Mandovi basin, which consists of the Aguada Bay, doesnot form part of the river-runoff computations in THMB because the bay’s bottomis below mean sea level and therefore forms part of the sea.

June 13, 2007 19:40 RPS rpb001ch01


1000

2000

3000

4000

5000

6000

7000

8000D

isch

arge

(M

cum

)Simulated Observed Panaji (Sim.)

Ganjem

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

200

300

400

500

600

700

800

Dis

char

ge (

Mcu

m)

Simulated Observed

Kulem

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

Figure 1.11 Simulated discharges at Ganjem and Panaji (top panel) and Kulem (bottompanel). The observed discharge is plotted for comparison. The shaded area is a measure ofthe variability in the observed discharge; the shading is done for one standard deviationon either side of the observed discharge. The thickness of simulated discharge curves atGanjem and Kulem show the impact of evapo-transpiration as it represents simulationsobtained by reducing and increasing the evapo-transpiration by one standard deviation.Note the negligible impact of evapo-transpiration and that the error in simulated dischargeis within the natural variability of the system (except in 1984 and 1988, when the blackcurve falls just outside the shaded region). (After Suprit and Shankar 2007.)

Runoff in the Mandovi increases almost two-fold from Ganjem to Panaji (fig-ure 1.11) (Suprit and Shankar 2007). A large fraction of this increase comes fromthe tributaries Khandepar (about 45%), Valvat (about 25%; includes Dicholi andKudnem rivers), and the Mhapsa (about 14%; includes Moide and Asnoda rivers);the balance 16% is direct flow into the estuary from the land adjoining it (figure 1.12).Note that for the reason given above, this runoff does not include that from the Sin-querim River, which flows directly into the Aguada Bay.

June 13, 2007 19:40 RPS rpb001ch01


0

1000

2000

3000

4000

5000

6000

7000

8000S

imul

ated

Dis

char

ge (

Mcu

m)

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

Mhapsa RValvat R

Panaji Ganjem

Khandepar R

Figure 1.12 Bar-chart showing the runoff into the Mandovi estuary from Ganjem (on theright) to Panaji (on the left) for 1992 as a function of distance (abscissa is the number of gridcells from Panaji to Ganjem) along the main channel of river. Runoff in the channel increasesalmost two-fold from Ganjem to Panaji, most of this increase coming from the runoff fromthe Khandepar, Valvat (including Dicholi and Kudnem), and Mhapsa (including Moide andAsnoda) rivers. The contributions of these tributaries are shown in black.

1.3.4 Runoff in the Zuari

Numerical modelling is also the only viable tool for estimating the runoff in theZuari, there being no runoff gauging stations in this basin. Preliminary results fromthe numerical model of Suprit and Shankar (2007) suggest that the average (over1981–1998) virgin (in the absence of control structures like the Selaulim Dam)annual runoff at the mouth of the Zuari near Cortalim, where it opens into theMormugao Bay, is about 2190Mcum. Note that the Zuari River in THMB ends atCortalim and excludes the Mormugao Bay. The runoff just upstream of Selaulimis 491Mcum, which matches the 75% annual dependable flow according to theobservations (475Mcum) reported in a public brochure by the Water ResourcesDepartment of the government of Goa; these measurements upstream of Selaulimwere made for a few years before the dam was constructed.

1.3.5 Implications of dams and channel geometry to flushing

The model simulations described above do not account for control structures likedams, and estimate what is called the “virgin flow” of the river. The only functional

June 13, 2007 19:40 RPS rpb001ch01


dam on the Mandovi is the Anjunem Dam on the Valvat or Nanda River, whichjoins the Mandovi downstream of Ganjem. The storage capacity of the dam is45Mcum, less than 1% of the runoff estimated at Panaji. Note that this dam has noimpact on the runoff at Ganjem.

On the Zuari, there exists a much bigger dam, the Selaulim Dam. With a storagecapacity of about 227Mcum (about 10% of the Zuari’s virgin flow at its mouth nearCortalim), this dam has a significant impact on the flow of the Zuari. The impact ofsuch control structures is more pronounced after the summer monsoon. The errorin ignoring them is comparable to the error in the observed runoff (measurementerror) or the error in the simulated runoff, both of which are ∼15% (Suprit andShankar 2007). After the summer monsoon, the natural runoff declines, but wateris released from the dam in a controlled manner, implying a greater runoff in thedry months of January–April than would exist without the dam.

Since the model river terminates at the point where the Mandovi and Zuari estuarieswiden into the Aguada and Mormugao bays, the rainfall over these bays, or thedirect flow into them from other smaller rivers like the Sinquerim, is not included inthe runoff estimates reported here. The rainfall over the bays, however, is expectedto be mixed differently than the runoff from the main river channel flowing intoit because the former represent a forcing at the surface and the latter represent avolume flux at the boundary.

The runoff in the Mandovi and Zuari estuaries is much greater than the volume ofthe river channel. The volume of the Mandovi (excluding the Aguada Bay) is about70Mcum, which is less than half the typical daily peak runoff (about 150Mcum)at Ganjem. The daily runoff at Panaji during a summer-monsoon burst would beeven greater. Hence, in an average year, the volume of freshwater flowing throughthe Mandovi exceeds the volume of the estuary by a factor of about 20 if usingthe Ganjem runoff and 40 if using the Panaji runoff. Most —- over 95% —- of thisfreshwater flux occurs during June–October, implying that the water in the riverchannel (till the Aguada Bay) is flushed out and renewed several times over duringthe summer monsoon. During the 26 July 1992 runoff burst of 168Mcum (measuredat Ganjem) (figure 1.9), the volume of freshwater flowing through the river was morethan twice its volume. Note that this burst of 168Mcum matches the 160Mcumvolume of the Mandovi when the Aguada Bay is included. Similar behaviour can beexpected in the Zuari, which has a smaller volume (less than 50Mcum) till Cortalim(matching its smaller catchment area in comparison with the Mandovi). Inclusionof the Mormugao Bay increases the volume of the Zuari to over 270 Mcum, but therunoff over a few days during July–August would still flush the Zuari completely.

Such flushing of the Mandovi and Zuari occurs during the summer monsoon everyyear. Physically, the implication is that the channel of the estuary is completelyflushed with freshwater during just a few days of the summer monsoon. Suchepisodes are expected to turn the estuarine water fresh from head to mouth.

Over the rest of the year, however, such flushing does not occur over the entirechannel because the runoff decreases considerably. At Ganjem on the Mandovi,

June 13, 2007 19:40 RPS rpb001ch01


for example, the average runoff during December–May is about 18Mcum, whichis much less than the estuary’s volume. For some distance from the head of theestuary, however, even this meagre runoff will tend to freshen the channel (over5–10 km), implying that salinity should decrease towards the head of the river evenduring the lean months.

1.4 WINDS

In the Mandovi and Zuari estuaries, the impact of winds is expected to be feltprimarily either through vertical mixing that waves generated on the surface wouldinduce or through the circulation set up owing to the stress exerted by winds.Neither of these impacts has so far been shown to be significant. For example, it isnot clear how the turbulence generated in the water column by tidal flow compareswith that generated by winds. Nonetheless, winds form an important environmentalvariable for any marine system. It is important to examine the annual cycle of windsto understand the environment in which the two estuaries exist.

The winds reported here were measured (for one year during 1 June 1996 to 31May 1997) near the mouths of the two estuaries; the measurements were made onthe campus of the National Institute of Oceanography at Dona Paula (see Map A)using an anemometer located on the terrace of the main building of the institute. Theanemometer was located at a height of about 50m above sea level. The recordeddata were ten-minute vector averages of wind speed (figure 1.13) and direction(figure 1.14). The data show striking differences between the summer monsoon andthe rest of the year. During the summer monsoon, winds are stronger and do notalways exhibit the distinct diurnal variation in wind speed seen after the monsoonwithdraws.

As seen in figure 1.13, on 8 June 1996 there was a distinct increase in wind speedheralding arrival of the monsoon. Panaji received the first rains of the season on thatday. The maximum wind speed recorded on that day was well over twice the dailymaximum speed during the first week of June. Also, the daily diurnal cycle of windspeed, with maximal wind around noon or in the afternoon, which was observedduring the first week of June, was absent on 8 June. In fact, the diurnal cycle washardly noticeable on most days from the onset of monsoon till about the middle ofAugust. During this time there was a period, approximately from mid-June to mid-July, when the wind was consistently from one direction, from about 330 degrees,and wind speed was roughly between 4 and 7ms−1. Climatological monthly-mean data generally report that winds at this time are westerly or southwesterly(see, for example, Shetye et al. 1985). The northwesterly or northerly angle ofwind direction observed in figure 1.14 may well be a local feature arising fromlocal terrain.

The strength of the winds decreases after the withdrawal of the summer monsoon.After the withdrawal and until the onset of the next monsoon, i.e., from October to

June 13, 2007 19:40 RPS rpb001ch01


05

1015

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32May 1997

05

1015

Apr 1997

05

1015

Mar 1997

05

1015

Feb 1997

05

1015

Jan 1997

05

1015

Dec 1996

05

1015

Nov 1996

05

1015

Oct 1996

05

1015

Sep 1996

05

1015

Aug 1996

05

1015

Jul 1996

05

1015

Jun 1996

Figure 1.13 Wind speed (ms−1) during 1 June 1996 to 31 May 1997. The horizontal axisgives date of the month shown at the bottom of each panel. Vertical axis gives wind speed.

the end of May or early June, the wind field is dominated by the sea breeze. Thewinds are strongest in the afternoon, when the cross-shore component is from seatowards land. During early morning, the winds are weaker and the cross-shore com-ponent is from land to sea. Though these characteristics do repeat from day to day,details differ. Figure 1.15 shows the hodograph, taken from Neetu et al. (2006), thatgives the average daily cycle of winds observed during 1–15 April 1997. Climatologi-cal data suggest that during April the large-scale wind field in the region is weak, and

June 13, 2007 19:40 RPS rpb001ch01


0

180

360

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32May 1997

0

180

360

Apr 1997

0

180

360

Mar 1997

0

180

360

Feb 1997

0

180

360

Jan 1997

0

180

360

Dec 1996

0

180

360

Nov 1996

0

180

360

Oct 1996

0

180

360

Sep 1996

0

180

360

Aug 1996

0

180

360

Jul 1996

0

180

360

Jun 1996

Figure 1.14 Wind direction (in degrees clockwise from north; north is 0 degrees) fromwhich the winds blew during 1 June 1996 to 31 May 1997. The horizontal axis gives date ofthe month shown at the bottom of each panel. Vertical axis gives direction.

hence sea-breeze circulation is expected to dominate coastal winds. In figure 1.15,wind vectors at 0600, 1200, 1800, and 0000 hours are identified. The wind speedusually picked up at 1000 hours (IST) and peaked around 1500 hours, the peak windspeed being about 5ms−1. By 2000h, the wind speed dropped to below 3ms−1, anddecreased slowly thereafter. Minimal wind speed of about 1.5ms−1 was observedat around 0600h. Before the onset of the sea breeze, the wind speed generally

June 13, 2007 19:40 RPS rpb001ch01


5

4

3

2

1

0

1

2

3

4

5

5 4 3 2 1 0 1 2 3 4 5

>Orientation ofcoastline

00 hr

06 hr

12 hr

18 hr

U (m s−1)

V (

m s

−1)

Figure 1.15 Sea breeze hodograph for the average winds during 1–15 April 1997. Thewind vectors at 0600, 1200, 1800, and 0000 hours are identified. Though there is day-to-dayvariation in the sea breeze (figure 1.13), on average, the sea breeze at about 1200 hours isapproximately from the northwest. At this time, the wind has a component perpendicularto the coastline and is oriented towards the coast (see orientation of the coastline marked).By 1800 hours the wind vector turns towards the right, to a direction that is along the coast,and continues to turn further while the magnitude decreases.

remained less than 2ms−1. The onset was marked by an abrupt change in winddirection: before 1000 hours, the wind was from about 90◦, but after onset, thedirection changed to about 300◦ (figure 1.14). During 1000–2000 hours, the winddirection changed slowly from 300◦ to about 360◦. The coastline in the vicinity ofGoa is oriented along approximately 340◦. Hence, the sea breeze recorded by theanemometer at a height of about 50m above sea level is almost along the coastline.At sea level, the breeze is expected to be at an angle somewhat lower than the angleat the height of the anemometer owing to veering due to frictional effects in theatmospheric boundary layer.

Sea breeze is a common mesoscale meteorological phenomenon in many coastalareas of the world, including the area along the coastline of the Indian subcontinent.The effect of the phenomenon is felt over both land and sea. Aparna et al. (2005),quoting earlier studies, noted that the landward extent along the Indian coast is110–150 km. They estimated that the sea breeze extends up to a distance of about180 km off Goa. This inference was made using an objective method based onoffshore decay of vector correlation.

June 13, 2007 19:40 RPS rpb001ch01


1.5 WAVES

Concomitant with the wind observations during June 1996 to May 1997, waveswere measured off the mouths of the Mandovi and Zuari estuaries using a Datawelldirectional waverider buoy in 23m of water. The instrument recorded 20-minuteaveraged wave spectra once every three hours. The most striking feature of theannual march of the spectra is the dramatically high wave energy observed duringthe summer monsoon when compared to the rest of the year. As seen in figure 1.16,

0

100

200

300

400

0

100

200

300

400

0.10.2

0.30.4

0.5

0.10.2

0.30.4

0.5

010

2030

4050

6070

8090

100

0.10.2

0.30.4

0.5

0 20 40 60 80 10010

0

200

300

400

Day

s fro

m 1

9 Fe

brua

ry,1

996

Frequency (Hz)

Ene

rgy

Den

sity

(m

2 H

z−1)

Figure 1.16 Spectral density (m2 Hz−1) as a function of frequency (Hz) at 3-hour intervalduring June 1996 to May 1997. White bands indicate missing data. Notice that most of theenergy is seen during the summer monsoon.

June 13, 2007 19:40 RPS rpb001ch01


peak spectral energy during the summer monsoon is about an order of magnitudelarger than that observed during the rest of the year. The wave spectrum at thistime is dominated by the swells that are generated over the open sea by the strongwinds of the summer monsoon.

The energy of the swells reduces after the withdrawal of the monsoon. Soon, the seabreeze becomes an important feature of the winds in the coastal areas. Wave datashow that the wave spectrum at this time often has two well-separated parts. Thelow-frequency part has energy with frequencies lesser than 0.15 Hz (i.e., periodslonger than about 7 seconds) and represents the swells. The high-frequency part,or the “sea”, represents the spectrum of the locally generated waves. As seen infigure 1.17, which presents the evolution of the spectrum during 1–15 April 1997,the sea spectrum often shows diurnal variability. This is in contrast to the spectrum of

0

0

0.10.2

0.30.4

0.5

0.10.2

0.30.4

0.5

0.0

0.5

1.5

2.0

2.5

0.0 0.5 1.0 1.5 2.0 2.5

1.0

14

13

12

1110

98

76

54

32

10

Dat

e du

ring

0115

Apr

il,19

97

Frequency(Hz)

Ene

rgy

Den

sity

(m2

Hz−1

)

Figure 1.17 Spectral density (m2 Hz−1) as a function of frequency (Hz) at 3-hour intervalduring 1–15 April 1997.

June 13, 2007 19:40 RPS rpb001ch01


swells. Neetu et al. (2006) have argued that the high-frequency spectrum is generatedby the sea breeze on the continental shelf off Goa.

Wave spectra in shallow areas serve as good indicators of energy available formixing of the water column. As seen above, the available energy is much largerduring the summer monsoon than during the rest of the year. This is expected tohave a major impact on vertical mixing near the mouths of the estuaries, in theAguada Bay in the Mandovi and in the Mormugao Bay in the Zuari (see Map A).After withdrawal of the monsoon, the two bays have lesser energy coming intothem from offshore, and the high-frequency part of this energy has a diurnal cycle.Contribution of energy from surface waves to vertical mixing in comparison toother available sources such as tidal stirring, is not known and needs to be explored,particularly in the two bays. It is expected that in the channels at the upstream endsof the two bays, the role of wind waves in vertical mixing would be significantlylower than that in the bays.

1.6 ESTUARINE REGIMES

As noted earlier, based on channel geometry, it is appropriate to divide each estuaryinto two parts: the bay and the channel. The former is wide, thus making along-estuary and cross-estuary physical processes similar in magnitude and importance.In the latter, along-estuary processes are by far the more important for temporalevolution of the estuary. The annual cycle of processes in each of these parts can bedivided into two phases: wet and dry. The wet phase is about 4–5 months long andis characterized by high influx of freshwater. The dry phase is 7–8 months long andthe influx of freshwater is either marginal or negligible. We can therefore dividethe evolution of the annual cycle of the estuary into the four regimes depicted infigure 1.18. In the two-part name of each regime, the first part distinguishes theregime with respect to runoff, and the second part with respect to geometry. Thefour regimes are:

• ‘Wet-Bay’: This regime represents the bay in each estuary when freshwater influxinto the bay is high, i.e., during the summer monsoon. When the influx is partic-ularly high, such as when the monsoon is in an active phase, the bay turns intoan almost entirely freshwater body. During lulls in the monsoon, tidal exchangeacross the wide mouth of the bay is expected to ensure that conditions similar tothose on the shelf off the bay would also exist in the bay. This is one of the tworegimes when mixing due to the influence of winds and waves is expected to beimportant. This regime is one of rapid temporal change in physical processes,and biogeochemical activity must adjust to this rapid change.

• ‘Wet-Channel’: This regime represents the channel in each estuary during thehigh runoff time of the summer monsoon. The channel at this time is primarilya freshwater body, except perhaps in the immediate vicinity of its bay-end. This

June 13, 2007 19:40 RPS rpb001ch01


0

10

20

30

40

50D

ista

nce

from

mou

th (

km)

J J A S O N D J F M A MTime (month)

WetBay

WetChannel DryChannel

DryBay

Figure 1.18 Schematic of the four regimes into which the Mandovi and Zuari estuaries canbe divided. See section 1.6 for details.

is the time of flushing of the estuary a few times over during the season. Eachflushing renews the water in the channel completely.

• ‘Dry-Bay’: This regime represents the bay during the dry season. Tides areexpected to make conditions in the bay similar to those that exist on the shelfoffshore. This period of about 8 months is the time of temporal evolutions thatis much slower than that seen during the ‘Wet-Bay’ regime. Winds, including theafternoon sea-breeze, are expected to contribute to vertical mixing in this regime.

• ‘Dry-Channel’: This is the most interesting of the four regimes. It is observed inthe channels during about 7–8 dry months after the withdrawal of the monsoon.The regime is dominated by physical processes that ensure continuous intrusionof salt from the bay-end of the channel towards the head. Tides play an importantrole in the intrusion. The salt intrudes into the pool of freshwater created bythe last episode of flushing during the summer monsoon. Of course, there iscontribution of freshwater to the pool from runoff after the withdrawal of themonsoon. This contribution is expected to be small, but we do not have reliableestimates of it.

The four regimes define the state of each of the estuaries based on factors that aredifferent from the internal working of the estuary. These regimes set the stage toput into perspective the internal working that forms the central theme of this book.

June 13, 2007 20:39 RPS rpb001ch02

2Tides and sea-level variability

S. R. Shetye, I. Suresh, and D. SundarNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

2.1 INTRODUCTION

Tides are a consequence of the gravitational pull of the Moon and the Sun exertedon the waters of the ocean as the three-body system consisting of the Earth, Moon,and Sun moves in a stable configuration under the influence of the gravitational pullof one another. Tides at any location on the Earth’s surface consist of a superpositionof oscillations with well-known periods. The periods depend on characteristics ofthe motion of the Sun–Earth–Moon system. The study of tides since the last fewhundred years has carefully documented the most important of these periods. In thejargon of tidal analysis, a period corresponds to a tidal constituent. There are nowwell-established procedures to determine the characteristics of a tidal constituentat a location, if a sufficiently long time series of sea level is available. A month-long time series is sufficient to determine the characteristics of about 30 importanttidal constituents at a location. Usually, information on half a dozen of the mostprominent (largest amplitude) constituents will suffice to understand the importantcharacteristics of the tide at the location.

While studying tides in an estuary, it is generally assumed that the tide at themouth of the estuary is a consequence of global-scale tidal dynamics to which thecontribution of the estuary is negligible. Given the tide at the mouth as a bound-ary condition, the characteristics of the tide within the estuary, as forced at theboundary, are then studied. We follow the same approach in this chapter. Section 2.2

29

June 13, 2007 20:39 RPS rpb001ch02


describes tides using a three-year record at Mormugao located at the mouth of theZuari estuary (Map A). In 1993, tide-pole measurements were carried out at 15locations in the Mandovi and Zuari estuaries for three days. The inferences derivedfrom an analysis of these data were that the amplitude of the tide, i.e., the range ofwater-level variation due to all tidal constituents, remains virtually unchanged in thechannels. The tide propagates from mouth to head with an average speed of about6 ms−1 (Shetye et al. 1995). However, the time-series of the data collected duringthis field experiment, being only three days long, did not permit determination ofthe characteristics of individual tidal constituents. During March–April 2003, tide-pole measurements similar to those during 1993 were carried out at 13 locations byrecording the water level once every fifteen minutes. The new time series, beingat least 30 days long, allowed the tidal analysis to determine 36 tidal constituents.The inferences reported by Sundar and Shetye (2005) differed from those reportedin Shetye et al. (1995). The new study, owing to a superior dataset, was able todetermine the characteristics of all the important tidal constituents in the two estu-aries. In section 2.3 we summarize the results of Sundar and Shetye and then goon to examine the reasons underlying the observed features (sections 2.4 and 2.5).The high runoff in the Mandovi and Zuari estuaries changes the mean sea level inthe two estuaries particularly at the upstream end; in section 2.6, the nature of thischange is described using the data collected in 1993.

2.2 TIDES AT THE MOUTH

The Survey of India maintains a tide gauge at Mormugao (Map A), an importantport. Typical variation of water level during a month as recorded by this gauge isshown in figure 2.1. Shetye et al. (1995) used hourly values of sea level recordedby this gauge during 5 June 1990 to 30 November 1993 to determine tidal charac-teristics. Harmonic analysis of the record gave amplitudes and phases of 69 tidalconstituents with periods between 6 months and 3 hours. Of these, the seven diur-nal and semi-diurnal constituents listed in table 2.1 had amplitudes in excess of5 cm. The amplitude and phase of each tidal constituent gets modified as the tidepropagates into the estuarine channels.

2.3 DATA OF MARCH–APRIL 2003 AND ANALYSIS

The 13 locations where tide pole measurements were carried out during March–April 2003 are shown in Map B. The 15-minute data collected at these locationswere analyzed to determine the tidal amplitude and phase difference of 28 majorconstituents and 8 related constituents. Of the 13 stations where the observationswere carried out, six were located in the Mandovi main channel, five in the Zuari

June 13, 2007 20:39 RPS rpb001ch02

TIDES AND SEA-LEVEL VARIABILITY 31

50

0

50

100

150

200

250

300S

ea L

evel

(cm

)

5 10 15 20 25 30Time (days)

Figure 2.1 Variation in water level recorded by the Survey of India tide gauge at Mormugaoduring January 1975.

Table 2.1 Diurnal and semi-diurnal tidal constituents at Mormugaowith amplitude greater than 5 cm.

Symbol of the Amplitudeconstituent Description Period (h) (cm)

O1 Principal lunar 25.82 16.17P1 Principal solar 24.07 9.32K1 Combined diurnal 23.93 31.98N2 Larger elliptical lunar 12.66 13.30M2 Principal lunar 12.42 55.65S2 Principal solar 12.00 20.23K2 Combined declinational 11.97 5.41

main channel, and two in the Cumbarjua Canal (Map B). The quality-controlleddata at these 13 locations are shown in figure 2.2. One feature seen in both estuariesis that the amplitude of water-level oscillation decreases drastically at the upstreamend. As will be seen later, this is due to an increase in the elevation of the twostations. Ignoring these stations for the time being, we examine the variation ofamplitude and phase lag in the two estuarine channels.

Of the 36 constituents whose amplitudes and phases were determined, six con-stituents —- M2, S2, N2, K1, O1, and P1 —- had an amplitude larger than 10cm at the

June 13, 2007 20:39 RPS rpb001ch02


0

100

200

300

Verem

Mandovi

0

100

200

300

Britona

0

100

200

300

Akkada

0

100

200

300

Volvoi

Obs

erve

d S

ea L

evel

(cm

)

0

100

200

300

Usgao

0

100

200

300

Ganjem

10 17 24 31 7 13March April

0

100

200

300

Zuari

Dona Paula

0

100

200

300

Cortalim

0

100

200

300

Borim

Obs

erve

d S

ea L

evel

(cm

)

0

100

200

300

Savordem

0

100

200

300

Sanguem

10 17 24 31 7 13March April

0

100

200

300

Cumbarjua

Banastarim

Obs

erve

d S

ea L

evel

(cm

)

10 17 24 31 7 13March April

Madkai

0

100

200

300

Figure 2.2 Observed sea level (cm) during the period of observation. The horizontal axisgives date and month. (After Sundar and Shetye 2005.)

June 13, 2007 20:39 RPS rpb001ch02


20

40

60

O1

K1

N2

M2

S2

MandoviA

mpl

itude

(cm

)

O1

K1

N2

M2

S2

Zuari

0

60

120

180

240

300

360

0 15 30 45

O1

K1

N2

M2

S2

Pha

se la

g (d

egre

e)

Distance from mouth (km)0 15 30 45

O1

K1

N2

M2

S2

Figure 2.3 Variation in amplitude (cm) and phase lag in degrees (referenced to IST) of M2,S2, N2, K1, and O1 in the main channels of the Mandovi and Zuari estuaries. (After Sundarand Shetye 2005.)

mouth of the two estuaries. Figure 2.3 shows how the amplitudes of five of theseconstituents (M2, S2, N2, K1, and O1) changed from mouth to head. Shetye et al.(1995), using three days of data, had inferred that tidal amplitude remains virtuallyunchanged in the two estuarine channels. From studies based on the month-longdata reported in Sundar and Shetye (2005), we find that behaviour of the ampli-tude in a channel depends on the constituent. The amplitude of the semi-diurnal

June 13, 2007 20:39 RPS rpb001ch02


constituents increased by about 20% in the Mandovi from mouth to head and byabout 30% in the Zuari; the increase in diurnal constituents was about 10% in theMandovi and 15% in the Zuari. Phase lag of both semi-diurnal and diurnal tidesincreased from mouth to head in both estuaries, and the increase was more for thesemi-diurnal than for the diurnal.

2.4 TIDAL DYNAMICS

It is known that in shallow estuarine channels that converge (cross-sectional areadecreases from mouth to head), three factors determine tidal propagation. The first isthe rate of decrease in channel cross-sectional area from mouth to head. More rapidthe decrease, greater is the tendency of the amplitude to grow. This is a consequenceof geometry: more confined the available cross-sectional area, greater is the increasein amplitude. Another factor is friction. Greater the friction, greater the decay of thetide as it propagates. Frictional dissipation usually varies among tidal constituentsbecause the nonlinear effects that lead to transfer of energy from one constituent toanother depend on the constituent. Numerical models of estuarine tidal circulationare usually successful in accurate simulation of tidal amplitude (see chapter 3). Whilesuch models are useful in applications, simpler, preferably analytical, models areuseful to gain a perspective of the dynamics underlying tidal propagation.

