Volume 10 2004 No. 1
ECOTROPICA 10: 1–14, 2004 © Society for Tropical Ecology

Mark-Oliver Rödel 1,2 & Raffael Ernst 2

University of Mainz, Institute of Zoology, Department of Ecology, Saarstrasse 21, D-55099 Mainz, Germany 2 Theodor-Boveri-Institute, Biocenter of the University, Department of Animal Ecology and Tropical Biology (Zoology III), Am Hubland, D-97074 Würzburg, Germany

Abstract : The need for standardization of field methods within and across studies has been recognized by the majority of ecologists throughout the world. However, comparable studies based on a standardized protocol are still scarce. We provide a guideline for effectively sampling and monitoring tropical forest amphibians and give recommendations for standardization. Based on a four-year study on amphibians in Taï National Park, Côte d’Ivoire, we evaluate commonly used techniques, and offer a catalogue of efficient techniques, along with suggestions for improvement of particular methods, with regard to study objectives and specific amphibian guilds. For simple short-term surveys we recommend visual and acoustic encounter surveys, accompanied by opportunistic trapping. The transect design introduced here proved to be most adequate for representative sampling and appeared to be appropriate for most studies that involve multivariate data. It is especially useful for long-term studies. Transects furthermore provide an effective method of investigating at least leaf litter frogs, not only at their breeding sites but also throughout the whole range of habitats used by them, thus generating a much more complete picture of an amphibian community than is possible with other methods. Accepted 4 February 2004. Key words : Tropical amphibian monitoring, standardization, method evaluation, Taï National Park, West Africa.

In theory there is no disagreement about the necessity to standardize ecological field methods and data acquisition in order to guarantee comparability between different studies, as well as to enhance the power of predictions resulting therefrom. The need for standardization of techniques across and within studies has continuously been emphasized by several authors (e.g., Heyer et al. 1994, Adis et al. 1998). Those who are involved in community ecological field research know well that this is a crucial factor when it comes to generalizing results in order to make them
* e-mail:

accessible to those who urgently need scientific guidelines to back up their practical efforts. This becomes especially important when looking at phenomena such as the well known world-wide amphibian decline, which seems to have affected even populations in pristine habitats and continues to affect or even wipe out whole populations with terrifying rapidity (Houlahan et al. 2000, Kiesecker et al. 2001, Pounds 2001). In these cases synergistic efforts are urgently needed, calling for unified methods and data output (Alford & Richards 1999, Parris 1999, DAPTF 2002). However, in reality there still is a lack of studies that can be compared without reservations, especially studies that were conducted in different parts of the world, although precisely this kind of comparison is 1

The first. More detailed descriptions of TNP are provided by Rompaey (1993). The maximum distance between transects was 6. Rodda et al. 5°50’N. iii) is efficient concerning time and money. Transects were arranged in pairs. Vonesh 2001. Our study sites are either pristine forest areas or have not been logged or cultivated since 1978. and iv) has a low environmental impact. Doan 2003) or total removal plots (Barbault & Trefaut Rodrigues 1978. The second. situated in western Côte d’Ivoire (5°08’6°07’N. 1979a. known as quadrat sampling (plot design). Selection of sites and design of sampling units We chose sites within existing macrohabitats (primary/ secondary forest) to establish a total of ten rectangular transects (six in primary forest. 2001) were a priori excluded from testing. All evaluations and recommendations are based on a study of tropical amphibian assemblages conducted over four years (1999–2002) in Taï National Park. on the basis of our experience in the field as well as with consecutive data analyses. The mean annual temperature is about 25° C (Rompaey 1993). four in secondary forest). that stretches west of the Dahomey Gap from Ghana into Sierra Leone and Guinea (Schiøtz 1967. Parts of the habitats on the park’s periphery were subject to more or less intense logging and cultivation until 1998 (P. The starting coordinate for each transect always marked the southeast corner to ensure identical geographic orientation between sampling units. We avoided extensive cutting and thus manipulation of important habitat features. Precipitation is distributed across two more or less distinct periods. known as transect sampling. PACPNT 2000). Our study site was located 23 km southeast of the small town of Taï and comprised about 30 km2 of primary and secondary rain forest around the Station de Recherche en Ecologie de Taï (SRET. Each had a north-south extension of 200 m and an east-west extension of 100 m. The rectangular transect design is a combination of two widely used standard techniques. map “Projet OMS Forêt de Taï”). Parren & de Graaf 1995. followed by a short dry season in July/August. 7°20’W). Given the lack of time for the development of effective conservation programs. We report on our experiences with already existing methods and with methods that we thought needed modifications to enhance their effectiveness and to enable the use of new statistical analysis methods. Richards 1996). such as standard plot sampling (Allmon 1991. It can be characterized as humid tropic seasonal (Riezebos et al. proved to be most effective for forest amphibians. The intention of this paper is thus to provide a guideline to choosing effective and comparable methods for the monitoring of tropical forest amphibian assemblages in general.RÖDEL & ERNST essential for academic and practical purposes alike. that ii) is easy to handle. For data acquisition the complete transect length of 600 m was subdivided in 25 m subunits (24 subunits/transect). transect sampling can provide similar data across habitat or dis- MATERIAL AND METHODS Study site The study was conducted in Taï National Park (TNP). Mean annual precipitation in the study area was 1806 mm during 1988–2002 (± 297mm. TNP is part of the Upper Guinea forest block. A number of recently published regional field monitoring manuals for amphibian surveys (Latin America: Lips et al. Poynton 1999). We aim to present a methodology that i) provides quantitative and qualitative data useful for rapid surveys as well as long-term monitoring programs. Floristically the TNP belongs to the Guinean-Congo-Region (Guillaumet 1967. b. data: Taï Monkey Project). Daily tempe2 . Every subunit was marked with numbered colored flag-tape. as well as their relative abundances and densities. Formenty pers. (1994) and PACPNT (2000). thus ensuring that all habitat types of a certain area within the very inhomogeneous forest were covered. also for para-ecologists. ratures vary between 20 and 33° C. The minimum distance between adjacent transects was 200 m. Riezebos et al. Transect paths were kept open so that walking at a constant speed was possible at all times. The major rainy season stretches from September to October. 1994. Africa: Howell 2002) reflect this need. as they may result in severe disturbance to the system under investigation. and unpubl. 2001. We thereby evaluated several methods that are already commonly used and present a set of techniques that. standardization of methods that yield broad-scale comparative data is now needed more than ever. consists of a series of small squares (quadrats) that are laid out at randomly selected sites within a habitat then thoroughly searched for amphibians.3 km. Thus widely used techniques. Lawson 1986. Côte d’Ivoire. The core dry season lasts from November until February/March. TNP is the largest remaining protected area of rain forest in West Africa. Whereas quadrat sampling can be used to determine the species present in a homogenous area. comm. A minor rainy season lasts from March/April to July. 6°47’-7°25’W). uses linear transects instead of squares.

