DIVERSITY OF ORTHOPTERA FROM BORNEAN LOWLAND RAIN FORESTTREES

O Society fbr Tropical Ecologv DIVERSITY FROMBORNEAN OF ORTHOPTERA LOWLAND RAINFORESTTREES Andreas Florenl, Klaus Riede2& Sigfrid Ingrisch2 lTheodor-Boveri-lnstitute, Departmentof AnimalEcologyand TropicalBiology,Biozentrum, Am Hubland,D-97074Wuerzburg,Germany 2 AlexanderKoenigZoologicalResearchlnstituteand Museumof Zoology,Adenauerallee160, 53113Bonn,Germany Absnact.The composition and colonization dynamics of arboreal Orthopteran communities were studied in a SE-Asian mixed dipterocarp lowland rain forest by insecticidal fogging, re-fogging after different periods oftime, and by srem eclecrors during different times of the year. In total, 2324 Orthoptera were collected by fogging, of which 87 .3o/o were nymphs. Imagines were sorted to 49 morphospecies, ofwhich 47 (96ok) were new to science.Almost 500/owere singlerons, and only three species, all belonging to the omnivorous subfamily Mogoplistinae (gents Ornebius), occurred with more than I 0 adult individuals. Orthoptera seemed to be randomly distributed in the canopy without showing any host-tree-specific adaptation. This is also indicated by the re-foggings, in which Orthoptera colonized the trees, according to the relative proportions ofsubfamilies and feeding guilds occurring in the canopy. No pioneer or climu speciescould be distinguished. ln the stem eclectors787 specimenswere caught, ofwhich 98.20lowere nymphs, mostly first and second instar, indicating migration into the canopy after egg development in the soil or lower forest strata. In contrast to the numerous nymphs of hemimetabolous taxa, lessmobile arthropods, mainly larvae of holometabolous groups, were almost completely lacking in the canopy. This is probably due to the high predation pressureof the ants which influence communiry composirion. Arepted 9 July 2001. Kry words: Ant predation, canopy fogging, colonization dynamics, diuersity, hemimetaboly, stem eclectors. INTRODUCTION The existenceofan extremely diverse arthropod fauna in the canopy oftropical lowland rain forest, as first suggestedby Roberts (1973) and Erwin (1982), has since been confirmed by various investigations(references,e.g., in Stork et dl. 1997, Linsenmair et al., in press),raising the question ofhow such high diversiry is maintained (e.g., Beaver 1979, Huston 1994, Linsenmair 1995, Pianka 1996, Price et al. 1995, distinguished: the Caelifera ("grasshoppers")which are diurnal and herbivorous, showing all varietiesof the specialist-generalistcontinuum (Rowell 1978), and "karydids" "crickets") and which the Ensilera (mainly are nocturnal and comprise a wider range of feeding habits, including herbivores, predators, and omnivores. To date, most studies have locused on Neotropical Caelifera, which were collected either by Fogging(Roberts I 973) or tree-felling,and netting of grasshoppers at ground level (Descamps1978, 1981). Rosenzweig1995). This is the locus of our project in SE Asia that aims at understanding the mechanisms Those studies revealeda species-richcanopy fauna and structuring arboreal arthropod communities. In this g a v e r h e f i r s r i n s i g h r si n t o c o m m u n i r y s t r u c r u r e .r e source partitioning (Amdddgnato 1997), and ethology paper we consider the Orthoptera, which usually account for berween I and 3 o/ooF the specimens in (Riede 1987). Although the Ensiferan launa oftrees tree communities (Stork 1991, Floren & Linsenmair is generally much more diverse than the Caeliferan 1997,'S7agner 1997). In contrast to previously ana- launa, the arboreal Ensilera remain largely unstudied. lJzed insect orders from the canopy, Orthoptera are Here we present for the first time an analysis of hemimetabolous. Their nymphs are usually highly complete Orthopteran communities from trees in a mobile and basicallybelong to the sameleeding guild SE-Asianlowland rain lorest,which were sampled by as the adults. Two major groups ofC)rthoptera can be fogging, re-fogging, and stem eclectors. \7e describe e-mail: fl oren@biozentrum. uni-wuerzburs.de cies, in terms of their diversiry and re-colonization the Orthoptera launa associatedwith three tree spe- 33 dynamics (of the primary fogged rrees) and ask how important tree trunks are as pathways [or canopy colonization. Furthermore we investigate whether one A. lagenocdrpatree was