Many authors have examined the characteristics of tidal propagation under theinfluence of the two factors mentioned above, geometry and friction. Friedrichs andAubery (1994) examined propagation of a tidal constituent in a converging channelas a unidirectional progressive wave. In this model, amplification due to geometriceffect is cancelled by friction to allow the tide to propagate as a free wave. Nayakand Shetye (2003) looked at the same problem by allowing a reflected wave to formwithin the estuarine channel. In this case, the characteristics of tidal propagation ofa constituent are determined by the combined effect of the incident wave (whichis like the progressive wave of Friedrichs and Aubrey, but is allowed to amplifyor decay under the combined influence of geometry and friction), and a reflectedwave that propagates in the direction opposite to that of the incident wave. Sureshand Shetye (2007) studied the propagation of tides in the Mandovi and Zuari usingdynamical relations similar to those in Nayak and Shetye (2003). In these relations,propagation of each tidal constituent into the estuary was modelled as an incidentwave of surface height and cross-sectionally averaged velocity forced by the tideat the mouth. In this model, a wave propagates in the channel under the influenceof pressure gradient, estuarine geometry, and friction. With the propagation of anincident wave, a reflected wave gets generated in the model. The estuarine geometry(width and depth) was modelled so as to mimic the observed geometry, i.e., theestuary had two parts, a bay and a channel. The former was wider, marginallydeeper, and more rapidly converging than the latter. The friction coefficient wastuned to simulate the observed behaviour of the constituent.

June 13, 2007 20:39 RPS rpb001ch02


Comparison between the observed and predicted values of amplitude and phasefor the most significant semi-diurnal and diurnal constituents, M2 and K1, suggeststhat this simple model captures the essence of tidal dynamics of the estuaries. Thewavelength of the incident wave in the model is of the order of a few hundredkilometres, much longer than the length of the estuary. Hence, consistent withobservations, it is expected that the water level would rise or fall with little phasedifference along the length of the estuary (observed maximal phase difference is∼ 60 degrees in semi-diurnal and ∼ 20 degrees in diurnal constituents). The phasespeed of the wave depends on a number of model parameters, including depth(deeper the channel, faster the propagation), rate of convergence of the channel,period of the incoming wave, friction, etc., and varies between 5 and 15 ms−1 inthe Mandovi and 8 and 30 ms−1 in the Zuari for the semi-diurnal constituents.Corresponding figures for the diurnal constituents are 6–17 ms−1 in the Mandoviand 5–14 ms−1 in the Zuari. The phase lag of the tide at a location in the channel isdetermined by superposition of the incident and reflected waves. The amplitude ofthe incident wave decreases from mouth to head, whereas that of the reflected waveincreases from mouth to head. Model computations showed that at a distance of30 km from the mouth, the incident wave is nearly 10 times larger than the reflectedwave in both estuaries. At a distance of 5 km from the head, the amplitude of thereflected wave is nearly half that of the incident wave in both estuaries. At the head,the two waves cancel each other to satisfy the boundary condition of no net flow(i.e., the dry season condition). The most useful characteristic of a model of thiskind is that it provides a useful conceptual framework to understand the observeddistribution of amplitude and phase in the estuaries in terms of propagation of twowaves.

2.5 DECAY OF TIDAL AMPLITUDE OWING TO INCREASEIN CHANNEL ELEVATION

Having examined the tidal propagation over the coastal plain, we now examinewhat happens at the upstream ends of the estuarine channels. This part of thechannel, being at the foothills of the Sahyadris, has a mean elevation above that ofthe coastal plain. The drastic drop in tidal amplitude seen at the upstream ends of thetwo estuarine channels is related to this increase in elevation. We now examine thisissue using the explanation given in Sundar and Shetye (2005). Each of the sevenpanels in figure 2.4 gives a section of the channel topography and water level alonga line drawn through the middle of the Mandovi channel. Both the topographyand the water level are based on observations: the topography is the same as thatdescribed in chapter 1, and the water surface depicted in the figure is based on datacollected during the 2003 observations. Two consecutive panels in the figure showhow the water level changed over a period of two hours during spring tide. Theseven panels together cover 12 hours. The figure helps to visualise how the waterlevel along the channel changes during a tidal cycle, and hence to understand how

June 13, 2007 20:39 RPS rpb001ch02


Figure 2.4 Each of the seven panels in this figure gives an instantaneous section of thechannel topography and water level along a line drawn through the middle of the Mandovichannel. Two consecutive panels are separated by two hours during spring tide. (AfterSundar and Shetye 2005.)

June 13, 2007 20:39 RPS rpb001ch02


the increase in channel elevation causes the tidal amplitude to drop at the upstreamend in an estuarine channel. As seen in figure 2.2, the tidal amplitude drops sharplyfrom Usgao to Ganjem in the Mandovi and from Sanvordem to Sanguem in theZuari. The tide reaches Ganjem in the Mandovi only when the tide is high enoughto overcome the effect of increased channel elevation. The impact of the tide isfelt at Ganjem in panels A, E, F, and G of the figure. In the rest of the panels theinfluence of the tide is not seen at Ganjem. In these cases the water level at Ganjemis determined merely by whatever little runoff exists in the river channel.

2.6 CHANGE IN MEAN SEA LEVEL BETWEEN WETAND DRY SEASON

During the summer monsoon, when runoff into the estuaries is high, the water-level variation is determined not only by the tide, but also by the variation in runoff.Impact of the runoff is particularly high in the upper reaches of the estuaries. Shetyeet al. (1995) measured the difference between mean water levels during two three-day long periods. One of these was during the dry season (7–9 April 1993) and theother during the wet season (19–21 August 1993). The difference as measured bythem is shown in figure 2.5. At the upstream end of the Mandovi the difference waswell over 1.5m and the tide was no more felt. It is expected that when an active spellof the monsoon occurs, this difference would be much higher. During such spells,

50

0

50

100

150

200

Diff

eren

ce in

mea

n se

a le

vel (

cm)

0 10 20 30 40 50 60Distance from mouth (km)

Mandovi

Zuari

Figure 2.5 Difference (cm) in mean sea level between 19–21 August 1993 (wet season) and7–9 April 1993 (dry season) as a function of distance from the mouths of the main channelsof the Mandovi and Zuari. (After Shetye et al. 1995.)

June 13, 2007 20:39 RPS rpb001ch02


the Mandovi becomes an extension of the rivers like the Mhadei and Khandeparthat supply freshwater to it. Marine influence on water-level variability disappears.

The influence of the runoff on water level decreases in the downstream directionas the channel, cross-section widens. As seen in figure 2.5, at the mouth of theMandovi, the mean water level was marginally lower during the wet season. In theZuari, the mean water level was lower by about 10 cm. Such a lowering of waterlevel is expected from the dynamics of large-scale currents along the west coastof India. The summer monsoon is the time of an equatorward current along thiscoast and upwelling occurs (Shetye et al. 1990). Coastal upwelling is expected to beaccompanied by lowering of sea level.

June 13, 2007 20:35 RPS rpb001ch03

3Numerical models

A. S. Unnikrishnan and N. T. ManojNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

3.1 INTRODUCTION

In a typical partially-mixed estuary, such as the Mandovi or Zuari, the tidally-averaged circulation consists of a downstream directed surface flow of fresher waterand an upstream-directed flow of more saline water at the bottom. This circulation,however, is much weaker than the circulation observed during a tidal cycle, partic-ularly when the tidal range is large (well over a metre). For practical applicationssuch as determination of the fate of pollutants and other passive and active tracers,it is often important to simulate this more energetic circulation within a tidal cycle.

Variability in estuaries is a function of three-dimensional space and time (x , y, z ,and t ). The partial differential equations that describe tidal circulation and salin-ity distribution in an estuary consist of the momentum equations, the equation ofcontinuity, and the conservation equation for salt. A description of these equationsis given in many text books on estuarine dynamics (Officer 1976). These differen-tial equations cannot be solved analytically, except in very special situations. Withthe development of computers and advances in numerical methods, many numer-ical models have been developed for solving complex three-dimensional estuarineproblems. Nevertheless, often, simpler models suffice.

The simplest type of these models is a one-dimensional model, often used forapplication to a narrow estuary. If wider sections are present in one such estuary,then such models can be supplemented by two-dimensional models. This chapterdescribes models of Mandovi and Zuari estuaries that follow both these approaches.

39

June 13, 2007 20:35 RPS rpb001ch03

40 A. S. UNNIKRISHNAN AND N. T. MANOJ

The next section describes a one-dimensional network model that was used to sim-ulate water level heights in the estuary, and section 3.3 describes a model thatuses the two-dimensional formulation in the wider sections of the estuaries and isone-dimensional in the narrower sections.

3.2 NETWORK MODEL OF MANDOVI–ZUARI ESTUARIES

In an early attempt at modelling of Mandovi–Zuari estuaries, Shetye and Murty(1987) used a one-dimensional model to simulate the annual cycle of salinity inthe Zuari estuary. The model was successful in simulating the observed averagesalinity over a tidal cycle during different months of a year along the entire stretchof the Zuari. This model did not resolve the tides. It looked at the tides as merelya mechanism that permits diffusion of salinity in the channels.

Unnikrishnan et al. (1997) resolved the tides in a model constructed by connectingall the major tributaries of Mandovi and Zuari and the Cumbarjua Canal. In thisnetwork model, each estuary or river is represented by a one-dimensional channel.Since the channels are connected, we refer to the model as a ‘network model’. Thismodel was used to simulate tides in the estuarine system, to study the dynamicsof tidal propagation, and to identify the changes introduced in water-level varia-tion by runoff during the wet season. The model used the following equations forconservation of momentum and mass in a one-dimensional channel.

∂U∂t

+ 2UA

q − 2bUA

∂η

∂t= −gA

∂

∂x(z0 + h + η) − g

U |U |AC 2R

, (3.1)

b∂η

∂t+ ∂U

∂x− q = 0, (3.2)

where t , x , U , η, q , g , h and R are, respectively, time, along-channel coordinate(increasing in upstream direction), along-channel transport, surface elevation withrespect to mean water level, freshwater influx per unit channel length, accelera-tion due to gravity, depth, and hydraulic radius; z0 is the height between bottomof the channel and an arbitrary datum below the bottom; bs is the width in whichalong-channel flow occurs, whereas width b includes mud flats, which act as storage.R = (A/Pr ), Pr being the wetted perimeter and A the area of cross section (excludingmud flats); C = (1.49/n)R1/6, where n is the Manning coefficient. The numericalscheme used by Harleman and Lee (1969) was used to solve the above equations.In this scheme, the continuity equation is solved at odd grid points to compute η

at the next time step and the momentum equation is solved at even grid points tocompute U . The original scheme of Harleman & Lee (1969) was developed for asingle channel. For developing a model for the Mandovi–Zuari estuarine system,Unnikrishnan et al. (1997) modified the above scheme for application to a network

June 13, 2007 20:35 RPS rpb001ch03

NUMERICAL MODELS 41

of channels. This was done by defining the equation of continuity at the junction ofthe network of more than two channels as follows:

�xbiηi ,t+�t − ηi ,t

�t= Ui−1,t − Ui+1,t − Uk ,t , (3.3)

where k is the adjacent grid point located in a channel connected to the mainchannel defined by grid points (i − 1), i , and (i + 1).

In the network model, as seen in figure 3.1, the Kushavati River is represented bya channel that is connected to the Zuari. The Mhapsa, Sinquerim, and Khandepar

X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

1

(61)

7

76

21 (

69)

Kus

hava

ti R

.

30

60

100

110

92

104

Sin

quer

im R

.

Mha

psa

R.

Val

vat a

nd D

icho

lim R

.

Kha

ndep

ar R

.

33 (

101)

31

35 (

77)

93 (

105)

51 (

93)

Zuari

Mandovi

Cum

barju

a C

analA

rabi

an S

ea

Figure 3.1 Schematic representation of the one-dimensional network model. The two mainestuaries, Mandovi and Zuari, the connecting channel, Cumbarjua Canal, and the majorrivers that join the estuaries are represented by a one-dimensional model each. In thenumerical staggered grid used, the stars (dots) indicate the computation points for velocity(surface elevation). The numbers shown indicate segment numbers. An intersection of twochannels is assigned two grid-points: the first refers to the main channel (estuary) and thesecond, shown in brackets, to an intersecting channel (river). The grid spacing is 2km. (AfterUnnikrishnan et al. 1997.)

June 13, 2007 20:35 RPS rpb001ch03


rivers are represented by a channel each, while Dicholim and Valvat rivers are rep-resented by a single channel and are connected to the Mandovi estuary. Each estu-arine channel is divided into a number of segments. The parameters that describe asegment are the equivalent depth and equivalent breadth, which are the represen-tative values for that particular segment. The distance between any two adjacentsegments or the grid space used is 2km and the time step used for numerical integra-tion is 110. The values of Manning coefficient (n) used varied from 0.025 to 0.030.When channel widths are less than 100m, n was set at 0.03; elsewhere, it was 0.02.

The tides simulated using the network model at various locations for the period,when observations were available in April (dry season) and August (wet) season inthe Zuari estuary, are shown in figure 3.2. No significant change in tidal ranges is

0

1

2

3

6 7 8 9 10 11Time (days)

Mormugao

0

1

2

3 Cortalim

0

1

2

3

Sur

face

ele

vatio

n (m

)

Loutulim0

1

2

3Sanvordem

0

1

2

3

Sanguem(a)

0

1

2

3

18 19 20 21 22 23Time (days)

Mormugao

0

1

2

3 Cortalim

0

1

2

3

Loutulim0

1

2

3Sanvordem

0

1

2

3

Sanguem(b)

Figure 3.2 Simulated surface elevations (one-dimensional network model) plotted withobserved surface elevations in the Zuari estuary during (a) 7–9 April 1993 (dry season) and(b) 19–21 August 1993 (wet season). Continuous (dotted) lines represent model (observed)values. (After Unnikrishnan et al. 1997.)

June 13, 2007 20:35 RPS rpb001ch03

NUMERICAL MODELS 43

found at Cortalim and Loutulim when compared to that at Mormugao, suggestingthat tides propagate upstream without undergoing any significant change for a dis-tance of about 40km (up to Sanvordem) from the mouth. At Sanguem, however,observations show a rapid decay during the wet season, which is simulated wellby the model. The characteristics of tidal propagation in the Mandovi estuary aresimilar to those in the Zuari, and are therefore not shown. Tides at Ganjem show arapid decay during August 1993, just as is observed at Sanguem in the Zuari.

The model reproduces the observed increase of mean sea level towards upstreamof the estuaries (Shetye et al. 1995) during wet seasons. When the river dischargeis high, a pressure gradient force develops from the head towards downstreamof the estuaries. In order to simulate this phenomenon, the model was run fordifferent discharge rates, namely, 0, 50, and 100m3 s−1. The runs were made fora period of M2 tide (12 hours and 42 minutes), which is the principal lunar tidalconstituent, and the values of surface elevation obtained during each time step wereused to compute the average value over the tidal period (figure 3.3). When riverdischarge is zero, there is no variation of mean sea level from mouth to head. In thesimulations with discharge rates of 50 and 100m3 s−1, an increase in mean sea leveltowards upstream is found. This increased mean sea level in the upstream directiongenerates a downstream directed current in wet seasons.

0

1

2

3

4

5

Mea

n se

a le

vel (

m)

0 10 20 30 40 50 60Distance (km)

0

50

100

Figure 3.3 Simulated mean sea level (one-dimensional network model) in the Zuari estuaryfor different river discharges (0, 50, and 100m3 s−1) using the network model. The simu-lations were made for an M2 tide (period of 12 hours and 42 minutes) and sea surfaceelevations obtained at each time step were averaged over one tidal period to compute themean sea level. (After Unnikrishnan et al. 1997.)

June 13, 2007 20:35 RPS rpb001ch03


3.3 HYBRID NETWORK TWO-DIMENSIONAL MODEL

A feature of the Mandovi and Zuari estuaries is that the width of each of the twoestuaries is large up to a distance of about 10km near the mouth, after which thechannel narrows considerably. Manoj and Unnikrishnan (2007) have developeda hybrid model that uses a two-dimensional model for the wide portion near themouths of the Mandovi and Zuari, and a one-dimensional network model for therest of the system. They used the model to study the tidal circulation and salinitydistribution.

To develop a two-dimensional vertically averaged model, the equations of momen-tum in x and y directions and the equation of continuity are integrated over thewater column in the vertical direction (see Leendertse and Gritton 1971 for details).The vertically averaged equations can be written as follows:

∂η

∂t+ ∂U

∂x+ ∂V

∂y= 0; (3.4)

∂U∂t

+ ∂

∂x(uU ) + ∂

∂y(vU ) = −g (h + η)

∂η

∂x+ fV

+ AH ∇2U − CDU

√U 2 + V 2

(h + η)2(3.5)

∂V∂t

+ ∂

∂x(uV ) + ∂

∂y(vV ) = −g (h + η)

∂η

∂y− fU

+AH ∇2V − CDV

√U 2 + V 2

(h + η)2(3.6)

∂S∂t

+ ∂

∂x(uS ) + ∂

∂y(vS ) = Kx

∂2S∂x2 + Ky

∂2S∂y2 (3.7)

Equation (3.4) is the continuity equation, equations (3.5) and (3.6) are the momen-tum equations in x and y directions respectively, and equation (3.7) is the conserva-tion equation for salt; x and y are the horizontal directions in a Cartesian co-ordinatesystem, η is the surface elevation with respect to mean sea level, h the still waterdepth, u and v are the vertically integrated velocity components in x and y direc-tions, g the acceleration due to gravity, f the Coriolis parameter, CD the bottomfriction coefficient, and AH the horizontal diffusion coefficient. U = (h + η)u andV = (h + η)v are the transports in x and y directions. S is the salinity and Kxand Ky are the diffusion coefficients in x and y directions respectively. The gridresolution is 500m and the time step used is 15s. The value chosen for Kx andKy are 100 and 10m2 s−1 respectively. The value of CD used is 0.002. Observedsalinities at Mormugao during April 1993 and August 1993 are used for prescribingopen boundary conditions along the open boundary at the mouth.

June 13, 2007 20:35 RPS rpb001ch03

NUMERICAL MODELS 45

One-dimensional models are connected to the two-dimensional model domainwhere the estuaries become narrow. Besides, Cumbarjua Canal and other tribu-taries are also represented by a one-dimensional model each. This entire networkmodel is referred to as a hybrid model. The equation of momentum and conti-nuity used for the one-dimensional models in this hybrid model are the same asthose described in the previous section. Additionally, we used an equation for theconservation of salt in the one-dimensional form (Harleman and Lee 1969):

A∂S∂t

+ ∂

∂x(uS ) = AKx

∂2S∂x2 . (3.8)

In the one-dimensional models also, the grid resolution is 500m, and the value ofthe diffusion coefficient is 100m2 s−1. The tidal currents simulated using the hybridmodel during a spring tide are shown in figure 3.4. The model runs made for a few

73 48' 73 51' 73 54'15 24'

15 27'

15 30'

73 48' 73 51' 73 54'15 24'

15 27'

15 30'0.5 m s-1

73 48' 73 51' 73 54'15 24'

15 27'

15 30'

(c)

73 48' 73 51' 73 54'15 24'

15 27'

15 30'

3 hours after high tide

73 48' 73 51' 73 54'15 24'

15 27'

15 30'

Zuari

73 48' 73 51' 73 54'15 24'

15 27'

15 30'Mandovi

73 48' 73 51' 73 54'15 24'

15 27'

15 30'

15 24'

15 27'

15 30'

15 24'

15 27'

15 30'

3 hours before high tide

15 24'

15 27'

15 30'0.5 m s-1

15 24'

15 27'

15 30'

(a)

15 24'

15 27'

15 30'

N

15 24'

15 27'

15 30'

Zuari

15 24'

15 27'

15 30'Mandovi

15 24'

15 27'

15 30'

High tide

0.5 m s-1

(b)

Zuari

Mandovi

73 48' 73 51' 73 54'73 48' 73 51' 73 54'

Low tide

73 48' 73 51' 73 54'

0.5 m s-1

73 48' 73 51' 73 54'

(d)

73 48' 73 51' 73 54'

Zuari

73 48' 73 51' 73 54'

Mandovi

73 48' 73 51' 73 54' E73 48' 73 51' 73 54'

Figure 3.4 Simulated tidal circulation in a two-dimensional model of the bays of the Man-dovi and Zuari estuaries during a spring tide. The timings of the occurrence of low tideand high tide are with respect to Mormugao: (a) three hours before the occurrence of lowtide, (b) at low tide, (c) three hours after low tide, and (d) at high tide. (After Manoj andUnnikrishnan 2007.)

June 13, 2007 20:35 RPS rpb001ch03


months, consisting of many spring and neap cycles, have shown that the magnitudesof currents in the downstream regions vary between 0.4 and 0.8ms−1.

The simulations (figure 3.5) using the hybrid model were made to study the lon-gitudinal salinity distribution during dry and wet seasons. The results indicate thatthe model reproduces the longitudinal variation of salinity very well during thedry season and reasonably well during the wet season. During the dry season,

0

10

20

30

40

Sal

inity

(ps

u)

0

10

20

30

40

Sal

inity

(ps

u)

0

10

20

30

40

19

0

10

20

30

40

(a)

0

10

20

30

40

20

0

10

20

30

40

21

0

10

20

30

40

22

0

10

20

30

40

7

0

10

20

30

40

(b)

0

10

20

30

40

8

0

10

20

30

40

9

0

10

20

30

40

10

0

10

20

30

40

Time (days)

0

10

20

30

40

Time (days)

0

10

20

30

40

Ganjem

0

10

20

30

400

10

20

30

40

Sonarbaag

0

10

20

30

400

10

20

30

40

Volvoi

0

10

20

30

400

10

20

30

40

Sarmanas

0

10

20

30

400

10

20

30

40

Aguada

0

10

20

30

40

0

10

20

30

40

Aguada

0

10

20

30

40

0

10

20

30

40

Sarmanas

0

10

20

30

40

0

10

20

30

40

Volvoi

0

10

20

30

40

0

10

20

30

40

Sonarbaag

0

10

20

30

40

0

10

20

30

40

Ganjem

0

10

20

30

40

Figure 3.5 Simulated (two-dimensional model) and observed distribution of salinity at var-ious locations in the Mandovi estuary: (a) 7–9 April 1993 (dry season) and (b) 19–21 August1993 (wet season). The continuous (dotted) lines represent model (observed) salinities. (AfterManoj and Unnikrishnan 2007.)

June 13, 2007 20:35 RPS rpb001ch03

NUMERICAL MODELS 47

the estuaries are well mixed, and during wet season, they are partially stratifiedin the downstream regions and nearly free from salt in the upstream regions. Thepresent study illustrates that the model is capable of reproducing the observedsalinity distribution fairly well and is useful for determining the longitudinal salin-ity distribution in the Mandovi and Zuari estuaries. The main inputs required forthe model runs are accurate boundary conditions: tides and salinity distributionalong the open boundary at the mouth and freshwater influx at the head of theestuaries.

3.4 CONCLUSIONS

All the efforts made so far at simulating the Mandovi–Zuari estuarine system treatthe estuary as a well-mixed estuary. While this is a good approximation to simulatethe horizontal distribution of active and passive variables of the estuaries, the nextchallenge lies in developing the capability to simulate distribution in the vertical. Itis hoped that this summary of work in mathematical modelling on the Mandovi andZuari estuaries carried out so far will induce researchers to take up this challenge.

June 13, 2007 18:54 RPS rpb001ch04

4Mixing and intrusion of salt

S. R. Shetye, G. S. Michael, and C. Pradnya Vishwas∗National Institute of Oceanography, Dona Paula, Goa 403 004, India.∗Department of Physics, Fergusson College, Pune 411 005, India.

4.1 INTRODUCTION

Estuaries are generally classified using two criteria: geological features and charac-teristics of circulation and mixing. As per the first criterion, the Mandovi and Zuariestuaries are coastal-plain estuaries because they are located on thealluvial coastal plain between the Sahyadris and the Arabian Sea (see the schematicin figure 1.2). Geological evidence from the continental shelf off Goa suggeststhat between 18,000 years BP (before present) when the last ice age occurred,and 7,000 years BP, the local sea level rose by approximately 110m and hasbeen at the present level since (Nair et al. 1979; Rao et al. 2003). As the sealevel rose, the sea invaded low-lying coastal river valleys and formed the presentday estuaries of Mandovi and Zuari. In this chapter, we examine classification ofthese estuaries from the point of view of their circulation and mixing as inferredprimarily from data on salinity. We begin the next section by examining the timeseries of vertical profiles of salinity collected at one location in the Mandovi.

4.2 TIME SERIES OF VERTICAL STRATIFICATION

On 1 February 1997, starting at about 1000 hours IST, observations on verticaldistribution of salinity and temperature were made once every half hour for 25 hours

49

June 13, 2007 18:54 RPS rpb001ch04


in the main channel of the Mandovi at Old Goa (see Map A for location). At thistime of the year, runoff in the Mandovi and its tributaries is negligible, and the flowis dominated by tides. It was spring tide during this field observation, as can be seenfrom figure 4.1, and this gave the expected water-level variation at Old Goa owingto the tide. To carry out the day-long time-series observations, a fishing trawlerwas stationed in the middle of the channel and vertical profiles of temperature andsalinity were collected using a SeaBird portable CTD profiler (Seacat).

The data were used to construct the contour plot in figure 4.2(a), which showsvariation of salinity as a function of depth and time. The location of the upperwater surface in the figure was determined with the help of a tide pole that waslocated at the jetty close to where the trawler was stationed. The most importantfeature seen in the data was that vertical stratification due to salinity depended onthe phase of the tide. During flow or ebb, when the water level rose or fell rapidly—-and, by inference, the tidal current was strong —- vertical stratification remainednegligible and the water column was well-mixed. As the water level that rose or felldue to the rising or falling tide slowed, and the tidal current weakened, stratificationbuilt up. It continued to build up as the tide turned. The stratification was erodedas the rise or fall of the tide picked up. Vertical variation in temperature during the

−1.25

−1.00

−0.75

−0.50

−0.25

0.00

0.25

0.50

0.75

1.00

1.25

Sea

leve

l (m

)

0 168 336 504 672Time (hours)01/02/1999 08/02/1999 22/02/1999 01/03/1999

Figure 4.1 Expected water level at Old Goa during February 1999. The symbols (circles)show the measured water level at Old Goa; these tide-pole observations were made on thedays when profiles of temperature and salinity were collected. The water level measured atOld Goa is super-imposed on the tide predicted at Mormugao (solid line) after shifting thelatter (in time) to make it coincide with the water level at Old Goa; the tide at Old Goa lagsthe tide at Mormugao by about 40 minutes. The four horizontal lines show the time whenvertical profiles of salinity and temperature were collected at Old Goa.

