~

SPONSORS

Hydrobiologiata from But, Central aDd West Africa with 1lI.........1 .t ,••• r.­other African countries, Fishery Scientists in the United 8tatel, C>neda"U.K., I ...and the Soviet Union.

F:DITOR

Dr. l. Okedi. Director, E.A.F.F.R.O., Jinja, Uganda.

~

1, I

I

EDITORIAL BOARD

Mr. M. Abolarin, Co-Manager, Kainji LakeProject, Lagos, Nigeria.Mr. l. Kambona, Chief Fisheries Officer,Dar es Salaam, Tanzania.Mr. J. Mubanga, DirectOl, Fisheries Division,Chilanga, Zambia.Dr. L. Obeng, Director, Institute of AquaticBiology, Achimota, Ghana.Mr. N. Odero. Director of Fisheries, Nairobi,Kenya.Mr. S. N. Semakula, Permanent Secretary,Ministry of Animal Resources, Uganda.

Professor W.B. Banap, Departmeot of Z0o­logy, University of Zambia, Luub.Mr. R.E. Morris, Director, E.A.M.P.R.O..Zanzibar.Dr. T. Pelr, Umnotoplche blbrbDZlll.Wien, Austria. .PlOfessor Mohamed Hyder, University ofNairobi, Kenya.Professor, A. F. De Bont, Univenitb deKinshasa. Kinshasa Xl, Republique Demo­cratique du Zaire

COMMI1

PROGRAMME

The African J6urnal of Tropical Hydrobiology and Fisheries will only accept oriainal andwell supported ideas on techniques, methodology and researeh findings from scientists, fisheryofficers,. fishery economists and sociologists.The Journal will therefore strengthen the African lesearch scientist by making researchmaterial available and also increasing the awareness and utility of aquatic resources.Its quality will conform to International standards, and will be published in English andFrench.

MANUSCRIPT ADDRESSManuscripts should be addressed to E.A.F.F.R.O., East African Freshwater Fisheries, ResearchOrganisation. East African Community, Box 343, linja, Uganda.

REPRINTSAuthors will receive IiO reprints frec or charge. Extra reprints may be procured on cost.

PUBLISHEREast African Literatme Bureau. P.O. Box 30022. Nairobi. Kenya.

ISSUESThe Journal consists or one volume a year, consisting of two issues with approximatelyeighty pages each.

SUBSCRIPTIONAnnual subscription within Ea,t Africa Sh. 35. Outside East Afriea. East African Sh. 70,US $ 10.00

SymI

APPROACHES TO THE EVALUATION AND MANAGEMENT OF

THE FISH STOCK IN KAINJI LAKE, NIGERIA

E. O. ITA

Kainji Lake Research Project, UNDP/FAO New Bussa, Nigeria

Present address: Department of Zoology, UnitJerlity of TorontoToronto, Ontario. Canada

INTRODUCTIONSeveral methods have been used in recent

years to evaluate fish stocks in fresh waters.HOLDEN (1964) states that fish productionin a lake depends on a chain of events relat­ing (a) nutrient salts, (b) the phytoplanktonand zooplankton, and (c) the fish. Attemptshave been made both in Kainji and elsewhereto assess the productivity of lakes by asampling programme at these three levels,but the nearer the end of the chain thesampling is carried out the more difficult itis to ensure that it is adequate. Since thereare many stages between the nutrient saltsand the final catch of fish, it is difficult tointerpret the resull~ of chemical analysis interms of fish production.

As gill-nets catch the majority of fishtaken from the lake, an important element ofthe research programme has been to deter­mine, for the commercially significant species,the selective characteristics of various meshsizes of gill-nets used at Kainji. This informa­tion can be used to determine the fractionsof the various stocks that are subjected tofishing and to estimate improvements thatmight be obtained by use of alternative meshsizes. It should also be noted that all otherinferences made from the gill-net survey pro-

gramme may be subject to serious bias unlessappropriate corrections are made fOr theselectivity characteristics of the nets u·sed.

This paper is an attempt to sumlIJarizebrielly the work accomplished to date, bothwith respect to the size of the pot~ntial

fishery and to the components of the piesentstocks which are currently ·subjected to thedominant gill-net fishery.

KAINJI LAKEKainji Dam was completed. in AiJgust,

1968, forming a lake with: surface of 1,270km'; volume 13.97 km'; maximum \.ength136 m; maximum depth 60 m; and meandepth II m. The lake has an annual /luctua­tion in level of 10m and a ratio of ca~acity

to discharge of I: 4. Fig. I shows lh~ out­line of the lake. There is a northern narrowstrip about 40 km long in which conditionsmay be lacustrine or riverine dependi~g onthe water level. The surrounding aiea iscovered with bush and cultivated land.

The middle portion, which is about 70%of the total surface area, has some extensive/lood plains on the eastern shore and ashallow central rise, over the /looded FogeIsland. Here the lake reaches its widestpoint of 24 km.

36 E. O. ITA

RESULTS

METHODSThe data us'

selectivity havgraded fleet ofpress) for estim.fleet consisted(63.5 mm), 3 I

(101.6 mm), 5 (inch stretched33 m long and

It has been asize the nets (would exhibitresearch nets. Ithe fishermenmethod of hannets and amongof hanging to s<

shape of the ,mesh size, it is J

length is primalthe mesh openilfore, that if a :the effective 501for the commerexhibit the sarra rather broad

The breedin~

were determinlalong the Httoshocker (Smithusing ne. Cl

pulses per seco.7 ill per secaneregular intervalon Fig. 1.

The southenits steep shore,forms about 2

10"50'

4"501

I

I

\';01

Somptmg stations

br young t(~h

IGill net Sa~pfir.9

stath:ns

ENTRA,L BASIN

••x X

.. .., \ I 10~5':J \:

I

NEW BUSSA.rF'1' ..,4-J<C1lnjl Dom

~\ 0 5I

10 " 20 Km.

