Técnica do Glossário Neológico Especializado

Herpetological Conservation and Biology 10(2):610–620. Submitted: 24 March 2015; Accepted: 3 August 2015; Published: 31 August 2015. IS HYALOSAURUS KOELLIKERI A TRUE FOREST LIZARD? DANIEL ESCORIZA1, 3 AND MAR COMAS2 1 Institute of Aquatic Ecology and Department of Environmental Science, University of Girona, Campus Montilivi, Faculty of Sciences, 17071 Girona, Spain 2 Estación Biológica de Doñana (EBD-CSIC), Avda, Américo Vespucio s/n, 41092 Sevilla, Spain 3 Corresponding author, e-mail: [email protected] Abstract.―Hyalosaurus koellikeri is the only member of the family Anguidae found on the African continent. It has been suggested that its distribution is dependent on forest habitats. Here, we investigated habitat selection of this secretive species in Morocco, at a fine spatial resolution. We examined the association between the presence of H. koellikeri and vegetation cover, substrate, topography, and climate. We then compared the habitat selection of this species and that of other lizards in the same region. Our results indicate that the proportion of tree crown cover is the variable that best explains the occurrence of this species, and to lesser extents, the climatic, substrate, and topographic conditions. However, this species is not associated with a particular class of forest and is present in both dry assemblages of evergreen oaks and junipers and in humid deciduous forests. Hyalosaurus koellikeri occupies forests more frequently than other genera of lizards, though it coexists with several generalist species in these habitats. This association with forest/woodland habitats could explain the sparse distribution of this species in Morocco and suggests that some isolated populations, like those of the Atlantic Plain, are vulnerable to extinction. Key Words.―Anguidae; conservation; Morocco; secretive species INTRODUCTION The anguids are a group of lizards typically found in the mesic environments of Eurasia and America (Campden-Main 1970; Stebbins 1985; Valakos et al. 2008). Hyalosaurus koellikeri (Fig. 1) is the only species present on the African continent, and is the product of a basal split within the family (Macey et al. 1999). This species is endemic to northwest Africa, and is found in central Morocco and extreme northwestern Algeria (Sindaco and Jeremĉenko 2008). In Morocco, it is a secretive and sparsely distributed species, apparently limited to herbaceous and forest habitats with humid conditions (Bons and Geniez 1996; Schleich et al. 1996; de Pous et al. 2011; Escoriza and Comas 2011). However, to date, no habitat selection studies have been undertaken at a fine spatial resolution. Knowledge of the habitat preference of the species is essential to its conservation, particularly considering the increasing land-use changes in the region (Nafaa and Watfeh 2000) and the known impact of habitat loss on anguid populations (e.g., Pseudopus apodus in Bulgaria; Beshkov and Nanev 2006). In this study, we characterized the habitats in which H. koellikeri occurs in Morocco. We measured various forest features related to stand maturity and structural complexity, both of which determine the composition of faunal assemblages in these habitats (Schonberg et al. 2004; Guénette and Villard 2005). The characteristics of the forest habitats were assessed at several scales, each Copyright © 2015. Daniel Escoriza All Rights Reserved. describing different ecological aspects of low-dispersal vertebrates, ranging from habitat use (smaller scale) to metapopulation stability (larger scale; Hanski 1991; Gamble et al. 2007). We also examined the substrate properties, given that soil is a component of the anguid habitat (Gregory 1980) and determines habitat suitability for several species of lizards (Kuhnz et al. 2005; Disi 2011). Finally, we evaluated the effect of topography and climate, because the occurrence of ectothermic vertebrates is explained by these variables, especially in warm and semiarid environments (Monger and Bestelmeyer 2006). The habitat selected by H. koellikeri was compared with that of other lizard genera inhabiting the same region. In central Morocco, lizards are particularly diverse and include lineages of very distinct biogeographical origins (Bons and Geniez 1996). They occupy most Moroccan ecoregions, from subtropical scrub to temperate montane grasslands (Schleich et al. 1996). However, forests are not favourable environments for many species of lizards because they are heliothermic (Huey 1974; Vitt et al. 1997). If H. koellikeri is a forest species, its association with this type of habitat should be greater than those of other groups of lizards. In this article, we report the habitat use by a little-known anguid species endemic to northwestern Africa. Our hypotheses were that the lizard is a forest species and is therefore more strongly associated with this type of habitat than are other lizards, 610 Escoriza and Comas.