Taxonomic composition of Coleoptera, Hemiptera (Heteroptera and Coleorrhyncha) and Mutillidae (Hymenoptera) at five different altitudes in Lamington National Park (Queensland, Australia) Frode 0DEGAARD Ola H. DISERUD Norwegian Institute for Nature Research (NINA), PB 5685 Sluppen, NO-7485 Trondheim, Norway. Email: [email protected] Citation: Odegaard, F. & Diserud, O.H. 2011 12 20: Taxonomic composition of Coleoptera, Hemiptera (Heteroptera and Coleorrhyncha) and Mutillidae (Hymenoptera) at five different altitudes in Lamington National Park (Queensland, Australia). Memoirs of the Queensland Museum- Nature 55(2): 359-374. Brisbane. ISSN 0079-8835. ABSTRACT This study describes the taxonomic composition of Coleoptera and selected groups of Hemiptera (Heteroptera and Coleorrhyncha) and Hymenoptera (Mutillidae) in the understorey of a subtropical rainforest along an altitudinal gradient in Lamington National Park, Queensland, Australia. The altitudinal gradient was subdivided into five zones (300, 500, 700, 900 and 1100 m above sea level (a.s.l.)) within each of which, four replicated sampling sites were established. A total of 16 783 individals from 1219 species of Coleoptera, 715 from 92 species of Hemiptera and 105 from 17 species of Mutillidae were collected from beating low vegetation. Total species richness and abundance were generally high throughout the gradient, with more than 3000 individuals from over 400 species at each of the five altitudinal zones, but significantly fewer species and individuals were present at higher elevations (900 and 1100 m a.s.l.). Different taxonomic groups showed various patterns of altitudinal zonation, with many groups restricted to the higher elevations, particularly at 1100 m. Of the species unique to one altitudinal zone, half were restricted to 1100 m. The results of the present study provide important base-line data upon which predictions can be made in early warning monitoring systems with regard to climatic change. (cid:9633) IBISCA, Hemiptera, Hymenoptera. The insect fauna of Australian rainforests is are key components to understand ecosystem nighly diverse with a large proportion of endemic patterns and processes (Huston & Gielbert 1996; tax4 (Naumann 2000; Austin et d. 2004). There is a Basset 2001). Differences or change in such number of studies on the community structures of assemblages may have implications for eco¬ bisects in these forests (e.g. Basset 1991; Kitching system functioning and conservation issues (e.g. Didham 1997; Watt et d. 1997; Basset et d. 2003; k Arthur 1993; Grove 2002; Stork & Grimbacher Stork et d. 2007; Chen et d. 2009). -006, Wilson et d. 2007a). However, these studies often restricted to certain taxonomic groups Assemblage structure and diversity of insect e fauna of particular tree species, habitats communities change substantially along altitud¬ 01 strata (Kitching et d. 2001). Large data sets of inal gradients (Hagvar 1976; Janzen et d. 1976; Insect assemblages from different tropical forests Stork & Brendell 1990; Stevens 1992; Olson Me m°irs of the Queensland Museum | Nature • 2011 • 55(2) • www.qm.qld.gov.au 359 0degaard & Diserud 1994; Andrew et al. 2003; Brehm & Fiedler 2003; tropics, to a large extent, consist of endemic taxa, Wilson et al. 2007a; Colwell et al. 2008). In general, (e.g. Bell et al. 2002) such scenarios of climate species diversity decreases with altitude (Stevens change may cause severe species extinctions. In 1992), however, a peak of diversity is often seen this case, it is particularly important to identify at mid elevations (Janzen et al. 1976; Olson 1994, taxonomic components susceptible to such Wilson et al. 2007b; Colwell et al. 2008). In addition, changes in order to develop early warning mon¬ diversity does not change consistently across itoring systems (Moritz et al. 2001). taxonomic groups (Stork & Brendell 1990). Low species richness at higher elevations may reflect MATERIAL AND METHODS the lower rates of invasion and higher rates of extinction of populations that colonise them, as Study area. This study was carried out along well as differences in abiotic factors (MacArthur an altitudinal transect in Lamington National 1972; Stevens 1992). Park, Queensland, Australia. Rainforests within The target taxa of