The Length of a Snake is It Continuous and Countable
Abstract
Body size and body shape are tightly related to an animal's physiology, ecology and life history, and, as such, play a major role in understanding ecological and evolutionary phenomena. Because organisms have different shapes, only a uniform proxy of size, such as mass, may be suitable for comparisons between taxa. Unfortunately, snake masses are rarely reported in the literature. On the basis of 423 species of snakes in 10 families, we developed clade-specific equations for the estimation of snake masses from snout–vent lengths and total lengths. We found that snout–vent lengths predict masses better than total lengths. By examining the effects of phylogeny, as well as ecological and life history traits on the relationship between mass and length, we found that viviparous species are heavier than oviparous species, and diurnal species are heavier than nocturnal species. Furthermore, microhabitat preferences profoundly influence body shape: arboreal snakes are lighter than terrestrial snakes, whereas aquatic snakes are heavier than terrestrial snakes of a similar length.
Introduction
An organism is an ensemble of many characteristics, among which body size is undoubtedly one of the most important. Body size is strongly related to the organism's physiology, energy requirements, ecology and life history (Calder, 1984; Shine, 1994a; Brown et al., 2004). Size is therefore the focus of many ecological and evolutionary studies, dating back to the inspiring studies of Bergmann (1847; translation in James, 1970) and Cope 1887) and continuing to our time (e.g. Hutchinson & MacArthur, 1959; Foster, 1964; Griffiths, 2012). Greater accessibility to data has enhanced the compilation of size data for some of the major vertebrate taxa ( Smith et al., 2003; Olden, Hogan & Zanden, 2007; Meiri, 2008; Olson et al., 2009), and these datasets have allowed us to search further for ecological and evolutionary processes involving size.
The body size of snakes has been studied in a variety of contexts, such as geographical variation (e.g. Ashton & Feldman, 2003; Olalla-Tárraga, Rodrıguez & Hawkins, 2006; Terribile et al., 2009; Amarello et al., 2010), range size (e.g. Bonfim, Diniz-Filho & Bastos, 1998; Reed, 2003), insularity (Boback, 2003), body size frequency distributions (Cox, Boback & Guyer, 2011), optimality (Boback & Guyer, 2003) and natural history (Pough & Groves, 1983). The measure of size in these studies, and in others, was length [ Olalla-Tárraga et al. (2006) converted all lengths to masses, but used the same equation for all species]. This is undoubtedly because the snout–vent length (SVL) and total length (TL) are the most common size measures reported for snakes. Mass, another measure of body size, is rarely reported, and thus is rarely used. Some authors have claimed that length is a more appropriate size proxy than mass (Boback, 2003; Boback & Guyer, 2003), because mass depends on factors such as breeding condition, season, health, and size of and time from the last meal (Meiri, 2010). Furthermore, length is probably highly correlated with mass (Kaufman & Gibbons, 1975; Guyer & Donnelly, 1990). However, being a linear measure, length may fail to explain variation in body shape (Meiri, 2010), and snakes, although all having elongated body forms (Lillywhite, 1987), are differently shaped (Pough & Groves, 1983; Aubret & Shine, 2009). Thus, length should be regarded carefully as a proxy for body size. Mass, however, is considered to be the best measure of body size in life history studies (Hedges, 1985). Mass is also the measure of choice in physiological studies (e.g. Seymor, 1987; Glazier, 2009; Bronikowski & Vleck, 2010), because various physiological rates (e.g. metabolic rate; Kleiber, 1947; Nagy, 2005) are mass dependent. Many ecological and physiological attributes vary allometrically with body mass (Peters, 1983; Schmidt-Nielsen, 1984). Hence, in many instances, mass will be a more useful proxy than length for body size. In addition, as lengths cannot be compared directly across groups of organisms that differ in shape, comparison between such groups that is based on length may prove to be misleading, but their masses can still be meaningfully compared (Hedges, 1985; Meiri, 2010).
