Clutch size in populations and species of cnemidophorines (Squamata: Teiidae) on the eastern coast of Brazil

We analyzed some reproductive aspects of 16 coastal populations, belonging to five lizard species (A. ocellifera, A. abaetensis, A. nativo, A. littoralis and C. lacertoides) from different restinga habitats along the eastern coast of Brazil. This study aimed to evaluate to what extent the reproductive aspects vary geographically and among species. For each female, we recorded the number of vitellogenic follicles, size and color of the largest follicle, presence and size of corpora lutea, and number and size of oviductal eggs. Clutch size of almost all coastal populations/species of Ameivula had little variation and most clutches were composed of two eggs. There was a significant relationship between female size and the mean clutch size when females from different species were pooled. Mean egg volume, among species, varied from 420 to 655 mm. Relative clutch mass varied from 0.129 to 0.159 and did not differ significantly among species. We concluded that the five coastal species studied (four bisexuals and one parthenogenetic) had similar reproductive characteristics. Most of them presented multiple clutches, low clutch size and low relative clutch mass, similar to other species in the genus and to unisexual and bisexual species of the Teiidae family.


INTRODUCTION
Differences in lizard reproduction traits such as clutch size, growth rates and adult sizes occur among species and populations within a species (e.g.Kiefer et al. 2008, Galdino andVan Sluys 2011).Such variations can occur in response to different local environmental conditions (e.g.Shine andElphick 2001, Brandt andNavas 2011), genetic differences among populations/species (e.g.Niewiarowski 1994) and also from evolutionary processes (e.g.Shine and Geer 1991).In Brazil, there are still few studies that focus on the reproductive traits among different populations within a lizard species (e.g.Kiefer et al. 2008, Galdino andVan Sluys 2011).This lack of knowledge limits our understanding of how various attributes influence the reproduction traits of the populations within a species along its geographical range.So, the changing profiles on the reproductive traits of lizards and the effects of the environment on them need to be better understood.
The reproductive patterns of some cnemidophorines species in tropical areas have been an object of interest in the last two decades, but this knowledge is still limited for many species http://dx.doi.org/10.1590/0001-37652014112212707-722 VANDERLAINE A. MENEZES and CARLOS F.D. ROCHA and populations (e.g.Vitt et al. 1997, Mesquita and Colli 2003a, Menezes et al. 2004, Balestrin et al. 2010, Santana et al. 2010).In South America, parthenogenesis in cnemidophorines is known for the Cnemidophorus lemniscatus complex in the Amazon Forest (Vanzolini 1970, Cole andDessauer 1993) and for Ameivula nativo in eastern Brazil (Rocha et al. 1997, Menezes et al. 2004).Ameivula nativo was the first example of a parthenogenetic species of cnemidophorines found south of the Amazon rainforest (Rocha et al. 1997(Rocha et al. , 1999)).The analysis of reproductive aspects in these cnemidophorines species may increase the understanding of how they interact with the environment in which they live in and allow for comparisons of ecological and physiological aspects of unisexual and bisexual species.
The clutch size is variable in cnemidophorines, with records of species with clutches of a single egg (as in C. arubensis -Vitt andBreitenbach 1993 andC. jalapensis -Colli et al. 2009) up to broods with several eggs (2 to 5 eggs in C. lemniscatus -Léon and Cova 1973).Different patterns of the ecology and life history of lizards can affect the clutch size of a given species.The body size of females is among the most obvious factors that can affect the size of clutch or the brood produced (e.g.Vrcibradic andRocha 2011, Winck andRocha 2012).Selective forces may either favor a larger clutch size, where each egg has a relatively small volume or, alternatively, may promote the production of a smaller clutch size, in which each egg has a relatively large volume (Lack 1954 apud Sinervo et al. 1992, Kiefer et al. 2008).