surface and depth. Repeated controls of identical transects on consecutive days were avoided to ensure independence of samples.or age-specific differences in habitat use.g. Sampling methods and sampling effort Habitat characterization. rectangular ones provide the possibility of easily up. These parameters included vegetation density in four strata (canopy: > 20 m. Edaphic parameters were registered as general substrate types according to Lieberoth (1982).. both in time and space.or downscaling data without neglecting local habitat diversity (e. the vegetation of all 25 m segments within a distance of about 100 cm left and right of the transect was recorded by counting the number of plants belonging to a certain category (diameter at breast height. thus providing data that are useful to test for sex. we characterized all habitats using several variables that were recorded at two defined points within each 25 m subunit (beginning and midpoint). In addition. for comparison of local with regional species pools). understory: < 0. dbh). We assumed that the number of plants with small dbh is greater in degraded. substrate moisture was determined in four categories during every transect walk. 1996. Specimens heard but not seen within this distance were not searched for and not included in the SVTS data sets (see below). Standardized visual transect sampling (SVTS). Pearman 1997). In order to quantify the availability of potential aquatic breeding sites. Sampling was interrupted for the duration of the recording in order not to overestimate locations in which animals had previously been captured. bush and shrub stratum: 0. divided into seven categories corresponding to particular densities. dbh = diameter of plants at breast height. To avoid duplicate recordings.. to assign them to particular guilds).AMPHIBIAN MONITORING IN TROPICAL FORESTS turbance gradients (Jaeger 1994. Categories of habitat variables measured on transects. whereas primary for- ests show increasing numbers of plants of larger dbh (Chatelain et al.g. In addition to species identity we recorded sex and snout-vent-length (SVL) of every individual. and to investigate correlations of species assemblages with environmental variables. Compared to linear transects.5 2 2.35 m/s). every aquatic habitat (lentic and lotic) located at a maximum distance of 25 m from either side of the transect was recorded with respect to type. To assess habitat preferences of particular species (e. Transects were intensively patrolled at a constant speed (0. Sampling was performed independent of prevailing weather conditions. lower tree stratum: 3–10 m. definition category 1 1.5 m). Jaeger & Inger 1994). vegetation density was measured in four strata (see text). Definitions of habitat variables are summarized in Table 1.5 m. thereby recording all amphibians within a distance of 100 cm from either side of the path. A simplified method originally developed by Braun-Blanquet (1964) for determination of vegetation coverage in vegetation analyses was used to estimate the percentage of leaf litter coverage. Usually four to six transects per day were sampled during daytime.30–0. secondary forests.5–1. captured frogs were marked by TABLE 1. A maximum of four transects was patrolled at night.5 3 3.5 4 5 6 7 vegetation density absent transition gaps predominating transition closed areas predominating transition closed substrate types forest soil arenaceous forest soil loamy soil arenaceous soil sabulose soil muddy soil swampy soil dbh [cm] 0–5 6–10 11–20 21–50 > 50 leaf cover (%) 0–20 21–40 41–60 61–80 81–100 3 . For a discussion of the advantages of rectangular sample units over quadratic or circular ones see Krebs (1989) and McCune & Grace (2002). As far as possible all individuals were captured. Additionally.

In those cases. Dodd 1991). Recaptures were excluded from the analyses. despite their potential abundance. VES can only provide information on the presence or absence of a species in an area but are inadequate for determining abundances. For practical reasons “man-hours” can be used. Pitfall traps with drift fences were installed in two of the primary forest transects from March to April 1999. data can be expressed in numbers of individuals per time and surface units. This technique has been formalized as the time-constrained technique of Campbell & Christman (1982).g. Thus. and additionally in only one of these transects from January 2001 to September 2002. Anurans that are strong jumpers (e. Drift fences consisted of durable green plastic gauze. Likewise microhabitat description was not always possible. and as the time-constrained searches of Corn & Bury (1990). which can be adjusted to the complexity of the habitats being sampled. As opposed to visual sampling. This device is useful to determine species richness of epigaeic organisms (Corn 1994). Throughout transect walks a combination of visual and acoustic techniques was applied at all times. this behavior can be exploited for acoustic monitoring. species composition. during SATS it was often impossible to determine individual parameters. Audio strip transects represent a commonly used method for acoustic monitoring (Zimmerman 1994). Total VES. Coding schemes used for individual recognition were not applied. 0.. This corresponds to 765 transect walks. 8. Calls from greater distances cannot be unambiguously identified and the chance of duplicate recordings in neighboring segments increases. may be underestimated when exclusively using visual techniques. VES and AES are frequently used for rapid assessments and the evaluation of larger areas.5 m to either side of the transect was defined. Ptychadena species) are more difficult to trap than terrestrial species that lack these abilities (e. From February 1999 to December 2000 we compiled 382. The resulting data are expressed in numbers of individuals of a certain species found in an area per unit time. 1994. In the transect corners. In concordance with these authors the methods were only used as a qualitative or semi-quantitative tool within the scope of our studies. Bufo species). thus creating 25 m x 25 m acoustic sampling plots.RÖDEL & ERNST toe clipping (Donnelly et al.e. Standardized acoustic transect sampling (SATS). 12 and 20). beginning with segment 1: segment 1.. An array of fences and traps consisted of a central trap (buckets: .g. Henle et al. Individuals below nine mm SVL were not marked due to their small size. and also for evaluating if species that are rarely found on a transect belong to a particular local species pool or represent single migrating specimens that have habitat preferences not covered by the transects.5 m high and stapled vertically onto wooden stakes.. The width of the acoustic transect depends on the ability to detect each species’ advertisement call. Removal of additional toes may decrease the recapture rate of marked individuals by more than 6 –18 % for each additional toe removed after the first (Parris & McCarthy 2001). and to estimate relative abundances of species within an assemblage (Crump & Scott 1994). For both SVTS and SATS. other than sex and species.5 4 hours of transect sampling. Due to the simplicity of the method. According to Corn & Bury (1990). keeping the transect routine identical for both techniques.and AES-effort was kept comparable to the time spent on SVTS and SATS surveys throughout the entire study period. VES and AES thus could be used to evaluate if the randomly chosen transects represented the regional species pool. Capture success may vary greatly between species (Corn & Bury 1990. At capturing sites a thorough description of the microhabitat was recorded. Counts can be used to estimate relative abundance of calling males. habitat analyses can be based on data obtained from the general habitat characterization for single transects or segments. as well as breeding habitat use and breeding phenology of species (Zimmerman 1994). only calls coming from the right-hand side in the first segments were registered (i. An area or habitat is searched systematically for individuals in a defined time period. Furthermore. We used these methods continuously in all habitats of the study area including the transect areas. Visual (VES) and acoustic (AES) encounter surveys. 1997). this technique allows the detecting of cryptic species that. following the characterization routine used for general habitat description. Trapping with pitfall or funnel traps along drift fences. individual marking should be restricted to studies in which individual recognition is indispensable. It can be used to determine the species richness of an area and the species composition of a local assemblage. For that reason a maximum recording distance of 12. Since in the majority of frog species males use species-specific calls to advertise their position to potential mates and rivals (Wells 1977).