this study include diverse this park can be classified into several structural and ecologically important components of types including warm subtropical, cool sub¬ the arthropod fauna. The beetles are the most tropical, warm temperate and cool temperate species rich and ecologically diverse order of rainforests (Williams et al. 1984; Laidlaw et insects (Lawrence et al 2000) with an estimated al. 2011). Rainfall averages 1830 mm with 100000 species in Australia (Yeates et al. 2003). most falling during the summer months and at However, knowledge of beetles is very poor higher elevations (see also Strong et al. 2011). in this continent with, approximately, only Insect samples were obtained from four plots 28300 species currently described (Lawrence & at each of five different elevations; 300, 500, Britton 1994). The Heteroptera include about 700, 900, and 1100 m above sea level (a.s.l.). 2100 described Australian species, which is "Lower elevations" and "higher elevations' were about half of the estimated Australian fauna defined as 300 to 700 m a.s.l., and 900 to 1100 (Cassis & Gross 1995, 2002). The velvet ants m a.s.l., respectively. Locations and elevations (Mutillidae) are wasps that are parasitic on other of the individual plots are given by Kitching et wasps, bees or ants, and they include several al. (2011). hundred largely unstudied and undescribed Collecting method. Insects were sampled by Australian species (Austin et al. 2004). beating vegetation, using a 1.5 m long beating The aims of the present study was to describe stick and al xl m nylon sheet for collecting the taxonomic composition of selected insect the falling material. Vegetation included all groups in the lower strata of a subtropical rain¬ structures, foliage and the trunks of large trees forest, and to build up knowledge on how insect to thin branches, both living and dead. Each communities change along altitudinal gradients. sample was obtained by beating all reach¬ In order to predict the effects of future climatic able vegetation on both sides of a 20 m long changes, it is essential to accumulate such base¬ transect starting just outside the 20 x 20 m line knowledge upon which predictions can standard IBISCA plots (see Kitching et al. 2011) be made. One of the predicted scenarios of climate and walking in a straight line away from the change is a faunal shift to higher elevations due to plot. A forest area of approximately 3 x 20 m elevated temperatures and reduced precipitation (60 m2), and a forest volume of 60 m2 x 3 m (Stork et al. 2007), which may result in iow- height (180 m3) was covered by each sample. A Iand biotic attrition and mountaintop extinctions total of 10 parallel samples (10 transects) were (Wilson et al. 2007b; Colwell et al. 2008; Chen et taken at each plot, all performed in different al. 2009). As tropical mountaintop biotas of the directions from the same starting point to prevent 1360 Memoirs of the Queensland Museum | Nature • 2011 • 55(2) Taxonomic composition of Coleoptera, Hemiptera and Mutillidae 700 600 (cid:9633) Hemiptera (/> (cid:9633)Adephaga 0) 500 o (cid:9633) Staphyliniformia d> Cl (cid:9633) other Polyphaga CO 400 Ho— (cid:9632) Cucujoidea 1— 0) 300 (cid:9633)Tenebrionoidea -Q E u (cid:9633) Chrysomeloidea 2 200 (cid:9633) Curculionoidea 100 (cid:9633) Mutillidae 300m 500m 700m 900m 1100m (cid:9632) Hemiptera (cid:9633)Adephaga EBStaphyliniformia (cid:9633)other Polyphaga (cid:9632)Cucujoidea (cid:9633)Tenebrionoidea (cid:9633) Chrysomeloidea (cid:9633)Curculionoidea (cid:9633)Mutillidae G-1. Total species richness; A, and abundances; B, of different taxonomic groups at the five altitudinal zones. ^ composite group 'other Polyphaga' includes the Scarabaeiformia, Elateriforniia, Bostrichiformia nd Cleroidea within the Cucujiformia. ^-beating of the same area. Catches from the Accordingly, a total of 600 samples was obtained u Parallel samples in each plot were pooled (10 samples x 20 plots x 3 seasons). ,re analyses. The sampling procedure was reP|icated at three different periods of the year, Material. Each sample of beaten material was ch representing a different season: spring (6 collected in a zip-lock bag and all samples were 2^ October 2006), autumn (8 to 28 March sorted into different target groups the