4°20'< I'" 4" 50'

N

'fID-SO'- .. f

, ,., ,

:{/

~:,"'" :

~ ./,'

" 1

,1"'- .)ij%I

---.'. I Jl .\I10'25' ~ J,.......,~

Fig"' Map of l'I.a.m/i LClk~ shOWing the major physical d I 'J i sions of the\(lk~ and sampling stations

Gill-net sampli.Detailed ana

net catches fn

FISH STOCK IN KAINH LAKE 37

9

th.

The southern portion is characterized byits steep shore-line and uncleared bush andforms about 20% of the surface area.

METHODSThe data used for the estimation of gear

selectivity have been extracted from thegraded fleet of gill-nets used by LELEK (inpress) for estimating the fish population. Eachfleet consisted of nets of 2 (50.8 mm), 2t(63.5 mm), 3 (76.2 mm), 3t (89.9 mm), 4(101.6 mm), 5 (127.0 mm) and 7 (177.8 mm)inch stretched mesh sizes each measuring33 m long and 3 m deep.

It has been assumed that for a given meshsize the nets of the commercial fishermenwould exhibit the ~ame selectivity as theresearch nets. However, nets constructed bythe fishermen vary considerably in themethod of hanging both from the researchnets and among fishermen. While the methodof hanging to some degree may influence theshape of the selectivity curve for a givenmesh size, it is likely that the modal selectionlength is primarily determined by the size ofthe mesh opening. It may be assumed, there­fore. that if a significant discrepancy exists,the effective selectivity of a given mesh sizefor the commercial fishery as a whole wouldexhibit the same mean retention length buta rather broader selection curve.

The breeding seastrns of the various fishwere determined by sampling young fishalong the littoral margin with an electricshocker (Smith-Root Electrofisher type VI)using D.C. current at 600 volts and 120pulses per second and with a pulse width of7 m per second. The 25 stations sampled atregular intervals with the shocker are shownon Fig. 1.

RESULTS

Gill-net samplingDetailed analysis of the results of the gill­

net catches from early 1960 to late 1970

has been given by TURNER (1970) andLELFK (in press). The sequence of eventsfor the past three years as indicated in pre­vious reports can be outlined as follo ...... s:

There was a boom in the fish populaHonin 1969. following impoundment, particularlyamong mernbers of the family Citharini:iae.The mean catch-per-unit that year was 31.6kg (wt.} per fleet per night and 95.0 (no.)per lIeet per night. The figure dropped toabout 11.2 kg (wt.) and 41.0 (no.) in 1970and 11.5 kg (wt.) 53.5 (no.) in 1971 (Tahle I).

Table 1. Catch-per-unit effort by number and weightbelween 1969 and 1971

I 1969 I t970 1971I

I----

No. of samples 5 4 4Mean c/e by no. 95.0 I 4t.O 54.0Mean cJe by wt. 36.6 kg It.2 kg I t t. 5kg

The family Citharinidae represented mostlyby Cirharirzus citltarus (Geotltoy St. Hilaire)and Citharirzus distichodoides Pellegrin,which formerly sustained the local fisheries,gradually dropped from 37% in 1969 to13.4% in 1970 and 11.87c in 1971.. Evidencefrom counts of local fishermen's catchesalso indicated the decreasing number ofCitharinus.

The family Characidae, represented mostlyby members of the genera /flestes and Hydro­cynus, became increasingly numerous risingfrom [6% in 1969 to 37.590 in 1970 and55.1 % in 1971. Obscrvation of local fisher­men's catches showed a shift from exclu­sively Citluuinus fisheries using 6-8 in meshnets to a mixed fleet with 2 and 3 in meshescatching large numbers of A/estes dentexsethente (Cuvier and Valenciennes), and A.baremoze (Joannis), together with youngerage groups of Hydrocyrzus forskali Cuvier,and H. lineatus Bleeker.

Sampling oj juvenilesShoreline sampling of the young of the

38 E. O. ITA

year also showed the ahundance of membersof the family Characidae in all Ihe micro·habitats (Table 2).

Table 2. The relative proportion in percent by numbeIof fish families caught by gill-Dds between1969 and 1971

Year 1969 1970 I 1971FamHy

11______

Cithatinidae I 37.0 I 13.4 11.8Characidae I 16.0 I 37.5 55.1Schilbeidae i 9.0

i9.8 9.5

Cyprinidae 9.0 I 0.4 4.6Mochokidae

I10,0 I 11.6 93

Others 1 19.0 I 143 I 9.7

Among the Characidae, Alestes bare­rnase, A. leuciscJ1s Gunther, and A.nurse Ruppell predominated in every month'ssample. B4rbus spp. and Labeo parvusBoulenger dominated the Cyprinid catch.The highest number of the young of Laheopseudocoubie Balche and Milton and L.eoubie Ruppel were caught during Septem­ber/October. Of the Bagridae. Chrysichthysauratus longifilis (Pfaff) were the mostnumerous and Tilapia galilaea Artedi. T.nilotica (Linnaeus), and T. :.illii Grcvaisdominated among the Cichlids. Few juvenilesof Citharinus occurred except in November(Tahle 3J when Over 100 individuals werecaught in stations north of the River Swashiup to Yelwa (Fig. I) and along the easternbank as far as the dam. The length fre­quencies of the fish caugh t with the shockerare summarized in Table 4.

Members of the genus Alesles apparendyhad more than one breeding season. Fishwithin the Jength group J-3 em were numer­ous in June, September and November witha peak in November. Citharinus appearedto breed in August since the few fry caughtin September measured less than 2 em, andthose caught in November measured morethan 2 em. Crysichthys and Tilapia bredthroughout the year, with a peak from Sep-

tember showing the breeding possibly tookplace in August or late July.