re-fogged on Four conse- c cutive mornings. differencesin life history featuresof holometabolous Stem photo-eclectors(emergencetraps) were installed on 47 tree trunks at 4 m height in order to ( I and hemimetabolous taxa have an effecr on the struc- quanti$/ arthropod immigration into the canopy along tree stems lrom the ground (these were 35 A. lagenocarpa, 4 A. subcaudata, 4 Depressaneruosa (Guttiferaceae),and 4 unidentified trees.Their bark was only slightly structured and not covered with c thalloid epiphytes.For the operational method ofthe eclector-traps see Bassetet al. (1997). The traps were c filled with 4o/o formalin solution and emptied every fwo weeks.A total of 171 sampleswas evaluared. l l, t u r e o F t r e e - d w e l l i n gc o m m u n i t i e s . MATERIALS AND METHODS Study site and samplingprocedures.Investigations were conducted in Kinabalu National Park, substation Poring Hot Spring, Sabah, Malaysia, Borneo (6"5'N, 11633' E), in a floristically rich mixed dipterocarp lowland Forest(500 to 700 m a.s.l.). The region is characterized by a rypical tropical climate with at least2000 ml precipitation per year (Kitayama 1992, Pfeiffer 1997). All investigations were done after the rainy seasons(which last from February ro May and August to end of November) during lour field stays between 1992 and 1993. An improved method of canopy logging was used Arthropod sorting and data anallsh. Orthoptera were separatedand sorted to morphospeciesby one o[ us (SI). However, most of the nymphs could be identified only to subfamily level, therefore evaluation is only based on adults. Classification lollowed the Orthoptera SpeciesFile (One & Naskrecki 1997). the Museum Koenig, Bonn or in the institutions men- individual trees. Since methodical details have been tioned in the resulting taxonomic publications. In addition, various persons kindly provided data from hand collecting and light-traps which were conrrasred were fogged For ten minutes with natural pyrerhrum with the fogging material. used as insecticide early in the morning and all ar- Shinozaki curves were calculated to comDare comm u n i t i e s a r r h e b e r a - d i v e r s i t yl e v e l ( S h i n o z a k i1 9 6 3 , thropods that had dropped into the funnels after rwo hours lollowing fogging were used in the analysis. Nineteen medium-sized trees of the canopy understory, representing three speciesfrom nvo lamilies, were fogged. These were ren Aporusa lagenocarpa,fle A. subcaudzta(both Euphorbiaceae),and foturXan- Achtziger et al. 1992). These are expected species accumulation curves and their steepnessprovides inlormation about the overall complerenessof the sampling effort (a more comprehensivediscussionis given in Floren & Linsenmair 2000). In addition, belonged to two species).A. lagenocarpawas rhe only speciesthat occurred in larger numbers in our re- Soerensentindex was used to measuresimilariry in speciescomposition berween communities (Magurran 1988). For a benveen comparison, the fogging data were standardized lor a crown projection of I m2 search area of which sufficient available individuals and a leafcover oF 100o/o.Bonferroni's correction was toph/lum afine (Polygalaceae) (we recognized only during the investigations that the Euphorbiaceae trees were lound that were not covered with lianas. All trees were smaller than 30 meters in height and similar with respectto insect habitat conditions, for example concerning the existenceofstem cavities, dead wood, and detritus accumulations, etc.; epiphytes were almost completely lacking (Floren & Linsenmair 1998a). Only one l. lagenocarpaflowered during one samp- ling period, however, without a noticeable elfect on the speciesabundance patterns. All trees showed continuous leaf flushes of low intensity. Ten trees (five A- lagenocarpa, three ,4. subcaudata, and wo X. ffinde) were re-fogged at least after six months, nvo l. subcaudata trees were fogged again alter a week, and 34 c T S c t I t c i c Specimensfrom the studies have been deposited in to sample arthropod communities associatedwith published elsewherewe give only basic information here (Floren & Linsenmair 1997,2000). The trees t used to adiust the level of significance for multiple pairwise comparisons (Lehmacher et al. 1991). RESULTS A total of 2324 individuals of Orthoptera was obtainedby fogging.The numberof