—Habitat of Hyalosaurus koellikeri. FIGURE 1. Adult Hyalosaurus koellikeri, Forêt de la Maâmora, Atlantic Plain of Morocco. (Photographed by Daniel Escoriza). and that the sparse distribution of the species could be Survey protocol and habitat characterization.―We attributable to changes in land use to agriculture. conducted several surveys from March to April and October to November during 2008–2013, as part of a study of north-African amphibian and reptile species MATERIALS AND METHODS (Escoriza 2010; Comas et al. 2014; Escoriza and Ben Study area.―The study area covered most of northern Hassine 2014, 2015). We selected survey sites to cover Morocco, between Essaouira to the south, Debdou to the most of the range of the species (rather than selection by east, and the Tingitana Peninsula to the north (Fig. 2). habitat condition) to avoid biases related to the The region falls within the dry-summer subtropical expectations of the researchers. We conducted sampling climate belt (Köppen classification), and the high- between 1100 and 1700 (local time) on sunny days altitude mountain ranges generate an important climatic (cloudiness <30%), and involved turning over stones and gradient that allows the growth of very heterogeneous small logs. Rock/log flipping and visual surveys are plant communities, combining Mediterranean, Tethyan, techniques commonly used in qualitative sampling of squamates (Lovich 2012). It also proved to be useful in and Sahelian elements (Charco 2001). surveys of other groups of ectothermic vertebrates that inhabit forest habitats (Crawford and Semlitsch 2007). The fieldwork was restricted to two person-hours in 2.5 ha, and ended when the selected portion of land was completely sampled, the time limit was reached, or a specimen of H. koellikeri was found. We also recorded other species of lizards and amphisbaenians found during the surveys. We defined absence using two criteria: no confirmed observation of H. koellikeri during the surveys or the absence of previous records in the area (10 km radius). For habitat characterization, we used the Food and Agriculture Organization (FAO) definition of forest: a land in which tree crown cover exceeds 10% of coverage, totalling a minimum of 0.5 ha and where trees reach a minimum height of 5 m at maturity (FAO. 2000. Available at http://www.fao.org/docrep/ [Accessed 20 FIGURE 2. The study region in Morocco (northwestern Africa). White circles: presence localities of Hyalosaurus koellikeri. Black March 2014]). Habitats composed of smaller dwarf diamonds: localities of absence. and/or stunted trees are categorized as wooded land. We 611 Herpetological Conservation and Biology delimited an area of 100 m2 around each point at which a specimen of H. koellikeri was found. Within this perimeter, we counted (n) all living trees (woody plants with a trunk diameter > 5 cm and height > 1.37 m; McKenny et al. 2006), and we measured the diameter at breast height (in cm, DBH; at 1.4 m). We calculated the tree density (ρ = n/100) and then, using DBH, estimated the tree basal area (BA = 0.00007854 * DBH²) in m², which is the portion of land surface occupied (Grove 2002). In the analyses, we used the ratio of summed BA to total ground surface (ΣBA/100). We also measured the height of every living tree (as defined above) using a clinometer. Three transects (50 × 2 m) were established, running from the specimen location and oriented north, southwest, and southeast (Guénette and Villard 2005). Along these transects, we counted all species of trees and shrubs (woody plants < 1.37 m tall) and classified them to species level according to Charco (2001). We used the counts of trees and shrubs to compute three indices of diversity (Keylock 2005): dominance, the Shannon-Wiener index (H), and the variation in Shannon-Wiener index (Hvar), an estimate of the spatial heterogeneity in plant composition. We calculated Hvar using a bootstrapped estimation (after 1000 replications) of the variation in the Shannon index when the north and the southwest transects were compared. For sites at which no H. koellikeri were found, we followed the same protocol so that the same number of sites was processed for both the presence and unconfirmed presence groups. We calculated diversity indices using the package PRIMER-E (PRIMER-E Ltd., Plymouth, UK). Geographic Information System (GIS) and remote sensing data.―We obtained site-specific data summarizing the climate, topography, substrate, and proportion of vegetation cover from GIS layers. We described the climatic conditions using two key variables that explain the composition of vertebrate assemblages in the Palaearctic ecozone (Escoriza and Ruhí 2014): the aridity index (mean annual precipitation /mean annual potential evapotranspiration) and the mean annual temperature (Hijmans et al. 2005; Consortium for Spatial Information. 