THE RAFFLES BULLETIN OF ZOOLOGY 2011 THE RAFFLES BULLETIN OF ZOOLOGY 2011 59(2): 181–194 Date of Publication: 31 Aug.2011 © National University of Singapore TAXONOMIC AND FUNCTIONAL DIVERSITY OF ANTS (HYMENOPTERA:FORMICIDAE) IN AN UPPER HILL DIPTEROCARP FOREST IN PENINSULAR MALAYSIA Nur-Zati A. Mustafa Entomology Branch, Forest Research Institute Malaysia 52109, Kepong, Selangor, Malaysia E-mail: [email protected] (Corresponding Author) Hannah M. W. Salim Department of Environmental Science, Policy, and Management, University of California, Berkeley, California, 94720, United States Christine Fletcher Forest Ecology Branch, Forest Research Institute Malaysia 52109, Kepong, Selangor, Malaysia Abd Rahman Kassim Water Quality Program, Forest Research Institute Malaysia 52109, Kepong, Selangor, Malaysia Matthew D. Potts Department of Environmental Science, Policy, and Management, University of California, Berkeley, California, 94720, United States ABSTRACT. – Studies of ant diversity provide valuable insights into the health and functioning of forest ecosystems, but ants are surveyed infrequently in many tropical forest ecosystems. The majority of lowland forests in Peninsular Malaysia was cleared for land development in the 1970s and 1980s, leaving upper hill dipterocarp forests as almost all remaining contiguous, primary forests in Peninsular Malaysia. The ant communities of these forests have not previously been documented. Our extensive survey of a 200 ha upper hill dipterocarp forest site in Temengor Forest Reserve captured 10,307 individual ants, representing 211 species from 60 genera and nine subfamilies. Myrmicinae was the most common ant subfamily censused, with Pheidole recorded as the most speciose genera (40 species) followed by Polyrhachis (19 species) and Camponotus (13 species). Generalised Myrmicinae was the most diverse functional group, followed by Cryptic Species, Subordinate Camponotini and Tropical Climate Specialists. Seven undetermined specimens of possibly new species were recorded, suggesting that the upper hill dipterocarp forest of Temengor Forest Reserve is home to numerous ant species that have not been documented before. Our fi ndings can be utilised to better understand the ant community composition and function of primary upper hill dipterocarp forests in Peninsular Malaysia as compared to other dipterocarp forests. Our results can also serve as a baseline to understand post-disturbance changes to ant community composition and functional diversity. KEYWORDS. – Ants, Upper Hill Dipterocarp Forest, Peninsular Malaysia. INTRODUCTION Given the importance of ants as bioindicators of disturbance and environmental stress (e.g., Andersen, 1997; Hoffmann Arthropods are often underappreciated, even though they & Andersen, 2003), it is valuable to go beyond identifying provide irreplaceable ecosystem services that must be taken ant specimens at a taxonomic level by also assigning them to into account when studying the functions of both intact functional groups. Although ants can be hard to identify due and disturbed ecosystems (Seastedt & Crossley, 1984). to their small size and their hyperdiversity (Longino et al., Ants are especially important ecosystem engineers that 2002), ant diversity and functional group surveys have been contribute to numerous crucial functions, including nutrient successfully employed by land managers to gain insight into cycling, decomposition of organic matter, soil aeration, the the health and functioning of ecosystems (Andersen, 1997; suppression of soil-borne diseases and pests, and the direct Andersen et al., 2002; Andersen & Majer, 2004). and indirect alteration of soil properties (Folgarait, 1998). 181 Nur-Zati et al.: Ants in Peninsular Malaysia Andersen (2000) made a careful study of the global ecology The objective of this study is to undertake what is, to our of rainforest ants to understand how the structure and function knowledge, the fi rst survey of ant diversity and function of ant communities differ across biomes. He classifi ed four in a pristine upper hill dipterocarp forest in Peninsular primary stressors of ants (low temperature, microhabitat Malaysia. Our fi ndings may be used for future comparisons structure and resource capture, nest site availability and of the ant communities of primary upper hill dipterocarp food supply), and then identifi ed seven ant functional groups forests in Peninsular Malaysia with the ant communities in based on ants’ response to competition, disturbance and stress dipterocarp forests at other elevations and across various (Appendix 1). Andersen utilised these functional group other environmental gradients. Additionally, since our classifi cations to analyse the distribution of ant functional study provides the only description of the ant community groups before and after disturbance in humid tropical in an undisturbed upper hill dipterocarp forest in Peninsular forests. He found a signifi cant increase in ants categorised Malaysia, our results can serve as a baseline to understand as Dominant Dolichoderinae, and a substantial decrease post-disturbance changes to ant community composition in individuals categorised as tropical climate specialists, and functional diversity. Our study is especially valuable indicating that the