SALAMANDRA 53(3) 405–412 15 AugusHt a2b0i1t7at useI SbSyN A l0g0y3ro6i–d3e3s 7m5archi Habitat use by peripheral populations of a lizard with a highly restricted distribution range, the Spanish algyroides, Algyroides marchi José Luis Rubio & Víctor Martín Departamento de Ecología, Universidad Autónoma de Madrid, Cantoblanco, 29049 Madrid, Spain Corresponding author: José Luis Rubio, e-mail: [email protected] Manuscript received: 15 August 2016 Accepted: 27 December 2016 by Stefan Lötters Abstract. Peripheral populations are considered vulnerable but important for conservation. The Spanish algyroides is a small and endangered lacertid lizard, endemic to a small area in the southeastern mountains of the Iberian Peninsula. It is a stenotopic species that will typically inhabit shaded and humid microhabitats in enclosed rocky situations. These habi- tat preferences relate to a very low thermal inertia and high evaporative water loss consistent with its small body size. Its patchy distribution is delimited by dryer and warmer lowlands. Newly detected localities seemed to expand its known dis- tribution range, providing a contact zone with core populations. We studied the habitat selection and conservation status of populations living in the border zone, produced a species distribution model of the new area, and compared border ver- sus core structural and environmental variables. The results confirmed the predicted occurrence, showing no differences in border vs. core selected habitat characteristics. From the perspective of habitat selection alone, edge effects and local adaptation seem insignificant. This could be related to the ‘hard-type edge’ and ‘two-patch system’ of the small distribu- tion of this species. Detected alterations of the habitat in this border area were mainly road and forest track construction, urbanisation, and livestock grazing. In the longer term, aridification of the area due to the global climate change could po- tentially gain importance, considering the dependency of the species on humid habitats, and the proximity of the edge of its distribution to the uninhabitable lowlands that determine its range limits. It is of great importance to identify the shape of the whole edge of this species’ range, effect in-depth evaluations, and monitor its conservation status. Key words. Edge populations, habitat selection, reptile conservation, species distribution model, Spain. Introduction The Spanish algyroides, Algyroides marchi Valverde, 1958, a small lacertid lizard, is an Iberian endemic with a Peripheral populations are considered of great impor- very restricted distribution. It is restricted to the south- tance for species conservation (e.g., Channel 2004) since eastern mountains of the Iberian Peninsula, the Pre- they would be more vulnerable to environmental fluctua- betic System (sensu Sánchez 1982; in general terms: Al- tions compared with those in the core areas. Populations at caráz, Cazorla and Segura ranges; Sánchez-Videgaín & the border of their species’ distribution would be typical- Rubio 1996, Rubio 2002, Carretero et al. 2010). These ly smaller, more isolated, less genetically diverse, and more occurrences amount to one of the smallest distribu- susceptible to inbreeding and genetic drift. Thus, they would tion ranges among continental European lacertid lizards be less resilient in the face of environmental changes (e.g., (< 5,000 km²). Within this area, suitable habitat seems to Lesica & Allendorf 1995, Böhme et al. 2006, Sexton be patchy, and some declines in both the quality of hab- et al. 2009, Bill & Gómez 2011, Cuervo & Moller 2013, itat and number of populations have been reported (Ru- Antunes et al. 2006, Volis et al. 2016, but see Gonza lez- bio 2002, Rubio et al. 2006). All this makes the species Guzman 2001, Channel 2004 for a discussion). On the very vulnerable; it is ranked as ‘Endangered’ by the IUCN other hand, populations adapted to peripheral environmen- (IUCN, Pérez-Mellado et al. 2008). tal conditions have been regarded as potential reservoirs of Algyroides marchi is a remarkably stenotopic species. It adaptive genetic variation (e.g., Blows & Hoffman 1993, typically occupies (geomorphologically) enclosed, rocky, Lessica & Allendorf 1995, Eckstein et al. 2006, Willi et shaded and humid localities (Rubio & Carrascal 1994). al. 2007, Ledoux et al. 2014), and local adaptation to condi- These characteristics would reflect a high sensitivity of the tions at the distribution limits has been reported and dis- species to high humidity and low temperature (Rubio & cussed in detail (e.g., Kawecki 2008, Bell & Gómez 2011). Carrascal 1994, García-Muñoz & Carretero 2013, © 2017 Deutsche Gesellschaft für Herpetologie und Terrarienkunde e.V. (DGHT), Mannheim, Germany Available at http://www.salamandra-journal.com 405 José Luis Rubio & Víctor Martín J. L. Rubio unpubl. data). This habitat preference has been The general configuration of the mountains, mainly related to the small size of the lizard (snout–vent length SW-NE, and the general altitude (700–2,100 m a.s.l.) ham- < 50 mm) with its inherent high surface/volume ratio, high per the crossing of