Gill-net selectionThe selectivity constants and the mean

selection lengths for some of the speciescaught with different mesh sizes of gill-nets(Table 5) were estimated using HOLT'S(1957) formula Lm-K theta where Lm isthe mean selection length, theta is the meshsize and K is constant.

The length distribution and log ratios ofcatches of each successive pair of nets forthe fish in Table 5.re shown in Appendices1-6. The selectivity constants used for theestimation of the mean selection lengthswere determined from the regression equa­lion of the log ratios, The logarithms of theratios shown in the appendices for the variousfish were plotted in Figs. 2-4. The regressionlines were fitted by the method of leastsquares. GULLAND and HARDING (1961)mention that if the selection curve is normaland of constant variance, then the plot shouldbe a straight line. The plots obtained hereare fairly straight for smaller fish and forSynodontis spp., but flattens out and becomesnearly horizontal for larger specimens ofCitharinus spp.

The normal selection curves which describeselection by meshing for the different meshsizes have been drawn (Fig. 5). For Citha­rinus, the curves had narrow peaks for the2 in. mesh. Two peaks were distinc t for the3 and 4 in. meshes. The first peaks werenarrower [han the second which were farbelow Ihe 50% retention lengths. Only a,ingle peak was observed for the 5 in. mesh.The double curves for 3 and 4 in. mesheswere fairly consislent with the flattening outof the points for larger fish on Fig. 2 for thesame meshes. GARROD (1961) explains thatoYer a greater part of the selection rangethe fish catches of a given mesh size arenormally distributed but at the limits of theselection ranges the numbers of fish caught

Table 3. The relatinumber of:shocker bet'o,1

1971 I Man

Family

Characidae 34.'Clupeidae 32.1Cyprinidae 16.:Cichlil.lae 9.1Bagridae 5.:Citharinidae OJOthers 2.:

Table 4. Numbers 0March and

1971Length group in em

Alestes baremozeA. nurseA. dentexA. leuciscusCitharinus citharusC. distichodidesDisticnodus sD·Chrysichthys auratusTilapia galifaeaLates nilotirus

(Linnaeus)Hydrocynus sp.Labeo senegalensis

Table j. The mean sfish is end(

Mesh size minch.Species

CitharimJs citharus

Synodontismembranocenous

S)'nodontis sD·

Lates

Hydrocynus

A/estes

Labeo

FISH STOCK IN ""'NJI LAKE 39

breeding possibly tooklate JUly.

Table 3. The relative proportion in percent bynumber of young fish caught with electricshocker between March and November 1971

Table 6. Relative proportion of fish in percent byweight in experimental catches arrangedaccording to trophic levelsperiod of Primary Secondary Tertiary

Table 5. The mean selection lengths in em of fish caught in different mesh sizes of gill-nets. The number offish is enclosed in brackets

1971 I March JuneIse~~~m-I

Novem- sampling I Con- Con- Con-

tx" 'sumers sumers sumersFamily r

Characi~1 June 1969·34.7 2n.5

I40.6 31.3 January 1970a 44.7 t~.2 1~.3

Qupeidae : 32.0 26.9 237 t4.7 April197Q-Cyprinidae 16.2 18.3 ! t4.8 8.1 October 1970aa 21.7 44.4 33.9Cichlidae 9.0 21.5 10.8 30.2 Pebruary 1971-Bagridae 5.3 4.3 7.6 10.8 May 1971aa 18.6 42.2 39.1Citharinidae 0.6 0.1 1.1 3.3Others 2.2 8.4 1.4 1.6 a Based on TURNER (1970)

aa Based on LELEK (in press)

Table 4. Numben of fish within specific length groups caught Wilh electric shocker along the shore betweenMarch and November 1971

21 em jlnd over

10 em and over(s. res~inatu)')

23 em and over

Approxirpate standardlength at maturity

Over 30 em

21-23 em and over

I 18 em and over

I(A. baJY'!!oze)

31 cm and overI(L. senegalensis)

33.30(10)

28.86(35)

1016 rnm 1127.0 mm 1177.8 mm I(4") (5") (7")

1556 I 28.15 I 31.58 I(4t4) (631) I (12)

24 67 25.67 Ii

(]67) (86)

23.73(88) i

29.30 29.85(50) (14)

23.50(32)

16.67(70)

22.67(204)34.86(121)

76.2 mm.(3")

19.71(191)

24.84I (200)

March June I Sept./October I November4--6: 7-9 1-3 I 4--6 I 7-9 t-3 I 4--61 7-9 1-3 ~'7-9

1---1-1--1--:---

13.1(21)

18.16(404)24.24(358)20.12(894)16.16

(90)

50.8 mm(2")

12.32(52)

1-3

A/estes

HydroC}'rnts

Synodontis sp.

Lates

Mesh size in inch.Species

Citltarinus citharus

Synodontismembranacenolfs

LabeQ

1971 1Length group in cm

A/estes baremoze 6 92 35 6 110 87 52 159

I

277 156 219 50A. nurse 17 I 13

1

24 92

I19 36 94 48 3 36 70 16

A. denfex

Ia 0 4 2 0 I 4 I 25 47 7 85 23

A. feuciscuj 23 185 10 147

I

192 2 37 866 : 118 77 264 28Citharinus cilharus 0 I 2

i0 0 0 0 4

I0 0 8 54 20

C. distiLhodides 0

I

0 0 0 0 0 0 0 1 0 20 IDistichodus sp. 0 2 0 0 1 1 0 t 9 0 8 0Chrysichthys aUrafu! 2 60 3 I 31 39