June 13, 2007 18:54 RPS rpb001ch04

MIXING AND INTRUSION OF SALT 51

0

1

2

3

4

5

6

7

8

Dep

th (

m)

0 2 4 6 8 10 12 14 16 18 20 22 24 26Time (hours)

20 21

2222

23

23

23

24

24

24

24

24

25

25

25

25

25

26

26

26

27

27

28

(b) February neap tide

0

1

2

3

4

5

6

7

8

Dep

th (

m)

0 2 4 6 8 10 12 14 16 18 20 22 24 26Time (hours)

21

2122

22

22

23 232324

24

24

2525

25

25

26

26

26

26

27

27

27

28

28

28

28

29

2930

3031

(a) February spring tide

Figure 4.2 Contour plot showing vertical structure of salinity (psu) in the Mandovi at OldGoa (marked by OL in Map B). A profile was collected once every half an hour. The tidewas measured using a tide pole attached to a nearby jetty. (a) Observations during springtide (1–2 February 1999); the horizontal axis gives time in hours starting from 1000 hourson 1 February. (b) Observations during neap tide (8–9 February 1999); the horizontal axisgives time in hours starting from 1000 hours on 8 February.

June 13, 2007 18:54 RPS rpb001ch04


observations was negligible. The maximum and minimum temperature observedanywhere in the water column during the field observation were 27.1◦C and 25.9◦C.Hence, as expected, it is the salinity field that provided density stratification.

Time-series observations of vertical structures were repeated on three other occa-sions, 8–9 February, 16–17 February, and 22–23 February 1999, at the same location.While the tidal range during the observations on 1–2 February was 2.02m, it was1.16m, 1.94m, and 1.58m respectively during the three successive field observa-tions that followed (see figure 4.1). The observed characteristics of vertical strati-fication on 16–17 February and on 22–23 February were similar to the variationon 1–2 February. On 8–9 February, i.e., during the neap tide, they were different(figure 4.2b). Salinity stratification was much better developed on this day: startingfrom about nine hours from the start of the observations till 18 hours from the start,the water column remained well stratified. During these nine hours, the water levelchanged by less than 70cm. In contrast, water level changed by as much as 2m in6 hours during the spring-tide observations shown in figure 4.2(a). Faster the rateof change of water level, and hence, by inference, larger the tidal current, greater isthe erosion of stratification. When the variation in water level is slow and currentsare weak, the scope for build-up of stratification is greater.

Of the four time series, three were similar to that in figure 4.2(a), showing a watercolumn that is well-mixed all the time, except when the tide turns. Strong stratifica-tion, such as that seen in figure 4.2(b), builds up only during neaps. During springs,the maximum difference in the water column was about 4psu, but such high valueswere rare, and most of the time the difference was less than 2psu. During neaps, themaximum difference was much higher, and it was not uncommon to see variationsof 6psu in salinity.

4.3 SECTION OF SALINITY ALONG THE MANDOVI ESTUARY

On 28 December 1999, at 1245 hours, a fishing trawler started from approximately35km from the mouth of the main channel of the Mandovi and moved towardsthe mouth near Aguada. The trawler reached the mouth at 1540 hours. The natureof the tide during this journey can be appreciated from the tidal variation shownin figure 4.3. En route, the trawler halted at 10 locations to collect data on verticalprofiles of salinity and temperature. These profiles allow us to construct a sectionof salinity along the Mandovi estuary. The section is shown in figure 4.4(a). Asimilar section that was made on 27 October 1997, two days after occurrence ofneap tide, is shown in figure 4.4(b). Spring tide occurred on 17–18 October and3 November. These two sections do not represent an instantaneous situation in theestuary. Nonetheless, they are useful to get an estuary-wide representation of thestratification.

June 13, 2007 18:54 RPS rpb001ch04


0

50

100

150

200

250

1912 hours

1219 hours

Pre

dict

ed ti

de (

cm)

21 22 23 24 25 26 27 28 29 30 31 1 2 3 4 5Time (days)December 1998 January 1999

Figure 4.3 Predicted sea level at Mormugao during 21 December 1998 to 5 January 1999.

4.4 SALT INTRUSION IN MANDOVI AND ZUARI ESTUARIES

Sections such as the ones shown in figure 4.4 are not easily available in the Mandoviand Zuari estuaries. It is known, however, that during the dry season, salinity keepsincreasing in the two estuaries, indicating intrusion of salt. Intrusion of salt in theZuari during the dry season and its flushing during the wet season was first doc-umented by Qasim and Sengupta (1981). Shetye and Murty (1987) modelled thisannual cycle of salinity in the Zuari using a tidally-averaged advection and diffu-sion model. In the model, the estuary was flushed of salt during the wet season byfreshwater entering the estuary at the upstream end. During the dry season, saltintruded upstream owing to horizontal diffusion.

The data on vertical stratification that we have discussed so far, namely, the fourtime series at Old Goa and two salinity sections, suggest a behaviour that is typicalof partially-mixed estuaries. Similar behaviour can be seen, for example, in theobservations reported by Miranda et al. (1998), who studied the Bertioga Channel,Sao Paulo, Brazil. Extensive observations reported by Townbridge et al. (1999)and by Geyer et al. (2000) for the Hudson estuary also reveal similar behaviour. Ahallmark of this behaviour is the difference (often called asymmetry) in stratificationand in circulation that arises between springs and neaps, and within a tidal cycle.The observations in the Mandovi offer an example of the former. It is known thatthey are a consequence of two mechanisms: baroclinic gravity-current formationand tidal streaming (see Stacey et al. 2001, for an overview).

June 13, 2007 18:54 RPS rpb001ch04


0

1

2

3

4

5

6

7

8

9

10

11

12

Dep

th (

m)

0 5 10 15 20 25 30 35 40 45Distance from mouth (km)

2.5

5

7.5

1012.5

1517.52022

.5

25

2527.5

27.53032.5

0

1

2

3

4

5

6

7

8

9

10

11

12

Dep

th (

m)


(15:40)

(15:25)

(15:14)

(15:03)

(14:53)

(14:40)

(14:11)

(13:32)

(13:11)(12:46)

(a)

28/12/98

0

1

2

3

4

5

6

7

8

9

10

11

12

Dep

th (

m)


2.5

5

7.51012.5

1517.5

20

22.5

2527.5

30

3032.5

32.5

(10:22)

(10:46)

(11:11)

(11:33)(12:04)

(13:24)

(13:50)(14:20)

(14:37)

(b)

27/10/97

Figure 4.4 Section of salinity along the length of Mandovi estuary. The horizontal axis givesdistance (km) from the mouth; the vertical axis gives depth in m. In a profile, the depthsat which salinity was measured are shown by crosses. The time of collection of profileis shown in hours at the bottom of the profile location. (a) Section based on 10 sectionscollected during 1245–1540 hours on 28 December 1998; the boat from which the profileswere collected moved towards the mouth. (b) Section based on nine sections collectedduring 1022–1437 hours on 27 October 1997; the boat moved from the mouth towards thehead.

June 13, 2007 18:54 RPS rpb001ch04


When tidal flow is weak, during neaps or when the tide turns, the baroclinic pres-sure gradient associated with the salinity gradient from mouth to head dominatesover the tidal barotropic pressure gradient. This leads to the formation of a gravity(i.e., baroclinic) current that moves saline water upstream near the bottom of theestuarine channel. Fresher water near the surface moves downstream. As a result,stratification builds up. It decays once the tidal current picks up and enough tur-bulence is generated by the interaction of the current with the channel bottom.Associated with the intermittent stratification is intermittent vertical shear. Gravity-current formation enhances shear during ebb and reduces it during flood. Thestrength and duration of the intermittent episodes of stratification and shear forma-tion and decay at a location in the estuary depend on many details that are relatedto the generation of turbulence at the location. The second mechanism that leads toan asymmetry in stratification and in shear is tidal straining (Simpson et al. 1990).In estuaries that have an along-channel salinity gradient and have a vertical shearin the water column, the shear causes fresher water to flow over more saline waterduring ebb. This enhances stratification, stabilizes the water column, suppressesturbulence, and the shear gets enhanced. During flood, the opposite happens: strat-ification is reduced, turbulence is enhanced, and shear is reduced. The asymmetryleads to residual circulation and net transport of salt into the estuary, observationalevidence for which has been given by Jay and Smith (1990) and Jay and Musiak(1994, 1996).

Salt flux into a narrow (in which variations in the cross-channel direction can beignored) estuary owing to advection depends on the vertical integral of the productof along-channel velocity and salinity. The two tidal asymmetries allow the magni-tude of the integral to differ between flood and ebb, or between neap and spring,thus permitting net salt flux into an estuary.

During 1993, water level and salinity were measured in the Mandovi and Zuari estu-aries at 15 locations (see Map C) for three days each during two different months,April and August. The April observations were carried out for 72 hours starting0000 hours on 7 April. The August observations were also carried out for 72 hoursstarting 0000 hours on 19 August. During these observations, salinity was measuredby collecting a water sample from the surface from the bank of the estuary onceevery hour. The April observations provided data on salinity during the second-last month of the dry season from October to May. The August observations pro-vided data on conditions during the second half of the four months of the wetseason from June to September. Figure 4.5 summarizes the data on salinity for theMandovi estuary. The lower panel in the figure permits appreciation of the saltintrusion into the estuary during the dry season. Figure 4.6 does the same for theZuari.

June 13, 2007 18:54 RPS rpb001ch04


0

5

10

15

20

25

30

35

August 1993S

alin

ity (

psu)

19 20 21

1

2

3

April 19937 8 9

1

2

3

4

0

5

10

15

20

25

30

35

Distance (km)

Sal

inity

(ps

u)

0 10 20 30 40 50 60

1 2 3 4 5

April 1993

August 1993

Figure 4.5 Top left: Salinity (psu) observed during 19–21 August 1993 (spring tide) in theZuari at the 5 locations, numbered 1–5, shown in Map C. The salinity corresponding toeach location is identified in the figure with the same number as in Map C. The horizontalaxis shown at the top of the figure gives the time of observation during the three days ofobservation. Top right: Same as top left except during 7–9 April 1993. Bottom: The twodark lines show the mean salinity (psu) during the three days of observations as a functionof distance (km) along the length of the channel during April and August. Arrowheads atthe top show the locations where observations were carried out. The grey bands above andbelow the dark line give the range of salinity (psu) during the observations.

June 13, 2007 18:54 RPS rpb001ch04


0

5

10

15

20

25

30

35

August 1993S

alin

ity (

psu)

19 20 21

6

7

8

April 19937 8 9

6

7

8

9

10

11

0

5

10

15

20

25

30

35

Distance (km)

Sal

inity

(ps

u)

0 10 20 30 40 50 60

6 7 8 9 10 11

April 1993

August 1993

Figure 4.6 Same as figure 4.5, except that the salinity (psu) was observed in the Mandovi atthe 6 locations, stations 6–11, shown in Map C. The salinity corresponding to each locationis identified in the figure with the same number as in Map C.

June 13, 2007 18:54 RPS rpb001ch04


4.5 CONCLUDING COMMENTS

The data presented in chapters 1 and 2 and in the present chapter allow us toclassify the Mandovi and Zuari estuaries. By virtue of their location and shape oftheir channels, they are coastal-plain estuaries with convergent channels. By natureof tides, with spring tides in excess of 2m, they are macrotidal. In view of the patternof rainfall experienced by the estuaries, they are positive estuaries during the wetseason. It appears that as the dry season progresses and freshwater influx decreases,the estuaries do not turn into negative estuaries even at the peak of the dry season.On the basis of data on vertical stratification presented in the present chapter, theMandovi and Zuari are partially-stratified estuaries.

Finally, Indian estuaries have often been described as “monsoonal”, but the precisemeaning of the word has not been stated. The description of runoff in the Mandoviand Zuari estuaries given in chapters 1, and of salinity variation given in the presentchapter, helps us to define the term better. At the end of the monsoon, the waterin the Mandovi and Zuari is completely changed due to replacement of the pre-monsoon water by freshwater flowing through the estuaries. This replacement iscomplete because the estuaries are flushed many times over. At the end of themonsoon, the estuarine processes start their eight-month-long dry-season evolution,in which horizontal mixing due to processes common in partially-mixed estuariesleads to mixing of freshwater at the upstream end, i.e., remnants of the water broughtin by monsoonal runoff, and the saline water at the sea end. During the four-month wet season, evolution of the estuary is dominated by spurts of freshwaterthat flush the estuary completely. This contrast between the temporal evolution ofMandovi and Zuari estuaries between the dry season and the wet season justifiestheir classification as “monsoonal”.

June 13, 2007 18:54 RPS rpb001ch05

5Variability of nitrateand phosphate

S. Sardessai and D. SundarNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

5.1 INTRODUCTION

Nitrate and phosphate are important elements, especially as nutrients for phyto-plankton and phytobenthos, of the biogeochemical system of an estuary. Observa-tions carried out during 2002–2003 as a part of an integrated coastal and marinearea management plan for Goa provided an opportunity to collect data to describetemporal and spatial variability of nitrate and phosphate in the Mandovi andZuari estuaries. An earlier study (Qasim and Sengupta 1981) on nutrients in theseestuaries reported that nitrate concentration decreases from head to mouth in thedry season. Qasim and Sengupta attributed this decrease to utilization of nitrateby the photosynthetic activity in the estuaries. In another study, De Sousa (1983)examined the behaviour of nutrients in the Mandovi estuary during the premon-soon season and observed that mining rejects form a source for nitrate. In thischapter, we describe observations on nitrate and phosphate carried out during2002–2003. The next section describes the sampling details. Sections 5.3 and 5.4discuss principal features seen in the observations and the processes underlying theobservations.

59

June 13, 2007 18:54 RPS rpb001ch05

60 S. SARDESSAI AND D. SUNDAR

5.2 OBSERVATIONS

Water samples were collected during April–May 2002, September 2002, and March2003 as shown in the schedule in table 5.1. Samples were taken at seven stations(M1–M7) in the Mandovi and at five stations in Zuari (Z1–Z5) (Map B). The locationscovered a stretch of about 50km in the Mandovi and 52km in the Zuari. Sampleswere also taken from one coastal station named MZ (Map B). Surface samples forsalinity and nutrients (nitrate, nitrite, and phosphate) were collected every threehours for 24 hours. Tidal data were collected every 15min for 36 hours so thatthere was an overlap between nutrient and tide observations. Salinity was estimatedusing a Guildeline salinometer (AUTOSAL 8400A). Nitrate and phosphate weremeasured by following the methods of Grasshoff et al. (1983).

5.3 VARIABILITY OF NITRATE

Figures 5.1 and 5.2 summarize the observed nitrate concentrations in the Mandoviand Zuari. Of the four periods of observation listed in table 5.1, nitrate concentra-tion was the highest during September 2002. Though September is the last monthof the four month-long normal monsoon season, runoff during this month is stillhigh compared to the other three periods of observation. The high nitrate concen-tration observed during this period is due to the runoff. The value at station M1,which is at the upstream end of the Mandovi estuary, was 4.4µM. At station M2,the downstream station closest to M1, nitrate concentration increased by 2µM to6.4µM. Such an increase could be attributed to one or more of the following threefactors.

• First, the Mandovi and Zuari estuarine banks have mangroves, which harboursediments rich in organic matter (Wafar 1987). Wafar et al. (1997) noted thatthe dissolved organic nitrates (DON) and dissolved organic phosphates (DOP)

Table 5.1 Schedule of sample collection in the Mandovi and Zuari estuaries.

Season Estuary Date of collection

Premonsoon Mandovi Spring tide (28–29/4/2002)Premonsoon Mandovi Neap tide (5–6/5/2002)Monsoon Mandovi 5–6/9/2002Premonsoon 2003 Mandovi 12–13/3/2003Premonsoon 2002 Zuari Spring tide (30/4–1/5/2002)Premonsoon 2002 Zuari Neap tide (7–8/5/2002)Monsoon Zuari 7–8/9/2002Premonsoon 2003 Zuari 14–15/3/2003

June 13, 2007 18:54 RPS rpb001ch05

VARIABILITY OF NITRATE AND PHOSPHATE 61

0

2

4

6

8

0 5 10 15 20 25 30 35 40 45 50

Nitr

ate

(μM

l−1)

Distance from mouth (km)

28/29 April 2002 5/6 May 20025/6 September 2002 12/13 March 2003

Figure 5.1 Variation of concentration of nitrate in the Mandovi estuary with distance fromthe mouth. The colour key that defines the period of observation is given at the top of thefigure.

0

2

4

6

8

0 5 10 15 20 25 30 35 40 45 50 55

Nitr

ate

(μM

l−1)


30/01 April/May 2002 7/8 May 20027/8 September 2002 14/15 March 2003

Figure 5.2 Variation of concentration of nitrate in the Zuari estuary with distance fromthe mouth. The colour key that defines the period of observation is given at the top of thefigure.

June 13, 2007 18:54 RPS rpb001ch05


released from the mangrove litter are important for sustaining the nutrient bud-get of the Mandovi and Zuari estuaries. Our observations also suggested thesignificant presence of ammonia and DON in these estuarine waters. Hence, thepassage of freshwater through mangroves could supply substantial amounts ofnitrogen to the estuaries.

• Second, nitrate concentration can increase owing to mining of iron and man-ganese ore in the vicinity of the Mandovi and Zuari estuaries. De Sousa (1983)has pointed out that mining rejects that end up in the estuaries can serve as asource of nitrates. The explosives used in mining operations also contain nitrates,and they end up in the mining rejects that the runoff carries to the estuaries.

• Third, the Mandovi and Zuari flow over a fertile coastal plain that supports ricecultivation and horticulture. Both these activities involve the use of fertilizers withnitrates as a component, and the runoff during the summer monsoon is expectedto carry some of these nitrates to the two estuaries.

During September 2002, nitrate concentration in the Mandovi did not change sig-nificantly between station M2 and M5, i.e., over a 20km stretch of the estuary (seeMap B). There was a drop in the concentration at the mouth. Climatological data(Naqvi et al. 2006, for example) show that nitrate concentration on the shelf isgenerally less than 1µM throughout the year. It is well known, however, that thewest coast of India experiences upwelling during the summer monsoon, leading toenhancement of nitrate. Hence, whenever an episode of upwelling occurs, it canbe expected that nutrient values would increase. Qasim and Sengupta (1981) havenoted that the impact of such an increase is restricted to less than 10km from themouth, covering mainly the bays of the two estuaries. During September 2002,mixing of the waters at the mouth and the high-nitrate waters farther upstream ledto a sharp increase in nitrates in the first few kilometres from the mouth of theMandovi.

Thus, at the end of the summer monsoon, the channels of the Mandovi and Zuarihave nitrate concentrations of the order of 6µM, which is at least a few times theconcentration found on the shelf. During this time, there is also enough light to sup-port primary productivity, which would then lead to productivity at higher trophiclevels and biogeochemical cycling of nutrients. Observations in the Mandovi dur-ing March 2003 showed that nitrate concentrations were less than 1µM throughoutthe estuary. It is expected that the observed decrease in nitrate concentration fromSeptember 2002 to March 2003 could have resulted from the following processes:

• First, the nitrate concentration on the shelf off the mouths of Mandovi andZuari remains less than 1µM throughout the dry season between withdrawalof the monsoon in September or October and the onset of the next monsoon inJune. Throughout this period, the properties of the waters at the mouth migrateupstream owing to horizontal mixing. One consequence of this mixing is a steadyrise in salinity in the channels of the Mandovi and Zuari. Since the concentration

June 13, 2007 18:54 RPS rpb001ch05


of nitrates remains low throughout the dry season, horizontal mixing could leadto a decrease in nitrates throughout the two estuaries.

• Second, biogeochemical cycling of nitrates is an important process in determin-ing levels of nitrate concentration in the two estuaries during the dry season. Aquantitative description of this process is not available at present. Expected toplay an important role in the cycling are two mechanisms. The first of these hasbeen mentioned earlier: the release of nitrates by mangroves on the banks of theestuaries. The second is the release of nitrates owing to suspension of sedimentsresulting from vertical mixing, which can be triggered off whenever tidal currentsare high. Release of nitrogen from mangroves or resuspension of sediments occurprimarily in the form of ammonia. The released ammonia quickly gets oxidizedto nitrate, a process called nitrification, in oxygen rich estuarine waters. Anotherpossibility is mixing events triggered by episodes of strong winds.

The data presented here show that the net result of the processes listed here —-biogeochemical cycling, horizontal mixing, and release of nitrates from mangrovesand sediments —- is the reduction in the concentration of nitrates from about 6µMin most of the estuary during the withdrawal phase of the monsoon to less than1µM in the last few months of the dry season. The role of horizontal mixing in thisdecrease of nitrate concentration can be appreciated by examining the behaviourof the estuary in a nitrate–salinity diagram (figure 5.3a). As seen from this fig-ure, the conditions in the main channel of the Mandovi, that is upstream of about10km from the mouth, during September 2002 are given by nitrates and salini-ties that are centred around 6µM of nitrate and 0psu. The state of this water inthe nitrate–salinity diagram is depicted by the semi-circle shown in the schematic(figure 5.3b). The state at the mouth during the same time is represented by thecircle centred on 0.5µM, and 35psu in the schematic. Had horizontal mixing beenthe dominant process in the evolution of nitrate concentration in the estuary, asis the case with salinity distribution, the state of the estuary in the nitrate–salinitydiagram would have always been represented in the dry season by points lying ona line joining the centres of the semi-circle and the circle, i.e., by the dashed linein the schematic (figure 5.3b). Observations during March 2003 and those duringApril–May 2002 show that the state of the estuary during these months cannot berepresented by the dashed line. Instead, it is represented by the solid horizontalline joining (0.5µM, 0psu) and (0.5µM, 35psu) (figure 5.3b). Hence, removal ofnitrate is much more than what might have been expected by horizontal mixingalone. The two other processes that are important during the dry season —- releaseof nitrates from mangroves and from sediments —- can only increase nitrate con-centration. That the observed line, the solid line in figure 5.3(b), is well below thedashed line implies that removal of nitrates by biogeochemical processes must beimportant in the evolution of nitrate in the estuary. Nitrate is removed from estu-arine water by phytoplankton and phytobenthos and by bacterial respiration insediments. Bacteria utilize nitrate as oxidant, in the absence of oxygen, during thedecomposition of organic matter. The intensity of this process will be greater in

June 13, 2007 18:54 RPS rpb001ch05


0

2

4

6

8

0 5 10 15 20 25 30 35 40

Nitr

ate

(μM

l−1)

(a)

Salinity (psu)


0

2

4

6

8

0 5 10 15 20 25 30 35 40

Nitr

ate

(μM

l−1)

Salinity (psu)

(b)

Figure 5.3 (a) Nitrate–salinity diagram of the waters of the Mandovi estuary. The colourkey that defines the period of observation is given at the top of the figure. (b) The semi-circlein the schematic shows the state of the channel during the summer monsoon in the channelof the Mandovi in the nitrate–salinity space. The circle indicates the state at the mouthduring the same time. The horizontal line gives the state of the estuary during March–May. The dashed line shows the curve that would have indicated the state of the estuary inMarch–May had horizontal mixing alone dominated evolution of nitrate in the estuary.

mangrove sediments that harbour very high populations of bacteria as the pres-ence of organic matter is very high. Thus figure 5.3 clearly reveals the dominationof nitrate removal over its release processes in the estuary. On average, the con-centration of nitrate during March 2003 was lower than the concentration during

June 13, 2007 18:54 RPS rpb001ch05


April–May 2002. Hence, there is significant interannual variability associated withthis variable.

Virtually everything that has been said above about the Mandovi applies to theZuari as well. Figure 5.2 should therefore be examined from the point of viewpresented above. In addition, it is worth noting here that runoff in Mandovi ishigher than that in the Zuari, but the volume of the Mandovi estuary is less thanthat of the Zuari. It is therefore expected that nitrate brought by runoff will play abigger role in the nitrate budget of Mandovi when compared to that of the Zuari.

5.4 VARIABILITY OF PHOSPHATE

Unlike nitrate, which had a distinctly higher concentration during September 2002compared to other periods of observation in both estuaries, the phosphate con-centration did not show much variation (figures 5.4 and 5.5). Concentration ofphosphate in river runoff, as seen from the values observed at stations M1 and Z1,was not significantly different from that observed at the mouths of the two estuariesduring September 2002. Other observations did not reveal any spatial or temporalpattern. Phosphate concentration is an important component of biogeochemicalcycling in an estuary. We have earlier concluded that this cycling must be active to

0

1

2

3

0 5 10 15 20 25 30 35 40 45 50

Pho

spha

te (

μM l−1

)



Figure 5.4 Variation of concentration of phosphate in the Mandovi estuary with distancefrom the mouth. The colour key that defines the period of observation is given at the topof the figure.

June 13, 2007 18:54 RPS rpb001ch05


0

1

2

0 5 10 15 20 25 30 35 40 45 50 55

Pho

spha

te (

μM l−1

)


30/01 April/May 2002 7/8 May 20027/8 September 2002 14/15 March 2003

Figure 5.5 Variation of concentration of phosphate in the Mandovi estuary with distancefrom the mouth. The colour key that defines the period of observation is given at the topof the figure.

remove nitrates from the two estuaries during the dry season. For instance, phos-phate can be released to water by stirring of mud (Sankaranarayanan and Qasim1969). Phosphate concentration too could be playing a role in the biogeochemicalprocesses. The data available at present, however, are insufficient for identifyingany pattern of variability or assigning a role to phosphate concentration in theseestuaries.

June 14, 2007 12:55 RPS rpb001ch06

6Phytoplankton diversity,biomass, and production

S. G. Prabhu Matondkar, Helga do R. Gomes∗,Sushma G. Parab, Suraksha Pednekar,and Joaquim I. Goes∗National Institute of Oceanography, Dona Paula, Goa, 403 004, India.∗Bigelow Laboratory for Ocean Sciences, West Boothbay Harbor, Maine, 04575, USA.

6.1 INTRODUCTION

Phytoplankton, chlorophyll a (Chl), and primary productivity (PP) are importantdeterminants of the standing stock and organic productivity of an estuary. Thesevariables have often been utilized to evaluate the overall health of an estuarineecosystem and in management strategies to ensure sustainable use of the estuaries.Earlier investigations on the three variables in the Mandovi and Zuari estuariesinclude studies by Devassy and Goes (1988, 1989) and Krishnakumari et al. (2002),who proposed a three-season regime in the biology of the estuaries; other authorsnoted a similar pattern in primary production (Dehadrai and Bhargava 1972; Bhat-tathiri et al. 1976; Devassy 1983). More recently, in 2002–2003, field observationswere carried out to measure phytoplankton composition, Chl, and PP in the twoestuaries to develop an estuarine and coastal zone management plan for Goa. In thischapter, we use these data to describe the spatial and temporal variability of thesethree variables with the objective of gaining insight into the processes regulatingprimary productivity in this estuarine ecosystem.