Because snake masses are rarely reported, they need to be estimated from the commonly reported measure, length. Pough (1980) presented the most comprehensive allometric equations for the estimation of snake mass from length. He proposed two general equations – one for the calculation of mass from SVL (mass = 6.6 × 10−4 × SVL3.02) and the other for the calculation of mass from TL (mass = 3.5 × 10−4 × TL3.02) (mass in grams, length in centimetres). However, Pough's pioneering equations are based on the length and mass data of only 13 species of colubrids and vipers from the southern USA (data from Kaufman & Gibbons, 1975) and, as such, may not generalize to other snake taxa, if the latter have different body shapes. Vipers, for example, are generally more robust than other snakes (Pough & Groves, 1983; and see below), and so are constrictors (see below). Moreover, body shape and hence mass–length relationships are influenced by phylogenetic affinities, life history traits and ecological characteristics (Pough & Groves, 1983; Shine, 1994a; Martins et al., 2001; Brown et al., 2004). Arboreal species, for example, are generally more slender and have longer tails than terrestrial species (Guyer & Donnelly, 1990; Lillywhite & Henderson, 1993; Martins et al., 2001), and burrowing or aquatic species may differ even more from above-ground active snakes (Murphy, 2007; and see below). One equation for the estimation of mass ignores any such difference, and these differences may be too great to be neglected.
More specific allometries will allow us to assess snake masses and the variation between different ophidian clades more accurately. Such equations have already been developed for lizards (Meiri, 2010; Pincheira-Donoso et al., 2011), and have been proven to be a better predictor of mass than is the single equation of Pough ( Pincheira-Donoso et al., 2011).
We compiled a database of 423 snake species in 10 families to develop new allometries for the estimation of snake mass from their SVL and TL, and evaluated the explanatory power of the equations depending on the measure used (SVL or TL). To overcome phylogenetic affinities and biological differences between clades, we generated different equations for different snake lineages. We then evaluated the predictive power of the equations and compared it with the predictive power of Pough's allometric equations. Finally, we attempted to estimate the influence of phylogenetic, ecological and life history traits on the relationship between mass and length. We specifically estimated the influence of reproduction mode, venomousness (function-based definition of venom, i.e. 'a toxic compound injected into prey or predator to cause rapid death or incapacitation'; Fry et al., 2012) and activity time on this relationship. As there is a relationship between microhabitat use and body shape (for example, arboreal species are lighter than terrestrial species; Lillywhite & Henderson, 1993; Martins et al., 2001), we added microhabitat use as a covariate in the model. We formulated four hypotheses:
-
Being elongated, snakes have limited abdominal space to hold their litter or eggs ( Bonnet et al., 2000). Because viviparous species, unlike oviparous species, hold their embryos until the end of their development, we predicted that viviparous species would be heavier than oviparous species, because females need more abdominal space to retain their embryos. Moreover, the evolution of viviparity is associated, in some species, with a shift towards larger female size relative to male size (Shine, 1994b), indicating that viviparous species may be relatively heavier than oviparous species.
-
In congruence with former studies (e.g. Guyer & Donnelly, 1990; Martins et al., 2001), we hypothesized that mass would decrease on a gradient from terrestrial to arboreal snakes, enabling the latter to move more easily on trees. We further predicted that aquatic species would be heavier, on average, than others to better retain heat.
-
Body shape may play a significant role in heat transfer (e.g. Spotila et al., 1973). For example, assuming a cylindrical shape and a specific density of 1 g cm−3, a 1130-mm (TL) and 94.7 g Dolichophis jugularis from Israel [Tel Aviv University Zoological Museum (TAUM) specimen #10075] has a surface/volume ratio of 12.2 mm−1, whereas a heavier (432.2 g) Daboia palestinae (TAUM #8186) with the same length has a surface/volume ratio of 5.7 mm−1. Nocturnal species are not exposed to solar radiation, and the heat exchange between them and the environment is influenced by conduction and convection (Heatwole & Taylor, 1987). Because nocturnal species face, on average, lower temperatures than diurnal species, we predicted that nocturnal species would benefit from a lower surface to volume ratio (i.e. would be heavier than diurnal species) to retain heat and reduce the rate of cooling (Slip & Shine, 1988).
-
Venom in snakes facilitates prey manipulation, swallowing and prey handling (Kardong, 1979). Hence, nonvenomous snakes must invest more power in killing their prey, either by biting or constricting. We predicted that nonvenomous snakes would therefore be heavier than venomous snakes of a similar length as a result of the more strongly developed muscles needed for prey handling.
Methods
Data
We accumulated data on length (SVL and TL, in millimetres) and body mass (in grams) for 423 snake species. Data are mainly from the published literature and from museum specimens housed at TAUM. We also measured live snakes at the Meier Segal's Garden for Zoological Research, Tel Aviv University. We used only literature length and mass data that were reported together for a species in the same publication and referred to the same individual or individuals. Because sexual size dimorphism (SSD) is common in snakes (Seigel & Ford, 1987; Madsen & Shine, 1993; Shine et al., 1998) and may affect the mass–length relationships (Kaufman & Gibbons, 1975), we collected data separately for males, females and unsexed individuals. We did not use data for females that were known to be gravid at the time of measurements. All museum specimens used were weighed and measured prior to preservation. When several measurements were available for the same species, we used the mean values of all data for the species. However, we did not average mass data if they were associated with SVL data for some individuals and with TL data for others. Instead, we created two datasets per species – one for SVL and mass and the other for TL and mass. We used only data from adults. Taxonomy follows Uetz (2011).