\17 173

,189 59 147 232 40

Ti/apia gal;laeu 26 I 19 ~ 49 117 50

I222 I 86 35 636 257 39

Lates ni/oticus !(Linnaeus) 0

!0 0 0 0 0

I9 4 1 0 0 1

Hydrocynus sp, I 0 , 0 I 0 1 I 0 0 t I 0 0 0 0Labeo senega/ensis

! •I , 2

I3 0 ! 0 0

I0 1 0 3 45 7,

I ,1 curves which describefor the different meshI (Fig. 5). For Cifha­narrow peaks for the: were distinct for theThe first peaks were~ond which were farlion lengths. Only a,d for the 5 in. mesh.3 and 4 in. meshes

'ith the flattening au t

fisb on Fig. 2 for the) (1961) explains that

the selection range~iven mesh size are: at the limits of the"hers of fish caught

)nstants and the meanr some of the splVCiest mesh sizes of gill-netsimated using HOLT'S-K tbeta wbere Lm is,ngth, tbeta is the meshnt.

ution and log ratios of~essive pair of nets for-e shown in Appendicesconstants used for thenean selection lengthsrn the regression equa­. The logarithms of thependices for the various'igs. 2-4. The regression

the method of leastand HARDING (1961)lection curve is norma]ce, then the plot shouldle plots obtained here. smaller fish and for.ttens out and becomes

larger specimens of

"OQ.

±+<t N

,~0

,,N ~W

• ..

,'''0.-·h~)

"' ;;,•

'"0

":;: ii1

zn

'"cD

:0

0_co

-t

o"

8

7_N -.•0

a0 <

'"~

,"

(-,- I:l.; •N

-rN

'"~•'"

'"~;::

-;:

0-~

/<>p I1-

;;~

•0N

'":;;

.",

•~.

~ai ,

N

:)-~~

20

1-3•" ~o3N

=

~ , •0

~

N

"

•;~

• 0~

•, 00

N0

"• ••~Ql 1+

"w

'"0

;

•~

'"0c

w

"

H

•N•"w"

•

w

:;;'"

""

'+

00 0"..~ '"• • "

"'10"'I<''1'"

VU'0"()p

FISH STOCK IN KAINJ! LAKE 41

em"I"

H

42 I!. O. ITA

I'00

"./1 ~ SO

+5-0-1

uz10w

0~

60 j 1O.!.l__ J~'"

+4-5-1

:t~ 50 --

8·

z

:: loge A

w

'"

+4·0-1 X

u~w '·ill

C

L

+3. s-I (.) = 10ge B

.,.I.Ie" 10 11. I" I

+3.0-1

./ /.~+2·5-1

+2.0

I I.loTU

X

I

+1.5...,

./ / •X X

+1.~~)

00

X

/' / ~ ..u'"j zt'J 1,

X

7 ~ 60/7 w

5O~4~"t- --

•

~ (20."

'"0 ~:.5

I u '"

~

~

~ ;./-1-0-1X

- '·5

I1'2 14 ",," IS

-2.0

/ j=-'j ,

~'?O00,

~ ..u~ 7.0$ 60

I 50r -- "),-:4I

~I I I

w ..I I , I30 31 32

u

I I

28 29

~

-5·0 1 i I26 21

w ,.I

22 23 24 25

L .20 21

I ,.f

,.LENGTH IN Cm.

for ..l..A1.a njlgtioys ,and lengths

'11 H T, II

' 01 log ratiosFig. 4 R~re5sion line!;different mesh sizes

t

caught by nets of

_SYNOOONTIS SSP

8 10 ,'12 14 16 ',e 20 12'14 26 18 }O 11

FISH LENGTH IN em.

,....,.20

1'!Ot--*----;,,----jo,~,'T---­

90..,...

10.3

8 10 11 14 16 \8 10 1.1 1'1 26 28 30 11 j4 36 le:'

FISH LENGTH IN em,

I "'TH SP.

32

:.II n jlatiou'

" .. 16 " 2. 22 24 2. 28 '3D 31,.

FIS"H LENGTH IN em.

•

ALESTES SPP.

~~90

~ ..u~ 7.0it ..~

50~ -- "'i9A~ ..\;1~ I·•

20I.." .. 16 18 10 12 H 16 18 :)0 )1'3"

I "ISH LENGTH IN C",

Fig.5 Relative selection ~urve$ (or different mesh sizes and species

44 E. O. ITA

are greater than would be predicted fromnormal distribution. This is shown by thesigmoid form of the regression of the loga·rithm of the catch ratios of the mesh com­parisons and is prohably caused by theentangling powers of the nc:ls.

Garrod's outline of the general theory ofgill-net selectivity states that:(a) If the growth of a fish is isometric, then

the selection for length by a given meshsize may be expected to be Ilurmallydistributed.

(b) When two gill-net units. A and B, arcfished simultaneously then the logarithmof the ratios of the catches as successivelength groups in the two twits will havea linear' rela tionship provided:(i) that the units differ very slightly

in mesh size (ie.. have widely over·lapping selective ranges).

(ii) that the standard deviation of theselection curve is the same for bothu/lits.

(c) The two selection curves will intersectat the: length where the ratio is unity andif a series of glll-net units are operatedsimultaneously the graph of this lengthagainst the average mesh size will belinear. passing through the origin.

The nearest approach to linearlty amongthe graphs was for Synodonri.\· spp. (Fig. 3)whose serrated opercular spines possiblyrestrict penetration of the fish in the netbeyond the operculum. For Lares niloticus(Fig. 4) the selection curves ifllersected abovethe zero line OIl the y axis. The relativeselection curves for these specks (Fig. 5).however, showed only single curves.

Girth was not recorded as a matter ofroutine, hence the length-girth rela tiunshipcould not he accurately estimated. Work onthis has beguu and the results will be dis­cussed in a later publication.

