1\ Kf sl

o l ) ) s b

L:$R-\l'Y

Research Paper 3l1977

i ! : - - - ; ' i

' r t ' " r ' i J \

_ t' . , ! ' ' l

I

FORESTS DEPARTIUIENTOF WESTERN AUSTRALIA

FOOD REQUIREMENTS OF MARDO(Waterhouse) AND

THE EFFECT OF FIRE ON MARDO

ABUNDANCE

by

R. HINDMARSH and J. D. MAJER *

SUMMARYThe effect of prescribed burning on the

availability and abundance of forest invertebrateshas been investigated to determine whether thechanged status of food supply infLuences theabundance of the nardo (Antechinus flg"'ilgt)following fire . AnaLysiTffi-ach conGntssuggests that the mardo is an opportunisticfeeder; the numericaLly predoninant invertebratesAraneae, Blattodea, Coleoptera and Hymenoptera arethe principal food sources. The habitats of theseinvertebrates indicate that the mardo feeds mainlyin the forest l i t ter layer.

The investigation revealed no correlat ionbetween the abundance of the mardo and that of theforest invertebrates during the post-firesuccession, suggesting that food avai labiLity isnot a major factor in mardo population changes.Other factors which might cause the obsenredvariations in rnardo abundance are briefly considered.

* Department of Biology, Western AustralianInst i tu te of Technology, Bent ley, W.A. 61A2.

a@c

INTRODUCTION

The mardo (Antechinus flaviPe: -(Waterhouse) ) is Fevalent in the forestsof southwest Western Austral ia. I t ispredominantLy insectivorous althoughLarrion is also eaten (Mar1.ow, 1 961 ) .0bservation has revealed that mardoabundance is considerably reducedfoLLowing wild or prescribed f ires in

the jarrah and karri forests. As thepost-f ire succession proceeds there is

; delayed recovery in the nunibers of

mardo compared with that of othersympatric mammals. The abundance of

mardos appears to be related to the t ime

lapse since the last burn (Christensen

& K inber , 1975) .

A similar Phenomenon has beenreported for A. swainsonii (ltlaterhouse)

and A. stuartii Macleay where no recovery

of nffie-burn levels wasobserved in the two years following the

1972 Nadgee wildf ire (Newsone et 4.,1 9 7 5 ) .

This studY examines the foodpreferences of the mardo in karri forests,

and the possibi l i ty that food avai labi l i ty

is the l imit ing factor in the post-f ire

recovery of the mardo PoPulation.

Food preferences have been el"ucidated

by examining the gut contents of mardos

trapped by the Forests Department inpfoll of, varying fire histories in the

Pemberton area. An attempt is made to

correlate these data with data

concerning the availability and abundance

of invertebrates trapped from similarforest areas so that an understanding of

the post-fire mardo population density

may be gained.

METHOD

Mardo collectionsTtre animals us ed for the gut content

analysis were trapped by officers of theForests Department in the Manj inup-Pernb erton karri forests in May 197 4 'Nrrnber and site data are sunmarized in

Table 1 .

Stomach content analYsisFrom 37 formalin-preserved mardos the

stomach and duodenum were removed then

weighed, slit open and the contents washed

int6 a petri aisfr with 70eo alcohol ' Food

fragnents were identified and scored using

a sfereo microscoPe, although because the

food was well nasticated, the i tems could

not be preciselY quanti f ied.

Plot selectionI t was not Possible to sanPLe the

original plots mentioned in Table 1 'because their logging and f ire histories

had changed since-the mardos were captured,(May 1974. Four sinilar areas of karri

forest which exhibited varying stages in

the f ire succession were selected for

study in May 1976. Detai ls are sunmarized

in Tab le 2 ,

Vegetation dataPlot vegetation was surveyed in May

1976 by estinating the Percentagevegetation cover for each stratumvert ical ly above two 100 m transectsacross the longitudinal axis of each plot '

I 'his method is a modification devised. by

lvluir (1977) of Beardr s ( 1 975) key. The

doninant vegetation species were taken to

be those with the highest percentage cover

in each stratum over both transects '

The volume of leaf l i t ter was

estimated by measuring the depth of litter

TABLE 1

site data for mardo colLections. Al l si tes areforested with karri and some marri

Locat ion Map reference Time s ince burn(vears)

Number of mardoscol lected

House Brook RoadPine Creek RoadCurd RoadHenwood RoadLocality unknown

1 15o sg t E . s4o- 24 t sl r s o s 1 , E : s 4 ? 1 8 ' s1 1 5 0 i e ' E : s 4 o g 1 ' s1 1 5 0 i s ' E , s 4 o 1 6 ' s

405

N . A .N . A .N . A .

