Constant seed size and mandible growth – a fundamental problem for granivorous ground beetle larvae (Coleoptera: Carabidae)

2006·1.2 MB·

by  Wilfried Paarmann

©EntomologicaFennica.11October2006
Constant seed size and mandible growth
– a fundamental problem for granivorous ground beetle larvae
(Coleoptera: Carabidae)
WilfriedPaarmann,NicolasFaust,ErikArndt,IngridLüchtrath&WolfgangRohe
Paarmann,W.,Faust,N.,Arndt,E.,Lüchtrath,I.&Rohe,W.2006:Constantseed
sizeandmandiblegrowth–afundamentalproblemforgranivorousgroundbee-
tlelarvae(Coleoptera:Carabidae).—Entomol.Fennica17:334–339.
Feeding on small tree seeds at fruit falls is a specific adaptation of harpaline
groundbeetlesandtheirlarvaeintropicalrainforests.Usingmandiblesastools,
theyhavetoperforatetheseedshelltoreachthenutritiousinterior.Theisometric
growth of larval mandibles, known from predatory species, would result in a
changing ratio between seed and mandible size during the course of develop-
ment.Thestablesizeofhosttreeseedsshouldselectanoptimummandiblesize,
similarforthethreelarvalinstarsinspermophageousspecies.Wefoundanin-
creasingtendency tomaintainthelengthoftheapicalmandiblepart(apex)in
seedfeedingspecies.ThesizeincreaseishigherinthespeciesfromAustralia,Af-
ricaandSouthEastAsiathanintheNewWorldspeciesofthegenusNotiobia.
Feedingexperimentshaverevealedthatlarvaeofspecieswithastrongerincrease
inapexgrowtharealsoabletodevelopaspredatorsofDrosophilalarvae.Ourre-
sultsfurthermorepointtoaninfluenceofbeetlesizeandshellhardnessofthehost
treeseedsontheapexsizeanditsgrowthrate.
W.Paarmann,N.Faust,I.LüchtrathandW.Rohe,HAWK,FakultätRessourcen-
management,Büsgenweg1A,D-37077Göttingen,Germany
E.Arndt,AnhaltUniversity,Department1,StrenzfelderAllee28,D-06406Bern-
burg,Germany
Received15January2006,accepted30June2006
1.Introduction shortageofpreyonthesoilsurface(Erwin1979),
while ample seed supply is available, although
Groundbeetlespeciesfeedingonsmalltreeseeds thisisrestrictedinspaceandtime.Manytreespe-
aresofaronlyreportedfromtropicallowlandrain cies in the Amazonian lowland forests produce
forests(Borcherdingetal.2000,Paarmannetal. suchabundantseedresource:Ficusspp.(Mora-
2001, 2002, 2003, Arndt & Kirmse 2002). The ceae), Bellucia spp., Loreya spp., Miconia spp.
ancestors of harpaline ground beetles were car- (Melastomatacea), Coussapoa asperifolia (Cec-
nivorousspecies(Beutel1997,ZettoBrandmayr ropiaceae), Goupia glabra (Celestracea) and
etal.1998).Thechangefromanoriginallycar- Vismiaguianensis(Clusiaceae)(Arndt&Kirmse
nivorouslifehabittospermatophagymaybethe 2002,Paarmannetal.2000,2002,2003).Inthe
result of selective forces caused by a general South-EastAsian(Borcherdingetal.2000),the
ENTOMOL.FENNICAVol.17 (cid:127) Carabidlarvaefeedingontreeseeds 335
Fig.1.Notiobiamandible.Arrowsindicateadditional
terebralteethtypicalforlarvalmandiblesofmostof
theseed-feedinggroundbeetlespeciesintropicalrain
forests.Apexsizeisdefinedasbeingthedistancebe-
tweenthetipofthemandibleandthetipofthe
retinaculum(seeTables1–2).