DISCUSSIONStudies on Kainji Lake as in other man-

made lakes have been directed lOward elu­cidating the changes in the fish population,investigating aspects of the biology of in­dividual species and estimating the potentialcatch of fish. These three considerationstaken together form the basis for the estab·lishment of a national development andmanagement policy.

Changes in fish populationMost man-made lakes undergo a sudden

expansion in fish population upon initialimpoundment. owing to high cuncentrationof nutrient elements from flooded bush. Thepopulation then declines fairly rapidly to anestablished state. ~ainji Lake took only threemonths to fill after impoundmenl and under­went the usual sudden upsurge of produc­tivity. Of the 25 major ,pecies of fislt inthe lake Citharinu.Y spp. dominated the catchof both the experimental and local fisher­men's catches. Depending mostly on plank­tun and detrital (oad, it exploited the largeamoum of nutrients and organte detrituswhich became available after nooding. TableI shows the decrease in number of Cirha­rinus in the catch. Juveniles of this fish werescanty in 1970 as few mature individualse~capcd the local fishermen's net. There wasa gradual increase in the number of juvenilescaughl in November, 1971 (Table 4) parti­cularly in the northern arm or the lake, but itis doubtful whether recruitment will increaseto any large extent because of substantialcatches of juveniles by local fishermen usingsmall mesh cast nets.

Citharinlls apparently requires swamphabitats, with plenty of organic matter, forspawning and feeding of lhe fry. Such areaswith extensive gentle slope and bottom richin detritus are mainly confined to limitedareas of the lake particularly alung the floodplains of the northern arm (Fig. 1) and theeastern flood plains of the middle region.These constitute approximately 20% of thelake area. The same. areas provide slJitable

breeding and nuWith the doclJ

from early 1970the relative pro)belonging to the6).

There was a I

of primarY cons!veys of local fishlthe increasing s1!lalce particularlymentioned. In gCi

lng programme Iof tertiary conSl

consumers and alsumers with seeonant in deep wattin shallow waters

Estimates of pot~

HENDERSONmate of potentialLake hased on oc.with that of the secompared the nut]for which TALLestima.tes or primeluded that Kainrange. From thisestimated assuminphic transfer efficboth mlcrophago

'Based OIl the repolakes and the profishes in the catcpropriate to aSSUI

approximately '"vorous, mkropha(average [ood·ch~

For an estimatedof 3.6 g carbOlsiderations suggemen of 35 to 1:

Estimates ba~

morphoedaphicdivided by mean

irected toward du­he fish population,the biology of in­lating the potentialree considerations)asis for the estab­development and

onundergo a suddenItion upon initial1igh concentrationflooded bush. The,irly rapidly to anke took only threedment and under­)Surge of produc­)pedes of fish inminated the catchand local fisher­mostly on plank­'plaited the large

organic detritus"flooding. Tablelumber of Citha­; of this fish wereature individuals's net. There wasmber of juveniles(Table 4) parti­

)f the lake, but itlent will increasee of substantialfishermen using

equires swampanic matter, forfry. Such areas

lild bottom richined to lim itedalong the floodFig. I) and themiddle region.,Iy 20% of the"ovide suitahle

breeding and nursery grounds for Tilapia.With the decline of the Citharinus fishery

from early 1970 there has been a change inthe relative proportions of the fish speciesbelonging to the various trophic levels (Table6).

There was a drop in the relative weightof primary consumers although recent sur·veys of local fishermen catches have revealedthe increasing significance of Tilapia in thelake particularly along the areas earliermentioned. In general the experimental fish­ing programme probahly took abollt 35%of tertiary consumers, 40% of secondaryconsumers and about 20% of primary con­sumers with secondary consumers predomi­nant in deep waters and primary consumersin shallow waters.

Estimates of potential catchHENDERSON (in press) proposed an esti­

mate of potential catch of fish from KainjiLake based on ecological considerations. Hewith that of the series of West African lakescompared the nutrient status of Kainji Lakefor which TALLING (1965) had obtainedestimates of primary production and con­cluded that Kainji should fall within therange. From this a range of fish yield wasestimated assuming reasonable ranges of tro­phic transfer efficiencies for herbivores. andboth microphagous and predaceous fishes.Based on the reported yields of other Africanlakes and the proportion of various types offishes in the catch of Kainji, it seemed ap·propriate to assume that yield is taken fromapproximately equal proportions of herbi­vorous, microphagous, and predaceous fishes(average food-chain length to the fishes of 2).For an estimated rate of primary productionof 3.6 g carbon/m'/day, the above con­siderations suggest a potential yield to fisher­men of 35 to 150 kg/ha fresh weight.

Estimates based on RYDER'S (1965)morphoedaphic index (total dissolved solidsdivided by mean depth) suggest an expected

FiSH STOCK IN ICA'NJI LAKE 45

catch of 35 kg/ha fresh weight or a totalestimate for the whole lake of 4,100 t perannum.

A preimpoundment estimate 5,400 t perannum with a possible increase to aboutdouble that figure was made by DAGETand (BAYAGBONA (1967). LELEK (inpress) arrived at a figure of 4,667 t perannum. HENDERSON modified the esti­mate using updated canoe counts to acurrent catch of 7,200 t per annum (in 1970).

The general conclusion in all cases is thatwith adequate resource management a yieldof 10,000 t per annum could be achieved.

Dr\.'elopment and management policyEVANS (1969) outlines some characteris­

tics predisposing a lake to either overfishingor underfishing. Among the characteristicsleading to overfishing are:(a) Low basic productivity of the lake.(h) Low catch·per·unit of effort.(c) Fishery primarily dependent on one

species.(d) Instability of the stock and question­

able recruitment.

Characteristics leading to underfishjng are:(a) Extensive littoral zone.(b) Light fishing intensity.(c) Use of selective gear and fishery based

upon individual etIort with low efficiencygear.