1 81576

N.A. data not avai lab le

Location Map reference Time since burn(vears)

House Brook RoadBarker RoadBarker RoadCrowea Road

i l so s9 f E . s4o 24 t sl l s o s 8 r E . s 4 o s l t s1 1 5 0 5 8 r E : s 4 o s 2 t s1 160 ozt E', s|o szt s

> 4 0> 2 0

50 . 5

TABLE 2

Site data for invertebrate collections. A11 sites areforested with karri and some marri

at 50 cm intervals aLong either side ofeach 100 n transect. T'he volume of l i t terper hectare was then calculated using thevalue for the mean depth.

The amount of dead wood was estimatedusing three size categories: 1-10 ci l ,10-50 cm and >30 crn dianeter. fiieoccurrence of wood was scored along a50 cm strip on either side of each 1 00 mtransect and an index of wood aburdanceestimated.

Invertebrate samplingFour conplenentary sanpling urethods

were empLoyed to obtain a reasonablycomplete census of relative invertebrateabundance in each plot.

(1) The epigaeic far:na was sampled byplacing pitfalL traps at 1 0 n intervalsalong each 100 m transect. Traps consistedof glass jars with a 4 cm mouth diametercontaining 10 nl of 70eo alcohoL . Trappingcontinued for 72 hours.

(2) The litter fauna was sarnpLed bytaking three 0. 1 m2 samples oe litterrandomly fron along each transect.Sanples were hand-sorted to remove largeanimals. SmalLer organisns were removedby placing the litter in a BerLese funnelin which the temperature was slowLyincreased fron roon temperature to 58oCover a period of 24 hours.

(3) Four shnrbs were randonly selectedalong each transect and the invertebratessarnpled using a 1 mZ beating tray.Sanpl.es were removed from the tray withan aspirator and pLaced in 70eo alcoholfor subsequent sorting.

(4) Hand searches were perforrned tosupplenent these three sampling methodsand to ensure that additional

rnicrohabitats, such as logs, were alsosampled. Tbo persons collected allinvertebrates along each transect for aperiod of 15 minutes.

RESULTSThe raw stomach content data of the

mardo are sunmarised in TabLe 3. The itemsincluded a wide variety of invertebrates ofwhich Araneae, Blattodea, Coleoptera andHymenoptera (excLuding ants) were the lnosta-bundant. One bird and one reptile liiltbwere found, supporting MarlowIs (1961)comment that carrion may be eaten. No plantmateriaL was found.

Ttre vegetation data for the four plotssurveyed are sr:rmnrized in TabLe 4.

TABLE 4

Surmary of vegetation datain study plots

Legend'l

. Trees2. MaLlee forrn3. Shnrbs4. Grasses and herbs5. Sedges6 . Ferns, mosses etc .

A. Dense (70-100eoB. Mid-dense (30- 70YoC. Sparse ( 0- 30%

coverage)r r

)' )

Location Timeburn

s1nce(years)

Vegetationsunmary

House BrookRoadBarker RoadBarker RoadCrowea Road

> 4 0> 2 0

50 . 5

A1/C3/C4/Cs/c6B1/C3/C4/CS/C6B1/C3/84/Cs/C6B1 /B,4/Cs/C0

TABLE 3

Stomach contents and biological data of mardos

-

N o . Sex Preservedweight

(e)

Stomachweight

(e)

Bodylength

( cm)

Stomach contents

29 1 0 . 0 070F 4 DermapLera, Blattodea, Coleoptera

Hymenoptera (WasP?)

AraneaeBlattodea, Araneae, ColeoPteraHeniptera ?Blattodea, HYmenoPtera (Bee?)

Homoptera (Fulgoroidea 6 nYmPhs)

CurculionidaeChi 1 opoda

Araneae

Araneae, Cutculionidae, Blattodea

Araneae, BlattodeaAraneae, DermaPtera, ColeoPteraB LattodeaColeoptera, Araneae, Blattodea?