Australian(Kowalski2004),andAfrican(Paar-
mannetal.unpubl.data)rainforests,Ficusspp.
play a similar role. Kowalski (2004) added
Dendrocnide photinophylla (Urticaceae) to the
groupofhostplantsintheAustralianrainforest.
Thebiggestseedsofthesehostplantsarethatof
Goupiaglabrawith2.3mmlengthand1.1mm Fig.2.Abrasionsofthelarvalmandible,causedby
width(Arndt&Kirmse2002). hardseedshells:second-instarmandiblesof
Lampetessp.1,fedwithDrosophilalarvae(up)and
Ontheirwaytoadapttoaseed-feedinghabit,
fedwithFicusbenjaminaseeds(down).
thebeetlespecieswithlarvaespecializedtoone
type or size of seed had to solve a fundamental
problem:theseedsremainthesamesizewhilethe
seed-shell perforating tools (mandibles) of the In this paper, we provide additional support
larvaeincreaseduringthedevelopmentfromthe for this hypothesis, using new material. While
first to the third instar. The functional part of a KirmseandArndt(2002)usedlarvaefromfield
nut-cracker,optimalforopeninganutofcertain catches(whichoftenshowabrasionsattheapex
size and shape, should be kept at the same size site,Fig.2)andcomparedthetropicalseedfeed-
during the general enlargement of the larva. ers with predatory species from temperate re-
Arndt&Kirmse(2002)werethefirsttoshowthat gions,weusedmandiblesoflarvaefromfeeding
thereisatendencyamongtheseed-feedinglarvae experiments with Drosophila larvae or wild
to reduce the enlargement of the apex (distance poppy (Papaver rhoeas) seeds, which do not
betweenthetipofthemandibleandthetipofthe causeabrasions.Furthermore,wecompareddata
retinaculum; Fig. 1) among different larval fromseed-feedingspecieswithdatafromapred-
stages. Consequently, the percentage of apex atory species, commonly appearing on fig fruit
growthbetweenthefirstandthethirdlarvalin- falls in the Australian rainforest. Arndt and
starscanbeusedasasimplemeasureforcharac- Kirmse (2002) measured only species from a
terisingthedegreeoflarvalspecialization,caused SouthAmericanrainforest;wealsoincludedspe-
bytheforcetokeeptheirnut-crackeratanopti- cies from Australian, African and South East
malsizeduringbodyenlargement. Asianrainforests.
336 Paarmannetal. (cid:127) ENTOMOL.FENNICAVol.17
2.Materialandmethods
Wemeasuredtheapexoflarvalmandiblesof11
ground beetle species originating fromfruit fall
sitesindifferenttropicallowlandrainforests.The
parentalstockofsevenNotiobiaspecieswascol-
lected during field studies in the Reverva Flo-
restal Adolpho Duke near Manaus (Amazonia,
Brazil); for details, see Paarmann et al. (2001,
2002,2003).TheparentalstockofLampetessp.1
wascollectedduringfieldstudiesinBrunei(Bor-
neo); for details, see Borcherding et al. (2000).
TheColeolissuspapuaandTrichotichnusstoreyi
breedingstockwascollectedintheNaturalRe-
serves‘LakeBarin’and‘LakeEacham’intheTa-
blelands of Queensland (Australia) during field
studiesin2000and2001.Platymetopussp.was
collectedduringFebruary2004intheKakamega
Forest,WesternKenya,Africa.
The mandibles used for the measurements
werecollectedduringfeedingexperiments;these
are described in detail by Paarmann (2002). To
estimatethegrowthoftheapex,wemeasuredthe Fig.3.Apexgrowthinthepredatoryspecies
distancebetweenthetipofthemandibleandthe Coleolissuspapuafromthefirsttothesecondandthe
tipoftheretinaculum(Fig.1)withamicroscope thirdinstars(up)andintheSouthAmericanseed
feederNotiobiaaulica(down).Whitelinesconnecttips
under10xmagnificationthatallowedaprecision
ofmandiblesandtipsofretinaculi.
