EFFECTS OF CLIMATE CHANGE ON GENETIC AND SPECIES DIVERSITY OF AMPHIBIANS AND REPTILES IN THE ALBERTINE RIFT

EFFECTS OF CLIMATE CHANGE ON GENETIC AND SPECIES DIVERSITY OF AMPHIBIANS AND REPTILES IN THE ALBERTINE RIFT Waswa Sadic Babyesiza 1 Akoth Sisiria 2 Lukwago Wilber 3 Isingoma Joseph 4 2017

1.1 Introduction Although amphibian populations are declining globally, amphibians are known to be an integral part of all ecosystems (Wake & Vredenburg, 2008), serving vital ecological roles such as being linked to various levels of the food chains; Amphibians are predators of invertebrates and also act as food for other vertebrates like birds and mammals (Gibbons et al., 2000). They are known to be very sensitive to physical, chemical and biological changes in their environment (Blaustein & Wake, 1995), which makes them good subjects for studying ecological conditions of an area. Compared to other vertebrates, herpetological studies in Uganda have been mostly basic. Inventories on amphibians in Uganda started as early as 1913 (Drewes, 1994). Work such as that has been on-going in different parts of Uganda. Recent detailed work in Uganda was carried out in the Rwenzori Mountains and is entitled The status of amphibians in Rwenzori Mountains National Park (RMNP) and Semliki National Park (SNP) (Behangana, 1995). A comparative study of amphibian diversity in Rwenzori Mountains National Park, the SangoBay area and Semliki National Park was also more recently conducted (Aguti, 2007). Although amphibians are understood as excellent biological indicators for environmental changes (Heyer et al., 1994), they have not been used as such in Africa until recently (Poynton, 1999; Rödel et al., 2002). There is a call for climate change effects research in relation to biodiversity within key biodiversity ecosystems to aid in decision making. This research will then enable the efforts geared towards amphibian conservation.. Research on the effect of climate change has been recently carried out and widely documented in Uganda. However, no clear studies have come out to show how climate change has impacted amphibians in key biodiversity habitats like the Albertine Rift. 1.2 Decline of amphibian populations Amphibians, like many other groups of organisms, are facing worldwide population declines, range contractions, and species extinctions. The single most important causes of amphibian

declines globally are habitat degradation and climate change resulting from human activities (Blaustein et al., 1998). Over the past two decades, amphibians have been the focus of increasing concern in the scientific literature and popular press because of numerous reports of population declines, range contractions and extinctions. Recent analysis of data on 157 species from 37 countries and 8 regions of the world suggest that amphibians have declined on global scale (Semlitsch, 2003). 1.2 Ecology Although amphibians and reptiles are found in a wide array of habitats, amphibians in particular are limited to areas where there is sufficient moisture for reproduction and survival. (Duellman &Trueb, 1986). The habitat of amphibians and reptiles is closely tied to the mode of reproduction (Clarke, 2000). Amphibians have porous skin and generally prosper in warm and damp places. Because virtually all physiological processes are temperature sensitive, ambient temperature is undoubtedly an important limiting factor (Pounds &Crump, 1994). Some species have extremely limited ranges while others are widely dispersed in a variety of localities and climatic regions. Some reptiles, crocodiles, turtles and lizards, have their modes of reproduction and sex of offspring dependent on temperature (Charnier, M. 1966). On the other hand, some reptiles such as Chamaeleo rudis live in very cold environments and produce live young. Amphibians have breeding, foraging and hibernation/aestivation sites. These sites may be temporally and spatially separated, and in such cases individuals must migrate to and from them in seasonal cycles (Msuya, 2001). Movements may occur only during a narrow range of environmental conditions and are often limited to relatively short periods during the annual activity period (Sinsch, 1990). 1.3 Factors influencing diversity and distribution of amphibians and reptiles Physical and biotic factors are important in structuring tropical, herpetological communities (Gascon, 1996). For instance, physical characteristics such as altitude, moisture and biotic characteristics such as prey abundance are known to positively correlate with litter amphibian abundance in some ecosystems (Lieberman, 1986; Allmon, 1991). Other biotic and abiotic factors known to influence amphibian diversity and distribution include tree cover.

