Odonalologica38(3):261-265 September1,2009 Differences in Odonataabundanceand diversity inpesticide-fished, traditionally-fishedand protectedareas inLakeVictoria, Eastern Africa (Anisoptera) D.J.Martins¹’² 1DepartmentofOrganismicandEvolutionaryBiology,Museum ofComparativeZoology, Harvard University, 26Oxford Street,Cambridge,MA,United States 2Nature Kenya(EastAfricaNatural HistorySociety), National Museums ofKenya, MuseumHill,Nairobi,Kenya [email protected] Received April29,2009 /ReviewedandAcceptedJune6,2009 Thegrowinguseofpesticidesforfishingisacurrentpracticeofconcernforbiodi- versityin EasternAfrica.Thereislittle informationavailable,however,ontheextent andeffectofthispracticeonconspicuousindicator groups like dragonflies.Odon., especiallyduringthe larvalstage,areparticularlyvulnerable topesticides.This sur- veyfound significantdifferencesin bothdragonflyabundance anddiversity inbays ofLakeVictoriathathadbeenfishedusingpesticides.Only 1-2dragonflyspp.canbe foundinareasroutinelypesticide-fishedincontrastwith >20 spp.inprotectedareas. Thissurvey highlightsthedetrimental effectofpesticidefishingoninvertebrates. INTRODUCTION Odonatahavewidely been viewedas usefulindicatorsofwater quality. They are particularlyvulnerabletopoorwater qualityinthelarvalstageboththrough directpoisoning, andas thelarvalstagesare obligate predators,through there- ductionof their prey which are also susceptible to poisons (CORBET, 1983). High concentrationsofinorganicpollutants andpesticides havebeen shownto affect different life-stages ofodonatesand application of insecticidesisknown to produce a direct effectof decreasein ©donatenumbersfrom variousstudies as summarisedby CORBET(1999). As brightly coloured, diurnaland active insects they are also easy to track and identify and thereforea usefulproxy forquality ofa given aquatic habitat 262 D.J.Martins (CLARK& SAMWAYS, 1996).Thismakesthemareliableindicatorofthehealth, stability andintegrityofa givenwetlandhabitat(CHOVANEC& WARINGER, 2005). Whiledifferencesinlife-historystrategiesmayresultinunevenspecies’ dis- tributions, the diversityandabundanceofmost commonspecies canbe tiedto the healthofthehabitat(WATSON etal. 1982). LakeVictoriais the world’ssecondlargestfresh-waterlakeand animportant areaforawiderange ofbiodiversity taxa including cichlidfishandbirds(BEN- NUN & NJOROGE, 1999).Thelake hassufferedfromanumberofecological catastrophes including theintroductionoftheNilePerch(PRINGLE,2005),the Fig. I. (A):diversityof dragonfliesatthreecategoriesofsitesin LakeVictoria,N =20sitesforeach category; - (B): meanabundance ofdragonfliesatthethree categoriesofsitesstudied inLakeVic- toria. — [errorbarsare+/-1 st,err.] Odonatadiversityinpesticide-fishedandother areasin LakeVictoria 263 waterhyacinthandmorerecently pollution (MARTINS,2006). Useofpesticides forfishingisone growingareaofconcern andthisstudysought toconductapre- liminary investigationoftheeffectofpesticides onthe diversityandabundance ofodonatesinLakeVictoria. MATERIALANDMETHODS All study siteswerelocated in theMwanza Gulf onLake Victoria. Threedifferent categoriesof sites werechosen: (i)areasthat werefished usingpesticides, (ii)areasthatwerefished traditionally (nopesticides,butusingnets, lines and fish-traps)and (hi)protectedareasthat wereneitherfished norhad pesticidesdirectlyapplied.Pesticide-fished areaswereidentified throughinterviews with lo- cal fishermen anddirect observation ofpoisonedfish and odonate larvae. Vegetationand habitat structurearoundall siteswas similarconsistingofarocky shorewithaquaticgrasses, Sesbania sp. andsmallpatchesofpapyrus, Odonate diversitywas quantifiedbyidentifyingthespeciespresentat each20differentsitesineach categoryover~l month(July)in2006and2008.This wasdonethrough direct observation and occasionally nettingand checkingwith fieldguides(TARBOTON&TAR- BOTON,2002; MILLER& MILLER,2003),Abundance ofodonates wasmeasured bycounting thetotalnumberofadultsflyingpast afixedpointfor 1hrbetween 10-11 a.m.in themorningonfive separatedaysatsitesforeachcategory.Thiswasthegeneralpeaktimeforinsectactivity. Onlyadult dragonflieswereidentified and countedastheywereeasier toidentify,damselflieswerenotincluded inthe survey.Dataweremaintained in spreadsheetsusingExcel andanalysedusingSTATA. RESULTS Significant differences were foundat allsites in both dragonfly diversityand abundance(Fig. 1A,B).Only2spp.werepresentinareasthatwerepesticide-fished