Coleoptera' AraneaeHomoptera, ColeoPtera

Araneae

Coleoptera (Nit iduLidae 4+) ,Symphyta larva""i t"t", Coleoptera (Nit idul idae)

Other ColeoPteraBirds foot and feathers,ColeoPtera

Coleopt era, (Nitidulidae)

B I attodeaAraneaeCurculionidae, HeteroPteraDermaptera' Chilopoda, Blattodea'

AraneaeB I attodeaAraneae, Blattodea, DiPloPodaHymenoptera (Ant or WasP) - --siaphyiinidae, Li zard leg (Skink)

Araireae, DiPtera (Mosquito?)

B lattodea

House Brook Rd.

Pine Creek Rd'

Curd Road

Henwood Road

Locali tyunknown

1 1 8

1 1 9120121

1 s 61571 5 8

1 5 91 8 61871 8 81 8 91 9 01 9 11921 9 51281 1 4

47

1 5 1

152

1 5 3

I 1 s s1 1 7 2I1 2 21 , 2 3l s gl a oIl s ol s 7Il s tIII z t olI zrsI| 2 1 7

| 228

| 2 2 1Il zzo

1 8 . 0 526.902 2 . 6 0

2 9 . 12 8 . 9 129.87

2 3 . 9 034.7031 .8929.3550 . 002 8 . 3 52 1 . 2 920.602 3 . 2 23 7 . 5 04 1 . 3 5

7 . 7 59 . 5 09 . 2 5

9 . 7 59 . 5 0

1 0 . 5 0

9 . 0 01 0 . 0 01 1 . 0 01 0 . 7 51 2 . 0 01 0 . 7 59 . 7 59 . 5 09 . 5 0

1 0 , 2 51 0 . 7 5

1 0 . 0 0

7 . 7 5

1 0 . 0 0

1 0 . 5 0

001

FMF

081

101

MMF

384

07370562867

11101101000

FMFMMMFFFMM

2 7 2 0

2 2 . 7 5

4 3 . 0 5

45 .75

2 3 . 5 019 .37

45.7223,7 ' � |2 8 . 9 05 0 . 4 0

2 3 . 8 52 1 . 6 5

2 4 , 9 0

7 . 7 58 . 5 0

1 1 . 2 59 . 2 5

1 0 . 2 51 1 . 5 0

9 . 7 51 0 . 0 0

9 . 7 5

4

6

4

6

0

0

F

M

M

FF

43

00

1373

1001

MFFM

74

00

FF

F

Myriapoda (DiPloPoda or

AraneaeBLattodea, HeteroPtera?,B 1 attodea

Chilopoda) ,

Coleopt era?

Scarabaeidae, . Araneae

Coleoptera (Nit iduridae), Araneae,

0ther ColeoPtera' HYmenoPteraBlattodea, Araneae (2) , ColeoPtera

Scorpionida?, Chilopoda? Heteroptera

Aran-eae , Larva (LePidoPtera?) ,Homoptera.

4

Marked inter-plot differencesoccurred in the understorey strata.The most recentlY burnt P1'ot hadno shnrb laYer, a herb and grasslayer of medium densitY andsparsely distributed sedges andferns. The Plot burnt five Yearspreviously was characterised bY amore dense herb and shnrb laYerof I fireweedf such as Croweaanqustifolia Sm. and Affi-

@r. The_ffiweedtffiffitey was considerabl'Y Lessdense twenty years after burningbut the Lower tree stratum wasmore dense. This trend continuedin the site burnt over fortYyears earl ier.

The volume of leaf litter wasdirect ly related to the firehistories of the sites; a gradualaccumuLation of litter coincidedwith the recovery from burning oneach si te (Fig. 1). This t rendwas also fotrnd with the dead wood

N I - l0 cm diometer

G-c>-E

lr,

fJ

oEuJFF=

oNUMBERFIGURE 1 :l i t ter in

5OF

20 40YEARS AFTER BURN

Ilstination of leafthe four studY areas.

llJ(Jz:)oz3GO

ooo]

ILoxuroz

NUMBER OF YEARS AFTER BURN

Eg l0 - 30 cm diometer

Estimation o F dead wood in

5

fl >30 cm diometer

the fortr study areas.

100

FIGURE 2:

although there was a substantial quantityof the smalLer size range of dead wood inthe recen t l y bu rn t p lo t (F ig . 2 ) : th i sprobably represents recentLy fallenfire-danaged timber.

The four sampling methods determiningthe abundance and distribution ofinvertebrates in the four study areas areset out in Appendixes l-IV.

