Morphological variation of Philodryas patagoniensis ( Girard , 1858 ) ( Serpentes , Dipsadidae ) from Brazil , based on the study of pholidosis , coloration and morphometric features

The current study aimed to verify the relationship between the patterns of coloration, the morphometrical features and pholidosis of specimens of Philodryas patagoniensis from Brazil, with the sexual dimorphism of this species. We studied specimens of P. patagoniensis deposited in several herpetological collections from Brazil. A total of 355 specimens were analyzed, of these 145 were males (87 adults and 58 juveniles) and 210 were females (134 adults and 76 juveniles). Adult specimens exhibit sexual dimorphism in snout-vent length, tail length, head length, number of ventral shields, and number of subcaudal shields. The analysis of variance showed that the adult females are significantly longer than adult males, both at snout-vent length and tail length. Females have a greater number of ventral shields (138-210) than males (151-200). The specimens studied also exhibit two distinct patterns of coloration unrelated to geographical or sexual variations.


Introduction
External morphological features as pholidosis, measures of body regions and patterns of coloration of specimens are fundamental for taxonomic studies of snakes (Thomas 1976, Di-Bernardo & Lema 1990, López & Giraudo 2008).Pholidosis is the arrangement or pattern of scales and shields on the body surface of reptiles and is one of features for identification of snakes, and is a feature that may have variations within the same species (Dowling 1951a, b).The differences between the phenotypes such as folidose, morphology and physiology reflect the functional capacity of the specimen, which in turn are adaptive to different environments (Arnold 1983, Greene 1986, Coddington 1988, Emerson & Arnold 1989, Arnold 1994, Garland & Losos 1994).
Differences in behavior may influence evolutionary changes, for example, the occurrence of populations or species different from the usual habitat, possibly exhibiting different behaviors, which may lead to selection of phenotypes that maximize the effectiveness of these behaviors and the animal's relationship to its habitat (Mayr 1963).
Research on sexual dimorphism performed in several species of snakes of the families Colubridae and Dipsadidae demonstrate that the females have a highest snout-vent length compared to males (Solorzano & Cerdas 1989, Rivas & Burghardt 2001, Zug et al. 2001), and males have a highest tail length than females (King 1989, López & Giraudo 2008, Matias et al. 2011).
The Patagonia green racer, Philodryas patagoniensis (Girard, 1858) (Dipsadidae), occurs in Argentina, Bolívia, Brazil, Chile, Paraguay and Uruguay (Thomas 1976, Peters et al. 1986, López & Giraudo 2008).This species was described by Girard (1858) as Callirhinus patagoniensis, based on two specimens collected in coastal region of Patagonia, Argentina.These two specimens were described as having mainly an olive-green coloration with black spots, 19 midbody rows of dorsal scales, divided cloacal and subcaudal shields, a divided nasal scale, a pair of loreal scales, two post-ocular scales, and one pre-ocular scale, with the third and fourth supralabial scales in contact with the orbit.This species was included into the genus Philodryas Wagler, 1830 by Hoge (1964), and redescribed by Thomas (1976).
The study of López & Giraudo (2008) with specimens of P. patagoniensis from Argentina is the only one that reveals significant differences in pholidosis of males and females of this species.Other studies with P. patagoniensis have addressed for ecological and natural history approaches (see Thomas 1976, Cei 1993, Achaval & Olmos 1997, Fowler & Salomão 1994a, b, 1995, Hartmann & Marques 2005, López & Giraudo 2008).Not only in specimens of P. patagoniensis, as in the whole tribe Philodryadini, from Brazil the studies with sexual dimorphism associated with pholidosis and morphometric features of this species are still scarce (Fowler & Salomão 1994a, b, 1995, Pontes 2007, López & Giraudo 2008, Zaher et al. 2008).
Therefore, the current study aimed to analyze the differences between sexual dimorphism and the patterns of coloration, the morphometrical features and pholidosis of P. patagoniensis from Brazil.
The specimens were analyzed between May 2011 and September 2012.A total of 355 specimens of Philodryas patagoniensis were analyzed, of which 145 were males (87 adult specimens and 58 juveniles) and 210 were females (134 adults and 76 juveniles) (see appendix).The speciemens analyzed in the current study were identified according to morphological diagnostic features proposed by Girard (1858), andThomas (1976).