specimens collected on treesfoggedfor the first time variedconsiderably berween7 and,238,representing beween0.72o/oand 3.4o/oof all foggedarthropods.The proportionof nymphs within primary fogging samplesfluctuated (nean 82.94t 19.59).This berweenB0 and 100o/o patternwasindependent ofthe month ofloggingand showedno seasonaliry. In the stemeclectors,787 spe- c € € C f t C I s T : DIV!,RSI'fY OF ORTHOPTEM FROM I]ORNEAN LO\flLANt) RAIN FORIST TREES cimens were collected, and the proportion of nymphs was significantly higher than in the fogging samples (mean 96.9 + 5.28, Mann \Whitney U-test, P< 0.001). All other taxa were represented by only a lew individuals or singletons in each sample.A remarkable ge- Table I comparesthe higher taxonomic composirion grasshoppers (Caelifera) which were represented by o[fogged Orthoptera versus those caught by eclector only 32 nymphs in the fogging samples. They occurred as single individuals or in small groups of up to lour (Acridoidea) or ten individuals (Eumasta- traps, ranked by the number of nymphs. Adults occurred only sporadically in the stem eclectors. Highest in abundance in both logging and eclector samples were the Podoscirtinae (hard-footed bush crickets). In the fogging samples,Mogoplistinae (scaly crickets) followed in second place, but had by far the highest number of adults. The differences between larval:adult proportions of the dominant Podoscirtinae and Mogoplistinae point to differences in the life cycle; however we cannot exclude the possibiliry that this result was influenced by dilferences in flight capabiliry (adult Podoscirtinae are fully winged flying species,while in Mogoplistinae the wings are reduced or absent and they are unable to fly). In the eclector traps the Mogoplistinae dropped to rank six, while the omnivorous Pteroplistinae rose ro rhe second rank. Pteroplistinaefirst-instar nymphs formed groups ranging between 7 and 25 individuals, suggesting that nymphs climbed up the stems immediately after hatching in the soil, lower vegetation, or from crevicesin the bark. In contrast, the low num- neral leature was the low percentage ofshort-horned coidea). Tetrigoidea always occurred as singlerons; three of the specimens lound in srem eclectors were adults. For the Ensifera, Thble 2 compares the number of adult speciesand their taxonomic status in the fogging samples with those collected by hand and by light-traps. A total of 127 morphospecies were difTerentiated,of which 85 were new;32o/o(4I1127) of all specieswere obtained by fogging only. Vithin the fogging samples the proportion of new species "only''497o was 989/0,while ofspecies collectedwith conventional methods were new. Only 8olo (6176) of tenigoniid and 2o/o (1144) of grylloid specieswere collected both by fogging and convenrional methods. Such a small overlap suggests the existence of a se"Canopy'' parate canopy fauna. taxa ('only fogging') which provided most new specieswere Pseudophylli- bers of Mogoplistinae nymphs in the eclectorsindi- nae, Meconematinae, Gryllacrididae, Podoscirtinae, and Mogoplistinae. Most other groups were collected predominantly by conventional merhods, indicating cate a preFerencefor the canopy throughout their life. a oreference for lower forest srrara. TABLE I . Ranksof Orthopteran groupsaccordingto the number of nymphs collectedby toggingand with stemeclectortraps.Guild assessment accordingto feedinghabits:H = herbivorous,O = omnivorous,P = predators. Fogging Rank Group I Podoscirtinae Cuild H 2 Mogoplistinae o 3 Meconematinae 4 Tiigoniinae P H? 5 Gryllacridoidea 6 Pteroplistinae o o 7 Phaneropterinae 8 Pseudophyllinae H H 9 Conocephalinae 10 Lipotactinae 11 Caelifera Total o P H Total Adults 826 545 3r8 161 171 101 84 5 7 L\) ) Stem-eclectors o/o Nymphs Adults 40 4,8 1 6 1 29,5 r0,7 34 22 r3,7 11 6,4 4 4,0 t4 16,7 12,3 7 - 2 L 66,7 786 384 284 139 160 97 70 t0 6 1 Total Adults 326 35 75 38 78 186 29 ll 1 0 2a 2324 295 2029 2 5 2 I I 3 787 t4 o/o Nymphs Rank Adults 0,6 14,3 324 30 - 3 8 2,6 76 0 , 5 r8 t _ 2 9 10 9,r 1 37.5 I o 4 5 3 2 7 8 9 5 773 35 TABLE 2. Speciesnumbers and taxonomic status of adult Ensifera collected by fogging and conventional techniques (hand collected and light-traps). Only logging Only hand & Collectedwith light sampling both methods Total New Total New Total New 2 4 5 0 0 0 0 0 0 1l 2 3 5 0 0 0 0 0 0 10 20 I6 