We classified species as either oviparous or viviparous and treated ovoviviparous species as viviparous, because females retain their eggs inside their bodies until hatching. We classified species as venomous or nonvenomous and treated opisthoglyphous (rear-fanged) species as venomous (a preliminary analysis found that our results are robust to treating them as venomous). We classified species as diurnal, nocturnal or cathemeral, and microhabitat preferences as aquatic (including semi-aquatics), burrowers (fossorial and semi-fossorial), terrestrial, terrestrial–arboreal and arboreal. We analysed this relationship for all snake clades together, controlling for phylogenetic affinities. We repeated the analysis for colubrids, a species-rich lineage which, unlike most snake families, contains much ecological and morphological diversity in the traits examined.
Data analysis
Prior to statistical analysis, we log10-transformed all weight and length data. In a preliminary analysis for each family, we found that allometries based only on males did not differ from those based only on females (not shown). Therefore, we analysed all data (i.e. males, females and unsexed individuals) together. To develop family-specific equations, we regressed mass separately on SVL and on TL for each lineage. To compare the predictive power of the equations using SVL or TL, we used only species for which we had both SVL and TL data from the same individual/individuals, and selected the best models according to the Akaike Information Criterion (AIC) (Wagenmakers & Farrell, 2004). To estimate the effect of ecological and life history traits on the relationship between mass and length, we used backwards-stepwise elimination and determined the best model as that in which all traits were statistically significant (at P = 0.05).
To evaluate the predictive power of the allometries, we randomly divided our data, using the 'rand' command in Microsoft Excel, into two, and developed allometries for the relationship between mass and SVL for each half. We then used these allometries to calculate the mass of the other half. We also calculated, for each species, the estimated mass using Pough's equation, and tested for the difference and percentage of deviation between the original mass, our predicted mass and Pough's predicted mass. All statistical tests were conducted using R 2.13.1 (R Development Core Team, 2011).
Phylogeny
To control for the effects of shared ancestry, we assembled a species-level phylogeny from published phylogenetic trees. The higher level branching (e.g. family and subfamily levels) is based on Lee et al. (2007) and Pyron et al. (2011) for Alethinophidia ('advanced snakes') and on Vidal et al. (2010) for Scolecophidia ('blind snakes'). As several phylogenetic hypotheses have been suggested recently, we also analysed the data following the higher level tree of Vidal et al. (xref ref-type="bibr" rid="bij2001-bib-0001">2010) to examine whether the results of our phylogenetic analysis are robust to the phylogenetic hypothesis we used. The tree of Vidal et al. (xref ref-type="bibr" rid="bij2001-bib-0001">2010) differs from that of Pyron et al. (2011) in its interpretation of the branching of several Alethinophidian clades, especially within the superfamily Colubridea. To compare the models based on these two phylogenies, we used only species for which we had all ecological and life history data. Preliminary analysis yielded qualitatively the same results (in terms of the predictors in the best model and their direction), and thus we present here only the results of the model which is based on the phylogeny of Lee et al. (2007) and Pyron et al. (2011). The results based on Vidal et al.'s (xref ref-type="bibr" rid="bij2001-bib-0001">2010) tree are presented in Appendix S3. Lacking branch lengths for most of the trees, we scaled branches to make the trees ultrametric using the cladogram transform in FigTree (Rambaut, 2010). We used phylogenetic generalized least-square (PGLS) regression to account for phylogenetic nonindependence. We adjusted the strength of phylogenetic nonindependence using the maximum likelihood value of the scaling parameter λ (Pagel, 1999) implemented in the R package CAIC (Orme, online, http://r-forge.r-project.org/projects/caic/). The scaling parameter λ varies between zero (no phylogenetic signal) and unity (strong phylogenetic signal, equivalent to phylogenetically independent contrast analysis).
