Substrate selection in larval Calopteryx splendens (Harris) (Zygoptera: Calopterygidae)

2008·1.6 MB·English

by  Ward, L

Odonatologica37(1):69-77 March I.2008
Substrate selection
inlarval Calopteryxsplendens (Harris)
(Zygoptera: Calopterygidae)
L. Ward¹* andP.J.Mil²l
1Askham BryanCollege,AskhamBryan,York,YD233FR,United Kingdom
2School ofBiologicalSciences,UniversityofLeeds,Leeds,LS29JT, United Kingdom
ReceivedJuly18,2007/RevisedandAcceptedAugust10,2007
Underexperimental conditions,the relationshipbetween substratecomposition,
with andwithoutthepresence ofanemergencesupport, andlarval distributionwas
investigated.Resultsrevealed thatC. splendenslarvaeshowedaclearpreferencefora
pebblesubstrate asopposedtosand orsilt,when all 3substrateswereequally avail-
able.However,thesubstrate typedecreasedinimportanceasthedensityoftheemer-
gence supportincreased. Resultssuggestthat themorphologicaladaptationsofC.
splendenslarvae,toclingtoasubstrate,canbeutilisedequallyinavegetatedhabitat
andahabitat predominatedbypebblesandcobbles.Thishasimplicationsforthedis-
persalofC.splendenstoareascontaining, traditionally,lessfavoured habitat. Range
expansionofC. splendensonitsnorthernborders,whereaquatichabitatcharacteris-
tics candiffermarkedlyfromwaterwaysinlowland southernEngland,isdiscussed.
INTRODUCTION
Thesubstrate inan aquatic environmentcan consist of varioustypes ofor-
ganicand inorganic materialsandprovides something sufficiently stable forin-
vertebratesto crawlon, cling to orburrow in (MINSHALL, 1984). Theleaves
andstemsofriparian vegetation andaquatic macrophytes constitutetheorganic
substrate. Thenature ofthe inorganic componentis largely determinedby the
underlying geology. Inorganic substratesare usually sedimentary materialsand
canrangefrommicroscopic siltandclay particles to largeboulders.
Apreliminary studyoninvertebrate-substraterelationships wasmadebyPER-
CIVAL& WHITE(1929),whoinvestigated themacroinvertebratefaunaassoci-
atedwithseven substratetypeswithdifferingamounts ofplantcover, instreams
*CorrespondingAuthor
70 L. Ward&PJ. Mill
innortheasternEngland. Morerecent studiesof aquaticinvertebratelarvalhab-
itat selection haveconcentratedon burrowing species, including gomphid odo-
nates (KEETCH & MORAN, 1966;HUGGINS&DUBOIS, 1982;SUHLING,
1994, 1996),and research onsubstrateselection withregard toodonatespecies
thatdo notburrowis scarce.
Previous work has shown that substrateparticle size and composition, and
the presence orabsence ofaquatic vegetation, can directly affect themicrodis-
tribution of aquatic insects (CUMMINS & LAUFF, 1969; SUHLING, 1996),
although otherfactorssuch as prey distributionandpredation, as wellas com-
petitionforspace,canresultina variationinabundancefromonesubstratetype
toanother(BAKER, 1981;BRUSVEN&ROSE, 1981;McPEEK, 1990). Inad-
dition,thebehaviourofovipositing femalesdetermines, tosomeextent,thetypes
ofresources availableto theiroffspring, andmanylarvalspecies are aggregated
initially asaresultofthe patchiness oftheoviposition substrate.Theseoffspring
normally selectmicrohabitatswithinthegeneral vicinityof theoviposition site
(JAENIKE &HOLT, 1991).Nevertheless, movementsbetweenmicrohabitatsare
commonplace duringlarvaldevelopment (CORBET.1962). MACAN(1966) re-
ported that larvalLestessponsadisperse soon afterhatching fromthemarginal
plants inwhichtheeggs werelaidandJOHANSSON (1978)showedthatEryth-
romma najas larvaemovefromopenmudbeneathoviposition sites toreedbeds.
Other factors implicated in the microdistributionof larvae includeseasonality
(LAWTON, 1970)andthelarvalstage(MIYAKAWA (1969) citedin CORBET,
1999).