1.3.1 Vegetation The most dominant form of vegetation in tropical forests is trees. The vertical and horizontal alignment of these trees directly influences amphibian assemblages. Vertically, the forest is stratified into layers, under storey and the middle and the top storey Bourgeron (1983). The under storey consists of decaying matter,the litter layer. This is the repository of all of the dead matter in the forest. As leaves, trees and other plants die, they fall from upper layers and land on the forest floor. Here a host of bacteria, fungi, worms, insects and other waste consumers process the litter to create new nutrient-rich soil (Parker & Brown, 2000). Most terrestrial amphibians are confined in this zone. Amphibians do not explore the top storey of the forest; however a great number of them are confined within the middle storey. This includes species in the genus Leptopelis, commonly known as the tree frogs (Channing & Howell, 2006) 1.3.2 Moisture Moisture availability is critical to amphibians, as they are biphasic and critically depend on availability of moisture for breeding and development (Crump, 2010). The seasonality in water levels directly influences the distribution of water-affiliated amphibians (Sally et al., 1998). This calls for their unpredicted distributions with varying moisture gradients away from water sources. Some amphibians also reproduce via direct development, an ecological and evolutionary adaptation that has allowed them to be completely independent from freestanding water. Almost all of these frogs live in wet tropical rainforests and their eggs hatch directly into miniature versions of the adult, passing through the tadpole stage within the egg (Duellman & William, 2012). Reproductive success of many amphibians is dependent not only on the quantity of rainfall, but the seasonal timing (Duellman & William, 2012). 1.3.3 Altitude Altitude is particularly important as it affects the geographical variation of other climatic and/or habitat factors (Dettki et al, 2003). Thus, altitude is a variable usually correlated with species distributions (Brito & Crespo, 2002) and with species richness (Soares et al, 2007).

There is a significant difference in amphibian species distributions with the differences in altitude (Morales et al, 2002). Only Anura and Urodela as amphibian orders have significant altitudinal differences (Brito &Crespo, 2002). This difference in distribution in amphibians is attributed to the fact that they are more specific to particular environmental conditions which are affiliated to altitude (Chefaoui et al, 2005). 1.4 Research problem There is insufficient knowledge on how climate change has affects the ecology, diversity and the distribution of amphibians and reptiles, particularly in developing countries (Stuart et al. 2004). Climate change is considered to be the most likely driver for change in reproductive seasons and phenology which in turn affects the diversity and distribution of different amphibian species. It has been noted however that it is unlikely that the change in climate itself is the principal source of amphibian declines but rather an enabling factor (Stuart et al. 2004). There is therefore the need to evaluate the relationship between changes in the physical climate system and how amphibians and reptiles are adapting and evolving to changes in their environment. 1.5 Research objectives 1.5.1 Main objective Assess the effect of climate change on the species diversity and spatial distribution of amphibians and reptiles in the Albertine rift. 1.5.2 Specific objective Determine amphibian and reptilian species richness and abundance in the different study areas Document the elevation distribution of amphibians and reptiles and how they respond to various changes in environmental attributes. 2.0 METHODS 2.1.1 Visual Encounter Surveys (VES)

VES involves moving through the habitat, turning logs or stones, inspecting retreats and watching out for surface-active species following the approach described by Howell (2002). VES also includes listening for and recording distinctive frog species-specific calls and actively searching the area with headlamps and torches. Calls of individual frogs were recorded using mobile phones and digital cameras as recorders. 2.1.2Pit fall traps Pitfall traps were set up with a drift fence, with 20 buckets in series set along a transect. The use of drift fences with pitfall traps is a common technique for studies of individual species and herpetofaunal communities. It has been used with success for amphibians and reptiles by Hare in, 2012. It was very useful the lower areas of RNP and the drier parts of SNP. The high altitudes of RNP were rocky and steep while some parts of SNP were in a flood plain. At these sites, the buckets fill up with water causing animals to swim away or drown. 2.2 Study sites 2.2.1 Rwenzori Mountains National Park (RNP) The Rwenzori Mountains are located in the Albertine rift valley where about 19% of amphibian species on the African continent occur (Behangana et al, 2003). The mountains are situated between latitudes 29 o 47 E and 30 o 11 E and longitudes 00 o 06 N and 00 o 46 N The mountain is about 90 km in length and covers 996 km 2 with an altitudinal range of 1,700 to 5,109 meters a.s.l. The mountains are endowed with a variety of vegetation types ranging from montane forests in the foothills, through a bamboo zone and a zone of giant tree heathers to Afro-alpine moorland and a high altitude zone below the snow line. The forest supports many animal and plant species of extreme conservation importance by virtue of their rarity and /or limited distributions. 2.2.2 Semliki National Park (SNP) SNP is a very dense forest and forms an eastern extension of the vast Ituri Forest. It was part of the forest continuum during the climatic upheavals of the Pleistocene. This is one of the richest areas for both flora and fauna in Africa with at least six species found nowhere else in Uganda