disturbance had a clear effect on ecosystem as timber harvesting and other anthropogenic land-use function. The study of ants in relation to their functional changes continue to occur in upper hill dipterocarp forests groups directly improves our ability to understand how in Peninsular Malaysia. anthropogenic disturbance affects the ecosystem services ants provide (King et al., 1998). MATERIAL AND METHODS Dipterocarp forests (dominated by trees in the family, Dipterocarpaceae) account for more than three-quarters of all Sampling location. – The study was conducted in Southeast Asian forests, representing substantial carbon sinks Compartment 44, Block 5 of the 9765 ha Perak Integrated and biodiversity pools in the region (Manokaran, 1995). In Timber Complex (PITC) (5° 24' N, 101° 33'E), a primary Malaysia, dipterocarp forests account for about 85% of the hill and upper hill dipterocarp forest concession within country’s forested areas (Kamaruzaman & Ahmad, 2003) Temengor Forest Reserve (TFR) (Fig. 1). TFR was declared a and are commonly composed of species from the genera Permanent Forest Reserve under Malaysia’s National Forestry Anisoptera, Dipterocarpus, Dryobalanops, Hopea, Shorea Act in 1984. It lies on the western portion of the Titiwangsa and Parashorea (Thang, 1987). Main Range of Peninsular Malaysia in the northern state of Perak, near the border of Thailand. TFR is located at an While the fl oristic composition and the charismatic megafauna altitude of 400–1000 m above sea level. Our sampling in of Malaysian dipterocarp forests tend to be well studied in PITC was carried out within an altitude range of 550–810 m both Malaysian Borneo and in Peninsular Malaysia, the in a forest dominated by Shorea platyclados (Sloot et Foxw), study of non-charismatic fauna lags signifi cantly behind Dipterocarpus costulatus (Pierre), Koompassia malaccensis (Chey et al., 1997; Godfrey et al., 1999). Despite several (Benth) and Intsia palembanica (Miq), which places our extensive ant studies that have been carried out in Malaysian sampling location within an upper hill dipterocarp forest Borneo (Chung, 1995; Brühl et al., 1998; Hashimoto et al., (Wyatt-Smith, 1963). 2001), very little is known about ant taxonomy or ecology in Peninsular Malaysia (Agosti et al., 1994). Only a few Sampling design. – Sampling occurred within a 200-ha area preliminary studies of ant diversity have ever been carried of PITC where 24 sampling plots were established (Fig. out in Peninsular Malaysia (e.g., Fiala et al., 1994; Liefke et 1). Each sampling plot consisted of a 20 × 80 m rectangle al., 1998). The most comprehensive study of ant diversity that contained four square subplots (20 × 20 m) delineated in Peninsular Malaysia took place in the lowland dipterocarp within them. Each subplot was further divided into four 10 forest of the Pasoh Forest Reserve (PFR) in Negeri Sembilan x 10 m quadrats. Each plot consisted of eight baits, eight (Bolton, 1998; Moog et al., 2003). A total of 427 species ground pitfall traps, eight arboreal pitfall traps and four sets from nine subfamilies have been documented at PFR (Malsch, of leaf litter sifting. 2000). Sampling methods. – Four sampling methods were employed However, the surveys of ant diversity at low-elevation PFR in order to thoroughly survey ant diversity, target ants are not likely representative of ant diversity in the rest of present in different strata and capture ants with different the peninsula. The majority of lowland forests in Peninsular functions. Sampling was conducted in Mar.2008 and Malaysia was cleared in the 1970s and 1980s for urban repeated in Mar.2009. The sampling methods chosen were development and agriculture, particularly oil palm and rubber baiting (Keeler, 1980), ground pitfall traps (Fichter, 1941), plantations (Liow et al., 2002; Abdullah & Nakagoshi, 2006). arboreal pitfall traps (Samson et al., 1997) and leaf litter As a consequence, almost all contiguous, primary forests in sifting (Olson, 1991). All methods were modifi ed to suit Peninsular Malaysia are now the less-accessible, steep hill the forest topography as detailed below. and upper hill dipterocarp forests found at higher elevations. We know little to nothing about the ant communities of these Baiting – Baiting ants is commonly used to estimate the forests, although we would expect them to differ from the composition and richness of ground-foraging ant fauna ant communities found in lowland forests (Sanders, 2002). (Greenslade, 1972). We placed a teaspoon of tuna and a cotton ball moistened with a 20% honey and 80% water 182 THE RAFFLES BULLETIN OF ZOOLOGY 2011 solution on transparent plastic plates in quadrats 2 and 4 tree trunks (5–20 cm dbh) at a height of 1.3 m in quadrats of each subplot, for a total of eight replicates in each plot. 1 and 3 of each subplot. Arboreal pitfall traps were baited After 60 min, visual counts of all ants at the baits were and fi lled with killing agent and left for 48 hours before completed and identifi cations were recorded to morpho- they were collected. species. Opportunistic hand collections of ants using soft forceps were also carried out. Leaf litter sifting – Ants are most commonly found in the leaf litter layer on top of the soil surface. Therefore, leaf Ground pitfall traps – Ground pitfall traps are generally litter sifting is particularly well-suited for use in structurally used to obtain an estimate of the abundance and species complex habitats with abundant leaf litter, such as our upper composition of ants active on the soil surface. Each ground hill dipterocarp forest study site (Bestelmeyer et al., 2000). pitfall trap was established in quadrats 1 and 3 in each We surveyed ant diversity within leaf litter in quadrat 3 of subplot by embedding a hard plastic cup (5-cm diameter) each subplot, for a total of four replicates in each plot. We in the soil until the mouth of the cup was even with the soil sieved leaf-litter for 30 min in each quadrat by screening surface. 