Recentinvestigation ofthedistributionofC.splendens intheUK suggests that
the species has undergone rangeexpansion (WARD & MILL, 2007). Indeed, in
2004 itwas reported in large numbers nearDalbeattie, south-west Scotland by
B.& R. Mearns(BATTY, 2007) havingpreviously beenunrecordedinScotland,
except forcasualsighting. In2006 C. splendens was recorded on thestony Urr
Waterin Kirkcudbrightshire, againinsouth-westScotland(BATTY, 2007). This
contradicts the widely heldview that C. splendens larvae preferentially inhabit
silt substrate.Previously consideredas a‘southernspecies’(e.g. BROOKS, 2004),
where thereareextensiveslowflowing rivers inlowlandhabitats, withwell-veg-
etatedbanks. C. splendens cannowbe consideredasmorewideranging. Indeed,
thepresenceofthespeciesinsouth-westScotland,onwaterwayswithcontrasting
physicochemical featuresto thosefurthersouth, suggests itcan adapt to condi-
tionswhichfacilitateitsmovement northward.
Morphologically, C. splendens larvaehaveanelongate abdomenand clasping
legs(CORBET, 1999).Theseadaptations facilitatetheirhabitationoflotichab-
itats by allowingthemto cling to asubstrate, such as vegetation stems.Indeed,
thelarvaehavebeen observedtoinhabittheroot andstemregions ofsubmerged
vegetation, insiltandmud-bottomedstreamsandrivers(MERRITT etah, 1996;
BROOKS, 2004).
Substrateselectionin Calopteryxsplendens larvae 71
Thisstudyaimsto determinetherelationship betweensubstratecompositions,
withandwithoutthepresenceofanemergencesupport,andthespatialdistribu-
tionofF-l andF-2instar larvaeof C. splendens.
METHODS
ThelarvaeofC.splendensusedforsubstratumselectionexperimentswerecollectedfromtheRiver
MedwayinKent,southeastEngland(suppliedbyBladesBiologicalSupplies). All larvaeweresorted
into instarsaccordingtohead width (C.Schutte,pers. comm.),andwereclassified asstagesF-l and
Each ofthree glassaquaria(0.48x0.36m)
weredivided by twolongitudinalblack perspex
sheets into three sections ofequal dimensions.
Within each sectiontherewerethree equalsized
areas, each containing one ofthree substrate
types: coarsegraveland small pebbles, medium
sand,andsilt/clay(aftertheWentworth Classifi-
cation ofsubstrate particlesize). Priortotheex-
perimentthesubstrataweresetoutintheaquaria
toadepthof0.02m and theaquariafilledwith
Fig. 1.AquariadesignusedforExperiment1. watertoadepthof0,12 m.Thewaterwasleft to
stand for 36 hours with airstones to aeratethe
water. In order tokeep the water aerated throughouttheexperiment,oneairstonewaspositioned
abovethe centresubstrate ineachblock ofthree,in such apositionthatthe larvaecould notcling
toit.Inallexperimentsawatertemperatureof 12°Canda 12:12hour 1ightidarkphotoperiodwere
maintained.
DuringallexperimentsthelarvaewerefedbyplacingoneChironomus larvaoneachsubstratetype,
unlessotherwisestated. C. splendenslarvaewerefedin thiswayonalternatedaystopreventtheset-
tlingofuneatenfooditems influencingsubstrate selection. Observationsofthepositionofthe lar-
vaewererecordedeveryhourfrom1000until 1800(GMT)fortendays,between FebruaryandApril
2001.
EXPERIMENT1.- An experimentwascarriedouttocontrolforlarvalpreferenceforposition
intheaquaria.Eachsectioncontained asinglesubstratetypeand onelarva wasplacedoneachsub-
strate(i.e.ineach section)inoneofthreestartingpositions(A,BorC).This wascarriedoutforeach
startingposition(Fig. 1).
This preliminaryinvestigationrevealed asignificant difference in the observed positions of C.
splendensintheaquaria(x2=16.47,d.f.=2,P<0.001).Atotalof40%oftheobservations (n=810)
werein themiddleregion(B),whereas 29.3%and30.7%ofthe observationswereinregionsAand
C,respectively.
EXPERIMENT 2, SUBSTRATE SELEC-
TION. - The followingexperiments were per-
formed todetermine therelative importanceof
larvalchoiceforsubstratetypeand/orvegetation.
Eachareawithin each sectioncontained adifer-
entsubstrate. Arandomisedorderoftheposition
ofthesubstrateswasusedinallexperiments(Fig.