2009. Available at http://www.cgiarcsi.org/ [Accessed 8 September 2013]). We measured topography with an index of terrain ruggedness (Riley et al. 1999). This index describes the variation in altitude within a 3 × 3 pixel grid and ranges between 0 (level terrain) and 959 (extremely rugged terrain). We examined substrate characteristics at the first standard depth (0–5 cm): bulk density (soil compactation; t/m3), cation exchange capacity (soil fertility; cmol/kg), soil organic carbon (g/kg), pH, sand (grain size 50–2000 µm) content, silt (grain size 2–50 µm) content, and clay (grain size < 2 µm) content (International Soil Reference and Information Centre. 2015. Available at http://www.isric.org/content/ data [Accessed 9 January 2015]). Loose soils have a bulk density lower than 1.3 t/m3, whereas medium compacted soils have a bulk density of 1.45 t/m3 (Chesworth 2008). We examined vegetation cover (percentage coverage per km²) for four classes: forest, shrubs, herbaceous vegetation, and cultivated vegetation, as provided by Global 1-km Consensus Land Cover (Tuanmu and Jetz. 2014. Available at http://www.earthenv.org/landcover. html [Accessed 15 September 2014]). These data were extracted with a spatial resolution of 30 arc-seconds (1 km) using the package QGIS vs 2.6.1 (Quantum-GIS Development Core Team. 2015. Available at http://qgis.osgeo.org [Accessed 10 January 2015]). We also examined vegetation cover at a higher spatial resolution from satellite and aerial photographs (Google Earth 2014). This approach allowed us to determine the proportion of tree crown cover in a ground surface of 0.0625 km2. We calculated the multispectral range based on a supervised classification routine implemented in the package MultiSpec vs.3.4 (Purdue Research Foundation 2014). Data analyses.―A Principal Coordinate Analysis (PCO) was used to generate an ordination plot of the presence and absence (or unconfirmed presence) data. First, we constructed a Euclidean distance matrix based on the variables obtained from the characterization of forest, climate, topography, and substrate. We then tested for differences between both groups using permutational multivariate analysis of variance (PERMANOVA), assessing the significance after 9,999 unrestricted permutations. The explanatory capacity of the environmental parameters was determined with a distance-based linear model, using a presence/absence matrix (constructed with Sørensen’s coefficient) as the dependent variable. To determine the variables that contributed most to the presence of H. koellikeri, we generated an optimal model using sequential tests, with the variables fitted as covariates (Anderson et al. 2008). We constructed the model using a step-wise procedure, using the Akaike Information Criterion (corrected for finite sample sizes; AICc) to select the best model (Burham and Anderson 2002). We determined significance with 99,999 permutations of the normalized predictor data. This analysis was performed using the package PRIMER-E (PRIMER-E Ltd., Plymouth, UK). We also investigated whether the distribution pattern of H. koellikeri could be explained by land-use changes. To do this, an ecological niche model was run, with 75% of the locations allocated to training and a regularization value of one, enabling linear, quadratic, and hinge features based on the number of presence data for the species (Anderson and Gonzalez 2011). This ecological 612 Escoriza and Comas.—Habitat of Hyalosaurus koellikeri. TABLE 1. Forest and climate class (UNEP 1997) for presence localities of Hyalosaurus koellikeri Latitude 34.25 34.21 34.09 34.06 34.05 34.05 34.01 33.98 33.87 33.62 33.60 33.56 33.48 33.47 33.44 33.37 33.22 32.52 31.26 Longitude -6.66 -6.57 -4.09 -6.54 -4.23 -4.16 -6.61 -3.01 -3.18 -5.31 -4.90 -5.20 -6.17 -5.13 -6.07 -5.26 -5.25 -6.02 -7.81 Elevation (m) 8 60 1,381 163 1,149 1,531 169 1,286 1,218 1,279 1,525 1,517 940 1,605 1,072 1,625 1,936 1,347 1,624 Dominant species Juniperus phoenicia Quercus suber Quercus ilex Quercus suber Quercus suber Quercus ilex Quercus suber Juniperus oxycedrus Quercus ilex Quercus ilex Quercus ilex Quercus canariensis Quercus ilex Fraxinus dimorpha Quercus suber Quercus ilex Quercus ilex Tetraclinis articulata Juniperus oxycedrus niche model included two climatic variables (aridity index, mean annual temperature) and four classes of vegetation cover (forest, shrubs, herbaceous, and cultivated vegetation). We calculated model accuracy based on the area under the receiver operating characteristic curve (AUC). This analysis was performed using the package MaxEnt 3.3.3k (Phillips et al. 2006). Finally, we