funnel traps (see Branch & Rödel 2003) were installed in a primary forest transect from 18 to 27 September 2002. They were checked at least every morning. SVTS = species recorded during visual transect walks. 220 mm bottom diameter) and two triangular fence segments (total length 16 m). data). In addition. Trap = pitfall and funnel traps. Pitfall trapping time summed to a total of approximately 4000 trap-days. unpubl. Funnel traps summed to a total of 384 trap-days. and in a forest fragment (Paulé-Oula 2) outside TNP from 30 August to 12 September 2002 (A. one on either side. Duct tape was used to reduce the diameter of the buckets and construct funnel-like openings in order to impede escaping from traps. Sampling success is referred to as number of species within families recorded. fossorial and aquatic species independent of families. results were analyzed for leaf litter. Data from each sampling method were compared with each other and to a list of species records that we gathered through January 1999 to September 2002 for the forest parts of the SRET region (Appendix 1). TaïF = all species recorded with all methods in forest habitats around SRET station. The data from pitfall and funnel traps can be expressed as number of individuals per trap-day. using a particular sampling method. which are situated south and west of TNP respectively (Rödel & Branch 2002. as sampling success and efficiency may vary with regard to the particular biology of a species. 285 mm top diameter. Each segment was tightened around a plastic bucket with an opening angle of 45°. FIG. we applied that method during a herpetological survey of the Haute Dodo and Cavally forests. Number of species per family recorded using different sampling methods. 1. In the course of this study. Additionally.. The ends of each segment were flanked with additional plastic buckets. arboreal. Branch & Rödel 2003). SATS = species recorded during acoustic transect walks. TC = species recorded during transect walks (SVTS and SATS combined).AMPHIBIAN MONITORING IN TROPICAL FORESTS 275 mm deep. AES/VES = visual and acoustic encounter surveys (only species that have been recorded in forest habitats around SRET station). Numbers above bars represent total number of species recorded with the respective method. 5 Ta ïF . Hillers et al. Traps were checked at least on a daily basis.

regularis. Rödel et al. Trap). H. 2. Phrynobatrachus calcaratus. compare Appendix 1). H. 2). two species were each listed in two guilds (compare Appendix). see Appendix 1). Phlyctimantis boulengeri ). p = 0.002. Four arboreal frogs were exclusively recorded with these methods (Hyperolius picturatus. mascareniensis. 2003. Phrynobatrachus fraterculus.. Bufo superciliaris) or simply absence in the area under investigation (e.g. 56 amphibian species are known to occur in TNP (Schiøtz 1967. RESULTS In total. rather than actual sampling method insufficiency. 2. 1.7 %) of nine families. TC. Seven species (Bufo maculatus. Rödel & Ernst 2000. VES/AES revealed 39 forest species (90. . Number of species per amphibian guild recorded using different sampling methods. SATS) we recorded a total of 15007 individual amphibians belonging to 37 species (86. Hyperolius guttulatus). 1. and between different amphibian families or guilds (Figs. SVTS.199. χ2 = 19. Hoplobatrachus occipitalis. and P. Ptychadena pumilio. Perret 1988. Numbers above bars represent total number of species recorded with the respective method. Effectiveness in detecting species within given families differed significantly between methods (Friedman test. For abbreviations see Fig.. Sampling efficiency in families containing higher numbers of genera and/or species was only marginally lower. Three leaf litter species (Amnirana occidentalis.g.3 % of TNP species). N = 10. thus making overall sampling success very high (Tab. 2002. 43 species were recorded in forest habitats within the study region (Fig.g.) or pitfalls only (Geotrypetes seraphini). nienokouensis. Two fossorial species were exclusively detected either with VES and pitfall traps (Hemisus sp. bibroni. Taking all methods into account.2 %) were recorded with pitfall and funnel trapping.. annulatus) were exclusively recorded during transect walks.) were recorded exclusively outside forest habitats (e. P.RÖDEL & ERNST FIG. clearing around SRET) and thus ignored in the analysis. AES/VES. Rödel 2000. Appendix 1). B. 1. fusciventris.0 % of TaïF) in eight families. tested for TaïF. SATS. P. During transect sampling (SVTS. P. Failure to detect particular species in this region was most probably due to their general scarcity (e. df = 5. Sampling success and data quality varied between different methods. sp. Using all methods we recorded 50 species in the region of the SRET (89. Families containing only 6 a few species in the TNP area showed the highest recording rates. Ten species (23.