2):
(a) Intheabsenceofaverticalsupport- Inthe
firsttank(1ato1c)oneindividual wasplaced
ineachpebblearea,inthesecondtank(2ato Fig.2.AquariadesignusedforExperiment 2.
72 L. Ward&P.J.Mill
2c) ineach silt areaandin the thirdtank (3ato3c)ineach sand area(Fig.2). Twotrials were
conductedwith ninelarvaeusedineach(18replicates).
(b) Inthepresenceoflowdensity‘vegetation’- Onepiece ofartificial‘vegetation’,inthe formof
adowelrod (5mmdiameter),wasplacedinthe centreofeach substrate (i.e,ineachofthe nine
areas) tosimulate sparse vegetation andtherebytoinvestigateifsparsevegetationismoreim-
portantthanasedimentsubstrate orvice versa.The experimentalprocedurewasthesameasin
experiment2(a).
(c) in thepresence ofhighdensity ‘vegetation’— Sixteen dowel rods werespacedevenly ineach
substrate (i.e.ineachofthe nineareas)tosimulatedense vegetationandhenceitsimportancein
relation tosparsevegetationand substrate.The experimentalprocedurewasthe sameasin ex-
periment2(a).
However, duetothehighmortalityrateoftheexperimentalanimals,onlyonetrialofexperiments
bandcwasperformed.Atotalof 12individuals wereused intheseexperiments,whichhad allbeen
used intheprevious substrate-onlyexperiment.
RESULTS
SUBSTRATESELECTION INTHEABSENCEOFAVERTICAL SUPPORT
Inthe absence ofa verticalsupport,the distributionof Calopteryxsplendens
differedsignificantly betweenthe threesubstratetypes(%2=832.45, d.f. = 2,P <
0.0001)withobservationsoflarvaeonthepebblesubstratebeing recordedsignifi-
cantly moreoften thanthoseoneitherthesiltorthesandsubstrates (Tab. I).
SUBSTRATESELECTION WITHA LOWDENSITYOFVERTICAL SUPPORTS
Theintroductionofaverticalsupport,intheformofone dowelrodperblock,
reducedtheproportion ofobservationsrecordedin thepebble substrateareasin
comparison to the absence ofa verticalsupport(Tab. I) butthenumberofob-
servationsinthepebble substrateareas(i.e. thoseonthesubstrateitselftogether
withthoseon thedowelrods inthatsubstrate) werestillsignificantly greaterthan
ineitheroftheothertwo substrateareas((%2= 182.31,d.f.=2, P<0.0001) (Tab.
Table I
Influence ofsubstrate and ofdowelrods (simulatedvegetation)onsub-
strate selection by larval Calopteryxsplendens. Data are larval occu-
pancy of each substrate area(%)with the number of observationsin
parentheses
Sedimentareachoice P
Pebbles Silt Sand
Substratum alone 66.9(1084) 19.8(321) 13.3(215) <0.001
1dowel rod 52.6(562) 26.0 (278) 21.4(228) <0.001
16dowel rods 32.1 (343) 32.0(342) 35.9(383) N.S.
N.S.=notsignificant
Substrateselectionin Calopteryxsplendenslarvae 73
I). However, the lar-
vae were recorded
significantly moreof-
ten onthedowelrods
(53.7%) than on the
substrate (46.3%) (x2
= 5.99, d.f. = 1, P =
0.01).
Ofthoseclinging to
thedowelrods, 38.2%
of the observations
were in the silt sub-
strateareascompared Fig. 3.Theproportionofobservations(n=1068)oflarvaeineachsub-
with34.2%and27.7% strateareainthe presence ofonedowelrodper substrate type.
in thepebble andsandsubstrate areas,respectively (Fig. 3);this differencewas
significant(x2=9.58,d.f. =2, P=0.01).Nevertheless, considering onlythoselar-
vae thatwere locatedonthesubstrateitself,theywere stillrecordedsignificantly
moreoftenonthepebble substrate(x2=367.67, d.f.= 2,P<0.0001) thanonei-
therthesand orsiltsubstrates(Fig. 3).
SUBSTRATESELECTIONWITHAHIGHDENSITYOFVERTIAL SUPPORTS
Withthe increasein density of vertical supports to 16dowelrods per block,
there was no significant differencebetween the numberof observations in the
threesubstrateareas(i.e. thoseonthe substrateitselftogetherwiththoseonthe
dowelrods inthatsubstrate) (x2= 3.07,d.f. = 2,P =0.22)(Tab. I).