67

June 14, 2007 12:55 RPS rpb001ch06

68 S. G. PRABHU MATONDKAR ET AL.

6.2 OBSERVATIONS

Water samples for phytoplankton taxonomy and cell counts, Chl, and rates of PPwere collected at stations M1, M2, M3, M4, M5, M6, and M7 in the Mandovi andZ1, Z2, Z3, Z4, and Z5 in the Zuari (Map A). The sampling schedule is shown intable 6.1. The stations covered a 40km stretch in each estuary. During each setof observations, eight samples at 3-hour intervals were collected during a 24-hourperiod, thus nearly covering a tidal cycle. The variation among the eight samplesprovided an estimate of the change seen at a location owing to horizontal movementof water parcels due to the tide. The observations carried out in September 2002are considered to be representative of conditions at the end of the high runoff of thesummer monsoon during June–September. The April–May 2002 observations, onthe other hand, are representative of conditions at the end of the dry season fromOctober to May, prior to the onset of the wet season.

For phytoplankton taxonomy and cell counts, 500ml water samples were fixed witha few drops of Lugol’s iodine, preserved in 3% buffered formaldehyde, and thenstored in dark and cool conditions until the time of analysis. Prior to microscopicanalysis, samples were concentrated to 5–10ml by siphoning the top layer with atube covered with a 10µm Nytex filter on one end. Sample concentrates were trans-ferred to a 1ml capacity Sedgwick-Rafter and counted using an Olympus® Invertedmicroscope (Model IX 50) at 200×magnification. Phytoplankton cell identificationswere based on standard taxonomic keys (Tomas 1997). The results are expressedas numbers of cells l−1.

Chl was measured by filtering 500ml water samples on to 47mm glass fibre filters(Whatman® GF/F), which were extracted overnight in 10ml of 90% acetone undercold and dark conditions. Chl concentrations in the extracts were determined witha Turner Designs® 10-000R fluorometer ( JGOFS 1994). Acidification of the extractsallowed corrections for phaeophytin. The fluorometer was periodically calibratedwith a spectrophotometer according to protocols recommended in JGOFS (1994).

Table 6.1 Schedule of observations in the Mandovi and Zuari estuaries.

Sampling timeNatureof tide Mandovi Zuari

Spring 28 (13:00h) to 29 (22:00h)April 2002

30 (12:00h) April to 1 (22:00h)May 2002

Neap 5 (12:00h) to 6 (22:30)May 2002

7 (11:00h) to 8 (22:00h) May2002

Spring 5 (10:50h) to 6 (10:00h)September 2002

7 (11:00h) to 8 (10:00h)September 2002

June 14, 2007 12:55 RPS rpb001ch06

PHYTOPLANKTON DIVERSITY, BIOMASS, AND PRODUCTION 69

Primary productivity was estimated using the 14C technique under simulated in situconditions ( JGOFS 1994). At the end of the incubations in running surface seawaterthat lasted 24 hours, samples were filtered on to 47mm, 0.45 micron cellulosenitrate filters, which were then exposed to HCl fumes and then transferred to liquidscintillation vials containing 10ml of scintillation cocktail. The activity in each vialwas determined in a Wallac® Scintillation Counter (Model no. 1409). All countswere quench-corrected prior to determinations of PP.

6.3 CHL, PHYTOPLANKTON CELLS, AND PP

Figure 6.1 gives Chl measured during spring, and neap tide measurements of Apriland May 2002. The two measurements were separated by about a week in eachestuary. The figure shows that, except for one or two locations, the variations in Chlover the one-week period were not significant. In general, values of Chl were highertowards the upstream end of the Zuari, whereas the highest values in the Mandovi

02468

1012141618

Chl

a (

mg

m−3

)

0 10 20 30 40 50Distance from mouth (km)

Mandovi Neap tide

02468

1012141618

Chl

a (

mg

m−3

)


Zuari Neap tide

02468

1012141618

Chl

a (

mg

m−3

)


Mandovi Spring tide

02468

1012141618

Chl

a (

mg

m−3

)0 10 20 30 40 50


Zuari Spring tide

Figure 6.1 Comparison of spring and neap tide phytoplankton biomass in the Mandoviand Zuari estuaries during April–May 2002.

June 14, 2007 12:55 RPS rpb001ch06


0

2

4

6

8

Phy

topl

ankt

on c

ount

s (c

ells

X 1

05l−1

)


Mandovi: September 2002

0

2

4

6

8

Phy

topl

ankt

on c

ount

s (c

ells

X 1

05l−1

)


Zuari: September 2002

0

2

4

6

8

Phy

topl

ankt

on c

ount

s (c

ells

X 1

05l−1

)


Mandovi: April 2002

0

2

4

6

8

Phy

topl

ankt

on c

ount

s (c

ells

X 1

05l−1

)


Zuari: April 2002

Figure 6.2 Phytoplankton cell counts in the Mandovi (left panels) and Zuari (right panels)during April (top) and September (bottom) 2002.

were in the middle portion at a distance of 15 to 25km from the mouth. Figure 6.2gives phytoplankton cells during the spring tide observations of April 2002 andSeptember 2002. The number of cells was generally lower during September thanduring April.

Taxonomic groups of the phytoplankton identified in the Mandovi and Zuari riversare presented in tables 6.2 and 6.3. In April, diatoms such as Nitzschia spp. andSynedra sp. dominated the freshwater zone of the Mandovi (table 6.2a), while Rhi-zosolenia spp. and Coscinodiscus spp. were abundant in the high-salinity zone. In theZuari, the diatoms Nitzschia spp. and Bacteriastrum sp. were the dominant organismsin the low-salinity region, while Coscinodiscus spp. was the important species in thehigh-salinity region (table 6.2b). In September 2002, Synedra sp. and Nitzschia sp.prevailed in the freshwater zone, while Dinophysis spp. and Coscinodiscus spp. werepredominant in the high-salinity zone of the Mandovi (table 6.3a). In the Zuariestuary during this time of the year, the freshwater stations were dominated byPheodactylum sp., while the saline stations showed the preponderance of Biddulphiasp. and Rhizosolenia spp. (table 6.3b).

June 14, 2007 12:55 RPS rpb001ch06


Table 6.2(a) Dominant phytoplankton species during April 2002 in theMandovi.

Station Dominant speciesnos.

M1 Peridinium sp. Prorocentrum sp. Synedra sp.

M2 Nitzschia spp. Melosira sp. Actinocyclus sp.

M3 Nitzschia spp. Melosira sp. Synedra sp.

M4 Rhizosolenia spp. Coscinodiscus spp. Nitzschia spp.

M5 Rhizosolenia sp. Nitzschia sp. Chaetoceros spp.

M6 Rhizosolenia spp. Coscinodiscus spp. Chaetoceros spp.

M7 Coscinodiscus spp. Cerataulina sp. Dinophysis sp.

Table 6.2(b) Dominant phytoplankton species during April 2002 in the Zuari.


Z1 Bacteriastrum sp. Nitzschia spp.

Z2 Coscinodiscus spp. Melosira sp. Nitzschia spp.

Z3 Nitzschia spp. Thalassionema sp. Rhizosolenia sp.

Z4 Nitzschia spp. Isthmia sp. Coscinodiscus sp.

Z5 Coscinodiscus spp Ceratulina sp. Nitzschia spp.

Table 6.3(a) Dominant phytoplankton species during September 2002 in theMandovi.


M1 Synedra sp. Enotia sp. Nitzschia spp.

M2 Nitzschia spp. Navicula spp. Synedra sp.

M3 Synedra sp. Nitzschia spp. Dinophysis sp.

M4 Thalassiosira sp. Dinophysis sp. Synedra sp.

M5 Thalassiosira sp. Dinophysis sp. Nitzschia spp.

M6 Dinophysis sp. Nitzschia spp. Peridinium sp.

M7 Coscinodiscus spp. Rhizosolenia spp. Nitzschia spp.

June 14, 2007 12:55 RPS rpb001ch06


Table 6.3(b) Dominant phytoplankton species during September 2002 in theZuari.


Z1 Phaeodactylum sp. Distephanus sp. Navicula spp.

Z2 Phaeodactylum sp. Nitzschia spp. Peridinium sp.

Z3 Planktoniella sp. Coscinodiscus sp. Phaeodactylum sp.

Z4 Biddulphia sp. Nitzschia spp. Thalassiosira sp.

Z5 Rhizosolenia spp. Chaetoceros spp. Ceratulina sp.

Table 6.4(a) Diversity index during April 2002 in the Mandovi. In the table thesamples that were collected at the beginning of the 24-hour period of observationsare shown in the column named “0 hours”. The next sets of samples were collected6, 12, and 18 hours later and are identified accordingly.

Station 0h 6h 12h 18hnos.

M1 2.8 – – –M2 1.7 2.3 3.4 1.7M3 2.5 2.5 2.2 3.0M4 3.3 3.1 – 2.5M5 2.5 2.1 2.1 2.2M6 3.2 2.6 2.5 2.1M7 3.2 3.0 2.8 2.2

Table 6.4(b) Same as table 6.4(a), but for September 2002.


M1 1.08 – – –M2 1.5 2.1 2.1 1.5M3 1.4 2.3 2.1 –M4 2.8 2.8 2.8 2.8M5 2.8 2.7 2.9 2.4M6 3.2 2.7 2.3 3.3M7 3.4 3.2 3.0 3.3

Diversity indices, which give a measure of phytoplankton species diversity duringa 24-hour period of observation, are presented in tables 6.4 and 6.5. Although theoverall range of values did not vary appreciably between the two rivers, subtle dif-ferences were observed between the two. In the Mandovi, for instance (table 6.4a),

June 14, 2007 12:55 RPS rpb001ch06


Table 6.5(a) Diversity index during April 2002 in the Zuari. In the table, thesamples that were collected at the beginning of the 24-hour period of observationsare shown in the column named “0 hours”. The next sets of samples were collected6, 12, and 18 hours later and are identified accordingly.


Z1 0.92 – – –Z2 1.82 1.69 1.72 1.96Z3 1.53 1.73 1.95 1.95Z4 2.61 2.61 2.34 2.14Z5 2.24 3.25 2.24 2.25

Table 6.5(b) Same as table 6.5(a), but for September 2002.


Z1 2.03 – – –Z2 1.18 1.48 1.27 3.03Z3 2.74 2.52 2.30 2.87Z4 3.15 2.72 2.50 –Z5 3.24 3.48 2.95 3.08

*– indicates samples not collected.

a low diversity index of 1.7 at station M2 in April 2002 was the result of a Nitzschiabloom in the freshwater zone. However, at all other stations, a diversity index >2was observed, indicative of the diverse nature of phytoplankton in April 2002. Dur-ing September, phytoplankton populations at stations M1, M2, and M3 were char-acterized by a low diversity index (<2.0). These three stations supported blooms ofSynedra sp. and Nitzschia spp. During April, the low value of 0.92 in the Zuari wasdue to the dominance of Bacteriastrum sp. and Nitzschia spp. at St. Z1 (table 6.5a). Thelow diversity index at Z2 in September was due to Coscinodiscus spp. and Melosirasp. blooms (table 6.5b).

Cluster analysis of phytoplankton species data based on similarity indices for theMandovi and the Zuari are given in figure 6.3 for the samples collected duringApril 2002 and in figure 6.4 for the samples collected during September 2002. Thefigures show that Nitzschia spp., Melosira sp., and Synedra sp. form a cluster witha preference for low-salinity water, while Coscinodiscus spp. and Rhizosolenia spp.formed another cluster in higher-salinity areas. In the Zuari, in April, Coscinodiscusspp. and Nitzschia spp. formed a separate cluster indicating their growth preferencefor the low-salinity zone, whereas the cluster of Rhizosolenia spp. and Thalassiosira sp.reflected their preference for the saline region. In September, when the salinity in

June 14, 2007 12:55 RPS rpb001ch06


Figure 6.3(a) Phytoplankton similarity index for the Mandovi estuary during April–May2002.

the Mandovi was low, Synedra sp., Nitzschia spp., and Dinophysis sp. formed a singlecluster, reflecting their preference for low salinities (figure 6.4a). In contrast, thecluster of Coscinodiscus spp., Gymnodinium sp., and Rhizosolenia spp. was due to theirpreference for higher-salinity waters that were present at the mouth of the Mandovi(figure 6.4a). In the Zuari, in September, Phaeodactylum sp. and Distephanus sp.formed a separate cluster due to their preference for low-saline waters (figure 6.4b).In contrast, the close association between Rhizosolenia spp. and Chaetoceros spp wasrelated to their preference for higher salinity.

June 14, 2007 12:55 RPS rpb001ch06


Figure 6.3(b) Phytoplankton similarity index for the Zuari estuary during April–May 2002.

Variations in Chl in each of the rivers closely resembled the variations of phyto-plankton cell counts (figure 6.5). Chl concentrations were clearly higher in both theMandovi and Zuari during April than in September. In April, Chl concentrationsvaried from 2.24 to 7.05mgm−3 in the Mandovi and from 1.74 to 11.44mgm−3

in the Zuari. During September, Chl a in Mandovi varied over a low and narrowrange of 0.09 to 1.25mgm−3 and from 1.72 to 3.97mgm−3 in the Zuari. Chl con-centrations were <2mgm−3 at stations closer to the mouth. A noteworthy featurein the observations was that Chl concentration in the Zuari during April 2002 washighest in the low-saline waters near the upstream end. This was different from

June 14, 2007 12:55 RPS rpb001ch06


Figure 6.4(a) Phytoplankton similarity index for the Mandovi estuary during September2002.

the conditions in the Mandovi, where Chl was generally higher in euryhaline (10to 30psu) waters, as can be seen in figure 6.5. In the Zuari, Chl increased frommouth to head during April–May 2002, leading to a gradient along the length ofthe estuary. By September 2002, however, the gradient no more existed (figure 6.5).Unlike the distribution of concentration of Chl, rates of PP in September 2002 didnot differ significantly from those in April 2002 (figure 6.6). Variation in the rates ofPP over a tidal cycle during September was small, implying that the spatial gradientin PP along the two estuaries was weak at this time.

June 14, 2007 12:55 RPS rpb001ch06


Figure 6.4(b) Phytoplankton similarity index for the Zuari estuary during September 2002.

6.4 DISCUSSION

Phytoplankton counts in both estuaries were higher in April than in September atthe end of the monsoon, when both the estuaries are flushed many times over. As aresult of this freshwater flushing and changes in salinity, the two rivers experiencethe most dramatic changes in phytoplankton species composition, as is evidentfrom the taxonomic data. Phytoplankton counts in the Mandovi during Septemberranged from 0.68–1.36 × 105 cells l−1; the range of cell numbers was much larger

June 14, 2007 12:55 RPS rpb001ch06


02468

1012141618

Chl

a (

mg

m−3

)



02468

1012141618

Chl

a (

mg

m−3

)



02468

1012141618

Chl

a (

mg

m−3

)


Mandovi: April 2002

02468

1012141618

Chl

a (

mg

m−3

)


Zuari: April 2002

Figure 6.5 Chlorophyll a concentration in the Mandovi (left panels) and Zuari (right pan-els) during April (top) and September (bottom) 2002.

in the Zuari (1.02–3.07 × 105 cells l−1), suggesting that the phytoplankton species inthe Zuari were more amenable to freshwater as compared to those in the Mandovi.The presence of a prominent community of freshwater phytoplankton was evidentfrom the high Chl values associated with low-salinity waters in the Zuari. These dataalso revealed a distinct population of euryhaline phytoplankton in the Mandovi,whose biomass was higher by a factor of two in comparison to the euryhalinepopulation in the Zuari. These features are evident in figure 6.7, which gives thedistribution of phytoplankton biomass with salinity in the Mandovi and Zuari. Inaddition, the rates of PP, normalized to unit biomass of phytoplankton, revealedthat phytoplankton communities present in the Zuari were photosynthetically moreefficient at low salinities than the populations from the Mandovi (figure 6.8). Theexistence of physiological differences between phytoplankton populations of theMandovi and the Zuari is important because it suggests that no similarities shouldbe assumed in the biological characteristics of two estuarine ecosystems merelybecause of geographical proximity.

June 14, 2007 12:55 RPS rpb001ch06


0

20

40

60

80

100

120

PP

(m

gC m

−3d−1

)



0

20

40

60

80

100

120

PP

(m

gC m

−3d−1

)



0

20

40

60

80

100

120

PP

(m

gC m

−3d−1

)


Mandovi: April 2002

0

20

40

60

80

100

120

PP

(m

gC m

−3d−1

)


Zuari: April 2002

Figure 6.6 Primary productivity in the Mandovi (left panels) and Zuari (right panels) estu-aries during April (top) and September (bottom) 2002.

0

5

10

15

20

0 10 20 30 40

Chl

a (

mg

m−3

)

Mandovi

Salinity (psu)

0

5

10

15

20

0 10 20 30 40

Chl

a (

mg

m−3

)

Zuari

Salinity (psu)

Figure 6.7 Distribution of chlorophyll in relation to salinity. Data plotted are from all theobservations that are listed in table 6.1. Points encircled are outliers.

The presence of freshwater-tolerant phytoplankton allows the Zuari to remain moreproductive upstream, possibly for a greater part of the year. This could make it anet exporter of organic material downstream and perhaps into the coastal waters.By and large, the Mandovi was dominated by a euryhaline population whose

June 14, 2007 12:55 RPS rpb001ch06


0

5

10

15

20

25

30

0 10 20 30 40

PP

Ch

l−1

(mg

C C

h l−1

d−1)

Mandovi

Salinity (psu)

0

25

50

75

100

0 10 20 30 40

Zuari

r2 = 0.86

Salinity (psu)

Figure 6.8 Comparison of Chl normalized PP in relation to salinity for the Mandovi (left)and Zuari (right). Data plotted are from all three phases of this study. The curve fitted to thescatter plot for the Zuari and the correlation coefficient shown in the top right corner of thepanel indicates the statistical significance of the relationship between the two variables forthe Zuari estuary. No such statistically significant relation was apparent between these twovariables for the Mandovi estuary.

biomass in both phases of our study was almost twice that of the Zuari. Thereis evidence from both controlled laboratory experiments (Qasim 1977) and fieldobservations (Devassy and Goes 1989), that coastal tropical marine phytoplank-ton species are capable of growing at faster rates when transferred to waters withslightly reduced salinity. The adaptability of tropical coastal phytoplankton to salin-ity reductions corresponding to conditions brought about by tidal mixing and riveroutflow could explain the higher biomass and production rates midway upstreaminto the Mandovi. This, together with close similarity to phytoplankton communi-ties at the mouth of the estuary and upstream, could imply that Mandovi is a netimporter of organic material.

It may be noted that the cell numbers observed during the present work are muchhigher than those reported from both the Mandovi and Zuari by Devassy andGoes (1988, 1989). This, however, may simply be attributed to the frequency ofsampling, which, in the present study, was much higher than the single monthlysamples undertaken by Devassy and Goes (1988, 1989). The phytoplankton speciesobserved during the present work and their diversity in both the Mandovi andZuari rivers also represent a change from the observations of Devassy (1983) andof Devassy and Goes (1989). The diversity indices of 0.96 during April and of 1.1to 1.7 during September noted by Devassy and Goes (1989) are much lower thanthe values observed by us for both the Mandovi and Zuari. Although differencesin frequency of sampling between the present study and the studies of Devassyand Goes (1989) could be a factor in the discrepancy of the diversity indices, it ispossible that the large presence of Trichodesmium erythraeum in April and the bloomsof Fragillaria oceanica and Ceratium furca in September could have contributed to

June 14, 2007 12:55 RPS rpb001ch06


the lower diversity indices observed by Devassy and Goes (1989). In the presentstudy, we did not observe these species in large numbers.

With growth and turnover rates of less than a day, phytoplankton are very sensi-tive to changes in the environment, and large variations in phytoplankton speciescomposition are often a reflection of significant alteration in ambient conditionswithin an ecosystem. This study, undertaken more than a decade after the seminalstudies of Devassy and Goes (1988, 1989) and others, indicates that phytoplanktoncharacteristics of the Mandovi and Zuari could be changing. More detailed and sys-tematic investigations of natural and anthropogenic variables that are contributingto changes in phytoplankton physiology and growth are required. Assessing changesin land-use patterns, influx of fertilizers from agricultural runoff, sewage intrusions,etc., in relation to phytoplankton standing stocks and productivity over the yearswill certainly provide vital information on the health of this estuarine ecosystem.This information will be critical for planning and management strategies requiredfor ensuring sustainable use of the two estuaries. Our study also emphasizes that itmay be difficult to extrapolate the findings from a single estuary to other estuariesalong the west coast of India in spite of their geographical proximity and similarityin physical features.

June 13, 2007 18:55 RPS rpb001ch07

7Zooplankton

R. Jyothibabu and N. V. MadhuNational Institute of Oceanography, Regional Centre, Kochi 682 018, India.

7.1 INTRODUCTION

Zooplankton, also known as secondary producers, are tiny floating or drifting ani-mals of aquatic systems. While a few zooplankton are capable of independent move-ment vertically (a behaviour called vertical migration), their horizontal movementis primarily determined by water currents. These animalas play an important rolein aquatic systems as intermediate members of the food web. In this chapter, wereview what is known about zooplankton in the Mandovi and Zuari estuaries. Thenext section examines the composition of zooplankton in the estuaries. Section 7.3describes the annual cycle of zooplankton biomass and abundance. The last sec-tion examines a couple of issues that have remained unaddressed and need to bepursued in these two estuaries.

7.2 ZOOPLANKTON COMPOSITION

The Mandovi and Zuari estuaries comprise zooplankton fauna of nearly 20 differ-ent taxonomic groups. A list of major groups of zooplankton, based on a year-longstudy in the two estuaries, is given in table 7.1 (after Padmavati and Goswami1996). Crustaceans in general, and copepods in particular, dominate the zooplank-ton community. Species belonging to the copepod families Acartidae, Pseudo-diaptomidae, Paracalanidae, and Acrocalanidae are common in these estuaries

83

June 13, 2007 18:55 RPS rpb001ch07

84 R. JYOTHIBABU AND N. V. MADHU

Table 7.1 Average numerical abundance (indm−3), unless noted otherwise, and percentagecomposition (in parenthesis) of major groups of zooplankton in the Mandovi and Zuariestuaries (from Padmavati and Goswami 1996).

Zooplankton groups Mandovi Zuari Cumbarjua canal

Hydromedusa 2.5 (0.15) 1.2 (0.08) 2.3 (0.3)Siphonophora 0.01 (0.0) 0.5 (0.03) 0.0 (0.0)Ctenophora 0.1 (0.01) 1.4 (0.09) 0.1 (0.01)Chaetognatha 10.6 (0.67) 21.1 (1.43) 28.3 (3.31)Cladocera 4.1 (0.26) 10.4 (0.73) 0.3 (0.03)Ostracoda 0.0 (0.0) 0.2 (0.01) 0.01 (0.0)Copepoda 1354 (86.11) 1242 (84.35) 696 (81.39)Amphipoda 0.2 (0.01) 0.1 (0.01) 0.2 (0.02)Lucifers 33.3 (2.12) 43.4 (2.95) 13.5 (1.57)Mysidacea 1.1 (0.06) 2.5 (0.16) 0.7 (0.08)Cumacea 0.2 (0.01) 0.5 (0.03) 0.3 (0.03)Invertebrate eggs 0.5 (0.03) 0.9 (0.06) 0.1 (0.01)Invertebrate larvae 154.3 (9.81) 131.6 (8.93) 109.8 (12.83)Pteropoda 0.0 (0.0) 0.1 (0.0) 0.0 (0.0)Copelata 4.5 (0.29) 9.7 (0.65) 0.8 (0.09)Fish eggs 4.0 (0.25) 4.7 (0.32) 0.04 (0.0)Fish larvae 2.8 (0.19) 2.0 (0.13) 2.7 (0.31)Miscellaneous 0.1 (0.01) 0.1 (0.01) 0.1 (0.01)

(Madhupratap 1987; Padmavati and Goswami 1996; Achuthankutty et al. 1997).About 49–55 species of copepods inhabit the estuaries (Goswami 1983;Achuthankutty et al. 1997). Other groups that are usually found in the zooplank-ton community are larvae of decapods and cirriped, sergestids, cladocerans, fisheggs, fish larvae, chaetognaths, copelates (appendiculareans), gastropod larvae,cumaceans, hydromedusae, and amphipods (see figure 7.1). Padmavati and Goswami(1996) have noted that the zooplankton composition does not differ appreciablybetween the Mandovi and the Zuari.

Since estuaries lie in the region of transition from marine to freshwater zones,they have a wide range in the distribution of salinity, which in turn forms a con-trolling factor in the distribution of zooplankton in Indian estuaries (Rao 1976a;Madhupratap 1987). Estuarine zooplankton from an estuary can be divided intothree components:

• marine species restricted to the seaward part of the estuary, and hence adaptedto the high-salinity range;

June 13, 2007 18:55 RPS rpb001ch07

ZOOPLANKTON 85

Figure 7.1 Major groups of zooplankton in the Mandovi and Zuari estuaries.

• organisms that live in brackish water; and• fresh water species of the upper estuary (Miller 1983).

Achuthankutty et al. (1997) have determined that in the Mandovi and Zuari estu-aries, 50% of the copepods live in the salinity range of 26–36psu, 30% live in theintermediate salinity range of 5–36psu, and 10% in the range 2–5psu. Figure 7.2provides a summary of prominent species of copepods in the two estuaries and therange of salinity in which they are found.

Goswami (1983) grouped copepods found in the two estuaries on the basis of theiroccurrence and abundance during different months of the year. He consideredthe following three groups:(a) species recorded either throughout the year or mostof the years, (b) species recorded during certain months, and (c) species seen onlyoccasionally. It was found that category (a) constituted 50–60% of the total copepodsfound in the estuaries. They were distributed in the entire stretch of both estuaries.The copepods of the category (b) were immigrants from the coastal waters and wereabundant from February to May. The third category included mostly freshwaterspecies occurring in water with salinity less than 0.5psu.

June 13, 2007 18:55 RPS rpb001ch07


Figure 7.2 Salinity ranges and the occurrence of different copepod species in the Man-dovi and Zuari estuaries. Dark areas indicate the ranges of greater abundance (fromAchuthankutty et al. 1997).

Ecological succession is a slow process of change in the number of individualsbelonging to a species of a community, or of the establishment of the new pop-ulations. In the latter case, the new species may gradually replace the originalinhabitants. This process is widespread in marine plankton communities, and is

June 13, 2007 18:55 RPS rpb001ch07

ZOOPLANKTON 87

mainly associated with seasonal changes in hydrography. Succession of copepodsin the Mandovi and Zuari estuaries begins after withdrawal of the monsoon by mid-October. At this time, the salinity in the estuaries begins to increase, particularlynear its mouth, and coastal copepods invade (Goswami 1983). Subsequently, thespecie diversity increases. It appears to peak at the end of the dry season, just beforethe summer monsoon sets in. With the onset of the monsoon begins the flushingof the estuaries by freshwater runoff and copepod biomass and diversity drop. Thelowest diversity and biomass of zooplankton in the two estuaries is found duringthe high-runoff and low-salinity time of the summer monsoon.

