Morphological and ecological variation of a tropical anoline lizard: are agonistic interactions shaping ecomorphological relationships?

Morphological and ecological variation of a tropical anoline lizard: are agonistic interactions... Abstract Ecologists have long been intrigued by which factors influence habitat use by an organism and how communities are structured. However, the links between habitat preferences, morphology, biotic interactions and community structure are still poorly understood. Moreover, interpopulation variation in ecomorphological relationships has usually been neglected. Here, we use a wide-ranging Anolis lizard, Anolis limifrons, to test whether interpopulation variation in morphology and habitat use is a function of interspecific agonistic interactions across the distribution of this species in Costa Rica. We found differences both in morphology and in habitat use among populations of A. limifrons, with populations from the Caribbean versant of Costa Rica having longer hind legs and perching lower than those from the Pacific versant. The intensity of interspecific agonistic interactions also varied across versants, with A. limifrons from Pacific sites displaying more often to congeners than those from the Caribbean. Agonistic interactions appear to be an important factor shaping habitat use and morphology. These findings can be explained by an interaction between phenotypic plasticity and ecological plasticity. Anolis limifrons, Costa Rica, ecomorphology, habitat use, mainland lizards, populations, structural habitat INTRODUCTION One cornerstone in ecological studies has been the understanding of the relationship between morphology and ecology (Wainwright & Reilly, 1994). Extensive research has been conducted across different taxa to predict ecological attributes (e.g. habitat use and dietary breadth) based on morphological traits (da Silva et al., 2014; Villalobos & Arita, 2014; Pereira Leitão et al., 2015). These ecomorphological studies rely on the assumption that a given phenotype will perform better in an environment with specific conditions than other phenotypes, affecting fitness and the direction of selection (Losos et al., 2000). Therefore, a tight link between habitat characteristics, morphology, ecology and behaviour has been proposed (Garland & Losos, 1994), influencing the habitat selection, interspecific interactions and trophic relationships of an organism (Strong et al., 1984). Most often, this link has been analysed in comparative studies among species (Ord & Klomp, 2014; Villalobos & Arita, 2014), although ecomorphological relationships are also present within a species (Irschick et al., 2005a, b; Chapman et al., 2015). Populations of the same species can occupy a wide range of habitats, where selective pressures could be very different (Kolbe, Larson & Losos, 2007; Lostrom et al., 2015). Therefore, high morphological variation is expected to occur among populations in traits that allow individuals to carry out relevant ecological tasks (Irschick et al., 2005a; Calderón-Espinosa, Ortega-León & Zamora-Abrego, 2013). Hence, differences in ecological attributes and behavioural responses among populations are also expected and to be predicted by phenotypic differentiation. Lizards have commonly been used to study the relationships between ecology, morphology and performance (Schulte et al., 2004; Irschick et al., 2005a, b; Goodman, Miles & Schwarzkopf, 2008; Crandell et al., 2014). Anoles, in particular, have received substantial attention because of the high diversity of the genus Anolis (~400 species; Losos, 2009) and high density of several species (Schoener & Schoener, 1980). However, most of our knowledge about Anolis stems from studies in the West Indies (Roughgarden, 1995; Losos, 2009) and, although anoline diversity is also high on the mainland (Nicholson et al., 2005), only a few ecomorphological studies have been conducted on mainland species (e.g. Pounds, 1988; Vitt et al., 2002; Irschick et al., 2005a, b; Velasco & Herrel, 2007; Logan et al., 2012; Siliceo-Cantero et al., 2016). Substantial differences in the ecomorphological relationships have been shown to occur between species from the West Indies and the mainland (Irschick et al., 1997). Anoles from the Greater Antilles have radiated into ecomorph classes defined according to their morphology, structural habitat and behaviour (Williams, 1972; Losos, 2009), but the same classes have not been recognized for species from the mainland (Irschick et al., 1997). Moreover, anoles from the Antilles show a significant relationship between hindlimb length and perch diameter, whereas most morphological features of mainland species are related to perch height (Irschick et al., 1997). A functional hypothesis that explains the above relationships assumes that the constraints imposed by local habitat characteristics have an impact on morphology, habitat use and mode of locomotion (Pounds, 1988; Harrison, Revell & Losos, 2015). The hypothesis predicts that relative limb length should be correlated positively with perch diameter, so that species can position their centre of gravity over the appropriate support (Pounds, 1988; Losos, 1990). Therefore, long-legged species should use broad perches and should run and jump more frequently than walk, whereas short-legged species should use narrow perches and move by walking. Given that wide-ranging species can occur in habitats with significant differences in structural characteristics, ecomorphological differences between populations of the same species should be detectable. This hypothesis, however, does not make predictions on the relationship between morphology, locomotion and perch height. Another explanation of differences in ecomorphological patterns among populations is that habitat use by individuals can be affected by the presence of competitors or predators (Calsbeek & Cox, 2010; McMillan & Irschick, 2010). Experimental evidence has shown that anoles can alter their habitat use, specifically perch height, depending on the presence and abundance of predatory species (Losos, Schoener & Spiller, 2004) or of aggressive congeners with similar habitat preferences (Rummel & Roughgarden, 1985; Leal, Rodríguez-Robles & Losos, 1998). Here, we examine three questions. Do populations of the same species differ in morphology, habitat use and locomotor behaviour? Is the functional hypothesis explaining differences among populations from a mainland species? Do congeners influence shifts in habitat use across populations of the same species? We use the widely distributed species Anolis limifrons (Cope, 1862) as a model to answer these questions. This common lizard spans from eastern Honduras to Panama on the Caribbean side, and from central Costa Rica to adjacent western Panama on the Pacific side, with populations found as high as 1340 m a.s.l. (Savage, 2002). In Costa Rica, the species occurs in both the Caribbean and Pacific versants, with one population also inhabiting Caño Island (Savage, 2002). Most research on this species in Costa Rica has been done on the Caribbean side, although some information from Pacific populations suggests differences in habitat use among versants. MATERIAL AND METHODS Study species and sites Anolis limifrons is a medium-sized (maximal body size 45 mm) and slender lizard with long hind legs. There is little sexual dimorphism; adult males have a short, white dewlap usually with an orange spot in the middle, whereas adult females exhibit a much shorter and immaculate dewlap (no dimorphism in size; Savage, 2002). This species is abundant in deciduous and evergreen forests, although high densities are also found in disturbed habitats (Andrews, 1991). Several aspects of its biology have been extensively investigated (Sexton, 1967; Ballinger, Marion & Sexton, 1970; Sexton et al., 1971; Andrews, 1979, 1991; Andrews, Stahl & Nicoletto, 1989); however, no ecomorphological study has yet been done. Research conducted primarily in Panama (Sexton & Heatwole, 1968) and on the Caribbean side of Costa Rica (Talbot, 1977) has shown that A. limifrons perches mostly on low vegetation (0.5–2.1 m) and uses a wide range of perch sites (e.g. fallen logs, tree trunks, stems, leaves and vines; Sexton, 1967; Savage, 2002). We conducted fieldwork at five sites in Costa Rica: two on the Caribbean versant (Pacuare Private Reserve, 10°12′N, 83°15′W; La Suerte Biological Station, 10°26′N, 83°47′W), two on the Pacific versant (Carara National Park, 09°46′N, 84°36′W; Golfito National Wildlife Refuge, 08°38′N, 83°09′W), and Caño Island Biological Reserve (08°42′N, 83°53′W) (Fig. 1), hereafter referred to as Pacuare, La Suerte, Carara, Golfito and Caño Island. Secondary tropical forests and similar climatic regimes occur at all sites, with average annual maximum temperatures ranging from 27 to 33 °C and average annual precipitation of 3500–5000 mm (Boza, 1996; Garber & Rehg, 1999; Lobo & Bolaños, 2005). The Sistema Nacional de Áreas de Conservación (SINAC) of Costa Rica approved all experimental procedures (permit number: 098-2006-SINAC). Figure 1. View largeDownload slide Map of Costa Rica showing the location of the five study sites, with an inset map of Central America showing the country’s location. Figure 1. View largeDownload slide Map of Costa Rica showing the location of the five study sites, with an inset map of Central America showing the country’s location. Habitat characteristics We measured the density of understorey vegetation at each study site to determine whether differences in habitat structure explain differences in ecomorphological patterns. We used a ‘vegetation profile board’ described by Nudds (1977) that consisted of a wooden board measuring 200 cm × 5 cm, with alternating vertical black and white bars 5 cm in width. The board was placed vertically within the forest 4 m away from the observer, and the number of bars covered by vegetation was counted. The proportion of bars covered by vegetation was used for subsequent analyses. Additionally, we measured the distance from a sampling point to the nearest tree > 10 cm in diameter at breast height (i.e. closest-individual method according to Cottam & Curtis, 1956) to determine the density of large trees. Both measurements were repeated 25 times at each site, with ~40 m between sampling points. Ecomorphological data We sampled lizards from August to December 2006 by walking along trails within the forest of each site, from 07:30 to 17:00 h. We searched for lizards by examining all vegetation and using the visual encounter surveys technique (Crump & Scott, 1994). Each individual captured was toe clipped for future identification, and the sex/age category (adult male, adult female and juvenile, with adults measuring ≥ 31 mm, according to Andrews, 1982) was determined. The following morphological data were recorded with a metallic ruler (to the nearest 0.5 mm): snout–vent length (SVL); lengths of the humerus, radius and hand (from the wrist to the tip of the fourth finger); the total length of the right fore leg; lengths of the femur, tibia and foot (from the heel to the tip of the fourth toe); the total length of the right hind leg; and tail length (only if undamaged). Also, we recorded the structural habitat (i.e. perch height and perch diameter) using a measuring tape (to the nearest 1 mm). For some analyses, perch height was divided into three discrete categories: 0–100, 101–200 and > 200 cm, and perch diameter was divided into four categories: 0–3, 3.1–6, 6.1–9 and > 9 cm. Focal observations Between nine and 11 individuals were observed at each site to determine the displacement and habitat use over a long period of time and to detect interactions with congeners. Random individuals were chosen to be used as focal animals, and observations were made for up to 90 min or until the subject was out of sight. During the period of observation, the mode of locomotion (run, jump, walk) of each motion bout was recorded, and perches used for ≥ 2 min were marked with flagging tape. At the end of the observation period, the focal animal was captured and the following data were recorded: SVL, sex/age category, height and diameter of the perches marked. Also, we used a scoring index of aggressiveness (from less to more aggressive behaviours) similar to that described by Hess & Losos (1991) as follows: escape when rival displayed aggressively (−2), escape when rival approached (−1), headbobs (1), dewlap extension (2) and mouth opening (3). The sum of scores of each individual was used for subsequent analysis. Agonistic interactions To control for interactions of A. limifrons with congeners, we carried out an experiment using a portable enclosure as a neutral arena. The enclosure (1.5 m long × 1.5 m wide × 1.5 m high) consisted of wooden poles as a