precipitation-bearing clouds from the heating/cooling rate, and high evaporative rate (Rubio & west, producing a strong gradient of aridity eastwards (to- Carrascal 1994). The mountain massif occupied by the wards the Murcia Region; Vidal-Abarca et al. 1987). The species would constitute a form of continental island, sur- study area contains a variety of lithological materials such rounded by lower warmer and drier territories that limit as limestone, dolomite and siliceous sands. The resultant the species distribution (Rubio 2002). carbonaceous character has produced a rugged landscape Recent surveys (Brakels et al. 2010, Carretero et al. with narrow valleys with a high degree of enclosure that 2010) revealed new localities of occurrence of the species provide shaded and humid microhabitats, often within (Moratalla Range, Murcia Region, and Taibilla and Zume- a more arid landscape matrix. The vegetation is mainly ta Ranges, Albacete Province) at the previously known composed of open forest formations of pines (Pinus ni­ southeastern limits. This new area is a generally arid region gra), juniper trees (Juniperus thurifera), oaks (Quercus ilex, where the populations appear to be rather isolated. The Q. pyrenaica, Q. coccifera), and Mediterranean shrubs and new data interestingly suggested some expansion of the pasturelands with poplar and walnut trees (Juglans regia) previously known general distribution area, and could help on the riverbanks (Alcaráz & Sanchez-Gómez 1988, to identify the shape of the peripheral limits in that sector. Gómez et al. 1992). They also opened the possibility that an important contact zone may exist between the core and the southeastern pe- ripheral populations. The vulnerability of A. marchi due to Study design its stenoicity and its reduced area of distribution would be greater in these border populations as they might be more We surveyed the study area during the activity period of exposed to external stress factors. The study of mitochon- the lizards for a total of 20 days, and registered with GPS drial markers (Carretero et al. 2010) revealed differences the exact UTM coordinates of the localities occupied by in these A. marchi edge populations, suggesting the exist- the species. We generated a distribution map based on 18 ence of old (Pleistocene) fragmentation/retraction process- occurrence localities over a 30-m resolution Digital Eleva- es in this area, although nuclear markers indicated a genet- tion Model (ASTER GDEM; NASA government; asterweb. ic flow between core and edge populations. Very prelimi- jpl.nasa.gov). For the study of habitat selection, we estab- nary estimates of abundance within border populations of lished sample plots of 25 × 25 m in 13 localities; one plot per A. marchi suggest that they could be less abundant com- locality. We located the centre of our sample plots at those pared to their core counterparts (Carretero et al. 2010). sites with the highest observed abundance of Algyroides In this paper, (1) we studied the distribution of marginal marchi individuals. Thirteen variables were measured in populations of A. marchi at its southeastern limits (Taibilla each sample plot: altitude, cardinal aspect, geomorpho- and Zumeta headwaters, Albacete province), (2) we carried logical enclosure, availability of water, percentage of area out a habitat selection research within the peripheral area occupied by cliffs, rocks (diameter > 70 cm) and stones and tested possible differences in habitat selection patterns (diameter < 70 cm), as well as percentage cover of trees, at a regional scale between core and peripheral popula- bushes, forbs, grams, litter and bare ground. We measured tions, (3) we then produced a distribution model to predict the altitude and the cardinal aspects by GPS-altimeter and the localities that were most favourable for its presence in compass respectively. The degree of enclosure was scored the study area, (4) we took into account only habitat prefer- as (1) flat terrain, (2) moderately sloped, (3) open gully and ences and preliminarily discuss our results for the species (4) closed gully. We calculated an index of water volume in the context of edge effects, and finally (5), we consider by multiplying length, average width and average depth of the main threats for the species in this area. any water body within the plot. We estimated covers visu- ally as percentages (after previous training). Additionally, we evaluated the degree of anthropogenic alteration in eve- Materials and methods ry location, ranking it by consensus among the authors in Study area four categories: (1) high, (2) medium, (3) low and (4) very low to nil. We systematically assessed the physical struc- The study of peripheral populations of Algyroides marchi ture of the landscape at 22 sites located at the intersections was conducted in July and September of 2012 for ten days of the UTM coordinates of 5 × 5 km. We measured the each in the southeastern periphery of its distribution, i.e., same variables using the same methods described above. at the headwaters of the rivers Taibilla and Zumeta (Al- By means of a Mann-Whitney U test (α ≤ 0.05), we