20 ml of water mixed with fragrance-free detergent batches of the leaf litter over a 100 × 80 cm white sheet was placed in the cup to serve as a killing agent. A coarse using a coarse (5 × 5 mm) rectangular (15 × 30 × 3 cm) wire wire net was placed over the mouth of the cup to exclude sieve. All sieved ants were collected in vials. larger organisms and to support a transparent plastic plate (15-cm diameter) approximately 2.5 cm above ground that Identifi cation and analysis. – Collected specimens were shielded the trap from rainfall. All pitfall traps were left in brought to the laboratory where pinning and identifi cation the fi eld for 48 hours before they were retrieved for species were conducted to the genus level according to Bolton (1994) identifi cation in the laboratory. and Hashimoto et al. (2001). When possible, specimens were identifi ed to the species level; when this was not possible, Arboreal pitfall traps – Extensive arboreal ant sampling can identifi cations were left as morpho-species. Identifi cations be carried out using fogging techniques that are expensive to the species level were made by direct comparison of and require a substantial investment of time and effort by collected specimens to the collection at Kagoshima University researchers (Delabie et al., 2000). A simpler and more (collector: Seiki Yamane). cost-effective way to analyse ant composition at the arboreal level is to use arboreal pitfall traps. This sampling method A rarefi ed species accumulation curve of individuals was is a modifi ed form of the ground pitfall trapping method created for each sampling year to determine if species (Bestelmeyer et al., 2000). Arboreal pitfalls were constructed in the site were adequately sampled (Brühl et al., 1998). by twist-tying disposable plastic cups with snap-on lids onto Species diversity estimations were calculated to determine the sampling effi ciency. Ant genera were later assigned to functional groups following Andersen’s (2000) classifi cations (Appendix 2), in order to characterise the ecosystem services provided by ants in TFR. All analyses were computed using the Vegan: Community Ecology Package (Oksanen et al., 2009) and Biodiversity R Package (Kindt & Coe, 2005) in R (R Development Core Team, 2009). RESULTS The four sampling methods (ground pitfall traps, arboreal pitfall traps, baiting and leaf litter sifting) that we employed to survey ant diversity in TFR led to the cumulative capture of 10,307 individuals, representing 211 species from 60 genera and nine subfamilies (see Appendix 2 for complete species list, abundances and functional classifi cations). The estimated total species richness ranged between 177.7 and 271.7, which yielded more than 70% sampling effi ciency (Table 1). Sampling effi ciency was calculated by dividing the number of actual species caught by the number of Fig. 1. a. Sampling occurred within a 200-ha area of the Perak Integrated Timber Complex (PITC) where 24 sampling plots were estimated species. A combined species accumulation curve established. b. Filled circle shows the location of PITC in the representing the sampling completed in both years approached Temengor Forest Reserve in northern Peninsular Malaysia. c. Each an asymptote after 40 plots were sampled (Fig. 2). Separate sampling plot consisted of a 20 × 80 m rectangle that contained species accumulation curves for the sampling carried out four square subplots (20 × 20 m) delineated within them. Each in 2008 and 2009 both begin to approach asymptotes after subplot was further divided into four 10 × 10 m quadrats. Open approximately 20 plots were sampled. circles indicate placement of arboreal pitfall and ground pitfall traps and fi lled circles indicate placement of baits and leaf litter sifting in each quadrat. 183 Nur-Zati et al.: Ants in Peninsular Malaysia Table 1. Species estimation of ants trapped in 2008 and 2009 in Temengor Forest Reserve, Peninsular Malaysia. Numbers in parentheses indicate sampling effi ciency*. Actual number of species Chao estimates First order Jackknife Bootstrap 2008 152 219.6 ± 23.3 (69.2) 210.4 ±15.6 (72.2) 177.7 ± 7.7 (85.5) 2009 153 208.1 ± 19.1 (73.5) 207.7 ± 14.0 (73.7) 178.6 ± 7.3 (85.7) All 211 264.4 ± 18.2 (79.8) 271.7 ± 13.2 (77.7) 238.4 ± 7.2 (88.5) *Sampling effi ciency was calculated by dividing the number of actual species caught by the number of estimated species (Brühl, 2001). Table 2. Species richness and abundance trapped in 2008 and 2009 in Temengor Forest Reserve, Peninsular Malaysia, listed by subfamily. Subfamily 2008 2009 Total Species Individuals Species