frame and screen mesh as the walls and roof. We placed a piece of a trunk (10 cm in diameter) with branches simulating natural perches in the middle of the enclosure and removed all leaflitter and other vegetation. We staged encounters by introducing an adult male of A. limifrons and an adult male of the most common congener observed at each site (see Results). We carried out nine to 11 trials at each site, and no individual was used more than once. The SVL of both animals was measured before each encounter. For each encounter, the numbers and types of interactions were recorded during a 90 min period. We calculated an index of aggressiveness as explained above. Statistical analysis We analysed variation in the characteristics of the habitat across sites by transforming the two variables measured to fulfil normality requirements. We squared the values of the density of understorey vegetation and obtained the square root for the distance to the nearest tree. We used these transformed variables as dependent variables in separate one-way ANOVAs, with population as the fixed factor. We also performed Bonferroni post hoc tests to examine the specific relationships between sites. We examined morphological variation across populations of A. limifrons. We removed the effect of body size by regressing each morphological variable against SVL, and the residual values were then used for subsequent analyses. Given that all morphological variables were highly correlated at all sites (Pearson correlation coefficients > 0.83, P < 0.001), we performed a principal component analysis (PCA, using varimax as the rotation method) to reduce the number of morphological variables. We excluded tail length from the analysis because of the large number of individuals with damaged or regenerated tails. We used the principal components (PCs) extracted from the analysis (with eigenvalues > 1.0) to compare the morphology of individuals among populations through a multiple analysis of variance (MANOVA), using PCs as response variables and population as the fixed factor, and subsequent analysis of contrasts to identify specific differences among populations. The structural habitat was also compared among populations. For data collected during visual encounter surveys, we performed correspondence analyses to compare separately the categories of perch height and perch diameter among populations, among versants and versants against Caño Island. We also looked for differences in perch height and perch diameter measured during focal observations. We averaged both measurements (perch height and perch diameter) for each individual and then log10-transformed these averages to fulfil normality requirements. We used these log10-transformed variables as input for one-way ANOVAs, with population as a fixed factor. We were interested in identifying relationships between the structural habitat and morphology, and potential differences among populations in their mode of locomotion. We performed a canonical correlation analysis to detect an overall relationship of structural habitat and morphological variables including data from all populations, with perch height and perch diameter as dependent variables and both PCs as independent variables. We repeated this type of analysis for each site and each versant. In addition, we compared the mode of locomotion of A. limifrons across sites using a correspondence analysis. Finally, interspecific agonistic interactions detected during focal observations and the controlled experiment were compared using one-way ANOVA and one-way ANCOVA, respectively. In both cases, we first log10-transformed the sum of scores from the index of aggressiveness obtained for each individual to fulfil normality requirements. For the analysis of data from focal observations, we used this log10-transformed variable as the dependent variable and population as a fixed factor. We also used the log10-transformed sum of scores recorded in the experiment as the dependent variable, population as a fixed factor and the log10-transformed sum of scores of the congener as the covariable. All statistical tests were carried out using SPSS version 15.0.1 (IBM©). RESULTS Habitat characteristics Differences between sites were detected for the density of understorey vegetation (F4,120 = 3.71, P = 0.007) but not for the distance to the nearest tree (F4,119 = 1.86, P = 0.123). Bonferroni post hoc comparisons demonstrated that differences in understorey vegetation were observed only between Golfito and Caño Island (Fig. 2). Figure 2. View largeDownload slide Mean ± SEM of the relative density of understorey (white bars) and distance (in centimetres) to the nearest tree > 10 cm in diameter at breast height (grey bars) at each study site. Figure 2. View largeDownload slide Mean ± SEM of the relative density of understorey (white bars) and distance (in centimetres) to the nearest tree > 10 cm in diameter at breast height (grey bars) at each study site. Ecomorphological variation A total of 555 A. limifrons were measured at all sites (Table 1). The first two PCs explained 78.15% of the morphological variation observed, with eigenvalues of 4.41 and 1.84, respectively. Principal component 1 had high loadings for measurements of the hind leg, and PC2 for measurements of the fore leg (see Supporting Information, Table S1). Differences among populations in the overall morphology were detected (Table 2). Analysis of contrasts demonstrated that both populations from the Caribbean versant did not differ in morphology, nor did both populations from the Pacific versant (Fig. 3). However, there were differences when comparing Caribbean vs. Pacific and each versant vs. Caño Island (Table 2, Fig. 3). In general, individuals from Caribbean populations had relatively shorter fore legs and relatively longer hind legs than those from Pacific populations, while individuals at Caño Island presented intermediate measurements (see Supporting Information, Table S2, Fig. S1). Table 1. Number of adult males, adult females and juveniles of Anolis limifrons and the most common congener captured at each study site Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  View Large Table 1. Number of adult males, adult females and juveniles of Anolis limifrons and the most common congener captured at each study site Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  View Large Table 2. MANOVA and analysis of contrasts comparing principal components extracted from morphological variables measured for five populations of Anolis limifrons Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  View Large Table 2. MANOVA and analysis of contrasts comparing principal components extracted from morphological variables measured for five populations of Anolis limifrons Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  View Large Figure 3. View largeDownload slide Morphological differences of Anolis limifrons populations based on principal component (PC) scores (PC1, hind leg measurements; PC2, foreleg measurements) with 95% confidence ellipses. Squares, Pacific versant; circles, Caribbean versant. Figure 3. View largeDownload slide Morphological differences of Anolis limifrons populations based on principal component (PC) scores (PC1, hind leg measurements; PC2, foreleg measurements) with 95% confidence ellipses. Squares, Pacific versant; circles, Caribbean versant. Similar results were obtained when perch height and perch diameter were compared across populations and versants (Table 3). We found that A. limifrons used mostly low perches (0–100 cm) in the Caribbean and high perches (> 200 cm) in the Pacific (Fig. 4A). At all sites, lizards preferred thin perches (0–3 cm), although Caribbean populations used thicker perches (> 6 cm) more often than those from the Pacific (Fig. 4B). Lizards from Caño Island used a wider range of perch heights and perch diameters compared with mainland sites (Fig. 4; see also Supporting Information, Table S2, Fig. S2). These results were confirmed, in part, through observations of focal subjects within each population, because differences among populations were detected for perch height (F4,44 = 9.80, P < 0.001) but not for perch diameter (F4,43 = 1.67, P = 0.174). Table 3. Results of χ2 tests obtained using categories of perch height and perch diameter measured for five populations of Anolis limifrons Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  View Large Table 3. Results of χ2 tests obtained using categories of perch height and perch diameter measured for five populations of Anolis limifrons Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  View Large Figure 4. View largeDownload slide Classification of Anolis limifrons populations according to perch height (A) and perch diameter (B) categories. Dimension 1 explains 96.08% (eigenvalue = 0.206) in A and 67.27% (eigenvalue = 0.066) in B of the χ2 value, and dimension 2 explains 3.92% (eigenvalue = 0.008) and 31.57% (eigenvalue = 0.031), respectively. Figure 4. View largeDownload slide Classification of Anolis limifrons populations according to perch height (A) and perch diameter (B) categories. Dimension 1 explains 96.08% (eigenvalue = 0.206) in A and 67.27% (eigenvalue = 0.066) in B of the χ2 value, and dimension 2 explains 3.92% (eigenvalue = 0.008) and 31.57% (eigenvalue = 0.031), respectively. Relationship between morphology, habitat use and locomotion A significant association (R = 0.34) between morphology and structural habitat (Wilks’s λ = 0.89, F4,1070 = 16.39, P < 0.001) for all populations of A. limifrons was found. Principal component 1 (hind leg measurements) and perch height accounted for most of the observed variation. The overall pattern found shows that individuals from all populations with larger values of hind leg measurements tended to perch lower (Supporting Information, Fig. S3). However, when considering geographical regions separately (i.e. Caribbean versant, Pacific versant and Caño Island) no significant correlation was found (Caribbean, R = 0.16, F4,566 = 1.98, P = 0.097; Pacific, R = 0.22, F4,318 = 1.92, P = 0.107; Caño Island, F4,170 = 1.97, P = 0.10). We also found significant differences in the mode of locomotion between sites (χ2 = 54.43, d.f. = 8, P < 0.001), with lizards from Caño Island running more, those from both Caribbean sites jumping more and those from both Pacific sites walking more often than individuals from other regions (Fig. 5). Figure 5. View largeDownload slide Classification of Anolis limifrons populations according to the mode of locomotion of individuals. Dimension 1 explains 94.63% (eigenvalue = 0.045), and dimension 2 explains 5.37% (eigenvalue = 0.003) of the χ2 value. Figure 5. View largeDownload slide Classification of Anolis limifrons populations according to the mode of locomotion of individuals. Dimension 1 explains 94.63% (eigenvalue = 0.045), and dimension 2 explains 5.37% (eigenvalue = 0.003) of the χ2 value. Agonistic interactions No differences in aggressiveness were detected between populations during observation of focal subjects (F3,36 = 1.24, P = 0.311). However, by controlling the interactions within an enclosure, we did find significant differences between populations (F3,35 = 6.06, P = 0.002). Overall, lizards from Caribbean populations were less aggressive towards but received more displays from the congener, whereas in the Pacific the opposite pattern was observed (Fig. 6). Figure 6. View largeDownload slide Sum of scores of the index of aggressiveness recorded during a manipulative experiment for male–male encounters of Anolis limifrons and the most common congener (Anolis humilis for La Suerte and Pacuare, Anolis cupreus for Carara, and Anolis polylepis for Golfito) found at each study site. Figure 6. View largeDownload slide Sum of scores of the index of aggressiveness recorded during a manipulative experiment for male–male encounters of Anolis limifrons and the most common congener (Anolis humilis for La Suerte and Pacuare, Anolis cupreus for Carara, and Anolis polylepis for Golfito) found at each study site. DISCUSSION We were interested to determine whether ecomorphological variation occurs along the distribution of a lizard and, if so, whether this variation is explained by a functional hypothesis of habitat use or by the influence of congeners affecting ecological preferences of the species. We found significant differences in morphology, structural habitat and mode of locomotion among populations of A. limifrons, and a significant but weak relationship between morphology and perch height. Interpopulation variation in ecomorphological patterns Morphological traits of A. limifrons differed among populations from Caño Island, the Pacific side and the Caribbean side of Costa Rica. These differences can be explained in two ways. First, they might