com- bacete Province, Spain; 38°07’ N, 2°20’ W; extension about pared the variables of the occurrence localities versus the 700 km²). This region is characterized by strong climatic systematically selected localities in the study area (headwa- contrasts (Sánchez 1982, Artiago 1984), with high daily ters of rivers Taibilla and Zumeta). We also compared the and annual variation of temperatures (annual range: 10– peripheral occurrence localities, as well as the systemati- 16°C), with dry and hot summers and rainy and freezing cally selected localities, with those in the core area (Alcaraz winters (mean annual precipitation: 350–650 mm). Range, Albacete Province; data available from a previous 406 Habitat use by Algyroides marchi study; Rubio & Carrascal 1994). Data analyses were per- model, the Area Under the receiver-operating characteris- formed with IBM SPSS Statistics 21. tic Curve (AUC) was computed using MAXENT’s internal To predict the most favourable localities of presence of validation procedure (Phillips & Dudík 2008). Jackknife the species at a regional scale, we generated a Species Dis- test and response curves were selected in a way to meas- tribution Model (SDM) (Elith & Leathwick 2009). The ure the importance and contribution of the variables to model was produced from occurrence records of A. marchi the models. After running MAXENT and obtaining prob- available from previous studies (Sanchez-Videgaín & ability distributions (from 0 to 1), the output distributions Rubio 1996, Carretero et al. 2010), and a set of envi- were reclassified in categories according to their degrees ronmental variables. These variables were obtained from of habitat suitability: low values (< 0.33), medium values the Worldclim dataset at a spatial resolution of 1 km2, and- (0.34–0.66), and high values (< 0.66). comprised annual precipitation, maximum temperature of the warmest month, and altitude, and another four topo- graphic variables derived from the Digital Elevation Model Results (DEM, 1 km² resolution; WorldClim dataset, http://www. Distribution, evaluation and habitat selection wordlclim.org): slope (%), curvature, aspect and insola- tion received in 2012. All transformations of variables were Our survey in the border area produced 18 new localities computed with ArcMap 9.3. of occurrence of Algyroides marchi (Fig. 1). These data al- Species distribution was modelled using the maxi- lowed us to draw a map of available potentially suitable mum-entropy machine learning SDM (Species Distribu- habitat in a potential contact zone. In general, the localities tion Model) algorithm MAXENT v. 3.3.3k (Phillips et al. were found mainly in relatively enclosed and humid gul- 2006), which is applicable to presence-only data and con- lies. The evaluation of anthropogenic alteration of the oc- sidered one of the most accurate SDM algorithms (Elith currence localities showed that 33% were exposed to very et al. 2011). We used the default settings of MAXENT with low or no impact, whereas the remaining localities showed logistic output that would provide an estimated suitabil- signs of greater impact, mainly due to the construction of ity between 0 and 1 (interpreted as the probability of pres- roads and tracks (60%), followed by urban constructions ence). In order to evaluate the predictive ability of the (22%) and livestock grazing (16%). Figure 1. New occurrences of Algyroides marchi (white dots) and systematically selected localities (dark grey dots) at the southeastern border of the distribution (headwaters of the rivers Taibilla and Zumeta, Albacete, Spain). Urban cores in black squares. Insert in the top left corner: the study area on the Iberian Peninsula. Map based on 30-m resolution; ASTER Global Digital Elevation ranging from 429 to 2,095 m a.s.l. 407 José Luis Rubio & Víctor Martín Table 1. Comparison of the structural characteristics (mean and standard deviation, abbreviated SD) of systematically selected lo- calities with those of occurrence of Algyroides marchi in the border populations at the headwaters of the rivers Taibilla and Zumeta (Abacete, Spain). Localities Localities Mann-Whitney with A. marchi systematically selected U test mean SD mean SD p Altitude (m) 1282.73 212.48 1366 193.25 0.191 Cliff cover (%) 29.91 21.10 0.00 0.00 0.000 Rocks cover (%) Rocks, diameter > 70 cm 29.73 23.64 10.46 25.64 0.002 Stones, diameter < 70 cm 9.00 10.11 7.77 14.96 0.257 Canopy cover (%) 17.91 14.44 23.47 22.00 0.601 Bush cover (%) 20.36 9.97 28.73 24.70 0.578 Forbs cover (%) 5.73 7.63 7.18 7.18 0.801 Grams cover (%) 7.90 14.56 20.73 26.09 0.087 Trash cover (%) 2.45 5.91 3.41 7.08 0.631 Bare ground (%) 0.34 0.65 6.27 8.08 0.001 Water volume index (%) 1.41 3.37 0.31 1.44 0.049 Degree of enclosure 3.45 0.52 1.82 0.85 0.000 Sample size 13 22 Table 1 presents the values of the variables in the locali- sign above), showed the main areas of habitat suitable for ties of occurrence of the species, and those in the locali- Algyroides marchi in the new border zone (Fig. 2A). The ties selected systematically at a regional scale in the new prediction for the border study area (Fig. 2B) shows con- border area (Taibilla and Zumeta headwaters). The Mann- sistency of the species’ occurrence zones with the scores Whitney U test results show that degree of enclosure, water predicted by the model as suitable. Sixteen localities match volume index, and percentage of surface occupied by rocks with the suitable habitats predicted for the species. Only and cliffs were all significantly elevated in the occurrence two localities offered habitats that were less adequate than localities of A. marchi. Bare ground was less common in predicted by the model in that they had lower suitabil- the occurrence localities than in those systematically se- ity scores. Additionally, our model predicts other zones lected. The cardinal aspects of the occurrence localities (Fig. 2) that could also be conducive to the presence of the tended towards the northern aspects (NW-N-NE, 62%). species (e.g., gullies in the south of the Taibilla Range and The remaining variables showed not significant differences. zones to the north of the Taibilla and Zumeta headwaters). Comparison of the structural variables of the occur- The applicability of our model was evaluated through rence localities in the core area with those in the border the AUC score (Area Under the Curve). We obtained an area showed a significantly higher degree of enclosure in AUC of 0.991, indicating a very precise prediction (Swets the occurrence localities in the border zone (mean = 3.54 1998). The most influential variables in the model were alti- vs. 2.16; U = 206.5; P = 0.001). Grass and bare ground were tude (39.4%) and the maximum temperature in the month less common in the occurrence localities in the border area of July (34.3%). The remaining variables contributed to a (U = 61.5, P = 0.032 and U = 18.0, P < 0.001, respectively). lesser extent to the model (slope = 10.4; curvature = 8.1; The remaining studied variables did not show any signif- annual precipitation = 6.6), with insolation and cardinal icant differences. On the other hand, our comparison of aspect contributing the least (0.8 and 0.4, respectively). In the variables that describe habitat availability (systemati- fact, altitude and maximum temperature are also the var- cally selected localities) in the core area vs. the border area, iables that better explain the model when they are used show a lower coverage of rocks in the border zone (U = 63, alone, but it is the maximum temperature in July that will P < 0.001). The available altitudes are higher in the border most affect the model when it is removed from it (Fig. 3). area (U = 295, P = < 0.001), and cliffs were absent in our sample of systematically selected localities (mean surface in core area = 10%). Discussion Considering the small expanse of the species’ distribution, Species Distribution Model the new 18 occurrence localities of Algyroides marchi found in the study area constitute a noteworthy expansion of its The SDM, obtained from occurrence data in the general known southeastern peripheral distribution. These data distribution area with a resolution of 1 km² (see Study de- increase our knowledge of the potential connectivity be- 408 Habitat use by Algyroides marchi Figure 2. Species Distribution Model (SDM) of Algyroides marchi with a resolution of 1 km². Dark grey areas represent suitability values > 0.66, medium grey areas suitability values 0.33–0.66, and light grey areas suitability values < 0.33. (A) Representation of the total distribution range as predicted by the model. The occurrences used to develop the model are presented in black. (B) Extension of the model for the study area. White dots represent main villages, and black dots locations where the species was found. tween populations, help to identify the shape of the species’ wards the southeastern distribution border (Vidal-Abarca distribution limits, and are important for future conserva- et al. 1987). The localities with very high degrees of enclosure tion plans (e.g., Burkey & Reed 2006). would provide enough humidity and suitable temperatures The results of this study confirm the predictions for the to allow the species to persist in this zone. Localities with the habitat selection patterns at a regional scale from previ- structural characteristics preferred by the species are scarcer ous studies (Rubio & Carrascal 1994, Carretero et al. in the border zone than in the more core-orientated range, 2010). In the study area, A. marchi appears mainly in locali- where A. marchi furthermore appears to use a wider range ties with rocky limestone gullies that provide water and a of habitats that provide the shaded and humid conditions high degree of enclosure, and, hence, low exposure to solar preferred by the species (crags, dolines, lapiés [lapiaz], large radiation. Such characteristics are frequently associated in screes etc.; Sanchez-Videgaín & Rubio 1996). the area with rocky riverbeds. The limestone materials that characterize the area facilitate a way of landscape modula- tion that will create numerous small caves, holes and crev- ices usable to the lizards; rocky screes with associated hu- midity and ambient moisture underneath are also found. Our comparison of