Thenumberofobservations oflarvae onthesubstrateitselfdecreased signif-
icantly (x 2= 226.65,
d.f. = 1, P < 0.001),
with 73% of the ob-
servations record-
ed on the dowelrods
and the number on
the latter was signifi-
cantly greater in the
sand substrate areas
(X2= 17.22,d.f. = 2,P
= 0.0002), thanin ei-
ther thepebble or silt
substrate areas (Fig.
4). Again, consider-
Fig. 4. Theproportionofobservations (n= 1068)oflarvaein each ing only those larvae
substrate areain thepresence of 16dowelrodspersubstrate type.
74 L.Ward&P.J.Mill
thatwere located on the substrateitself, the numberof observationsof larvae
on thepebble substratewas stillsignificantly greaterthan oneither sandor silt
(X2= 13.56,d.f. = 2,P= 0.001) (Fig. 4).
DISCUSSION
Inthe current study, Calopteryx splendens larvae showed a clear preference
forapebble substrate as opposed to sandor silt,when all threesubstrates were
equallyavailable.AsimilarpreferencehasbeenfoundforPteronarcys californica
(Plecoptera), Arctopsyche grandis(Trichoptera), Ephemerella grandis(Ephemero-
ptera)andBrachycentrus sp.(Trichoptera), undercomparable laboratorycondi-
tions(BRUSVEN&PRATHER, 1974).Furthermore,undernaturalconditions,
aquatic insects withrelatively long larval lifecycles of two to three years,such
as some dragonflies andstoneflies(Pteronarcys, Paragnetina),are commonly as-
sociated with bouldersubstrates, possibly because theyare more likely to find
shelter during highwater flow(DE MARCH, 1976;MINSHALL, 1984). Inad-
dition,severalauthorshavestatedthathabitationofstonysubstrates mayresult
in reducedpredation by fish(SUHLING, 1994;McPEEK, 1990;BRUSVEN &
ROSE, 1981).Inadditionto predation,otherimportant factorsin microhabitat
selection arecompetition (SUHLING, 1996) andthe physical characteristics of
the habitat.
Calopteryx larvae have a high oxygen requirement, which effectively limits
membersofthe genusto lotichabitats(WILDERMUTH, 1994).Furthermore,
ERIKSEN(1966) hasshownthatthe presenceofsiltand finesand can cause a
significant reductionintheoxygencontentwithinthesubstratum.Inaddition, silt
maysignificantly affect thehabitability of asubstratumby altering watermove-
ment and food quality(MINSHALL, 1984). Heavy silting generally results in
reducedinsect species diversityand productivity (MINSHALL, 1984). Several
authors havenotedthatthepresenceoflargeamounts ofsiltandsandcoincide
withareductionin theabundanceand diversityofstream insects(CHUTTER,
1969;MINSHALL,I984;NUTTAL,1972;PENNAK&VANGERPEN, 1947).
Thisisofsignificance to C. splendens whoselarvae are carnivorous.
Severalauthorshavereported a gradualincreaseinspeciesabundancefrombare
substrateto thepresenceofvegetation (PERCIVAL & WHITE,1929;WHITE-
HEAD, 1935;MINCKLEY, 1963)anditisestablishedthatthepresenceofaquat-
ic plants can lead to increaseddiversityof faunainaquatic habitats(MACAN,
1977;MACAN& MAUDSLEY, 1968),thusproviding a ready foodsupply for
predatory invertebratessuch as C. splendens.
Thecurrent studyshowedthatthesubstratetypedecreasedinimportance asthe
density ofdowelrods increased. C. splendens are generallyconsideredtoinhabit
silty areas (BROOKS, 2004), although thisobservationpresumably resultsfrom
thepresenceofaquatic vegetation insiltsubstrates, whichlarval C.splendens use
Substrateselectionin Calopteryxsptendens larvae 75
forcover, asanemergencesupportand, intheadultstage,as aperch.Wherethe
flowrate ofariver ismorerapid, sedimentdoesnot settleandtheresultantsub-
strateispredominantly pebbles andboulders.Theseconditionsarenot conducive
toeasyestablishmentof vegetation communities.Incontrast, in slowerflowing
regions, wheresilt settles outon theriverbed,plant communitiesare moreread-
ilyestablished.
