The Natural History Journal of Chulalongkorn University 3(1): 9-15, April 2003 ©2003 by Chulalongkorn University Sampling Biodiversity in Bornean Frogs Robert F. INGER * Department of Zoology, The Field Museum, Chicago, USA ABSTRACT.–Analysis of biodiversity is attracting increasing interest within the public arena, with many articles appearing in newspapers and popular magazines. This change is understandable because of the relationship of biodiversity to conservation in an increasingly stressed global environment. There is equal interest in biodiversity and related issues in scientific circles because of their relationship to questions in ecology, evolutionary biology, and conservation biology. The term biodiversity is often used as though it meant only the numbers of species (species richness), but the diversity of life is much more complex than that. The scientific literature on biodiversity is centered on three main subjects: lists of species occurring in a small region, comparisons of the species occurring in different environments within a region, and comparisons of the species occurring in different regions. Unfortunately, the scientific literature on biodiversity is frequently marred by errors in sampling. This paper will illustrate some aspects of biodiversity drawing upon work on the frogs of Borneo. My studies of the frogs of Borneo were designed to explore local and regional diversity, the impact of environmental differences on regional diversity, and the variation between species in the patterns of life cycles. Examples of these aspects of biodiversity are given. KEY WORDS: biodiversity; sampling; Borneo; frogs words, they mean local or regional species INTRODUCTION richness. But there are many more faces to biodiversity. In addition to species richness, Suddenly, it seems as if the entire world is biodiversity also encompasses diversity in talking about biodiversity. Certainly this is a habitat relations, in life cycles, and various very significant topic because it relates so aspects of behavior. The focus in this paper directly to our growing interest in conservation will be on the patterns within one group of in a stressed global environment. But what do animals in one general environment: frogs of we mean by biodiversity? Do we all mean the the rain forests of Borneo, a large island same thing? How do we study it? Are all the straddling the equator and having North-South studies of biodiversity useful? In this paper I and East-West dimensions greater than 1100 km. hope to provide some, at least partial, answers to those questions. By biodiversity, most people mean simply SAMPLING WEAKNESSES the number of species at some place. In other Before reporting the results of the work in * Corresponding author. Borneo, I wish to make a few comments about E-mail: [email protected] sampling weaknesses. My experience as a 10 NAT. HIST. J. CHULALONGKORN UNIV. 3(1), APRIL 2003 Number of frogs observed per night on Sg. Serbong 80 70 gs 60 o fr 50 f o r 40 e b 30 m u N 20 10 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 Night of observation FIGURE 1. Variation in numbers of frogs observed (all species combined) on 36 nights over the same 620-meter transect on Sungai Serbong, Nanga Tekalit, Sarawak, during the course of 12 months (1962/63). reviewer of manuscripts and reader of if sampling had been carried out on both slopes published papers tells me that many of the during the same period or on both slopes at published statements about biodiversity several times of the year? Unfortunately, this cannot be supported because of very poor type of sampling deficiency is not uncommon, sampling procedures. Let me cite a few rendering a portion of the literature on examples. biodiversity unreliable. A recent study of a nature reserve in Ignoring important sources of variation is Madagascar (Raxworthy et al., 1998) compiled another common sampling error. Several an inventory of the amphibians and reptiles on students at a university proudly told me about a the eastern and western slopes of a massif. The class exercise they had conducted to determine authors concluded that “ 30 species are restricted the differences in the frog species using ponds to a single slope only (between1000and1700m in two environments: open fields and adjacent elevation)...suggesting significant differences in rain forest. They showed me their results and community composition.” Sampling on the asked my opinion. I asked them how many eastern slope took place 18 October - 28 No- ponds they had sampled in each environment vember, 1994, while sampling on the western the answer was one! I then asked them if they slope took place 25 January - 11 February, thought their results would have been the same 1995. Amphibians are known to vary activity if they had sampled two ponds in each in response to variations in rainfall and environment. That question puzzled them. temperature, but the authors give no indication Unfortunately, it also puzzled their professor of these variables of weather. No one would who was standing next to us. These students challenge the observation that 30 species were and their professor had assumed that every not found on both slopes, but the conclusion pond in the forest would have the same species, that 30 species are “restricted to a single slope” that there was no variation within habitat types, must be challenged. We may ask if the that the species in each environment were observed differences between the species lists essentially uniformly distributed within environ- are artifacts of differences in sampling periods. ments, like jam spread over a slice of bread. Would the observed differences have appeared Although this example is drawn from my F. INGER – SAMPLING BIODIVERSITY IN BORNEAN FROGS 11 A Limnonectes leporinus number seen per night Sg. Serbong 20 18 16 s 14 g ro 12 f f 10 o . 