reflect genetic differentiation. Costa Rican mountain ranges divide the territory into two clearly defined versants (Fig. 1), creating a geographical barrier for a variety of organisms (Savage, 2002). This barrier could have interrupted the gene flow for A. limifrons, so that populations from both versants could have been relatively isolated and different phenotypes could have adjusted to different conditions. Pacific populations of this species from Costa Rica and Panama have previously been referred to as Anolis biscutiger (Taylor, 1956; Echelle, Echelle & Fitch, 1971; Fitch, 1973), although most researchers do not accept this species as valid (Savage & Bolaños, 2009). Although genetic differences among populations remain to be tested, it seems unlikely that these differences, if present, represent species differentiation. Morphological variation found here is not high enough to account for different species, and other researchers have been unable to find major differences between populations of both versants (Savage, 2002). In addition, it is unlikely that gene flow has been interrupted for this species, because the altitudinal gradient that A. limifrons occupies (0–1340 m a.s.l.) could allow individuals to cross versants (Fig. 1). A second possible explanation is that morphological variation could be the result of adaptive phenotypic plasticity. Natural selection would favour phenotypes that perform better in habitats with particular conditions, accentuating morphological differences between populations facing different selective pressures (Losos et al., 2000). It is unknown whether A. limifrons exhibits high phenotypic plasticity, but this characteristic has been reported previously for other anoline species (Losos et al., 2000) and could be a widespread feature in the genus Anolis (Kolbe & Losos, 2005). Variation in structural habitat use was also found among populations from the geographical regions (low perches in the Caribbean, high perches in the Pacific and wide range of perch heights on Caño Island). Variation in morphology and habitat use is also linked with variation in locomotor behaviour. Lizards from each geographical region exhibited a distinct mode of locomotion (Caribbean populations prefer to jump, Pacific populations walk more, and on Caño Island individuals run more often). Locomotor behaviour is an expression of morphological attributes and the characteristics of the vegetation available to the individuals (Losos, 2009; Harrison et al., 2015). Thus, our study confirms that populations with long hind legs and short fore legs and using dense, low perches jump more frequently (Caribbean). However, the predictions that populations with short legs and using narrow perches walk more often (Pacific), and populations with long legs and using broad perches tend to run (Caño Island), are partly confirmed. This ecomorphological variation is thus not fully explained by the functional hypothesis proposed for the West Indies species. Factors explaining ecomorphological patterns Different selective pressures could be acting on populations of these regions. Anoline lizards can use a particular structural habitat in order to maximize the visibility of their territory (Scott et al., 1976), avoid predators (Schoener, Spiller & Losos, 2002) or reduce interspecific competition (Pacala & Roughgarden, 1985; Losos, Marks & Schoener, 1993). Therefore, forest structure, predation intensity and the frequency of interspecific agonistic interactions could be major selective forces shaping habitat use by A. limifrons. We did not measure predation intensity in this study, so we do not know its potential variation across sites, and further research is needed. The density of the understorey vegetation and the distance to the nearest tree, in contrast, were similar among almost all study sites (only Golfito and Caño Island differed in their values of density of the understorey vegetation), such that it seems unlikely that forest structure influences the variation in structural habitat use. We did detect differences in the number and type of interspecific agonistic interactions across populations. Consequently, the presence of congeners and their interspecific agonistic interactions could be impacting the structural habitat use of A. limifrons, a phenomenon that has been referred to as agonistic character displacement (Grether et al., 2009; Dufour, Herrel & Losos, 2018). In this case, Caribbean populations are adapted to use low perches, because A. limifrons is not affected by the presence of other congeners (Talbot, 1979). However, Pacific populations have shifted their preferences to use higher perches, because interactions with larger congeners (e.g. Anolis polylepis and Anolis cupreus, both with maximal SVLs of 57 mm; Savage, 2002), already using low perches, can be costly. Another source of support for this conclusion comes from the comparison between the mainland and Caño Island. On the island, A. limifrons is the only anoline lizard present, and evidence from the present study demonstrates that the species experienced an ecological release, with individuals occupying different parts of the habitat. Interspecific agonistic interactions are common between anoles (Ortiz & Jenssen, 1982; Hess & Losos, 1991) and other lizard species (Korner, Whiting & Ferguson, 2000) and can be a factor shaping the direction of selection. Ecomorphological relationships The ecomorphological relationships found in the present study seem to confirm information previously reported on mainland anoline species (Irschick et al., 1997). The morphological characteristics of anoles from Greater Antilles are correlated mainly with perch diameter, whereas the correlation is stronger with perch height for mainland species. This poses a problem, because no satisfactory explanation exists for such a correlation. One functional hypothesis has been proposed to explain the positive relationship of perch height and the number of lamellae found for mainland anoles, referring to a greater capacity to cling to the substrate of species that use higher perches (Irschick et al., 1997). However, this hypothesis does not explain the positive correlation found here between hind leg measurements and perch height. Even more intriguing is the fact that no apparent relationship was found when considering geographical regions of A. limifrons separately. An ecomorphological relationship was found (i.e. hind leg measurements and perch height) when all populations were pooled and the species was treated as a whole, but this relationship became non-significant when each geographical region was analysed independently. This finding has two important consequences. First, it calls attention to ecomorphological studies comparing interspecific patterns, because the ecomorphological data should take into account population differences of each species considered (Losos, 2009). Second, it reflects the finding that the degree of phenotypic variation does not match the degree of ecological variation, with morphological differences among populations being smaller than ecological ones. This means that if no genetic differences are present among populations, the phenotypic plasticity is smaller than the ecological plasticity for A. limifrons. Considerations for evolutionary ecology The extent to which morphology constrains ecological breadth has major implications for the direction of selection and how communities are structured (Strong et al., 1984; Pounds, 1988; Losos et al., 2003). If the selective pressures producing morphological changes persist through time and gene flow is interrupted, then local adaptation can occur and, eventually, speciation. However, if populations have not evolved genetically to adapt to different environments and phenotypic plasticity is present within a species, then selection within a population may not lead to evolutionary change (Losos, 2009). This means that if phenotypically plastic species are capable of occupying different environments with novel conditions, then shifts in ecological preferences (e.g. structural habitat) could be expected without genetic differentiation. Moreover, for some species, such as A. limifrons, the magnitude of ecological shifts among populations appears to be higher than phenotypic changes. Shifts in ecological preferences have been shown for many Anolis lizards (Jenssen, 1973; Rummel & Roughgarden, 1985; Losos et al., 1993, 2004), so that both phenotypic and ecological plasticity could be widespread phenomena within this genus. The major roles that vegetation structure and predation pressure can play in morphological and ecological variation are undeniable. However, other factors can also have a strong influence, such as interspecific agonistic interactions. Therefore, it becomes essential to consider all biotic and abiotic factors shaping the evolutionary pathways of a species. These pathways can vary greatly throughout the distribution of a species, because different selective pressures can be acting at different sites. Species showing both phenotypic and ecological plasticity are more capable of overcoming these costs, so that coexistence with closely related species is possible. However, the intensity of selective pressures varies not only spatially but also temporally, and thus community structure could be a rather dynamic process. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher's web-site: Table S1. Principal component loadings of residual morphological variables (lengths). Tail length is not included because of the high proportion of individuals with damaged tails. Significant loadings are in bold. Table S2. Mean ± SD (minimal and maximal values) of morphological (in millimetres) and structural habitat (in centimetres) variables recorded for Anolis limifrons at each study site. Figure S1. Mean ± SEM of the principal component (PC) scores (PC1, grey bars; PC2, white bars) extracted from morphological variables recorded for Anolis limifrons at each study site. Figure S2. Number of individuals found for each category of perch height (left panels) and perch diameter (right panels) in five populations of Anolis limifrons: Carara (A), Golfito (B), Caño Island (C), La Suerte (D) and Pacuare (E). Figure S3. Relationship of the first principal component (PC1) scores extracted from morphological variables and perch height recorded for five populations of Anolis limifrons. Black regression line corresponds to all data pooled together; coloured regression lines correspond to each population. ACKNOWLEDGEMENTS We are thankful to Gilbert Barrantes, Gerardo Ávalos, Mahmood Sasa and José Manuel Mora for their criticisms and suggestions to improve the manuscript. We are also grateful to the staff from the study sites and to Viviana Arguedas for her support and help in the field. Jonathan Losos, Martin Whiting, Daniel Noble, Dani Chandrasoma, Kerrie Wechmann, Siobhan Dennison and Pau Carazo also contributed valuable comments. We also thank three anonymous reviewers for their helpful comments. This study received donations from The Holly Hill Charitable Trust and Idea Wild and the logistic support of the School of Biology, University of Costa Rica. REFERENCES Andrews RM. 1979. Reproductive effort of female Anolis limifrons (Sauria: Iguanidae). Copeia  1979: 620– 626. Google Scholar CrossRef Search ADS   Andrews RM. 1982. Patterns of growth in reptiles. In: Gans C, Pough HF, eds. Biology of the reptilia. Vol. 13 Physiology D: physiological ecology . New York: Academic Press, 273– 320. Andrews RM. 1991. Population stability of a tropical lizard. Ecology  72: 1204– 1217. Google Scholar CrossRef Search ADS   Andrews RM, Stahl SJ, Nicoletto PF. 1989. Intra-population variation in age of sexual maturity of the tropical lizard Anolis limifrons in Panama. Copeia  1989: 751– 753. Google Scholar CrossRef Search ADS   Ballinger RE, Marion KR, Sexton OJ. 1970. Thermal ecology of the lizard, Anolis limifrons with comparative notes on three additional Panamanian anoles. Ecology  51: 246– 254. Google Scholar CrossRef Search ADS   Boza MA. 1996. Costa Rica, Parques Nacionales . San José, Costa Rica: Incafo. Calderón-Espinosa ML, Ortega-León AM, Zamora-Abrego JG. 2013. Intraspecific variation in body size and shape in an Andean highland anole species, Anolis ventrimaculatus (Squamata: Dactyloidae). Revista de biologia tropical  61: 255– 262. Google Scholar PubMed  Calsbeek R, Cox RM. 2010. Experimentally assessing the relative importance of predation and competition as agents of selection. Nature  465: 613– 616. Google Scholar CrossRef Search ADS PubMed  Chapman BB, Hulthén K, Brönmark C, Nilsson PA, Skov C, Hansson LA, Brodersen J. 2015. Shape up or ship out: migratory behaviour predicts morphology across spatial scale in a freshwater fish. Journal of Animal