(Davenport, 1996). It is the only protected area in Uganda composed primarily of tropical lowland forest. SNP is situated in the remote corner of extreme west of Uganda on the border of the Democratic Republic of Congo within the western arm of the east African Rift Valley. The geographical coordinates are 0 44 o -0 53 o N- 29 57-30 11 o E. SNP (220 km 2 ) occupies a flat to gently undulating landform ranging between 618 and 760 meters a.s.l. It is both the only lowland tropical rain forest in East Africa and has been classified as semi-deciduous. Figure 1 (a): Survey points in SNP (b): Survey points in RNP Data analysis Species diversity Species richness between Rwenzori and Semliki were compared using sample-based rarefaction curves. Rarefaction curves are created by randomly re-sampling the pool of samples multiple times and then plotting the average number of species found in each sample as a function of number of individuals (Colwell, 2013). Rarefaction accounts for species richness beyond the reference sample and allows comparisons between study areas.

Species-habitat association Small mammal community composition and association was measured using multipatt function in R analysis package. According to Miquel De C_aceres (2013) function multipatt is the most commonly used function for species habitat association and community characterization. It allows determining lists of species that are associated to particular groups of habitats (or combinations of those). The different habitats were first clustered using non-hierarchical cluster analysis in R the clustering yielded groups of habitats depending on their similarity in-terms of species composition. Relationship between environmental variables and amphibian and reptiledistribution To assess how environmental parameters influence the pattern of amphibian and reptile communities among study sites in both Rwenzori and Semliki, Canonical Correspondence Analysis (CCA) was used. CCA is a correspondence analysis (CA) in which weighted multiple regression is used to represent the axes as linear combination of the explanatory variables. It is a multivariate method for explaining the relationships between biological assemblages of species and their environment by extracting environmental gradients from ecological data-set (ter Braak & Verdonschot, 1995). The gradients are the basis for describing and visualizing the habitat preferences of species.

Fig 2 (a): Spatial distribution of B. xenorhinum(b): Spatial distribution of C. rudis 3.0 Results Overall 622 individuals were recorded for both the two study sites (361 from Rwenzori and 263 from Semliki). At 361 individuals, RNP with 37 species (21 amphibians and 16 reptiles) was richer than Semliki with 28 species (19 amphibians and 9 reptiles).species rarefaction curves show that in RNP, asymptote was reached while for SNP it is still increasing (Fig 3).The smoothed averages of these individual curves represent the statistical expectation of species accumulation curve per study area. Rwenzori Species 0 5 10 20 30 Semuliki 0 50 100 200 300 Number of individuals Fig 3: Rarefaction curves comparing sample size and species accumulation for the two sites Diversity was overall high but slightly higher in Semliki (2.8) compared to that of Rwenzori (2.7). Renyi diversity plots also showed SNP to be more diverse compared to RNP (Fig 4).

Fig 4: Renyi diversities for SNP and RNP, the dots show the values for sites, and the lines the extremes (min max) and median in the data set for example in SNP diversity ranged maximum compared to RNP where diversity ranged mostly minimally. Calculations of relative species abundance showed Bradypodion xenorhinum and Chamaeleo johnstoni at 28% and 20% respectively to be the most abundant in RNP while Xenopus victorianus and Leptopelis christyiat 14% and 13% respectively as the most abundant in SNP. A plot for relative species abundance showed that most species were rare (fig 5). Relative species abundances follow very similar patterns over a wide range of ecological communities. When plotted as a histogram of the number of species represented by 1, 2, 3,..., n individuals usually fit a hollow curve, such that most species are rare, (represented by a single individual in a community sample) and relatively few species are abundant (represented by a large number of individuals in a community sample) Fig 5.