localities of A. marchi occurrence in the border study area with the localities occupied by the spe- cies in the core area shows no variation in habitat use, indi- cating that habitat conditions in the localities occupied by A. marchi do not vary significantly throughout its distribu- tion. Only the degree of enclosure was significantly high- er in the peripheral study area than in the core area. The relatively arid and flattened matrix between suitable locali- Figure 3. Contribution of every variable when used alone to cal- culate the model (dark grey), and when removed from it and ties that characterize the landscape in the border range at keeping the remaining variables to calculate the model with a the Taibilla and Zumeta headwaters could explain the high resolution of 1 km² (light grey). The last bar represents the full degree of enclosure found in the localities of occurrence of model. Key to variables: (A) altitude in m above sea level; (B) cur- A. marchi in this peripheral zone. The Foehn effect gener- vature index of the surface; (C) cardinal direction; (D) slope; ated by the Prebetic Mountains blocking the westerly winds (E) mean annual precipitation; (F) insolation; (G) maximum (see area of study) create arid environmental conditions to- temperature in July. 409 José Luis Rubio & Víctor Martín The species’ preference for a habitat with such charac- 2014). Diaz et al. (2006), for example, reported stable dif- teristics confirms in this area the importance of temper- ferences in life history traits in Psammodromus algirus, a ature and humidity in its distribution proposed in previ- large lacertid lizard with a wide distribution in the Iberi- ous studies (Rubio & Carrascal 1994, Carretero et an Peninsula (e.g., two vs. single clutches) between edge al. 2010, García-Muñoz et al. 2010). Small lizards like and core populations (see also Sinervo 1990). Although A. marchi have a fast rate of heat exchange with the en- further study is needed on the A. marchi periphery, our vironment (e.g., Stevenson 1985, Carrascal et al. 1992, data show a ‘hard-type edge’ sensu Stamps et al. (1987) or, Garrick 2008) and a high relative evaporative water loss as classified by Kawecki (2008), a ‘two-patch system’ (a (Mautz 1982) due to their high surface-to-volume ratio. little-studied model). The aridity of the surrounding land The short time available for diel and annual activity, de- would limit this species’ distribution, and has been related rived from the environmental conditions in enclosed lo- to extinction-recolonisation processes in the past, as ap- calities, may be of great importance in relation to the small parently is reflected in the genetic differences reported for size of the species, as well as for its reproductive strategy: the extant border populations (Carretero et al. 2010). typically a small cultch size in a single oviposition period This last work also demonstrated, from nuclear markers, (Rubio & Palacios 1986). Its selection of large rocks pro- the existence of gene flow between edge and core popula- vides opportunities for thermo- and water regulation (e.g., tions, a factor able to counteract local selection (Barton Huey et al. 1989, Shah et al. 2004). The cool, humid and & Partridge 2000, Lenormand 2002, Kawecki 2008). moist conditions under these rocks furthermore provide a More specific studies about peripheral versus core popu- good source of arthropod prey (mainly small spiders and lations are needed to investigate edge effects in A. marchi: flies; J. L. Rubio unpubl. data). Although a forest-dwelling on demography, life history, body condition, and biometry character has been attributed to the species (Arnold 2002, (including morphology and asymmetry), as are accurate Carretero et al. 2010), the results of our habitat selection studies of populations genetics. studies portray A. marchi more as a species of shaded rocky In the light of our habitat selection study and species sites where water will be available, not necessarily one of distribution model, the high predictability of occurrenc- forest situation. High tree cover, and the associated humid es of A. marchi facilitates the development of management and moist microclimate might help in the dispersal of this plans for the conservation of this species. The immediate species, however. risk factors for A. marchi populations in the study area The model prediction for the study area sees occurrence are related to direct alterations of their habitats: mainly localities and the highest values of suitability coincide, and the construction of roads and forest tracks, urban devel- thus provides a map relevant to conservation planning. opment, and livestock grazing. The alteration of the veg- The model provides an image concordant with the habi- etation could also affect suitably shaded microclimates. tat selection study, and is coherent with the SDM in the The effect of livestock grazing, poorly known in lizards general distribution (Carretero et al. 2010). It signalled (Blevins & With 2011, but see Behroo et al. 2015 for vi- the importance, at various geographical scales, of a tall and pers), may have some impact on this