Ecology  84: 1187– 1193. Google Scholar CrossRef Search ADS PubMed  Cope, ED. 1862. Contributions to Neotropical saurology. Proceedings of the Academy of Natural Sciences of Philadelphia  14: 176– 88. Cottam G, Curtis JT. 1956. The use of distance measures in phytosociological sampling. Ecology  37: 451– 460. Google Scholar CrossRef Search ADS   Crandell KE, Herrel A, Sasa M, Losos JB, Autumn K. 2014. Stick or grip? Co-evolution of adhesive toepads and claws in Anolis lizards. Zoology  117: 363– 369. Google Scholar CrossRef Search ADS PubMed  Crump ML, Scott NJJr. 1994. Visual encounter surveys. In: Heyer WR, Donnelly MA, McDiarmid RW, Hayek L-AC, Foster MS, eds. Measuring and monitoring biological diversity: standard methods for amphibians . Washington: Smithsonian Institution Press, 84– 92. Dufour CMS, Herrel A, Losos JB. 2018. Ecological character displacement between a native and an introduced species: the invasion of Anolis cristatellus in Dominica. Biological Journal of the Linnean Society  123: 43– 54. Google Scholar CrossRef Search ADS   Echelle AA, Echelle AF, Fitch HS. 1971. A comparative analysis of aggressive display in nine species of Costa Rican Anolis. Herpetologica  27: 271– 288. Fitch HS. 1973. A field study of Costa Rican lizards. University of Kansas Science Bulletin  50: 39– 126. Google Scholar CrossRef Search ADS   Garber PA, Rehg JA. 1999. The ecological role of the prehensile tail in white-faced capuchins (Cebus capucinus). American Journal of Physical Anthropology  110: 325– 339. Google Scholar CrossRef Search ADS PubMed  Garland TJr, Losos JB. 1994. Ecological morphology of locomotor performance in squamate reptiles. In: Wainwright PC, Reilly SM, eds. Ecological morphology: integrative organismal biology . Chicago: University of Chicago Press, 240– 302. Goodman BA, Miles DB, Schwarzkopf L. 2008. Life on the rocks: habitat use drives morphological and performance evolution in lizards. Ecology  89: 3462– 3471. Google Scholar CrossRef Search ADS PubMed  Grether GF, Losin N, Anderson CN, Okamoto K. 2009. The role of interspecific interference competition in character displacement and the evolution of competitor recognition. Biological Reviews  84: 617– 635. Google Scholar CrossRef Search ADS PubMed  Harrison AS, Revell LJ, Losos JB. 2015. Correlated evolution of microhabitat, morphology, and behavior in West Indian Anolis lizards: a test of the habitat matrix model. Behaviour  152: 1187– 1207. Google Scholar CrossRef Search ADS   Hess NE, Losos JB. 1991. Interspecific aggression between Anolis cristatellus and A. gundlachi: comparison of sympatric and allopatric populations. Journal of Herpetology  25: 256– 259. Google Scholar CrossRef Search ADS   Irschick DJ, Carlisle E, Elstrott J, Ramos M, Buckley C, Vanhooydonck B, Meyers J, Herrel A. 2005a. A comparison of habitat use, morphology, clinging performance and escape behaviour among two divergent green anole lizard (Anolis carolinensis) populations. Biological Journal of the Linnean Society  85: 223– 234. Google Scholar CrossRef Search ADS   Irschick DJ, Vanhooydonck B, Herrel A, Meyers J. 2005b. Intraspecific correlations among morphology, performance and habitat use within a green anole lizard (Anolis carolinensis) population. Biological Journal of the Linnean Society  85: 211– 221. Google Scholar CrossRef Search ADS   Irschick DJ, Vitt LJ, Zani PA, Losos JB. 1997. A comparison of evolutionary radiations in mainland and Caribbean Anolis lizards. Ecology  78: 2191– 2203. Google Scholar CrossRef Search ADS   Jenssen TA. 1973. Shift in the structural habitat of Anolis opalinus due to congeneric competition. Ecology  54: 865– 869. Google Scholar CrossRef Search ADS   Kolbe JJ, Larson A, Losos JB. 2007. Differential admixture shapes morphological variation among invasive populations of the lizard Anolis sagrei. Molecular Ecology  16: 1579– 1591. Google Scholar CrossRef Search ADS PubMed  Kolbe JJ, Losos JB. 2005. Hind-limb length plasticity in Anolis carolinensis. Journal of Herpetology  39: 674– 678 Google Scholar CrossRef Search ADS   Korner P, Whiting MJ, Ferguson JWH. 2000. Interspecific aggression in flat lizards suggests poor species recognition. African Journal of Herpetology  49: 139– 146. Google Scholar CrossRef Search ADS   Leal M, Rodríguez-Robles JA, Losos JB. 1998. An experimental study of interspecific interactions between two Puerto Rican Anolis lizards. Oecologia  117: 273– 278. Google Scholar CrossRef Search ADS PubMed  Lobo J, Bolaños F, eds. 2005. Historia natural de Golfito . Santo Domingo de Heredia, Costa Rica: Editorial INBio. Logan ML, Montgomery CE, Boback SM, Reed RN, Campbell JA. 2012. Divergence in morphology, but not habitat use, despite low genetic differentiation among insular populations of the lizard Anolis lemurinus in Honduras. Journal of Tropical Ecology  28: 215– 222. Google Scholar CrossRef Search ADS   Losos JB. 1990. Ecomorphology, performance capability, and scaling of West Indian Anolis lizards: an evolutionary analysis. Ecological Monographs  60: 369– 388. Google Scholar CrossRef Search ADS   Losos JB. 2009. Lizards in an evolutionary tree: ecology and adaptive radiation of anoles . Oakland, California: University of California Press. Losos JB, Creer DA, Glossip D, Goellner R, Hampton A, Roberts G, Haskell N, Taylor P, Ettling J. 2000. Evolutionary implications of phenotypic plasticity in the hindlimb of the lizard Anolis sagrei. Evolution  54: 301– 305. Google Scholar PubMed  Losos JB, Leal M, Glor RE, De Queiroz K, Hertz PE, Rodríguez Schettino L, Lara AC, Jackman TR, Larson A. 2003. Niche lability in the evolution of a Caribbean lizard community. Nature  424: 542– 545. Google Scholar CrossRef Search ADS PubMed  Losos JB, Marks JC, Schoener TW. 1993. Habitat use and ecological interactions of an introduced and a native species of Anolis lizard on Grand Cayman, with a review of the outcomes of anole introductions. Oecologia  95: 525– 532. Google Scholar CrossRef Search ADS PubMed  Losos JB, Schoener TW, Spiller DA. 2004. Predator-induced behaviour shifts and natural selection in field-experimental lizard populations. Nature  432: 505– 508. Google Scholar CrossRef Search ADS PubMed  Lostrom S, Evans JP, Grierson PF, Collin SP, Davies PM, Kelley JL. 2015. Linking stream ecology with morphological variability in a native freshwater fish from semi-arid Australia. Ecology and Evolution  5: 3272– 3287. Google Scholar CrossRef Search ADS PubMed  McMillan DM, Irschick DJ. 2010. Experimental test of predation and competition pressures on the Green Anole (Anolis carolinensis) in varying structural habitats. Journal of Herpetology  44: 272– 278. Google Scholar CrossRef Search ADS   Nicholson KE, Glor RE, Kolbe JJ, Larson A, Hedges SB, Losos JB. 2005. Mainland colonization by island lizards. Journal of Biogeography  32: 929– 938. Google Scholar CrossRef Search ADS   Nudds TD. 1977. Quantifying the vegetative structure of wildlife cover. Wildlife Society Bulletin  5: 113– 117. Ord TJ, Klomp DA. 2014. Habitat partitioning and morphological differentiation: the Southeast Asian Draco lizards and Caribbean Anolis lizards compared. Oecologia  175: 651– 666. Google Scholar CrossRef Search ADS PubMed  Ortiz PR, Jenssen TA. 1982. Interspecific aggression between lizard competitors, Anolis cooki and Anolis cristatellus. Zeitschrift für Tierpsychologie  60: 227– 238. Google Scholar CrossRef Search ADS   Pacala SW, Roughgarden J. 1985. Population experiments with the Anolis lizards of St. Maarten and St. Eustatius. Ecology  66: 129– 141. Google Scholar CrossRef Search ADS   Pereira Leitão R, Sánchez-Botero JI, Kasper D, Trivério-Cardoso V, Morais Araújo C, Zuanon J, Pellegrini Caramaschi É. 2015. Microhabitat segregation and fine ecomorphological dissimilarity between two closely phylogenetically related grazer fishes in an Atlantic Forest stream, Brazil. Environmental Biology of Fishes  98: 2009– 2019. Google Scholar CrossRef Search ADS   Pounds JA. 1988. Ecomorphology, locomotion, and microhabitat structure: patterns in a tropical mainland Anolis community. Ecological Monographs  58: 299– 320. Google Scholar CrossRef Search ADS   Roughgarden J. 1995. Anolis lizards of the Caribbean: ecology, evolution and plate tectonics . New York: Oxford University Press. Rummel JD, Roughgarden J. 1985. Effects of reduced perch-height separation on competition between two Anolis lizards. Ecology  66: 430– 444. Google Scholar CrossRef Search ADS   Savage JM. 2002. The amphibians and reptiles of Costa Rica: a herpetofauna between two continents, between two seas . Chicago: University of Chicago Press. Savage JM, Bolaños F. 2009. A checklist of the amphibians and reptiles of Costa Rica: additions and nomenclatural revisions. Zootaxa  2005: 1– 23. Schoener TW, Schoener A. 1980. Densities, sex ratios, and population structure in four species of Bahamian Anolis lizards. Journal of Animal Ecology  49: 19– 53. Google Scholar CrossRef Search ADS   Schoener TW, Spiller DA, Losos JB. 2002. Predation on a common Anolis lizard: can the food-web effects of a devastating predator be reversed? Ecological Monographs  72: 383– 407. Google Scholar CrossRef Search ADS   Schulte JA2nd, Losos JB, Cruz FB, Núñez H. 2004. The relationship between morphology, escape behaviour and microhabitat occupation in the lizard clade Liolaemus (Iguanidae: Tropidurinae: Liolaemini). Journal of Evolutionary Biology  17: 408– 420. Google Scholar CrossRef Search ADS PubMed  Scott NJJr, Wilson DE, Jones C, Andrews RM. 1976. The choice of perch dimensions by lizards of the genus Anolis (Reptilia, Lacertilia, Iguanidae). Journal of Herpetology  10: 75– 84. Google Scholar CrossRef Search ADS   Sexton OJ. 1967. Population changes in a tropical lizard Anolis limifrons on Barro Colorado Island, Panama Canal Zone. Copeia  1967: 219– 222. Google Scholar CrossRef Search ADS   Sexton OJ, Heatwole H. 1968. An experimental investigation of habitat selection and water loss in some anoline lizards. Ecology  49: 762– 767. Google Scholar CrossRef Search ADS   Sexton OJ, Ortleb EP, Hathaway LM, Ballinger RE, Licht P. 1971. Reproductive cycles of three species of anoline lizards from the Isthmus of Panama. Ecology  52: 201– 215. Google Scholar CrossRef Search ADS   Siliceo-Cantero HH, García A, Reynolds RG, Pacheco G, Lister BC. 2016. Dimorphism and divergence in island and mainland Anoles. Biological Journal of the Linnean Society  118: 852– 872. Google Scholar CrossRef Search ADS   da Silva JM, Herrel A, Measey GJ, Vanhooydonck B, Tolley KA. 2014. Linking microhabitat structure, morphology and locomotor performance traits in a recent radiation of dwarf chameleons. Functional Ecology  28: 702– 713. Google Scholar CrossRef Search ADS   Strong DRJr, Simberloff D, Abele LG, Thistle AB, eds. 1984. Ecological communities: conceptual issues and the evidence . Princeton: Princeton University Press. Talbot JJ. 1977. Habitat selection in two tropical anoline lizards. Herpetologica  33: 114– 123. Talbot JJ. 1979. Time budget, niche overlap, inter- and intraspecific aggression in Anolis humilis and A. limifrons from Costa Rica. Copeia  1979: 472– 481. Google Scholar CrossRef Search ADS   Taylor EH. 1956. A review of the lizards of Costa Rica. University of Kansas Science Bulletin  38: 3– 322. Velasco JA, Herrel A. 2007. Ecomorphology of Anolis lizards of the Chocó region in Colombia and comparisons with Greater Antillean ecomorphs. Biological Journal of the Linnean Society  92: 29– 39. Google Scholar CrossRef Search ADS   Villalobos F, Arita HT. 2014. Morphological diversity at different spatial scales in a Neotropical bat assemblage. Oecologia  176: 557– 568. Google Scholar CrossRef Search ADS PubMed  Vitt LJ, Avila-Pires TCS, Zani PA, Espósito MC. 2002. Life in shade: the ecology of Anolis trachyderma (Squamata: Polychrotidae) in Amazonian Ecuador and Brazil, with comparisons to ecologically similar anoles. Copeia  2002: 275– 286. Google Scholar CrossRef Search ADS   Wainwright PC, Reilly SM, eds. 1994. Ecological morphology: integrative organismal biology . Chicago: University of Chicago Press. Williams EE. 1972. The origin of faunas: evolution of lizard congeners in a complex island fauna: a trial analysis. Evolutionary Biology  6: 47– 89. © 2018 The Linnean Society of London, Biological Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biological Journal of the Linnean Society Oxford University Press

Morphological and ecological variation of a tropical anoline lizard: are agonistic interactions shaping ecomorphological relationships?

Loading next page...
 
/lp/ou_press/morphological-and-ecological-variation-of-a-tropical-anoline-lizard-M0q2D6JqM1
Publisher
The Linnean Society of London
Copyright
© 2018 The Linnean Society of London, Biological Journal of the Linnean Society
ISSN
0024-4066
eISSN
1095-8312
D.O.I.