The sexual size dimorphism (SSD) was calculated according to the index adapted from Gibbons & Lovich (1990): SSD = mean SVL of the larger sex divided by the mean SVL of the smaller sex, where negative values indicate that males are larger than females.Individuals missing all or part of the tail were not used in the analyses that employed SVL and TL values.Females with SVL < 670 mm and males with SVL < 525 mm were classified as juveniles, according to Pontes (2007).
The relationship of sex and/or sexual maturity with the NVS, NSS, SVL, and TL was determined by the Wilcoxon test.The same test was used to relate the HL with the sex of the specimens and to relate the NVS, NSS, SVL, TL, TBL, HL, WH and IOW to the patterns of coloration.
The data was tested for homogeneity of the variances by the Levene test before other statistical analyses.The significance level in all the tests was α = 0.05.The statistical analyses were carried out with the R for Windows program.
To determine the sex, in all specimens analyzed was verified the absence or presence of hemipenis by eversion, or through a small median longitudinal section in the region of the first subcaudal shields according to Yuki (1994).
The juvenile and adult females have a higher NVS that than males (Table 1; W = 3179.5,p = 0.005 for juveniles; W = 10062, p < 0.001 for adults).Adult females have a longer SVL than adult males (Table 1; W = 10606.5,p < 0.001) and the SVL of juvenile snakes does not differ between the sexes (Table 1; W = 2390, p = 0.330).
Males have a greater NSS than females, according to the analysis of covariance (Table 1; W = 748, p = 0.001 for juveniles; W = 1923.5,p = 0.001 for adults), however the adult females have a greater TL than males (Figure 1) (Table 1; W = 6995, p = 0.007).The TL of the juvenile specimens did not present a statistical difference between the sexes (Figure 1) (Table 1; W = 1981.5,p = 0.381).
The sexual size dimorphism (SSD) index in P. patagoniensis was positive in this study, confirming that females are larger than males (1.004).
The analysis of covariance between the TL and SVL of males and females (R 2 = 0.7848; F = 638.3)shows that the TL is influenced both by SVL and by sex, with a good approximation between the sexes based on the TL and SVL values (Figure 1).
The HL of the females (n=208) ranged between 11.3 to 44.2 mm, while for the males (n=144) it ranged between 7.7 to 33.1 mm.The HL of the females was significantly greater that than HL of males (W = 20925, p = 0.001).
Three models were also tested in the analysis of covariance between the HL and SVL of males and females.Model 3 presented the highest significance (R 2 = 0.8867; F = 2740), with the following coefficients (Intercept: -280.7827,p = <0.001;HL: 38.6754, p = < 0.001).Therefore, the head length is closely related to the snout-vent length of the specimens, however, there was no influence of sex in any of the models (Figure 1).
According to the R 2 value (R 2 = 0.006) found in the regression analysis, there was no correlation between the variables NVS, NSS, and TBL, indicating that these variables are not closely related.
The regression analysis also showed that the TL of these specimens corresponds to 24% (R 2 = 0.840) of the TBL.This means that the TL value  can be predicted from the TBL value.The head length is corresponds to 1.8% (R 2 = 0.903) of the TBL.Thus it is possible to obtain any value of TBL or HL from the other pair.
We observed two patterns of coloration in the specimens studied herein: (1) dorsum dark brown, with black spots only on the posterior region of the scales, a clear and delineated black belly, and orange coloration on the head, and (2) uniform brown head and dorsum, light greenish-brown without black spots on the posterior portion of the dorsal scales, a light belly.These patterns (with or without spots) were used for comparisons of the all specimens studied herein (Figure 2) (Table 2).
The correlations of these two patterns of coloration between the NVS, NSS, TBL, SVL, TL, WH, HL, and IOW were performed according to the Wilcoxon test.We observed that only the NVS and NSS variables were not significant.Among the variables that showed significant results, only the variable IOW showed higher mean for the patterns of coloration 2. For all other variables, the mean was higher for the patterns of coloration 1 (Table 3).