Individuals Species Individuals Amblyoponinae – – 1 1 1 1 Cerapachyinae 1 2 – – 1 2 Dolichoderinae 5 97 6 264 7 361 Ectatomminae 2 62 3 54 4 116 Formicinae 42 593 36 780 50 1373 Myrmicinae 76 3936 0 3305 111 7300 Ponerinae 24 494 26 656 35 1150 Proceratiinae 1 1 – – 1 1 Pseudomyrmicinae 1 2 1 1 1 3 Total 152 5187 153' 5061 211 10307 Overall, the subfamily Myrmicinae, yielded the highest Amblyoponinae, Cerapachyinae, Proceratiinae, and number of species (111 species), followed by Formicinae (50 Pseudomyrmicinae (Table 2). The species richness of species), Ponerinae (35 species), Dolichoderinae (7 species) both Myrmicinae and Formicinae were higher in 2008 and Ectatomminae (4 species). Only one species from each than in 2009, but the species richnesses of Ponerinae and of the four subfamilies Amblyoponinae, Cerapachyinae, Dolichoderinae were higher in 2009. The same species Proceratiinae and Peudomyrmicinae was trapped (Table 2). in Pseudomyrmicinae (Tetraponera attenuate) was found The relative abundance of the nine subfamilies according to in both years. Amblyopone reclinata in the subfamily both total individuals and number of species did not differ of Amblyoponinae was found only in 2009, while the between years, with Myrmicinae ranking highest, followed subfamilies Cerapachyinae and Proceratiinae were found by Ponerinae, Formicinae, Dolichoderinae, Ectatomminae, only in 2008. Of the total number of species trapped in both sampling years, 72.0% was captured in 2008 and 72.5% was captured in 2009. The dominant species composition was similar in the two sampling years. At the genus level, 63.3% of the 60 genera was trapped in both sampling years. The total species richness captured in the two sampling years was also very similar, with 152 species captured in 2008 and 153 species captured in 2009. Pheidole was recorded as the most speciose genera (42 species, 19.9%; Table 3), followed by Polyrhachis (19 species; 9.0%) and Camponotus (13 species; 7.1%). The most abundant species was Lophomyrmex bedoti (Emery) with a total of 3265 individuals caught throughout the sampling period, followed by Crematogaster sp C with 683 individuals captured (Fig. 3). Both of these species belong to Myrmicinae. The third most abundant species was Odontoponera transversa (Smith) in the subfamily of Ponerinae with 330 individuals caught. Euprenolepis procera (Emery) was the most abundant ant in the Formicinae subfamily (187 individuals), while Dolichoderus thoracicus (Smith) was the most abundant Dolichoderinae (154 Fig. 2. Species accumulation curves with error bars for sampling in Temengor Forest Reserve, Peninsular Malaysia in 2008 [Δ], 2009 individuals). Of the 211 species trapped, 48.8% of species [ο] and combined [■]. had a total number of captures of fi ve or fewer individuals. 184 THE RAFFLES BULLETIN OF ZOOLOGY 2011 Table 3. Distribution of species by genus for ants sampled in Temengor Forest Reserve. Number of species is followed by percentage of total species in parentheses. Genus Total Genus Total Genus Total Pheidole 42 (19.9) Platythyrea 3 (1.4) Discothyrea 1 (0.5) Polyrhachis 19 (9.0) Echinopla 2 (0.9) Emeryopone 1 (0.5) Camponotus 15 (7.1) Euprenolepis 2 (0.9) Lepisiota 1 (0.5) Crematogaster 10 (4.7) Meranoplus 2 (0.9) Lophomyrmex 1 (0.5) Hypoponera 10 (4.7) Myrmoteras 2 (0.9) Lordomyrma 1 (0.5) Tetramorium 9 (4.3) Ponera 2 (0.9) Mayriella 1 (0.5) Pachycondyla 8 (3.8) Recurvidris 2 (0.9) Myopias 1 (0.5) Leptogenys 5 (2.4) Paraparatrechina 1 (0.5) Odontomachus 1 (0.5) Myrmicaria 5 (2.4) Acanthomyrmex 1 (0.5) Odontoponera 1 (0.5) Pheidologeton 5 (2.4) Acropyga 1 (0.5) Oecophylla 1 (0.5) Strumigenys 5 (2.4) Amblyopone 1 (0.5) Carebara 1 (0.5) Gnamptogenys 4 (1.9) Anochetus 1 (0.5) Prenolepis 1 (0.5) Myrmecina 4 (1.9) Aphaenogaster 1 (0.5) Proatta 1 (0.5) Paratopula 4 (1.9) Cataulacus 1 (0.5) Pseudolasius 1 (0.5) Technomyrmex 4 (1.9) Cerapachys 1 (0.5) Pyramica 1 (0.5) Monomorium 4 (1.9) Cladomyrma 1 (0.5) Solenopsis 1 (0.5) Dolichoderus 3 (1.4) Cryptopone 1 (0.5) Tetraponera 1 (0.5) Cardiocondyla 3 (1.4) Dacetinops 1 (0.5) Vollenhovia 1 (0.5) Nylanderia 3 (1.4) Diacamma 1 (0.5) Eurhopalothrix 3 (1.4) Dilobocondyla 1 (0.5) Twenty-fi ve species were captured in an abundance of 100 or Species (38 species; 18%), Subordinate Camponotini (36 more individuals; 16 of these species belong to Myrmicinae, species; 17.1%), Tropical Climate Specialists (28 species; while fi ve belong to Ponerinae, three to Formicinae and one 13.3%), Opportunists (28 species; 13.3%), Specialist to Dolichoderinae. Singletons of 38 different species were Predators (23 species; 10.9%), Dominant Dolichoderinae captured, representing fi ve subfamilies: Amblyoponinae, (three species; 1.4%) and Cold Climate Specialist (one Dolichoderinae, Formicinae, Myrmicinae and Ponerinae. species; 0.5%).When taking into account overall abundance, Tropical Climate Specialists were most abundant (3967 Generalised Myrmicinae was the most diverse functional individuals), followed by Generalist Myrmicinae (2659 group (54 species; 25.6%; Fig. 4), followed by Cryptic d e ur pt a C s al u d vi di n of I er b m u N Fig. 4. Number of species (bars) and number of individual (lines) of ants from different functional groups. C: Cryptic species; DD: Fig. 3. Species rank abundance for number of ants caught during Dominant Dolichoderinae, GM: Generalised Myrmicinae; TCS: the sampling in Temengor Forest Reserve, Peninsular Malaysia in Tropical Climate Specialists; O: Opportunists; SC: Subordinate 2008 [Δ], 2009 [ο] and combined [■]. Componitini; SP: Specialist Predators. 