10.1093/biolinnean/bly049
Publisher site
See Article on Publisher Site

Abstract

Abstract Ecologists have long been intrigued by which factors influence habitat use by an organism and how communities are structured. However, the links between habitat preferences, morphology, biotic interactions and community structure are still poorly understood. Moreover, interpopulation variation in ecomorphological relationships has usually been neglected. Here, we use a wide-ranging Anolis lizard, Anolis limifrons, to test whether interpopulation variation in morphology and habitat use is a function of interspecific agonistic interactions across the distribution of this species in Costa Rica. We found differences both in morphology and in habitat use among populations of A. limifrons, with populations from the Caribbean versant of Costa Rica having longer hind legs and perching lower than those from the Pacific versant. The intensity of interspecific agonistic interactions also varied across versants, with A. limifrons from Pacific sites displaying more often to congeners than those from the Caribbean. Agonistic interactions appear to be an important factor shaping habitat use and morphology. These findings can be explained by an interaction between phenotypic plasticity and ecological plasticity. Anolis limifrons, Costa Rica, ecomorphology, habitat use, mainland lizards, populations, structural habitat INTRODUCTION One cornerstone in ecological studies has been the understanding of the relationship between morphology and ecology (Wainwright & Reilly, 1994). Extensive research has been conducted across different taxa to predict ecological attributes (e.g. habitat use and dietary breadth) based on morphological traits (da Silva et al., 2014; Villalobos & Arita, 2014; Pereira Leitão et al., 2015). These ecomorphological studies rely on the assumption that a given phenotype will perform better in an environment with specific conditions than other phenotypes, affecting fitness and the direction of selection (Losos et al., 2000). Therefore, a tight link between habitat characteristics, morphology, ecology and behaviour has been proposed (Garland & Losos, 1994), influencing the habitat selection, interspecific interactions and trophic relationships of an organism (Strong et al., 1984). Most often, this link has been analysed in comparative studies among species (Ord & Klomp, 2014; Villalobos & Arita, 2014), although ecomorphological relationships are also present within a species (Irschick et al., 2005a, b; Chapman et al., 2015). Populations of the same species can occupy a wide range of habitats, where selective pressures could be very different (Kolbe, Larson & Losos, 2007; Lostrom et al., 2015). Therefore, high morphological variation is expected to occur among populations in traits that allow individuals to carry out relevant ecological tasks (Irschick et al., 2005a; Calderón-Espinosa, Ortega-León & Zamora-Abrego, 2013). Hence, differences in ecological attributes and behavioural responses among populations are also expected and to be predicted by phenotypic differentiation. Lizards have commonly been used to study the relationships between ecology, morphology and performance (Schulte et al., 2004; Irschick et al., 2005a, b; Goodman, Miles & Schwarzkopf, 2008; Crandell et al., 2014). Anoles, in particular, have received substantial attention because of the high diversity of the genus Anolis (~400 species; Losos, 2009) and high density of several species (Schoener & Schoener, 1980). However, most of our knowledge about Anolis stems from studies in the West Indies (Roughgarden, 1995; Losos, 2009) and, although anoline diversity is also high on the mainland (Nicholson et al., 2005), only a few ecomorphological studies have been conducted on mainland species (e.g. Pounds, 1988; Vitt et al., 2002; Irschick et al., 2005a, b; Velasco & Herrel, 2007; Logan et al., 2012; Siliceo-Cantero et al., 2016). Substantial differences in the ecomorphological relationships have been shown to occur between species from the West Indies and the mainland (Irschick et al., 1997). Anoles from the Greater Antilles have radiated into ecomorph classes defined according to their morphology, structural habitat and behaviour (Williams, 1972; Losos, 2009), but the same classes have not been recognized for species from the mainland (Irschick et al., 1997). Moreover, anoles from the Antilles show a significant relationship between hindlimb length and perch diameter, whereas most morphological features of mainland species are related to perch height (Irschick et al., 1997). A functional hypothesis that explains the above relationships assumes that the constraints imposed by local habitat characteristics have an impact on morphology, habitat use and mode of locomotion (Pounds, 1988; Harrison, Revell & Losos, 2015). The hypothesis predicts that relative limb length should be correlated positively with perch diameter, so that species can position their centre of gravity over the appropriate support (Pounds, 1988; Losos, 1990). Therefore, long-legged species should use broad perches and should run and jump more frequently than walk, whereas short-legged species should use narrow perches and move by walking. Given that wide-ranging species can occur in habitats with significant differences in structural characteristics, ecomorphological differences between populations of the same species should be detectable. This hypothesis, however, does not make predictions on the relationship between morphology, locomotion and perch height. Another explanation of differences in ecomorphological patterns among populations is that habitat use by individuals can be affected by the presence of competitors or predators (Calsbeek & Cox, 2010; McMillan & Irschick, 2010). Experimental evidence has shown that anoles can alter their habitat use, specifically perch height, depending on the presence and abundance of predatory species (Losos, Schoener & Spiller, 2004) or of aggressive congeners with similar habitat preferences (Rummel & Roughgarden, 1985; Leal, Rodríguez-Robles & Losos, 1998). Here, we examine three questions. Do populations of the same species differ in morphology, habitat use and locomotor behaviour? Is the functional hypothesis explaining differences among populations from a mainland species? Do congeners influence shifts in habitat use across populations of the same species? We use the widely distributed species Anolis limifrons (Cope, 1862) as a model to answer these questions. This common lizard spans from eastern Honduras to Panama on the Caribbean side, and from central Costa Rica to adjacent western Panama on the Pacific side, with populations found as high as 1340 m a.s.l. (Savage, 2002). In Costa Rica, the species occurs in both the Caribbean and Pacific versants, with one population also inhabiting Caño Island (Savage, 2002). Most research on this species in Costa Rica has been done on the Caribbean side, although some information from Pacific populations suggests differences in habitat use among versants. MATERIAL AND METHODS Study species and sites Anolis limifrons is a medium-sized (maximal body size 45 mm) and slender lizard with long hind legs. There is little sexual dimorphism; adult males have a short, white dewlap usually with an orange spot in the middle, whereas adult females exhibit a much shorter and immaculate dewlap (no dimorphism in size; Savage, 2002). This species is abundant in deciduous and evergreen forests, although high densities are also found in disturbed habitats (Andrews, 1991). Several aspects of its biology have been extensively investigated (Sexton, 1967; Ballinger, Marion & Sexton, 1970; Sexton et al., 1971; Andrews, 1979, 1991; Andrews, Stahl & Nicoletto, 1989); however, no ecomorphological study has yet been done. Research conducted primarily in Panama (Sexton & Heatwole, 1968) and on the Caribbean side of Costa Rica (Talbot, 1977) has shown that A. limifrons perches mostly on low vegetation (0.5–2.1 m) and uses a wide range of perch sites (e.g. fallen logs, tree trunks, stems, leaves and vines; Sexton, 1967; Savage, 2002). We conducted fieldwork at five sites in Costa Rica: two on the Caribbean versant (Pacuare Private Reserve, 10°12′N, 83°15′W; La Suerte Biological Station, 10°26′N, 83°47′W), two on the Pacific versant (Carara National Park, 09°46′N, 84°36′W; Golfito National Wildlife Refuge, 08°38′N, 83°09′W), and Caño Island Biological Reserve (08°42′N, 83°53′W) (Fig. 1), hereafter referred to as Pacuare, La Suerte, Carara, Golfito and Caño Island. Secondary tropical forests and similar climatic regimes occur at all sites, with average annual maximum temperatures ranging from 27 to 33 °C and average annual precipitation of 3500–5000 mm (Boza, 1996; Garber & Rehg, 1999; Lobo & Bolaños, 2005). The Sistema Nacional de Áreas de Conservación (SINAC) of Costa Rica approved all experimental procedures (permit number: 098-2006-SINAC). Figure 1. View largeDownload slide Map of Costa Rica showing the location of the five study sites, with an inset map of Central America showing the country’s location. Figure 1. View largeDownload slide Map of Costa Rica showing the location of the five study sites, with an inset map of Central America showing the country’s location. Habitat characteristics We measured the density of understorey vegetation at each study site to determine whether differences in habitat structure explain differences in ecomorphological patterns. We used a ‘vegetation profile board’ described by Nudds (1977) that consisted of a wooden board measuring 200 cm × 5 cm, with alternating vertical black and white bars 5 cm in width. The board was placed vertically within the forest 4 m away from the observer, and the number of bars covered by vegetation was counted. The proportion of bars covered by vegetation was used for subsequent analyses. Additionally, we measured the distance from a sampling point to the nearest tree > 10 cm in diameter at breast height (i.e. closest-individual method according to Cottam & Curtis, 1956) to determine the density of large trees. Both measurements were repeated 25 times at each site, with ~40 m between sampling points. Ecomorphological data We sampled lizards from August to December 2006 by walking along trails within the forest of each site, from 07:30 to 17:00 h. We searched for lizards by examining all vegetation and using the visual encounter surveys technique (Crump & Scott, 1994). Each individual captured was toe clipped for future identification, and the sex/age category (adult male, adult female and juvenile, with adults measuring ≥ 31 mm, according to Andrews, 1982) was determined. The following morphological data were recorded with a metallic ruler (to the nearest 0.5 mm): snout–vent length (SVL); lengths of the humerus, radius and hand (from the wrist to the tip of the fourth finger); the total length of the right fore leg; lengths of the femur, tibia and foot (from the heel to the tip of the fourth toe); the total length of the right hind leg; and tail length (only if undamaged). Also, we recorded the structural habitat (i.e. perch height and perch diameter) using a measuring tape (to the nearest 1 mm). For some analyses, perch height was divided into three discrete categories: 0–100, 101–200 and > 200 cm, and perch diameter was divided into four categories: 0–3, 3.1–6, 6.1–9 and > 9 cm. Focal observations Between nine and 11 individuals were observed at each site to determine the displacement and habitat use over a long period of time and to detect interactions with congeners. Random individuals were chosen to be used as focal animals, and observations were made for up to 90 min or until the subject was out of sight. During the period of observation, the mode of locomotion (run, jump, walk) of each motion bout was recorded, and perches used for ≥ 2 min were marked with flagging tape. At the end of the observation period, the focal animal was captured and the following data were recorded: SVL, sex/age category, height and diameter of the perches marked. Also, we used a scoring index of aggressiveness (from less to more aggressive behaviours) similar to that described by Hess & Losos (1991) as follows: escape when rival displayed aggressively (−2), escape when rival approached (−1), headbobs (1), dewlap extension (2) and mouth opening (3). The sum of scores of each individual was used for subsequent analysis. Agonistic interactions To control for interactions of A. limifrons with congeners, we carried out an experiment using a portable enclosure as a neutral arena. The enclosure (1.5 m long × 1.5 m wide × 1.5 m high) consisted of wooden poles as a frame and screen mesh as the walls and roof. We placed a piece of a trunk (10 cm in diameter) with branches simulating natural perches in the middle of the enclosure and removed all leaflitter and other vegetation. We staged encounters by introducing an adult male of A. limifrons and an adult male of the most common congener observed at each site (see