of 5 µm. Only mandibles from larvae fed with
Drosophila larvae or wild poppy were used to
avoidaninfluenceofabrasions.Forthedescrip- thestudiedspecies(subfamilyHarpalinae).Dif-
tive statistics and the statistical analysis (F-test, ferencesingrowthwereconnectedtothefeeding
Student’s t-test) we used Microsoft Excel and habitsofdifferentspecies.Thepredatoryspecies
WinSTAT.SEM(scanningelectronmicroscope) C.papuadoubleditsapexlengthfromthefirstto
photos were taken with the help of the Max the third instar, while the apex length was only
PlanckInstitute,Göttingen.Theassessmentcrite- 17.1% longer in the third vs. in the first larval
riaoffoodqualityforthedevelopmentofground instarintheseed-feedingN.aulica(seealsoTa-
beetle species was described in detail in ble1).Fig.3illustratesthesetwodifferenttypes
Paarmann (2002). We used a score system be- of mandible apex growth. The data of all mea-
tween 0 (larva died without feeding) and 10 surementsarepooledinTable1.
points(thehatchedbeetlesurvivedforatleasttwo Inall,66%(±5.1%SD)ofallseedseatendur-
months). To describe food quality, we used the ingdevelopmentwereconsumedduringthethird
average score points, together with the percent- larval instar in all the seed-feeding species
age of larvae that reached 10 points. We also [pooled data from all ten studied spermato-
counted the number of seeds consumed during phageousspecies;n=90;tenindividualsofeach
eachlarvalinstar. speciesexceptN.aulica(n=7)andN.maxima(n
=3)].Therefore,weassumedthattheapexshould
reach its optimal size at this stage; hence this
3.Results measurewasusedasthebaseofcomparisons(Ta-
ble2).
Wefounddifferencesinthesizeenlargementof IfofferedDrosophilalarvaeasprey,thenon-
theapicalpartofthemandibles (apex) between Americanspeciesreachedhighaveragefoodsuit-
ENTOMOL.FENNICAVol.17 (cid:127) Carabidlarvaefeedingontreeseeds 337
Table1.Apexsize(µm)ofthethreelarvalinstars(withSDandn,i.e.,thenumberofmeasuredmandiblesateachinstar)insev-
eralexaminedcarabidspeciesofthesubfamilyHarpalinae.Coloelissuspapuaisapredatoryspecies,alltheotherspeciesare
spermophageous.Continent=continentoforigin(Au=Australia,Af=Africa,SEAs=South-EastAsia,SAm=SouthAmerica).
Species 1.instar 2.instar 3.instar Continent
Size SD n Size SD n Size SD n
ColoelissuspapuaDarlington 260.0 5.3 12 366.0 15.8 15 531.8 30.4 17 Au
TrichotichnusstoreyiBaehr 302.5 14.9 12 375.8 9.0 12 490.0 11.3 12 Au
Platymetopussp. 418.5 7.5 20 507.7 9.2 20 600.5 13.1 20 Af
Lampetessp.1 406.7 6.6 21 453.5 10.4 20 531.5 18.1 20 SEAs
NotiobiaflavicinctusErichson 411.3 11.0 28 475.7 17.0 30 526.1 17.2 33 SAm
Notiobiapseudolimbipennis
Arndt 582.5 17.7 12 645.7 19.1 14 726.2 32.3 13 SAm
NotiobiaumbriferaBates 328.1 16.0 16 362.5 13.4 16 401.1 15.2 19 SAm
NotiobianebrioidesPerty 446.5 10.0 17 467.9 21.2 14 530.7 17.3 14 SAm
NotiobiaaulicaDejean 335.0 14.6 16 355.6 13.8 18 392.3 21.3 13 SAm
NotiobiaglabrataArndt 395.3 14.6 17 411.3 19.6 16 458.3 11.5 18 SAm