Discussion
The wide range observed in the NVS and NSS in the specimens studied herein was also observed by Thomas (1976) and D'Agostini (1998) in other species of the genus Philodryas, such as P. aestiva (Duméril, Bibron & Duméril, 1854), P. olfersii (Lichtenstein, 1823) and P. viridissima (Linnaeus, 1758).Thomas (1976) studied P. patagoniensis also found a large variation in the NVS and NSS (151 to 194 ventral shields and 68 to 120 subcaudal shields) than that observed in the current study.
The results obtained on the NSS and NVS of the adults of P. patagoniensis are in accordance with observed by López & Giraudo (2008) in the same species from Argentina.These authors observed that the adult males of P. patagoniensis have a higher NSS than the adult females, while the adult females have a higher NVS than the adult males.The higher body size and NVS in females of this species can be a reflective of the optimization of reproductive potential, since the females needs a higher body volume to produce a higher number of eggs or large size of eggs (Mebert 2011).In the current study we observed a significant difference in the NSS between males and females (Table 1).We observed that the adult females of P. patagoniensis have a largest body size that than adult males (Table 1).A similar pattern of females were reported in several other species of snakes (Solorzano & Cerdas 1989, Shine 1993, Rivas & Burghardt 2001, Gregory 2004, Pinto & Fernandes 2004, Matias et al. 2011, Mesquita et al. 2011, Henao-Duque & Ceballos 2013).
Several authors observed that the fertility is considered one of the main selective agents for the larger size of females for some species of snakes; however, the largest size of males in another species of snakes can be considered as a significant feature of males of species that possesses the behaviour of competition for territory and for females in reproductive period (Shine 1978, 1993, 1994, 2000, Zug et al. 2001, Matias et al. 2011, Henao-Duque & Ceballos 2013).
We found a high value of sexual size dimorphism index (SSD), which can possibly indicate the absence of territorial behaviour in P. patagoniensis, since the species of snakes that exhibiting the behaviour of territorial and reproductive dispute usually have SSD < 0.05 or with negative values of this index, according Shine (1978Shine ( , 1994) ) and Oliveira (2008).The same applies to the species P. agassizii and P. nattereri, since they have SSD >0.05 and there are no indications of territorial behavior (Marques et al. 2006, Mesquita et al. 2011).Shine (1978) says that the late maturity in female snakes can represent a trade off in which the sexual maturity is delayed until their body size allows the production of sufficiently numerous and large eggs, and highest fertility.Sex differences in snake body size adjustments are considered that may lead to a higher success reproductive (Pizzatto et al. 2007).Therefore, the sexual dimorphism probably exists due to the benefits of larger body size of female so that, according to Pizzatto et al. (2007), the natural selection has favored females with larger body size.Studies of snakes have shown that on the average females are larger than males (approximately 15%) in species where males do not compete for mates, Shine (1994).The sexual dimorphism occurring in the SVL of P. patagoniensis indicates than the sexes reach maturity at different sizes; however, according to Fowler & Salomão (1995) it is known that growth is not a continuous process.Biota Neotrop., 17( 1  According to Mesquita et al. (2010), in specimens of Oxybelis aeneus (Wagler in Spix, 1824) (Colubridae) from Brazil the growth rate declines after sexual maturity and therefore females reach maturity at a larger average size than males; although adult males with larger tails that adult females are most commonly pattern found in snakes, due to the presence of hemipenis and retractors muscles (Shine 1993).However, we observed that this pattern of tail sizes of males and females in P. patagoniensis (Table 1) not correspond with reported in most species of snakes.We can not find arguments that could support a hypothesis about females having a larger tail than males.Perhaps in the future, with the increase of studies on morphometric patterns in other species of neotropical snakes, not yet studied, could elucidate this fact.López & Giraudo (2008) found out that there is difference in the count and size of the shields of females and adults males of P. patagoniensis.Females have more ventral shields and males have more subcaudal shields.Vanzolini & Brandão (1944, 1945) found the occurrence of that dimorphism in Bothrops alternatus Duméril, Bibron & Duméril, 1854, with the same pattern (females have a higher number of dorsal scales and ventral shields, and males have a higher number of subcaudal shields), however, as in the current study these authors also do not find a justification that would explain this fact.
The sexual dimorphism can be observed in different ways in snakes (Rivas & Burghardt 2001).For several species the sexual dimorphism is related to volume and size of body (Shine 1993(Shine , 1994)); the shape and size of head is another feature for diagnostic of sexual dimorphism (Camilleri & Shine 1990, Vincent et al. 2004); the size of visceral mass and glands also is a important feature for detection of sexual dimorphism (Kissner et al. 1998); but some authors affirms that the patterns of coloration rarely is related to sexual dimorphism (Shine 1993, Marques & Sazima 2003).
In the current study we can affirm that not exist a significant difference in the pattern of coloration between males and females of P. patagoniensis.
The specimens that we examined from the state of Roraima have not been reported previously in the literature; therefore the current study provides data that expands geographic distribution of this species in Brazil (Figure 3).
We can affirm that the specimens of P. patagoniensis studied herein had a great variation in the folidosis.Sexual dimorphism in relation to SVL, TL, NVS, NSS, and HL was observed.Adult females are larger than adult males in SVL and TL.However, in juveniles these differences do not occur.Females have a greater number of ventral shields than males, and males have a greater number of subcaudal shields.Also we can affirm that the specimens of P. patagoniensis studied have two distinct patterns of coloration, however without geographic and sexual specificity for these patterns.

Figure 1 .
Figure 1.(A) Relation between the tail length and number of subcaudal shield of juveniles females and males.(B) Relation between the tail length and number of subcaudal shield of adults females and males.(C) Relation between the tail length and snout-vent length of females and males.(D) Relationship of the snout-vent length, head length and sex of the specimens.

Figure 3 .
Figure 3. Distribution of studied specimens of Philodryas patagoniensis in 18 states of Brazil and the Federal District.

Table 2 .
Percentage of specimens from different Brazilian states according to the color pattern, sex and sexual maturity in the specimens of the Philodryas patagoniensis.

Table 3 .
Total number, means, standard deviation, variation and statistical results between the coloration patterns with variable: number of ventral and subcaudal shields (NVS and NSS), total length (TBL), snout-vent length (SVL), tail length (TL), head length (HL), head width (WH) and inter-ocular width (IOW) of Philodryas patagoniensis.Statistical significance represented by * in the table.