DIETARY ANALYSIS OF THREE SPECIES OF THE GENUS ANOLIS (SAURIA: DACTYLOIDAE) IN “LOS TUXTLAS,” VERACRUZ, MEXICO

— Anoles have been studied by researchers to a greater extent than any other group of lizards. Their high diversity has led them to colonize a variety of niches, making them an ideal model group for evaluating ecological hypotheses such as dietary niche overlap. This work analyzes the stomach contents of 73 individuals from three species of the genus Anolis: A. barkeri (34), A. sericeus (17), and A. tropidonotus (22) occurring in the vicinity of Los Tuxtlas, Veracruz. Analyses performed included Shannon’s index in its log form to calculate dietary diversity, the Jaccard index to estimate the dissimilarity of the species’ diets, and Schoener’s index to measure dietary overlap. The results suggest that A. barkeri (10.08) hast the most generalist diet, followed by A. sericeus (8.75) and A. tropidonotus (5.78). Schoener’s index showed a considerable amount of diet overlap between A. barkeri and A. sericeus (0.76). We conclude that the three focal species show a generalist feeding behavior in times of abundant prey, such as the rainy season in which this study was conducted. This may lead to a reduction of interspecific competition explaining why we did not observe dietary niche displacement between these three sympatric species of Anolis.


INTRODUCTION
The genus Anolis is a group of lizards that has been used extensively in ecological and evolutionary studies given their large diversity (Losos, 2009). The genus contains approximately 400 species (Poe et al., 2017), with a distribution extending from southeastern United States, through Mexico, and into parts of South America, as well as the Caribbean islands (Losos, 2009). The majority of ecological and evolutionary studies on Anolis have focused on the island populations, for which the interspecific interactions and trophic levels have had important evolutionary consequences (Losos, 2009). Similarly, data on the dietary composition of island Anolis species have been widely reported in the literature (Wolcott, 1923;Schoener et al., 1968;Lister, 1981;Reagan, 1986;Dial & Roughgarden, 2004;Herrel et al., 2004), while the diets of mainland species have been less studied (Andrews, 1971;Sexton, 1972;Fitch, 1972Fitch, , 1975Lister & Garcia-Aguayo, 1992;Seliceo-Cantero & Garcia, 2015). Sympatric Anolis species are thought to exhibit different diets in order to partition resources when living in sympatry, leading to trophic niche partitioning (Losos, 2009). However, few studies have tested this hypothesis in mainland Anolis species.
This work presents the results of a dietary analysis of three sympatric species of Anolis found on the biosphere reserve "Los Tuxtlas" in Veracruz, Mexico. The focal species include A. barkeri, which occupies a semiaquatic habitat on rocks and debris near streams and rivers (Birt et al., 2001); A. sericeus, which tends to be found in trees and bushes of more arid areas (Henderson & Fitch, 1975); and A. tropidonotus, a strictly terrestrial lizard commonly found in the leaf litter (Henderson & Fitch, 1975).

MATERIAL AND METHODS
Organisms for this study were collected by RCV and OFV, at the Tropical Biological Station "Los Tuxtlas", UNAM during the rainy season (May to November). The stomach contents of 73 individuals from the three focal species were analyzed: A. barkeri (34), A. sericeus (17), and A. tropidonotus (22). Stomach items were identified to order-level, and the total volume of stomach contents, as well as each individual item, was calculated using volumetric displacement in a test tube. The frequency of the presence of prey items was calculated per Anolis species, effectively homogenizing the diet for each species and avoiding count bias for parts of prey in an individual's stomach.
The resulting frequency table was inputted into the "vegan" package (Oksanen et al., 2015) in R (R Core Team, 2013) to calculate Shannon's Index in its log form (effective numbers) in order to compare the dietary diversity between each species (Moreno et al., 2010). Furthermore, Jaccard's dissimilarity index (1 -Jaccard's Index) was calculated to understand which species are least similar in respect to the incidences of prey that they consumed. Finally, values of trophic niche overlap between species were calculated using Schoener's Index implemented in the package "spaa" (Zhang, 2013).

RESULTS
A total of 14 distinct prey items were found in the stomach contents of the observed lizards. The results show that A. barkeri had the most number of food items in their diet, followed by A. sericeus, and lastly A. tropidonotus (Table 1). The effective number of species (1D) is shown in Table 2. Anolis barkeri was found to have the most diverse diet, followed by A. sericeus, and lastly A. tropidonotus ( Table 2).
The results also show that the largest trophic niche overlap occurs between the pair A. barkeri and A. sericeus, followed by the pair A. barkeri and A. tropidonotus. The least amount of trophic niche overlap was found between A. sericeus and A. tropidonotus (Table 3).
Lastly, the results of the Jaccard's Index show that A. sericeus and A. tropidonotus had the most similar diet, followed by A. sericeus and A. barkeri. The least similar diet was found between A. barkeri and A. tropidonotus (Table 4).

DISCUSSION
The results of this study support the findings from Losos et al. (2009) in relation to the generalist dietary components observed in Anolis. These lizards tend to adopt a generalist diet when a prey item is found abundantly and can be easily captured. Regarding the three Anolis species studied in this work, arachnids were the most abundant prey item.
Our data revealed that A. barkeri is the most generalist of the three species, corroborating the findings of Birt et al. (2001). Second in prey diversity was A. sericeus, nested between the generalist A. barkeri and the specialist A. tropidonotus. These results suggest that the semiaquatic habitats of A. barkeri provide a larger diversity of foraging grounds and access to a variety of different prey (Birt et al., 2001). When Schoener's values of overlap were compared, A. barkeri and A. sericeus were found to have the highest value (0.76), indicating that although A. barkeri has the most expansive diet, the species shows preference for the same terrestrial prey as A. sericeus, and not for prey found in aquatic environments. This could be an artifact of the season in which the lizards were collected (rainy season, between May and November), given that some tropical invertebrates reach their highest abundance during the rainy season (Didham & Springate, 2003;Wolda, 1989;Santana et al., 2015). Although their diets overlapped, partitioning of dietary components between these two species was not observed, likely due to the high prevalence of prey items.
The results of the dietary components dissimilarity analysis (Jaccard Index) suggest that these three sympatric species of Anolis are not competing for food resources, as suggested by Losos (2009). It is likely that an abundant availability of prey items is related with a high diet similarity between species. Even though they display some similarities, it seems that variation in existing habitat and the opportunistic condition of feeding behavior lead to the avoidance of competition for food between these three species.
The results should be considered as preliminary given the small sample size and the limited seasonality covered. Several studies have found that the diet of Anolis species varies intraand interspecifically between different habitats of the same locality (i.e. A. stratulus Reagan, 1986;A. conspersus Schoener, 1967;A. sagrei, A. distichus, A. carolinensis, and A. angusticeps Schoener, 1968;A. opalicnus Floyd & Jensen, 1983;and A. aeneus Stamps et al., 1997). Another limiting aspect of this study was the classification of prey items to their order level, which may introduce bias into the analyses, particularly in the conclusions of dietary similarities (Greene & Jaksić, 1983). Further studies should include a larger sample size, sampling throughout a broader temporal scale, higher taxonomic resolution of dietary components, and an evaluation of available prey at each site.

CITED LITERATURE
Andrews, R.M. 1971. Structural habitat and time budget of a tropical   Tabla 4. Índice de incidencia de Jaccard entre las tres especies de Anolis analizadas.