7 0.8 0.47 specimens per trap-day in contrast to 0.0 36. No additional species were detected with funnel traps. SVTS.0 100. but very low in arboreal species (6. A distance of 1 m (visual) and 12. Trap). tested for TaïF. VES/AES.7 % of recorded arboreal species).5 m (acoustic) proved to work equally well in all habitats investigated.0 0.0 86. quantitative data for all arboreal species were best assembled with SATS.3 We could not detect a statistical difference in sampling efficiency between different amphibian guilds using the methods employed (Friedman test.0 81.0 100.0 0. nocturnal leaf litter frogs (arthloeptids.7 100. presence and abundance of purely aquatic species and tadpoles of all species were underestimated with all methods used here. the fossorial Geotrypetes seraphini was recorded with pitfall traps only.0 100. Likewise.0 0. but that concerned mostly single specimens.0 AES/VES 0. For other abbreviations see text and Fig.0 100. p = 0. Percentage of species within an amphibian family or guild that were recorded using a particular method. Most arboreal frogs could be recorded with SVTS.0 0. asytlosternids) were best recorded with SATS. in a primary forest transect with none in the forest fragment (Hillers et al.0 100.0 80.067.8 0.0 100.5 100. We only captured a few leaf litter frogs. this was most probably due to the small overall sample size and high heterogeneity of species numbers in families and guilds (Bortz et al. Schaich & Hamerle post hoc multiple comparisons revealed no significant differences between methods at α = 0.0 100.0 100.294. unpubl.03 specimens per trap-day in pitfall traps).7 0. TC.0 76. Reference is the number of species that were recorded using all methods in forested habitats around the SRET station (TaïF).0 0.0 0.0 100.0 100.0 77. In Haute Dodo and Cavally forests we captured 38 % of the recorded amphibian species (16 of 42) with a combination of pitfall and funnel traps.g.0 68. In contrast.0 0.0 SATS 0. data). 7 .0 100. 1990).0 100.0 77. Most specimens were captured with funnel traps (0.AMPHIBIAN MONITORING IN TROPICAL FORESTS TABLE 2. which is mute. Family / guild Caecilidae Pipidae Bufonidae Hemisotidae Ranidae Petropedetidae Astylosternidae Arthroleptidae Hyperoliidae Rhacophoridae leaf litter arboreal fossorial aquatic TaïF 1 1 2 1 5 11 1 3 17 1 22 18 2 3 TC 0.0 100.5 % of recorded leaf litter frogs).0 100.0 100.0 100.) vegetation density. χ2 = 10. In TNP funnel traps were ineffective.0 33.0 50. As in TNP. other than based on time units. As aquatic sites were only taken into account. Diurnal ground-dwelling frogs were best recorded quantitatively using SVTS.05.8 100.0 100. Branch & Rödel 2003).4 0.0 100.0 Trap 100.0 50.0 76. The traps proved to be inadequate in qualitative and quantitative sampling since even small species such as Phrynobatrachus villiersi (SVL: 10–15 mm) managed to escape regularly (several direct observations).0 66.0 100.0 100. SATS.0 100. Quantification of these data.0 100. when being a part of the transect line itself.5 100. Trapping success was highest in leaf litter frogs (56. However.2 66.0 31.0 100.0 100. all were recorded using other methods as well.0 0.0 72. VES and AES revealed highest species numbers. transect walks made it possible to search for amphibians in a very comparative way. neither between given family or guild.7 100.0 64.4 100. 1.8 0. With the exception of Acanthixalus sonjae. N = 4.0 SVTS 0. was difficult as detectability of species varied considerably with (e.0 100.0 100.. All other species were recorded with VES/AES (Rödel & Branch 2002. df = 4.0 40.0 100.

species not encountered during transect sampling. inaccessible areas than in areas that are easy to monitor. The percentage of species within supraspecific taxonomic units. In addition. e. such as species of the genus Acanthixalus. Schiøtz 1999. Although the combination of AES and VES provided an adequate approximation to real conditions (see also Doan 2003). Some taxa. spending more time in complex. 1996).. However. The advantage of transect sampling was especially obvious when focusing on leaf litter species. VES and AES yielded highest species numbers. Rödel 2003. but is certainly just as well suited for the investigation of tropical anuran assemblages in general. and failed to detect aquatic and fossorial species. these methods do not seem to sample a given fauna adequately well (Allmon 1991. Compiling data on presence and abundance of aquatic species and tadpoles was outside the scope of this study and these species and stages were inadequately sampled with all methods used here. acoustic sampling appeared to be indispensable since these species were more readily detected by their calls than 8 by sight. Standard plot sampling and total removal plots were not tested. We found it impossible to quantify VES and AES other than in relation to time. such as species groups. Rödel & Ernst 2003). Migrating pipid frogs can also be easily recorded with pitfall and funnel traps. such as the Hylidae. Thus even time-based quantification of these data is questionable. Rödel et al. When including relative abundance measurements these methods were insufficient..RÖDEL & ERNST DISCUSSION Choice of methods should reflect the best compromise between i: practicability. ii: specific aims. Although widely used and recommended. Nonetheless. (1995) a simple comparison of species numbers may not be sufficient to evaluate the efficiency of different sampling methods. Therefore transect sampling can be considered to be the method of choice when sampling leaf litter anurans. Vonesh 2001.1. pitfall traps and drift fences . and iv: environmental impact. Pitfall trapping was only useful in detecting fossorial species that were not encountered during transect walks. According to Pearman et al. In this study this has been done by determining relative species richness with reference to families or guilds. iii: efficiency.2.g. a combination of SVTS and SATS. SVTS was generally the best method for detecting diurnal leaf litter frogs. Arthroleptis sp. whereas visual sampling appeared to be superior to acoustic sampling. VES and AES are useful tools for the compilation of species inventory lists. (1994). Semlitsch et al. Bury & Corn 1987. can then be compared to the real conditions in an area or habitat. as an adequate measure. genera or families. especially in long-term field studies within temperate regions (e. The combined transect sampling methods provided a close approximation to the real presence and abundance of species. In general SATS was very efficient in detecting nocturnal leaf litter frogs and treefrogs but less successful in detecting diurnal leaf litter frogs. We found no additional species and only very few specimens by funnel trapping. during rapid assessment surveys (for recent West African examples see Rödel & Branch 2002. or Phrynobatrachus alticola. They suggest using relative species richness within taxonomic entities. We found it more realistic to apply a (clearly subjective) measure for habitat complexity on which to base collection effort during VES/AES. (1997) and Rödel (1998a). such as the very abundant Arthroleptis sp. Olson et al. 2003). but appear not to be appropriate when standardized quantitative sampling is required. this only holds true if species numbers alone are considered. Cardioglossa leucomystax. Simple presence/absence data of aquatic species and stages can best be gathered by direct observation (surfacing individuals) and with dip-netting. as they impose a severe disturbance on the system under investigation and thus were not suitable to our goals. This was also true for most of the arboreal species belonging to the families Hyperoliidae and Rhacophoridae. Doan 2003). especially arboreal and fossorial. as well as SVTS alone. could be detected exclusively using visual sampling methods. and will likewise hold true for other arboreal anurans throughout the tropics. which are mute (Drewes 1984.g. The methods tested within this study proved to perform with varying degrees of efficiency. were nearly as successful in detecting species. in order to sample particularly secretive leaf litter species. The differences in efficiency of visual versus acoustic sampling varied between taxa. Methods of collecting quantitative data for aquatic amphibians and for tadpole assemblages have been described by Shaffer et al. data acquisition varies considerably depending on habitat type in VES and even AES. The methods VES and AES in combination with pitfall traps with drift fences were useful in sampling additional. determined by using a particular sampling method. Whereas transect walks provide the possibility of monitoring amphibians in a very comparative way. However.