Results
Our dataset (Appendix S1) includes 423 species (336 species for SVL and 337, partially overlapping species for TL). It contains both the smallest known snake (Tetracheilostoma carlae, SVL = 93.7 mm, TL = 99.4 mm, 0.6 g; Hedges, 2008) and a few of the largest species (green anacondas, reticulated, Burmese and African rock pythons, up to SVL = 3370 mm, TL = 3735 mm and 21.75 kg in our database). We calculated family-specific allometric equations for all Alethinophidian families with N ≥ 6 species for which we had length and mass data. The infra-order Alethinophidia is represented in our data by seven families and the infra-order Scolecophidia is represented by three (however, because of the small sample size, we grouped all Scolecophidians and generated one equation for the entire clade). The family mass–length allometries are shown in Table 1.
Table 1.
Mass–length allometries for snake families and the infra-order Scolecophidia
| Family | Measure | N | Slope | SE | colspan="2">95% confidence interval | Intercept | SE | R 2 | |
|---|---|---|---|---|---|---|---|---|---|
| All | SVL | 336 | 2.786 | 0.063 | 2.661 | 2.910 | −5.773 | 0.172 | 0.853 |
| TL | 337 | 2.597 | 0.073 | 2.452 | 2.744 | −5.465 | 0.208 | 0.789 | |
| PGLM | SVL | 241 | 2.578 | 0.079 | 2.422 | 2.734 | −5.148 | 0.254 | 0.834 |
| TL | 235 | 2.443 | 0.092 | 2.261 | 2.625 | −4.989 | 0.313 | 0.777 | |
| Boidae | SVL | 15 | 2.776 | 0.279 | 2.171 | 3.380 | −5.500 | 0.814 | 0.883 |
| TL | 13 | 2.856 | 0.331 | 2.126 | 3.585 | −5.886 | 0.982 | 0.871 | |
| Colubridae | SVL | 166 | 2.520 | 0.088 | 2.346 | 2.693 | −5.143 | 0.239 | 0.834 |
| TL | 154 | 2.340 | 0.098 | 2.146 | 2.534 | −4.912 | 0.278 | 0.792 | |
| Elapidae | SVL | 26 | 2.453 | 0.215 | 2.009 | 2.897 | −4.892 | 0.606 | 0.844 |
| TL | 49 | 2.407 | 0.188 | 2.028 | 2.786 | −4.819 | 0.539 | 0.777 | |
| Homalopsidae | SVL | 11 | 3.631 | 0.792 | 1.837 | 5.425 | −7.713 | 2.128 | 0.700 |
| TL | 8 | 3.617 | 0.767 | 1.738 | 5.496 | −8.027 | 2.130 | 0.787 | |
| Lamprophiidae | SVL | 35 | 3.232 | 0.201 | 2.823 | 3.641 | −7.092 | 0.542 | 0.886 |
| TL | 31 | 2.821 | 0.229 | 2.351 | 3.290 | −6.286 | 0.646 | 0.839 | |
| Pythonidae | SVL | 6 | 2.630 | 0.301 | 1.793 | 3.467 | −5.124 | 0.972 | 0.950 |
| TL | 12 | 2.611 | 0.260 | 2.031 | 3.190 | −5.131 | 0.848 | 0.909 | |
| Viperidae | SVL | 60 | 2.655 | 0.107 | 2.440 | 2.869 | −5.165 | 0.293 | 0.913 |
| TL | 51 | 2.910 | 0.156 | 2.596 | 3.223 | −6.013 | 0.438 | 0.877 | |
| Infra-order Scolecophidia | SVL | 17 | 2.985 | 0.553 | 1.804 | 4.165 | −6.381 | 1.254 | 0.660 |
| TL | 18 | 3.068 | 0.489 | 2.031 | 4.104 | −6.596 | 1.144 | 0.711 |
| Family | Measure | N | Slope | SE | colspan="2">95% confidence interval | Intercept | SE | R 2 | |
|---|---|---|---|---|---|---|---|---|---|
| All | SVL | 336 | 2.786 | 0.063 | 2.661 | 2.910 | −5.773 | 0.172 | 0.853 |
| TL | 337 | 2.597 | 0.073 | 2.452 | 2.744 | −5.465 | 0.208 | 0.789 | |
| PGLM | SVL | 241 | 2.578 | 0.079 | 2.422 | 2.734 | −5.148 | 0.254 | 0.834 |
| TL | 235 | 2.443 | 0.092 | 2.261 | 2.625 | −4.989 | 0.313 | 0.777 | |