NotiobiamaximaArndt 417.5 32.0 4 437.5 26.3 4 440.0 8.2 4 SAm
abilityscores,similartothepurepredatoryspe- cinctus/N.pseudolimbipennis),significantlydif-
ciesC.papua(Table2).Theaveragescoresofthe feredintheirapexsize(Table2).Therewasno
Notiobiaspeciesthatfedonfruitflieswerelower, overlapbetweentheapexsizesofthesetwospe-
andonlytwoofthemsurvivedtotheadulthood. cies. The medium apex size of the third larval
Intwobeetlepairsofdifferentbodysize,each instarofthesmallerspecies,N.flavicinctus,was
adapted to the same seed type (N. umbrifera/N. distinctlylowerthantheapexsizeofthefirstlar-
aulicatoMiconia,andN.glabrata/N.maximato val instar of the bigger species, N. pseudo-
Bellucia),wefoundsimilarapexsizesinthethird limbipennis (Student’s t-test, p <0.0001). The
larvalinstarofthedifferentspecies(Table2).A seedsoftheirhostplantswereallinthesamesize
third pair, specialized on fig seeds (N. flavi- range (length 0.8–1.2 mm, width 0.7–0.8 mm;
Table2.Apexgrowthfromthefirsttothethirdlarvalinstar(%enlargement).Thecolumnsarefromlefttoright:
Species(notethatC.papuaisapredatoryspecies);%apexg.=%apexgrowth;Scorep.=scorepoints;%10
p.=%larvaethatreached10points(seetext);Elytra=lengthofelytrainmm;SD=standarderrorforthelength
ofelytra;n=numberofspecimensmeasuredforthelengthofelytra;Apexsize=apexsizeinthethirdinstarin
µm(forthreebeetlespeciespairs).Italics–specializedonfigseeds,underlined–specializedonMiconiaseeds,
bold–specializedonseedsofBelluciaandLoreya.
Species %apexg. Scorep. %10p. Elytra SD n Apexsize
Coloelissuspapua 104.6 8.4 35 8.0 0.2 6 –
Trichotichnusstoreyi 52.0 8.1 56 5.5 0.2 19 –
Platymetopussp. 43.5 8.8 40 – – – –
Lampetessp.1 30.7 8.2 37 6.2 0.1 21 –
Notiobiaflavicinctus 27.7 5.4 0 6.0 0.3 33 526.3
Notiobiapseudolimbipennis 24.7 5.5 0 8.2 0.3 26 726.2
Notiobiaumbrifera 22.2 2.4 0 5.4 0.3 12 401.1
Notiobianebrioides 18.9 6.0 0 8.2 0.5 21 –
Notiobiaaulica 17.1 5.9 23 6.8 0.5 22 392.3
Notiobiaglabrata 16.0 7.5 48 7.2 0.3 20 458.3
Notiobiamaxima 5.4 – – 9.0 0.3 21 440.0
338 Paarmannetal. (cid:127) ENTOMOL.FENNICAVol.17
Fig.4.Seeds(botanicallydefinedasnuts)offigspeciesfromtheAmazonianrainforest
nearManaus(Brazil).Fromlefttoright:Ficusguianensis,F.subapiculata,F.donell-
smithii,F.hebetifolia,F.guianensis-complex,F.albert-smithii,F.mathewsiiandF.
greiffiana.
Fig.4).Duringthethirdlarvalinstar,larvaeofthe sophila larvae. The lowest apex growth was
bigger species N. pseudolimbipennis opened on found in N. maxima, the biggest species among
average233.9seeds(±51.2SD;n=32),121.1% thestudiedseedfeeders.Itseemsplausibletoas-
more of Ficus seeds (pooled data for F. donell- sume that the biggest species should show the
smithii, F. guianensis and F. subapiculata) than strongestreductioninapexgrowth,comparedto
the smaller species N. flavicinctus [mean 105.1 thesmallerspecies.Notiobiamaximaisspecial-
seeds(±19.1SD;n=31)]. izedtoBelluciaandLoreyaseeds(Paarmannet
al.2002).ThesmallerspeciesN.glabrataisalso
veryabundantatfruitfallsofthesetreespecies.