Along the areas of restinga in the eastern coast of Brazil four species of Ameivula and one of Contomastix are known to occur: C. lacertoides (in southern Brazil), A. littoralis (endemic to the state of Rio de Janeiro), A. nativo (in the states of Espírito Santo and Bahia), A. abaetensis (in the states of Bahia and Sergipe) and A. ocellifera (widely distributed in northeastern Brazil) (Menezes and Rocha 2013).Although many studies have contributed to increase the knowledge on reproduction of lizards of these genera (e.g.Léon and Cova 1973, Schall 1983, Vitt 1983, Magnusson 1987, Vitt et al. 1997, Mesquita and Colli 2003a, b, Mojica et al. 2003), the information on reproductive aspects of their populations along the coast of Brazil is still very limited (e.g.Menezes et al. 2004), and sufficient data is not available for the comparison of populations and species.The analysis of how reproductive parameters vary among species and among populations of the same species can allow us to understand adaptive responses in the natural history of species and any differences between bisexual and unisexual species within these parameters.
In this study we analyzed some reproductive aspects of 16 coastal populations belonging to the five aforementioned species of the genus Ameivula and Contomastix (Harvey et al. 2012) from different restinga habitats along approximately 3000 km of extension of the eastern coast of Brazil.The study's goals were (1) to evaluate to what extent the reproductive aspects (mean clutch size produced by females, minimum size at maturity for females and males) vary geographically and among species; (2) to determine the mean volume of eggs and the relative clutch mass (clutch mass/female mass) of females of cnemidophorines in each locality; (3) to evaluate the extent to which body size influences female clutch and egg size in each population/ species; (4) to determine if there are differences between bisexual and unisexual species in the relationship between body size and clutch size; and (5) to determine if there is sexual dimorphism in body size for lizards in the studied populations.

STUDY AREAS
Fieldwork was carried out in 15 areas of restinga habitats from Genipabu, in the state of Rio Grande do Norte, northeastern Brazil, to Florianópolis, in the state of Santa Catarina, southern Brazil along approximately 3000 km of coast (Fig. 1, 2).REPRODUCTIVE TRAITS OF CNEMIDOPHORINE SPECIES  Restingas are sandy strips located between the sea and the mountains of the Brazilian eastern seaboard and are part of the Atlantic Forest biome.This type of habitat originated in the Quaternary, resulting from successive marine regressions which occurred throughout the Holocene and Pleistocene periods (Muehe 1983, Perrin 1984, Suguio and Tessler 1984).

SAMPLING METHODS AND DATA ANALYSES
Collections were all done during the rainy season (October-May), within three years (2004)(2005)(2006), depending on the area), in order to reduce the possible influence of weather and seasonality on reproductive patterns.We followed the procedures of the Society for the Study of Amphibians and Reptiles outlined in the Guidelines for Use of Life of Amphibians and Reptiles in Field Research, which recommend the use of anesthetics prior to euthanizing the animals.Lizards were collected with rubber bands or pellet rifles, euthanized with ether and immediately fixed in 10 % formalin.We measured the snout-vent length (SVL) of the lizards using a Vernier caliper (to the nearest 0.1 mm) and weighed them using a Pesola spring balance (to the nearest 0.001g) or a dynamometer (to the nearest 0.5 g).
In the laboratory, we dissected individuals for gonadal analysis.For each female, we recorded the number of vitellogenic follicles in each ovary, presence of corpora lutea, and number and size of oviductal eggs.Follicles were considered vittelogenic when they were yellow and > 1.4 mm in diameter.We measured the length and width of each egg and estimated its volume using the formula for the ellipsoid: Females with vitellogenic follicles, oviductal eggs and/or corpora lutea were considered reproductive.
For males, the length and the width of each testicle were measured, and the volume of each testicle was estimated using the formula for the ellipsoid.To determine if males were sexually mature, the left testicle and epididymis of each male were dehydrated, subjected to histological sections of 5 µm and stained with hematoxylineosin.The presence of spermatozoa in seminiferous tubules and/or epididymis was used to differentiate between adults and juveniles.