incontrast withtraditionally-fished (10 spp.)andprotected areas (23 spp.) (Tab. I). Odonatediversity differedsignificantly between all sites (Fig. 1A, ANOVA F =86.6p> 0.0001). Thereweresignificant differencesinodonateabundance 257 betweenpesticide-fished andthetwoothercategories of sites(Fig. 1A,ANOVA F212= 12.07p> 0.0013).Odonateabundancedidnotdiffersignificantlybetween traditionally-fished sitesandprotected areas(Fig. 1B,Bonferronipost-hoc testp >1.0),but diversitydid(p > 0.0001) DISCUSSION Theresults ofthissurvey show thatthere are cleardifferencesbetweenareas thatare pesticide-fished and areas thatare not subject to directapplication of pesticides onLakeVictoria. Thetwo species thatwereobservedinthepesticide- -fishedareasare Pantalaflavescens, thatiswidelyknownto migratebothlocally andregionally (TARBOTON &TARBOTON, 2002)andBrachythemis leucostica, whichweobservedtobethemostcommon dragonflyin theregion. B.leucostica was observedto lay many hundredsofeggs atthe baseof vegetation surround- ing the baysandinletsof thelake.Thisreproductive strategy maybe the reason why it isableto survive as it isbothproducing large numbers ofoffspring and 264 D.J.Martins TableI Diversity ofdragonflyspeciesobserved atthedifferent sites SSppeecciieess PPeessttiicciiddee--ffiisshheedd TTrraaddiittiioonnaallllyy--ffiisshheedd PPrrootteecctteeddaarreeaa IIccttiinnooggoommpphhuussffeerrooxx XX AAnnaaxxeepphhiippppiiggeerr XX AAnnaaxxssppeerraattuuss XX AAnnaaxxiimmppeerraattoorr XX XX PPhhyylllloommaaccrroommiiaappiiccttaa XX BBrraacchhyytthheemmiissllaaccuussttrriiss XX BBrraacchhyytthheemmiisslleeuuccoossttiiccaa XX XX XX PPaallppoopplleeuurraalluucciiaa XX XX PPaallppoopplleeuurraappoorrttiiaa XX XX RRhhyyootthheemmiisssseemmiihhyyaalliinnaa XX XX OOrrtthheettrruumm ccaaffffrruwmn XX OOrrtthheettrruumm cchhrryyssoossttiiggmmaa XX AAcciissoommaappaannoorrppooiiddeess XX XX PPaannttaallaaffllaavveesscceennss XX XX XX PPhhiilloonnoommoonn lluummiinnaannss XX XX SSyymmppeettrruummffoonnssccoolloommbbiiii XX TTrriitthheemmiisshheeccaattee XX TTrriitthheemmiiss kkiirrbbyyii XX TTrriitthheemmiissaarrtteerriioossaa XX XX CCrrooccootthheemmiisseerryytlhhrraaeeaa XX XX utilisingmicro-habitatsthatmaybemoreshelteredfromtheeffectsofpesticide application. Theuseofpesticides inand aroundLakeVictoria needs tobeaddressedat a regional level.Pesticidesare widelyavailableinlake-sidetowns,such asMwanza, andsometimes soldin shops thatdeal in fishing gearand tackle.It isapparent thatit isafew unscrupulous fishermenwhouse this technique andnot thema- jority. Several differentpesticides are usedandmore work needs tobe doneon thedifferentkindsofpesticides being used, including through both surveys of indicatorgroups such as dragonflies on LakeVictoriaandmore detailedanaly- sis ofwater and aquatic habitats. ACKNOWLEDGEMENTS This studywasmadepossiblethroughfieldassistance fromstudentsinthe ‘EcologyoftheGreat Lakes andAfrican Savannahs’coursein 2006and 2008 including:L. ARMSTRONG, C.ALLEN, A. CRANDALL-RAINES, A.FARRAND, T.KUCHNER, C. LAVERRIERE and N.LEWIS. Thanks totheparticipatingfishermen and toProfessors G. MIKELL and B.D. HOFFMANof GeorgetownUniversityandG.MUTALEMWAofSt.AugustineUniversity, Tanzaniaforlogistical assistance. Odonatadiversity inpesticide-fishedand otherareasinLake Victoria 265 REFERENCES BENNUN,L.A.& P.NJOROGE, 1999. ImportantbirdareasinKenya.EastAfrica Nat. Hist.Soc., Nairobi. CHOVANEC, A.&J.WARINGER,2005.Dragonflies(Insecta:Odonata)asindicators oftheeco- logicalhealth ofwetlandecosystems. Verb.int. Ver Limnol. 29:422-425. CLARK,T.E.&M.J.SAMWAYS, 1996.Dragonflies(Odonata)asindicators ofbiotopequalityin theKrugerNational Park,SouthAfrica. J.appl. Ecol. 33: 1001-1012. CORBET.P.S., 1983. Abiologyofdragonflies.Classey,Fanngdon. CORBET,P.S., 1999. Dragonflies:behavior andecologyofOdonata. Comstock,Cornell Univ. Press, Ithaca/NY. MARTINS,D.J.,2006. LakeVictoria,giantindistress.SwaraJ. E.Afr. WildlifeSoc. 3:34-48. MILLER,P.&K. MILLER,2003. East Africandragonflies:aguideforresidents,students andvisi- tors.Nature Kenya,Nairobi. PRINGLE, R.M.,2005.Theoriginsofthe NileperchinLakeVictoria. BioScience 55;780-787. TARBOTON, W.& M,TARBOTON, 2002. A fleldguidetothe dragonfliesofSouth Africa. War- wick, Modimolle. WATSON,J.A.L.,A.H,ARTHINGTON&D.L. CONRICK,1982.Effectofsewageeffluentondrag- onflies (Odonata)ofBulimbaCreek,Brisbane. Austr.J.MarineFreshw. Res. 33; 517-528.