185 Nur-Zati et al.: Ants in Peninsular Malaysia individuals; Fig. 4). Cold Climate Specialists were the least A comparison of our fi ndings at TFR to the fi ndings of abundant and diverse with only one species (Prenolepis sp Malsch (2000) at PFR shows that similar ant communities two of NZA) and two individuals trapped. inhabit upper hill dipterocarp forests and lowland dipterocarp forests (Table 4). Only a higher percentage of Myrmicinae and Ponerinae were found at TFR as compared to PFR. The DISCUSSION small differences in the taxonomic structure of PFR and TFR might be due to the fact that different methods were used Species accumulation curves calculated from sampling in in this study as compared to the study conducted in PFR. 2008 and 2009 indicate that more sampling would need A second study conducted by Malsch (2000) that looked to be carried in order to adequately assess the entire ant specifi cally at the diversity of ground foraging ants in PFR community in TFR. However, when sampling from both using a standard quadrat sampling yielded 120 species, in years were combined, the species accumulation curve almost comparison to the 143 ground foraging ant species collected reached an asymptote, indicating that at least 40 plots are using ground pitfall trap, baiting and leaf litter sifting that needed to adequately sample a 200-ha area of forest. The were found in this study. Therefore, despite the rapid nature species estimation also indicates that the sampling effi ciency of our ant survey in TFR, we managed to capture about half falls above 70% throughout the study, which supports the of the overall ant species diversity of PFR and a greater effi ciency of the trapping methods used (Table 1). Finally, number of grounds foraging ant species. This suggests the similarity of the separate species accumulation curves for that pristine upper hill dipterocarp forests in Peninsular 2008 and 2009 demonstrates the repeatability and reliability Malaysia may be as diverse as lowland dipterocarp forests. of our sampling methods. However, in the future, a more intensive study should be carried out at TFR replicating the methods used by Malsch Comparison to previous studies. – Taking into account the et al. (2000) in order to capture more of the rare upper hill specimens trapped by all four sampling methods (ground forest specialist species. pitfall traps, arboreal pitfall traps, baiting and leaf litter sifting) in both sampling years, ants from the following nine Functional groups. – A large number of the species identifi ed subfamilies were identifi ed: Amblyoponinae, Cerapachyinae, (55 species) are categorised as Generalised Myrmicinae Ectatomminae, Formicinae, Myrmicinae, Dolichoderinae, (GM). GM ants occur abundantly and ubiquitously in primary Ponerinae, Proceratiinae and Pseudomyrmicinae. Bolton et forests throughout the tropics (Andersen, 2000). The most al. (2006) documented 21 subfamilies throughout the world, abundant species at our study site was Lophomyrmex bedoti with twelve subfamilies occurring in the Indo-Australian (Emery) in the subfamily Myrmicinae. This generalist region. A series of studies carried out in Malaysian Borneo, predator species is expected to be highly abundant in tropical indicated the presence of all 12 Indo-Australian subfamilies forests and is commonly found on the forest fl oor where they of ants (Hashimoto et al., 2001). Similarly, a comprehensive forage (Rigato, 1994). Their heterogeneous diets enable them study that sampled ants across multiple strata in Pasoh to survive in complex environments and dominate forest ant Forest Reserve (PFR), an unlogged lowland dipterocarp communities (Moffett, 1986). forest reserve in Peninsular Malaysia, also documented all 12 subfamilies (Malsch, 2000). The most diverse genus at TFR was Pheidole (42 species); these GM ants are also hyperdiverse and found to be abundant Our failure to document three of the 12 Indo-Australian in most tropical forest sites worldwide (Wilson, 2003). This subfamilies, Aenictinae, Dorylinae and Leptanillinae, which genus plays an important role in ecosystem functioning as it are composed of topsoil species, would likely have not contains ant species that range from omnivores to scavengers occurred at TFR if more thorough topsoil sampling had been to predators of other small insects. Eguchi (2001) reported conducted (Bolton, 1994; Malsch, 2000). Aenictinae and that there are at least 52 species of Pheidole on Borneo. Dorylinae are army ants, which are highly mobile and diffi cult Polyrhachis was the second most diverse genus (19 species) to catch (Wilson, 1964), while Leptanillinae are subterranean at TFR. This genus is diverse and commonly abundant, species and very rarely recorded (Bolton, 1990). This calls but these Subordinate Camponotini (SC) ants are generally attention to the importance of employing several trapping submissive to the Dominant Dolichoderinae (DD) and are methods simultaneously when attempting to catalogue the ant ecologically segregated