8 o N 6 4 2 0 1 4 7 10 13 16 19 22 25 28 31 34 Night of observation B Pedostibes hosii number seen per night Sg. Serbong 25 20 s g o 15 r f f o . 10 o N 5 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 Night of observation FIGURE 2. Numbers of individuals of each of two species seen on Sungai Serbong, Nanga Tekalit, Sarawak, on 36 nights spread over 12 months (1962/63). A. Limnonectes leporinus (Andersson). B. Pedostibes hosii (Boulenger). personal experience, it can be duplicated, three nights within a five day interval, or some unfortunately, by looking at the published similar short period. Papers of this sort assume literature. that all species of frogs present in the region are A common type of publication is a paper active on all nights, without variation between listing species of frogs (or other animals) species in activity patterns.However,itis known occurring in some small region, a national that some species of frogs call and breed, and wildlife reserve or an administrative district. are therefore exposed to observation, only at Reading the methods section of such papers, the beginning of a rainy season whereas others one often sees that the authors collected on call and breed over most of the season (vide 12 NAT. HIST. J. CHULALONGKORN UNIV. 3(1), APRIL 2003 Heyer, 1973). Even in some aseasonal rain large portions of the forest. Field studies of the forest environments, breeding activity of some frog fauna were designed to explore local and species is episodic whereas it is continuous in regional diversity, the impact of environmental others (Inger and Tan, in prep.). An example of differences on regional diversity, fluctuations in variation in activity of frogs in an aseasonal local population sizes, and diversity in life climate is shown in Figure 1. The data in cycles and behavior. Most of my data come Figure 1 include all species of frogs observed from 10 sites distributed in Sarawak and Sabah; on a 600-meter transect along an 8-m wide sampling at each of these sites extended for at stream in Sarawak, Borneo, over a period of least four weeks (Inger and Voris, 1993), with oneyear. Differencesbetweenspecies in activity additional sampling at four additional sites. patterns along the stream dealt with in Figure.1 Local and regional species diversity.–Work are sharp(Fig.2). A sampling period of three was carried out at Nanga Tekalit, Sarawak nights along the Sungai Serbong, Sarawak, (1°38'N/113°34'E), a site of lowland (100-230 would certainly include Limnonectes leporinus m), hilly primary rain forest at three periods: but probably omit Pedostibes hosii from the 366 days in 1962/63, 93 days in 1970, and 30 fauna of that stream. Clearly, to estimate species days in 1984. Fifty four species of frogs were richness of any tropical region of any size it is observed during 1962/63. From the species essential that the sampling period be designed to accumulation curve for that year (Fig. 3), it is cover interspecific variation in activity patterns obvious that additional species of frogs were to and seasonal patterns in weather. be expected. During the shorter period of field work in 1970, 45 species of frogs were found, including six species that were not seen in EXPLORING BIODIVERSITY OF FROGS OF 1962/63. In the still shorter period of work in BORNEO 1984, only 35 species were observed, all of them species that had been seen in 1962/63 but Borneo is a very large island, approximately three that had not been seen in 1970. 1200 km from north to south and from east to Field work was also carried out at Segaham, west, with great variation in topography and Sarawak (2°44'N/113°55'E), an area of hilly maximum elevation of 4100 m. The island primary rain forest with slightly steeper straddles the equator and climate is essentially topography than that at Nanga Tekalit, 128 km aseasonal. Until relatively recently, the entire distant. During the 61 days at Segaham, 47 island was covered with evergreen forest, species of frogs were found, all but four of though economic development has now cleared TABLE 1. Relation between topography and breeding patterns of frogs at Bornean localities. The number of days of sampling were: Labang–128 days; Segaham–61 days; Nanga Tekalit–366 days. Geographic coordinates of the three localities are given in the text. Locality Topography Breeding microhabitat Streams In current1 Out of current2 Ponds3 Uncertain No. of species Labangflat 0 10 23 1 Segaham Steep, hilly 10 14 14 5 N. Tekalit hilly 13 19 24 5 Notes : 1 In riffles and torrents. 2 In side pools, leaf drifts, and bank potholes. 3 Pools on forest floor as well as water containing tree holes. F. INGER – SAMPLING BIODIVERSITY IN BORNEAN FROGS 13 Species accumulation of frogs at Nanga Tekalit, 1962 /63 60 s 50 g o r f f 40 o es 30 ci e p 20 s . o N 10 0 1 9 7 5 3 1 9 7 5 3 1 9 7 5 3 1 9 7 5 1 3 5 7 9 0 2 4 6 8 9 1 3 5 7 8 0 2 1 1 1 1 1 1 2 2 2 2 2 3 3 Cumulative days FIGURE 3. Species accumulation curve of frogs observed at Nanga Tekalit, Sarawak, during 12 months in 1962/63. Relative abundances Sg. Serbong 0.3 e 0.25 1962 c n 1970 da 0.2 n 1984 u ab 0.15 e v ati 0.1 el R 0.05 0 1 2 3 4 5 6 7 8 Species (see legend) FIGURE 4. Year to year fluctuation in observed relative abundances of eight dominant species of frogs on Sungai Serbong, Nanga Tekalit, Sarawak. Adapted from Voris and Inger 1995. them also found at Nanga Tekalit. Significantly, 77 km distant. The forest at Labang is the samples from the two areas shared all the developed on very flat land <100 m asl and the species that are known to breed in streams streams there have very slow current and silty having moderate to strong currents. The bottoms. During the 128 days spent at Labang, significance of topography on the species 35 species of frogs were found, none of them composition of these faunas is evident when one species that breed at riffles or torrents; on the compares the frogs found at Segaham with those other hand, species with these habits were found at Labang, Sarawak (3°21'N/113°27'E), common at Nanga Tekalit and Segaham (Inger 14 NAT. HIST. J. CHULALONGKORN UNIV. 3(1), APRIL 2003 and Voris, 1993). Two-thirds of the species also important for assessing the likelihood of found at Labang breed in rain pools on the long range dispersal affecting geographic forest floor, but less than a third of the species distribution of species (Inger and Voris, 2001). found at Segaham (Table 1). Movements of individual frogs of a number of Fluctuations in numbers.