Mean clutch size was estimated based on the number of oviductal eggs or vitellogenic follicles.The simultaneous occurrence of vitellogenic follicles and oviductal eggs or of vitellogenic follicles and corpora lutea was considered as evidence of the production of at least two clutches per female per reproductive season (multiple clutches).
Relative clutch mass (RCM) was calculated as clutch mass divided by total mass (body mass + clutch mass).The effect of female body size (SVL) on clutch size for each population and on clutch size, egg volume and relative clutch mass (arcsine-transformed) for each species (pooling data from different populations for each species) was estimated using simple regression analysis or, when values were not normally distributed, Spearman Correlation (Zar 1999).
The differences in mean clutch size, egg volume and relative clutch mass among different populations of a species and among females of different species were tested by analysis of covariance (ANCOVA, using female body size as the covariate) followed by Scheffe tests for multiple comparisons (Zar 1999).
The differences in adult male and female SVL among populations/species were also evaluated using ANOVA and post-hoc Scheffe test to determine what pairs of populations/species differed in their coastal distribution (Zar 1999).An ANOVA and post-hoc Scheffe test were also used to compare minimum size at maturity (MSM) values among lizard populations/species.The relationships between female MSM and mean SVL and clutch size were tested for the species studied using simple regression analysis (Zar 1999).REPRODUCTIVE TRAITS OF CNEMIDOPHORINE SPECIES All data were tested for homocedasticity of variances and for normality of distributions before performing statistical analyses.Due to the wide variation in SVL and clutch size of some populations, these variables were log-transformed.

INTRASPECIFIC COMPARISONS
Among populations within each species, mean adult female SVL differed significantly for A. nativo (ANOVA, F 4,132 = 5.298, r 2 = 0.138, P < 0.05) and for A. ocellifera (ANOVA, F 3,63 = 3.469, r 2 = 0.142, P < 0.05).Among A. nativo populations, females from Setiba (ES) were significantly larger than those of Guriri (ES) (Post Hoc Scheffe, P < 0.05) and Guaratiba (BA) (Post Hoc Scheffe, P < 0.05) (Fig. 3).Because of the small sample of adult females (n = 3), the population of Prado was not included in this analysis.Among A. ocellifera populations, the Post Hoc Scheffe did not indicate significant differences in female mean SVL, although the probability value for females from Barra dos Coqueiros (SE) and Guarajuba (BA) was near significance (P = 0.076).As a result of the small sample size of adult females (n = 3), the population of Genipabu (RN) was not included in this analysis.
The minimum size at maturity (MSM) was similar among populations of a given species (Fig. 3 and 4).The mean clutch size differed significantly among populations of A. littoralis (ANCOVA, F 2,34 = 3.860, r 2 = 0.187, P < 0.05), with Grussaí (RJ) and Maricá (RJ) being the populations that differed from each other (Post Hoc Scheffe, P < 0,05), and for C. nativo (ANCOVA, F 4,76 = 2.752, r 2 = 0.169, P < 0.05) (Fig. 3).Post Hoc Scheffe did not indicate differences in CS among populations of C. nativo, although the difference in CS between the populations of Comboios (ES) and Maraú (BA) was close to significance (Post Hoc Scheffe, P = 0.077).The population of Prado (BA) was not included in this analysis due to the small sample size of adult females (n = 3).
Except for the populations of A. nativo from Guaratiba (Regression Analysis, F 1,35 = 6.528, r 2 = 0.157, P < 0.05) and of A. abaetensis from Guarajuba (Regression Analysis, F 1,3 = 13.414,r 2 = 0.817, P < 0.05), there was no relationship between clutch size and female SVL for any of the populations studied (all P values > 0.05).When data of CS from different populations were pooled for each species, there was a positive and significant effect of female size on clutch size for A. nativo (Regression Analysis, F 1,84 = 4.487, r 2 = 0.051, P < 0.05).