440

360

320

240

r60

120

80

Ttre resuLts show no general trend ofinvertebrate abr.mdance foLlowing fire.Amphipoda, Isoptera, Heteroptera andDiptera (Larvae) were not found in thesampLes from the recently burnt atea,suggesting that theY are the nostfire-sensit ive. The Pseudoscorpionida,Chilopoda, Blattodea, DermaPtera'Hymenoptera (excluding ants),Staphylinidae and Curculionidae increased

ooo-i5

oo

3.9Eou-

o? '?t rb

6 9o-=o o

.i3 -goo-oq,

EU

oooco

o-ooo.o. go

rf,FoJo-

=ou,u-

olLJJo-=a.ntuI

\u,colrlFE,IJ.J

zll-

od,BJ6=f,z

t , r ' LlJ.t

k ?# U qluE :t"= O2 * - l o; on o: t 1 5H == z 2 0z

Glb

o. . C ^

3 go hb 5 gE e !o o c

E € E €b . - b f I

> . o - o YT o ( J f

< 9

?oto: os € so ' F O

_ !

& ; eo - o - Eo o hu . ; d

o* g F r

s g r E !- E € n 5 o € b

EgE TEE Td o A A 9 = a

oT Eo r o

/ r Lv U g o

o 2 0 - ob i 9 pt f o =. F 9 6 -L r L I C )

H v ' , q o g € €g o o ! o 6 6& E € & E + &f Bs 5S€'.B

FREQUENCY OF FOOD ITEMS

FIGURE 3: Comparison of invertebrates obtained from plots unburnt for 2A or 40years (upper part of graph) with the frequency of food itens in nardo stomachs( lower part of graph) .

6

-l

in abr.rndance shortly after the burn"Other groups were reduced in abundancefollowing the fire but increased againas the post-f ire succession progressed.These included Opil iones, fsopoda,Diplopoda, Collenbola and .Lepidoptera(larvae). A finaL group showed a peak inthe plot burnt five years previously andthen decLined in abundance; incl.uded inthis group were the Scorpionida, Acarina,Araneae, Anphipoda, Gryllidae, Homoptera,Fo:rnicidae, Carabidae and Scarabaeidae.The significance of these interplotdifferences was tested by the chi-squarenethod and the results are shown inAppendix V.

A conparison of invertebrate frequencyin the nardo stomachs and that of theinvertebrates collected from the unburnt(for 20 and 40 years) plots is shown inFigure 3. A Spearmants rank correlationcoefficient was calculated for the twosets of ranked data. The obtained valueof r , was 0.358. By test ing th is va lueusing Students t it nay be concLuded thatthese variables are positively correlated( p = 0 . 0 5 ) .

DISCUSSION

Although the entire stomach contentsof the mardo were not identified, theidenti ty of a suff icientLy Largeproportion was determined to allow a fairdegree of certainty in interpreting themardot s feeding preferences.

Many of the species eaten by the nardowere collected on the ground or in thelitter layer, which indicates that itfeeds mainly on invertebrates from theforest floor. Wakefield and Warneke(1967) note that the characterist ics ofmardo feet are well adapted for scansorialact ivi ty so the possibit i ty of arborealfeeding cannot be eliminated. It isLikely that the fulgorid nymphs eaten byspecimen 158 (Table 5) were obtained fromthe shrub layer.

The positive correlation betweeninvertebrates present and invertebratesconsumed suggests that the mardo is anopportr:nistic feeder. This correspondswith Kitchener and Chapmanrs (1977)observations on Phascoga.lg caluraGould which feeds on most invertebrateswhich are present. Tho points shouLdalso be noted when considering thisopportunism. One is that there appearsto be a nininum size Limit for food

consuned. No CoLlembola were found inthe gut despite their high abundancein the environment (Fig. 3) . The smallestfood i terns, Nit idul idae beetles , wereapproximately 3 nn long suggesting thatthere is a si ze f.init be low whi ch it isnot economical to expend energy onobtaining a food item. Secondly, it isnoteworthy that ants were found in onlytwo mardos (FiS. 3) despite their highabundance in the forest ecosystem.Presunably the defence rnechanisurs of antsare relatively effective against thispredator.