7.3 SEASONAL VARIATION

Figure 7.3 provides an overview of the annual cycle of zooplankton biomass andabundance in the Mandovi and Zuari estuaries; the figure is based on the datareported in Padmavati and Goswami (1996). From the monthly data of seven loca-tions (see figure 7.4 for locations), no clear seasonal cycle is apparent, but a consistentfeature seen in both biomass and abundance is that they attain their lowest annualvalue during June–August. This is the time when the summer monsoon is active,runoff is high, and salinity is at its lowest during the year. Other than this one fea-ture, that too a rather weak one, there is no distinguishing pattern to the variability.The low abundance and biomass during the monsoon could be a consequence ofthe following three processes.

• First, there is the physical mechanism of flushing of the estuary because of thelarge runoff that replaces the water in the estuary many times over during therainy season. This effect is expected to be particularly high in the Mandovi, whoserunoff is higher and volume lower than that of the Zuari.

• Second, salinity is low in the estuary at this time of the year. As noted earlier,50% of the copepods live in the salinity range of 25–26psu. Such salinities canbe expected only near the mouth of the two estuaries in the region of the bays.

• The third factor is availability of food: phytoplankton concentration is at its lowestduring the rainy season (see chapter 6).

Figure 7.3 shows that the average zooplankton biomass is significantly higher duringthe dry months of October–May than during June–August, but there is no patternto the evolution towards the increased biomass of the dry season from the lowvalues of the rainy season. A couple of stations in figure 7.3 show a sudden spurtin biomass during September, when the withdrawal of the monsoon begins; thereare, however, other stations that do not show this feature. There is also no easilyrecognizable pattern to the evolution of biomass or of abundance during the entireeight-month-long dry season.

June 13, 2007 18:55 RPS rpb001ch07


0.0

0.2

0.4

0.6

0.8

1.0B

iom

ass

(ml m

-1) M1

J F M A M J J A S O N D0

4

8

Abu

ndan

ce(x

103

ind

m-3

)0.0

0.2

0.4

0.6

0.8

1.0

Bio

mas

s (m

l m-1

) M2


4

8

Abu

ndan

ce(x

103

ind

m-3

)

0.0

0.2

0.4

0.6

0.8

1.0

Bio

mas

s (m

l m-1

) M3


4

8

(25.7)

Abu

ndan

ce(x

103

ind

m-3

)

0.0

0.2

0.4

0.6

0.8

1.0

Bio

mas

s (m

l m-1

) C3


4

8

Abu

ndan

ce(x

103

ind

m-3

)

0.0

0.2

0.4

0.6

0.8

1.0

Bio

mas

s (m

l m-1

) Z1


4

8

Abu

ndan

ce(x

103

ind

m-3

)

0.0

0.2

0.4

0.6

0.8

1.0

Bio

mas

s (m

l m-1

) Z2


4

8

Abu

ndan

ce(x

103

ind

m-3

)

0.0

0.5

1.0

1.5

2.0

Bio

mas

s (m

l m-1

) Z3


4

8

12

Abu

ndan

ce(x

103

ind

m-3

)

Figure 7.3 Zooplankton biomass (mlm−3) and abundance (indm−3) at stations M1, M2,and M3 in the Mandovi, Z1, Z2, and Z3 in the Zuari, and at location C1 in Cumbarjua canal.These locations are shown in figure 7.4. The data used in the figure above were collectedduring January–December 1990. Note the high value of abundance at station M2; this onevalue was well beyond the range of the axis. Data source: Padmavati and Goswami (1996).

7.4 DISCUSSION

Zooplankton samples are usually collected using a plankton net with a mesh sizeof 200µm. This mesh retains the so called mesozooplankton but excludes groupsof species smaller than 200µm (microzooplankton) that comprise developmentalstages of the mesozooplankton. These microzooplankton are capable of exploiting

June 13, 2007 18:55 RPS rpb001ch07

ZOOPLANKTON 89

7342' E 7348' E 7354' E 7400' E

1518' N

1524' N

1530' N

1536' N

M1

M2

M3

Z1 Z2

Z3

C1

Arabian Sea

Cu

mb

arju

a C

anal

Zuari

Mandovi

0 5 km

Figure 7.4 Location of stations from where data have been used in figure 7.3.

pico and nanoplankton that are inefficiently utilized by other, larger zooplankton.They also act as a significant food source for a variety of larger zooplankton andvertebrate predators (Robertson 1983; Stoecker and Capuzzo 1990), thereby actingas a trophic intermediate between pico/nanoplankton and mesozooplankton. Themicrozooplankton is known to be crucial for transfer of organic carbon from het-erotrophic bacteria to higher trophic levels—- the microbial loop (Azam et al. 1983).There are not many studies on microzooplankton from the Mandovi and Zuariestuaries. Available information, based on Gauns (2000), indicates the presenceof a rich microzooplankton community in the Mandovi and Zuari estuaries. Thecommunity is qualitatively and quantitatively dominated by ciliated protozoans(e.g., tintinnids), with nearly 60 species recorded (figure 7.5). Their abundance ishigher during the dry season (in excess of 30000indl−1) than the wet season (lessthan 10000indl−1). This perhaps indicates a key role for salinity in controlling the

June 13, 2007 18:55 RPS rpb001ch07


Figure 7.5 Common ciliates of microzooplankton, in part carrying lorica (house).

microzooplankton community (similar to that of larger zooplankton). On the otherhand, this could mean that with high runoff, the production of microzooplanktonwithin the estuary gets flushed out during the wet season. In recent years, a numberof studies have highlighted the role played by microzooplankton in Indian estuaries(see, for example, Godhantaraman 1994; Jyothibabu et al. 2006; Madhu et al. 2007).There is a need to take up similar studies in the Mandovi and Zuari estuaries.

There is also a need to understand processes that control zooplankton abundance.This would require establishing a link between zooplankton and their predators inthe Mandovi and Zuari estuaries. In summary, our understanding of zooplanktonin the two estuaries remains inadequate at this stage. Critically needed are fieldexperiments that can describe the zooplankton field, as well as the rates of growthand mortality, with more certainty than the present data are capable of doing.

June 14, 2007 15:42 RPS rpb001ch08

8Benthic macrofauna

Z. A. Ansari, S. Sivadas, and B. S. IngoleNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

8.1 INTRODUCTION

Benthic communities play a critical role in trophic relationships by providing majorsources of energy to economically and ecologically important demersal fishes. Theirdiverse morphology and ability to adapt to various habitats make them importantas food for large benthic organisms (McIntyre 1977; Gerlach 1978) and in recir-culation of nutrients (Kristensen et al. 1985; Tamaki and Ingole 1993). The dis-tribution of estuarine benthos varies both temporally and spatially (Rainer 1981)owing to the patchiness of their occurrence. Parulekar et al. (1980) and Ansariet al. (1986) have studied the distribution of macrobenthos in the Mandovi andZuari estuaries; Parulekar et al. (1982) investigated the association of macrobenthoswith the demersal fishery of the estuary. The macrofauna in these waters hasbeen empirically defined as animals retained by 0.5mm screens. Many organ-isms can thus be seen only on close inspection, while other may weigh severalgrams while fresh. Benthic macrofauna are good indicators of estuarine conditionsbecause they are relatively sedentary at the sediment–water interface and withindeeper sediments (Dauer and Conner 1980). The abundance of benthic animalsin an area is closely related to its environment and reflects the characteristics ofan ecological niche (Ansari et al. 2003). With this in mind, field observations onbenthic macroinvertebrates in the Mandovi and Zuari estuaries were carried outduring 2002–2003 as a part of a study to develop an integrated coastal zone man-agement plan for Goa. This chapter summarizes the salient findings from theseobservations.

91

June 14, 2007 15:42 RPS rpb001ch08

92 Z. A. ANSARI ET AL.

8.2 MATERIALS AND METHODS

A total of 15 stations shown in Map B were sampled from the Mandovi (nine stations)and Zuari (six stations) estuaries during 5–8 May 2002, 5–9 September 2002, and12–14 April 2003. Sediment samples in duplicate were collected using a van Veengrab of 0.04m2 area. Macrofaunal samples were washed and sieved using a 500µmsieve in running water and stored for sorting in a shore-based laboratory. The faunawere preserved in neutralized 5% formalin. All the animals were then identifiedup to the lowest possible taxa level using the keys available for the Indian coast.Macrofauna was counted and biomass was estimated by the wet weight method.Population density was converted into indm−2 and biomass was expressed as wetweight (g m−2).

8.3 RESULTS

The observed distributions are given in tables 8.1–8.6. The average density ofmacrofauna in the Mandovi estuary ranged from 670 to 1236indm−2. The highestvalues were recorded during May 2002 and lowest during September 2002(figure 8.2). Macrobenthic average density in Zuari ranged from 581–1900indm−2,with the highest values during April 2003 and lowest during September 2002(figure 8.2). The macrofaunal community at all the stations was dominated bypolychaetes, which constituted more than 55% of all individuals recorded. In the

Table 8.1 Macrofaunal density (ind m−2) in the Mandovi estuary during May 2002.

Taxa MZ M7 M6 M5 M4 M3 M2 M1 Mean

Polychaeta 150 1425 150 80 325 175 1500 75 485Oligochaeta 0 50 0 0 0 0 0 25 9Bivalvia 50 50 75 180 50 25 50 0 60Gastropoda 0 25 50 25 25 50 0 0 22Isopoda 0 0 0 25 25 0 25 0 9Amphipoda 100 50 0 0 3250 25 100 25 444Tanaidacea 0 0 50 0 475 0 25 0 69Cumacea 0 25 0 25 0 0 0 0 6Decapoda 25 25 50 50 0 50 0 0 25Sipunculida 0 0 25 25 0 0 0 0 6Nemertea 25 25 50 0 0 25 0 0 16Echiurida 0 0 25 25 0 0 0 0 6Others 175 75 125 75 25 50 75 25 78Total (indm−2) 525 1750 600 510 4175 400 1775 150 1236Biomass (gm−2) 14.9 15.4 11 28.2 18.4 6.6 19.8 5.3 14.95

June 14, 2007 15:42 RPS rpb001ch08

BENTHIC MACROFAUNA 93

Table 8.2 Macrofaunal density (indm−2) in the Mandovi estuary during September 2002.

Taxa M7 M6 M5 M4 M3 Mean

Polychaeta 225 1050 100 100 25 300Bivalvia 300 75 1050 0 25 290Gastropoda 25 25 0 0 0 10Amphipoda 75 75 50 0 0 40Cumacea 75 25 0 0 0 20Decapoda 25 0 0 0 0 5Nematode 0 0 25 0 0 5Total (indm−2) 725 1250 1225 100 50 670Biomass (gm−2) 9.18 0.89 14.58 0.005 3.305 5.59

Table 8.3 Macrofaunal density (indm−2) in the Mandovi estuary during April 2003.

Taxa MZ M7 M6 M5 M4 MN M3 M2 Mean

Obelia colony 0 0 0 0 25 0 0 0 3Nematoda 0 0 0 0 25 0 0 0 3Polychaeta 1575 325 825 425 900 125 575 975 716Oligochaeta 0 0 0 175 50 25 0 25 34Bivalvia 25 0 25 150 50 550 75 200 134Gastropoda 0 0 0 0 0 0 25 125 19Isopoda 0 50 0 0 50 0 75 25 25Amphipoda 25 25 0 25 350 25 0 25 59Harpacticoida 25 0 0 0 0 0 0 0 4Cumacea 25 0 0 0 50 200 25 100 50Pcynogonida 0 0 0 0 0 0 0 0 0Decapoda 0 25 75 25 50 25 0 0 25Cirripiedia 0 0 0 0 0 0 0 0 0Echiniodermeta 50 0 0 0 0 0 0 0 6Others 0 0 50 0 0 0 0 0 6Total (indm−2) 1725 425 975 800 1550 950 775 1475 1084Biomass (gm−2) 5.75 0.42 0.93 25.87 3.41 98.55 0.43 3.98 17.42

Mandovi estuary, biomass values ranged from 5.59 to 17.42gm−2 (figure 8.1), withthe highest values recorded during April 2003 and lowest during September 2002.Macrofaunal biomass in the Zuari estuary ranged from 3.38 to 10.42gm−2

(figures 8.2). A total of 68 taxa were identified during the study. Polychaete domi-nated the macrobenthic diversity. Secondary polychaetes, representing 58% of thetotal polychaete fauna, were the largest group.

June 14, 2007 15:42 RPS rpb001ch08


Table 8.4 Macrofaunal density (indm−2) in the Zuari estuary during May 2002.

Taxa Z5 Z4 Z3 Z2 Z1 Mean

Polychaeta 2500 325 150 100 155 646Oligochaeta 25 0 0 0 25 10Bivalvia 25 0 125 50 0 40Gastropoda 0 0 50 0 75 25Amphipoda 25 25 75 50 25 40Tanaidacea 125 25 25 50 0 45Cumacea 0 0 0 0 0 0Decapoda 200 25 50 0 25 60Sipunculida 0 0 25 0 0 5Nemertea 0 125 0 0 0 25Echiurida 0 0 25 0 0 5Others 200 25 0 50 0 55Total (indm−2) 3160 550 525 300 305 968Biomass (gm−2) 24.2 2.2 12.6 5.5 7.6 10.42

Table 8.5 Macrofaunal density (indm−2) in the Zuari estuary during September2002.

Taxa Z5 Z4 Z3 Z2 Mean

Polychaeta 825 90 900 25 460Bivalvia 0 25 175 0 50Gastropoda 25 25 0 0 13Amphipoda 25 0 50 0 19Cumacea 0 10 0 0 3Nematode 150 0 0 0 38Total (indm−2) 1025 150 1125 25 581Biomass (gm−2) 22.61 0.41 8.36 0.08 7.86

8.4 DISCUSSION

During the 2002–2003 observations, the macrofaunal density and biomass showedseasonal change. The lowest values of both density and biomass were recordedduring September 2002 in both the estuaries (see figures 8.1 and 8.2). Given softsediments, this is most likely a consequence of the low salinity due to the summermonsoon. Panikkar (1969) proposed partial or complete destruction of tropical estu-arine fauna during the summer monsoon, followed by an annual re-population ofthe estuaries and backwaters after its withdrawal and the resulting salinity increase.Parulekar et al. (1980) undertook a large seasonal investigation of more than half ofeach of the two rivers (downstream parts) during 1971–1973 with the same methodsas our study. They found that salinity had the largest influence on abundance in the

June 14, 2007 15:42 RPS rpb001ch08


Table 8.6 Macrofaunal density (indm−2) in the Zuari estuary during April 2003.

Taxa Z5 Z4 Z3 Zn Z2 Mean

Obelia colony 0 475 0 50 0 105Nematoda 0 50 0 25 0 15Polychaeta 1600 775 100 2675 250 1080Oligochaeta 0 0 0 25 0 5Bivalvia 0 275 0 100 50 85Isopoda 75 0 0 475 150 140Amphipoda 75 550 0 300 0 185Harpacticoida 0 25 0 0 0 5Cumacea 0 0 0 425 25 90Pcynogonida 0 25 0 0 0 5Decapoda 0 0 0 50 0 10Echiniodermeta 0 850 0 0 0 170Others 0 0 0 25 0 5Total (indm−2) 1750 3025 100 4150 475 1900Biomass (gm−2) 0.97 8.73 0.06 6.58 0.59 3.38

Mandovi, while in the Zuari it was a combination of environmental factors. The r2

values of their multiple correlations were rarely >0.25 (see also Alongi (1990), forobservations made elsewhere in the tropics).

Macrofaunal density during April and May 2002 was higher in the Zuari estuarythan in the Mandovi, but biomass was higher in the Mandovi than in the Zuari(compare figures 8.1 and 8.2). The higher biomass in the Mandovi estuary was dueto the presence of bivalves (see tables 8.1–8.3). Molluscs are known to be dominantin regions with a sandy bottom, whereas polychaetes prefer muddy substratum.The highest bivalve density in the present study was observed in the Mandovi atstation M5 (tables 8.1–8.3), where the sediment was sandy. The stations in Zuari,which had silty clay, were dominated by polychaetes. These results agree withthose of Parulekar and Dwivedi (1973), who observed higher benthic biomass inthe Mandovi estuary and attributed this to the more favourable substratum.

A comparison of the present data with those reported earlier (where the same sam-pling gear was used in all these studies) indicates that the abundance and distributionof fauna has changed. Polychaetes and bivalves accounted for more than 70% ofthe fauna during 1972–1973. During 1982–1983 and 1992–1993, the polychaetescontinued to be the dominant fauna, but crustaceans replaced bivalves. The sameis the case observed during the 2002–2003 observations. The reasons behind thesechanges are not known. Nevertheless, one cannot avoid noting that two activitiesthat can have an impact on benthic populations have been on the rise during thelast few decades. The first of these is mining of iron and manganese. Initiated in1950, the last couple of decades have seen a consistent increase in this activity.Mining rejects often end up in the estuary and are therefore expected to contribute

June 14, 2007 15:42 RPS rpb001ch08


0

1000

2000

3000

4000

5000

Den

sity

(in

d m

−2)


April 2003

0

1000

2000

3000

4000

5000

Den

sity

(in

d m

−2)


September 2002

0

1000

2000

3000

4000

5000D

ensi

ty (

ind

m −2

)


May 2002

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


April 2003

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


September 2002

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


May 2002

Figure 8.1 Macrofaunal density (left panels; indm−2) and biomass (right panels; gm−2)in the Mandovi estuary during May 2002 (top), September 2002 (middle), and April 2003(bottom). The horizontal axis gives distance from the mouth (km). In May 2002 and April2003, one station on the shelf (hence, negative distance on the axis) was also sampled. Thevertical bars indicate range of observed values at the respective stations.

to a change in the substratum. The precise implications of such a change are notknown. The second activity that disturbs the substratum is mining of sand from thebed of the estuaries. With the increase in building construction in Goa, mining ofsand is also on the rise, and we do not know the impact it has had on benthic life.

June 14, 2007 15:42 RPS rpb001ch08


0

1000

2000

3000

4000

5000

Den

sity

(in

d m

−2)


April 2003

0

1000

2000

3000

4000

5000

Den

sity

(in

d m

−2)


September 2002

0

1000

2000

3000

4000

5000D

ensi

ty (

ind

m −2

)


May 2002

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


April 2003

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


September 2002

0

20

40

60

80

100

120

Bio

mas

s (g

m −2

)


May 2002

Figure 8.2 Macrofaunal density (left panels; indm−2) and biomass (right panels gm−2)in the Zuari estuary during May 2002 (top), September 2002 (middle), and April 2003(bottom). The horizontal axis gives distance from the mouth (km). In May 2002 and April2003, one station on the shelf (hence, negative distance on the axis) was also sampled. Thevertical bars indicate range of observed values at the respective stations.

Nonetheless, studies that have earlier noted a change in both abundance and diver-sity of benthic populations are worth noting here. Parulekar et al. (1980) reportedthe presence of 111 species (0.5 mm mesh size) of macrobenthos during 1972–1973. The data from 1982–1983 on abundance and species occurrence reported by

June 14, 2007 15:42 RPS rpb001ch08


Ansari et al. (1986) showed a decline in the number of species with only 67 recorded.Polychaeta were the dominant fauna during 1972–1973 and were followed byBivalvia. During 1982–1983, however, crustaceans replaced bivalves as the seconddominant group. Observations carried out in 1992–1993 showed a further declinein the number of species, the new number being 48. The total number of benthicorganisms also declined. The observations carried out during 2002–2003 showedthat there was an increase in the number of species from 1992–1993. A total of 67species were recorded and bivalves again constituted the second dominant groupafter polychaetes. The dominant polychaete species during the present study wereMediomastus sp., and Cossura sp. During the earlier studies, Diopatra neapolitana,G. alba, and P. pinnata were dominant. Three species, namely M. casta, C. fluviatlis,and A. chilkensis, endemic to the two estuaries earlier, were either absent or recordedin low numbers at various stations during 2002–2003. The number of crustaceanshas oscillated: the average density during 1972–1973 was 1604indm−2, it decreasedto 573indm−2 in 1992, but rose to 1073indm−2 in 2002–2003.

In essence, while the present data are too meagre to arrive at a statistically significantconclusion regarding all aspects of abundance and diversity of benthic populationsin the Mandovi and Zuari estuaries, there is empirical evidence to support twoconclusions. The first is that benthic macrofauna in the two estuaries go through anannual cycle of destruction and rejuvenation. Indications are that the destructionoccurs when the salinity drops owing to runoff during the summer monsoon, as alsoshown by earlier works cited above. Rejuvenation follows withdrawal of the freshwater. The second conclusion is that rejuvenation differs from year to year. Whileit has been reported that there have been long-term changes in benthic populationsin the Mandovi and Zuari estuaries, and that these changes can be associated withenvironmental changes due to mining of ore on land and of sand from the estuarinechannels, firm evidence is lacking.

June 15, 2007 12:52 RPS rpb001ch09

9Distribution of ironand manganese

Analia M. Mesquita and Sujata KaisaryNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

9.1 INTRODUCTION

An important industry in Goa is mining of iron and manganese ore. Active mininggoes on at present at a number of locations distributed along a belt that stretchesfrom north to south in the central part of the state. The present mining sites areshown in figure 9.1. The iron ore from these sites is transported in open trucks toa number of nearby convenient locations on the banks of the Mandovi and Zuariestuaries. From these locations the ore is transferred to barges, which take the oreto the Mormugao Port, from where it is shipped to other places.

The mining operations and the process of transporting the ore generate dust that isthen carried by winds to nearby areas. The wind-borne transport of dust is particu-larly strong during the dry season. It is expected that some of the dust would settlein the waters of the two estuaries, leading to significant concentrations of dissolvedand particulate iron and manganese. Also contributing to these concentrations iswater-borne transport of dust and sediments by streams that flow through the areasaround active mining sites. This transport is particularly high during the summermonsoon. While the mining operations are suspended during this season becauseof rains (hence wind-borne transport is minimal), the runoff over mining areas,which ultimately ends in the Mandovi and Zuari estuaries, becomes a significantsource of iron and manganese in the estuaries. In view of this, it is of interest to

99

June 15, 2007 12:52 RPS rpb001ch09

100 ANALIA M. MESQUITA AND SUJATA KAISARY

Figure 9.1 Map of Goa showing areas of types of ores and mining activities. (Courtesy:Goa State Pollution Control Board, Goa.)

June 15, 2007 12:52 RPS rpb001ch09

DISTRIBUTION OF IRON AND MANGANESE 101

determine concentrations of dissolved and particulate iron and manganese in thewater column in the estuaries and in the sediments. In this chapter, we report onthe observed concentrations.

9.2 SAMPLING AND ANALYTICAL METHODS

The schedule of sampling of water for data discussed in this chapter is given intable 9.1. During each period of observation, water samples were collected at sta-tions M1–M7 in the Mandovi and at Z1–Z5 in the Zuari (Map B). Station MZ , whichrepresents the regime off the mouths of the two estuaries, was also worked. Dur-ing each period of simultaneous observations (stations were occupied by separateboats), which lasted 24hours, a water sample was collected every three hours.Hence, at each station during each period of observation, there were eight sam-ples. The variability during the 24-h period gives a measure of change due to tidalmovement of water. Each water sample was collected using a Niskin bottle at adepth of approximately 3m from the surface. Sediment samples were collectedonce during a 24-h observation period by using a Van Veen grab of 0.04m2 area.

Sample collection, filtration, treatment, and analysis for iron and manganese wascarried out according to the protocols of ultra-trace-metal analysis (Bruland et al.1979; Danielson 1980) using atomic absorption spectroscopy (AAS). The accuracyof the dissolved metal analyses was found to be ±5% using certified referencewaters CASS-3 and NASS-4 (supplied by National Research Council of Canada).Extractable metal concentrations in suspended particulate matter (SPM; > 0.45µm)were analyzed after the acid digestion of SPM and drying of the residue to a constantweight on membrane filters. The same digestion procedure was performed on filterswithout particles to find procedural blanks. Dried bottom sediment samples weredigested with a mixture of hydrofluoric-nitric-perchloric acids. The residue wasdissolved in dilute hydrochloric acid and metal concentrations were determined byAAS. Analysis of the certified reference materials PACS-1 and BCSS-1, followingthe same protocol, showed good recoveries (∼95%) of the metals.

Table 9.1 Schedule of sampling.

Estuary Period of observation

Mandovi 28–29 April 2002 (spring tide)Mandovi 5–6 May 2002 (neap tide)Mandovi 5–6 September 2002Zuari 30 April–1 May 2002 (spring tide)Zuari 7–8 May 2002 (neap tide)Zuari 7–8 September 2002

June 15, 2007 12:52 RPS rpb001ch09


9.3 RESULTS AND DISCUSSION

Figures 9.2 and 9.3 summarize the observed spatial and temporal distribution ofiron and manganese respectively. From the figures, no clear pattern emerges. Onaverage, concentration of particulate iron in the Mandovi and Zuari vary from neg-ligible concentration to about 50µgl−1. The range for dissolved iron too is similar.Dissolved manganese also varied from negligible concentrations to about 50µgl−1.and concentration of particulate manganese has a similar range. The observedconcentrations in sediments are shown in figure 9.4. Again there is no clear tem-poral or spatial pattern. The range of concentration of iron in the two estuaries is0–20mgg−1, and that for manganese is 0–15mgg−1.

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

15028/29 April 2002

Mandovi

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

15030/01 April/May 2002

Zuari

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

1507/8 May 2002

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

1505/6 May 2002

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

1505/6 September 2002


Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100



Figure 9.2 Variation in dissolved (red line) and particulate (blue line) iron (µgl−1) in theMandovi estuary (left panels) and the Zuari estuary (right panels) with distance from themouth (km) of each estuary. The period of observations is identified at the upper righthand corner of each panel. The vertical bars indicate the range of observed values at therespective stations.

June 15, 2007 12:52 RPS rpb001ch09

DISTRIBUTION OF IRON AND MANGANESE 103

While there is no clear pattern for dissolved and particulate iron or manganesein the two estuaries, there is a weak tendency that is noticeable. The values oftentend to be higher in the middle estuary, that is, somewhere mid-way between thehead and the mouth of the estuary. This tendency implies that the source of ironand manganese is largest somewhere in the middle estuary, and not in the waterthat enters the estuary at the upstream end, or in the water on the shelf. Thisdoes make sense in view of the known sources: wind-borne dust or water-bornematerial, both arising from mining operations. The mines are distributed over awide area (figure 9.1). As the material carrying iron and manganese is expected toreach the estuaries after following a chaotic path, it is not surprising that there isno clear spatial or temporal pattern to the distribution. It is also not surprising that

Dis

solv

ed (

μg l−1

)

0

50

100P

artic

ulat

e (μ

g l−1

)

0

50

100

15028/29 April 2002

Mandovi

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

15030/01 April/May 2002

Zuari

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

1507/8 May 2002

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100

1505/6 May 2002

Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100



Dis

solv

ed (

μg l−1

)

0

50

100

Par

ticul

ate

(μg

l−1)

0

50

100


0 10 20 30 40Distance from mouth (km)

Figure 9.3 Variation in dissolved (red line) and particulate (blue line) manganese (µgl−1)in the Mandovi estuary (left panels) and the Zuari estuary (right panels) with distance fromthe mouth (km) of each estuary. The period of observations is identified at the upper righthand corner of each panel. The vertical bars indicate the range of observed values at therespective stations.