Fig 5: Relative species abundance of amphibians and reptiles sampled from Rwenzori-Semliki national parks showing the universal hollow curve. The different micro habitats from the two survey areas were clustered in R, the clustering produced eight different groups that are summarized in Fig 6. The groups are a result of association of different species with different habitat variables. Fig 6: An agglomerative Cluster dendrogram for habitat clustering depending on amphibian and reptile community composition

An indicator species analysis was completed using R analysis package and an indicator species list for each site group (or site group combination) was obtained. Five species were significantly associated with one, two or three of the groups, Chamaeleo rudis was significantly associated with group 3 at P=0.032, Leptopelis flavomaculatus was significantly associated with group 7 at P=0.003, Phrynobatrachus sp was significantly associated with group 8 at P=0.046, Bradypodion xenorhinum (Plate 1 (a)) was significantly associated with group 4 at P=0.019 while Chamaeleo johnstoni (Plate 1 (b)) was significantly associated with groups 4 and 6 at P=0.004. Plate 1 (a): Bradypodion xenorhinum (b) Chamaeleo johnstoni Comparison of species incidence with previous surveys The current survey recorded 37 (21 amphibians, 16 reptiles) and 28 (19 amphibians, 9 reptiles) amphibians and reptiles for RNP and SNP respectively (Appendix 1). The species are listed together with those known from studies in these sites by Mathias Behangana between 2002 and 2003. 15 species of amphibians were reported from previous studies but not in the current study, while three (Ptychadena anchietae, Leptopelis flavomaculatus and Xenopus victorianus) species where not recorded by the surveys of Mathias Behangana between 2002 and 2003. Hyperolius castaneus reported in SNP by Behangana (1996), was recorded at 2822m asl in RNP (Appendix 3).

Relationship between environmental variables and amphibian and reptile distribution It was assumed that environmental variables influence amphibian and reptile community structure. The assumed influence was evaluated using CCA ordination technique testing 3 constraining variables including elevation, relative humidity and temperature. The best explanatory model for amphibian and reptile community structure included two and three most important gradients for SNP and RNP respectively (table 1, figures 4 a and b). Table 1: Constraining environmental variables in the first two canonical axes (a) Semliki CCA1 CCA2 r2 Pr(>r) Temperature (Temp) 0.59859 0.80105 0.774 0.025* Relative Humidity (RH) -0.90735-0.42037 0.854 0.008** (b) Rwenzori Temperature 0.87991-0.47514 0.931 0.002** Relative Humidity 0.91545 0.40243 0.809 0.018* Elevation (ELEV) -0.86980 0.49341 0.922 0.005** The variables in table 1 are significant with strong positive correlations thus, very important in explaining species distribution. However, the partitioning of variance in the model indicates that the proportion explained by environmental variables is 23% for Semliki and 45% for Rwenzori. The permutation test indicate that the relationship is not significant (p = 0.254 and 0.051) for SNP and RNP respectively. By far, the result suggests that the selected environmental variables may not be influencing the distribution pattern of the species alone but rather with other contributing factors.

(a) (b) Figure 7: Triplots for amphibian and reptile species distribution pattern in (a) SNP and (b)rnp. Numbers represent sites, text-species (appendix 2) and arrows-environmental variables. The distribution pattern is more even across sites in relation to constraining variables in (a) while in (b), more species are clustered close to the centre (figure 7). The lengths of vectors indicate the strength of their correlation and arrow tips represent the magnitude of the impact of on species distribution; thus, Leptopelis kivuensis is associated with relative humidity in site 1 (figure 74a) while Amietia rwenzorica is correlated with elevation (figure 7b).. The variables are significant (table 2) with strong positive correlation in the first two axes. However, the proportion explained by environmental variables is only 26%, suggestive of a weak relationship although the permutation test indicates that the relationship is significant (p=0.001). Generally, the result suggests that selected environmental variables are not sufficient in explaining the distribution pattern of amphibian and reptile species by themselves, other factors might be critical.