examined whether this species is associated with forest habitats more frequently than are other species of lizards occurring in the same region. To do this, we compared the habitat characteristics based on the records that we collected during the surveys. We defined habitats by the four classes of vegetation cover (forest, shrubs, herbaceous, and cultivated vegetation). We compared the observed associations with those obtained under the null distribution (after 99,999 iterations) using an equiprobable constraint to randomize the occurrences. If the length of the longest run is greater than that obtained under the null distribution, the occurrences are aggregated (Gotelli and Entsminger 2009). We performed this analysis using the Runs tests provided in the package Ecosim 7.72 (Gotelli and Entsminger 2009). The contribution of the four classes of vegetation to lizard occurrence in the region was examined with PCO and a similarity percentage (SIMPER) analysis, based on a standardized matrix of Bray-Curtis distances (Clarke and Gorley 2006). These analyses were carried out by the package PRIMER-E (PRIMER-E Ltd., Plymouth, UK). Forest class Evergreen needleleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen needleleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Deciduous broadleaved Evergreen broadleaved Deciduous broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen broadleaved Evergreen needleleaved Evergreen needleleaved Climate class Semi-arid Semi-arid Semi-arid Semi-arid Semi-arid Semi-arid Semi-arid Semi-arid Semi-arid Dry sub-humid Semi-arid Humid Semi-arid Humid Dry sub-humid Humid Humid Semi-arid Semi-arid in March and April, all under a rock or log. Site characterization revealed that H. koellikeri was present in forest and/or woodland (66% average tree crown cover per 0.0625 km2; Appendix 1), mainly composed of evergreen oaks (Quercus ilex and Q. suber; 68.43%) and junipers (Juniperus oxycedrus, J. phoenicia, and Tetraclinis articulata; 21.05%), whereas deciduous forests (Q. canariensis and Fraxinus dimorpha) together accounted for no more than 10.52% of the sites containing H. koellikeri (Table 1). These forests and woodlands contained trees of variable sizes, and were usually dominated by a few plant species (Appendix 1). The substrate was loosely compacted, slightly acidic, and contained abundant organic matter (Appendix 1). The topographic and climatic conditions varied widely among these sites (Table 1 and Appendix 1). The PCO ordination plot indicated that the two groups (sites at which H. koellikeri was present, those from which its presence was not confirmed) were distributed unevenly in ecological space (Fig. 3). PERMANOVA confirmed the significant differences between these groups (pseudo-F1,36 = 6.773, P < 0.001). Sequential tests indicated that the most important variable explaining the occurrence of H. koellikeri was the proportion of tree crown cover, and this variable alone explained 40% of the observed variance. The best model to explain the occurrence of H. koellikeri (75% of the observed variance) included the proportion of tree crown cover, climate (aridity and mean temperature), substrate (organic soil and pH), terrain ruggedness, and tree/shrub diversity (Table 2). The ecological niche model produced an AUC of RESULTS 0.941, indicating a very high predictive value. This We found H. koellikeri at 19 sites (Fig. 2) and model indicates that the populations of H. koellikeri can obtained 154 records for 17 genera and 28 species of be divided into three main clusters (Atlantic Plain, lizards. Most of the H. koellikeri specimens were found Middle and High Atlas mountain ranges, and Tell Atlas 613 Herpetological Conservation and Biology FIGURE 3. Principal coordinates ordination plot showing the distribution in the environmental space of presence/absence localities. 250: % of tree crown cover at 250 m resolution; Bulk: soil bulk density (t/m3); Cation: soil cation exchange (cmol/kg); Organic: organic soil matter (g/kg); pH: soil pH; Sand, Silt and Clay: % in soil; Aridity: aridity index; Tº C: mean annual temperature (º C); Rugg: terrain ruggedness; Basal: tree basal area (m2) in 100 m2; Height: tree average height (m) in 100 m2; Density: tree density in 100 m2; Taxa: number of species of trees/shrubs; N: total number of tree/shrubs. DOM: Dominance of tree/shrubs species; H: Shannon-Wiener index of tree/shrubs species; Hvar: variation in Shannon-Wiener index of tree/shrubs species. mountains), isolated by broad areas of unsuitable habitat (Fig. 4). The ecological niche model also showed that the occurrence of H. koellikeri is mainly dependent on vegetation cover (Table 3). The analysis of null models indicated that the occurrence H. koellikeri is associated with forest more than would be expected if the