GOODYEAR (2000) compared theenvironmentalrequirements of larval C.
splendens and C. virgo.Ingeneral thepredominant substratewhereC. splendens
wasfoundwassiltand detritus,although atthreeoutoften sites, sandandgrav-
elwas thedominantsubstratetype.GOODYEAR (2000) concludedthat,forC.
splendens, thegeneralpatternisastrongpreference, butnotanecessity,formuddy
bottoms,wherereadyestablishmentofmacrophyteandriparianplant communi-
tiesoccur.
MINSHALL(1984) suggested thatcertainspecies ofEphemeroptera, Tricho-
pteraand Odonataarerestricted to vascularplants, partlybecausethey provide
a substrateforclinging to. Additionally, since oxygen levels are much lower in
silt andmudsubstrates(see above) ahighoxygen intakemaybeachieveddirect-
ly from aplant surfacesince, during daylight, plants produce oxygen. However,
forspecies with aparticularly highoxygenrequirement, suchas Calopteryx virgo
(MERRITT etah, 1996;GOODYEAR,2000), whichare predominantly found
on apebble substrate, theflowrateof theriveris generallymorerapid, thus gen-
erating therequired oxygen leveland hencetheneed foroxygen produced from
aquatic vegetation is reduced.
As aspecies inhabiting lotichabitats, larval C. splendens haveadaptations for
clinging to asubstrateandhavebeendescribed as thigmotactic inhabit(COR-
BET, 1999).Thus,intheabsenceofaverticalsupport,such asavegetation stem,
pebbles wouldbe thefavouredhabitatbecausethey provide afirmsubstratefor
attachment.
Formerly consideredasouthernspecies, there is no doubtthatthe northern
rangeofC. splendens hasexpandedintonorth-easternEngland and,fromitsbase
innorth-westernEngland, intosouthernScotland(CLARKE, 1999; JEFFRIES,
2001;WARD & MILL, 2004). Rivers inthe northofEngland and inScotland
tendtobefast flowingovermuchoftheirlengthand, consequently, thesubstrate
particle size tendsto be greaterinthis sortoflotichabitat. C. splendens larvae
havethepotential tothriveinthis habitatsincethemorphologicaladaptation to
clingingto asubstratecanbeutilisedequally ina vegetated habitatandahabitat
predominated bypebbles andcobbles.
ACKNOWLEDGEMENTS
LWwassupportedbytheJohnHenryGamerScholarship,UniversityofLeeds.
76 L. Ward&P.J.Mill
REFERENCES
BAKER,R.L.,1981. Behavioural interactions and useoffeedingareasbynymphsofCoenagrion
resolutum (Coenagrionidae:Odonata).Oecologia49: 353-358.
BATTY,R,2007. Changesin Scottish dragonflies.DragonflyNews51:20-21.
BROOKS,S„ 2004.FieldguidetothedragonfliesanddamselfliesofGreatBritainandIreland. [2ndedn]
BritishWildlife Publishing,Hook,Hampshire,UK.
BRUSVEN, M.A.&K.V PRATHER, 1974. Influence ofstreamsedimentsondistribution ofma-
crobenthos. J.ent. Soc. Br. Colomb. 71:25-32.
BRUSVEN, M.A.& S.T.ROSE, 1981. Influence ofsubstrate compositionand suspended sedi-
ment oninsectpredationby theTorrent Sculpin, Cottusrhoteus. Can. J. Fish, aquat. Sci.
38: 1444-1448.
CHUTTER,F.M., 1969. Theeffectsofsilt and sandonthe invertebratefaunaofstreamsand riv-
ers.Hydrobiologia34; 57-76.
CLARKE,D., 1999. TheoutpostpopulationsoftheBanded Demoiselle Calopteryx splendens(Har-
ris) intheSolwayFirtharea,Cumbria:historicalperspectiveand recentdevelopments.J.Br.
DragonflySoc. 15:33-38.
CORBET, P.S., 1962. Abiologyofdragonflies.Witherby,London.
CORBET,P.S., 1999. Dragonflies:behaviour andecologyofOdonata. HarleyBooks,London.
CUMMINS,K..W.&G.H.LAUF, 1969.The influence ofsubstrateparticlesizeonthemicrodistri-
bution ofstreammacrobenthos. Hydrobiologia34: 145-181.
DE MARCH,B.G.E., 1976.Spatialandtemporalpatternsinmacrobenthic streamdiversity.J.Fish.
Res. BoardofCan.33: 1261-1270.
ERIKSEN,C.H., 1966. Benthicinvertebrates and somesubstrate-current-oxygeninterrelationships.