June 15, 2007 12:52 RPS rpb001ch09


Iron

(m

g g−1

)

0

4

8

12

16Mandovi

Iron

(m

g g−1

)

0

4

8

12

16Zuari

Man

gane

se (

mg

g−1)

0

4

8

12

16


Man

gane

se (

mg

g−1)

0

4

8

12

16


April 2002 September 2002

Figure 9.4 Variation in sediment Fe and Mn(mgg−1) in Mandovi and Zuari estuaries as afunction of distance from the mouth (km).

the sediments did not show any systematic pattern. After deposition on the surfaceof the bottom of the estuary, the sediments would have been moved around inthe estuary by tidal flows, until somehow they reach subsurface levels, where theywould not be vulnerable to further movement. On the other hand, a mid-estuarineincrease could also result from the release or re-precipitation of reduced or oxidizedforms of iron and manganese from organic rich sediments; the relative importanceof these processes is unknown.

It is worth noting that the samples that were analyzed here were derived from surfacesediments. It is possible that a core collected in the estuary would exhibit a sharpchange in iron and manganese distribution with depth because mining of iron andmanganese in Goa started around 1950. If our hypothesis, that the concentrationof metals in the estuaries today is primarily due to mining, is correct, then thereshould be a distinct reduction in concentration in sediments that were depositedprior to the 1950s.

June 14, 2007 13:0 RPS rpb001ch10

10Distribution of tributyltin(TBT) in the Mandovi estuary

Narayan B. BhosleNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

10.1 BIOFOULING AND ITS EFFECTS

It is a common experience that when water is left in a bucket or in any othercontainer for a long period of time, the area of the bucket in contact with waterbecomes slimy. This slimy layer is known as biofilm. Different steps are involved inthe development of a biofilm. The first step is adsorption of dissolved organic matteronto a surface, thereby conditioning it. The conditioned surface is then colonized bymicro-organisms, including bacteria, microalgae, fungi, protozoa, etc. Developmentof a biofilm on a surface may induce the settlement and metamorphosis of larvae ofinvertebrates such as barnacles, mussels, oysters, and algal spores. This results in thedevelopment of macrofouling visible to the naked eye. Biofilm and macrofoulingtogether are defined as biofouling. The best example of biofouling is the growthyou see on the hull of a ship. Biofouling on the hull or any other industrially usefulmaterial can cause considerable damage. It can increase frictional drag on the ship,thereby increasing fuel consumption, induce metal corrosion, impair heat-transferefficiency of heat conductors, block flow in pipelines, reduce the life span of offshoreplatforms, etc. Such effects can cause huge economic losses to several industries,including naval and commercial shipping, nuclear and thermal plants, offshore oilproduction platforms, and many other installations in coastal areas. Several billiondollars are spent worldwide to control the problem of biofouling.

105

June 14, 2007 13:0 RPS rpb001ch10

106 NARAYAN B. BHOSLE

10.2 CONTROL OF BIOFOULING

To reduce economic losses due to biofouling, ship hulls and other surfaces arecoated with antifouling paints containing toxic chemicals. The most commonly usedantifouling compound is tributyltin (TBT) (Hoch 2001). Use of TBT-based antifoul-ing paints helps the shipping industry save about US$ 6 billion per year, which,in turn, results in fuel saving of about 7 million tonnes per year. TBT compoundsare also used as stabilizers in the PVC industry, plastic additives, industrial cata-lysts, insecticides, fungicides, bactericides, and wood preservatives. Various sourcesresponsible for the release of TBT compounds in a marine environment are shownin figure 10.1. There is no doubt that use of TBT results in several economic ben-efits. In addition, TBT usage in the shipping industry also offers some indirect

Figure 10.1 Sources of organotin compounds for human exposure (modified from Hoch2001).

June 14, 2007 13:0 RPS rpb001ch10

DISTRIBUTION OF TBT IN THE MANDOVI ESTUARY 107

Figure 10.2 Distribution and fate of TBT compounds in aquatic environment (modifiedfrom Hoch 2001).

environmental benefits such as reduced production of greenhouse gases and sup-pression of bioinvasion of biofouling organisms. Nevertheless, when released intothe marine environment, TBT undergoes several chemical and biological processes(figure 10.2). The most important concern in the use of TBT is that it is highly toxicto non-target organisms when released into marine waters.

10.3 ENVIRONMENTAL EFFECTS OF TBT

Use of organotin such as TBT has increased several fold during the last two decades.TBT compounds used in industrial applications, especially as a constituent of anti-fouling paints, have resulted in a high abundance of organotins such as TBT andother butyltin and phenyltin compounds in the water, sediment, and tissues oforganisms from several aquatic and marine environments; most affected are har-bour and near-shore areas (Hwang et al. 1999; Hoch 2001). It is well documentedthat TBT at the level of a few ngl−1 (nanograms per litre) is detrimental to non-targetorganisms. Some of these effects include high larval mortality, shell deformation,reduced reproduction in many invertebrates, imposex (change of sex, i.e., a femaledeveloping male sex organ) gastropods, including dogwhelks, retardation of growthin mussels, and inhibition of photosynthesis and growth in microalgae. Owing to

June 14, 2007 13:0 RPS rpb001ch10


these concerns, many countries, particularly in the European Union, USA, andJapan have enforced strict laws to regulate use of TBT.

In the early 1980s, the first catastrophic environmental impact of TBT was noticedin Arcachon Bay, France, where the oyster fishery nearly collapsed. Abnormalthickening of the shells was observed in juvenile Pacific oysters from the bay. Theseabnormalities were attributed to the release of TBT from antifouling point. Alzieu(2000) observed a connection between TBT levels and oyster-shell anomalies.

At about the same time, a decline in the population of dogwhelk Nucilla lapillusin southeast England was attributed to TBT contamination. Field and laboratorystudies confirmed that TBT was highly toxic to non-target organisms. It was realizedthat TBT was one of the most toxic compounds ever introduced by man into themarine environment.

10.4 LEGISLATIVE RESTRICTIONS

Owing to these concerns, many countries, particularly the European Union, USA,and Japan, have established strict laws to regulate use of TBT.

On seeing the near-complete destruction of shell fishery in the Arcachon Bay, theFrench Government formulated laws to regulate the use of TBT-containing paints inJanuary 1982 for vessels shorter than 25m in length. Subsequently, similar actionwas taken by the governments of the United Kingdom in 1985, USA in 1988,and Canada in 1989. When it was realized that there was a strong connectionbetween levels of TBT in the oyster tissues and deformities in oysters, use of TBT inantifouling paint was banned in Australia in 1989. In Japan, use of TBT in antifoulingpaints and aquaculture nets was banned or restricted. In view of the toxic effects,use of TBT was completely banned in Austria, Sweden, and Switzerland.

Such legislative action to restrict or ban use of TBT has resulted in benefits to marinefarms. These measures were instrumental in decreasing levels of TBT in water, sedi-ment, and organisms (Hoch 2000). In Great Britain, water quality improved enoughto permit resumption of oyster farming (Alzieu 2006). Decrease in TBT in waterled to a decrease in TBT levels in oysters. There was also a decrease in imposexin the gastropod population in Great Britain. These efforts were instrumental indecreasing levels of TBT in aquaculture farms. In the countries that did not imposeany restrictions or regulations, concentrations of TBT remained relatively high,particularly in harbours with high levels of shipping activity.

India is one of the fastest developing countries in Asia. In India, TBT compounds areused in antifouling paints to control the growth of fouling organisms on commercialand naval ships as well as on other industrial structures that use seawater. There is nolegal ban on the usage of TBT compounds in India. There is also little informationavailable on the distribution of TBT compounds in water, sediments, and animals

June 14, 2007 13:0 RPS rpb001ch10


collected from the coast of India. The Mandovi estuary is one of the very few marineareas of the country where some information on TBT levels is available. Hence,this chapter summarises what is known of TBT in the estuary.

10.5 STUDY AREA AND SAMPLE COLLECTION

The Mandovi estuary is used for transport of iron and manganese ore from the hin-terland to Mormugao harbour. Along the Mandovi estuary, there are several smalljetties and wharfs. Jetties are used for berthing fishing vessels and barges. Smallworkshops that are involved in repair, maintenance, and construction of small shipsand barges are located on the banks of the river. The estuary has mangroves alongits banks. They support several flora and fauna that are either resident or migra-tory, and act as nursery grounds and nutrient sources for a variety of economicallyand ecologically important organisms, including fishes, prawns, and many otherinvertebrates.

As a part of our study on TBT, seawater samples were collected using 5l (litre)Niskin samplers at a depth of about 1m from 10 locations in the Mandovi estu-ary (figure 10.3). The samples were filtered onto GF/F filters to remove suspendedparticulate matter. Sediment samples from the same locations (figure 10.3) werecollected using van Veen grab, and stored at −20◦C until used for analysis. Sedi-ments were lyophilized, powdered, and stored at −20◦C until analysis. Glasswareused was cleaned with chromic acid and washed with distilled water, followed byUV-Milli-Q purified water, and dried in an oven. Before use, the glassware wasrinsed with dichloromethane and kept covered with aluminium foil until use.

73°42'E 73°48'E 73°54'E 74°00'E 74°06'E

15°24'N

15°30'N

15°36'N

1

2

34

5

6 7

89

10

Cu

mb

arju

a C

analNerul

Betim

Old Goa

Pilgao

Aguada Bay Panaji

Figure 10.3 Map showing sampling sites along the Mandovi estuary.

June 14, 2007 13:0 RPS rpb001ch10


10.6 EXTRACTION OF TBT COMPOUNDS

A known volume of un-filtered seawater was transferred to a teflon separating fun-nel. To the funnel, dichloromethane, 6%NaBH4, and tripropyltin chloride (TPrT),the last as internal standard, were added. The contents of the funnel were mixedand then extracted for 10 minutes. After separating the organic layer the samplewas subjected to extraction once again. Both the organic extracts were combined,dried using anhydrous sodium sulphate, and filtered through Whatman No 1 filterpaper. The volume of the organic extract was reduced to 100µl using nitrogen gas.The standard was prepared following the same procedure.

In order to extract TBT compounds from sediments or SPM, a known amountof sediment sample, or a GF/F filter containing SPM, was transferred to a testtube. To the test tube 0.1% of NaOH in methanol and a known amount ofTPrT as an internal standard were added. 0.1% of NaOH in methanol served as ablank. The standard was prepared following the same procedure but without sam-ple. The sample, standard and the blanks were vortexed for one hour to extract TBT.Then hexane and NaBH4 were added to all the tubes and the reaction mixture wasvortexed for 15 minutes and the hexane layer containing derivatized compoundswas collected. The extraction step was repeated thrice and all the extracts werepooled together. The hexane extract was dried over anhydrous sodium sulphateand concentrated to 100µl using nitrogen gas, and analysed by gas chromatography.

10.7 ANALYSIS OF TBT COMPOUNDS BY GAS CHROMATOGRAPHY

Separation and quantification of butyltin compounds were performed by a capillarygas chromatograph (Agilent HP6890 Series model). The gas chromotograph wasequipped with a flame photometric detector (FPD), a tin-specific (610) filter, anda HP 5 capillary column (5 metres). One microlitre sample or standard mixtureor blank was injected using a programmable on column injector when the initialtemperature was 60◦C. After 2 minutes, the oven temperature was programmed to230◦C at the rate of of 20◦C per minute and held at this temperature for 8 minutes.Nitrogen was used as a carrier gas (one mlmin−1). The injector was operated intrack oven mode, while the FPD detector was maintained at 250◦C with hydrogenand air flowing at 80 and 105mlmin−1, respectively. Quantification of each peak ina sample was done by using the data handling system installed in the instrument.

10.8 BUTYLTINS IN WATERS AND SUSPENDED PARTICULATE MATTEROF THE MANDOVI ESTUARY

TBT was present in seawater samples collected from most of the stations that weresampled in the Mandovi estuary (figures 10.4a, b), but the concentration of TBT

June 14, 2007 13:0 RPS rpb001ch10


0

50

100

150

200

250

300

TB

T (

ng l-1

)

(a) Water

Station no.

1 2 3 4 5 6 7 8 9 10

0

5

10

15

20

25

30

35

TB

T (

ng l-1

)

(b) SPM

Station no.1 2 3 4 5 6 7 8 9 10

0

10

20

30

40

50

60

70

80

TB

T (

ng l-1

)

(c) Sediment

Station no.1 2 3 4 5 6 7 8 9 10

Figure 10.4 Distribution of TBT in (a) water, (b) suspended particulate matter, and (c) sed-iments, at various stations in the Mandovi estuary. The connection between station numbersand station locations is as follows. 1: Panaji Harbour; 2: Fishing Jetty at Betim; 3: ChoraoFerry; 4: Old Goa; 5: Piligao; 6: Sarmanas; 7: Amona; 8: Mayana-1; 9: Mayana-2; 10: Pali.The thin vertical bars indicate the standard deviation.

varied from station to station. For both water and SPM samples, no particular trendwas recorded for the distribution of TBT in the Mandovi estuary.

The concentration of TBT varied from 4.9ngg−1 to 49ngg−1 dry wt of sediment ofthe Mandovi estuary. Concentration of TBT was relatively higher at stations 8, 9,

June 14, 2007 13:0 RPS rpb001ch10


and 10 (figure 10.3). These higher concentrations of TBT in the sediments collectedfrom stations such as 8, 9, and 10 in the Mandovi estuary probably indicate theeffect of berthing of barges and small fishing boats at these locations.

10.9 IMPLICATIONS OF THE DISTRIBUTION OF TBT COMPOUNDS

Most of the locations sampled in the Mandovi estuary are influenced by activitiessuch as dry docking, repair and maintenance of small ships and barges, and con-struction and painting of barges. Therefore, depending on the intensity of theseactivities at the sampled locations, the concentration of TBT varied. The observedTBT levels in water and sediments of the Mandovi estuary are comparable to thosereported from other coastal areas of the world subject to shipping activity. Theconcentration of TBT in the waters of the Mandovi is similar to that recorded forthe waters of other estuaries and coastal waters of the world before the TBT ban.A high concentration of TBT was observed at a few stations — for example, thesediment at stations 5–10 had high TBT concentration. The major source for suchlocalized contamination of TBT is expected to be the release of TBT by barges andsmall boats used for fishing and recreational activities such as motor boating. Sta-tions 5–10 are located in the vicinity of jetties and small ship repair workshops thatconstruct, repair, overhaul, sandblast, and paint small commercial vessels. Suchactivities lead to release of TBT compounds through waste waters and effluentsinto the estuary. Moreover, the observed spatial distribution pattern of TBT mayimplicate the influence of bacterial degradation, tides, and currents at the samplingsites of the study.

TBT compounds have both lipophilic and ionic properties. Owing to their lipophilicnature, TBT compounds dissolve in lipids, whereas in ionic interactions, TBT com-pounds bind to macromolecules. Therefore, good relationships between TBT andtotal lipids and/or organic carbon helps in identifying the role of these factors incontrolling the abundance of TBT compounds in marine waters. In the Mandoviestuary, we did not observe any relationship between TBT compounds and organiccarbon or total lipids. This implies that organic carbon and lipids may produce vari-able effects on the distribution of TBT compounds, probably due to the influenceof some local factors such as quality of organic matter and the waves, and currentsin the study area.

Whatever may be the factors involved in controlling the distribution of TBT com-pounds in marine waters, the water and sediment samples of the Mandovi estuaryappear to contain substantial amounts of TBT compounds. TBT is considered to betoxic to many marine organisms at levels of about 100ngl−1. Even at levels less than10ngl−1, TBT can cause endocrine disruption, resulting in imposex, especially ingastropods (Hwang et al. 1999; Hoch 2001). The observed levels of TBT compoundsin the Mandovi are higher than those known to induce imposex, and, in fact, impo-sex has been recorded in gastropods collected from Mormugao harbour. Bhosle

June 14, 2007 13:0 RPS rpb001ch10


(2006) reported that oysters, mussels, clams, and fish collected from the nearbyMormugao harbour were contaminated with TBT, and that TBT can suppress nor-mal functioning of the immune system of mammals. In view of this, consumptionof TBT-contaminated fish and shellfish is a matter of concern. There is thereforeneed for legislation to regulate use of TBT compounds in the estuaries and coastalwaters of India.

June 13, 2007 20:6 RPS rpb001ch11

11Sewage-pollution indicatorbacteria

N. Ramaiah, V. Rodrigues, E. Alvares, C. Rodrigues,R. Baksh, S. Jayan, and C. MohandassNational Institute of Oceanography, Dona Paula, Goa 403 004, India.

11.1 INTRODUCTION

Microbiologists the world over detect sewage contamination of aquatic habitatsby enumerating coliform groups of bacteria (Brock et al. 1994). As is universallyaccepted, higher the sewage contamination (either through indiscriminate, deliber-ate, accidental, or regular/routine disposals), higher will be the number of coliformsin environmental samples. Further, microbiologists rely on the principle that higherthe incidence of sewage indicator bacteria in any environment, higher will be thechances for human pathogenic bacteria to be present (Brock et al. 1994). Also,bacterial metabolism is such that if a particular group, say E. coli, is the dominantbacterium in the sewage discharges, it can compete with and outgrow the nativemicroflora. This can lead to increased levels of indicator bacteria in the water bod-ies, and the loads of human pathogenic bacteria may well exceed both ecologicaland human acceptable limits.

Raw sewage disposal into the Mandovi and Zuari estuaries has been a commonpractice throughout the history of the estuaries. Treatment of sewage from majorcities like Panaji before its disposal into the estuary is a recent development. Withincreasing population on the banks of the estuaries, the amount of sewage dumpedin the estuary has also increased. It is therefore of interest to determine what the

115

June 13, 2007 20:6 RPS rpb001ch11

116 N. RAMAIAH ET AL.

levels of pollution indicator bacteria are owing to sewage disposal. This informationwould help determine if careful waste treatment and disposal procedures are neededto safeguard the natural environment.

In this chapter, we describe the spatial distribution and annual cycle of sewage-pollution- indicator and human-pathogenic bacteria in water and sediment samplesin the Mandovi and Zuari estuaries. We then discuss the accepted norms for theseindicators in developed countries and reflect on the measures that we need to put inplace to improve the conditions in the estuaries.

11.2 OBJECTIVES AND METHODOLOGY

We monitored microbiological parameters to understand spatial and temporal vari-ations in the abundance of sewage-pollution-indicator and human-pathogenic bac-teria. For assessing bacteriological quality, samples were collected and analyzedfrom 12 different locations shown in Map B: M1, M2, M3, M4, M5, M6, and M7 inthe Mandovi, and Z1, Z2, Z3, Z4, and Z5 in the Zuari. Sampling was carried outduring three seasons representing pre-monsoon (28 April to 8 May 2002), sum-mer monsoon (5–8 September 2002), and post-monsoon (12–15 March 2003). Theschedule of observations is given in table 11.1. At each location, water sampleswere collected every three hours for 24 hours. The eight samples collected overa 24-hour period allowed us to examine the variability in bacterial abundance atdifferent locations in these estuaries.

Water samples were collected using Niskin samplers. A van Veen grab was usedfor sediment sampling. Only one sediment sample was analyzed from each loca-tion during each of the three seasons. After collection, the samples were storedon ice and transported to the laboratory for analysis (usually) within 3 hours. Stan-dard and established microbiological methods (American Society for Microbiology1957) were followed for all the microbiological analyses. Water samples were ana-lyzed using the standard membrane filtration technique and the filters were placedon specific media. Suitable dilutions of sediment samples were prepared and tworeplicate aliquots were spread-plated on to agar plates. All plates were incubated at37◦C, and final counts of colony forming units were recorded after 48, or in someinstances, 72 hours of incubation.

The pollution-indicator bacteria enumerated from these samples are total coliforms(TC) and total fecal coliforms (TFC). The human pathogens, Escherichia-coli -likeorganisms (ECLO) and Streptococcus faecalis (FS), were also enumerated. The TCand ECLO were enumerated on EcoliO157 agar (Hi-Media, Mumbai). All coloniesformed on this medium were counted as TC. Typical, blue, convex, entire, 2 mmdiameter colonies were counted as ECLO. The FS and TFC were enumeratedby using rapid enterococci agar (Hi-Media, Mumbai). All colonies appearing onthis medium were counted as TFC. Bluish, entire, convex, small (less than 2 mm

June 13, 2007 20:6 RPS rpb001ch11

SEWAGE-POLLUTION INDICATOR BACTERIA 117

Table 11.1 Sampling schedule followed for enumeration of bacterial populations duringthis study.

Estuary Sampling dates Sampling strategy

Mandovi 28–29 April 2002 Water samples collected every three hours coveringa 24-hour period during spring tide

Mandovi 5–6 May 2002 Water samples collected every three hours coveringa 24-hour period during neap tide

Mandovi 5–6 September 2002 Water samples collected every three hours coveringa 24-hour period

Mandovi 12–13 March 2003 Water samples collected every three hours coveringa 24-hour period

Zuari 30 April–1 May 2002 Water samples collected every three hours coveringa 24-hour period during spring tide

Zuari 7–8 May 2002 Water samples collected every three hours coveringa 24-hour period during neap tide

Zuari 7–8 September 2002 Water samples collected every three hours coveringa 24-hour period

Zuari 14–15 March 2003 Water samples collected every three hours coveringa 24-hour period

Table 11.2 Mean ± standard deviation of counts of bacterial populations (no ml−1) during28–29 April 2002 at sampling locations in the Mandovi (M2–M7) and Zuari (Z2–Z5) estuaries.The bacterial counts are the mean values from at least eight samples from each period ofobservations. See table 11.1 for the sampling schedule followed in this study. Only onesample was collected at stations denoted with “#”. See Map B for station locations fromwhere water and sediment samples were collected.

Total Total fecal Escherichia StreptococcusStation coliforms coliform coli faecalis

M#1 15 5 9 2

M2 14 ± 10 12 ± 10 0.7 ± 0.8 9 ± 7M3 20 ± 17 15 ± 13 0.6 ± 0.6 13 ± 13M4 12 ± 6 37 ± 12 2 ± 2 27 ± 15M5 7± 5 20 ± 10 1 ± 1 11 ± 7M6 10 ± 7 20 ± 13 1 ± 1 16 ± 14M7 24 ± 27 1 ± 0.7 1 ± 1 0.4 ± 0.6Z#

1 15 3 0 0Z2 28 ± 14 32 ± 21 2 ± 2 30 ± 22Z3 18 ± 13 30 ± 22 2 ± 2 28 ± 24Z4 15 ± 7 24 ± 18 3 ± 3 23 ± 20Z5 8 ± 6 2 ± 2 0.1 ± 0.2 0.6 ± 1

June 13, 2007 20:6 RPS rpb001ch11


Table 11.3 Same as table 11.2, except during 5–6 May 2002.

Total Total fecal Escherichia StreptococcusStation coliforms coliform coli faecalis

M2 17 ± 8 16 ± 13 0.6 ± 0.8 10 ± 10M3 12 ± 6 14 ± 9 0.4 ± 0.5 8 ± 6M4 30 ± 48 23 ± 21 2 ± 3 3 ± 6M5 9 ± 10 17 ± 19 5 ± 7 9 ± 14M6 20 ± 25 15 ± 12 8 ± 11 9 ± 7M7 50 ± 62 2 ± 7 1 ± 1 0.2 ± 0.6Z2 25 ± 7 12 ± 7 6 ± 5 3 ± 3Z3 21 ± 18 14 ± 5 2 ± 2 6 ± 4Z#

4 43 8 2 7Z5 35 ± 10 0.1 ± 0.2 0.1 ± 0.2 0.04 ± 0.1

Table 11.4 Same as table 11.2, except during 5–6 September 2002.

Total Total fecal Escherichia StreptococcusStation coliforms coliforms coli faecalis

M#1 1900 800 0 1

M2 299 ± 242 46 ± 76 6 ± 4 36 ± 60M3 85 ± 61 14 ± 15 5 ± 3 6 ± 2M4 165 ± 112 24 ± 9 11 ± 5 14 ± 7M5 130 ± 73 10 ± 9 7 ± 5 4 ± 4M6 92 ± 38 84 ± 100 9 ± 5 55 ± 82M7 160 ± 190 165 ± 180 8 ± 7 161 ± 180Z#

1 57 383 0 358Z2 59 ± 29 18 ± 12 2 ± 3 8 ± 5Z3 94 ± 63 30 ± 42 5 ± 3 10 ± 8Z4 188 ± 296 1000 ± 94 2 ± 3 42 ± 56Z5 587 ± 1460 59 ± 34 2 ± 2 14 ± 13

diameter) colonies were counted as FS. Based on a recent study (Samant 2006),we have sorted out the uncertainty of ‘like organisms’ by employing an array ofbiochemical tests. By subjecting over 500 isolates designated as LO to a set of the12 most relevant biochemical tests, we found that 72.2% of ECLO are EC and76% of FSLO are FS. Using this result, “nearly true” percentages of EC and FSfrom the data on ECLO and FSLO have been presented in this chapter. Tables11.2–11.5 summarize the mean counts of different populations of bacteria in thewater samples. Table 11.6 gives their counts in sediments collected from differentlocations during this study.

June 13, 2007 20:6 RPS rpb001ch11

SEWAGE-POLLUTION INDICATOR BACTERIA 119

Table 11.5 Same as table 11.2, except during 12–13 March 2003.

Total Total fecal Escherichia StreptococcusStation coliforms coliforms coli faecalis

M#1 23 5 0 1

M2 5 ± 3 2 ± 2 0.1 ± 0.3 0.7 ± 0.7M3 6 ± 5 2 ± 3 0.3 ± 0.7 0.3 ± 0.5M4 8 ± 5 10 ± 9 0.6 ± 0.7 6 ± 5M5 4 ± 3 2 ± 2 0.5 ± 1 1 ± 1M6 93 ± 175 26 ± 68 0.6 ± 2 1 ± 2M7 36 ± 105 46 ± 136 1 ± 3 0.9 ± 3Z#

1 23 1 2 0.8Z2 37 ± 19 1 ± 1 1.56 ± 2.19 0.8 ± 0.8Z3 30 ± 22 2 ± 1 4 ± 4 0.7 ± 0.6Z4 11 ± 2 1 ± 1 3 ± 4 0.5 ± 0.7Z5 1 ± 2 0.6 ± 1 0 0

Table 11.6 Abundance (no × 105 g−1 dried sediment) of pollution-indicator and human-pathogenic bacteria in sediment samples collected from various locations in the Mandovi(M1–M7) and Zuari (Z1–Z5) during different seasons. See Map B for station locations.