–At Nanga Tekalit species have been explored at Nanga Tekalit, three streams (8-15 m wide) were set aside as Sarawak. “experimental” streams, and frogs were not The home ranges of several large (snout- removed from those streams during 1962/63. vent > 80 mm) species of frogs –Limnonectes Frogs were recorded over a stretch of 600 m on ibanorum, and L. leporinus –at Nanga Tekalit 36 nights on each of these three streams. In were small: median values 5.2-9.8 m for both each of the next two periods of work at Nanga species (Inger, 1969). These ranges were Tekalit, each of these streams was surveyed determined by observations of marked five nights; specimens were removed and individuals made over a period of 12 months. preserved during these two years. Although no For these two species, the net movement (that two streams can be exactly identical in channel is, the distance between the points of first and shape, bottom types, etc., the same species of last captures) for those individuals we had frogs dominated the assemblages on all three under observation for at least six months (180 streams, suggesting that the environments were days) varied from 1 to 151 meters, with 10 of generally similar. The similarity in terms of 19 being less than 11 m. In contrast, net the species occurring on the streams was quite movements of 18 individuals of Bufo asper high–0.73-0.89 (1.0 = complete identity) –for (snoutvent of adults 80-125 mm) under obser- between stream, within year comparisons. vation for at least 180 days were larger, with Indeed, the similarity was as high as the only five of 18 less than 11 m and five ranging between year, within stream comparisons –0.73- from 134 to 465 m. 0.90 (Inger and Voris, 1993). Both Limnonectes ibanorum and L. The relative abundances of species on these leporinus are known only from Borneo, streams was also quite similar. The between whereas Bufo asper is known from Borneo, stream, within year similarities were 0.68-0.96 Sumatra, and parts of Southeast Asia at least as and the between year, within stream similarities far north as 15°N in western Thailand. The 0.72-0.88 (Inger and Voris, 1993). However, greater movement of individuals of B. asper variation in relative abundances within species may have enabled this species to disperse over time was significant (Voris and Inger, between the continent and Borneo during the 1995). Figure 4. shows this temporal variation relatively short periods of sea regression during in eight common species on one stream. the Pleistocene (see maps in Voris, 2000). Movements of individuals.–The fluctuations in abundances shown in Figure 4. are strong SUMMARY indication that over long periods of time local populations of any species may go extinct. My conclusion is a simple one. It is However, the existence of some species at possible to obtain data that will permit one to many sites scattered over large areas of Borneo make statements about biodiversity that suggest that either local extinctions have not encompass local and regional species richness, affected these species at those sites or that behavioral diversity, and even historical immigration from other sites has restored any diversity. However, the validity of those populations that may have gone extinct in the statements, and the degree of confidence one past. If the latter phenomenon has occurred, may have in them, depend entirely on the then extent of movements of individuals is an appropriateness of the sampling procedures. It important factor. Information on movements is follows, of course, that conservation policies F. INGER – SAMPLING BIODIVERSITY IN BORNEAN FROGS 15 and actions also depend on the quality of our Raxworthy, C. J., F. Andreone, R. A. Nussbaum, sampling procedures. N. Rabibisoa and H. Randriamahazo. 1998. Amphibians and reptiles of the Anjanaharibe-Sud LITERATURE CITED Massif, Madagascar: Elevational distribution and regional endemicity. In A floral and faunal Heyer, W. R. 1973. Ecological interactions of frog inventory of the Rserve Spciale d’Anjanaharibe- larvae at a seasonal tropical location in Thailand. Sud, Madagascar: With reference to elevational J. Herpetol. 7: 337-361. variation. ed: S. M. Goodman. Fieldiana: Zool. Inger, R. F. 1969. Organization of communities of 90: 79-92. frogs along small rain forest streams in Sarawak. Voris, H. K. 2000. Maps of Pleistocene sea levels J. Animal Ecol. 38: 123-148. in Southeast Asia: Shorelines, river systems, Inger, R. F. and H. K. Voris. 1993. A comparison time durations. J. Biogeogr. 27: 1153-1167. of amphibian communities through time and Voris, H. K. and R. F. Inger. 1995. Frog abundances from place to place in Bornean forests. J. along streams in Bornean forests. Conserv. Biol. Tropical Ecol. 9: 409-433. 9: 679-683. Inger, R. F. and H. K. Voris. 2001. The biogeo- graphical relations of the frogs and snakes of Received: 9 December 2002 Sundaland. J. Biogeogr. 28: 863-891. Accepted: 6 Febuary 2003