species, considering rough relief in the maintenance of high humidity, through its potential sensitivity to environmental changes. How- attracting precipitation, and reduced temperatures that ever, many occurrence localities are relatively inaccessible would minimize the summer thermal stress. to livestock. In addition, recent observations of individu- Algyroides marchi exhibits little capacity to colonize als, mainly juveniles, in pastures (Carretero et al. 2010, suboptimal sites (although juveniles might use them for J.L. Rubio unpubl. data) suggest that this vegetation might dispersal; see below). The results of the SDM obtained play a role in their dispersal, and grazing will not be a ma- from the data of the general distribution together with jor problem as long as overexploitation would not lead the results of the study of habitat selection indicate that to bare ground; however, further research on this issue is A. marchi will select the same habitats, at a regional scale, needed. Although A. marchi has sometimes been observed throughout its entire distribution range, including the pe- on the walls of old isolated houses (Carretero et al. 2010, riphery. Edge effects and local adaptations are thus unap- J.L. Rubio unpubl. data), the process of construction itself parent in these populations, unlike in other reported in- would obviously drastically alter occurrence localities, and stances (e.g., Díaz et al. 2006, Vergeer & Kunin 2013, but synanthropic species, such as domestic cats and rodents, see, e.g., Bjorndal & Balten 2010, Peñalver et al. 2010). are surely threats. The risk of forest fire is always present We hypothesize that this might be linked to the reduced in this area. Other threats reported from other areas in the and sharply edged expanse of the distribution range of distribution range of A. marchi have not been found here, A. marchi, which is restricted to the Prebetic Mountains. In but should nevertheless be considered for future manage- this regard, Holt et al. (2005), following Carter & Prince ment measures in this border zone. The list includes stream (1981), indicated the possibility that individuals would ex- canalising and/or silting by erosion after logging, replace- perience the same environmental conditions in core and ment of traditional stone ponds (usable by the lizards) with suitable patches in edge habitats. Larger distribution rang- metal or plastic tanks, small dams, and tourism facilities es may present greater spatial variation and environmen- (Rubio & Carrascal 1994, Sánchez-Videgaín & Rubio tal gradients that thus can promote local adaptation (sensu 1996, Carretero et al. 2010). The study area is partly locat- Fortin et al. 2005; Kawecki 2008, but see Ledoux et al. ed within protected natural spaces (SPA of Alcaráz, Mundo 410 Habitat use by Algyroides marchi and Segura Ranges), but, considering the afore-mentioned Blows, M. W. & A. A. Hoffman (1993): The genetics of central local impact factors, any management plan should focus and marginal populations of Drosophila serrata. I. Genetic vari- on not altering humidity, moisture and temperature of the ation for stress resistance and species borders. – Evolution, 47: 1255–1270. shaded rocky sites characteristic of the species’ occurrence. In a longer term, a potential aridification of the area due Böhme, M. U., N. Schneeweiss, U. Fritz, M. Schlegel & T. U. Berendonk (2006): Small edge populations at risk: genetic di- to global climate change (Summer et al. 2003) could gain versity of the green lizard (Lacerta viridis viridis) in Germany importance considering the dependency of the species on and implications for conservation management. – Conserva- humid habitats, and the proximity of the edge of its distri- tion Genetics, 8: 555–563. bution to unsuitable lower, more flattened, and more arid Brakels P., F. Koopmans & J. L. Rubio (2010): First record of the areas that determine its range limit. Studies with climatic Spanish algyroides, Algyroides marchi (Lacertidae) in the Mur- change sceneries will allow evaluating local perturbations cia region, Southeast Spain. – Herpetology Notes, 3: 49–52. of the habitat. Although SDMs have limitations (possible Burkey, T. V. & D. H. Reed (2006): The effects of habitat fragmen- unaccounted biological factors, change of climatic varia- tation on extinction risk: Mechanisms and synthesis. – Song- bles with time, or other methodological factors; Thuiller klanakarin Journal of Science and Technology, 28: 9–37. et al. 2004, Sinclair et al. 2010, Sanchez-Fernandez Carter, R. N. & S. D. Prince (1981): Epidemic models used to ex- et al. 2011), this tool surely will speed up finding new plain biogeographical distribution limits. – Nature, 293: 644–645. A. marchi localities in the future (cf. Guisan & Zimmer- Channell, R. (2004): The conservation value of peripheral popu- mann 2000). As far as border populations are concerned, lations: the supporting science. – Proceedings of the Species at it will be of great interest to identify the shape of the whole Risk 2004 Pathways to Recovery Conference, Victoria, Canada. distribution edge of this species, evaluate it in depth, and