Total coliforms Total fecal coliforms Escherichia coli Streptococcus faecalis

Station 1 2 3 1 2 3 1 2 1 2 3

M1 1.4 0.3 1 – 7.3 0.1 – – – – –M2 – – – – – 0.4 – – – – –M3 1.5 1.2 2.9 – 2.4 0.3 – 0.6 – – 0.1M4 219 1.5 – 0.5 0.3 – – 0.03 – 3 –M5 183 – 5.4 23.5 3.7 11.5 – – – – 0.8M6 7.3 0.1 2.1 291 0.6 20.9 – – – – –M7 17.6 0.4 10.7 7.4 4.8 87.5 – – – 1.9 1.8Z1 9.6 9.2 21 – 0.9 0.6 – 0.03 – 0.5 –Z2 – – 0.1 – – 0.1 – – – – 0.1Z3 – – – – – – – – – – –Z4 47.7 – – 15.9 0.02 – 6.4 – 1.3 0.1 –Z5 31.7 0.3 0.1 20.1 0.01 70 – – 5.1 0.2 –

1: April–May 2002; 2: September 2002; 3: March 2003. Escherichia coli were not detected inany of the samples collected during March 2003. ‘–’ implies that the given bacterial groupwas not detected.

11.3 DISCUSSION AND CONCLUSIONS

Indian standards categorizing natural water bodies for safe uses are not available.There are two well-known standards that have been used extensively in Europe andin the United States of America. The European standard, known as the “European

June 13, 2007 20:6 RPS rpb001ch11


Blue Flag Beach Criteria” (Anonymous 2002), recommends that coliform countsin excess of 5 perml in natural water are unsafe for bathing, but the standardgoes on to recommend that 5 to 100 coliform per ml are “allowed a few timesduring the season”. In essence, total coliform counts exceeding 5 perml have seriousimplications for bathers and fishers of the region. Except for gardening and use influshing of toilets, freshwater with counts higher than 5 is unacceptable for domesticuses such as washing utensils, feeding farm animals, poultry, etc. The standard in theUnited States of America has been defined by the US Environmental ProtectionAgency (USEPA). USEPA sets limits of 2 fecal or 10 total coliforms per ml ofseawater (Dufour 1984; Fujioka 2002).

As can be noted from tables 11.2–11.5, by both standards listed above, the waters ofthe Mandovi and Zuari are unfit for bathing. During September 2002 (table 11.4), thecounts of TC exceeded 100 perml at many locations. The TFC, EC, and FS were allhigher during this period of observation than those observed during other samplingperiods. This is primarily due to excessive land run-off containing raw sewage andfecal debris that support the proliferation of coliform bacteria examined. During theother observations too, there were hardly any samples that had counts of bacteriathat would be considered safe.

Every effort leading to reduction in sewage-pollution-indicating bacteria andpathogenic microbes has to be promoted and implemented. Installation of sewagetreatment plants at all the domestic settlements, avoidance of indiscriminate dis-posal of other organic wastes, and effective waste treatment measures are required.This will not only safeguard the interests of tourism-related uses (such as swim-ming, recreational fishing, surfing, water-scooter riding, etc.), but also help maintainhealthy natural ecosystems in these estuaries. Steps must be put in place to controlthe flux of raw sewage and related pollutants in these estuaries.

June 14, 2007 13:1 RPS rpb001ch12

12The khaznam of Goa

S. N. de SousaGoa State Pollution Control Board, Patto, Panaji, Goa 403 001, India

12.1 INTRODUCTION

The khaznam (khazan lands) of Goa are agricultural lands that are subject to inunda-tion by a neighbouring river from which they are protected by bundhs. (Khaznam,pôiim, mansô, mélgam, and motté are the plural forms of khazan, pôim, manas, mélog, andmottó, respectively, in Konkani.) These salty, low-lying, flat lands were originallymangrove swamps or mudflats lying along both the banks of the rivers of Goa. Thepresence of organic peat deposits at depths of 1.5–2.5m in the middle of these fieldssuggests that they were covered with a thick growth of mangrove forests some 6000years before present (Mascarenhas and Chauhan 1998). The early settlers of thisplace, who came down from the Ghats, found these lands to be fertile, reclaimedthem by constructing mud bundhs all along the river and started cultivating them.Cosme Jose Costa (undated) compares the khaznam of Goa to the ‘polders’ of TheNetherlands; the polders are agriculturally productive, flat, low lands reclaimedfrom the sea or river by constructing dikes. A difference between the polders andkhaznam is that in some polders the water is drained out with the help of pumps(steam/diesel/electric) or windmills, while in the khaznam of Goa, the water drainsout by itself through specially made sluice gates at low tide. Coastal lands like khaz-nam with paddy cultivation and fish farming are found in other parts of the westcoast of India — West Bengal and Bangladesh — but they are named differently:khar or kharvat in Maharashtra, gazzani in Karnataka, and pokkali in Kerala (Alvares2002). The agricultural lands (bheels) developed in the Ganga Delta in Bangladeshare also known as polders.

121

June 14, 2007 13:1 RPS rpb001ch12

122 S. N. DE SOUSA

The purpose of this article is to describe the working of the khaznam of Goa. Earlierdescriptions of these lands, for example Rubinoff (2001) and Sonak et al. (2005),have focused on their socio-economics. The emphasis in the present article is onthe processes behind the working of a khazan. In the next section, we provide a briefoverview of how these lands were established. Section 12.3 describes the physicalfeatures of a khazan. Section 12.4 examines a typical Goan khazan, the Corlim khazan.Section 12.5 concludes the chapter.

12.2 A BRIEF HISTORY

There are no written records to show when the early settlers occupied the presentland of Goa, and how and when they established the khaznam. This could only beachieved through slow and persistent hard work, perhaps stretching over a coupleof centuries. Xavier (1852) postulates that it was sometime between the 8th and 9thcentury A.D that the first settlers came down from the Ghats, occupied the lands,and developed the khaznam. The polders of The Netherlands came into existenceduring the 12th century. One can visualize a pre-reclamation scenario in whicha tidal estuary would have been surrounded along both its banks by mangroves.Probably it is in this state that the nomadic tribes from North Canara, who camedown the hills and settled themselves in the high lands, found the banks of theGoan estuaries. They reclaimed the marshy lands of the banks by constructing asystem of mud bundhs (to protect the lands from tidal waters) and sluice gates (todrain out water during low tide). As the settlements grew they moved to otherparts. They formed a system of gaumkary (village associations) to own and cultivatethe land collectively. They gave themselves a set of laws (mandavôlli ), punishedthe defaulters with dhônd (fine), while rewarding those who abided faithfully by therules. They developed the reclaimed flat lands into khaznam with a system of internaldrainage consisting of pôiim (internal water bodies), advónam (mud partitions acrossa pôim), and manas (sluice gate). They divided these lands (khaznam) into large plots(mélgam) and rented them to farmers for cultivating paddy on payment of sidau(annual rent), which was payable in kind (paddy), thus generating funds for thegaumkary. The profit earned, after deducting the cost of maintenance, including thesalaries of employees of gaumkary, was distributed among the gaumkar in the form ofzónn (jonos) and shares (acções). The zónn were hereditary, every male child of thegaumkar being eligible after attaining the age of 18, while the acções were shares,which were transferable (could be sold, gifted, mortgaged, etc.).

12.3 THE PHYSICAL FEATURES OF KHAZNAM

The khaznam of Goa together occupy an area of about 18,000ha of fertile lowlands,and consist in general of four main components (see figure 12.1): bundh (dyke), manas

June 14, 2007 13:1 RPS rpb001ch12

THE KHAZNAM OF GOA 123

Figure 12.1 Schematic of a khazan. The manas is essentially a valve that permits water toflow from pôiim to estuary, but not vice versa.

(sluice gate), rice fields (elevated land for farming), and pôiim (internal water bodiesconstituting drainage channels). Discussed below are each of the components.

12.3.1 Bundh

The bundh (figure 12.1) is made of the clayey soil (chikol ) dug up from the marshlandsor mudflats. It is about 2–2.5m high (from the water level of pôim) and about 1–1.5mwide at the top (about 3m wide at the base). The total combined length of thekhazan bundhs of Goa is about 2000km. One peculiar characteristic of chikol is thatit becomes hard when dry and is not easily washed out with the monsoon rainsor tidal water. Generally, no laterite stone is used in the construction of the bundhbecause the use of rough laterite stone is more damaging to the structure of thebundh, unless the laterite stone is dressed and the joints are sealed with mortar.The bundh protects the khazan from inundation with brackish water at high tide,and maintains water level in the khazan during the monsoon. A dense growth ofmangroves on the side of the estuary next to the bundh helps not only to protect itfrom erosion due to wave action, but also acts as a trap for sediment and nutrientsin the estuary. The mangroves also form a nesting ground for birds, both local andmigratory. Their thick underwater root system serves as a shelter for fish during thespawning season and protects the immature fish and young prawns from predators.

June 14, 2007 13:1 RPS rpb001ch12

124 S. N. DE SOUSA

12.3.2 Manas

The manas (sluice gate) is a simple mechanism that allows water from the pôim todrain out into the neighbouring estuary during low tide, while automatically closingto prevent brackish water from entering the khazan during high tide (figure 12.1).An opening (about 10m wide) is made in the bundh at a strategically located placethat depends on relief of the land and is located at the point of lowest elevation.The sides of this opening are strengthened with laterite masonry work (or nowadayswith cement concrete) and fitted with the manas. It helps to maintain a suitable waterlevel in the khazan during the monsoon and to keep the rice fields dry and free fromsaline water during summer.

A manas consists of a thick wooden frame (about 30cm×20cm cross-section) withwooden shutters (about 5cm thick). Figure 12.2 shows a sketch of a typical manasframe and shutters. The width of the manas and the number of its shutters vary(normally 6–8) depending on the size of the khazan. The entire structure is made ofmatti wood, which is resistant to decay from continuous exposure to saline water,and also to the action of woodborers. The entire wooden structure of the manas iscoated with a layer of dhik (extract of cashew nut shells), which acts as an antifouling

Figure 12.2 Sketch showing the construction of a manas.

June 14, 2007 13:1 RPS rpb001ch12


Figure 12.3 Estuary-side view of the Corlim manas Mangueiral, with closed shutters at hightide.

agent and protects the wood from decay due to continuous exposure to brackishwater. Application of dhik is repeated every six months. The wooden structureis then fitted into the opening in the bundh (figure 12.3), in such a manner thatthe shutters can only open into the estuary, thereby allowing water to flow out ofthe pôim. The manas thus acts as a one-way valve. The shutters, however, are notcompletely watertight and allow some leakage of estuarine brackish water into thepôim (figure 12.4). This linkage, however, has no significant impact on the workingof the system, and can be minimized through good workmanship of the manasstructure. The normal lifespan of a manas is three years, after which it is replacedwith a new structure. In the event of any part of the manas or even the entirestructure breaking down, emergency measures are taken to either repair or replaceit. In such emergencies, a 1–1.25m thick dyke (ganddó ) made of wet river clay,reinforced with laterite stones and tree trunks, is built across the manas, opening onthe estuary side at a distance of about 10m from the manas. This keeps the tidal wateraway from the manas, allowing free access to the workplace to carry out necessaryrepairs.

During high tide, as the water level in the estuary starts rising, it exerts pressure onthe shutters of the manas, forcing the shutters to close automatically. The brackish

June 14, 2007 13:1 RPS rpb001ch12

126 S. N. DE SOUSA

Figure 12.4 Pôim-side view of the manas in closed position at high tide. Note the estuarinewater leaking through the shutters.

water is prevented from entering the khazan. Some brackish water from the estuarycan seep into the khazan through the bundh during high tide. When the tide reversesduring ebb, the water level in the estuary starts falling. When the water level in theestuary falls just below that in the pôim, the shutters of the manas open, allowing thewater accumulated in the pôim to drain out into the estuary (figure 12.5).

During the monsoon, once the weeding operations in the rice fields are completed(towards the end of July), the normal functions of the manas are blocked by usingaddambó (thick wooden poles put across the manas) to prevent the shutters fromopening at low tide. This helps to maintain knee-deep water level in the rice fieldsbecause the rice crop needs continuous water supply. This blocking is done aftertaking into consideration the amount of rainfall at the time and the height of therice saplings: it is necessary to ensure that the rice crops do not get submerged. Incase there is continuous heavy rainfall when the addambó is in place, and there isfear of the crops getting submerged, the addambó is temporarily removed to allowthe water level to fall to the desired level. In normal circumstances, the addambóremains in position till the end of September or till the paddy is ready for harvesting.

June 14, 2007 13:1 RPS rpb001ch12


Figure 12.5 Estuary-side view of the manas at low tide, with the shutters just opened.

12.3.3 Rice fields

These are the flat, low-lying lands of soft soil where rice grown during the monsoonseason (kharif crop) forms the main crop. Winter crops consisting of vaingann (rabicrop of rice), beans (chôunlli ), watermelons, chillies, onions, sweet potatoes, veg-etables, pulses, etc. are grown under irrigation on the lands not affected by salinewater.

These agricultural flat lands are divided into mélgam (plots) of 0.5–1.0ha area each.The mélgam usually have regular rectangular shapes with the boundaries markedwith laterite stones fixed at the four corners. The motté (elevated mud paths onemetre wide and one metre high) criss-crossing the khazan provide access to thefarmers from the village road to the various mélgam, each mélog having direct accessto the mottó on at least one side. This allows the farmer a free passage to his fieldwithout having to pass through the fields belonging to other farmers. Although onlyone crop (kharif ) is grown, in general, in these fields, the farming operations keepthe farmer occupied practically round the year; the field having to be prepared forthe ensuing season.

June 14, 2007 13:1 RPS rpb001ch12

128 S. N. DE SOUSA

After about 5–10 years, pests (weeds, snails, leeches, field rats, etc.) grow in thekhazan, making cultivation of rice uneconomical. When such a situation arises, therice field is inundated with brackish water for about 3–4 months ( January to end-April). At the end of this period, the water is drained out and the fields are allowed todry, giving sufficient time to farmers to plough the dry fields. This operation, calledhanddém, leaves the soil salty and consequently, only a local, salt-resistant varietyof rice (kôrngutt ) can be grown during the first season after the handdém. During thehanddém period, the entire khazan acts as a large pisciculture pond where shrimps,white prawns, tiger prawns, mullets, crabs, and different types of tasty fish growabundantly.

12.3.4 Pôim

Pôiim are large internal water bodies that are interlinked with each other and con-nected to the neighbouring estuary through the manas (see figure 12.1). During sum-mer, most of them are half-filled with brackish water from the estuary, the waterhaving leaked out through the gaps in the manas and also through the many bhôm(holes) in the bundh, while the others are almost dry. During the monsoon, they getflooded with freshwater from rainfall and runoff. These water bodies also serve asdrainage channels for the village storm water, connecting the estuary to the villagedrains through channels (valls). These 1–1.5m deep water bodies having irregulargeometric shapes serve as water receptacles for the khazan, holding all the brackishwater leaking through the manas at high tide, the water subsequently draining outduring low tide. They have an almost flat bottom and vertically cut sides, which areoften lined with dressed laterite stones.

De-silting of the pôim on a regular basis forms one of the important parts ofthe maintenance of a khazan. De-silting involves digging out the chikôl from thebottom of the pôim and staking it on its banks till it is dry. The dry chikôl isthen spread over the neighbouring rice fields. This operation helps to maintainthe water-carrying capacity of the pôim, while the chikôl spread in the fields hasdual functions — maintaining the level of the fields and acting as a naturalfertilizer.

Different types of tasty brackish-water fish are grown and captured in the pôiim.Fish species include mullets, pearl spot, catfish, chonnak, milkfish, tilapia, forsandi,etc. White prawns, tiger prawns, and crabs are also found in plenty. A variety ofpoisonous sea snakes living in holes pose a constant threat to the fishermen ventur-ing into the pôim. Fishing activity in the khazan has traditionally been an extensionof the agricultural system, a by-product of the reclamation of the wetland. Greaterimportance was earlier given to the promotion of agriculture and its improvement.With the passage of time, however, fishing started gaining importance owing tothe high demand for fish. At present, fishing appears to be the main activity in thekhaznam as it is more lucrative and involves less labour.

June 14, 2007 13:1 RPS rpb001ch12


Figure 12.6 Estuary-side view of the manas at low tide, with open shutters and the fishingnet in position.

Fishing at the manas is done using a special type of net (bag net). As the shutters ofthe manas open and the water starts flowing out into the estuary during low tide, thenet is fixed at the mouth of the manas and secured in open position with the help ofpoles and ropes (figure 12.6). Normally, fishing at the manas is carried out at nightto catch prawns. Fishing in the pôiim is done using different types of nets — kanttalli(gill net), paguer, rampôn, ormól, ênddi, zalli, etc.

12.4 THE CORLIM KHAZAN: A TYPICAL GOAN KHAZAN

12.4.1 Geography of the Corlim khazan

Corlim is a village forming part of the Tiswadi Island of Ilhas Taluka of Goa, lyingabout 13km to the east of Panaji. It is bounded on the east by the Cumbarjua Canal,on the west by the villages of Ela and Carambolim, on the north by the village ofGandaulim, and on the south by the village of Carambolim and the CumbarjuaCanal.

June 14, 2007 13:1 RPS rpb001ch12

130 S. N. DE SOUSA

Figure 12.7 The Corlim khazan. Water level was measured using tide poles at locations #6(in the pôim) and #7 (in the estuary) (see figure 12.8a). Surface water samples were collectedfrom location #1 (in the estuary) and #2–#5 (in the pôim) for salinity analysis (see figure12.8b).

The Corlim khazan (figure 12.7) is situated towards the south and southeast end ofthe village, bounded on the south and southeast by the Cumbarjua Canal, fromwhich it is protected by a bundh, and in the southwest by the Carambolim khazan.The Corlim khazan, in fact, consists of two khaznam — Novem khazan in the east andMangueiral in the west, each one having its own manas and a pôim. Together, the twokhaznam occupy an area of about 182ha, out of which the cultivable area (rice fields)consists of about 153ha, the remaining area of 29.4ha being occupied jointly by thetwo pôiim. Assuming an average depth of 1 m, the pôiim can hold roughly 294 ×103 m3 (294 million litres) of water. The manas of Novem Khazan is known as ‘SamTiago’ (A in figure 12.7), while that of Mangueiral bears the name of its khazan, i.e.,‘Mangueiral’ (B in figure 12.7) (Mangueiral is a Portuguese word meaning ‘MangoOrchard’).

June 14, 2007 13:1 RPS rpb001ch12


The two pôiim are separated from each other by a 2m wide advónn (C), whichprovides the farmers access to the fields lying on the other side of the pôim. A 2-mwide gap across the advónn allows free exchange of water between the two pôiim. Anetwork of motté criss-crossing the khazan and linking the village road (H) providefree access to farmers to any field. An ancient fort (khótt), with circular watchtowersat regular intervals, cuts through the Mangueiral khazan and enters into the khazanof Carambolim village. Though only the ruins of the fort can be seen at present, athick-walled, approximately 3-m-wide masonry gate made of huge laterite stonesis still found in good condition at the point where the fort enters the Carambolimkhazan. This gate was probably being used by the inhabitants of the walled city ofOld Goa to move in and out, and to bring in food supplies and other materials afterlanding at the Mangueiral landing site. The relative positions of the two mansô —Sam Tiago and Mangueiral (A and B), advónn (C), khótt (D), bundh (F), vall (G), andthe nearest village road (H) are shown in figure 12.7, which also shows the khazanrice fields and the network of pôiim.

The khaznam are protected from the tidal waters of the Cumbarjua Canal by abundh (F), at the two extremities of which lie the two mansô, Sam Tiago (A) andMangueiral (B). The total length of the bundh is approximately 1.6km. Each manasconsists of eight shutters made of 2-inch thick planks of matti wood. The pôiim areconnected to the storm drains of the village through valls (G), which help to drainout the rainwater from the village into the estuary. During the monsoon, the entiresystem acts as a large aquaculture farm and breeding ground for brackish waterfish because there is bountiful food in the form of insects and their larvae (flies,caterpillars, locusts, and fish larvae), so that when the addambó is removed, thereis a big haul of fish at the manas lasting for a few days immediately following theremoval of addambó. The associated ‘Festa de Novidades’ is celebrated on 24 Augusteach year.

Most of the mines in Goa are situated in the basin of river Mandovi. The ore,after being mined at the mining sites situated in the hinterland, is transported bytrucks and staked on the banks of the river, from where it is loaded on to barges andtransported, via Aguada bar mouth, to Mormugao for export. During the monsoon,however, the barge traffic through the Aguada bar mouth is closed owing to theformation of a sand bar at the mouth. Consequently, all the barge traffic in theMandovi River — the traffic is much less during this season — is diverted to the ZuariRiver via the Cumbarjua Canal. The bow-shock waves caused by the movementof barges in the narrow channel causes damage to the bundh structure owing to theerosive action of the waves, leading to breaches (khavttém) in the bundh. One suchkhavttém was present in the bundh of the neighbouring khazan of Carambolim villageduring the course of this study. According to Mr. Damodar Phadte, Chairman,Corlim Khazan Tenants’ Association, all the efforts of the State Government to repairthe damaged bundh had failed owing to procedural delays between the inspectionand preparation of cost estimates to sanction the work order, leading to amplificationof the damage and escalation of cost estimates. Consequently, the whole process

June 14, 2007 13:1 RPS rpb001ch12

132 S. N. DE SOUSA

had to be repeated all over again. This has led to continuous inundation of theCarambolim khazan over a long period of time. During the spring high tide, thewater level in the inundated field rises so high that the brackish water overflowsinto the neighbouring Corlim khazan, as discussed below.

12.4.2 Tidal variation in the pôiim

To check the effectiveness of the bundh–manas–pôiim system in Corlim khazan, threesets of field observations were carried out to map the salinity distribution and vari-ation in water level in the pôiim and the Cumbarjua Canal.

During the first observation, surface water samples were collected from 5 locations(marked #1–#5 in figure 12.7) during the first week of June 2005, just before theonset of the monsoon. The water samples were analyzed for salinity. During thesecond observation carried out on 13–14 October 2005, variations in water levelwere measured in the estuary (Cumbarjua Canal) and the Mangueiral pôim using tidepoles fixed about 10–12m apart in the two water bodies across the manas (figure 12.7,locations #6 and #7). Water level measurements were made simultaneously fromthe two tide poles at 15 minutes intervals for 25 hours. Figure 12.8(a) shows thevariation of water level in the two water bodies. While the water level in the estuaryshowed large variations with two peaks and two troughs in a day and a rangeof 163cm, the level in the pôim varied by just 16cm. The shutters of the manasremained closed as long as the water level was higher in the estuary. The shuttersopened when the water level in the estuary fell below that of the pôim. When thefirst low water occurred at 1430 hours on 13 October 2005, the shutters of manas didnot open because the water level in the estuary did not fall below that in the pôim.The shutters opened during the next two low waters that occurred at 0145 hoursand 1430 hours on 14 October 2005. A similar situation arises during every neaptide when the water level in the estuary does not fall below that in the pôim evenduring low tide. Under such conditions, there is continuous influx of brackish waterfrom the estuary into the pôim owing to leakage through the manas.

At the start of the flood tide (point D in figure 12.8a), when the shutters of the manaswere in open position, as the water level in the estuary started rising, the water levelin the pôim continued to fall till the same level (E) was reached in both the waterbodies. As soon as the water level in the estuary rose above that in the pôim, theshutters closed. At this point, the water level in both water bodies started rising, butat different rates: very slowly in the pôim and rapidly in the estuary. This continuedtill the tide peaked in the estuary (F). At this point, the water level in the estuarystarted falling, but the level in the pôim continued to rise till both water bodies cameto the same level. When the level in the estuary fell below that in the pôim (G), theshutters opened and the water level in the pôim started falling.

In general, there are two factors causing a rise in water level in the pôim whenthe shutters are closed: (a) flow of groundwater into the pôim, and (b) leakage of

June 14, 2007 13:1 RPS rpb001ch12


2

4

6

8

10

12

Sal

inity

(ps

u)

12 24 36Time (hours)

(b)

13Oct05 12:00 14Oct05 12:00

estuarypoim

50

100

150

200

250T

ide

heig

ht (

cm)

(a)

A

B

C

D

E

F estuarypoim

Figure 12.8 Observations in the estuary (red; location #7) and pôim (blue), during13–14 October 2005. (a) Water level (cm) and (b) salinity (psu). See figure 12.7 for themeasurement locations.

brackish water from the estuary through the manas and bundh. In addition, there areinputs from rain and surface runoff, which, however, are significant only during themonsoon and during the period immediately following the monsoon (see chapter 1).The water level in the estuary is controlled mainly by the tide during the summermonths, with the river flow having an enhancing effect during the flood tide. During

June 14, 2007 13:1 RPS rpb001ch12

134 S. N. DE SOUSA

the monsoon, however, both river flow and surface runoff exert significant influenceon the water level in the estuary during both flood and ebb tides.

12.4.3 Salinity distribution in the pôiim

In the first set of observations made on 4 June 2005, i.e., just before the onset of themonsoon, surface water samples were collected from 5 different locations (locations#1–#5 in figure 12.7). Location #1 is in the Cumbarjua Canal and locations #2–#5are situated in the pôiim (locations #2–#4 in the pôim of Novem Khazan, which isunder the influence of manas Sam Tiago (A), and location #5 in the pôim underthe influence of manas Mangueiral (B)). The water samples were analyzed using the‘Autosal’ salinometer.

Salinity was highest (31.31psu) at location #1 in the estuary. Salinity decreased inthe pôim with increasing distance from the manas. It was 31.12psu at location #2, justinside the manas, and 30.31psu at location #4, which is the sampling point farthestfrom the manas, in the pôim Novem Khazan. Salinity at location #3 was 30.90psu.Location #5, which comes under the influence of the Mangueiral manas showeda relatively higher salinity (30.44psu). Generally, the salinity in the neighbouringestuary (Cumbarjua Canal) is highest during this time of the year, just before theonset of the monsoon. Also, during this time, the freshwater contribution fromgroundwater is negligible. Consequently, salinity in the pôiim is highest at this timeduring the year owing to leakage of brackish water through the manas and bhôm inthe bundh. The decrease in salinity with increasing distance from the manas indicatesthat the leakage does not overwhelm the freshwater accumulated earlier in the pôiim.