I 4 2 12 2 1 r3 4 0 0 0 0 1 0 0 I o 4 0 0 0 0 0 0 0 1 5 TETTIGONIOIDEA Phaneropterinae Pseudophyllinae Meconematinae Mecopodinae Conocephalinae-Conocephalini Conocephalinae-Agraecini C o n o c e p h a ln i a e - C o p i p h o rni i Hexacentrinae Lipotactinae Sum o/onew species GRYI-LOIDEA Podoscirtinae Pteroplistinae Mogoplistinae Trigoniinae Nemobiinae Gryllinae Itarinae Eneopterinae Oecanthidae Myrmecophilidae Gryllotalpidae Sum t2 2 5 4 I 0 0 0 0 0 0 24 o o/onew species TOTAL 7o new species t2 2 5 4 I 0 0 0 0 0 0 24 I00o/o o 3 2 0 0 3 2 4 3 2 0 0 0 I I 2 U 0 I I I 2 19 0 I 10 53o/o 0 83o/o I I 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 41 40 98o/o with almost all speciesbeing rare as adults. The majority of all specieswere singletons (49%) and only three species (all Mogoplistinae) were lound with more than ten individuals; Ornebius marginatus represented by 67 individuals, O. flori by 48, and O. 78 38 49o/o I Total I B New 26 19 2 0 9 6 6 4 0 2 0 1 3 7 2 0 l 0 2 2 43 76 57o/o l6 4 5 4 4 2 4 I I I 2 44 l00o/o 100o/o Composition and recolonization dynamics of the Orthopteranfauna. The rank-abundance plot ofall foggings (Fig. 1) gives the rypical pattern o[species distribution known From tropical lowland rain forests, 36 48o/o 91o/o 7o new species GRYI-LACzuDIDAE I 59 6 I 0 0 7 0 0 1 28 Total l6 4 5 4 I I 2 0 0 I I 35 80% I 7 100% l00o/o 7 880/o 127 85 67o/o rubidusby 1 1 individuals. Even the more abundant speciesdid not aggregateon particr,rlartrees,occurring on almost every tree with some individuals irrespective ofthe tree species(Fig. 1). The steep increaseofthe Shinozaki curves (Fig. 2) and the low Soerensen indices (Table 3) demonstrate that speciesoverlap among individual trees was low, and that overall s a m p l i n ge f l o r t w a s n o t s u f l i c i e n tt o r e p r e s e not r e s t i mate total speciesnumber at the regional or even local ' t DlvflRsl'l Y ol- oRl HoP'lf-RA I-RoN{ BoRN!_AN l.o\(/l.ANt) RAIN FORI-_S] RI-.I-.S o o c G o c f -o Omebiusrubidus o) o J 11 13 15 17 1921 23 25 27 29 31 33 35 37 39 41 43 45 47 49 RankedSpecies FIG. 1. Rank-abundance curve plotted for the Orthoptera oFall fogs, 60 o a-uJ ll oo @ 'o6 O 3 0 oa c o 2 0 o c) o- .^ x tu 0 15 20 Samples= q 25 FIG. 2. Shinozaki curves computed for primary logged trees, lor all foggings, and lor all loggings excluding the most numerous Ornebius spp.q = number of discrete samples (fogged trees),ES(q) = expected number of speciesin q samples. 2a FI-OREN f7',41-. TABLE 3. Mean beta-diversities berween the fogged trees expressedby the Soerensen index. Standard de- when the composition of the feeding guilds benveen the various times of foggings were compared (Fig. 4b). viations in brackets. Guild classification corresponds to that ofThble I and Tlee species A. lagenocarpa (n =10) A. subcaudata (n =5) (" = 4) X. afine All trees Soerensen 0 . 1 6( 0 . 1 7 ) 0 . 1 6( 0 . r 4 ) 0.28 (0.34) 0 . 2 0( 0 . 2 1 ) distinguishes omnivores, herbivores, and predators. Only predators in the daily samples differed significantly from the other foggings (K-W-Anova, 12 = 9.39, P = 0.024), causedby an increasein Meconematidae which were the lastestnewcomers (seeFig.4). DISCUSSION Characterization of Orthopteran communities. Arthropod diversiry in tropical rainforests, and in particuscale. The Shinozaki curves are far from reaching lar in the canopy, has received much attention during an asymptote, independently of whether all fogging samples (primary- and re-foggings), only primary the last decade, but analysis has been mainly based on Coleoptera and Formicidae (e.g., Erwin 1982, foggings, or only sampleswithout the most common Stork 1991, Basset1996, Stork er al. 1997, Linsen- Ornebius spp. were included in the calculation. Adding species captured in the re-fogging expe- mair et a|.2001). Ve here present data on the Orthopteran fauna of three Bornean lowland rain forest riments always resulted in a larger number of new spe- tree species,considering both adults and nymphs. The cies, without changing the steepnessof the curve. constant high percentage ofnymphs within each tree Fig. 3 shows the standardizednumbers of Orth- crown community was independent of the month of opterans From the trees fogged lor the first time and fogging and indicates continuous reproduction in those re-fogged at various intervals of time. The