4.Discussion Its larvae show a very similar apex size in the
thirdlarvalinstar(Table2;duetothelowsample
Alowerdegreeofapexgrowthwasfoundinlar- sizeofN.maxima,nostatisticswereperformed).
vae of all the studied spermophageous ground To reach this size, the smaller species needs an
beetle species, compared to the predatory ones apex growth of 16%, compared to 5.4% in the
[Table2;seeArndt&Kirmse(2002)forfurther biggerone.Similarconditionswerefoundinthe
dataofcarnivorousandomnivorousspecies].If speciespairN.aulicaandN.umbrifera(Table2,
thereductioninapexgrowthcorrelateswiththe underlined numbers). Both of these species are
degreeofadaptationtoseedfeedingfromorigi- specialized on Miconia seeds (Paarmann et al.
nally predatory species, this adaptation is more 2002);theyalsohaveverysimilarapexsizesin
advanced in the South American species of the the third instar (Student’s t-test, p = 0.2) and a
genusNotiobiathaninthestudiedspeciesfrom strongerapexgrowthinthesmallerspecies.The
the other continents. A possible explanation of significant differences of about 50 µm between
this phenomenon could be a larger seed and a thetwospeciespairs(t-test:for N. maxima:N.
lower prey resource at fruit falls in the Amazo- aulica and N. umbrifera, p <0.001 and for N.
nianlowlands,comparedtothesituationinsimi- glabrata:N.aulicaandN.umbrifera,p<0.001)
larforestsonothercontinents.Thiscorresponds mayreflectthenicheseparationamongthediffer-
alsowiththepercentageofseed-feedingindivid- ent Melastomatacea seed feeding specialists
uals in the carabid assemblages: 92–98% in (Miconia and Bellucia/Loreya, respectively),
Brazil vs. 39% in South-East Asia and 24% in found earlier by Paarmann et al. (2002) in the
Australia (Borcherding et al. 2000, Kowalski field.Thetwospecies,N.pseudolimbipennisand
2004,Paarmannetal.2001,2002). N. flavicinctus, specialized on fig seeds (Paar-
WhilethepredatoryspeciesC.papuaandthe mannetal.2001,Paarmann2003)alsodifferin
seed-feedingspeciesfromAfrica, Australiaand theirbodysize(Table2).Incontrasttothespecies
Borneoreachhighaveragescoreswhenfedwith pairs adapted to Miconia or Bellucia, the apex
fruitflylarvae(Table2),theNotiobiaspeciesdo sizeofthefig-seedfeedersdifferedsignificantly
notshowaclearcorrelationbetweenapexgrowth (Table2,numbersinItalics).Thisresultcouldbe
and theability to develop feeding only on Dro- interpreted as a clear niche separation – except
ENTOMOL.FENNICAVol.17 (cid:127) Carabidlarvaefeedingontreeseeds 339
thattheseedsofthedifferentfigspeciesintheir AbhandlungenderNaturwissenschaftlichenVereini-
habitatareallinthesamesizerange(Fig.4),and gungHamburg(NF)31.164pp.
Borcherding,R.,Paarmann,W.,NyawaS.B.&Bolte,H.
bothspeciesreproduceatfruitfallsofthesame
2000:Howtobeafigbeetle?Observationsofground
Ficusspecies(Paarmannetal.2001).Feedingex-
beetles(Col.,Carabidae)associatedwithfruitfallsina
periments(Paarmann2002)showedthatseedsof rainforestofBorneo.—Ecotropica6:169–180.