Table 2: Constraining environmental variables in the first two canonical axes Variables CCA1 CCA2 r2 Pr(>r) Temperature 0.99764 0.06870 0.8986 0.001 *** Relative Humidity 0.99927 0.03831 0.6525 0.007 ** Elevation -0.91430-0.40504 0.9499 0.001 *** However, figure 8 portrays a different pattern for combined data. The random loading of sampling sites resulted into 4 clusters (marked with blue and red circles). Figure 8: A triplot for amphibian and reptile species distribution pattern in RNP and SNP Sites in RNP (blue circles) clustered in relation to elevation while sites in SNP formed a single cluster (red circle) in relation to temperature (figure 8). The clustering suggests that elevation and temperature are the most important constraining variables in structuring species distribution in RNP and SNP respectively. Discussion The patterns of hepertile diversity observed among study sites indicate that landscape structure distinctively affects species richness. The results suggest that differences in land cover characteristics and gradient have important effects on biodiversity in the study area. The

contribution of habitat elements to hepertilian species richness among study sites in areas with high elevation and low elevation suggest a more frequent exchange of species among heterogeneous habitats which are the low lands than in the homogeneous habitats in high elevations. Taxa-specific differences in habitat requirements and dispersal abilities among species are the likely causes for the differing species richness between the different study sites. Sixteen endemic reptile species occur in the Rift, of which Virunga National Park contains the highest number (11), followed by Rwenzori Mountains National Park (9) and Nyungwe National Park (8). Only two were recorded for this study -Bradypodion xenorhinum and Chamaeleo johnstoni. This could be attributed to the fact that most hepertiles in this region have not been exclusively studied therefore listed as Data Deficient by IUCN. However species such as Chamaeleo rudis need to be given a higher conservation category because their Area of occupancy and quality of habitat is reducing because of climate change effects and continued encroachment on RNP. Bradypodion xenorhinum was observed to be very abundant between 1500-2300m a.s.l, much as this is its preferred range it can be argued that most of the forest bellow has been cleared for farming or heavily disturbed. Our results have indicated that much as temperature, relative humidity and elevation may play a role in the distribution of species their effect is low; this could be because species distribution is influenced by many factors that act synergistically. Therefore to better quantify the effect of climate change on hepertiles, it s imperative to study various environmental factors at the same time. One notable effect of climate change is the distribution of species, along a gradient when climatic temperatures change, some species tend to move further up or below (examples), it was observed that species such as Chamaeleo rudis with a known limited distribution, it had records bellow 2000masl contrary to the known distribution between 2000-4000masl fig 2, some species now have more clumped or limited distribution for example the density of B. xenorhinum drastically falls below 1500masl nevertheless this study has provided more baseline data to be based on for future assessments for the various species. It was also observed that amphibian records were few and sparsely distributed above 2000m, which is contrary to the known fact that as conditions get warmer more species move up the elevation

but explains the fact that climatic change impacts are mostly taxa specific and result from a combination of various factors such as changes in precipitation, emerging and reemerging infections and habitat modification. Only one species of amphibian (Hyperolius castaneus) was recorded above 2000m asl in the sedges or flooded swamps in Kicucu. It is important to note that much as Hyperolius castaneus was the only species recorded above 2000masl, it was also recorded with low frequency. This could be because at elevations beyond 2000m amphibian diversity and abundance significantly shifts from widely distributed low land species to mostly endemic highland species such as Hyperolius castaneus (Sinsch et al 2011). The ongoing changes in climatic conditions could also explain the low frequency of amphibians at high elevations given their specific habitat requirements. More amphibian species were recorded below 1500m with the peak being in SNP. Low land habitats are more heterogeneous thus providing microhabitats for different species. Arthroleptis adolfifderichi a species known to occur in montane regions of eastern Democratic Republic of Congo, Uganda, Rwanda and Burundi with a lower elevation limit of 1780m asl was recorded in Semliki at 688. This then confirms the fact that some of the Albertine rift hepertile species are poorly known and need more detailed studies for proper evaluation by IUCN. Plate 2: Some of the Species from the genus Arthroleptis collected in SNP