num- most species.Despite quite a substantial sampling effort, relatively lew adult specimens(representingonly ber oF individuals within daily re-loggingswas significantly lower than those ofthe primary fogged trees (Mann-Whitney U-test, P = 0.007) and also than those of the re-loggings after 6 months (M-!? U-test, P -- 0.002). The number of individuals caught in the daily samplings was not significantly different from c o that obtained by weekly sampling due to an outliet o o o and the same was true for all Orthopterans collected (I' weekly and in comparison to those collected after more than six months. However, because of the lew samplesinvolved this lack of significanceshould not be taken too seriously.The proportion of nymphs was equally high in trees fogged for the first time (87o/o) and in those re-fogged after one week and after six months, but lower in samplesderived from daily retoggings (66%). For all loggings the majority of specimens captured belonged to five subfamilies(seeFig.4a), which together contributed between 760/o of all specimens o o .N E o c o a t:f ;l L ol 41 2' 0i il+ I r c t I t I -Tt F"S 1 ' - Daily s T weFt<ty >6 month after six months. Only the Podoscirtinaeand Meco- FIG. 3. Mean numbers oFOrthopteran individuals calculated for a crown projection of 1m2 and a leaf cover of 1000/o.Primary foggings (n = l9) are com- nematinae showed significant variation in numbers under different fogging regimes (for the Podoscirtinae: pared with re-foggings after different intervals of time. Daily: Four re-foggings of a single tree on four con- K-W-Anova, y2 = 11.34, P = 0.010; for the Meco'Unknown' nematinae: X2 = 9.46, P = 0.024) (the secutive days; l7eekly: re-foggings offour trees after one week; > 6 month: re-foggings often Aporusa-trees group was not considered).A similar pattern emerged after six months. in the daily samples and 90o/o in the samples taken 38 F F t a I I I DIVERSITY OF ORI'HoP'I't,RA FII,OI\{BORNI.-AN I,OVI,AND RAIN FORI.-S'I1'REES not been sampled from the ground (Riede unpublished), and were new to science(lngrisch 1998), thus indicating that they are confined to the canopy. Re- o cent studies reveal that at least some speciesof the 6 Podoscirtinae do not need to go to the soil in any part o o oftheir life cycle, lor example for egg deposition (In- ! _c s oTi@niinae sGryllilddoidea a0EE Fog 1 Daily Weekly > 6 month grisch I 997). Howeveq for most speciesobservations on life habits are lacking. The lact that we also collected juveniles oFall instarsofthese speciesby fogging suggest they are really canopy specialists. Our data do not allow us to draw any conclusions Orthabout the existenceof a tree-species-specific opteran community. However, host tree adaptation might emerge if sampling effort were significantly extended and if nymphs, which belong to the same feeding guilds as adults, could be included in the ana- a o lysis. However, an adequate consideration of nymphs would require comprehensive taxonomic and autecological studies.A preliminary analysisof a further ! s Fog 1 Daily Weekly > 6 month 39 foggings in the mature forests, which include ten additional tree species,does not show a change in the pattern describedabove (Floren unpublished). Trees FIG. 4. Relative proportion of a) Orthopteran subfamily- and b) leeding guild composition in foggings in general seem to provide similarly lavorable conditions for omnivorous species,and no differences in habitats- like the presenceofepiphylls, stem cavities, carried out after diflerent periods of time. Labeling offogging categoriesas in Fig. 3. aggregationso[ detritus, etc. - were lound (Floren & Linsenmair 1998a). The failure to identify a treespecies-specific Orthopteran communiry corresponds 13o/oof all specimens)were collected and assignedto morphospecies. Our sampling revealedsurprising species richness:all fogged Grylloidea and 91% of Tet- with the findings lor other taxa, such as Coleoptera (e.g., Floren & Linsenmair 1998b, lTagner 1999), Ichneumonidae (Horstmann er al. 1999), or Formicidae (Floren & Linsenmair 1997, 2000), but see the study by Amdddgnato (1997) on Acrididae of an Amazonian rain forest. The question of how many ar- tigonioidea belonged to hitherto undescribed taxa. Even conventional collection methods (light-trapping, thropod speciesare tree specialistsis still intensively