distribution of the species were random (observed runs = 6, simulated runs = 2.15, P < 0.001). A SIMPER analysis showed that H. koellikeri appeared more frequently in forest habitats than other lizard species (Table 4), although there was an important overlap with other genera, as shown on PCO first factorial plane (Appendix 2). DISCUSSION Hyalosaurus koellikeri is a remarkable lizard endemic to northwestern Africa, with a discontinuous distribution (Schleich et al. 1996). In this study, we investigated the habitat preferences of the species to clarify the factors TABLE 2. Optimal model obtained by distance-based linear model sequential tests explaining the occurrence of Hyalosaurus koellikeri. underlying its distribution. The regional occurrence of The abbreviation Cumul. = cumulative proportion of explained lizards can be successfully described based on models variance. that include climate, vegetation, and substrate descriptors, because these variables are related to the AICc Pseudo‒F P Cumul. thermoregulatory efficiency of reptiles (Melville et al. Tree crown 198.68 24.12 < 0.001 0.40 2001; Jácome-Flores et al. 2015). Our findings also cover at 250 m indicate that the occurrence of H. koellikeri in central+Aridity index 190.85 10.76 0.002 0.54 western Morocco was mainly explained by the presence +Soil organic 188.15 4.99 0.029 0.6 of forest/wooded land, and to a lesser extent by other abiotic factors, including substrate, topography, and +Soil pH 186.04 4.41 0.043 0.65 climatic conditions. This anguid does not have a strong +Terrain 184.14 4.25 0.048 0.69 preference for a specific habitat composition and was ruggedness found occupying both dry forests of junipers and +Tree/shrub 182.12 4.40 0.045 0.73 evergreen oaks and humid deciduous forests. In contrast diversity with the results obtained by de Pous et al. (2011), in our +Mean annual 181.45 3.24 0.083 0.75 temperature study the presence of H. koellikeri in closed deciduous forest was more marginal (10.52%). 614 Escoriza and Comas.—Habitat of Hyalosaurus koellikeri. FIGURE 4. Ecological niche model obtained based on two climatic variables (aridity index and mean annual temperature) and four classes of vegetation cover (forest, shrubs, herbaceous, and cultivated vegetation). Black circles: presence localities of Hyalosaurus koellikeri. These forests/woodlands can be composed of trees of variable heights that are spaced densely to very sparsely. These results indicate that H. koellikeri selects a habitat similar to that described for Pseudopus apodus, which inhabits open forests of pines and evergreen oaks (Rifai et al. 2005; Franzen et al. 2008; Valakos et al. 2008). Hyalosaurus koellikeri is not a substrate specialist, tolerating wide variations in the proportions of sand, silt, and clay. On the Atlantic Plain, it can occur in soils containing a high proportion of sand (82%), but in other regions, this proportion is lower, at around 47%. However, in most of the sites studied, the substrate was only slightly compacted (below 1.39 t/m3), rich in organic matter and acidic, typical of forested areas (Islam and Weil 2000). Compared with other genera of lizards, H. koellikeri was more frequently associated with forests, although other species of lizards also occur in this type of habitat. Both stenoecious species (e.g., Chalcides lanzai) and generalist species (e.g., Agama bibronii, Tarentola mauritanica, Trogonophis wiegmanni) occur sympatrically with H. koellikeri. The presence of H. koellikeri in forested habitats is partly attributable to the ability of anguids to forage while their body TABLE 3. Relative contributions of the environmental variables to the ecological niche model for Hyalosaurus koellikeri (see Fig. 4). Variables Forest Cultivated vegetation Annual temperature Herbaceous vegetation Shrubs Aridity index Percentage contribution 51.5 29.0 17.0 2.1 0.3 0.1 Permutation importance 9.7 44.5 15.6 27.7 2.5 0.1 temperatures are low (Avery 1982; Capula and Luiselli 1993). Like other lizards, H. koellikeri displays heliothermic behavior. Nevertheless, it is active at dawn, at dusk, and on rainy days (Schleich et al. 1996). Activity with relatively low body temperatures could be related to the lower vulnerability of anguids to predation than other lizards, because they have heavily armoured skins (Hailey 1984; Meek 1986). The association between H. koellikeri and the presence of forest/woodlands means that this species is vulnerable to factors that regulate the extent of this biome. In the last 30,000 y, the area covered by northwestern African forests has undergone major contractions (RodríguezSánchez and Arroyo 2008; Alba-Sánchez et al. 2015). These contractions reached the peak during the hyperarid periods associated with glaciation, when the forests were reduced to the Atlas Mountains (Rhoujjati et al. 2010). The survival of H. koellikeri in a few refugia at that time would explain its relatively low haplotype diversity (de Pous et al. 2011), similar to that described for other mesophilic species occurring in the region (Recuero et al. 2007). We believe that our results have important implications for the conservation of this species. The ecological niche model indicates that some populations located on the Atlantic Plain (forêt de Maâmora and Sidi Boughaba) are probably isolated from the core distribution of the species. The Atlantic Plain was covered by forests, woodland, and scrub before the onset of agriculture (United States Geological Survey. 