In the second set of observations carried out at the Mangueiral manas on 13–14 Octo-ber 2005 soon after the removal of addambó, surface water samples were collectedfrom the two tide pole locations #6 and #7 across the manas at hourly intervals for25 hours, and the samples were analyzed for salinity. As the observations were inprogress, a heavy post-monsoon shower lasting for about 2–3 hours lashed the areabetween 1500 and 2000 hours. The results of salinity measurements are shown infigure 12.8(b). During the first half of the tidal cycle, the salinity in the pôim (5.31–11.35psu) was consistently higher than that of the estuary (3.03–9.27psu), with bothlocations showing salinity maxima and minima that did not coincide with the highand low water in the estuary, as would normally be the case. The salinity in bothestuary and pôim dropped steeply and remained low when the manas remainedclosed. Incidentally, this period also coincided with the shower, and consequently,the drop in salinity may indicate the localized effect of rainfall and surface runoff.The moment the shutters of the manas opened, allowing water to flow out from thepôim into the estuary, the salinity at both locations shot up to the maximum levels(10.94–11.35psu and 6.51–9.27psu in the pôim and estuary, respectively), indicat-ing that the water in the pôim is relatively more saline than that in the estuary. Thiswas an altogether unexpected observation, as the general impression is that theestuarine water is always more saline than that in the pôim.

June 14, 2007 13:1 RPS rpb001ch12


The most likely reason for the above observation is the following. Immediatelyafter the monsoon, salinity in the estuary is low and its value depends on localrainfall, runoff, and groundwater flow. Depending on the relative contribution ofthese sources of freshwater, salinity in the estuary can sometimes be lower than thatin the pôiim at this time.

The above discussion shows that salinity distribution in the pôim during the post-monsoon season is not controlled by a simple process of mixing of estuarine brack-ish water and fresh groundwater, but by several factors that have to be accounted for.

• Leakage of estuarine water through the manas during high tide introduces thishigher-salinity water into the pôim. This higher-salinity water remains trappedin the pôim in the vicinity of the manas owing to the near-stagnant nature of thewater in the pôim: water circulation in the pôim is almost non-existent, the outflowof water from the pôim during low tide being very small compared to its volume.

• Rainfall, surface runoff (whenever it occurs), and groundwater inject freshwaterinto the pôim and reduce the salinity there. After the withdrawal of the monsoon,leakage across the manas increases the salinity in the pôiim continuously.

• A large breach (khavttém) in the bundh of the neighbouring khazan of Caram-bolim village had led to the inundation of this khazan with high-salinity (∼31psu)brackish water from the Cumbarjua Canal during the summer months. Dur-ing the spring high waters, this high-salinity water overflowed into the Corlimkhazan, inundating the low-lying fields around the Mangueiral pôim, while duringthe neap tides, the fields dried up. Successive inundations followed by drying ofthese fields and evaporation have rendered their soil highly saline, leaving a thinlayer of crystals of salt at the surface. Factors such as these need to be taken intoaccount to understand the salinity variation in the pôiim.

• During the monsoon, although the fields are flooded with freshwater, the saltsleaching out from the soil are retained within the system owing to the presenceof the addambó at the manas. After the removal of addambo, the low-lying fieldsaround the pôim continue to remain under ankle-deep water for a long time.Continuous evaporation of this water over large shallow areas renders it highlysaline. This water, which is contiguous with the pôim, enters the pôim during lowtide and raises the salinity there.

12.5 CONCLUDING REMARKS

The passing of the Goa, Daman, and Diu Agricultural Tenancy Act, Fifth Amend-ment, 1976, by the Goa Legislative Assembly (Government of Goa 1976), gave thefarmers ownership rights over the lands cultivated by them. This right gave thema sense of security and freedom from the strict laws of ‘Código das Comunidades’,with its clauses for punishment with heavy fine and loss of land for any irregu-larities in the cultivation of the tenanted lands. The farmers became conscious of

June 14, 2007 13:1 RPS rpb001ch12

136 S. N. DE SOUSA

their rights, but not of their duties. They started neglecting the cultivation of thelands, and even misused it for purposes other than agriculture — like pisciculture,real estate development, etc. At present, about 80% of the fields are left fallow,only 20% being cultivated. According to Rubinoff (2001), the practice of rice pro-duction and fish farming in Goa is now reversed. In the past, the fishing activitieswere secondary as the sluice gates existed to protect the khaznam from inundation:the fields did not exist for the gates. The fields have now become secondary, andfishing has become one of the major economic activities in the khaznam. The rea-sons for the recent neglect of agriculture in Corlim khazan are the same as thosementioned above. One additional factor specific to this area is that Corlim being anindustrial area, with many industries opening their factories here, there are plentyof job opportunities for both skilled as well as unskilled workers. Consequently,people prefer to work in industries with a secure job, rather than to work in thefields exposed to the elements of nature.

It is sad to note here that these rather special systems, the khaznam of Goa, whichhave survived and even thrived for over a millennium, may well be on their wayout. If this happens, the khaznam will become yet another victim of the fast evolvingsocio-economic environment of our times. It is necessary that we take measures toensure that at least a few examples of this intricate system survive in the traditionalway as samples of our heritage. We should also ensure that we record, to the bestof our ability, how the system works. Steps such as these will remind the Goancommunity of the sophisticated agriculture–fishery system that has been a part ofits culture, and serve to expose the tourists who visit Goa to an eco-friendly systemin which man worked with nature to sustain life.

June 13, 2007 18:57 RPS rpb001_epilogue

Epilogue

As noted in the prologue, the primary objective of putting together this book hasbeen to compile the information and knowledge that exists on the Mandovi andZuari estuaries —- two of the most studied estuaries of India. The 12 chapters of thisbook provide an overview of what we know, and in comparison, of what remainsto be discovered. The purpose of this chapter is to reflect on both these issues, andon how we could discover more about other estuaries of India.

The description of the Mandovi and Zuari estuaries provided in this book revealssome distinct characteristics that we summarize here,

• There are advantages in looking at each estuary as consisting of two parts: a bayand a channel. Conditions in the bay are strongly influenced by what happenson the shelf. What happens in the channel is influenced by what goes on in thebay at its downstream end and (in the freshwater source, or river) at its upstreamend.

• Both the bay and the channel go through a distinct annual cycle. There are twodistinct phases to this cycle,

• The wet phase (roughly June to mid-October) is the time of flushing the estu-aries many times over, thus expelling everything that existed in the water (salt,plankton, etc.); benthic organisms disappear due to the change.

• The dry phase (mid-October to May) is the time of rejuvenation. Phytoplanktongrows; so does zooplankton, and benthic organisms re-colonize the bed. Saltmigrates upstream into the channels of the two estuaries.

In essence, two spatial regimes, each with two distinct temporal variations, dominatethe behaviour of these estuaries. This book provides a glimpse of the behaviour.

Yes, it is only a glimpse, because we are not yet ready for more. While we do havea fair idea of some of the physical processes —- tides, for example —-, there are other

137

June 13, 2007 18:57 RPS rpb001_epilogue

138 EPILOGUE

aspects we know little about. This book has not discussed tidal currents in the estu-aries because we do not have good records to describe them. The channels of theestuaries are too much in use to permit deployment of current meter moorings forsufficiently long periods to allow proper tidal analysis (about a month). The avail-able description of stratification is at best rudimentary, and no effort has yet beenmade to quantify the upstream migration of salt during the dry phase. We only havea glimpse of how some of the fundamental variables associated with biogeochemi-cal cycling —- nitrates, phytoplankton, and zooplankton —- behave. Needed are data,more extensive and more systematically collected, to describe the spatial and tem-poral variability revealed in the chapters of this book. Nevertheless, these glimpsescan help design new field and laboratory experiments to describe the evolution ofthe estuaries with greater confidence than is possible today.

Another challenge today is to describe quantitatively the processes revealed inthese chapters using mathematical models. Such efforts have not progressed beyondsimulation of tides, and, may be, vertically averaged salinity in the two estuaries. It istime to take the next step, and launch efforts to model biogeochemical cycling, eventhough we are not particularly happy with the description of the cycling processesthat are operating in the estuaries. Perhaps the efforts to model the processes willprovide new insights that would help in designing field experiments to describebiogeochemical cycling more comprehensively and efficiently.

In summary, though the material presented in this book documents the advancesmade since Qasim (2003) in our understanding of the Mandovi and Zuari estuaries,it is evident that much still remains to be done.

We hope this book will inspire studies such as this on the Mandovi and Zuariestuaries. Our fondest hope, however, is that the faculty of a university department,or a college located near an estuary, would, after reading this book, consider theestuary a natural laboratory that can be used to formulate student research projects.

Editors

June 13, 2007 20:15 RPS rpb001_ref

References

Achuthankutty C. T., Ramaiah N. and Padmavati G. (1997) Zooplankton variability andcopepod assemblage in the coastal and estuarine waters of Goa along the central westcoast of India; Intergovernmental Oceanographic Commission, 1 workshop report no.142 (eds) Pierrot-Bults A. C. and Vander Spoel S., UNESCO, Paris, pp 1–11.

Alongi D. (1990) The ecology of tropical soft-bottom benthic ecosystems; Oceanography andMarine Biology: An Annual Review 28 381–496.

Álvares Claude (2002) Fish, Curry and Rice (Revised 4th edn.). Goa Foundation.

Alzieu C. (2000) Environmental impact of TBT: the French experience; Science of the TotalEnvironment 258 99–102.

Alzieu C. (2006) Effects of tributyltin pollution on oyster industry: the Arcachon Bay case;In: Multiple dimensions of global environmental change (ed.) Sangeeta Sonak, TERI Press,New Delhi, pp 444–458.

American Society for Microbiology (1957) Manual of Microbiological Methods; ASM Press,Washington DC.

Anonymous (2002) Guidance notes to the European Blue Flag Beach Criteria; 29p.

Ansari Z. A., Ingole B. S. and Parulekar A. H. (1986) Spatial and temporal changes inbenthic macrofauna from Mandovi and Zuari estuaries of Goa, west coast of India;Indian Journal of Marine Sciences 15 223–229.

Ansari Z. A., Ingole B. S. and Furtado R. (2003) Response of benthic fauna to different pollu-tants: Some case studies and relevance of benthos to environmental impact assessment;In: Recent advances in environmental sciences (ed.) Hiremath K. G. (Discovery PublishingHouse) pp 422–428.

139

June 13, 2007 20:15 RPS rpb001_ref

140 REFERENCES

Aparna M., Shetye S. R., Shankar D., Shenoi S. S. C., Mehra P. and Desai R. G. P. (2005)Estimating the seaward extent of sea breeze from QuickSCAT scatterometry; GeophysicalResearch Letters 32 doi:10.1029/2005GL023107.

Azam F. T., Fenchel J. G., Field J. S., Gray L. A., Mayer R. and Thingstad F. (1983) Theecological role of water column microbes in the Sea; Marine Ecology Progress Series 10257–263.

Bhosle N. B. (2006) Butyltin compounds in biofilms and marine organisms from DonaPaula Bay west coast of India; In: Multiple dimensions of global environmental change (ed.)Sangeeta Sonak (New Delhi, TERI Press) pp 432–443.

Brock T., Madigan M. T., Martinko J. M. and Parker J. (1994) Biology of Microorganisms(7th edn.). Prentice Hall, New Jersey.

Bruland K. W., Franks, Knauer G. A. and Martin J. H. (1979) Sampling and analyticalmethods for the determination of copper, cadmium, zinc and nickel at nanogram perliter level in seawater; Analytica Chimica Acta 105 233–245.

Bhattathiri R. M. S., Devassy V. P. and Bhargava R. M. (1976) Production at different trophiclevels in the estuarine system of Goa; Indian Journal of Marine Sciences 5 83–86.

Coe M. T. (2000) Modeling terrestrial hydrological systems at the continental scale: Testingthe accuracy of an atmospheric GCM; Journal of Climate 13 686–704.

Costa, Cosme Jose (undated) Goa and her khajans. Instituto Menezes Braganca.

Danielsson L.-G. (1980) Cadmium, cobalt, copper, iron, lead, nickel and zinc in Indianocean waters; Marine Chemistry 8 199–225.

Dauer D. M. and Conner W. G. (1980) Effect of moderate sewage input on benthic poly-chaete populations; Estuarine and Coastal Marine Science 10 335–362.

de Sousa S. N. (1983) Studies on the behaviour of nutrients in the Mandovi estuary duringpremonsoon; Estuarine, Coastal and Shelf Science 16 299–308.

Dehadrai P. V. and Bhargava R. M. S. (1972) Seasonal organic production in relation toenvironmental features in Mandovi and Zuari estuaries, Goa; Indian Journal of MarineSciences 1 52–56.

Devassy V. P. (1983) Plankton production associated with cold water incursion into theestuarine environment; Mahasagar 16 221–233.

Devassy V. P. and Goes J. I. (1988) Phytoplankton community structure and succession ina tropical estuarine complex (central west coast of India); Estuarine, Coastal and ShelfScience 27 671–685.

Devassy V. P. and Goes J. I. (1989) Seasonal patterns of phytoplankton biomass and pro-ductivity in a tropical estuarine complex (west coast of India); Proceedings of the IndianAcademy of Sciences (Plant Sciences) 99 485–501.

June 13, 2007 20:15 RPS rpb001_ref

REFERENCES 141

Dufour A. P. (1984) Bacterial indicators of recreational water quality; Canadian Journal ofPublic Health 75 49–56.

ETOPO2 (2007) Online documentation. http://www. ngdc. noaa. gov/mgg/fliers/ 06mgg01. html .

Friedrichs C. T. and Aubrey D. G. (1994) Tidal propagation in strongly convergent channels;Journal of Geophysical Research 99 3321–3336.

Fujioka R. (2002) Microbial indicators of marine recreational water quality; In: Manual ofEnvironmental Microbiology, 2nd edn. (eds) Hurst C. J., Crawford R. L., Knudsen G.,McIneney M. J. and Stetzenbach L. D., American Society for Microbiology Press,Washington DC, pp 234–243.

Gauns M. (2000) Role of microzooplankton in the food chain dynamics of some tropicalmarine environments; Ph. D. Thesis, Goa University, India, pp. 220.

Gerlach S. A. (1978) Food chain relationship in subtidal silty sand marine sediment and therole of meiofauna in stimulating bacterial productivity; Oecologia 33 55–69.

Geyer W. R., Townbridge J. H. and Bowen M. M. (2000) The Dynamics of a Partially MixedEstuary; Journal of Physical Oceanography 30 2035–2048.

GLOBE (2004) The Global Land One kilometer Base Elevation (GLOBE) Digital Eleva-tion Model. Version 1. 0. National Oceanic and Atmospheric Administration, NationalGeophysical Data Center, http://www.ngdc.noaa.gov/mgg/topo/globe.html.

Godhantaraman N. (1994) Species composition and abundance of tintinnids and copepodsin the Pichavaram mangroves (South India); Ciencias Marinas 20 371–391.

Goswami S. C. (1983) Coexistence and succession of copepod species in the Mandovi andZuari estuaries, Goa; Mahasagar 16 251–258.

Government of Goa (1976) Goa, Daman and Diu Agricultural Tenancy Act, Fifth Ammend-ment (GDD 17 of 1976, S3). Government Gazette (Extraordinary) No. 29, dated14/10/1976.

Grasshoff K., Ehrhardt M. and Kremling K. (1983) Methods of Seawater Analysis. VerlagChemie, New York, NY, 419 pp.

Harleman D. R. F. and Lee C. H. (1969) The computation of tides and currents in estuariesand canals; Technical bulletin 16, M. I. T., Massachusetts, 264 pp.

Hoch M. (2001) Organotin compounds in the environment – an overview; Appl.Organometallic Chemistry 16 719–743.

Hwang H. M., Oh J. R., Kahng S. H. and Lec K. W. (1999) Tributyltin compounds in mussels,oysters, and sediments of Chinhae Bay Korea; Marine Environmental Research 47 61–70.

Jay D. A. and Musiak J. M. (1994) Particle trapping in estuarine tidal flows; Journal ofGeophysical Research 99 445–461.

June 13, 2007 20:15 RPS rpb001_ref

142 REFERENCES

Jay D. A. and Musiak J. M. (1996) Internal tidal asymmetry in channel flows: Origins andconsequences; In: Mixing Processes in Estuarine and Coastal Seas (ed.) Pattiaratchi C.,AGU, Washington D. C., pp 211–249.

Jay D. A. and Smith J. D. (1990) Residual circulation in shallow estuaries, 2, Weakly stratifiedand partially mixed estuaries; Journal of Geophysical Research 95 733–748.

JGOFS (1994) Protocols for the Joint Global Ocean Flux Study ( JGOFS) core measure-ments, Manual and guides 29, Scientific Committee on Oceanic Research, UnitedNations Educational, Scientific and Cultural Organization, Paris, 170 pp.

Jyothibabu R., Madhu N. V., Jayalakshmi K. V., Balachandran K. K., Shiyas C. A., MartinG. D. and Nair K. K. C. (2006) Impact of fresh water influx on microzooplanktonmediated food web in a tropical estuary (Cochin backwaters), India; Estuarine, Coastaland Shelf Science 69 505–515.

Krishnakumari L., Bhattathiri P. M. A., Matondkar S. G. P. and John J. (2002) Primaryproductivity in Mandovi–Zuari estuaries in Goa; Journal of the Marine Biological Associationof India 44 1–13.

Kristensen E. M., Jensen M. H. and Anderson T. K. (1985) The impact of polychaete (Nereisvirens Sars) burrows on nitrification and nitrate reduction in sediments; Journal of Exper-imental Marine Biology and Ecology 85 75–91.

Leendertse J. J. and Gritton E. C. (1971) A water quality simulation model for well mixedestuaries; Vol. II, computation procedure; R-708-NYC, The New York City Rand Institutetechnical report, 53 p.

Madhu N. V., Jyothibabu R., Balachandran K. K., Honey U. K., Martin G. D., Vijay J. G.,Shiyas C. A., Gupta G. V. M. and Achuthankutty C. T (2007) Monsoonal variability inplanktonic standing stock and abundance in a tropical estuary (Cochin Backwaters –India); Estuarine, Coastal and Shelf Science 3 54–64.

Madhupratap M. (1987) Status and strategy of zooplankton of tropical Indian estuaries: areview; Bulletin of Plankton Society of Japan 34 65–81.

Manoj N. T. and Unnikrishnan A. S. (2007) Tidal circulation and salinity distribution inMandovi–Zuari estuaries; A numerical model study (unpublished manuscript).

Mascarenhas A. and Chauhan O. S. (1998) A note on ancient mangroves of Goa, centralwest coast of India; Indian Journal of Marine Sciences 27 473–476.

McIntyre A. D. (1977) Effect of pollution on inshore benthos; In: Ecology of marine benthos(ed.) Coull B. C. (University of South Carolina Press) pp. 301–312.

Miller C. B. (1983) The zooplankton of estuaries; In: Estuaries and enclosed seas (ed.) KetchumB. H. (Amsterdam: Elsevier Scientific Publishing Company) pp. 103–149.

Miranda de L. B., Castro de B. M. and Kjerfve B. (1998) Circulation and Mixing Due toTidal Forcing in the Bertioga Channel, Sao Paolo, Brazil; Estuaries 21 204–214.

June 13, 2007 20:15 RPS rpb001_ref

REFERENCES 143

Nair R. R., Hashimi N. H. and Gupta M. V. S. (1979) Holocene limestones of part of thewestern continental shelf of India; Journal of Geological Society of India 20 17–23.

Naqvi S. W. A, Naik H., Jayakumar D. A., Shailaja M. S. and Narvekar P. V. (2006) Seasonaloxygen deficiency over the western continental shelf of India; In: Past and Present watercolumn anoxia (ed.) Neretin L. N., NATO Sci. Ser. IV: Earth and Environmental Science 64,Dordrecht, The Netherlands: Springer, 195–224.

Nayak R. K. and Shetye S. R. (2003) Tides in the Gulf of Khambhat, west coast of India;Estuarine, Coastal and Shelf Science 57 249–254.

Neetu S., Shetye S. R. and Chandramohan P. (2006) Impact of sea breeze on the wind-seasoff Goa, west coast of India; Journal of Earth System Science 115 229–234.

Officer C. B. (1976) Physical oceanography of estuaries (and associated coastal waters); (NewYork: John Wiley & Sons) 465 p.

Padmavati G. and Goswami S. C. (1996) Zooplankton ecology in the Mandovi–Zuariestuarine system of Goa, west coast of India; Indian Journal of Marine Sciences 25268–273.

Panikkar N. K. (1969) New perspectives in estuarine biology; Proc. all India Symp. inEstuarine Biology, Madras, 27–30 December, 8 pp.

Parulekar A. H., Dhargalkar V. K. and Singbal S. Y. S. (1980) Benthic studies in Goa estuaries.3. Annual cycle of Macrofaunal distribution, production and trophic relation; IndianJournal of Marine Sciences 9 189–200.

Parulekar A. H. and Dwivedi S. N. (1973) Ecology of benthic production during southwestmonsoon in an estuarine complex of Goa. Recent Researches in Estuarine Biology. (ed.)Natarajan R. (Delhi: Hindustan Publ. Co.) pp 21–30.

Parulekar A. H., Harkantra S. N. and Ansari Z. A. (1982) Benthic production and assessmentof demersal fishery resources of the Indian seas; Indian Journal of Marine Sciences 11107–114.

Qasim S. Z. (1977) Biological productivity of the Indian Ocean; Indian Journal of MarineSciences 6 122–137.

Qasim S. Z. and Sengupta R. (1981) Environmental characteristics of the Mandovi–Zuariestuarine system in Goa; Estuarine, Coastal and Shelf Science 13 557–578.

Qasim S. Z. (2003) Indian Estuaries. Allied Publishers, 420 pp.

Rainer S. (1981) Temporal pattern in the structure of Macrobenthic communities of anAustralian estuary; Estuarine and Coastal Marine Science 13 597–619.

Rao T. S. S. (1976a) Salinity and distribution of brackish warm water zooplanktonin India estuaries; Proceedings of the Symposium on warm water zooplankton,196–204.

June 13, 2007 20:15 RPS rpb001_ref

144 REFERENCES

Rao V. P., Montaggioni L., Vora K. H., Almeida F., Rao K. M. and Rajagopalan G. (2003)Significance of relic carbonate deposits along the central and southwestern margin ofIndia for late Quaternary environmental and sea level changes; Sedimentary Geology 15995–111.

Rao Y. P. (1976b) Southwest Monsoon; India Meteorological Monograph (Synoptic Meteorology),No. 1/1976, Delhi, 376 pp.

Robertson J. R. (1983) Predation by estuarine zooplankton on tintinnid ciliates; Estuarine,Coastal and Shelf Science 25 581–591.

Rubinoff J. A. (2001) Pink Gold: Transformation of backwater aquaculture on Goa’s KhazanLands; Economic & Political Weekly, 36:13, pp. 1108–1114. Memorial Library Period-ical: AP E19 A543.

Samant D. (2006) Abundance of pollution indicator and certain human pathogenic bacteriain Mandovi–Zuari estuaries; M. Sc. Thesis. University of Goa, 63p.

Sankaranarayanan V. N. and Qasim S. Z. (1969) Nutrients of the Cochin backwaters inrelation to environmental characteristics; Marine Biology 2 236–247.

Shankar D., Kotamraju V. and Shetye S. R. (2004) A quantitative framework for estimatingwater resources in India; Current Science 86 543–552.

Shetye S. R., Gouveia A. D., Shenoi S. S. C., Sundar D., Michael G. S., Almeida A. M. andSantanam K. (1990) Hydrography and circulation off the west coast of India during thesouthwest monsoon 1987; Journal of Marine Research 48 359–378.

Shetye S. R., Gouveia A. D., Singbal S. Y., Naik C. G., Sundar D., Michael G. S. andNampoothiri G. (1995) Propagation of tides in the Mandovi–Zuari estuarine network;Proceedings of the Indian Academy of Sciences (Earth and Planetary Sciences) 104 667–682.

Shetye S. R. and Murty C. S. (1987) Seasonal variation of salinity in the Zuari estuary,Goa, India; Proceedings of the Indian Academy of Sciences (Earth and Planetary Sciences) 96249–257.

Shetye S. R., Shenoi S. S. C., Antony M. K. and Krishnakumar V. (1985) Monthly meanwind stress along the coast of the north Indian Ocean; Proceedings of the Indian Academyof Sciences (Earth and Planetary Sciences) 94 129–137.

Simpson J. H., Brown J. Matthews J. and Allen G. (1990) Tidal Straining, Density Currents,and Stirring in the Control of Estuarine Stratification; Estuaries 13 125–132.

Sonak S., Kazi S. and Abraham M. (2005) Khazans in troubled waters (TERI Press, TheEnergy Research Institute, New Delhi).

Stacey M. T., Burau J. R. and Monismith S. G. (2001) Creation of residual flows in a partiallystratified estuary; Journal of Geophysical Research 106 17,013–17,037.

June 13, 2007 20:15 RPS rpb001_ref

REFERENCES 145

Stoecker D. K. and Capuzzo J. M. (1990) Predation on protozoa: its importance to zoo-plankton; Journal of Plankton Research 12 891–908.

Sundar D. and Shetye S. R. (2005) Tides in the Mandovi and Zuari estuaries, Goa, westcoast of India; Journal of Earth System Science 114 493–503.

Suprit K. and Shankar D. (2007) Resolving orographic rainfall on the Indian west coast;International Journal of Climatology, in press.

Suresh I. and Shetye S. R. (2007) An analytical model for the propagation of tides in theMandovi and Zuari estuaries, Goa, west coast of India (unpublished manuscript).

Tamaki A. and Ingole B. S. (1993) Distribution of juvenile and adult ghost shrimps, Cal-lianassa japonica Ortmann (Thalassinidea), on an intertidal sandflat: Intraspecific facili-tation as a possible pattern-generating factor; Journal of Crustacean Biology 13 175–183.

Tomas C. R. (1997) Identifying Marine Phytoplankton (New York: Academic Press) 858 pp.

Townbridge J. H., Geyer W. R., Bowen M. M. and Williams A. J. III (1999) Near-bottom Tur-bulence Measurements in a Partially Mixed Estuary: Turbulent Energy Balance, Veloc-ity Structure, and Along-Channel Momentum Balance; Journal of Physical Oceanography29 3056–3072.

Unnikrishnan A. S., Shetye S. R. and Gouveia A. D. (1997) Tidal propagation in theMandovi–Zuari estuarine network, west coast of India: Impact of freshwater influx;Estuarine, Coastal and Shelf Science 45 737–744.

Wadia D. N. (1975) Geology of India (4th edn.). Tata McGraw-Hill Publishing CompanyLimited, New Delhi.

Wafar S. (1987) Ecology of mangroves along the estuaries of Goa; Ph. D. Thesis, KarnatakaUniversity, Dharward.

Wafar S., Untawale A. G. and Wafar M. V. M. (1997) Litter fall and energy flux in a mangroveecosystem; Estuarine, Coastal and Shelf Science 44 111–124.

Xavier, Filipe Neri (1852) Bosquejo Historico das Comunidades das Aldeas dos Conselhosdas Ilhas, Bardez e Salcete. Part I Govt. of Goa, pp 90.

Download - June 22, 2007 12:51 RPS rpb001 fm - National Institute of ... 22, 2007 12:51 RPS rpb001_fm vi CONTENTS Chapter 6 Phytoplankton diversity, biomass, and production S. G. Prabhu Matondkar,

Top Related