| Boidae | SVL | 15 | 2.776 | 0.279 | 2.171 | 3.380 | −5.500 | 0.814 | 0.883 |
| TL | 13 | 2.856 | 0.331 | 2.126 | 3.585 | −5.886 | 0.982 | 0.871 | |
| Colubridae | SVL | 166 | 2.520 | 0.088 | 2.346 | 2.693 | −5.143 | 0.239 | 0.834 |
| TL | 154 | 2.340 | 0.098 | 2.146 | 2.534 | −4.912 | 0.278 | 0.792 | |
| Elapidae | SVL | 26 | 2.453 | 0.215 | 2.009 | 2.897 | −4.892 | 0.606 | 0.844 |
| TL | 49 | 2.407 | 0.188 | 2.028 | 2.786 | −4.819 | 0.539 | 0.777 | |
| Homalopsidae | SVL | 11 | 3.631 | 0.792 | 1.837 | 5.425 | −7.713 | 2.128 | 0.700 |
| TL | 8 | 3.617 | 0.767 | 1.738 | 5.496 | −8.027 | 2.130 | 0.787 | |
| Lamprophiidae | SVL | 35 | 3.232 | 0.201 | 2.823 | 3.641 | −7.092 | 0.542 | 0.886 |
| TL | 31 | 2.821 | 0.229 | 2.351 | 3.290 | −6.286 | 0.646 | 0.839 | |
| Pythonidae | SVL | 6 | 2.630 | 0.301 | 1.793 | 3.467 | −5.124 | 0.972 | 0.950 |
| TL | 12 | 2.611 | 0.260 | 2.031 | 3.190 | −5.131 | 0.848 | 0.909 | |
| Viperidae | SVL | 60 | 2.655 | 0.107 | 2.440 | 2.869 | −5.165 | 0.293 | 0.913 |
| TL | 51 | 2.910 | 0.156 | 2.596 | 3.223 | −6.013 | 0.438 | 0.877 | |
| Infra-order Scolecophidia | SVL | 17 | 2.985 | 0.553 | 1.804 | 4.165 | −6.381 | 1.254 | 0.660 |
| TL | 18 | 3.068 | 0.489 | 2.031 | 4.104 | −6.596 | 1.144 | 0.711 |
PGLM, phylogenetic general linear model; SE, standard error; SVL, snout–vent length; TL, total length.
Log mass/log length allometries for different snake families and infra-order Scolecophidia.
Table 1.
Mass–length allometries for snake families and the infra-order Scolecophidia
| Family | Measure | N | Slope | SE | colspan="2">95% confidence interval | Intercept | SE | R 2 | |
|---|---|---|---|---|---|---|---|---|---|
| All | SVL | 336 | 2.786 | 0.063 | 2.661 | 2.910 | −5.773 | 0.172 | 0.853 |
| TL | 337 | 2.597 | 0.073 | 2.452 | 2.744 | −5.465 | 0.208 | 0.789 | |
| PGLM | SVL | 241 | 2.578 | 0.079 | 2.422 | 2.734 | −5.148 | 0.254 | 0.834 |
| TL | 235 | 2.443 | 0.092 | 2.261 | 2.625 | −4.989 | 0.313 | 0.777 | |
| Boidae | SVL | 15 | 2.776 | 0.279 | 2.171 | 3.380 | −5.500 | 0.814 | 0.883 |
| TL | 13 | 2.856 | 0.331 | 2.126 | 3.585 | −5.886 | 0.982 | 0.871 | |
| Colubridae | SVL | 166 | 2.520 | 0.088 | 2.346 | 2.693 | −5.143 | 0.239 | 0.834 |
| TL | 154 | 2.340 | 0.098 | 2.146 | 2.534 | −4.912 | 0.278 | 0.792 | |
| Elapidae | SVL | 26 | 2.453 | 0.215 | 2.009 | 2.897 | −4.892 | 0.606 | 0.844 |
| TL | 49 | 2.407 | 0.188 | 2.028 | 2.786 | −4.819 | 0.539 | 0.777 | |
| Homalopsidae | SVL | 11 | 3.631 | 0.792 | 1.837 | 5.425 | −7.713 | 2.128 | 0.700 |
| TL | 8 | 3.617 | 0.767 | 1.738 | 5.496 | −8.027 | 2.130 | 0.787 | |
| Lamprophiidae | SVL | 35 | 3.232 | 0.201 | 2.823 | 3.641 | −7.092 | 0.542 | 0.886 |
| TL | 31 | 2.821 | 0.229 | 2.351 | 3.290 | −6.286 | 0.646 | 0.839 | |
| Pythonidae | SVL | 6 | 2.630 | 0.301 | 1.793 | 3.467 | −5.124 | 0.972 | 0.950 |
| TL | 12 | 2.611 | 0.260 | 2.031 | 3.190 | −5.131 | 0.848 | 0.909 | |
| Viperidae | SVL | 60 | 2.655 | 0.107 | 2.440 | 2.869 | −5.165 | 0.293 | 0.913 |
| TL | 51 | 2.910 | 0.156 | 2.596 | 3.223 | −6.013 | 0.438 | 0.877 | |
| Infra-order Scolecophidia | SVL | 17 | 2.985 | 0.553 | 1.804 | 4.165 | −6.381 | 1.254 | 0.660 |