Carrascal, L. M., P. López, J. Martín & A. Salvador (1992): monitor its conservation status; this will enable us to de- Basking and antipredator behaviour in a high altitude lizard: tect and anticipate anthropogenically caused fluctuations implications of heat-exchange rate. – Ethology, 92: 113–54. so that preventive measures can be taken. Carretero, M. A., F. Ceacero, E. García-Muñoz, N. Sillero, M. I. Olmedo, P. L. Hernández-Sastre & J. L. Rubio (2010): Seguimiento de Algyroides marchi. Informe final. Monografías Acknowledgements SARE. Asociación Herpetológica Española – Ministerio de Me- dio Ambiente y Medio Rural y Marino, Madrid. We thank M. Luzón and C. Díez for their help with fieldwork, and Cuervo, J. J. & A. P. Møller (2013): Temporal variation in popula- C. Mata and J. Traba for their comments on the SDM. tion size of European bird species: Effects of latitude and mar- ginality of distribution. – PLoS ONE, 8: e77654. Díaz, J. A., J. Pérez-Tris, D. Bawens, D. Pérez-Aranda, R. Car- References bonell, T. Santos & J. L. Tellería (2006): Reproductive per- formance of a lacertid lizard at the core and the periphery of Alcaráz, F. & P. Sánchez-Gómez (1988): El paisaje vegetal de la the species’ range. – Biological Journal of the Linnean Society, provincia de Albacete. – Al-Basit, 24: 94–4. 92: 87–96. Antunes, A., R. Faria, W. L. Johnson, R. Guyomard & P. Alex- Eckstein, R. L., R. A. O’Neill, J. Danihelka, A. Otte & W. Köh- andrino (2006): Life on the edge: The long-term persistence and ler (2006): Genetic structure among and within peripheral contrasting spatial genetic structure of distinct brown trout life his- and central populations of three endangered floodplain violets. tories at their ecological limits. – Journal of Heredity, 97: 193–205. – Molecular Ecology, 15: 2367–2379. Arnold, E. N. (2002): Field guide to the reptiles and amphibians of Elith, J. & J. R. Leathwick (2009): Species distribution models: Britain and Europe. – Collins, London. Ecological explanation and prediction across space and time. Artiago, A. (1984): El clima de la provincial de Albacete. – E.U.O, – Annual Review of Ecology, Evolution, and Systematics, 40: Albacete, Spain. 677–697. Barton, N. & L. Partridge (2000): Limits to natural selection. – Elith, J., S. J. Phillips, T. Hastie, M. Dudík, Y. E. Chee & C. J. BioEssays, 22: 1075–1084. Yates (2011): A statistical explanation of Maxent for ecologists. – Diversity and Distributions, 17: 43–57. Bell, G. & A. Gómez (2011): Adaptation and evolutionary rescue in metapopulations experiencing environmental deterioration. Fortin, M. J., T.H. Keitt, B. A. Maurer, M. L. Taper, M. K. – Science, 332: 1327–1330. Dawn & T. M. Blackburn (2005): Species’ geographic ranges and distributional limits: pattern analysis and statistical issues. Behrooz, R., M. Kaboli, E. Nouran, M. Ahmadi, A. A. Shaba- – Oikos, 108: 7–17. ni, M. Yousefi, A. Asadi & M. Rajabizadeh (2015): Habitat modeling and conservation of the endemic Latifi’s viper (Monti­ García-Muñoz, E. & M. A. Carretero (2013): Comparative eco- vipera latifii) in Lar National Park, northern Iran. – Herpeto- physiology of two sympatric lizards. Laying the groundwork logical Conservation and Biology, 10: 572–582. for mechanistic distribution models. – Acta Herpetologica, 8: 123–128. Bjorndal, K. A. & A. B. Balten (2010): Hawksbill sea turtles in seagrass pastures: Success in a peripheral habitat. – Marine Bio- Garrick, D. (2008): Body surface temperature and length in rela- logy, 157: 135–145. tion to the thermal biology of lizards. – Bioscience Horizons, 1: 136–142. Blevins, E. & K. A. With (2011): Landscape context matters: local habitat and landscape effects on the abundance and patch oc- Gonzalez-Guzman, L. I. & D. W. Mehlman (2001): Developmen- cupancy of collared lizards in managed grasslands. – Landscape tal stability across the breeding distribution of the scissor-tailed Ecology, 26: 837–850. Flycatcher (Tyrannus forficatus). – Ecology Letters, 4: 444–452. 411 José Luis Rubio & Víctor Martín Gómez, R., J. J. Presa & M. D. García (1992): Estudio faunístico Shah B., R. Shine, S. Hudson & M. Kearney (2004): Experimen- y ecológico de los Caelifera (Orthoptera, Insecta) de la Sierra tal analysis of retreat-site selection by thick-tailed geckos Ne­ del Taibilla (Albacete). – Universidad de Castilla-La Mancha, phrurus milii. – Austral Ecology, 29: 547–552 Murcia. Sánchez, J. (1982): Geografía de Albacete. Factores del desarrol- Guisan, A. & N. E. Zimmermann (2000): Predictive habitat distri- lo económico de la provincia y su evolución reciente. Tomo I. bution models in ecology. – Ecological Modelling, 135: 147–186. CSIC. Serie I. Ensayos Históricos y Científicos, nº 12. – Instituto de Estudios Albacetenses, Albacete, Spain. Holt, R. D., T. H. Keitt, M. A. Lewis, B. A. Maurer & M. L. Ta- per (2005): Theoretical models of species’ borders: single spe- Sánchez-Fernández D., J. M. Lobo & O. L. Hernández-Man- cies approaches. – Oikos, 108: 18–27. rique (2011). Species distribution models that do not incor- porate global data misrepresent potential distributions: A case Huey, R. B., C. R. Peterson, S. J. Arnold & W. P. Porter (1989): study using Iberian diving beetles. – Diversity and Distribu- Hot rocks and not-so-hot rocks. Retreat site selection by gar- tions, 17: 163–171. ter snakes and its thermal consequences. – Ecology, 70: 931–44. Sánchez-Videgaín, J. & J. L. Rubio (1996): Atlas provisional de IUCN (2010): European Species on the Red List. – Available at los anfibios y reptiles de las Sierras Prebéticas albacetenses. – http://www.iucnredlist.org/europe. Al-Basit, 38: 5–30. Kawecki, T. J. (2008): Adaptation to marginal habitats. – Annual Sexton, J. P., J. McInture, A. L. Angert & K. J. Rice (2009): Evo- Review of Ecology, Evolution, and Systematics, 39: 321–42. lution and ecology of species range limits. – Annual Review of Ledoux, J., D. Aurelle, N. Bensoussan, G. Marshchal, J. Fe- Ecology, Evolution, and Systematics, 40: 415–36. ral & J. Garrabou (2014): Potential for adaptive evolution at Sinclair, S. J., M. D. White & G. R. Newell (2010): How useful species range margins: constraining interactions between cor- are Species Distribution Models for managing biodiversity un- al populations and their environment in a changing Ocean. – der future climates? – Ecology and Society, 15: 8. Ecology and Evolution, 5: 1178–1192. Sinervo, B. (1990): The evolution of maternal investment in liz- Lenormand, T. (2002): Gene flow and the limits to natural selec- ards: an experimental and comparative analysis of egg size and tion. – Trends in Ecology & Evolution, 17: 183–189. its effects on offspring performance. – Evolution, 44: 279–294. Lesica, P., & F. W. Alendorf (1995): When are peripheral pop- Stamps, J. A., M. Buechner & V. V. Krishnan (1987): The effect ulations valuable for conservation? – Conservation Biology, 9: of edge permeability and habitat geometry on emigration from 753–760. patches of habitat. – The American Naturalist, 129: 533–552. Mautz, W. J. (1982): Patterns of evaporative water loss. – pp. 443– Stevenson, R. D. (1985): Body size and limits to the daily range 481 in: Gans, C. & C. H. Pough (eds): Biology of the Reptilia, of body temperature in terrestrial ectotherms. – The American vol. 12. – Academic Press, New York. Naturalist, 125: 102–117. Peñalver, M., P. Aragón, M. C. Breedveld & P. S. Fitze (2016): Summer, G. N., R. Romero, V. Homar, C. Ramis, S. Alonso & E. Microhabitat selection in the common lizard: implications of Zorita (2003): An estimate of the effects of climate change on biotic interactions, age, sex, local processes, and model trans- the rainfall of Mediterranean Spain by the late twenty first cen- ferability among populations. – Ecology and Evolution, 5: 1–14. tury. – Climate Dynamics, 20: 789–805. Pérez-Mellado V., R. Márquez & I. Martínez-Solano (2008): Swets, K. (1988): Measuring the accuracy of diagnostic systems. – Algyroides marchi. The IUCN Red List of Threatened Species. Science, 240: 1285–1293. Version 2009.2. – IUCN, Gland, Switzerland. Thuiller, W., L. Brotons, M. B. Araújo & S. Lavorel (2004): Phillips, S. J. & M. Dudík (2008): Modeling species distribution Effects of restricting environmental range of data to project cur- with Maxent: New extensions and a comprehensive evaluation. rent and future species distributions. – Ecography, 27: 165–172. – Ecography, 31: 161–175. Vergeer, P. & W. Kunin (2013): Adaptation at range margins: com- Phillips, S. J., R. P. Anderson & R. E. Schapire (2006): Maxi- mon garden trials and the performance of Arabidopsis iyrata mum entropy modeling of species geographic distributions. – across its northwestern European range. – New Phytologist, Ecological Modeling, 190: 231–259. 197: 98–1001. Pleguezuelos, J. M., R. Márquez & M. Lizana (2002): Atlas y Vidal-Abarca, M. R., C. Montes, L. L. Ramírez-Díaz& L. M. libro rojo de los anfibios y reptiles de España. – DGCN-AHE, Suárez (1987): El clima de la cuenca del rio segura (s.e. de Es- Madrid. paña): factores que lo controlan. – Anales de Biología, 12: 11–28. Rubio, J. L. & L. M. Carrascal (1994): Habitat selection and con- Willi, Y., J. van Buskirk, B. Schmid & M. Fisher (2007): Genetic servation of an endemic Spanish lizard Algyroides marchi (Rep- isolation of fragmented populations is exacerbated by drift and tilia: Lacertidae). – Biological Conservation, 70: 245–250. selection. – Journal of Evolutionary Biology, 20: 534–542. Rubio, J. L. (2002): Algyroides marchi Valverde, 1958. Lagartija de Volis, S., D. Ormanbekova, K. Yermekbayev, M. Song & I. Valverde. – pp. 193–195 in: Pleguezuelos, J. M., R. Márquez Shulgina (2016): The conservation value of peripheral popula- & M. Lizana (eds): Atlas y libro rojo de los anfibios y reptiles de tions and a relationship between quantitative trait and molecu- España. – DGCN-AHE, Madrid. lar variation. – Evolutionary Biology, 43: 26–36. Rubio, J. L. & F. P. Palacios (1986): The reproductive cycle of Algy­ roides marchi. – pp. 473–477 in: Roček, Z. (ed.): Studies in Her- petology. – Societas Europea Herpetologica, Prague. Rubio, J. L., J. M. Pleguezuelos, M. Moreno & S. Honrubia (2006): Parque natural de las sierras de Cazorla, Segura y Al- caraz. Áreas importantes para los anfibios y reptiles de España. – Ministerio de Medio Ambiente, Madrid, Spain. 412