JOURNAL OF TROPICAL BIOLOGY AND CONSERVATION 6 : 89 – 101, 2010 89 Ant diversity of Maliau Basin Conservation Area, Sabah, Malaysia Sukarman Sukimin, Maryati Mohamed and Hassan Aris Institute for Tropical Biology and Conservation, Universiti Malaysia Sabah, Locked Bag 2073, 88999 Kota Kinabalu, Sabah, Malaysia. ABSTRACT. The purpose of this study is to the ecosystem. Some ants are known as seed determine the diversity of ants in the Maliau dispersers, decomposers, nutrient recyclers Basin Conservation Area. Collection was and are also a source of food for other animals made using fi ve different sampling protocols: (e.g. birds and mammals). Some even exhibit Winkler Bag along 100m length transect, mutualistic relationship with plants and other Pitfall Trap made in a belt transect 100m animals (Holldobler & Wilson 1990; Alonso, length, Pitfall Trap in 5x5m plot, Manual in 2000; Hashimoto & Homathevi, 2003). 100m transect length, and Manual Random. A total of 210 morphospecies from 10 The ant fauna of Borneo is very diverse subfamilies were identifi ed to at least genus and unique. The island has nine subfamilies, level. The species composition showed a 94 genera and more than 1000 described comparatively high species diversity in species (Hashimoto & Homathevi, 2003). the subfamily Myrmicinae, followed by Worldwide, there are 16 subfamilies, ca. 300 Formicinae and Ponerinae. Manual collection genera and ca. 15,000 described species of in 100 m length recorded the highest number ants (Bolton, 1995). Thus, Borneo represents of species (96). Whereas, Pitfall Trap in 100m ca. 30% of ants genera and ca. 7% of ants length recorded the lowest number of species species, although Borneo covers less than 0.2% of the earth’s land surface. There are (43). The Winkler Bag was the most effective six genera of ants endemic to Borneo, namely in collecting individuals (2,992) followed by Bregmatomyrma, Epelysidris, Ishakidris, Pitfall Trap in 5x5m plot and Pitfall Trap in100 Loweriella, Secostruma, Tetheamyrma, m length (2,341). Generally, the Pitfall Trap (Hashimoto & Homathevi, 2003). recorded the highest number of individuals as in most of the previous studies. This area has Several studies done in Malaysia, lower ant abundance but is higher in species especially in Sabah, showed that ants could richness. be used as an indicator of changes in the environment. Maryati (1994) compared ant Keywords: Ants, diversity, Ginseng Camp, species composition and suggested that an area Maliau Basin. not ideal for ants, such as degraded forests or highlands, showed a high representation of INTRODUCTION species in the subfamily Ponerinae, followed by Formicinae. Ants are a major component of arthropod fauna in tropical rainforests in terms of species The objective of the research is to reveal richness and biomass, and play important roles the ant diversity in Maliau Basin Conservation as predators, prey, detritivores and mutualisms Area (MBCA), particularly around Ginseng with plants in the forest ecosystem (Holldobler Camp. As the ant diversity of MBCA has never & Wilson, 1990). This insect is not only diverse been studied previously, this study would give in species composition and high in abundance, the fi rst checklist of ants and its diversity in the but is also important for the functioning of area. 90 ANT FAUNA FROM GINSENG CAMP MATERIAL AND METHODS with diluted soapy water, ¼ way up into each cup. The soapy water was made by mixing Study site one teaspoon of dish washing liquid soap with 125ml of water. The soapy water was added to Three study sites were selected as sampling reduce surface tension so that ants that fall into sites: (1) Around Ginseng Camp (582 – 700 m the water will sink to the bottom of the cup. a.s.l.), (2) Along the trail of Ginseng Camp to Traps were made in the middle of each 5m that Maliau Waterfall (GCMW; 256 – 780 m a.s.l.) have been measured. Traps were left for 24 and (3) Along the trail of Ginseng Camp to hours. Ants were collected the next day using a Agathis Camp (GCAC; 460 – 700 m a.s.l.). tea sieve and left inside a coded vial with 75% ethanol. This process was repeated for three Ant Collection days. Four different baits (corned beef, tuna in soil, sweet/candy and control) were used. This Ants were collected using fi ve different method was used for collecting nocturnal and sampling protocols as follows: diurnal ants. Winkler Bag in 100 m length Pitfall Trap in 5x5m plot All leaf litter in the 5cm topsoil layer, which A smaller quadrate (5m x 5m) was marked. can be easily scraped from the more compact The small quadrate was divided into 25 squares soil (in one subplot 1m x 1m) was sieved each (1m x 1m). A hole was dug at the center (1cm x 1cm size). Branches and twigs were of each square. These