| TL | 18 | 3.068 | 0.489 | 2.031 | 4.104 | −6.596 | 1.144 | 0.711 |
| Family | Measure | N | Slope | SE | colspan="2">95% confidence interval | Intercept | SE | R 2 | |
|---|---|---|---|---|---|---|---|---|---|
| All | SVL | 336 | 2.786 | 0.063 | 2.661 | 2.910 | −5.773 | 0.172 | 0.853 |
| TL | 337 | 2.597 | 0.073 | 2.452 | 2.744 | −5.465 | 0.208 | 0.789 | |
| PGLM | SVL | 241 | 2.578 | 0.079 | 2.422 | 2.734 | −5.148 | 0.254 | 0.834 |
| TL | 235 | 2.443 | 0.092 | 2.261 | 2.625 | −4.989 | 0.313 | 0.777 | |
| Boidae | SVL | 15 | 2.776 | 0.279 | 2.171 | 3.380 | −5.500 | 0.814 | 0.883 |
| TL | 13 | 2.856 | 0.331 | 2.126 | 3.585 | −5.886 | 0.982 | 0.871 | |
| Colubridae | SVL | 166 | 2.520 | 0.088 | 2.346 | 2.693 | −5.143 | 0.239 | 0.834 |
| TL | 154 | 2.340 | 0.098 | 2.146 | 2.534 | −4.912 | 0.278 | 0.792 | |
| Elapidae | SVL | 26 | 2.453 | 0.215 | 2.009 | 2.897 | −4.892 | 0.606 | 0.844 |
| TL | 49 | 2.407 | 0.188 | 2.028 | 2.786 | −4.819 | 0.539 | 0.777 | |
| Homalopsidae | SVL | 11 | 3.631 | 0.792 | 1.837 | 5.425 | −7.713 | 2.128 | 0.700 |
| TL | 8 | 3.617 | 0.767 | 1.738 | 5.496 | −8.027 | 2.130 | 0.787 | |
| Lamprophiidae | SVL | 35 | 3.232 | 0.201 | 2.823 | 3.641 | −7.092 | 0.542 | 0.886 |
| TL | 31 | 2.821 | 0.229 | 2.351 | 3.290 | −6.286 | 0.646 | 0.839 | |
| Pythonidae | SVL | 6 | 2.630 | 0.301 | 1.793 | 3.467 | −5.124 | 0.972 | 0.950 |
| TL | 12 | 2.611 | 0.260 | 2.031 | 3.190 | −5.131 | 0.848 | 0.909 | |
| Viperidae | SVL | 60 | 2.655 | 0.107 | 2.440 | 2.869 | −5.165 | 0.293 | 0.913 |
| TL | 51 | 2.910 | 0.156 | 2.596 | 3.223 | −6.013 | 0.438 | 0.877 | |
| Infra-order Scolecophidia | SVL | 17 | 2.985 | 0.553 | 1.804 | 4.165 | −6.381 | 1.254 | 0.660 |
| TL | 18 | 3.068 | 0.489 | 2.031 | 4.104 | −6.596 | 1.144 | 0.711 |
PGLM, phylogenetic general linear model; SE, standard error; SVL, snout–vent length; TL, total length.
Log mass/log length allometries for different snake families and infra-order Scolecophidia.
Analysis of covariance revealed that intercepts differed significantly between families, both in the relationship between mass and SVL (F 1,7 = 22.466, P ≪ 0.001, Fig. 1) and in the relationship between mass and TL (F 1,7 = 31.197, P ≪ 0.001). Family slopes differed in the relationship between mass and TL (F 1,7 = 2.101, P = 0.043), but not in the relationship between mass and SVL (F 1,7 = 1.665, P = 0.117).
Figure 1.
A, Mass–snout to vent length (SVL) relationship for snake families in our dataset. B, Mass–SVL relationship for snake families. All species were fitted with parameter estimates for a terrestrial, diurnal and oviparous snake. See Table 3 for parameter estimates.
Mass was usually explained better by SVL than by TL (Table 2). When the same individuals were used to derive allometric equations for both length measures, only in elapids and, to a lesser extent, in homalopsids, did TL explain more of the variation in mass than SVL. Because SVL is generally a superior predictor for mass (Table 2), we used SVLs to examine the factors influencing the length–mass relationship.