There was no relationship between the minimum size at maturity and mean clutch size for the different populations/species of Ameivula studied (pooled data from all populations/species) (Regression Analysis, F 1,12 = 1.322,P = 0.273).
There was no significant differences in mean SVL between adult males and females for the set of species studied, except for A. littoralis (pooled data from three populations, F 1,95 = 5.389, r 2 = 0.054, P < 0.05), in which males were larger than females.

DISCUSSION
The reproductive traits of a population/species such as reproductive period, clutch size, egg size, testes volume and minimum size at maturity can be affected by different environmental factors such as temperature (Cruz 1996, Ramirez-Bautista et al. 2000, Rezende-Pinto et al. 2009), photoperiod (Ramirez-Bautista and Vitt 1998, Rezende-Pinto et al. 2009), rainfall (Rodríguez-Ramírez and Lewis 1991, Ramirez-Bautista et al. 2000, Van Sluys et al. 2010) and food availability (Fitch 1982).Occasional intra and interspecific differences in some of those characteristics can also be partially attributed to the differences in body size of the females (Vitt andCongdon 1978, Van Sluys et al. 2010).In the present study, the data indicated some intrapopulational differences in mean clutch size for A. littoralis and for A. nativo, similar to those reported for populations of other cnemidophorines species from different geographical areas, such as C. lemniscatus (e.g.Léon and Cova 1973,  Magnusson 1987, Castro 1994, Mojica et al. 2003) and A. ocellifera (e.g.Vitt 1983, Cruz 1996, Mesquita and Colli 2003b).However, when data were grouped, the mean clutch size was considerably similar (approximately two eggs) for all species, with the exception of A. abaetensis in Guarajuba, BA (χ = 4.0 eggs).The relatively high value of mean clutch size for A. abaetensis found in the present study needs to be further studied, since the sample of reproductive females was relatively small (n = 5) and Dias et al. (2002) in the restinga of Abaeté, in the state of Bahia, recorded for the same species a mean clutch size of 1.4 eggs.
The literature on the species of cnemidophorines in South America shows that the clutch size for species in the genus ranges from 1 to 5 follicles/ eggs, with most being around 2 eggs (Fig. 7).Thus, according to our study, the clutch size of the species of Ameivula and Contomastix in restinga habitats on the east coast of Brazil is similar to that found for most cnemidophorines species studied in South America such as C. cryptus in the Brazilian Amazon (χ = 1.9, Vitt et al. 1997) and C. lemniscatus in Colombia (χ = 2.2, Mojica et al. 2003) (Fig. 7).
The small clutch size in cnemidophorines has been associated with the energetic cost of REPRODUCTIVE TRAITS OF CNEMIDOPHORINE SPECIES RCM tends to be low in active foraging lizards and varies little among individuals or species.
There was no difference in the relative clutch mass among species studied.When compared with other species of cnemidophorines in South America, the values of RCM in this study are within the expected pattern (Fig. 7).When compared with sedentary foraging lizards (χ = 0.228 + 0.009, n = 86, Vitt and Price 1982), the RCM of species studied was comparatively low (χ = 0.140 + 0.044, n = 31, all species together).RCM represents a balance between the advantages of a high investment in reproduction and the costs associated with it in relation to foraging and escape behavior (Vitt and Price 1982), as has been demonstrated for some lizard species whose running speed was impaired by an increase in clutch mass (Shine 1980).
The minimum size at maturity for females was similar among populations of a given specie of coastal Ameivula and Contomastix (A. littoralis: 54.1 to 57.7 mm, A. nativo: 46.9 to 51.1 mm, A. ocellifera: 45.9 to 50.5 mm).Also, among the studied species, female sexual maturity is reached at similar sizes (C.lacertoides: 51.1 mm, A. littoralis: 54.1 mm, A. nativo: 46.9 mm, A. abaetensis: 54.0 mm, A. ocellifera: 45.9 mm).Although the relationship between the minimum size at maturity and clutch size was not significant, there is a perceptible biological tendency in the species/populations studied for those who had larger minimum size at maturity to have slightly larger clutches.