from them due to their large body diversity of an area. Our results also highlight the potential limitations of attempting a rapid ant diversity assessment in sizes and nocturnal foraging (Andersen, 2000). DD diversity such a biologically diverse tropical forest. and abundance recorded in this study were low (three species; 1.4% of all individuals). DD prefer open and hot habitats, and Within the nine subfamilies found at TFR, we captured in undistributed tropical rainforests, this habitat is found in 211 species, half of the number of species found in PFR forest gaps and in the canopy layer. Hence, it was caught where a comprehensive inventory of the lowland primary less in the ground-level traps. More DD would likely have forest ant community was conducted (Malsch, 2000). It been collected, if the canopy had been fogged. The low is important to note that the sampling effort employed in abundance and richness of DD captured in TFR also imply the study at PFR was much greater and involved the use of that the forest’s structure is relatively intact. Winkler Extractors, the Berlese collection method, pitfall traps, baiting, the single rope technique and hand collection Tramp species of ants are closely associated with human over a 10-month time period. activity and often nest in human structures (Schultz & 186 THE RAFFLES BULLETIN OF ZOOLOGY 2011 Table 4. Taxonomic structure of ants in Temengor Forest Reserve (this study) and Pasoh Forest Reserve (Malsch, 2000). Subfamily % Species TFR % Species PFR % Genera TFR % Genera PFR Myrmicinae 52.6 40.7 46.6 39.0 Formicinae 23.7 28.2 20.7 18.2 Ponerinae 16.6 14.9 20.7 19.5 Dolichoderinae 3.3 5.9 3.4 6.5 Ectatomminae 1.9 0.4 3.4 1.3 Cerapachyinae 0.5 2.9 1.7 1.3 Proceratiinae 0.5 1.4 1.7 3.9 Pseudomyrmicinae 0.5 1.4 1.7 1.3 Amblyoponinae 0.5 1.2 1.7 5.2 Aenictinae 0.0 2.5 0.0 1.3 Dorylinae 0.0 0.2 0.0 1.3 Leptanillinae 0.0 0.2 0.0 1.3 Andersen, 2000); only two tramp species were found in ACKNOWLEDGEMENTS low abundances at TFR (Technomyrmex albipes (Forel) and Tetramorium pacifi cum (Mayr)). Even though these species The Government of Malaysia, Forest Research Institute are most likely native to Malaysia, their low abundance at Malaysia (FRIM), and the Conservation of Biodiversity the study site provides insight into TFR’s overall health. Project (CBioD) enabled us to conduct fi eld sampling through These species are classifi ed as Opportunists, meaning they co-funding by UNDP-GEF (MAL/04/G31) and ITTO [PD pre-dominate sites when disturbance limits the productivity 165/02 Rev.3 (F)]. Our key partners in the CBioD Project, of other species (Andersen et al., 2000). Finally, no invasive including the Perak Integrated Timber Complex, Perak State species were trapped, providing further support for the Development Corporation (SEDC), Forestry Department classifi cation of TFR as an undisturbed primary forest. Peninsular Malaysia Headquarters and State Forestry Departments of Peninsular Malaysia granted us permission New species. – Seven undetermined specimens were to conduct sampling in Peninsular Malaysia. This Project discovered at TFR, which possibly correspond to new species. was conducted in collaboration with Duke University and These seven undetermined specimens were only found in was also supported by the National Science Foundation of TFR, as compared to other undetermined species listed in the United States through a Graduate Research Fellowship Appendix 1, which were already present in the collection to Hannah M. W. Salim. We would like to thank the Head of Seiki Yamane at Kagoshima University. The fi nding of of Entomology Branch at FRIM, Laurence G. Kirton, for these species suggests that the upper hill dipterocarp forest his guidance and our fi eld assistants, Noor Azmy Mohd of TFR harbors a high diversity of ant species that have not Noor, Wan Shaina Mazuin, Mohd Shah Fadir Ishak and been thoroughly studied before. These specimens appear Nafaruding Che Nan for their hard work. We would also to be Cryptic Species, which are small and predominantly like to thank Prof. Seiki Yamane (Kagoshima University) and Myrmicinae and Ponerinae that nest and forage primarily Dr. Bakhtiar Effendy Yahya (University Malaysia Sabah) for in soil, litter and rotting logs. They are most diverse and their supervision of species identifi cation and involvement abundant in forested habitats and are a major part of the in improving this manuscript. leaf litter ant communities in tropical rainforests (Andersen, 2000). As Cryptic Species, they are able to avoid detection; their methods of crypsis include camouflage, nocturnal LITERATURE CITED activity, subterranean lifestyles, transparency and mimicry (Zuanon & Sazima, 2006). Abdullah, S. A. & N. Nakagoshi, 2006. Changes in landscape spatial pattern in the highly developing state of Selangor, Peninsular Conclusions. – Primary upper hill dipterocarp forests harbour Malaysia. Landscape and Urban Planning, 77(3): 263–275. diverse ant communities, and it is necessary to utilise a Agosti, D., M. Mohamed & C. Y. C. Arthur, 1994. Has the diversity combination of sampling methods to ensure the accurate of tropical ant fauna been underestimated? An indication from representation of an entire community. Our results can be leaf litter studies in a West Malaysian