The trends in the invertebrate samplessupport Springett I s (1976) findings thatfire has a marked effect on invertebrateabundance. It is possible that the suddenreduction of invertebrates following firecontributes to the accompanying depletionof mardo populations. However, Christensenand Kinber (1975) note that on dry uplandsit takes more than 20 years for the mardopopulat ion to return to pre-f ire levels.Since many of the invertebrate groups havereturned to their forrner popuLation levelswithin five years (Appendixes I-IV) foodavailability does not seem to be theprincipal limiting factor on mardoregeneration. Furthermore, since themardo is an opportr.rristic feeder, it wouldbe able to switch to feeding on those preyitens which are abundant following fire.

fiie results indicate that otherfactors which nay Limit the regenerationof the mardo population should beinvestigated. Table 4 shows changes inthe forest vegetation structure foLlowingfire. The fire-induced clearing mayrender the nardo more conspicuous topredators or place it r:nder increasedheat or water stress. However, the denseffireweedt regrowth wouLd probably reducethese problems quite rapidLy, so thisseems unl ikeLy to be a l imit ing factor.

The data presented here suggest thatthe factors monitored cannot be solelyresponsible for the long term reductionin mardo nunbers foLlowing fire. It ispossible that the changes in thevertebrate fauna distribution andpopulation following fire (Christensenand Kinber, 1975; Newsome et a7., 1975)alter the competitive balanFbetweenspecies which in turn effect and supporta decreased mardo population forconsiderable periods. One otherpossibiLity is that fire may produce along term disnrption of the mardobreeding pattern.

ACKNOWLEDGEMENTS

We wouLd like to thank David and JoanOsborne and Susan Posmus for theirassistance in the field and in sortingthe sanples. We are also particularl'ygrateful to Peter Kimber and PerChristensen of the Forests Department,Manj irnup for suggesting the proj ect andfor providing facil i t ies'while this workwas carried out.

REFERENCESBEARD, J. S . ( 1975) . fite vegetation survey

of Western AustraLia. Vegetatio3 0 , 1 7 9 - 1 8 7 .

CI IRISTENSEN, P.C. and KIMBER, P.C. (1975) .Effect of prescribed burning on thefLora and fauna of south-west AustraLianforests. f Managing Terrestrial'Ecosystemst eds. Kikkawa, J . and Nix,H.A. Proceedings of the EcologicalSociety of Austra l ia 9, 85- 1 06.

KITCFIENER, D.J. and CHAPMAN, A. C. (1977) ,The mammals of Bendering Reserve.In Biological survey of the WesternAustralian wheatbelt. Part II I .Records of the ltrestern Austral'ianMuseum Supplenent No. 4 ( in press). '

MARI0W, B.J. (1 961 ) . ReProductivebehaviour of the narsuPial mouse'Antechinus flavipes (Waterhouse)

e develoPment ofthe pouch young. Australian Journalof ZooLogy 9, 203-218.

MUIR, B"G. (1977). Vegetation and habitatof Bendering Reserve. In Biologicals.urvey of the Western AustraLianwheatbelt. Part II. Records of theWestern Australian Museum SupplementN o . 3 "

NEWSOI,IE, A.E., McILROY , J. and CATLING, P.(1 975) . The effects of an extensivewildfire on popuLations of twentygrotrnd vertebrates in south-eastAustralia. tManaging TerrestrialEcosystemsr eds. Kikkawa, J. and Nix,H.A. Proceedings of the EcologicalSociety of Austral ia 9, 107-123.

SPRINGETT, J.A. (1976). The effect ofprescribed burning on the soiL faunaand on litter deconposition in WesternAustralian forests . AustralianJournal of Ecolog)r 1 , 77-82.

WAKEFIELD, N. A. and WARNEKE, R.M. (1967) .Some revision in Antechinus(Marsupialia) 2lffin Naturalist84 , 69 -99 .

Total nrrurber of

APPENDIX 1

invertebrates trapped by the Pitfall trap nethod

Class Order Family Years after burn

0 5 20 40

Arachnida

Crustacea

DiplopodaChilopodaInsecta

0LigochaetaGastropoda

ScorpionidaPseudoscorpionidaOpil.ionesAcarinaAraneaeAnphipodaIsopoda

ColLenbolaOrthoptera: GryLl.idaeBlattodeaDentapteraI sopteraHonoPteraHeteropteraLepidoptera (Lawae)DipteraDiptera (Lawae)Hyurenoptera : Fornicidae