generalization of data with regard to phenology or abundances is less possible. for other abbreviations see text and Fig. Transect walks are time intensive. When considering the relatively high costs and time investment due to the high level of maintenance required (Parris 1999). 1. accounted for 72. amphibian guild and environmental impact. none of them uniquely found by this method. SVTS and SATS have been proven to be the only methods that provided quantitative data on forest amphibians with regard to space and time. e. the frequency of sampling can be reduced when general phenological traits of the species recorded have been clarified. however. Costs with regard to time spent for data acquisition was highest in trapping and in transect walks. Data can be used for comparisons between habitats. one should test the efficiency at a particular site in advance before installing these traps on a broad scale. Data from both trapping methods can. With the exceptions of aquatic and fossorial species they pro- vided data that cover the whole community very well. be quantified in relation to time. The impact of traps and drift fences was difficult to judge. SVTS and SATS provided data on species and habitats und thus may reveal changes in composition and species abundance. VES and AES had probably the lowest environmental impact. If the investigation is interrupted for some time. VES and AES are less time consuming. Doan 2003). TC leaf litter frogs arboreal frogs fossorial amphibians aquatic frogs qualitative data quantitative data environmental impact costs (time) costs (material) utility (ecology) utility (conservation) +++ ++ – +++ +++ +++ low high low very high very high SVTS +++ ++ – +++ +++ ++ low high low high high SATS ++ ++ – – +++ ++ low high low high high AES/VES +++ +++ + +++ + + very low low very low low high Trap + – +++ + + – medium very high low very low low QS/TR +++ – – – +++ +++ very high high low very low low 9 .. as well as environmental changes. 2001. In our experience. experience of investigator and prevailing weather conditions (see Branch & Rödel 2003). We observed (e. Parris 1999. Theoretically it might well be that trapping will result in dislocating at least some species from the area where traps have been established. Vonesh 2001. (2001) report on similar experiences using this method in a herpetological survey in Guyana. the fossorial Hemisus perreti and the aquatic Silurana epitropicalis.) that leopard (Panthera pardus) tracks TABLE 3. traps have to be uninstalled or closed and reopened when starting again. Burger et al. However trapping success depends very much on choice of site.g.6 % of all individuals captured with this method.AMPHIBIAN MONITORING IN TROPICAL FORESTS proved to be the least effective method. if not on amphibians then on other organisms. in theory. Donnelly et al. Rodda et al. A comparison of trapping data with data other than those of the observer in question seems to be difficult. seasons and years. Effectiveness of methodology in regard to type of data required. Thus data should be gathered throughout a whole season to obtain a thorough knowledge of the local fauna. However only two species. Traps have to be checked at least daily. They collected only six species (out of a total of 132 species of amphibians and reptiles being recorded). QS/TR = quadrat sampling/total removal sampling (data from Allmon 1991. However. this method cannot be recommended for standardized sampling of tropical anurans.g. Transects will have an environmental impact. as trapping success per trap per time unit. (2004) had high trapping success with pitfalls in Gabon when considering numbers of individuals. in consecutive field seasons. It is advisable to perform them randomly independent of prevailing weather conditions.