2013. http://www. rmgsc.cr.usgs.gov/ Available from ecosystems [Accessed 28 September 2014]), indicating the presence of suitable conditions for the species in most of this region until historical times. However this landscape transformation becomes very intense in the 615 Herpetological Conservation and Biology TABLE 4. Comparison on the association between occurrence and vegetation cover between the genus Hyalosaurus and other lizard genera occurring in the region, obtained by similarity percentage routine. Cover Type H. koellikeri Other genus Average dissimilarity Contribution% Forest Cultivated Shrubs Herbaceous 31.18 42.44 24.82 1.56 19.81 40.42 33.54 6.23 15.26 14.00 13.56 3.59 32.88 30.18 29.21 7.73 last hundred years (Emberger 1939; Mhirit and Blerot 1999), making these populations highly vulnerable to extinction. Other peripheral populations are already very rare and may even be extinct. In Essaouira, Morocco, this species has not been recorded since 1931(Bons and Geniez 1996). However, in the Middle and High Atlas Mountains, there are still vast areas with suitable conditions for the species, where H. koellikeri could maintain stable populations in the long term. Acknowledgments.―Fieldwork in Morocco was authorized by various scientific permits provided by the Haut Commissariat aux Eauxet Forêts et à la Lutte Contre la Désertification (HCEFLCD /DLCDPN / CFF), Rabat, Morocco. This work was partly supported by a grant from the Mohamed bin Zayed Species Conservation Fund (Project 13057819) and a Severo Ochoa contract supported Mar Comas (ref: SVP-2014-068620). LITERATURE CITED Alba-Sánchez, F., J.A. López-Sáez, D. Nieto-Lugilde, and J.C. Svenning. 2015. Long‐term climate forcings to assess vulnerability in North Africa dry argan woodlands. Applied Vegetation Science 18:283–296. Anderson, M.J, R.N. Gorley, and K.R. Clarke. 2008. PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods. University of Auckland, Auckland, New Zealand. Anderson, R.P., and I. Gonzalez. 2011. Species-specific tuning increases robustness to sampling bias in models of species distributions: an implementation with MaxEnt. Ecological Modelling 222:2796‒2811. Avery, R.A. 1982. Field studies of body temperatures and thermoregulation. Pp. 93‒166 In Biology of the Reptilia, Vol. 12. Gans, C., and F.H. Pough (Eds.). Academic Press, New York, USA. Beshkov, V., and K. Nanev. 2006. Amphibians and Reptiles in Bulgaria. Pensoft, Sofia, Bulgaria. Bons, J., and P. Geniez. 1996. Amphibiens et Reptiles du Maroc. Asociación Herpetológica Española, Barcelona, Spain. Burnham, K.P., and D.A. Anderson. 2002. Model Selection and Multivariate Inference: A Practical Information and Theoretic Approach. Springer, New York, New York, USA. Campden-Main, S.M. 1970. The first record of Ophisaurus gracilis (Gray) (Sauria: Anguidae) in South Vietnam. Herpetologica 26:17‒18. Capula, M., and L. Luiselli. 1993. Ecology of an alpine population of the Slow Worm, Anguis fragilis Linnaeus, 1758 thermal biology of reproduction. Herpetozoa 6:57‒63. Charco, J. 2001. Guía de los Árboles y Arbustos del Norte de África. Ediciones Mundo Árabe e Islam, Madrid, Spain. Chesworth, W. 2008. Encyclopedia of Soil Science. Springer, Dordrecht, The Netherlands. Clarke, K.R., and R.N. Gorley. 2006. PRIMER v6: User Manual/Tutorial. PRIMER-E, Plymouth, UK. Comas, M., D. Escoriza, and G. Moreno-Rueda. 2014. Stable isotope analysis reveals variation in trophic niche depending on altitude in an endemic alpine gecko. Basic and Applied Ecology 15:362‒369. Crawford, J.A., and R.D. Semlitsch. 2007. Estimation of core terrestrial habitat for stream‐breeding salamanders and delineation of riparian buffers for protection of biodiversity. Conservation Biology 21:152–158. de Pous, P., E. Mora, M. Metallinou, D. Escoriza, M. Comas, D. Donaire, J.M. Pleguezuelos, and S.Carranza. 