Results). We carried out nine to 11 trials at each site, and no individual was used more than once. The SVL of both animals was measured before each encounter. For each encounter, the numbers and types of interactions were recorded during a 90 min period. We calculated an index of aggressiveness as explained above. Statistical analysis We analysed variation in the characteristics of the habitat across sites by transforming the two variables measured to fulfil normality requirements. We squared the values of the density of understorey vegetation and obtained the square root for the distance to the nearest tree. We used these transformed variables as dependent variables in separate one-way ANOVAs, with population as the fixed factor. We also performed Bonferroni post hoc tests to examine the specific relationships between sites. We examined morphological variation across populations of A. limifrons. We removed the effect of body size by regressing each morphological variable against SVL, and the residual values were then used for subsequent analyses. Given that all morphological variables were highly correlated at all sites (Pearson correlation coefficients > 0.83, P < 0.001), we performed a principal component analysis (PCA, using varimax as the rotation method) to reduce the number of morphological variables. We excluded tail length from the analysis because of the large number of individuals with damaged or regenerated tails. We used the principal components (PCs) extracted from the analysis (with eigenvalues > 1.0) to compare the morphology of individuals among populations through a multiple analysis of variance (MANOVA), using PCs as response variables and population as the fixed factor, and subsequent analysis of contrasts to identify specific differences among populations. The structural habitat was also compared among populations. For data collected during visual encounter surveys, we performed correspondence analyses to compare separately the categories of perch height and perch diameter among populations, among versants and versants against Caño Island. We also looked for differences in perch height and perch diameter measured during focal observations. We averaged both measurements (perch height and perch diameter) for each individual and then log10-transformed these averages to fulfil normality requirements. We used these log10-transformed variables as input for one-way ANOVAs, with population as a fixed factor. We were interested in identifying relationships between the structural habitat and morphology, and potential differences among populations in their mode of locomotion. We performed a canonical correlation analysis to detect an overall relationship of structural habitat and morphological variables including data from all populations, with perch height and perch diameter as dependent variables and both PCs as independent variables. We repeated this type of analysis for each site and each versant. In addition, we compared the mode of locomotion of A. limifrons across sites using a correspondence analysis. Finally, interspecific agonistic interactions detected during focal observations and the controlled experiment were compared using one-way ANOVA and one-way ANCOVA, respectively. In both cases, we first log10-transformed the sum of scores from the index of aggressiveness obtained for each individual to fulfil normality requirements. For the analysis of data from focal observations, we used this log10-transformed variable as the dependent variable and population as a fixed factor. We also used the log10-transformed sum of scores recorded in the experiment as the dependent variable, population as a fixed factor and the log10-transformed sum of scores of the congener as the covariable. All statistical tests were carried out using SPSS version 15.0.1 (IBM©). RESULTS Habitat characteristics Differences between sites were detected for the density of understorey vegetation (F4,120 = 3.71, P = 0.007) but not for the distance to the nearest tree (F4,119 = 1.86, P = 0.123). Bonferroni post hoc comparisons demonstrated that differences in understorey vegetation were observed only between Golfito and Caño Island (Fig. 2). Figure 2. View largeDownload slide Mean ± SEM of the relative density of understorey (white bars) and distance (in centimetres) to the nearest tree > 10 cm in diameter at breast height (grey bars) at each study site. Figure 2. View largeDownload slide Mean ± SEM of the relative density of understorey (white bars) and distance (in centimetres) to the nearest tree > 10 cm in diameter at breast height (grey bars) at each study site. Ecomorphological variation A total of 555 A. limifrons were measured at all sites (Table 1). The first two PCs explained 78.15% of the morphological variation observed, with eigenvalues of 4.41 and 1.84, respectively. Principal component 1 had high loadings for measurements of the hind leg, and PC2 for measurements of the fore leg (see Supporting Information, Table S1). Differences among populations in the overall morphology were detected (Table 2). Analysis of contrasts demonstrated that both populations from the Caribbean versant did not differ in morphology, nor did both populations from the Pacific versant (Fig. 3). However, there were differences when comparing Caribbean vs. Pacific and each versant vs. Caño Island (Table 2, Fig. 3). In general, individuals from Caribbean populations had relatively shorter fore legs and relatively longer hind legs than those from Pacific populations, while individuals at Caño Island presented intermediate measurements (see Supporting Information, Table S2, Fig. S1). Table 1. Number of adult males, adult females and juveniles of Anolis limifrons and the most common congener captured at each study site Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  View Large Table 1. Number of adult males, adult females and juveniles of Anolis limifrons and the most common congener captured at each study site Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  Species by site  Sex/age category  Total  Male  Female  Juvenile  Pacuare  A. limifrons  64  42  40  146  A. humilis  10  10  5  25  La Suerte  A. limifrons  62  40  49  151  A. humilis  14  8  6  28  Carara  A. limifrons  34  18  20  72  A. cupreus  4  3  23  30  Golfito  A. limifrons  53  28  15  96  A. polylepis  31  9  7  47  Caño Island  A. limifrons  31  35  24  90  Total A. limifrons  244  163  148  555  Total all species  311  197  195  703  View Large Table 2. MANOVA and analysis of contrasts comparing principal components extracted from morphological variables measured for five populations of Anolis limifrons Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  View Large Table 2. MANOVA and analysis of contrasts comparing principal components extracted from morphological variables measured for five populations of Anolis limifrons Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  Test  Wilks’s λ  F  d.f.  P-value  All sites  0.42  73.69  8, 1098  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  0.10  0.13  2, 294  0.878  Pacific sites (Carara vs. Golfito)  0.98  2.08  2, 165  0.128  Caribbean vs. Pacific  0.42  322.51  2, 462  < 0.001  Caribbean vs. Caño Island  0.62  118.53  2, 384  < 0.001  Pacific vs. Caño Island  0.85  22.90  2, 255  < 0.001  View Large Figure 3. View largeDownload slide Morphological differences of Anolis limifrons populations based on principal component (PC) scores (PC1, hind leg measurements; PC2, foreleg measurements) with 95% confidence ellipses. Squares, Pacific versant; circles, Caribbean versant. Figure 3. View largeDownload slide Morphological differences of Anolis limifrons populations based on principal component (PC) scores (PC1, hind leg measurements; PC2, foreleg measurements) with 95% confidence ellipses. Squares, Pacific versant; circles, Caribbean versant. Similar results were obtained when perch height and perch diameter were compared across populations and versants (Table 3). We found that A. limifrons used mostly low perches (0–100 cm) in the Caribbean and high perches (> 200 cm) in the Pacific (Fig. 4A). At all sites, lizards preferred thin perches (0–3 cm), although Caribbean populations used thicker perches (> 6 cm) more often than those from the Pacific (Fig. 4B). Lizards from Caño Island used a wider range of perch heights and perch diameters compared with mainland sites (Fig. 4; see also Supporting Information, Table S2, Fig. S2). These results were confirmed, in part, through observations of focal subjects within each population, because differences among populations were detected for perch height (F4,44 = 9.80, P < 0.001) but not for perch diameter (F4,43 = 1.67, P = 0.174). Table 3. Results of χ2 tests obtained using categories of perch height and perch diameter measured for five populations of Anolis limifrons Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  View Large Table 3. Results of χ2 tests obtained using categories of perch height and perch diameter measured for five populations of Anolis limifrons Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  Test  Perch height  Perch diameter  χ2  d.f.  P-value  χ2  d.f.  P-value  All sites  117.25  8  < 0.001  52.57  12  < 0.001  Caribbean sites (Pacuare vs. La Suerte)  4.54  2  0.103  4.82  3  0.185  Pacific sites (Carara vs. Golfito)  2.56  2  0.278  2.23  3  0.525  Caribbean vs. Pacific  112.48  2  < 0.001  31.49  3  < 0.001  Caribbean vs. Caño Island  20.08  2  < 0.001  14.01  3  0.003  Pacific vs. Caño Island  35.46  2  < 0.001  33.07  3  < 0.001  View Large Figure 4. View largeDownload slide Classification of Anolis limifrons populations according to perch height (A) and perch diameter (B) categories. Dimension 1 explains 96.08% (eigenvalue = 0.206) in A and 67.27% (eigenvalue = 0.066) in B of the χ2 value, and dimension 2 explains 3.92% (eigenvalue = 0.008) and 31.57% (eigenvalue = 0.031), respectively. Figure 4. View largeDownload slide Classification of Anolis limifrons populations according to perch height (A) and perch diameter (B) categories. Dimension 1 explains 96.08% (eigenvalue = 0.206) in A and 67.27% (eigenvalue = 0.066) in B of the χ2 value, and dimension 2 explains 3.92% (eigenvalue = 0.008) and 31.57% (eigenvalue = 0.031), respectively. Relationship between morphology, habitat use and locomotion A significant association (R = 0.34) between morphology and structural habitat (Wilks’s λ = 0.89, F4,1070 = 16.39, P < 0.001) for all populations of A. limifrons was found. Principal component 1 (hind leg measurements) and perch height accounted for most of the observed variation. The overall pattern found shows that individuals from all populations with larger values of hind leg measurements tended to perch lower (Supporting Information, Fig. S3). However, when considering geographical regions separately (i.e. Caribbean versant, Pacific versant and Caño Island) no significant correlation was found (Caribbean, R = 0.16, F4,566 = 1.98, P = 0.097; Pacific, R = 0.22, F4,318 = 1.92, P = 0.107; Caño Island, F4,170 = 1.97, P = 0.10). We also found significant differences in the mode of locomotion between sites (χ2 = 54.43, d.f. = 8, P < 0.001), with lizards from Caño Island running more, those from both Caribbean sites jumping more and those from both Pacific sites walking more often than individuals from other regions (Fig. 5). Figure 5. View largeDownload slide Classification of Anolis limifrons populations according to the mode of locomotion of individuals. Dimension 1 explains 94.63% (eigenvalue = 0.045), and dimension 2 explains 5.37% (eigenvalue = 0.003) of the χ2 value. Figure 5. View largeDownload slide Classification of Anolis limifrons populations according to the mode of locomotion of individuals. Dimension 1 explains 94.63% (eigenvalue = 0.045), and dimension 2 explains 5.37% (eigenvalue = 0.003) of the χ2 value. Agonistic interactions No differences in aggressiveness were detected between populations during observation of focal subjects (F3,36 = 1.24, P = 0.311). However, by controlling the interactions within an enclosure, we did find significant differences between populations (F3,35 = 6.06, P = 0.002). Overall, lizards from Caribbean populations were less aggressive towards but received more displays from the congener, whereas in the Pacific the opposite pattern was observed (Fig. 6). Figure 6. View largeDownload slide Sum of scores of the index of aggressiveness recorded during a manipulative experiment for male–male encounters of Anolis limifrons and the most common congener (Anolis humilis for La Suerte and Pacuare, Anolis cupreus for Carara, and Anolis polylepis for Golfito) found at each study site. Figure 6. View largeDownload slide Sum of scores of the index of aggressiveness recorded during a manipulative experiment for male–male encounters of Anolis limifrons and the most common congener (Anolis humilis for La Suerte and Pacuare, Anolis cupreus for Carara, and Anolis polylepis for Golfito) found at each study site. DISCUSSION