Table 2.
Amount of variation explained and Akaike Information Criterion (AIC) scores for allometries of mass on snout–vent length (SVL) and total length (TL)
| Family | Measure | R 2 | AIC |
|---|---|---|---|
| All (245) | SVL | 0.837 | 94.555 |
| TL | 0.773 | 175.774 | |
| All (corrected for family) | SVL | 0.881 | 29.100 |
| TL | 0.858 | 72.263 | |
| Boidae (10) | SVL | 0.917 | 5.378 |
| TL | 0.880 | 8.691 | |
| Colubridae (128) | SVL | 0.822 | 23.942 |
| TL | 0.774 | 54.344 | |
| Elapidae (12) | SVL | 0.529 | 11.787 |
| TL | 0.562 | 10.902 | |
| Homalopsidae (6) | SVL | 0.799 | 3.730 |
| TL | 0.806 | 3.531 | |
| Lamprophiidae (30) | SVL | 0.886 | −6.566 |
| TL | 0.839 | 3.773 | |
| Viperidae (44) | SVL | 0.905 | −27.145 |
| TL | 0.899 | −24.068 | |
| Infra-order Scolecophidia (15) | SVL | 0.634 | 11.982 |
| TL | 0.622 | 12.454 |
| Family | Measure | R 2 | AIC |
|---|---|---|---|
| All (245) | SVL | 0.837 | 94.555 |
| TL | 0.773 | 175.774 | |
| All (corrected for family) | SVL | 0.881 | 29.100 |
| TL | 0.858 | 72.263 | |
| Boidae (10) | SVL | 0.917 | 5.378 |
| TL | 0.880 | 8.691 | |
| Colubridae (128) | SVL | 0.822 | 23.942 |
| TL | 0.774 | 54.344 | |
| Elapidae (12) | SVL | 0.529 | 11.787 |
| TL | 0.562 | 10.902 | |
| Homalopsidae (6) | SVL | 0.799 | 3.730 |
| TL | 0.806 | 3.531 | |
| Lamprophiidae (30) | SVL | 0.886 | −6.566 |
| TL | 0.839 | 3.773 | |
| Viperidae (44) | SVL | 0.905 | −27.145 |
| TL | 0.899 | −24.068 | |
| Infra-order Scolecophidia (15) | SVL | 0.634 | 11.982 |
| TL | 0.622 | 12.454 |
Number of species in each category is given in parentheses.
Table 2.
Amount of variation explained and Akaike Information Criterion (AIC) scores for allometries of mass on snout–vent length (SVL) and total length (TL)
| Family | Measure | R 2 | AIC |
|---|---|---|---|
| All (245) | SVL | 0.837 | 94.555 |
| TL | 0.773 | 175.774 | |
| All (corrected for family) | SVL | 0.881 | 29.100 |
| TL | 0.858 | 72.263 | |
| Boidae (10) | SVL | 0.917 | 5.378 |
| TL | 0.880 | 8.691 | |
| Colubridae (128) | SVL | 0.822 | 23.942 |
| TL | 0.774 | 54.344 | |
| Elapidae (12) | SVL | 0.529 | 11.787 |
| TL | 0.562 | 10.902 | |
| Homalopsidae (6) | SVL | 0.799 | 3.730 |
| TL | 0.806 | 3.531 | |
| Lamprophiidae (30) | SVL | 0.886 | −6.566 |
| TL | 0.839 | 3.773 | |
| Viperidae (44) | SVL | 0.905 | −27.145 |
| TL | 0.899 | −24.068 | |
| Infra-order Scolecophidia (15) | SVL | 0.634 | 11.982 |
| TL | 0.622 | 12.454 |
| Family | Measure | R 2 | AIC |
|---|---|---|---|
| All (245) | SVL | 0.837 | 94.555 |
| TL | 0.773 | 175.774 | |
| All (corrected for family) | SVL | 0.881 | 29.100 |
| TL | 0.858 | 72.263 | |
| Boidae (10) | SVL | 0.917 | 5.378 |
| TL | 0.880 | 8.691 | |
| Colubridae (128) | SVL | 0.822 | 23.942 |
| TL | 0.774 | 54.344 | |
| Elapidae (12) | SVL | 0.529 | 11.787 |
| TL | 0.562 | 10.902 | |
| Homalopsidae (6) | SVL | 0.799 | 3.730 |
| TL | 0.806 | 3.531 | |
| Lamprophiidae (30) | SVL | 0.886 | −6.566 |
| TL | 0.839 | 3.773 | |
| Viperidae (44) | SVL | 0.905 | −27.145 |
| TL | 0.899 | −24.068 | |
| Infra-order Scolecophidia (15) | SVL | 0.634 | 11.982 |
| TL | 0.622 | 12.454 |
Number of species in each category is given in parentheses.