same 'O and summer (14 to 30 January 2008). day. The target taxonomic groups, which consist Mem oirs of the Queensland Museum | Nature • 2011 • 55(2) 361 0degaard & Diserud of a wide range of feeding guilds, included The altitudinal distribution of total species beetles (Coleoptera), true bugs (Heteroptera richness showed a peak at 700 m (Fig. la), including Coleorrhyncha) and velvet ants a pattern that seems to be driven by many (Mutillidae, Hymenoptera). All specimens of taxonomic groups including Cucujoidea, these groups were dry mounted, labelled, sorted Tenebrionoidea, Chrysomeloidea and Mutilli- into morphospecies and databased. The collection dae (Fig. 2). There was significantly lower is stored in the first author's collection at the species richness at higher elevations (Fig. 1, see Norwegian Institute for Nature Research (NINA). also Tables 1, 2). This pattern was reflected in the species richness of most taxonomic groups Statistical analyses. The data were first collated including the Cucujoidea, Tenebrionoidea and into nine taxonomic groups (Hemiptera, Ade- other Polyphaga, with the exception of Staphy- phaga, Staphyliformia, 'other Polyphaga', Iiniformia whose species richness increased at Cucujoidea, Tenebrionoidea, Chrysomeloidea, higher elevations (Fig. 2, Table 2). Curculionoidea, Mutillidae). Individual groups were analysed separately for species richness The number of individuals was largest at and abundance. The groups were treated 500 m with significantly less individuals at as independent of each other, so testing the higher elevations (900 and 1100 m) (Fig. 1, see effect of altitude on their responses has been also Tables 1, 2). These patterns were driven performed on each group separately. The by differences in several taxonomic groups variation between plots of the same altitude and including StaphyUniformia, Cucujoidea and the the seasonal variation were not incorporated in Tenebrionoidea whose abundance decreased these analyses, but may introduce a systematic at higher elevations (Fig. 3). The abundance of bias in the estimated variance within the alt¬ Curculionoidea increased gradually all along itudes. In these analyses we considered altitude the elevation gradient, and species richness was to be a categorical variable with five levels, so significantly greater at the higher compared to the test was a standard one-way ANOVA. We the lower altitudinal group (Table 2). also performed post-hoc two-sample equal Particular families showed very pronounced variance t-tests between pairs of adjacent patterns of altitudinal zonation. The large number altitudinal zones and low (300, 500 and 700 of weevils at higher elevations (Fig 3) were m) versus high (900 and 1100 m) elevational mainly caused by the wood-boring subfamiliy groups. All statistical analyses were performed Cryptorhynchinae (see Appendix 1). The beetle with the free statistical software R (R Development families Byrrhidae and Phloeostichidae, and the Core Team 2008). bug families Peloridiidae, Enicocephalidae and Schizopteridae were almost exclusively found at RESULTS 1100 m, while no families were restricted to the lower elevations. High abundance and species The material gave a total of 16783 Coleoptera richness of Staphyliniformia and Lygaeidae individuals sorted to 1219 species of 70 families. were also found at higher altitudes (Appendix The Hemiptera material consisted of 715 spec¬ 1). The Adephaga (Carabidae in this case) were imens sorted to 92 species of 13 families of also significantly more abundant and species the suborder Heteroptera and one family of rich at 1100 m compared to 900 m (Table 2). Coleorrhyncha (Peloridiidae). Hymenoptera material consisted of 105 velvet ants (Mutillidae) A total of 185 species (14%) were exclus¬ belonging to 17 species. The number of species ively found at 1100 m, and when we also and abundance of each family collected at each include species restricted to 900 m (105 spp) elevation are presented in Appendix 1. and both 900 m and 1100 m (61 spp), a total 1362 Memoirs of the Queensland Museum | Nature • 2011 • 55(2) Taxonomic composition of Coleoptera, Hemiptera and Mutillidae Hemiptera Adephaga p u o r g r e p s e i c e p s r o r e o m u N Staphyliformia Other Polyphaga p u o r g r e p s e i c e p