We were interested to determine whether ecomorphological variation occurs along the distribution of a lizard and, if so, whether this variation is explained by a functional hypothesis of habitat use or by the influence of congeners affecting ecological preferences of the species. We found significant differences in morphology, structural habitat and mode of locomotion among populations of A. limifrons, and a significant but weak relationship between morphology and perch height. Interpopulation variation in ecomorphological patterns Morphological traits of A. limifrons differed among populations from Caño Island, the Pacific side and the Caribbean side of Costa Rica. These differences can be explained in two ways. First, they might reflect genetic differentiation. Costa Rican mountain ranges divide the territory into two clearly defined versants (Fig. 1), creating a geographical barrier for a variety of organisms (Savage, 2002). This barrier could have interrupted the gene flow for A. limifrons, so that populations from both versants could have been relatively isolated and different phenotypes could have adjusted to different conditions. Pacific populations of this species from Costa Rica and Panama have previously been referred to as Anolis biscutiger (Taylor, 1956; Echelle, Echelle & Fitch, 1971; Fitch, 1973), although most researchers do not accept this species as valid (Savage & Bolaños, 2009). Although genetic differences among populations remain to be tested, it seems unlikely that these differences, if present, represent species differentiation. Morphological variation found here is not high enough to account for different species, and other researchers have been unable to find major differences between populations of both versants (Savage, 2002). In addition, it is unlikely that gene flow has been interrupted for this species, because the altitudinal gradient that A. limifrons occupies (0–1340 m a.s.l.) could allow individuals to cross versants (Fig. 1). A second possible explanation is that morphological variation could be the result of adaptive phenotypic plasticity. Natural selection would favour phenotypes that perform better in habitats with particular conditions, accentuating morphological differences between populations facing different selective pressures (Losos et al., 2000). It is unknown whether A. limifrons exhibits high phenotypic plasticity, but this characteristic has been reported previously for other anoline species (Losos et al., 2000) and could be a widespread feature in the genus Anolis (Kolbe & Losos, 2005). Variation in structural habitat use was also found among populations from the geographical regions (low perches in the Caribbean, high perches in the Pacific and wide range of perch heights on Caño Island). Variation in morphology and habitat use is also linked with variation in locomotor behaviour. Lizards from each geographical region exhibited a distinct mode of locomotion (Caribbean populations prefer to jump, Pacific populations walk more, and on Caño Island individuals run more often). Locomotor behaviour is an expression of morphological attributes and the characteristics of the vegetation available to the individuals (Losos, 2009; Harrison et al., 2015). Thus, our study confirms that populations with long hind legs and short fore legs and using dense, low perches jump more frequently (Caribbean). However, the predictions that populations with short legs and using narrow perches walk more often (Pacific), and populations with long legs and using broad perches tend to run (Caño Island), are partly confirmed. This ecomorphological variation is thus not fully explained by the functional hypothesis proposed for the West Indies species. Factors explaining ecomorphological patterns Different selective pressures could be acting on populations of these regions. Anoline lizards can use a particular structural habitat in order to maximize the visibility of their territory (Scott et al., 1976), avoid predators (Schoener, Spiller & Losos, 2002) or reduce interspecific competition (Pacala & Roughgarden, 1985; Losos, Marks & Schoener, 1993). Therefore, forest structure, predation intensity and the frequency of interspecific agonistic interactions could be major selective forces shaping habitat use by A. limifrons. We did not measure predation intensity in this study, so we do not know its potential variation across sites, and further research is needed. The density of the understorey vegetation and the distance to the nearest tree, in contrast, were similar among almost all study sites (only Golfito and Caño Island differed in their values of density of the understorey vegetation), such that it seems unlikely that forest structure influences the variation in structural habitat use. We did detect differences in the number and type of interspecific agonistic interactions across populations. Consequently, the presence of congeners and their interspecific agonistic interactions could be impacting the structural habitat use of A. limifrons, a phenomenon that has been referred to as agonistic character displacement (Grether et al., 2009; Dufour, Herrel & Losos, 2018). In this case, Caribbean populations are adapted to use low perches, because A. limifrons is not affected by the presence of other congeners (Talbot, 1979). However, Pacific populations have shifted their preferences to use higher perches, because interactions with larger congeners (e.g. Anolis polylepis and Anolis cupreus, both with maximal SVLs of 57 mm; Savage, 2002), already using low perches, can be costly. Another source of support for this conclusion comes from the comparison between the mainland and Caño Island. On the island, A. limifrons is the only anoline lizard present, and evidence from the present study demonstrates that the species experienced an ecological release, with individuals occupying different parts of the habitat. Interspecific agonistic interactions are common between anoles (Ortiz & Jenssen, 1982; Hess & Losos, 1991) and other lizard species (Korner, Whiting & Ferguson, 2000) and can be a factor shaping the direction of selection. Ecomorphological relationships The ecomorphological relationships found in the present study seem to confirm information previously reported on mainland anoline species (Irschick et al., 1997). The morphological characteristics of anoles from Greater Antilles are correlated mainly with perch diameter, whereas the correlation is stronger with perch height for mainland species. This poses a problem, because no satisfactory explanation exists for such a correlation. One functional hypothesis has been proposed to explain the positive relationship of perch height and the number of lamellae found for mainland anoles, referring to a greater capacity to cling to the substrate of species that use higher perches (Irschick et al., 1997). However, this hypothesis does not explain the positive correlation found here between hind leg measurements and perch height. Even more intriguing is the fact that no apparent relationship was found when considering geographical regions of A. limifrons separately. An ecomorphological relationship was found (i.e. hind leg measurements and perch height) when all populations were pooled and the species was treated as a whole, but this relationship became non-significant when each geographical region was analysed independently. This finding has two important consequences. First, it calls attention to ecomorphological studies comparing interspecific patterns, because the ecomorphological data should take into account population differences of each species considered (Losos, 2009). Second, it reflects the finding that the degree of phenotypic variation does not match the degree of ecological variation, with morphological differences among populations being smaller than ecological ones. This means that if no genetic differences are present among populations, the phenotypic plasticity is smaller than the ecological plasticity for A. limifrons. Considerations for evolutionary ecology The extent to which morphology constrains ecological breadth has major implications for the direction of selection and how communities are structured (Strong et al., 1984; Pounds, 1988; Losos et al., 2003). If the selective pressures producing morphological changes persist through time and gene flow is interrupted, then local adaptation can occur and, eventually, speciation. However, if populations have not evolved genetically to adapt to different environments and phenotypic plasticity is present within a species, then selection within a population may not lead to evolutionary change (Losos, 2009). This means that if phenotypically plastic species are capable of occupying different environments with novel conditions, then shifts in ecological preferences (e.g. structural habitat) could be expected without genetic differentiation. Moreover, for some species, such as A. limifrons, the magnitude of ecological shifts among populations appears to be higher than phenotypic changes. Shifts in ecological preferences have been shown for many Anolis lizards (Jenssen, 1973; Rummel & Roughgarden, 1985; Losos et al., 1993, 2004), so that both phenotypic and ecological plasticity could be widespread phenomena within this genus. The major roles that vegetation structure and predation pressure can play in morphological and ecological variation are undeniable. However, other factors can also have a strong influence, such as interspecific agonistic interactions. Therefore, it becomes essential to consider all biotic and abiotic factors shaping the evolutionary pathways of a species. These pathways can vary greatly throughout the distribution of a species, because different selective pressures can be acting at different sites. Species showing both phenotypic and ecological plasticity are more capable of overcoming these costs, so that coexistence with closely related species is possible. However, the intensity of selective pressures varies not only spatially but also temporally, and thus community structure could be a rather dynamic process. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher's web-site: Table S1. Principal component loadings of residual morphological variables (lengths). Tail length is not included because of the high proportion of individuals with damaged tails. Significant loadings are in bold. Table S2. Mean ± SD (minimal and maximal values) of morphological (in millimetres) and structural habitat (in centimetres) variables recorded for Anolis limifrons at each study site. Figure S1. Mean ± SEM of the principal component (PC) scores (PC1, grey bars; PC2, white bars) extracted from morphological variables recorded for Anolis limifrons at each study site. Figure S2. Number of individuals found for each category of perch height (left panels) and perch diameter (right panels) in five populations of Anolis limifrons: Carara (A), Golfito (B), Caño Island (C), La Suerte (D) and Pacuare (E). Figure S3. Relationship of the first principal component (PC1) scores extracted from morphological variables and perch height recorded for five populations of Anolis limifrons. Black regression line corresponds to all data pooled together; coloured regression lines correspond to each population. ACKNOWLEDGEMENTS We are thankful to Gilbert Barrantes, Gerardo Ávalos, Mahmood Sasa and José Manuel Mora for their criticisms and suggestions to improve the manuscript. We are also grateful to the staff from the study sites and to Viviana Arguedas for her support and help in the field. Jonathan Losos, Martin Whiting, Daniel Noble, Dani Chandrasoma, Kerrie Wechmann, Siobhan Dennison and Pau Carazo also contributed valuable comments. We also thank three anonymous reviewers for their helpful comments. This study received donations from The Holly Hill Charitable Trust and Idea Wild and the logistic support of the School of Biology, University of Costa Rica. REFERENCES Andrews RM. 1979. Reproductive effort of female Anolis limifrons (Sauria: Iguanidae). Copeia  1979: 620– 626. Google Scholar CrossRef Search ADS   Andrews RM. 1982. Patterns of growth in reptiles. In: Gans C, Pough HF, eds. Biology of the reptilia. Vol. 13 Physiology D: physiological ecology . New York: Academic Press, 273– 320. Andrews RM. 1991. Population stability of a tropical lizard. Ecology  72: 1204– 1217. Google Scholar CrossRef Search ADS   Andrews RM, Stahl SJ, Nicoletto PF. 1989. Intra-population variation in age of sexual maturity of the tropical lizard Anolis limifrons in Panama. Copeia  1989: 751– 753. Google Scholar CrossRef Search ADS   Ballinger RE, Marion KR, Sexton OJ. 1970. Thermal ecology of the lizard, Anolis limifrons with comparative notes on three additional Panamanian anoles. Ecology  51: 246– 254. Google Scholar CrossRef Search ADS   Boza MA. 1996. Costa Rica, Parques Nacionales . San José, Costa Rica: Incafo. Calderón-Espinosa ML, Ortega-León AM, Zamora-Abrego JG. 2013. Intraspecific variation in body size and shape in an Andean