2011. Elusive but widespread? The potential distribution and genetic variation of Hyalosaurus koellikeri (Günther, 1873) in the Maghreb. Amphibia-Reptilia 32:385–397. Disi, A.M. 2011. Review of the lizard fauna of Jordan: (Reptilia: Sauria). Zoology in the Middle East 54:89‒102. Emberger, L. 1939. Aperçu général sur la végétation du Maroc. Pp. 40–157 In Ergebnisse der Internationalen Pflanzengeographischen Exkursion durch Marokko und Westalgerien, 1936. Rubel, E., and W. Ludi (Eds.). Verlag Hans Huber, Bern, Switzerland. Escoriza, D. 2010. Ecological niche modeling of two Afrotropical snakes: is the Sahara desert a true barrier for these species? Revista Española de Herpetologia 24:93‒100. Escoriza, D., and M. Comas. 2011. Nova cita de Hyalosaurus koellikeri (Squamata: Anguidae) a la regió de Debdou (Nord-est de Marroc). Butlletí de la Societat Catalana d’Herpetologia 19:122‒124. 616 Escoriza and Comas.—Habitat of Hyalosaurus koellikeri. Escoriza, D., and J. Ben Hassine. 2014. Phenotypic variability in larvae of two species of Mediterranean geometric morphometrics. African Journal of Herpetology 63:152‒165. Escoriza, D., and J. Ben Hassine. 2015. Niche partitioning at local and regional scale in the North African Salamandridae. Journal of Herpetology 49:276–283. Escoriza, D., and A. Ruhí. 2014. Macroecological patterns of amphibian assemblages in the Western Palearctic: implications for conservation. Biological Conservation 176:252‒261. Franzen, M., M. Buβmann, T. Kordges, and B. Thiesmeier. 2008. Die Amphibien und Reptilien der Südwest-Türkei. Laurenti verlag, Bielefeld, Germany. Gamble, L.R., K. McGarigal, and B.W. Compton. 2007. Fidelity and dispersal in the pond-breeding amphibian, Ambystoma opacum: implications for spatio-temporal population dynamics and conservation. Biological Conservation 139:247‒257. Gotelli, N. J., and G.L. Entsminger. 2009. EcoSim: Null Models Software for Ecology. Version 7. Acquired Intelligence Inc. & Kesey-Bear, Jericho, Vermont, USA. Gregory, P.T. 1980. Physical factor selectivity in the fossorial lizard Anguis fragilis. Journal of Herpetology 14:95‒99. Grove, S.J. 2002. Tree basal area and dead wood as surrogate indicators of saproxylic insect faunal integrity: a case study from the Australian lowland tropics. Ecological Indicators 1:171‒188. Guénette, J.S., and M.A. Villard. 2005. Thresholds in forest bird response to habitat alteration as quantitative targets for conservation. Conservation Biology 19:1168‒1180. Hailey, A. 1984. Thermoregulation and activity metabolism of the armoured anguid Ophisaurus apodus. British Journal of Herpetology 6:391‒398. Hanski, I. 1991. Single‐species metapopulation dynamics: concepts, models and observations. Biological Journal of the Linnean Society 42:17‒38. Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones, and A. Jarvis. 2005. Very high resolution interpolated global terrestrial climate surfaces. International Journal of Climatology 25:1965‒1978. Huey, R.B. 1974. Behavioral thermoregulation in lizards: importance of associated costs. Science 184:1001‒1003. Islam, K.R., and R.R. Weil. 2000. Land use effects on soil quality in a tropical forest ecosystem of Bangladesh. Agriculture, Ecosystems & Environment 79:9‒16. Jácome-Flores, M.E., M.C. Blazquez, V.J. Sosa, and Y. Maya. 2015. Type of soil and temperature range explain the preferred habitat and current distribution of the endemic lizard Aspidoscelis hyperythra in spadefoot toad: an approach using linear southern Baja California peninsula. Journal of Arid Environments 113:126‒133. Keylock, C.J. 2005. Simpson diversity and the ShannonWiener index as special cases of a generalized entropy. Oikos 109:203–207. Kuhnz, L.A., R.K. Burton, P.N. Slattery, and J.M. Oakden. 2005. Microhabitats and population densities of California legless lizards, with comments on effectiveness of various techniques for estimating numbers of fossorial reptiles. Journal of Herpetology 39:395‒402. Lovich, R. 2012. Techniques for reptiles in difficult-tosample habitats. Pp. 167–196 In Reptile Biodiversity: Standard Methods for Inventory and Monitoring. McDiarmid, R.W., M.S. Foster, C. Guyer, J.W. Gibbons, and N. Chernoff (Eds.). The University of California Press, Berkeley, California, USA. Macey, J.R., J.A. Schulte II, A. Larson, B.S. Tuniyev, N. Orlov, N., and T.J. Papenfuss. 1999. Molecular phylogenetics, tRNA evolution, and historical biogeography in anguid lizards and related taxonomic families. Molecular Phylogenetics and Evolution 12:250‒272. McKenny, H.C.,W.S. Keeton, and T.M. Donovan. 2006. Effects of structural complexity enhancement on Eastern Red-Backed Salamander (Plethodon cinereus) populations in northern hardwood forests. Forest Ecology and Management 230:186‒196. Meek, R. 1986. Field body temperature of the Glass Lizard Ophisaurus apodus in Yugoslavia. AmphibiaReptilia 7:43‒49. Melville, J., I.I. Schulte, and A. James. 2001. Correlates of active body temperatures and microhabitat occupation in nine species of central Australian agamid lizards. Austral Ecology 26:660‒669. Mhirit, O., and P. Blerot. 