traps were left overnight broken during the sieving process to collect and were collected the next day. ants nesting in them. The fi ne leaf litter from sifting was then placed in a debris bag and Hand collection in 100 m length taken to the base camp and then placed into a mesh bag (4mm x 4mm size). For extraction, A transect of 100m x 2m was set up and they were left in the Winkler’s Bag in a secure divided into sections of 5m each. Two persons location, sheltered from wind and rain. As then collected all ants seen within 30 minutes the debris dried up slowly, ants crept out of using fi ne tip forceps; one person collected the mesh and fell into the alcohol. Samples along the left side and another on the right were then removed from Winkler Bags after side. It took 10 hours for two persons to set the 72 hours. This time frame is recommended transect and collect samples. Collection was for collecting ants because by then, most of conducted from 9 am to 4 pm for two days. the species would have fallen into the 75% All ants collected were preserved in vials ethanol at the bottom of the bag (Bestelmayer containing 75% ethanol. et al., 2000; Ward, 2000). Manual Random Pitfall Trap in100 m length Ants were randomly collected using fi ne tip Pitfall trapping along a transect was conducted forceps. Specimens were placed into coded along a 100m x 2m belt transect. Each transect vial with 75% ethanol. This sampling protocol was divided into 20 sections, each 5m long. was used at all the sites. The other protocols The pitfall traps (using plastic cups) with were done only around Ginseng Camp. a diameter of 8cm and 15cm depth were installed in the ground, at the same level with the forest fl oor. The pitfall traps were fi lled SUKARMAN SUKIMIN ET AL. 91 RESULTS The diversity of ant communities based on sampling protocols used Ant Composition The TM (Manual in 100m transect) recorded the A total of 9,317 individuals of ants were highest species number (96 species) followed sampled in this study. This comprised of by MR (Manual Random) which recorded 210 species from 52 genera belonging to 95 species. The lowest number came from ten subfamilies (Table 1, Appendix 1; Plate the PT100m (Pitfall Trap in 100m transect) 1), namely Aenictinae, Amblyoponinae, protocols with only 43 species (Figure 1). Cerapachyinae, Dolichoderinae, Dorylinae, Ectatomminae, Formicinae, Myrmicinae, Myrmicinae, Formicinae and Ponerinae Ponerinae and Pseudomyrmicinae. The were dominant in term of species number in this study. Species composition of Myrmicinae Table 1. Number and % of genera and species of was sampled most successfully by PT5x5m each subfamily obtained for all sampling protocols. compared with other protocols. The species of family Formicinae were sampled most successfully by MR (Figure 2). Species of Ponerinae were sampled the most by the TM method (x2=42.405, df=7, p=0.000; Chi- Square test. Shannon-Wiener (H’) and Evenness (E) index The number of species recorded for each sampling protocols showed that TM recorded the highest number of species (96 species) with the value H’=3.93 (Table 3). The lowest number subfamilies with the highest representation of species recorded was by the PT100m (43 were Myrmicinae with 80 species, followed species). Interestingly the Shannon-Weiner by Formicinae (66 species) and Ponerinae index shows that PT100m had more diverse (35 species). These subfamilies dominated ant species than PT5x5m. This was due to samples collected, accounting for more than evenness: the community of ants from PT100m 84% of the total number of genera and 89% of was more even (E=0.62) than PT5x5m. Even the total species. The Dorylinae included only though the PT5x5m protocol recorded higher one species, and the subfamily Leptanillinae number of species (49) compared to PT100m was not recorded in this study. The taxonomic (42), the community of ants in PT5x5m arrangement and species name followed protocol was not even; this resulted in an H’ Bolton’s catalogue (1995) and Bolton (2003). value lower than PT100m. PT100m method was diverse because of the evenness of ants, After ranking (Table 2), four genera while the PT5x5m was diverse because of the included more than fi ve percent of all species species richness. (Polyrhachis: 27 species, Pheidole: 16 species, Crematogaster: 15 species and Camponotus: Figure 3 shows that the longest tail is 13 species). Twenty-two of 52 genera exhibited by TM, which indicates that the TM comprised only one species. sampling protocol is more diverse than other protocols in term of species richness. From 92 ANT FAUNA FROM GINSENG CAMP Table 2. The ant genera ranked after number of species. The proportion of species number of the genus to total species number (205 species) is given in %.