N. N’Dri. Branch. Pp. 5: 37–53. Gabon: Species turnover along an elevational gradient.. R. fond. 2 ed.” Working in TNP was kindly permitted by the Commandant K. Field techniques for herpetofaunal community analysis. Rödel. & S.J.S.” of the Republic of Ivory Coast. Hayek. After using transects intensively for some time. Verteilungsfreie Methoden der Biostatistik. 1982. 1990. Pp. Lips et al. & M. Alford. 1979b. Channing. The presented transect design proved to be adequate for representative sampling and appeared to be appropriate for most studies involving multivariate structure data. Lienert. Bortz. noire. M.. Washington & London. W. 364 pp. 1964. Global amphibian declines: a problem in applied ecology. Linsenmair.. Hofer. 1999. U. & K. TROPENBOS . All this support is gratefully acknowledged. Standard methods for amphibians.Y. & P. 2003. Various administrative services and lodging facilities in TNP were provided by the “Centre de Recherche en Ecologie” and the “Projet Autonome pour la Conservation du Parc National de Taï.. Wildlife Research Report 13. Herpetological survey of the Haute Dodo and Cavally forests. Springer Verlag. Glos. G.G. Memoir 28: 145–186. L. Ecotropica 4: 93–97. Part II: Trapping results and reptiles. 109–117 in Heyer. Grundzüge der Vegetationskunde..o. & M. ACKNOWLEDGMENTS This study is part of the BIOLOG-program of the German Ministry of Education and Science (BMBF. Sci. Afr. and B. R. Trop.A. J. J.. Boehnke. Observations sur la reproduction et la dynamique des populations de quelques anoures tropicaux IV. U.E.. 193–200 in Scott. 865 pp.S.R... 1994.A. Springer Verlag.. A plot study of forest floor litter frogs. For simple short-term surveys we recommend VES and AES.W. Corn. Observations sur la reproduction et la dynamique des populations de quelques anoures tropicaux III.. J. Arthroleptis poecilonotus. Syst.RÖDEL & ERNST were more often seen on the transect path than in undisturbed forest. The access permit to TNP was issued by the “Ministère de la Construction et de l’Environnement. Jr. & K. Richards. R. consequent long-term effects on forest structure.1964. 01 LC0017). A. & R. Ouoro provided invaluable help during fieldwork. Brazil.. Basset... western Ivory Coast. W. Corn. Phrynobatrachus accraensis. Furthermore transects provide an effective method of investigating leaf litter frogs in particular. & M. thus generating a much more complete picture of an amphibian community than is possible with other methods.J. Ecol. Trefaut Rodrigues. 30: 133–165. Bury. Berlin a. 20: 64–77. Straight-line drift fences and pitfall traps. Rev. Salamandra 39: 91–110. REFERENCES Adis. Foster. Ptychadena maccarthyensis et Ptychadena oxyrhynchus. A 41: 417–428. 7: 503–522. P. 1997. Wien. McDiarmid. Calif. Canopy fogging of an overstory tree – recommendations for standardization. Hammond.): Herpetological communities. (eds. C. Christman. Advantages and disadvantages of different methods are summarized in Tab. H. Bull. J. R..R. P.S.W. Y. 1979a. Acad. Rödel 1998a). Braun-Blanquet. W. Terre Vie 32: 441–452. for specific questions (cf. Branch. Evaluation of pitfall trapping in northwestern forests: trap arrays with drift fences. Barbault.B. Chatelain. Inst. 3. Hillcoat provided valuable comments on the manuscript.. Trefaut Rodrigues. Halliday. Conserv. Barbault. M. and the “Taï Monkey Project” provided logistic support. Observations sur la reproduction et la dynamique des populations de quelques anoures tropicaux I. 1991..C. W. not only on their breeding sites but throughout the whole range of habitats used by them. Gautier. Trop.. SVTS and SATS are especially useful for long-term studies. Ann. J. Amphibians and reptiles of Monts Doudou. & S... Donnelly. & M. Research permission was given by the “Ministère de l’Enseignement Supérieur et de la Recherche Scientifique. Spichiger. However. Olson et al. The “Projet Autonome pour la Conservation du Parc National de Taï”.D. 2004. Smithsonian Institution Press.. Barbault. This might affect densities of other mammals including possible seed dispersers with. Ecol. L. possibly accompanied by opportunistic trapping. Ecol. and C. Central Amazon. & A. 2003) transects should be supplemented by specifically investigating aquatic breeding sites with the appropriate set of standardized techniques (Heyer et al. Allmon. R. Transect cutting and the use of transects should therefore be done with the necessary precautions. Department of the Interior. 1998. (ed.A. A recent history of forest fragmentation in southwestern Ivory Coast.): Measuring and monitoring biological diversity.Côte d’Ivoire helped with transportation. Trefaut Rodrigues. Gbamlin.M.. Sér.-O. 1987. 1996. 1994.S. Floren.” G.P. 1978.R.-A. Burger. R. Biodiv. 10 . T.. Fish and Wildlife Service 13. Project W08 BIOTA-West. J. Pflanzensoziologie. Wildlife Management 51: 112–119. in the worst case. & M. a well visible path becomes established that at least in primary forests remains so for years. 939 pp. Campbell.

S. Kuzmin. Lips. Foster (eds. Foster (eds. Complex causes of amphibian population declines.R. Smithsonian Institution Press. SSAR.K. + 15 plates. L. 1994.. L. Foster (eds.-A. 300 pp.. 97–102 in Heyer. B. Sampling methods for terrestrial amphibians and reptiles. M.R. vi.): Naturschutzrelevante Methoden der Feldherpetologie... Reeve.o. Quadrat sampling. C.-A.. systems and processes. Pearman. Quantitative evidence for global amphibian population declines. Herpetol.W. R. & M. Plant ecology in West Africa. Tropenbos Series 13.AMPHIBIAN MONITORING IN TROPICAL FORESTS Corn.. Meyer. 1999. Foster... Rheinbach. Hayek. Smithsonian Institution Press. Which methods are most effective for surveying rain forest herpetofauna? J. M. Kennzeichnung und Eigenschaften der landwirtschaftlich genutzten Böden der DDR.R. 3. C. & R. W..B. Velasco. 2001. Nature 410: 681–684.. & R. R.C. 115 (7). and the Seychelles Islands.E. J. 2003.und Tierschutzsicht. & J. Biol. M. http://www. Parris. PACPNT 2000.C. B.. Alford.M. K.K.. MjM Software Design.M. Parris. L.. Society for Northwestern Vertebrate Biology. & M. Pearman..L.-A. Krebs. W. 18 pp. Pp.R. Auflage. 1967. Drewes. New York a.E. Pp. Identifying effects of toe clipping on anuran return rates: the importance of statistical power. Kasparek Verlag. Jaeger.P.A. Review: amphibian surveys in forests and woodlands.. McDiarmid. Hayek. 1991.R. W.. Standard methods for amphibians. Foster. A. Schmidt-Loske. & R. Donnelly. M.. & L..A.) 2002. Amphibian monitoring in Latin America: A protocol manual.pdf Donnelly.S. K.K. DAPTF 2002. Schmidt. The quest for natural forest management in Ghana. Witters. Bury. 364 pp. J. John Willey & Sons Chichester. G. 134 pp.. C. McDiarmid.K.P. 2002.J.C. Standard methods for amphibians. 1994. Reaser. Juterbock.. Wageningen.. http://www..B.) 1994. Measuring and monitoring biological diversity. Visual encounter surveys.D.. L.A. & C. A.A. K. Scott.. Donnelly. Flore du Parc National de Taï (Côte d’Ivoire).R. & G.A.W. 1984.cnah. Jaeger. Lawson. Forest Service. 37: 72–81. Recherches sur la végétation et la flore de la région du Bas-Cavally (Côte d’Ivoire). de Graaf. Belden. Berlin.R.. R. Donnelly. (eds. Tropical amphibian monitoring: a comparison of methods for detecting inter-site variation in species composition. Mertensiella 7. K. Conserv. J. Washington & London. 1989.. Inger. 103–107 in Heyer. Grace. & A.S. African forest biodiversity. Standard methods for amphibians. Blaustein.. C.S. C. MN.A. New York. Hayek.A. Donnelly. 1986..W. & M. A field survey manual for vertebrates. Lieberoth. Herpetologica 51: 325–337. McCarthy.S.-L. M. 2001. Biol.W. Parren.W. Drift fence-associated sampling bias of amphibians at a Florida sandhills temporary pond. R. 25: 296–301. Leonard. W. 17–44 in Davis. Bayero University. (eds. Côte d’Ivoire and Liberia. 133–184 in Henle. Aufbau. J. Chen.htm. 2000. & R. Iwokrama. Madagascar. 161 pp.S. Lips. Washington & London. Amphibia-Reptilia 22: 275–289. & M. 320 pp. Gleneden Beach. R.. R. Northwest Fauna Number 4. Ibáñez. R.E.htm. D. Nigeria. Earthwatch Institute (Europe). & M.. Bury. Oregon. Nature 404: 752–755. J. & R.M. Houlahan.R. M.-A. 199 pp. K.. R. McDiarmid. Kiesecker. Conserv.B. 2003. Bender 1997.. Veith (eds. Techniques for marking amphibians. Smithsonian Institution Press. Correlates of amphibian diversity in an altered landscape of Amazonian Ecuador. López.A. Sci. Guillaumet.. J. W. Pap. Smithsonian Institution Press. Standard methods for amphibians. Kuhn.): Measuring and monitoring biological diversity. Findlay. & L. Hayek. 364 pp. Young.. 1997. T..): Measuring and monitoring biological diversity. Declining Amphibian Population Task Force.. Heidelberg. Smithsonian Institution Press..C. Guyer. Dodd. Ecological traits predicting amphibian population declines in Central America. (ed. 17: 1078–1088.A. Doan. 1997. Department of Agriculture. J. Shoreview.. & N. 1994. Podloucky.H... Pp.): Measuring and monitoring biological diversity.O. 2002. 20 247 pp. & N. M.... Hayek..B. Standard methods for amphibians.): Measuring and monitoring biological diversity. 1982.R.S. L.H. McDiarmid. Transect sampling. Washington & London.T. 277–284 in Heyer. K. McDiarmid.. P.. Pp. G. 2001.G. 84–92 in Heyer.R.. Mémoires O. Occ. B. 139: 1–70. Howell. Acad. 378 pp.M. Individualerkennung und Markierung mitteleuropäischer Amphibien und Reptilien: Übersicht und Bewertung der Methoden..iwokrama..R. R. 364 pp. Oxford. Harper Collins Publishers. Washington & London.M. Herpetofauna of the Iwokrama Forest. Empfehlungen aus Natur. 357 pp.C. Herpetol. I. Olson.R. Bodenkunde.W. 654 pp. General Technical Report.S. 1995. 364 pp. P. A... M. Sampling amphibians in lentic habitats.C. Donnelly. P. 2001. Pp. J. Entstehung. Analysis of ecological communities. A. Crump. Jr.M. Henle. 364 pp. Watson.L. A phylogenetic analysis of the Hyperoliidae (Anura): Treefrogs of Africa. PNW-GTR-256.F.S. & M. K. Pp.J. U. 1994. Washington & London. W. & S. Calif. 11: 1211–1225. McCune. 11 . Contemporary Herpetology 1: http://www.G. Ecological Methodology.. & M. K. Amphibians and reptiles: the herptiles.. Watkins. U.B.