lowland rain forest. Tropical Biodiversity, 2(1): 270–275 used to compare the ant communities of primary upper hill dipterocarp forests in Peninsular Malaysia with dipterocarp Andersen, A. N., 1997. Functional groups and patterns of forests in other floristic zones and at other elevations. organization in North American ant communities: A comparison with Australia. Journal of Biogeography, 24: 433–460. Additionally, since our study provides the only description of the ant community in an undisturbed upper hill dipterocarp Andersen, A. N., 2000. A global ecology of rainforest ants: Functional forest in Peninsular Malaysia, our fi ndings can serve as a groups in relation to environmental stress and disturbance. In: baseline to understand post-disturbance changes to ant species Agosti, D., J. Majer, L. Alonso & T. Schultz (eds.), Measuring and Monitoring Biological Diversity: Standard Methods for composition and functional diversity. 187 Nur-Zati et al.: Ants in Peninsular Malaysia Ground-Living Ants. Smithsonian Institution Press, Washington Godfray, H. C. G., O. T. Lewis, J. Memmott, 1999. Studying insect D. C. Pp. 25–34. diversity in the tropics. Philosophical Transactions of The Royal Society Biological Sciences, 354(1391): 1811–1824. Andersen, A. N., B. D. Hoffmann, W. J. Muller & A. D. Griffi ths, 2002. Using ants as bioindicators in land management: Greenslade, P. J. M., 1972. Comparative ecology of four tropical Simplifying assessment of ant community responses. Journal ant species. Insectes Sociaux, 19: 195–212. of Applied Ecology, 39(1): 8–17. Hashimoto, Y., S. Yamane & M. Mohamed, 2001. How to design Andersen, A. N. & J. D. Majer, 2004. Ants show the way Down an inventory method for ground-level ants in tropical forests. Under: Invertebrates as bioindicators in land management. Nature and Human Activities, 6: 25–30. Frontiers in Ecology and the Environment, 2: 291–298. Hoffmann, B. D. & A. N. Andersen, 2003. Responses of ants to Bestelmeyer, B. T., D. Agosti, L. E. Alonso, C. R. F. Brandão, disturbance in Australia, with particular reference to functional groups. Austral Ecology, 28: 444–464. W. L. Brown Jr., J. H. C. Delabie, & R. Silvestre, 2000. Field techniques for the study of ground-dwelling ants: An overview, Kamaruzaman, M. P. & W. M. S. W. Ahmad, 2003. Forest harvesting description and evaluation. In: D.Agosti, J. Majer, L. E. Alonso practices towards achieving sustainable forest management in & T. Schultz (eds.), Ants: Standard Methods for Measuring Peninsula Malaysia. In: International Expert Meeting on the and Monitoring Biodiversity. Smithsonian Institution Press, Development and Implementation of National Codes of Practice Washington D. C. Pp. 122–144. for Forest Harvesting—Issues and Options. 17–20 Nov.2003. Bolton, B. 1990. The higher classifi cation of the ant subfamily Keeler, K. H., 1980. Distribution of plants with extrafl oral nectaries Leptanillinae (Hymenoptera: Formicidae). Systematic in temperate communities. American Midland Naturalist, 104: Entomology, 15:267–282. 274–280. Bolton, B., 1994. Identifi cation Guide to the Ant Genera of the Kindt, R. & R. Coe, 2005. Tree Diversity Analysis. A Manual and World. Harvard University Press, Cambridge, Massachusetts. Software for Some Common Statistical Methods for Biodiversity 232 pp. and Ecological Analysis. World Agroforestry Centre (ICRAF), Nairobi. 196 pp. Bolton, B., 1998. A preliminary analysis of the ants of Pasoh Forest Preserve. In: Lee, S. S., Y. M. Dan, I. D. Gauld & King, J. R., A. N. Andersen & A. D. Cutter, 1998. Ants as bioindicators of habitat disturbance: Validation of the functional J. Bishop (eds.), Conservation and Development of Forest group model for Australia’s humid tropics. Biodiversity and Resources. Forest Research Institute Malaysia, Kuala Lumpur. Conservation, 7: 1627–1638. Pp. 84–95. Liefke, C., W. H. O. Dorow, B. Hölldobler & U. Maschwitz, 1998. Bolton, B., G. Alpert, P. S. Ward & P. Naskrecki, 2006. Bolton’s Nesting and food resources of syntopic species of the ant genus Catalogue of Ants of the World: 1758–2005. CD-ROM. Harvard Polyrhachis (Hymenoptera, Formicidae) in West-Malaysia. University Press, Cambridge, Massachusetts. Insectes Sociaux, 45: 411–425. Brühl, C. A., G. Gunsalam & K. E. Linsenmair, 1998. Stratifi cation Liow, L. H., N. S. Sodhi & T. Elmquist, 2002. Bee diversity along a of ants (Hymenoptera, Formicidae) in a primary rain forest in disturbance gradient in Tropical Lowland Forests of South-East Sabah, Borneo. Journal of Tropical Ecology, 14: 285–297. Asia. Journal of Applied Ecology, 38(1): 180–192. Brühl, C.A., 2001. Leaf litter ant communities in tropical lowland Longino, J. T., J. Coddington & R. K. Colwell, 2002. The ant rain forests in Sabah, Malaysia: Effects of forest disturbance fauna of a tropical rain forest: Estimating species richness three and fragmentation. PhD-thesis. Julius-Maximilians-Universität, different ways. Ecology, 83(3): 689–702. Würzburg. Malsch, A., 2000. Investigation of the diversity of leaf-litter Chey, V. K., J. D. Holloway & M. R. Spreight, 1997. Diversity of inhabiting ants in Pasoh, Malaysia. In: Agosti, D., J. Majer, moths in forest plantations and natural forests in Sabah. Bulletin L. Alonso & T. Schultz, (eds.), Sampling Ground-Dwelling of Entomological Research, 87(4): 371–385. Ants: Case Studies from the Worlds’ Rain Forests (Bulletin Chung, A. Y. C., 1995. Common Lowland Rainforest Ants of Sabah. No. 18). School of Environmental Biology, Curtin University. Forestry Department, Sabah. 