discussed (Erwin 1982, Basset 1996, Basser et al. hand collecting, and stem eclectors)revealed49o/ooF 1996, Mawdsley & S species as new. As is rypical for tropical insect faunas, the majo- rcrk 1997, Oedegaard 2000). In addition, the lack oFthe existenceof a tree-specificcommuniry is sup- riry of Orthoptera speciesoccurred in small numbers ported by our result that Orthoptera were obviously and sampling effort was much too small to represent sampled at random lrom a large speciespool: neither the taxonomic nor the guild composition of Orth- local speciesdiversity or even to assessthe necessary Further sampling effort required lor reliable estimates. The highest ranks in speciesrichnessand abundance among fogged species- considering adults as well as nymphs - were occupied by rwo cricket subfamilies - Podoscirtinae and Mogoplistinae. However, only opteran communities had changed significantly in the re-fogging samples. Orthoptera colonize trees rather slowly compared to Coleoptera and Diptera, which can reach original abundances already after one day (Floren & Linsenmair 1998a). Vhile the total two speciesof Mogoplistinae (genus Ornebius),which number of Orthoptera dropped significantly in the are generalist feeders,occurred on all tree individuals. daily foggings, the proportion o[adults and nymphs did not change. The predatory Meconematinae we- These specieswere lacking in the stem eclectors, have 39 "canopy re the lastest colonizers. However, relative proporti- most analyses oF ons showed no significant differences from the ori- to adult holometabolic insects.Already in our first ginal conditions already after one week, indicating fogging samples we recognized that nymphs of holo- that most specimens enter tree crowns from neigh- metabolous species were almost completely lacking in the trees (Floren & Linsenmair 1997,1998a,b, boring treesaccording to their frequency in the spe- faunas" have been limited cies pool, without tree-species-specific featuresbeing 1999). Their absencefrom the tree crowns was con- important. From our data we were not able to di- firmed by additional hand collecting, close observations in the trees (Riede, unpublished), and further stinguish either pioneer or climax species,such aswere identified by Amdddgnato & Descamps (1980) in larly and in high abundances,always outnumbering our investigation, Podoscirtinae and Mogoplistinae adults. This difference can be explained by the high xa, especiallyof the numerically dominant Hemiptera and Orthopteroidea sensu latu, are lound regu- were dominant there. Podoscirtinaeincluded almost predation pressureofthe ants, that dominate all trees rwice as many larvae as Mogoplistinae, but Mogo- and patrol in the canopy regularly (Floren submitted). which belong to the genera Ectatoderus and Ornebius and which are specifically diflerent from the species occurring on Mt. Kinabalu. This might indicate a high degreeof endemism for the fauna of the region oFMount Kinabalu, which is a center of diversiry in SE Asia (Ashton 1989, Myers et a|.2000). The extreme rariry of short-horned grasshoppers (Caetifera) in our sampleswas surprising and is in contrast to Amazonian and Costa Rican rain lorests ( R o b e r t s 1 9 7 3 , D e s c a m p s1 9 8 1 , f u e d e 1 9 9 3 , A m d - Ba most of which attacked lessmobile prey. By contrast, the nymphs of hemimetabolous groups are highly mobile and thereforeable to escapelrom predatory ants. Similarly, eggs have to be protected by endophytic or lJa subterranean oviposition. Nymphs which hatch in the soil or lrom the lower vegetation layers could then climb up to the canopy to complete their life rycle. This might be an explanation lor the higher per- Ba centageof nymphs observed in our eclector samples compared with the logging samplesand points to the rized by a vivid coloration, stout legsand body form, "dendroand protuberant eyes, summarized as the p h i l o u s " l i f e l o r m b y D e s c a m p s( 1 9 7 6 , 1 9 8 1 ) . M a n y importance of tree trunks as routes to reach the ca- speciesare brachypterous or apterous. Some species nematinae) endophytic oviposition prevails,and the have developed endophytic and epiphytic egg-laying same could be true for the cricket sublamily Podos- strategies,which would allow them to spend their whole liFe cycle - from egg to adult - in rhe canopy (Amdddgnato 1997). An analysisof museum speci- c i r r i n a e ,w h i c h w a s r h e d o m i n a n t t a x o n i n o u r i n v e - tains similar dendrophilous specieswhich are, however,very rare (Riede I 993). This is corroborated by Asl Experimental feeding with caterpillarsdemonstrates the high predation pressureexerted by canopy anrs, ddgnato 1997). Arboreal grasshoppersare characte- mens shows that the SE-Asian grasshopper launa con- Ar foggings. In contrast, nymphs of hemimetabolous ta- large lorest gaps and anthropogenic clearings. 