M.RÖDEL & ERNST Perret. M.. M. Cambridge. 2001. Rodda. van R. 36: 561–571. Structure and dynamics of an amphibian community.E. 1994. West Africa. a global perspective. M.W. Schiøtz. Treefrogs of Africa. Genève 41: 275–294. and Seychelles. Sci. 2001. 1988.G. PhD Thesis. Frankfurt /M. Ernst. Pp 108–113 in Girardin. Vooren. Woodward. Veith. R. 37: 43–52. 2001.W.. Rödel. M. Rev. Salamandra 38: 245–268.): Patterns of distribution of amphibians. 1999. The tropical rain forest. R. Koné. Ernst.L. Herpetol.. & W.P. U. J.. Ivory Coast. Hayek.R. A new Phrynobatrachus from the Upper Guinean rain forest. J. Herpetozoa 16: 23–39. M.-O. Rapport de Centre Suisse de la Recherche Scientifique. McDiarmid. Gibbons. 2000.. 2003. & J.S. 1994.-A. Guillaumet 1994. Ernst.. Richards. M. D. Herpetol.H. 633 pp. Kaulquappengesellschaften ephemerer Savannengewässer in Westafrika. Salamandra 39: 91–110. Côte d’Ivoire.E.D...... Smithsonian Institution Press. M... Rödel. J. 575 pp. M. Smithsonian Institution Press. 1993. 1996. M.R. 130–141 in Heyer. an ecological study.): Etat des recherches en cours dans le Parc National de Taï (PNT).A. 364 pp. Rödel. & R.A. 92–97 in Heyer. Standard methods for amphibians. Herpetol. & R. First record of the genus Acanthixalus Laurent. Spolia zool. Foster (eds.-O.. B. 1996. 2000.-O. H. Biotropica 33: 502–510. & R.C.-O. 195 pp. 32: 24-30. J.. Pp. Semlitsch. E.K. & J. Tropenbos Series 8.P. Rödel. herpetofauna 22 (125): 9–16. & M. Pounds. 1944 from the Upper Guinean rain forest.. Johns Hopkins University Press. A spatial gradient analysis. sp. Le Parc National de Taï. Bufo taiensis n. Hayek. & G. M.W. J. eine neue Kröte aus dem Taï-Nationalpark. 1998b.): Measuring and monitoring biological diversity.S. Les espèces de Phrynobatrachus (Anura.. Arch. B. Donnelly..-L. Vonesh. Riezebos. A.J. Schiøtz. Donnelly. J.A. Foster (eds. 217–448 in Long-term studies of vertebrate communities. Campbell. A. Gascon. Ernst. Wageningen. 2002.. I.-O. Rödel. Wageningen. Herpetozoa 11: 19–26.-O. West Africa. The amphibians of Mont Sangbé National Park. R. Fritts. Rödel. K. R. Distribution of amphibians in Sub-Saharan Africa. Audio strip transects.. Les communautés d’amphibiens dans le Parc National de Taï.. Altig. Edition Chimaira.B. 142 pp.A. Sempervira. M.-A.-O. 323 pp. Cambridge University Press. including the description of a new species. 1999. A reproductive mode so far unknown in African ranids: Phrynobatrachus guineensis Guibé & Lamotte. J.. & M. Pechmann. Scott. Elfenbeinküste..K. W. 1961 breeds in tree holes (Anura: Ranidae). Herpetological survey of the Haute Dodo and Cavally forests. 9. Forest gradients in West Africa. Alford.. Baltimore. E. Branch. A. Duellman (ed.D.): Measuring and monitoring biological diversity. Edition Chimaira. Frankfurt/M.D. 1998a. Part I: Amphibians. including a description of a new reproductive mode for the genus. Academic Press.S. Rompaey. Wells. Zimmerman. Kosuch.W.. Abidjan. Nature 410: 639–683.R.R. L. Shaffer. & Y. Climate and amphibian declines. J.. O. G. 1977. 2003. 1967.L. & R. Tano (eds. P. S. Pp 483–539 in W.C. Rödel. 2002.H.. The amphibians of Marahoué and Mont Péko National Parks. McDiarmid. Pp. western Ivory Coast. Patterns of richness and abundances in a tropical African leaf-litter herpetofauna. Washington & London.C.. Rödel. The social behaviour of anuran amphibians..A. 2003. 364 pp. 12 . Evidence from a 16-year study of a natural pond. Standard methods for amphibians. Rödel.. A high validity census technique for herpetofaunal assemblages. 350 pp.-O. Quantitative sampling of amphibian larvae. Washington & London. Côte d’Ivoire. 25: 1–346.. Animal Behaviour 25: 666–693. Ranidae) à éperon palpépral.. W. The treefrogs (Rhacophoridae) of West Africa. Musei Haun. Madagascar. Ivory Coast. Poynton. Richards.R. R. Les anoures comme bio-indicateurs de l’état des habitats. Pp. & T. L.W.H.-O.