highland anole species, Anolis ventrimaculatus (Squamata: Dactyloidae). Revista de biologia tropical  61: 255– 262. Google Scholar PubMed  Calsbeek R, Cox RM. 2010. Experimentally assessing the relative importance of predation and competition as agents of selection. Nature  465: 613– 616. Google Scholar CrossRef Search ADS PubMed  Chapman BB, Hulthén K, Brönmark C, Nilsson PA, Skov C, Hansson LA, Brodersen J. 2015. Shape up or ship out: migratory behaviour predicts morphology across spatial scale in a freshwater fish. Journal of Animal Ecology  84: 1187– 1193. Google Scholar CrossRef Search ADS PubMed  Cope, ED. 1862. Contributions to Neotropical saurology. Proceedings of the Academy of Natural Sciences of Philadelphia  14: 176– 88. Cottam G, Curtis JT. 1956. The use of distance measures in phytosociological sampling. Ecology  37: 451– 460. Google Scholar CrossRef Search ADS   Crandell KE, Herrel A, Sasa M, Losos JB, Autumn K. 2014. Stick or grip? Co-evolution of adhesive toepads and claws in Anolis lizards. Zoology  117: 363– 369. Google Scholar CrossRef Search ADS PubMed  Crump ML, Scott NJJr. 1994. Visual encounter surveys. In: Heyer WR, Donnelly MA, McDiarmid RW, Hayek L-AC, Foster MS, eds. Measuring and monitoring biological diversity: standard methods for amphibians . Washington: Smithsonian Institution Press, 84– 92. Dufour CMS, Herrel A, Losos JB. 2018. Ecological character displacement between a native and an introduced species: the invasion of Anolis cristatellus in Dominica. Biological Journal of the Linnean Society  123: 43– 54. Google Scholar CrossRef Search ADS   Echelle AA, Echelle AF, Fitch HS. 1971. A comparative analysis of aggressive display in nine species of Costa Rican Anolis. Herpetologica  27: 271– 288. Fitch HS. 1973. A field study of Costa Rican lizards. University of Kansas Science Bulletin  50: 39– 126. Google Scholar CrossRef Search ADS   Garber PA, Rehg JA. 1999. The ecological role of the prehensile tail in white-faced capuchins (Cebus capucinus). American Journal of Physical Anthropology  110: 325– 339. Google Scholar CrossRef Search ADS PubMed  Garland TJr, Losos JB. 1994. Ecological morphology of locomotor performance in squamate reptiles. In: Wainwright PC, Reilly SM, eds. Ecological morphology: integrative organismal biology . Chicago: University of Chicago Press, 240– 302. Goodman BA, Miles DB, Schwarzkopf L. 2008. Life on the rocks: habitat use drives morphological and performance evolution in lizards. Ecology  89: 3462– 3471. Google Scholar CrossRef Search ADS PubMed  Grether GF, Losin N, Anderson CN, Okamoto K. 2009. The role of interspecific interference competition in character displacement and the evolution of competitor recognition. Biological Reviews  84: 617– 635. Google Scholar CrossRef Search ADS PubMed  Harrison AS, Revell LJ, Losos JB. 2015. Correlated evolution of microhabitat, morphology, and behavior in West Indian Anolis lizards: a test of the habitat matrix model. Behaviour  152: 1187– 1207. Google Scholar CrossRef Search ADS   Hess NE, Losos JB. 1991. Interspecific aggression between Anolis cristatellus and A. gundlachi: comparison of sympatric and allopatric populations. Journal of Herpetology  25: 256– 259. Google Scholar CrossRef Search ADS   Irschick DJ, Carlisle E, Elstrott J, Ramos M, Buckley C, Vanhooydonck B, Meyers J, Herrel A. 2005a. A comparison of habitat use, morphology, clinging performance and escape behaviour among two divergent green anole lizard (Anolis carolinensis) populations. Biological Journal of the Linnean Society  85: 223– 234. Google Scholar CrossRef Search ADS   Irschick DJ, Vanhooydonck B, Herrel A, Meyers J. 2005b. Intraspecific correlations among morphology, performance and habitat use within a green anole lizard (Anolis carolinensis) population. Biological Journal of the Linnean Society  85: 211– 221. Google Scholar CrossRef Search ADS   Irschick DJ, Vitt LJ, Zani PA, Losos JB. 1997. A comparison of evolutionary radiations in mainland and Caribbean Anolis lizards. Ecology  78: 2191– 2203. Google Scholar CrossRef Search ADS   Jenssen TA. 1973. Shift in the structural habitat of Anolis opalinus due to congeneric competition. Ecology  54: 865– 869. Google Scholar CrossRef Search ADS   Kolbe JJ, Larson A, Losos JB. 2007. Differential admixture shapes morphological variation among invasive populations of the lizard Anolis sagrei. Molecular Ecology  16: 1579– 1591. Google Scholar CrossRef Search ADS PubMed  Kolbe JJ, Losos JB. 2005. Hind-limb length plasticity in Anolis carolinensis. Journal of Herpetology  39: 674– 678 Google Scholar CrossRef Search ADS   Korner P, Whiting MJ, Ferguson JWH. 2000. Interspecific aggression in flat lizards suggests poor species recognition. African Journal of Herpetology  49: 139– 146. Google Scholar CrossRef Search ADS   Leal M, Rodríguez-Robles JA, Losos JB. 1998. An experimental study of interspecific interactions between two Puerto Rican Anolis lizards. Oecologia  117: 273– 278. Google Scholar CrossRef Search ADS PubMed  Lobo J, Bolaños F, eds. 2005. Historia natural de Golfito . Santo Domingo de Heredia, Costa Rica: Editorial INBio. Logan ML, Montgomery CE, Boback SM, Reed RN, Campbell JA. 2012. Divergence in morphology, but not habitat use, despite low genetic differentiation among insular populations of the lizard Anolis lemurinus in Honduras. Journal of Tropical Ecology  28: 215– 222. Google Scholar CrossRef Search ADS   Losos JB. 1990. Ecomorphology, performance capability, and scaling of West Indian Anolis lizards: an evolutionary analysis. Ecological Monographs  60: 369– 388. Google Scholar CrossRef Search ADS   Losos JB. 2009. Lizards in an evolutionary tree: ecology and adaptive radiation of anoles . Oakland, California: University of California Press. Losos JB, Creer DA, Glossip D, Goellner R, Hampton A, Roberts G, Haskell N, Taylor P, Ettling J. 2000. Evolutionary implications of phenotypic plasticity in the hindlimb of the lizard Anolis sagrei. Evolution  54: 301– 305. Google Scholar PubMed  Losos JB, Leal M, Glor RE, De Queiroz K, Hertz PE, Rodríguez Schettino L, Lara AC, Jackman TR, Larson A. 2003. Niche lability in the evolution of a Caribbean lizard community. Nature  424: 542– 545. Google Scholar CrossRef Search ADS PubMed  Losos JB, Marks JC, Schoener TW. 1993. Habitat use and ecological interactions of an introduced and a native species of Anolis lizard on Grand Cayman, with a review of the outcomes of anole introductions. Oecologia  95: 525– 532. Google Scholar CrossRef Search ADS PubMed  Losos JB, Schoener TW, Spiller DA. 2004. Predator-induced behaviour shifts and natural selection in field-experimental lizard populations. Nature  432: 505– 508. Google Scholar CrossRef Search ADS PubMed  Lostrom S, Evans JP, Grierson PF, Collin SP, Davies PM, Kelley JL. 2015. Linking stream ecology with morphological variability in a native freshwater fish from semi-arid Australia. Ecology and Evolution  5: 3272– 3287. Google Scholar CrossRef Search ADS PubMed  McMillan DM, Irschick DJ. 2010. Experimental test of predation and competition pressures on the Green Anole (Anolis carolinensis) in varying structural habitats. Journal of Herpetology  44: 272– 278. Google Scholar CrossRef Search ADS   Nicholson KE, Glor RE, Kolbe JJ, Larson A, Hedges SB, Losos JB. 2005. Mainland colonization by island lizards. Journal of Biogeography  32: 929– 938. Google Scholar CrossRef Search ADS   Nudds TD. 1977. Quantifying the vegetative structure of wildlife cover. Wildlife Society Bulletin  5: 113– 117. Ord TJ, Klomp DA. 2014. Habitat partitioning and morphological differentiation: the Southeast Asian Draco lizards and Caribbean Anolis lizards compared. Oecologia  175: 651– 666. Google Scholar CrossRef Search ADS PubMed  Ortiz PR, Jenssen TA. 1982. Interspecific aggression between lizard competitors, Anolis cooki and Anolis cristatellus. Zeitschrift für Tierpsychologie  60: 227– 238. Google Scholar CrossRef Search ADS   Pacala SW, Roughgarden J. 1985. Population experiments with the Anolis lizards of St. Maarten and St. Eustatius. Ecology  66: 129– 141. Google Scholar CrossRef Search ADS   Pereira Leitão R, Sánchez-Botero JI, Kasper D, Trivério-Cardoso V, Morais Araújo C, Zuanon J, Pellegrini Caramaschi É. 2015. Microhabitat segregation and fine ecomorphological dissimilarity between two closely phylogenetically related grazer fishes in an Atlantic Forest stream, Brazil. Environmental Biology of Fishes  98: 2009– 2019. Google Scholar CrossRef Search ADS   Pounds JA. 1988. Ecomorphology, locomotion, and microhabitat structure: patterns in a tropical mainland Anolis community. Ecological Monographs  58: 299– 320. Google Scholar CrossRef Search ADS   Roughgarden J. 1995. Anolis lizards of the Caribbean: ecology, evolution and plate tectonics . New York: Oxford University Press. Rummel JD, Roughgarden J. 1985. Effects of reduced perch-height separation on competition between two Anolis lizards. Ecology  66: 430– 444. Google Scholar CrossRef Search ADS   Savage JM. 2002. The amphibians and reptiles of Costa Rica: a herpetofauna between two continents, between two seas . Chicago: University of Chicago Press. Savage JM, Bolaños F. 2009. A checklist of the amphibians and reptiles of Costa Rica: additions and nomenclatural revisions. Zootaxa  2005: 1– 23. Schoener TW, Schoener A. 1980. Densities, sex ratios, and population structure in four species of Bahamian Anolis lizards. Journal of Animal Ecology  49: 19– 53. Google Scholar CrossRef Search ADS   Schoener TW, Spiller DA, Losos JB. 2002. Predation on a common Anolis lizard: can the food-web effects of a devastating predator be reversed? Ecological Monographs  72: 383– 407. Google Scholar CrossRef Search ADS   Schulte JA2nd, Losos JB, Cruz FB, Núñez H. 2004. The relationship between morphology, escape behaviour and microhabitat occupation in the lizard clade Liolaemus (Iguanidae: Tropidurinae: Liolaemini). Journal of Evolutionary Biology  17: 408– 420. Google Scholar CrossRef Search ADS PubMed  Scott NJJr, Wilson DE, Jones C, Andrews RM. 1976. The choice of perch dimensions by lizards of the genus Anolis (Reptilia, Lacertilia, Iguanidae). Journal of Herpetology  10: 75– 84. Google Scholar CrossRef Search ADS   Sexton OJ. 1967. Population changes in a tropical lizard Anolis limifrons on Barro Colorado Island, Panama Canal Zone. Copeia  1967: 219– 222. Google Scholar CrossRef Search ADS   Sexton OJ, Heatwole H. 1968. An experimental investigation of habitat selection and water loss in some anoline lizards. Ecology  49: 762– 767. Google Scholar CrossRef Search ADS   Sexton OJ, Ortleb EP, Hathaway LM, Ballinger RE, Licht P. 1971. Reproductive cycles of three species of anoline lizards from the Isthmus of Panama. Ecology  52: 201– 215. Google Scholar CrossRef Search ADS   Siliceo-Cantero HH, García A, Reynolds RG, Pacheco G, Lister BC. 2016. Dimorphism and divergence in island and mainland Anoles. Biological Journal of the Linnean Society  118: 852– 872. Google Scholar CrossRef Search ADS   da Silva JM, Herrel A, Measey GJ, Vanhooydonck B, Tolley KA. 2014. Linking microhabitat structure, morphology and locomotor performance traits in a recent radiation of dwarf chameleons. Functional Ecology  28: 702– 713. Google Scholar CrossRef Search ADS   Strong DRJr, Simberloff D, Abele LG, Thistle AB, eds. 1984. Ecological communities: conceptual issues and the evidence . Princeton: Princeton University Press. Talbot JJ. 1977. Habitat selection in two tropical anoline lizards. Herpetologica  33: 114– 123. Talbot JJ. 1979. Time budget, niche overlap, inter- and intraspecific aggression in Anolis humilis and A. limifrons from Costa Rica. Copeia  1979: 472– 481. Google Scholar CrossRef Search ADS   Taylor EH. 1956. A review of the lizards of Costa Rica. University of Kansas Science Bulletin  38: 3– 322. Velasco JA, Herrel A. 2007. Ecomorphology of Anolis lizards of the Chocó region in Colombia and comparisons with Greater Antillean ecomorphs. Biological Journal of the Linnean Society  92: 29– 39. Google Scholar CrossRef Search ADS   Villalobos F, Arita HT. 2014. Morphological diversity at different spatial scales in a Neotropical bat assemblage. Oecologia  176: 557– 568. Google Scholar CrossRef Search ADS PubMed  Vitt LJ, Avila-Pires TCS, Zani PA, Espósito MC. 2002. Life in shade: the ecology of Anolis trachyderma (Squamata: Polychrotidae) in Amazonian Ecuador and Brazil, with comparisons to ecologically similar anoles. Copeia  2002: 275– 286. Google Scholar CrossRef Search ADS   Wainwright PC, Reilly SM, eds. 1994. Ecological morphology: integrative organismal biology . Chicago: University of Chicago Press. Williams EE. 1972. The origin of faunas: evolution of lizard congeners in a complex island fauna: a trial analysis. Evolutionary Biology  6: 47– 89. © 2018 The Linnean Society of London, Biological Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Biological Journal of the Linnean SocietyOxford University Press

Published: May 14, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off