'We compared our results with logged material from Sulawesi(seeKnight & Holloway 1990). As in plistinae were more numerous as adults. There were four speciesof Mogoplistinae (probably undescribed) RT Acl nopy. On the other hand, in severaltettigoniid subfamilies (Phaneropterinae, Pseudophyllinae, Meco- Ba Be Dt stigations.This could mean that many of the arboreal speciescomplete their life rycles in the canopy. However, given the present state of knowledge these Dt interpretations must remain speculative,since biolo- Dr fogging studies by Stork (1991) in Brunei, and the gical information on tropical Ensifera is generally scarce(e.g., Bailey & Rentz 1990). additional 39 loggings of 10 tree specieslrom seven lamilies in Kinabalu Park (Floren unpublished). The ACKNO\vLEDGMENTS rariry oF CaeliFeradoes not seem to be a sampling 'W'e Er to park staff For generous support. The logged ma- Fl artifact. Among the possible causesof this rariry Riede (1 993) discusseshigher predation pressureand canopy structure as well as food plant availabiliry and patatabiliry. Howevet we must admit that Caeliferan scarcity in SE-Asian rainlorestsis still an enigma. Dffirences between ho/ometabolousand hemimetabolous arthropods that ffict 40 community composition.-lo date, are grateFul to the Director of Sabah Parks, Datuk Ali Lamri, lor authorization of our research,and terial from Sulawesiwas kindly lent by Judith Marshall (Natural History Museum London). tWe thank C.H.F. Rowell For critical comments on the manuscript. This study was supported by the German ResearchCouncil (DFC), Li l50l13 14. DIVI'.RSlIY Ol ORfHOP I I'-RAIjR()NI IIORNI:AN t-O\(l.ANl) IIAIN I()RflST TRI'-!'S Floren, A., & K.E. Linsenmair. 1998a. Diversity and reco- REFERENCES lonisation ofarboreal Formicidae and Coleoptera in a Achtziger, R., Nigmann, U., & H. Zwdlfer. 1992. Rarefaction-Methoden und ihre Einsatzmtiglichkeiten bei der zootikologischen Zustandsanalyse und Bewertung von Biotopen. Z. Okol. Natursch. l: 89 105. Amddignato, C., & M. Descamps. 1980. Evolution des populations d'orthoptdres d'Amazonie du Nord-Ouest dans les cultures rraditionelles et les formations secondaires d'origine anthropique. Acrida 8: 1-33. Amdddgnato, C. 1997 . Diversity oF an Amazonian canopy grasshopper community in relation to resource partitioning and phylogeny. Pp. 281-319 in Stork, N.E.' Adis J., S{ R.K. Didham (eds.). Canopy arthropods. London. Ashton, PS. 1989. Speciesrichness in tropical forests. Pp. 239-251 lz Nielson, H., & B. Lauritz (eds.).Tiopical forests. l,ondon. 'fhe Tettigoniidae their Bailey,V. J., & D. C. Rentz.I 990. diversity and importance in biological research.Pp. I 7 'Ibttigoin Barley, \q.J., & D.C.F. Rentz (eds.). The niidae.Berlin. Basset,Y. 1996. Local communities ofarboreal herbivores in Papua New Guinea: predictors of insect variables. Ecology 77: 1906-1919. Basset,Y., Samuelson,C.A., Allison, A, & S.E. Miller. 1996. How many speciesofhost-specific insectsfeed on a species oftropical tree? Biol. J. Linn. Soc. 39:201-216 Basset,Y., Springatc, N.D., Aberlenc, H.P' & B. Delvare. 1997. A review ofmethods for sampling arthropods in tree canopies. Pp.27 52lz Stork, N.E., AdisJ., & R.K. Didham (eds.). Canopy arthropods. London. Beaver, R.A. 1979. Host specificity of temperate and tropical animals. Nature 281: 139-141 Descamps, M. 1976. l.e peuplement acridien d'un piri- mBtre d'Amazonie colombienne. Ann. Soc. Entomol. 12:291-318. Descamps, M. 1978. Etude des Ecosystcms(iuyanais III. Acridomorpha dendrophiles (Orthoptera, Caclifera). Ann. Soc.Entomol. 14:301-349. Descamps,M. 1981. La laune dendrophile neotropicale.VI. Diagnoses gdndriques ct spdcifiques d'Acridoidea de la rigion de Manaus (Orthoptcra). Ann. Soc. Entomol. 