1999. Le Grand Livre de la Forêt Marocaine. Mardaga, Sprimont, Belgique. Monger, H.C., and B.T. Bestelmeyer. 2006. The soilgeomorphic template and biotic change in arid and semi-arid ecosystems. Journal of Arid Environments 65:207‒218. Nafaa, R., and A. Watfeh. 2000. Holocene and actual degradation of the environment in the Mamora forest (Morocco). International Journal of Anthropology 15:263‒270. Phillips, S.J., M. Dudik, and R.E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190:231‒259. Purdue Research Foundation. 2014. MultiSpec vs. 3.4. Purdue Research Foundation, West Lafayette, Indiana, USA. Recuero, E., A. Iraola, X. Rubio, A. Machordom, and M. García-París. 2007. Mitochondrial differentiation and biogeography of Hyla meridionalis (Anura:Hylidae): 617 Herpetological Conservation and Biology an unusual phylogeographical pattern. Journal of Biogeography 34:1207‒1219. Rhoujjati, A., R. Cheddadi, M. Taïeb, A. Baali, and E. Ortu. 2010. Environmental changes over the past 29,000 years in the Middle Atlas (Morocco): a record from Lake Ifrah. Journal of Arid Environments 74:737–745. Rifai, L., M.A. Baker, D. Al Shafei, A. Disi, A. Mahasneh, and Z. Amr. 2005. Pseudopus apodus (Pallas, 1775) from Jordan, with notes on its ecology. Herpetozoa 18:133‒140. Riley, S.J., S.D. DeGloria, and R. Elliot. 1999. A terrain ruggedness index that quantifies topographic heterogeneity. Intermountain Journal of Sciences 5:23‒27. Rodríguez-Sánchez, F., and J. Arroyo. 2008. Reconstructing the demise of Tethyan plants: climate‐driven range dynamics of Laurus since the Pliocene. Global Ecology and Biogeography 17:685‒695. Schleich, H.H., W. Kästle, and K. Kabisch. 1996. Amphibians and Reptiles of North Africa. Koeltz Scientific Books, Koenigstein, Germany. Schonberg, L.A., J.T. Longino, N.M. Nadkarni, S.P. Yanoviak, and J.C. Gering. 2004. Arboreal ant species richness in primary forest, secondary forest, and pasture habitats of a tropical montane landscape. Biotropica 36:402‒409. Sindaco, R., and V.K. Jeremĉenko. 2008. The Reptiles of the Western Palearctic. Belvedere, Latina, Italy. Stebbins, R.C. 1985. A Field Guide to Western Reptiles and Amphibians. Houghton Mifflin Company, Boston, Massachusetts, USA. Tuanmu, M.N., and W. Jetz. 2014. A global 1‐km consensus land‐cover product for biodiversity and ecosystem modelling. Global Ecology and Biogeography 23:1031–1045. UNEP (United Nations Environment Programme). 1997. World Atlas of Desertification 2nd Edition. UNEP, London, UK. Valakos, E.D., P. Pafilis, K. Sotiropoulos, P. Lymberakis, P. Maragou, and J. Foufopoulos. 2008. The Amphibians and Reptiles of Greece. Chimaira, Frankfurt am Main, Germany. Vitt, L.J., P.A. Zani, and A. Lima. 1997. Heliotherms in tropical rain forest: the ecology of Kentropyx calcarata (Teiidae) and Mabuya nigropunctata (Scincidae) in the Curuá-Una of Brazil. Journal of Tropical Ecology 13:199‒220. DANIEL ESCORIZA (right) is a Researcher from the University of Girona, Spain. His research focuses on the ecology of invasive frogs and Mediterranean salamanders. He has a Ph.D. in amphibian ecology (2015) from Girona University (Spain). MAR COMAS (left) is a Researcher from the Estación Biológica de Doñana EBD-CSIC (Spain). She has studied montane amphibian and reptiles from Morocco and Spain. Currently she is working on the ecology of frogs 618 Escoriza and Comas.—Habitat of Hyalosaurus koellikeri. and lizards along an altitudinal gradient in Sierra Nevada (Spain). (Photographed by Daniel Escoriza). APPENDIX 1. Comparison between the localities of presence/absence of Hyalosaurus koellikeri for several habitat dimensions: vegetation cover, substrate characteristics, climate, topography, and forest structure. Tree crown cover at 250 m Soil bulk density Soil cation exchange Soil organic Soil pH Sand% Silt% Clay% Aridity index Mean annual temperature Terrain ruggedness Tree ΣBA/100 Tree mean height Tree density Tree/shrubNtaxa Tree/shrub N Tree/shrub Dominance Tree/shrub Shannon-Wiener index Tree/shrub variation in Shannon-Wiener index mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range mean range 619 Presence 66 33‒99 1.26 1.0‒1.39 8.32 2.0‒20.0 11.79 6.0‒23.0 5.41 3.7‒6.2 63.89 47‒82 14.16 4‒30 20.58 9‒35 0.47 0.24‒0.72 14.0 10.0‒18.3 240.5 31.1‒651.1 0.13 0.01‒0.53 7.46 2.31‒13.53 0.04 0.01‒0.06 4.6 2.0‒15.0 68.2 21‒155 0.49 0.14‒0.88 0.96 0.23‒2.27 0.45 0.001‒0.85 Absence 24 0‒93 1.39 0.89‒1.70 4.58 0‒10.0 7.11 5.0‒11‒0 5.82 4.19‒8.10 76.53 59‒85 10.58 4‒23 12.26 3‒34 0.52 0.27‒0.94 17.0 11.3‒19.2 148.9 4.0‒510.9 0.08 0.0‒0.95 2.83 0.0‒15.84 0.01 0.0‒0.09 3.4 1.0‒14.0 33.3 1‒131 0.66 0.17‒1.0 0.64 0.0‒2.05 0.72 0.04‒1.0 Herpetological Conservation and Biology APPENDIX 2. Principal coordinates ordination plot, showing the association between vegetation cover and occurrence of several lizard genera in the region. 620