Spec. Publ. PymatuningLab. Ecol.4:98-115.
JEFFRIES,M.,2001. TheNorthumbrian frontierofthe Banded Demoiselle Calopteyx splendens
(Harris). J.Br.DragonflySoc. 17;5-58.
GOODYEAR, K.G.,2000.Acomparisonoftheenvironmentalrequirementsoflarvaeofthe Band-
ed Demoiselle Calopteryx splendens(Harris) and the BeautifulDemoiselle C. virgo(L.).J.
Br. DragonflySoc. 16:33-51.
HUGGINS, D.G. & MB. DUBOIS, 1982.Factors affectingthe microdistribution of twospecies
ofburrowingdragonflylarvae,with notesontheir biology(Anisoptera:Gomphidae).Odo-
natologica11:1-14.
JAENIKE,J.& R.D.HOLT, 1991.Geneticvariation forhabitatpreference: evidence andexplana-
tions. Am. Nat. 137(Suppl.):67-90.
JOHANSSON, O.E., 1978. Co-existence of larval Zygoptera(Odonata)commontothe Norfolk
Broads. Oecologia32: 303-321.
KEETCH, D.P.&V.C. MORAN, 1966. Observations onthebiologyofnymphsofParagomphus
cognatus(Rambur)(Odonata:Gomphidae).I.Habitatselection inrelation tosubstratepar-
ticle size. Proc. R ent.Soc. (A)41: 116-122,
LAWTON,J.H., 1970.ApopulationstudyonlarvaeofthedamselflyPyrrhosomanymphula(Sulzer)
(Odonata:Zygoptera). Hydrobiologia36: 33-52.
MACAN,XT., 1966.Predation by Salmotruttoinamoorland fishpond.Verb.int. Ver. theor. angew.
Limnol. 16: 1081-1087.
MACAN,XT.&R,MAUDSLEY, 1968. TheinsectsofthestonysubstratumofWindermere. Trans.
Soc. Br. Ent. 18; 1-18,
McPEEK, M.A., 1990. Determination ofspeciescompositionin the Enallagmadamselflyassem-
blagesofpermantlakes. Ecology71:83-98.
MERRITT,R„ N.W.MOORE& B.C.EVERSHAM, 1996. AtlasofthedragonfliesofBritainand
Ireland. [ITEResearch Publ. 9],HMSO,GreatBritain.
Substrateselection inCalopleryx splendemlarvae 77
MINCKLEY,W.L., 1963. The ecology ofa spring stream: Doe Run,Meade county,Kentucky.
Wildl. Monogr. 11: 1-24.
MINSHALL,G.W.,1984.Aquaticinsect-substratumrelationships.In:VH.Resh&D.M.Rosenberg,
[Eds], Theecologyofaquaticinsects. Praeger,New York,
NUTTAL,P.M.,1972. The effectsofsanddepositionupon themacroinvertebrate fauna oftheriver
Camel,Cornwall.Freshw. Biol. 2: 181-186.
PENNAK,R.W.& E.D.VAN GERPEN, 1947. Bottom faunaproductionand physical natureof
thesubstrate inanorthColorado trout stream.Ecology28:42-48.
PERCIVAL,E.&H.WHITE, 1929. A quantitativestudyofthefaunaofsometypesofstreambed.
J.Ecol. 17:282-314.
SUHLING,F., 1994. Spatialdistribution ofthelarvae ofGomphuspulchellusSelys (Anisoptera:
Gomphidae).Adv.Odonatol.6: 101-111.
SUHLING,R, 1996. Interspecificcompetitionandhabitatselection bytheriverine dragonflyOny-
chogomphusuncatus.Freshw. Biol. 35:209-217.
WARD,L.&P.J.MILL,2004.Distribution oftheBandedDemoiselle Calopteryxsplendens(Harris)
in northern England:anexampleofrangeexpansion?J.Br. DragonflySoc. 20:61-69.
WARD,L.& P.J.MILL,2007.Longrange movementsbyindividuals asavehicle forrangeexpan-
sioninCalopteryxsplendens(Odonata;Zygoptera). Europ.J. Ent. 104: 195-198.
WHITEHEAD,H., 1935.AnecologicalstudyoftheinvertebratefaunaofachalkstreamnearGreat
Driffield,Yorkshire. J.anim. Ecol. 4:58-78.
WILDERMUTH,H„ 1994. Habitatselektion bei Libellen, Odonatologica6: 223-257.