OthersColeoptera: Staphylinidae

ScolytidaeCurculionidaeCarabidaeScarabaeidaeOthersLarvae

01 045

1 00011

11721

2 10100

1 00

1 31

3709

1 11 544

302

2l ,2

2420t 31 101

6144

1 90520I0

558

2708

1 92 10001

0322

l 61

. 130

4931

1 4000262

2S6

1 401

1 10

1 4511

0251

1 8260q

050120203

1 3374401319100

1

, I4aL

APPENDIX II

Total. nurnbei of invertebrates cot lected from litter sanpLes

CIass 0rder Fq$ily Years after burn

50 20 40

Arachnida

Crustacea

DiplopodaChilopodaInsecta

OligochaetaGastropoda

ScorpionidaPseudi scorpionidaOpilionesAcarinaAraneaeArnphipodaIsopoda

Col lenbol aOrthoptera: GrYlLidaeB 1 attodeaDermapteraIsopteraHomopteraHeteropteraLepidoptera (Larvae)DipteraDiptera (Larvae)Hymenoptera : Formicidae

OthersColeoptera: StaPhYlinidae

Scol.ytidaeCurculionidaeCarabidaeScarab aeidaeOthersLarvae

000050200700000031020000001000

201

1 11 92

1 144

5116000220311202404800

0149

1 49

' 4

2 17

1 2035015

1 014

1 50201006

1 401

3 l8 lo l

1 6 |2 l

1 8 |41 |3 l0 lo ls l0 lo l2 l1 lo l0031000201310

1 0

APPENDIX I I I

Total number of invertebrates collected by beating shrubs

Years after burnOrder Family

Arachnida

Cnrstacea

DiplopodaChi lopodaInsecta

Oligoch aetaGastropoda

0005

620

11210

1280400

1 10

1 31 003

1 411

1 6002000

000

1 78 2

0000

1 1 1110030

1 11 401

1 500400

6 3000

ScorpionidaPs eudoscorpionidaOpil.ionesAcarinaAraneaeAnphipodaI sopoda

CollembolaOrthoptera: GrylLidaeBlattodeaDermapteraI sopteraHonopteraHeteropteraLepidoptera (Lawae)DipteraDiptera ([arvae)Hymenoptera: Formicidae

OthersColeoptera: StaphYlinidae

Sco lytidaeCurculionidaeCarabidaeScarabaeidaeOthersLarvae

0001

270000

550000502

1 000

3730800

2 7000

0003

1020

2S00

1 200002354039101004000

1 1

APPENDIX'IV

Total nr-rmber of invertebrates col.Lected by the unit search nethod

Class Order Fanil.y Years after burn

0 5 20 40

Arachnida

Crtrstacea

DiplopodaChilopodaIns ecta

OL igochaetaGastropoda

ScorpionidaPseudoscorpionidaOpi lionesAcarinaAraneaeAnphipodaIsopoda

CoL lembolaOrthoptera: GryllidaeB I attodeaDermapteraI s opteraHomopteraHeteropteraLepidoptera (Larvae)DipteraDiptera (Larvae)Hymenoptera : Forrnicidae

OthersCoLeoptera: Staphylinidae

ScoLytidaeCurcuLionidaeCarabidaeScarabaeidaeOthersLarvae

110190

1 6941 700

2 3201000030002100020

1002907

s642060

1 61 50100

980003000010

1001

1 50,3

614

. 00400000009010

1 1012109

0100

1 11

1 6163

2201000000000000000100

1 2

Conparison of pooled,foyr sample plots

invertebrate collections fron theusing the chi- square urethod

Class -0rder Farnily x2

Arachnida

Cnrstacea

DiplopodaChiLopodafnsecta

+0Ligochaeta+Gastropoda

+Scorpionida+PseudoscorpionidaOpi lionesAcarinaAraneaeAmphipodaI sopoda

CoL l embol a+Orthoptera: GryllidaeB lattodeaDe:maptera

+IsopteraHomoptera

+HeteropteraLepidoptera (Larvae)Diptera

+Diptera LanraeHynenoptera : Formicidae

0thersColeoptera: Staphylinidae

+ScolytidaeCurculionidaeCarabidaeScarabaeidaeOthers

+Lanrae

relsse nl1 7 . 3 8 *

1 6 . 5 0 n s

9 . 5 84 1 , 2 248 .0016.67

243.362 1 2 . 1 2

7 . 8 2v 8 . 3 1

1 5 . 2 33 . 7 9

1 8 1 . 3 01 1 . 0 950. 55

5 . 8 01 3 . 2 33 2 . 3 2

1 7 1 . 0 2

ns* ** *

ns* ** *

ns* *

nsns

* *

ns* *

nsns* ** *

)X- (criticaL) for df=9 at

+ No test done since e >5

59o19o

and

leveL = 76.92 *leveL = 21 .67 **

T >20