17: 3 1I 3 3 0 . Erwin, T.L. 1982.'liopical liorcsts:'l'heir richness in Colcoptera and other arthropod species.Coleopts. Bull. 361: 74 75. andrecol997 Diversiry A., & K.E.Linsenmair. Floren, lonisation dynamics ofselccted arthropod groups on dif- lowland rain forest in Sabah, Malaysia. Selbyana 192: I55-161. 1998b.Non-equilibrium Floren,A., & K.E. l.insenmair. communities of Coleoptera in trees in a lowland rain forest of Borneo. Ecotropica 4: 55-67 . Floren, A., & K.E. Linsenmair. 2000. Do ant mosaicsexist in pristine lowland rain forests?Oecologia 123: 129-137. Horstmann, K., Floren, A., & K.E. Linsenmair. 1999. High of Ichneumonidae from the canopy of species-richness a Malaysian rain forest. Ecotropica 5: l-12. Huston, M.A. 1994. Biological diversity. Cambridge. Ingrisch, S. I 997. Taxonomy, stridulation and development of Podoscirtinaefroml'hailand (lnsecta:Ensifera:Grylloidae; Mogoplistinae). Senckenb. biol. 7 7 : 47 -7 5. Ingrisch, S. 1998. New Mogoplistinae from the Kinabalu region in Sabah, North Borneo (Ensifera, Grylloidea, Mogoplistidae). Senckenb. biol. 77 : 225 234. Kitayama, K. I 992. An altitudinal transectofthc vegetation on Mount Kinabalu, Borneo. Vegetatio lO2:149-171. Knight, \7.J., & J.D. Holloway. 1990. Insects and the rain forest of South east Asia (Vallacea). l,ondon. Lehmacher, W., W'assmer,G., & Il Reitmeir. 1991. Procedures of two-sample comparisons with multiple endpoints controlling the experimentwise error rate. Biometrics 47: 5ll-521. Linsenmair, K.E. 1995 Tropische Biodiversitdt - vom ursichlichen Verstdndnis der strukturellen und funktionellen Komplexitdt noch weit entfernt. Rundgespr:iche Komm. Okol. 10: I l5-1 26. Linsenmair, K.8., Davis, A.J., Fiala, B., & M.R. Speight. ln press.Tiopical forest canopies: ecology and managemcnt. Dordrecht. Magurran, A.E. 1988. Ecological diversiry and its mcasurement. Princeton. Mawdsley, N.A., & N.F,: Stork. 1997. Host-specificiry and the effective specialisation of tropical beetles Pp. 104-130 iz Stork, N.Fl., AdisJ., & R.K. Didham (eds.). Cianopy arthropods. London. Myers, N., Mittermeier, R.A., Mittermeier, C.C.' Fonseca' G.A.B., & J. Kent. 2000. Biodiversiry hotspots for conservation priorities. Nature. 403: 853-858. Oedegaard, F. 2000. How many speciesofarthropods? L,rwin's estimate revised.Biol. J. Linn. Soc. 7l:583-597. Otte, D., & ll Naskrecki. 1997. Orthoptera SpeciesOnline. http://viceroy.ecb.uconn.edu/Orthoptera. Pfciftbr, M. 1997. Sozialstrukturund Verhaltensbkologieder Ferenttree species in a lowland rain fbrest in Sabah, Ma- Riesenanreisen Camponotus gigas Latreille 1802 im Re- laysiawith specialreferenceto Formicidae. Pp.344-382 lz Stork, N.E,., Adis J., & R.K. Didham (eds.). Cano- genwald Malaysias auf Borneo. Ph. D. Thesis, Univer py arthropods. I-ondon. sity of Vucrzburg. Berlin. Pianka, E.R. 1996. Evolutionary ecology. New York. 4l H.ORr,.N fr,,11_. Price,W.P, Diniz, I.R., Morais,H.C., & E.S.A.Marques. 1995.The abundance ofinsectherbivorespecies in the tropics: the high local richnessof rare species.Biotropica27:468478. fuede,K. 1987.A comparative studyof matingbehaviour in someNeotropicalgrasshoppers Acridoidea.Ethology 76:265-296. Riede,K. 1993.Abundancedifferences ofphytophagoustro(Orthoppical insects,the caseof forest grasshoppers tera: Caelifera).Pp.2l1-128 rz Barthlott,V., Naumann,C.M., Schmidt-Loske, K., & K.L. Schuchmann (eds.).Animal-plant-interactions in tropical environments.Bonn, Roberts,H. 1973.ArborealOnhopterain rhe rain forestof Costa Rica collectedwith insecticide:A report on the Acrididae, including new species.Proc. grasshoppers Acad.Nat. Sci. Phil. 125:49-66. Rosenzweig, M.L. 1995.Species diversiryin spaceand time. Cambridge. 4l Rowell,H.F. 1978.Foodplant specificiryin Neotropicalrain forestacridids.Entomol. Exp.Appl. 24:61-662. Shinozaki,K. 1963. Note on the species-area curve. Proc. lst Ann. Meet. Ecol.Soc.Japan5. Tokyo. Stork,N. E. 1991.The compositionof the arthropodfauna of Bornean lowland rain forest trees.J. Tiop. Ecol. 7: l 61-l 80. Stork,N.E., Adis,J., & R.K. Didham. 1997.Canopy arthropods.London. Vagner, T. 1997.The beetlefauna ofdifferent tree species in forestsofRwanda and Emt Zaire. Pp. 169-183 in Stork,N.E., Adis J., & R.K. Didham (eds.).Canopy arthropods.London. 'Wagner, T. 1999. Arboreal chrysomelidcommuniry srructure and faunaloverlapbemeendifferentrypesofforests in Central Africa. Pp. 247-270 lz Cox, M.L. (ed.). Advancesin Chrysomelidaebiology.Leiden.