References Aguti, C. (2007).Comparative study of amphibian diversity in Rwenzori mountains national park, Sango bay area and Semliki national park, Uganda. MSc Thesis, Zoology Environment and Natural Resources of Makerere University. Allmon, W.D. (1991). A plot study of forest floor litter frogs, central amazon, Brazil. Journal of Tropical Ecology 7:503-522. Behangana, M. (1995).The status of Amphibians in Rwenzori Mountains Western Uganda. MSc Thesis, Zoology Department, Makerere University, Kampala, Uganda. Behangana, M. (1996). Amphibians and reptiles (HERPS) of Rwenzori Mountains and Semliki Forest. In: The Rwenzori Mountains National Park, Uganda (Eds H. Osmaston, J. Tukahirwa, C. Basalirwa and J. Nyakaana). Department of Geography, Makerere University, Kampala, Uganda. Behangana, M., Meirte, D., Plumptre, A. J., Howell, K., Stuart, S. and Hinkel, H. (2003).In Plumptre.A. J., Behangana. M., Davenport. T. R.B., Kahindo. C., Kityo. R., Ndomba. E., Nkuutu. D., Owiunji.I.O., Ssegawa.P. and Eliu. G. (Eds). The Biodiversity of the Albertine Rift. Alberine rift technical reports series number. Bourgeron, Patrick S. (1983). "Spatial Aspects of Vegetation Structure". In Frank B. Golley. Tropical Rain Forest Ecosystems. Structure and Function. Ecosystems of the World (14A ed.). Elsevier Scientific. pp. 29 47. ISBN 0-444-41986-1. Blaustein, A. R., D. B. Wake. 1995. The puzzle of declining amphibian populations. Scientific American 272:52-57 Blaustein, A. R., Hoffman, J.M., Kiesecker, J. M., Chivers, D. P., Hokit, D. G., Marco, A., Belden, L. K., and Hatch, A. (1998).Effects of Ultra-violet radiation on amphibians: Field experiments. Am. zool.38: 799-812. Brito, J.C. & Crespo, E.G. (2002). Distributional analysis of two vipers (Vipera latastei and V. Seoanei) in a potential area of sympatry in the Northwestern Iberian Peninsula, Eagle mountain Publishing Co., Lancaster. Channing, A. & Howell, K.M. (2006). Amphibians of East Africa. Cornell University Press, Ithaca, NY. Charnier, M. (1966). Action de la température sur la sex-ratio chez l'embryon d'agama agama (Agamidae, Lacertilien). C. R. Soc. Biol. Paris 160, 620-622. data to identify environmental correlates. Conservation Biology15: 903 913.

Chefaoui, R.M., Hortal, J. & Lobo, J.M. (2005). Potential distribution modeling, niche characterization and conservation status assessment using GIS tools: a case study of Iberian Copris species. Biological Conservation 122: 327-338 Clarke, G. P. (2000). Climate and climatic history In Burgess, N. D. and G. P. Clarke (eds.). Coastal Forests of Eastern Africa. IUCN The world Conservation Union. Colwell, R.K. (2013) EstimateS: Statistical estimation of species richness and shared species from samples. Version 9. User's Guide and application published at: http://purl.oclc.org/estimates. Crump, M.L. (2010). Amphibian diversity and life history. In Amphibian ecology and conservation a handbook of techniques, K. C. Dodd Jr. (Ed.). pp. 1 19. Oxford University Press, Oxford, U.K. Davenport, T., Howard P., and Baltzer, M. (Eds). (1996).Forest department Biodiversity reports. Dettki, H., Lofstrand, R. & Edenius, L. (2003). Modelling habitat suitability for moose in c oastal northern Sweden: empirical vs process oriented approaches. Journal of Amphibian biology 32:549-556. Duellman, W.E. and Trueb, L. (1986).Biology of Amphibians. New York: McGraw- Hill. 670 pp. Gascon, C. (1996). Amphibian litter fauna and river barriers in flooded and non-flooded Amazonian rainforests. Biotropica 28: 136-140. Heyer, W. R., Donnely, M. A., Mcdiarmid, R. W., Hayek, L. A.C. and Foster. M. S. (1994) (Eds).Measuring and monitoring Biological Diversity; Standard methods for Reptiles and Amphibians. Smithsonian Institution Press, PA. Howell, K. (2002). Amphibians and Reptiles. The Hepertiles. In: African Forest Biodiversity: a field survey Manual for vertebrates (Ed. Davies, G.) Earth watch Institute (Europe), London. pp. 17-44. Kelly M. Hare in (2012).Inventory and monitoring toolbox: herpetofauna pitfall trapping DOCDM-760240 Lieberman, S.S. (1986). Ecology of the leaf litter herpetofauna of a neotropical rain forest: la selva, costa rica. Acta zool. Mex., ns. 15:1 71. Morales, J.J., Lizana, M. & Del Arco, C. (2002). Análisis de la distribución altitudinal de la herpetofauna en las sierras de Segovia. Revised Episode of Herpetology 16: 29-42