(d) Seasonal fishing pattern; fishing hasedupon a small segment of the stock andspawning of fish at least once beforeentering the fishery.

Considering these factors Kainji Lakecould be said to have a low basic produc­tivity. The fishery was initially dependent onCitharinus caught with large mesh gill-nets,hut with the decline in the population ofCitharinus, fishennen have resorted to mixedfleets with smaller meshes for Alestes spp.and juveniles of Hydrocynlls. LELEK (inpress) states that doubling the effort would

46 E. O. ITA

~

not necessarily double the catch and at pre­sent the catch-per-unit effort in the lake islow, approximately II kg per fleet of 7 netseach measuring 33 x 3 m, with 2 (50.8 mm),2! (63.5 nlm), 3 (76.2 mm), 3! (89.9 mm),4 (101.6 mrn), 5 (127.0 mm) and 7 (177.8mm) inch stretched mesh sizes.

Because of its more regular shape Kainjihas a more limited area of littoral zone uti­lized for fishing by comparison with VoltaLake. The· area of Kainji Lake is approxi­mately t that of Volta Lake but the numberof fishermen in Kainji Lake is four timesthat of Volta Lake (BAZIGOS 1971). Thefishery is, however, based on individualeffort. Data on mortality are at present notavailable because of lack of information onthe growth rate of fish and inability to deter­mine age of individual fish. It seems, how·ever, that the future recruitment of Citharinusand Tilapja is promising because of theabundance of the young of these speci",in recent months (Table 4).

Co~sidering stability on the basis of thepoints outlined by COOPER (1966) it isdoubtful whether Kainji Lake could beclassified as a stable reservoir, a factorwhich should augment production. TURNER(1970) is of the opinion that the fisheriespotential of Kainji Lake appears to be limiteddue to the low nutrient content of inflowingwaters, the declining available nutrients fromthe lake basin aftcr the first filling andprobably ·the high rate of volume displace­ment. He believes, however, that the presentcharacter of the dominant food web towardprimary consumer fish species probablyrealizes the greatest possible fisheries poten­tial of the lake.

It is reasonable to assume that there isno immediate danger of overexploitation inKainji Lake. EVANS (1969) states thatpeople frequently fear the danger of over­exploitation, but this is seldom a seriousthreat in tbose fisheries where the catch isharvested on an individual fisherman basis

with relatively low efficiency gear. The timehas come, however, when some sort ofcontrol may need to be considered for KainjiLake.

Altbougb information on growth of fishin the lake is lacking, the present classifica­tion of the approximate length of the variousfish at first maturity, together with theirrespective selection ranges can provide suit­able guidelines to management. The currentuse of 7 inch mesh and over by local fisher­men should be encouraged for open waterfishing of Citharinus, Lates and Laboe spe­des. For bottom sets, 3 inch mesh and overwould be suiTable for catching Synodontisspecies. For surface sets. in deep waters,3 inch mesh could be used for catchingAlestes although there is a danger of elimi­nating the young of HydrocYllus which in­cidentally move in shoals with Alestes dentexand A. baremozc. Six inch mesh and overwould be suitable for shore-line fishing forTilapia without necessarily eliminating theyoung of other species.

LAGLER (969) points out that if thedevelopment of the fishery of a new reservoiris allowed to be haphazard, the net yield islikely to be far below the maximum sustain­able, and overexploitation of at least somekey stocks is bound to occur. It is reason­able. therefore, that in a multiple speciesfishery where optimal harvesting requiresthat a variety of gear (with differing selec­tivity characteristics) be used, managementmust rely heavily on advisory programmesfor fishermen with less emphasis on regula·tion. The latter must necessarHy be com­plicated and difficult to enforce.

SUMMARYEstimates of potential yield for Kainji

Lake, and the methods of analysis by earlierworkers are discussed. Also summarized isthe state of the fishery after impoundment,between 1969 and 1971, based on experi­mental gill·net catches.

Recent Sal

along the litlof the comnspawned suafishing mort:the use of Sl

men, presen~

establishmentThe result

based on HeThe data ha'menta) gill-",nets betweentions based 0

made to ensof the fish sp'in catch-peN

RESUMECe docume

potentiel de nmethodes d',sp&:ialistes anII resume aus

REFERENCES

Ba7:igos, G. P.Lake. FAGSF/NR,24

Cooper, G. P.poundment~

104: t-21.Daget, ]., and

Dams Pro:NEDEOOLondon.

Evans, W. A. lfished or uLake Rese;:

Garrod, D. J. (of nylonGraham. J.203.

Gutland, J. A..tion of CIllJgill-nets. J.215·22.

Henderson, F.catch of fgeneral ec<l

FISH STOCK IN KA1NJI LAKE 47

~fficiency gear. The timer, when some sort ofbe considered for Kainji

Ltion on growth of fish19, the present classifica­Late length of the vadousty, together with theirranges can provide suit­.anagement. The currentmd over by local fishcr­)uraged for open water" Lates and Laboe spe­s, 3 inch mesh and over'or catching Synodontis: sets, in deep waters,

be used for catchingre is a danger of elimi­Hydrocynus which in­

oals with A/estes dentexIn inch mesh and over)f shore-line fishing forossarily eliminating the:ies.points out that if the)hery of a new reservoir~azard, the net yield is. the maximum sustain·ation of at least someto occur. It is reason·in a multiple species

11 harvesting requiresr (with differing selec­be used, managementadvisory programmess emphasis on regula-

necessarily be com­to enforce.

tial yi~ld for Kainjiof analysis by earlierAlso summarized is

I after impoundment,71, based on experi-

Recent sampling of the young of the yearalong the littoral margin indicates that mostof the commercially important species havespawned successful1y in the lake. An intensefishing mortality of juvenile fish, owing tothe use of small mesh nets by local fisher­men, presents a possible threat to the futureestablishment of the fish in the lake.