F.guianensisandF.donell-smithiiareofsimilar Erwin,T.L.1979:Thoughtsontheevolutionaryhistoryof
high value to both beetle species. Ficus sub- groundbeetles:hypothesesgeneratedfromcompara-
tivefaunalanalysesoflowlandforestsitesintemper-
apiculataappearstobeabetterhostspeciesforN.
ateandtropicalregions.—In:Erwin,T.L.,Ball,G.E.,
flavicinctusthanforN.pseudolimbipennis,while
Whitehead, D. R. & Halpern, A. L. (eds.), Carabid
F.hebetifoliarepresentstheopposite.Thesedif- beetles:theirevolution,naturalhistoryandclassifica-
ferences may be caused by differences in shell tion:539–587.Dr.W.Junk,TheHague.635pp.
hardness.Thenecessityforthebiggerspeciesto Kowalski, S. 2004: Potentielle Wirtsbäume des samen-
fressenden Laufkäfers Trichotichnus storeyi (Col.,
open a comparable high number of fig seeds
Carabidae)imaustralischenRegenwald.—Diploma
(121.1%morethanN.flavicinctus)couldleadto
thesis, Fachhochschule Hildesheim, Holzminden,
much stronger abrasions, followed by a loss of Göttingen.70pp.
functionality(compareFig.2).Thehigherscore Paarmann, W., Adis, J., Stork, N. E., Gutzmann, B.,
valuesN.pseudolimpipennisreached,iffedwith Stumpe,P.,Staritz,B.,Bolte,H.,Küppers,S.,Holz-
kamp,K.;Niers,C.&daFonseca,C.R.V.2001:The
theseedsofF.hebetifolia,maybetheresultofa
structureofgroundbeetleassembladges(Coleoptera:
higherpowerinthebiggerfirstlarvalinstarcom-
Carabidae)atfigfruitfalls(Moraceae)inaterrafirme
pared to the small first instar larvae of N. rainforestnearManaus(Brazil).—JournalofTropi-
flavicinctus. calEcology17:549–561.
Paarmann, W., Gutzmann, B., Stumpe, P., Bolte, H.,
Küppers.S.,Adis.J.,Stork,N.E.&daFonseca,C.R.
Acknowledgements. B. Bleher, N. Farwig (Johannes-
V.2002:Thestructureofgroundbeetleassemblages
Gutenberg-UniversitätMainz)andJ.Holstein(Staatliches
(Coleoptera: Carabidae) at fruit falls of Melasto-
Museum für Naturkunde, Stuttgart) collected living
mataceaetreesinaBrazilianterrafirmerainforest.—
ground beetle specimens in Kenya and organized their
Biotropica34:368–375.
transporttoGermany.Thephotosweretakenunderthead-
Paarmann,W.2002:Determinationoffoodqualityforsuc-
viceandwiththeequipmentofM.StukeandK.Müller
cessfuldevelopmentofseed-feedinggroundbeetles
(Max-Planck-Institut für Biophysikalische Chemie,
(Col.,Carabidae)fromtropicallowlandrainforests.
Göttingen).
—In:Szyszko,J.etal.(eds.):HowtoprotectorWhat
weknowaboutcarabidbeetles:45–57.Agricultural
UniversityPress,Warszawa.378pp.
Paarmann, W., Adis, J., Stork, N. E., Stumpe, P., Gutz-
References mann,B.&Holzkamp.K.2003:Findingthehosttree
species of Notiobia nebrioides Perty (Coleoptera,
Carabidae),amemberoftheseed-feedingguildatfruit
Arndt,E.&Kirmse,S.2002:Adaptationtoseed-feeding falls in Amazonian non-inundated lowland rain-
ingroundbeetles(Coleoptera:Carabidae:Harpalini) forests.—JournalofNaturalHistory37:839–844.
ofSouthVenezuela.—StudiesonNeotropicalFauna ZettoBrandmayr,T.,Giglio,A.,Marano,I.&Brandmayr,
andEnvironment31:205–216. P.1998:Morphofunctionalandecologicalfeaturesin
Beutel,R.1997:(cid:1)berPhylogeneseundEvolutionderCo- carabid(Coleoptera)larvae.—AttiMuseoRegionale
leoptera (Insecta), insbesondere der Adephaga. — diScienzeNaturali,Torino:449–490.