Miquel De C_aceres (2013). How to use the indicspecies package (ver.1.7.1), Centre Tecnol_ogic Forestal de Catalunya. Ctra. St. Lloren_c de Morunys km 2, 25280, Solsona, Catalonia, Spain. Msuya, C. A. (2001). Habitats, distribution and feeding of amphibians in Zaraninge Forest Reserve, Tanzania. A case study of the ecology of a community of amphibians in coastal forest ecosystem in Bagamoyo District, Tanzania. PhD Thesis, University of Dar es Salaam. Parker, G.G. and Brown, M.J. (2000). Forest canopy stratification: is it useful? The American Naturalist 155:473-484. Pounds, J. A. and M. L. Crump (1994). Amphibian declines and climate disturbance: The case of the golden toad and the Harlequin frog. Conserv. Biol. 8:72-85. Poynton, J.C. (1999).Distribution of Amphibians in sub-saharan Africa, Madagascar, and Seychelles. Pp 483-539. In Duellman, W.E. (Ed). Patterns of distribution of Amphibians; a global perspective. The Johns Hopkins University Press, Baltimore. Rödel, M.O., Lötters, S., Kosuch, J.and Veith, M., (2002).Amphibian monitoring in Africa (Côte d Ivoire, Kenya) and Asia (Borneo), Froglog, 49, 2. Sally, P., Nelson, R.J. & Lauzon, K. (1998). Hexactinellid sponge ecology: growth rates and seasonality in deep water sponges. Journal of Experimental Marine Biology and Ecology 230:111 129 Semlitsch, R. D. (2003). Amphibian Conservation. Washington DC: Smithsonian Institution. Sinsch, U. (1990). Migration and orientation in anuran amphibians. Ethology Ecology and Evolution 2: 65-79. Sinsch, U., Greenbaum, E., Kusamba, C. & Lehr, E. (2011). Rapid assessment of montane anuran communities in the Albertine Rift: Hyperolius castaneus Ahl, 1931 as a focal species for conservation. African Zoology, 46, 320 333. Soares, C. & Brito, J. (2007). Environmental correlates for species richness among amphibians and reptiles in a climate transition area. Biodivers Conserv 16:1087 1102. doi:10.1007/s10531-006-9070-5 Stuart, S., Chanson, J.S., Cox, N.A., Young, B.E., Rodrigues, A.S.L., Fishman, D.L. & Waller, R.W. (2004). Status and trends of amphibian declines and extinctions worldwide. Science 306: 1783-1786.

ter Braak, C.J.F. & Verdonschot, P.F.M. (1995). Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquatic Science 57: 255 289. Wake DB, Vredenburg VT (2008). Are we in the midst of the sixth mass extinction? A view from the world of amphibians. Proc Natl Acad Sci USA. 2008;105:11466 11473. http://dx.doi.org/10.1073/pnas.0801921105