The results of gill-net selection studiesbased on HOLT'S (1957) method are given.The data have been extracted from experimental gill-net catches with graded fleets ofnets between 1969 and 1971. Recommenda­tions based on the above studies have beenmade to ensure a succ~sful establishmentof the fish species in the lake and an increasein catch-per~unit effort in subsequent years.

RESUMECe document a trait aux evaluations du

potentiel de rendement du lac Kainji et auxmethodes d'analysc utilis6es par certainss¢Cialistes anterieurement a ces evaluations.II resume aussi 1a situation des peches iJ. la

REFERENCES

Dazigos, G. P. (1971). Frame Survey at KainjiLake. FAO Fish. lech. Pap., WS/Bl154 Ft:SFINK, 24, P. M.

Cooper, G. P. (1966). Fish production inim­poundmems. Mich. Dept. Cons. Res. Repl.,t04: 1-21.

Daget, J., and N. E. Mayag'bona (1961). NigerDams Project, Vol. 6, Part 8, Fisheries.KEDEOO and Bal'four Beaty, Hague andLondoll.

Evans, W. A. (1969). Is Volta Lake being over­fished or underfished? Progress report, VoltaLak.e Research Projecl.

Garrod, D. 1. (1961). The selection characteristicsof nylon gill-nets for Tilapia esculelllaGraham. J. Cons, Int. Explor. Mer., 26: 191­203.

GulIand, I. A., and D. Harding (196·1). The selec­tion of Clar;as mossamblcus (peters) by nylongill-nets. J, Cons. btl. E.rplor. Mer .• 26 (2):215-22.

Henderson, F. (in press). Estimation of potentialJ;atch of fish from Kainji Lake based ongeneral ecology of the lake. FAD Fish lech.

suite de la mise en eau, entre 1969 et 1971,baste sur les prises experimentales aux filetsmaillants.

Un echantillonnage des jeunes poissons deJ'anne", effectue recenilllent Ie long du lit·toral, indique que la plupart des especesd'importance commerciale ont fraye avecsucces dans le lac. Un taux eleve de morta­lire de peche des jeunes poissons, dO iJ. l'em­ploi de petits filets par les pecheurs indigenes,est une menace a I'etablissment du poissondans Ie lac.

Le document donne 105 resultats des etudessur la selectivire des filets maillants baseessur la methode de HOLT (1957). Les don­nees sont extraites des prises experimentaleseffectuees au moyen de gammes de filets dediverses dimensions entre 1969 et 1971. Ledocument presente des rlXommendations,basees sur les etudes ci-dessus, en vue<i'assurer J'etablissement des especes de pois­son dans Ie lac et l'augmentation des prisespar unite d'effort dans les annees. successives.

Pap. Fonnerly an appendix to June, 1971,progress report, Kainji Lake Research Projef;t.

Holden, M. J. (1%4). Estimating the fisheriespotential of African takes. Ne'W" Scientist,No. 411, pp. 20-23-

Holt, S. 1. (1957). A method for determining gearselection and its application. Joint Sci. Meet.ICNAF. IrES and FAO, Pap. No. 51.5,Lisbon.

Lagler, K. F. (1%9). Man-made Lakes. Planningand Development. F AO, Rome.

LeIck. A. (in -press). FiS'h population of the KainjiLake, trend of their development and utinza­tion. Tenninal Report to FAO, Rome.

Ryder, R. A. (1965). A method for estimating thepotential fish production of north-temperalelakes. Trans. Am. Fish Soc., 94 (3); 214-18.

Tailing, 1. F. (1 (5). The photosynthetic activjtyof the phytoplankton of the East Africanlakes. Int. Rpvue ges. Hydrobiol., 50 (1):1-32.

Turner. J. L. ([970). The fish populati<ln of newlyimpounded Kainji Lake. FAD Tech. ReportI FI; SFiNlR 24.

48 E. O. ITA

Appendix 1. Length distribution and log ratios for Citharinus cilharus and C. distichodoides caught in fleets ofnylon gill-nets with 2, 3.4 aad 5 in stretched meshes between 1969 and 19710t =number of fish of length (L)

adj = number of fisl1 adjusted according to effort

AB I e I D I IStandard . ~ Loge B/A Loge CjB Loge D/e

Length em r (5 nels) 3" (14nels) I 4" (29 nels) 5" (53 nets)

n, adj nj adj n, adj n,10 5 53.0II 23 243.812 33 349.813 33 349.814 .19 201.4 2 7.57 -3.281515 1 10.6 10 37.86 1.273316 4 42.4 27 102.21 0.871017 I 10.6 45 170.36 4 7.31 2.7777 -3.149418 75 283.93 34 62.14 -1.519519 80 302.86 190 347.24 0.136620 32 121.14 230 420.34 • 1.244521 25 94.64 206 376.48 1.3811 -5.238922 27 102.21 86 157.17 2 0.4309 -4.365323 19 71.93 36 65.79 24 --0.0891 -1.008724 80 30.29 33 60.31 75 0.6888 0.218125 14 53.00 72 131.l9 182 0.9095 0.324526 29 109.79 82 149.86 209 0.3125 0.332327 22 83.29 79 144.38 219 0.5504 0.416428 8 30.29 71 129.76 172 1.4555 0.281429 5 18.93 36 65.79 90 1.2459 0.313230 2 7.57 18 32.90 57 1.4695 0.549731 3 11.36 5 9.14 26 --0.2174 0.045832 5 9.14 17 0.620733 4 7.31 2 1.296434 I35 136 2