Trap = species recorded with pitfall and funnel traps. Perret 1988) but not recorded by us. TP = all species recorded on transects in primary forest. SRET = species recorded with VES/AES on the clearing around the SRET station. VES/AES = species recorded via visual and acoustic encounter surveys in forest habitats near SRET. TC = all species recorded on transects. TS = all species recorded on transects in secondary forest. 5 = treated as full species and not as subspecies of Hyperolius fusciventris due to sympatric occurrence in TNP (see Rödel & Branch 2002). 2003). 2 Petropedetidae Phrynobatrachus accraensis Phrynobatrachus gutturosus Phrynobatrachus fraterculus Phrynobatrachus guineensis 3 Phrynobatrachus phyllophilus Phrynobatrachus liberiensis Phrynobatrachus alticola Phrynobatrachus alleni Phrynobatrachus plicatus TaïF G SRET TC TP TS VES/AES Trap guild 1 0 0 0 0 0 0 1 f 1 0 0 1 1 0 1 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 0 1 0 0 1 1 1 1 1 1 0 1 1 1 1 1 1 0 1 1 0 0 0 0 1 0 0 0 1 1 1 0 0 1 1 1 0 0 0 0 0 0 0 1 1 0 0 1 1 0 0 0 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 0 0 1 1 1 0 0 0 0 1 0 0 0 1 1 1 1 1 1 1 1 0 0 0 1 0 0 0 0 1 1 0 0 0 0 0 1 1 0 1 0 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 1 0 1 0 0 0 0 1 1 0 1 1 0 1 1 1 1 1 1 1 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 aq l l l l l f aq l l aq l l l l l l l l l l l. 4 = Acanthixalus sonjae lives aquatically in large water filled tree holes but forages in surrounding vegetation (Rödel et al. 3 = Phrynobatrachus guineensis is a leaf litter frog but reproduces in small water-filled tree holes (Rödel 1998b). l = leaf litter species. 5°25’ N. List of amphibian species recorded in Taï National Park (TNP). 1 = known from TNP (Schiøtz 1967. TaïF = species recorded in forest habitats around the SRET station. 2 Ranidae Hoplobatrachus occipitalis Amnirana albolabris Amnirana occidentalis Aubria occidentalis Ptychadena pumilio Ptychadena bibroni Ptychadena mascareniensis Ptychadena superciliaris Ptychadena aequiplicata Ptychadena longirostris Ptychadena sp. a = arboreal species. aq = aquatic species. other than transects. 2 = determination not possible at present time.AMPHIBIAN MONITORING IN TROPICAL FORESTS APPENDIX. G = species recorded in southern TNP (Guiroutou. a l l l l l 13 . species Gymnophiona Caecilidae Geotrypetes seraphini Anura Pipidae Silurana tropicalis Bufonidae Bufo regularis Bufo maculatus Bufo togoensis Bufo taiensis Bufo superciliaris Hemisotidae Hemisus sp. 6 = recorded north of area under investigation. 7°10’ W). f = fossorial species.

aq a 14 .RÖDEL & ERNST Appendix continued species Phrynobatrachus calcaratus Phrynobatrachus taiensis 1 Phrynobatrachus villiersi Phrynobatrachus annulatus Astylosternidae Astylosternus occidentalis Arthropleptidae Cardioglossa leucomystax Arthroleptis sp. 2 2 Hyperoliidae Leptopelis hyloides Leptopelis occidentalis Leptopelis macrotis Hyperolius concolor Hyperolius guttulatus 6 Hyperolius picturatus Hyperolius sylvaticus Hyperolius zonatus Hyperolius fusciventris 5 Hyperolius lamtoensis 5 Hyperolius nienokouensis Hyperolius wermuthi 1 Hyperolius chlorosteus Afrixalus dorsalis Afrixalus nigeriensis Afrixalus vibekae Kassina lamottei Phlyctimantis boulengeri Acanthixalus sonjae 4 Rhacophoridae Chiromantis rufescens total 56 species in TNP TaïF 0 0 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 43 G 1 0 1 0 1 1 1 1 1 1 1 1 0 1 1 1 0 1 1 0 1 1 1 0 1 0 0 1 41 SRET 0 0 0 0 0 0 1 0 1 0 0 1 0 1 0 0 1 0 0 0 0 1 0 0 0 0 0 0 16 TC 0 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 0 1 1 1 1 1 0 1 1 37 TP 0 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 1 0 0 1 0 1 1 1 0 0 1 33 TS VES/AES Trap 0 0 1 0 1 1 1 1 1 1 1 0 0 0 1 1 0 0 0 0 1 1 1 0 0 0 1 1 28 0 0 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 39 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 guild l l l l l l l l a a a a a a a a a a a a a a a a l a a. 12 Arthroleptis sp.

Sign up to vote on this title
UsefulNot useful