Appendix 1: Comparison of amphibian and reptile records for current study and previous known studies Species Rwenzori now Rwenzori prev Semliki now Semliki prev Acanthocercus_atricollis Adolfus_jacksoni Afrixalus fulvovittatus x x Afrixalus_osorio Agama_agama Amietia_angolensis x x Amietia_ruwenzorica x x Amietophrynus gutturalis x Amietophrynus kisoloensis x x Amietophrynus_maculatus x Amietophrynus_regularis Amietophrynus_vittatus x x Amnirana_albolabris x Arthroleptis_adolfifriederici x Atheris_nitschei Bitis_arietans Chamaeleo_cf._adolfifriderici Chamaeleo_ellioti Chamaeleo_johnstoni Chamaeleo_rudis Chamaeleo_xenorhinus Cnemaspis_quatuorseriatus Grayia_tholoni Hapsidophris_lineata Hapsidophrys_smaragdina Hemidactylus_brooki Hemidactylus_sp Hoplobatrachus_occipitalis x x Hyperolius acuticeps x x Hyperolius discodactylus x Hyperolius_cinnamomeoventris x Hyperolius_kivuensis x x Hyperolius_lateralis x Hyperolius_castaneus x Hyperolius_viridiflavus x Kassina_senegalensis x x Leptopelis_christyi x x Leptopelis_kivuensis

Leptopelis_flavomaculatus Mabuya_maculirabris Mabuya_striata Naja_melanoleuca x Phlyctimatis verrucosus x Phrynabarachus bequaerti x x Phrynobatrachu_auritus Phrynobatrachus acutirostris cf x Phrynobatrachus asper x Phrynobatrachus dalcqi x Phrynobatrachus versicolor x Phrynobatrachus_acridoies x x Phrynobatrachus_dendrobates x x Phrynobatrachus_graueri x x Phrynobatrachus_natalensis x x Phylothamnus_angolensis Psamophis_mossambicus Ptychadena christyi x Ptychadena porosissima x Ptychadena_anchiatae Ptychadena_chrysogaster x Ptychadena_mascareniensis x Rhampholeon_boulengeri Schoutedenalla schubotzi x Scoutedenela discodactyla x Xenopus fraseri x Xenopus ruenzoriensis x Xenopus wittei x Xenopus_laevis x x Xenopus_victorianus Appendix: 2 Species list (Triplot) Acanatri Adlfs_jc Afrxls_s Amt_ngln Amt_rwnz Amtphryns_r Amtphryns_m Amtphryns_v Amnrn_lb Arthrlp_ Acanthocercus-atricollis Adolfus_jacksoni Afrixalus_osorio Amietia_angolensis Amietia_ruwenzorica Amietophrynus_regularis Amietophrynus_maculatus Amietophrynus_vittatus Amnirana_albolabris Arthroleptis_adolfifriederici

Athrs_nt Atheris_nitschei Bts_rtns Bitis_arietans Chamadol Chamaeleo_cf.adolfifriderici Chml_llt Chamaeleo_ellioti Chml_jhn Chamaeleo_johnstoni Chml_rds Chamaeleo_rudis Chml_xnr Chamaeleo_xenorhinus Cnmsps_q Cnemaspis_quatuorseriatus Gry_thln Grayia_tholoni Hplbtrc_ Haplobatrachus_occipitalis hpsdphr_ hapsidophrys_lineata Hpsdphr_ Hapsidophrys_smaragdina Hmdctyls_b Hemidactylus_brooki Hmdctyls_s Hemidactylus_sp Hyprls_ct Hyperolius_acuticeps Hyprls_cn Hyperolius_cinnamomeoventris Hyprls_k Hyperolius_kivuensis Hyprls_l Hyperolius_lateralis Hyprls_s Hyperolius_sp Hyprls_v Hyperolius_viridiflavus Kssn_sng Kassina_senegalensis Lptpls_c Leptopelis_christyi Lptpls_k Leptopelis_kivuensis Lptpls_s Leptopelis_sp Mby_mclr Mabuya_maculirabris Mby_strt Mabuya_striata Nj_mlnlc Naja_melanoleuca Phrynbtrchs_c Phrynobatrachus_acridoies Phrynbtrchs_r Phrynobatrachus_auritus Phrynbtrchs_d Phrynobatrachus_dendrobates Phrynbtrchs_n Phrynobatrachus_natalensis Phrynbtrchs_s Phrynobatrachus_sp Phylthm Phylothamnus_angolensis Psmphs_m Psammophis_mossambicus Ptychdn_n Ptychadena_anchietae Ptychdn_c Ptychadena_chrysogaster Ptychdn_m Ptychadena_mascareniensis Rhmphln_ Rhampholeon_boulengeri Xnps_vct Xenopus_victorianus