I I

Appendix 2. Lengtl2.3.4

StandardLength em 2'

n10 2'11 7'12 913 414 ~

15 216171819202122232425262728293031

FISH STOCK IN KAINJI LAKE 49

r:hodoidey caught in fleets ofod 1971

Appendix 2. Length distribution and log ratios for Synodontis spp. caught in fleets of nylon gill-nets with2, 3. 4 and .s in stretched meshes between 1969 and 1971.

nl =number of fish of length (L)adj =number of fish adjusled according to effort

l/A ILoge CjB Loge D/C

-3.1494-1.51950.13661.24451.38110.4309

-0.08910.68880.90950.31250.55041.45551.24591.4695

-0.2174

-5.2389-4.3653-1.0087

0.21810.32450.33230.41640.28140.31320.54970.04580.62071.2964

A I B C DLoge B/A LOBe CjB Loge D/CStandard

Length em 2" (11 oets) 3" (13 nets) 4" (28 nelS) 5'" (7 nets)

nt adj nl adj n, nt adj10 27 68.7311 79 201.0912 95 241.8213 49 124.7314 26 66.18 5 10.77 -1.815915 23 58.55 9 19.38 -1.105916 4 10.18 22 47.38 1.537917 4 10.18 20 43.08 1.443118 2 5.09 17 36.62 1.973719 2 5.09 18 38.77 2.030820 1 2.55 13 28.00 14 2.3967 -0.069321 11 23.69 63 0.978122 8 17.23 124 1.973923 9 19.38 139 1,970524 3 6.46 246 6 24.0 1.3374 -2.327625 1 2.15 192 16 64.0 4'.4929 -1.098826 I 2.15 96 25 100.0 3.7997 0.040827 2 4.30 36 11 44.0 2.1252 0.200828 28 16 64.0 0.826829 28 2 8.0 1.253130 7 8 32.0 1.511031 4 16.0

•

50 E. O. ITA

Appendix 3. Length distribution and log ratios for Lates nilo/icus oaught in fleets of nylon gill~nets with2,3 and 4 in stretched meshes between 1969 and 1971.

01 =number of fish of length (L)actj =number of fish adjusted according to effort

I

A B

Ic

IStandard Loge BfA LogeC/B

Length em 2' (56 Dets) )" (60 nets) 4" (10 Dets)

Dl adj D1 D1 adj13 6 6.4314 18 19.2915 126 135.0016 126 135.0017 138 147.8618 150 160.7119 102 109.2920 78 83.57 24 -1.248021 60 64.29 72 0.J.l3)22 54 57.86 111 0.651523 36 38.57 108 0.029724 30 32.14 99 1.125325 27 28.93 84 1.066126 9 9.64 42 1.471927 6 6.43 27 6 36.0 1.4352 0.287728 3 3.21 12 9 54.0 1.3\89 1.504329 1 1.07 6 12 72.0 1.7254 2,485230 9 9 54.0 1.792231 3 9 54.0 0.588032 2 12.0

Appendix 4. Length dislskali andnylon gill­meshes beDI=numtadj=numing to eJJ(

StandardLength em 2

D

20 1421 2222 8423 9124 20725 8726 7727 8428 1429 2130 1431 732 233 3343536373839404142

FISH STOCIC IN ICAINJI LAKE 51

Appendix 5. Length distribution of Ales/es spp.caught in fleets of nylon gill~ne!s with2 and 3 in stretched meshes between1969 and 1971

nl=number of fish of length (L)adj=nurnber of fish adjusted according to effort

Standard ------Length em 2" (23 nets) 3' (35 nets)

n' adj n'14 6 9.1315 18 27.3916 15 22.8317 15 22.8318 45 68.4819 181 275.4320 252 383.4821 189 287.6122 Jl4 17J.4823 n 50.22 324 21 31.96 325 3 4.57 326 I 1.52 1427 I J.5~ 1428 6329 4930 3531 2132 733 1

leets of nylon gill-nets with

1

18033i15:97:53611952 0.287789 1.5(}4354 2.4852

1.79220.5880

Appendix 4. Length distribution of H)'droc)'nus fOT~

skali and H. linea/us caught in fleets ofnylon gill-nets with 2 and 3 in stretchedmeshes between 1969 and 1971n 1 :=c-numbcr of fish of length (L)adj =number of fish adjusted accord­ing to effort

A BStandard

Length cm 2' (49 nels) 3' (42 nets)

nl nl adj20 1421 2222 8423 9124 20725 8726 7727 84 11 12.8328 14 14 16.3329 21 J1 12.8JJO 14 28 J2.67JI 7 29 33.8J32 2 JO 35.0033 3 J6 42.0034 80 93.3335 55 64.1736 36 42.0037 28 32.6738 II 12.8339 14 16.3340 11 12.8341 3 J.5042 3 J.50

•

A B

52 E. O. ITA

A I B

DEVEI

BURUI

parM. NGOMDirecteurDepartmen

GENERAlLe lac T

d'une pecbbasee surartjficielle ~

"Ndagala"etait praticcalme. Grnlpar les peelIe Ndagalaepuisettes ~

coutumieretonne par I

Des 194<hydrobiololloppa un ,peche dansil nommapeehe. Dessieur Collafaibles reneles premielfurent leol<quee de no

et

L. HALIN<Proiet Pech

•

nJ adj r TIl

2 7.872 7.87

16 62.9368 267.4766 259.60 1245 177.00 1213 51.13 128 31.47 166 23.60 20R 31.47 215 19.67 581 3.93 882 7.87 154

7840131414131010

121314151617181920212223242526272829303132

StandardLength em

Appendix 6. Length distribution and Jog ratiosLtJbeo pseudocoubie and L. serrega/ensiscaught in fleets of nylon gill·nets with2 and 3 in stretched meshes between1969 and 1971

n'~number of fish of length (1.)adj =number of fish adjusted according to effort

-----~_I 2' (15 nets) I 3' (59 nets)

0:

..,i

LE PEeRVne unit<