68 pp. Pp. 31–40. Delabie, J. H. C., B. L. Fisher, J. D. Majer & I. W. Wright, 2000. Manokaran, N., 1995. South-east Asian dipterocarp forest Sampling effort and choice of methods. In: Agosti, D., J. Majer, ecosystems. In: IUFRO XX World Congress sub plenary sessions L. Alonso & T. Schultz (eds.), Ants: Standard Methods for Caring for the Forest: Research in a Changing World. 6–12 Measuring and Monitoring Biodiversity. Smithsonian Institution Aug.1995. Tampere, Finland. Press, Washington D. C. Pp. 145–154. Moffett, M. W., 1986. Observations of Lophomyrmex ants from Kalimantan, Java and Malaysia. Malayan Nature Journal, 39: Eguchi, K., 2001. A revision of the Bornean species of the ant genus 207–211. Pheidole (Insects: Hymenoptera: Formicidae: Myrmicinae). Tropics Monograph Series, 2: 1–154. Moog, J., B. Fiala, M. Werner, A. Weissfl og, L. G. Saw & U. Maschwitz, 2003. Ant-plant diversity in Peninsular Malaysia, Fiala, B., H. Grunsky, U. Maschwitz & K. E. Linsenmair, 1994. with special reference to the Pasoh Forest Reserve. In: Okuda, Diversity of ant-plant interactions: Protective efficacy in T., N. Manokaran, Y. Matsumoto, K. Niiyama, S. C. Thomas Macaranga species with different degrees of ant association. & P. S. Ashton (eds.), Pasoh: Ecology of a Lowland Rain Oecologia, 97: 186–192. Forest of Southeast Asia. Springer-Verlag, Tokyo, New York. Fichter, E., 1941. Apparatus for the comparison of soil surface Pp. 459–494. Arthropod populations. Ecology, 22(3): 338–339. Oksanen, J., R. Kindt, P. Legendre, B. O’Hara, G. L. Simpson, Folgarait, P. J., 1998. Ant biodiversity and its relationship to P. Solymos, M. H. H. Stevens, & H. Wagner, 2009. Vegan: ecosystem functioning: A review. Biodiversity and Conservation, Community Ecology Package. R package version 1.15-4. http:// 7(9): 1221–1244. CRAN.R-project.org/package=vegan<http://cran.r-project. org/package=vegan> 188 THE RAFFLES BULLETIN OF ZOOLOGY 2011 Olson, D. M. 1991. A comparison of the effi cacy of litter sifting Seastedt, T. R. & D. A. Crossley Jr., 1984. The influence of and pitfall traps for sampling leaf litter ants (Hymenoptera, arthropods on ecosystems. BioScience, 34(3): 157–161. Formicidae) in a tropical wet forest, Costa Rica. Biotropica, Thang, H. C., 1987. Forest management systems for tropical high 23: 166–172. forest, with special reference to Peninsular Malaysia. Forest R Development Core Team, 2009. R: A Language and Environment Ecology and Management, 21(1): 3–20. for Statistical Computing. R Foundation for Statistical Wilson, E. O., 1964. The true army ants of the Indo-Australian Computing, Vienna, Austria. http://www.R-project.org/ area (Hymenoptera: Formicidae: Dorylinae). Pacifi c Insects, Rigato, F., 1994. Revision of the myrmicine ant genus Lophomyrmex, 6(3): 427–482. with a review of its taxonomic position (Hymenoptera: Wilson, E. O., 2003. Pheidole on the New World: A dominant, Formicidae). Systematic Entomology, 19: 47–60. hyperdiverse ant genus. Harvard University Press, Cambridge, Samson, D. A., E. A. Rickart & P. C. Gonzales, 1997. Ant diversity Massachusetts. 818 pp. and abundance along an elevational gradient in the Philippines. Wyatt-Smith, J., 1963. Manual of Malayan Silviculture for Inland Biotropica, 29: 349–363. Forests. Malayan Forest Record No. 23. Forest Research Sanders, N. J., 2002. Elevational gradients in ant species richness: Institute, Kepong. Area, geometry, and Rapoport’s rule. Ecography, 25(1): Zuanon, J. & I. Sazima, 2006. The almost invisible league: Crypsis 25–32. and association between minute fishes and shrimps as a Schultz, T. R. & A. N. Andersen, 2000. The interactions of ants possible defense against visually hunting predators. Neotropical with other organisms. In: Agosti, D., J. Majer, L. Alonso Ichthyology, 2006(2): 219–214. & T. Schultz (eds.), Ants: Standard Methods for Measuring and Monitoring Biodiversity. Smithsonian Institution Press, Washington D. C. Pp. 35–44. 189 Nur-Zati et al.: Ants in Peninsular Malaysia Appendix 1. The seven functional groups of ants, their abbreviations and their descriptions (adapted from Andersen, 2000). Functional Group Abbreviation Description Dominant Dolichoderinae DD Abundant and highly aggressive species that exert a strong competitive infl uence on other ants; favour hot and open habitats. Subordinate Camponotini SC Behaviourally submissive to DD; ecologically segregated from DD because of large size and nocturnal foraging Tropical-, Hot- & TCS, HCS, CCS Found in arid zones (HCS), humid tropics (TCS) or cool-temperate zones Cold-Climate Specialists (CCS); CCS and TCS are unspecialised, exist where DD are rare; HCS coexist with DD Cryptic Species C Small species; found in soil, litter, rotting logs; most diverse/abundant in forests Opportunists O Unspecialised, poorly competitive; predominate where stress or disturbance limit other ants Generalised Myrmicinae GM Ubiquitous in warmer regions; compete with DD Specialist Predators SP Medium to large size; only interact with other ants through direct predation because of special diets and small populations 190