For most species of lizards there is a general trend to increase clutch/litter size with the increase in female body size, both intra and interspecifically (Dunham et al. 1988, King 2000, Kiefer et al. 2008, Vrcibradic and Rocha 2011), and Ameivula and Contomastix spp.are no exception (e.g.Vitt 1983, Vitt et al. 1997, Mesquita and Colli 2003b, Mojica et al. 2003).Body size affected clutch size when females of different species were analyzed together, suggesting that larger females in general tend to produce larger clutches.Although the size of the female had a significant effect on clutch size produced, this effect was relatively small (38%) as a result of the little variation in clutch size both within and between species (1 to 5 eggs).Among all the studied areas, only for the populations of A. nativo in Guaratiba and of A. abaetensis in Guarajuba (both in the state of Bahia) were found a significant effect of the body size on the clutch size of females.When the data for all populations were analyzed together, the effect of the female SVL on the clutch size remained significant for A. nativo, albeit relatively weak (6%).
Sexual dimorphism has been reported in different populations and species of lizards; in some species, larger body size may be favored in females, due to its influence on clutch size and egg volume (e.g.Fitch 1981, Anderson and Vitt 1990, Rocha 2008).However, in other species, females are smaller than males, suggesting that they employ a greater investment of energy in the production of a clutch, thus having a smaller amount of energy available for growth (e.g.Rocha 1992, Van Sluys 1998).Also, in some cases females reach smaller sizes than males as a result of intrasexual selection, thus males obtain advantages with larger body size during agonistic interactions with other males (Fitch 1981, Vitt and Cooper 1985, Rocha 1992).In the present study, in general, we did not observe a clear tendency for sexual dimorphism in SVL for the species studied, with the exception of A. littoralis, where males were significantly larger than females.
We concluded that the five species of cnemidophorines studied (four bisexual and one parthenogenetic) had similar reproductive characteristics.Species presented small clutch size and low relative clutch mass; three of them had multiple clutches per year, similar to other species/ populations of the genus studied previously (e.g.Vitt 1983, Dearing and Schall 1994, Cruz 1996) and to other unisexual and bisexual species of the closely related genus Aspidoscelis (sensu Reeder et al. 2002) (e.g.Schall 1978, Ramírez-Bautista et al. 2000).VANDERLAINE A. MENEZES and CARLOS F.D. ROCHA Long-term studies of populations of these species are needed in order to gain knowledge about the survival of adults, the frequency of clutches per year, the time required for development from birth to sexual maturity, and environmental factors that may cause annual changes in mean clutch size of a given population/species as well as other important information about their reproductive biology.O volume médio dos ovos, entre as espécies, variou 420 a 655 mm3.A massa relativa da ninhada variou de 0,129 a 0,159 e não diferiu significativamente entre as espécies.

Figure 1 -
Figure 1 -Distribution of Ameivula and Contomastix species studied in the restingas along the eastern coast of Brazil.

FFigure 4 -
Figure 4 -Mean snout-vent length (SVL, in mm), minimum size at sexual maturity (MSM, in mm), mean clutch size, relative clutch mass (RCM), volume (in mm 3 ), length and width (in mm) of eggs, and testis volume (in mm 3 ) for five cnemidophorines species studied along the eastern Brazilian coast.Values are represented by arithmetic mean + one standard deviation with the range in parentheses and sample size below.Simple regression analyses between clutch size (CS), RCM and egg volume and female SVL.Results of Analyses of Variance for differences in SVL between males and females.Analysis of Variance (ANOVA) of female SVL and Analysis of Covariance (ANCOVA) of CS among populations of each species.* Spearman rank correlation