s r o r e o m u N Pt • 2. (This page and opposite) Box-plots for species richness of the nine taxonomic groups and all croups combined ('total') at five different altitudinal zones. The composite group 'other Polyphaga' includes e ^carabaeiformia, Elateriformia, Bostrichiformia and Cleroidea within the Cucujiformia. ^errr oirs of the Queensland Museum | Nature • 2011 • 55(2) 3631 0degaard & Diserud Cucujoidea Tenebrionoidea p u o gr r e p s e ci e p s or er b m u N Chrysomeloidea Curculionoidea p u o gr r e p s e ci e p s or er b m u N Mutillidae Total p u o gr r e p s e ci e p s or er b m u N 364 Memoirs of the Queensland Museum | Nature • 2011 • 55(2) Taxonomic composition of Coleoptera, Hemiptera and Mutillidae TABLE 1. P-values from one-way ANOVA tests for differences in the number of individuals and species richness of the nine taxonomic groups and all groups combined ('total') among the five altitudinal zones. The composite group 'other Polyphaga' includes the Scarabaeiformia, Elateriformia, Bostrichiformia and Cleroidea within the Cucujiformia. Group individuals species Hemiptera 0.2451 0.7579 Adephaga 0.0912 0.0392* J^taphyliformia 0.0000 *** 0.0009 *** Other Polyphaga 0.6258 0.0057** Cucujoidea 0.0000 *** 0.0000 *** Tenebrionoidea 0.0001 *** 0.0000 *** jEhrysomeloidea 0.0228 * 0.0320 * J7urculionoidea 0.0058 ** 0.0142* Mutilidae 0.05407 0.0682 = Total 0.0488* 0.0121 * TABLE 2. Significant P-values (p<0.05) representing differences in species richness (spp.) and abundance (ind.) between pairs of different altitudinal zones, or groups of altitudinal zones for nine taxonomic groups and all groups combined. Altitudinal zones of 300, 500 and 700 m were grouped as lower elevation (low ^• ) /and 900 and 1100 m grouped as higher elevation (high el.). The composite group 'other Polyphaga' deludes the Scarabaeiformia, Elateriformia, Bostrichiformia and Cleroidea within the Cucujiformia. 300 m vs. 500 m 500 m vs. 700 m 700 m vs. 900 m 900 m vs. 1100 m low el. vs. high el. _ ind. spp. ind. spp. ind. spp. ind. spp. ind. spp. Hemiptera n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. Adephaga n.s. n.s. n.s. 0.0108 n.s. n.s. 0.0385 0.0233 n.s. n.s. ^aphylmiformia n.s. 0.0298 0.0001 n.s. n.s. n.s. n.s. n.s. 0.0004 0.0003 other Polyphaga n.s. n.s. n.s. n.s. n.s. 0.0087 n.s. n.s. n.s. 0.0002 Cucujoidea n.s. 0.0482 n.s. n.s. <0.0001 0.0001 n.s. n.s. <0.0001 <0.0001 Tenebrionoidea 0.0012 n.s. n.s. n.s. 0.0124 0.0007 0.0434 n.s. 0.0189 <0.0001 Chrysomeloidea n.s. n.s. 0.0459 n.s. n.s. 0.0374 0.0453 n.s. n.s. 0.0041 ^^ulionoidea n.s. n.s. n.s. 0.0185 n.s. n.s. n.s. n.s. 0.0018 n.s. ^utillidae n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. ^• groups n.s. n.s. n.s. n.s. n.s. 0.0062 n.s. n.s. 0.0085 0.0255 Mem oirs of the Queensland Museum | Nature • 2011 • 55(2) 365 0degaard & Diserud of 351 species (26.4%) were restricted to higher insects. However, larger taxonomic scales may elevations. However, as rare species may be mask changes in species composition nested found at one particular elevation by chance, within the same taxonomic group or feeding we repeated the calculations for species with guild (Grimbacher & Stork 2007). In order to more than 5 individuals in the total samples. detect such changes, it is necessary to scale The pattern remained the same with as many down the taxonomic resolution of the target as 86 (16.5%) out of 522 species restricted to groups as much as possible, ideally to the higher elevations (900 and 1100 m). Of the 72 species level. For instance, overall abundance species (with more than 5 individuals) unique of Staphyliniformia decreased with increased to one elevation, half (50 %) were restricted to altitude. Only species-level identification 1100 m (Fig. 4). revealed the intriguing pattern that some extremely common species predominate at lower altitudes, while the higher elevations DISCUSSION are characterised by a unique and species rich The present study found decreasing abund¬ fauna. On the other hand, if species assemblages ance and species richness of insects at respond similarly to environmental factors, higher elevations which agrees with most responses may be readily detectable for nested studies of species diversity along altitudinal taxonomic groups or feeding guilds. Hence, gradients (Stork & Brendell 1990; Stevens the identification of proper target taxa or 1992). However, the span of elevations in the guilds for early warning monitoring systems is present study was probably not large enough a key stone in order to detect changes, such as to see the prominent diversity declines such as climate warming. those normally seen at even higher elevations, The present study is based on samples from e.g. between 1500 and 3500 m a.s.l. (Brehm & only one forest stratum (understorey) and Fiedler 2003; Wilson et al 2007b; Chen et al particular tree species where processes may not 2009). Decrease in diversity with altitude may be representative of other strata or tree species. be explained by parameters such as climatic Tree species may be affected differentially factors and metapopulation structures by climatic changes and elevated C02 levels (Hagvar 1976; Janzen et al. 1976; Stork & due to their unique characteristics such as Brendell 1990; Stevens 1992; Olson 1994; different root structures, hydraulic properties Andrew et al. 2003; Brehm & Fiedler 2003). In or photosynthetic rates (Stork et al. 2007). addition, the vegetation structure at the The effects of climate change may be more 1100 m plots differs from the others by having pronounced in canopy than the understorey, a significantly smaller number of tree species for example, through changes in leaf traits. and a much more prominent epiphyte flora. However, the structures of feeding guilds of However, the total effects of these contrasting beetles in the canopy of tropical rainforests in factors on species diversity are unknown, northern Australia did not differ from that at and investigation of individual parameters is ground level (Grimbacher & Stork 2007). beyond the scope of the present paper. A large proportion of the species in this study Community structure of insects may be meas¬ were restricted to higher elevations. Due to the ured at larger taxonomic scales. For instance, if large proportion of rare species, however, the the distributions of specific resources change exact proportion of high-altitude specialists along altitudinal gradients, relative abundances remains unknown. In the case of this study, the of particular feeding guilds may be used to true percentage of species restricted to higher capture differences in community structure of elevations should probably lie somewhere be- 1366 Memoirs of the Queensland Museum | Nature • 2011 • 55(2) Taxonomic composition of Coleoptera, Hemiptera and Mutillidae Hemiptera Adephaga p ou0 gr 6 er 0 p5 s vidual 40 ndi30 er of i 20 d m0 Nu 1 0 300 500 700 900 1100 Altitude Staphyliformia Other Polyphaga group 150 er p viduals 100 di n of i0 er 5 b m u N 0 Cucujoidea 0 oup 20 gr s per 150 al u d vi00 ndi 1 or i ber 50 m u N 300 500 700 900 1100 300 500 700 900 1100 Altitude Altitude Me foil's of the Queensland Museum | Nature • 2011 • 55(2) 367 0degaard & Diserud Chrysomeloidea Curculionoidea Mutillidae Total Q. O3 cn 0 Q. (0 00 O > V c 6 l_ 00 E 0 z 300 500 700 900 1100 Altitude FIG. 3. (This page and opposite) Box-plots of abundances of the nine taxonomic groups and all groups combined (Total') at five different altitudinal zones. The composite group 'other Polyphaga' includes the Scarabaeiformia, Elateriformia, Bostrichiformia and Cleroidea within the Cucujiformia. tween 16.5 % and 26.4 %. A substantial increase associated with the moist environment in the in sampling intensity may be necessary to de¬ Nothofagiis-forest. scribe the elevational preferences of all taxa, The results of the present study may serve but singletons found in our study also provide as important base-line data upon which pred¬ valuable information as many almost certainly ictions can be made in early warning mon¬ prefer the elevation from where they were itoring systems with regard to climatic change. recorded, based on knowledge of the general If levels of moisture and precipitation decrease feeding habit of the groups to which they belong. in these forests as predicted from climatic The species restricted to higher elevations are models (Foster 2001), species restricted to generally dominated by moss-feeders and taxa certain ranges of altitudes may shift their 1368 Memoirs of the Queensland Museum | Nature • 2011 • 55(2)