Factors influencing the length–mass relationship
Nonphylogenetic analysis
The nonphylogenetic model included 336 species (Appendix S1). The best model for snake mass included SVL (slope = 2.668, SE = 0.066), family (Fig. 1), microhabitat use, reproduction mode (Fig. 2) and diel activity. The model explained 92.7% of the variation in mass. There were no interactions between SVL and any parameter. In colubrids viviparous species were heavier than oviparous species, but this difference was marginally nonsignificant (intercept difference = 0.069, t = 1.761, P = 0.0793). Diurnal species were heavier than nocturnal species, but this difference was also marginally nonsignificant (intercept difference = 0.057, t = 1.815, P = 0.070). Shape was strongly correlated with microhabitat use, with aquatic species significantly heavier than all other snakes (P ≪ 0.001 in all cases) and arboreal species significantly lighter than others (P ≪ 0.001 in all cases). The difference between terrestrial and terrestrial–arboreal species was marginally nonsignificant (P = 0.095), and both terrestrial and terrestrial–arboreal species did not differ significantly from burrowing species. Parameter estimates of the full model are shown in Table 3.
Figure 2.
A, Mass–snout to vent length (SVL) relationship for oviparous and viviparous snake species in our dataset: white, oviparous species; black, viviparous species. B, Mass–SVL relationship for oviparous and viviparous colubrids and viperids species. All species were fitted with parameter estimates for a terrestrial and diurnal snake. See Table 3 for parameter estimates.
Table 3.
Parameter estimates for nonphylogenetic model. Slope and intercept values for oviparous, diurnal and terrestrial species in each family. The intercept value differences should be added for the calculation of other traits [for example, the intercept of a viviparous (0.069), nocturnal (–0.057) and arboreal (–0.312) viperid is –5.110 + 0.069 – 0.057 – 0.312]
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.668 | ||
| Colubridae | −5.510 | ||
| Boidae | −5.072 | ||
| Elapidae | −5.459 | ||
| Homalopsidae | −5.324 | ||
| Lamprophiidae | −5.552 | ||
| Pythonidae | −5.177 | ||
| Viperidae | −5.110 | ||
| Scolecophidia | −5.642 | ||
| Viviparous | 0.069 | ||
| Cathemeral | −0.030 | ||
| Nocturnal | −0.057 | ||
| Aquatic | 0.163 | ||
| Arboreal | −0.312 | ||
| Burrower | 0 | ||
| Terrestrial-arboreal | −0.068 |
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.668 | ||
| Colubridae | −5.510 | ||
| Boidae | −5.072 | ||
| Elapidae | −5.459 | ||
| Homalopsidae | −5.324 | ||
| Lamprophiidae | −5.552 | ||
| Pythonidae | −5.177 | ||
| Viperidae | −5.110 | ||
| Scolecophidia | −5.642 | ||
| Viviparous | 0.069 | ||
| Cathemeral | −0.030 | ||
| Nocturnal | −0.057 | ||
| Aquatic | 0.163 | ||
| Arboreal | −0.312 | ||
| Burrower | 0 | ||
| Terrestrial-arboreal | −0.068 |
Table 3.
Parameter estimates for nonphylogenetic model. Slope and intercept values for oviparous, diurnal and terrestrial species in each family. The intercept value differences should be added for the calculation of other traits [for example, the intercept of a viviparous (0.069), nocturnal (–0.057) and arboreal (–0.312) viperid is –5.110 + 0.069 – 0.057 – 0.312]
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.668 | ||
| Colubridae | −5.510 | ||
| Boidae | −5.072 | ||
| Elapidae | −5.459 | ||
| Homalopsidae | −5.324 | ||
| Lamprophiidae | −5.552 | ||
| Pythonidae | −5.177 | ||
| Viperidae | −5.110 | ||
| Scolecophidia | −5.642 | ||
| Viviparous | 0.069 | ||
| Cathemeral | −0.030 | ||
| Nocturnal | −0.057 | ||
| Aquatic | 0.163 | ||
| Arboreal | −0.312 | ||
| Burrower | 0 | ||
| Terrestrial-arboreal | −0.068 |
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.668 | ||
| Colubridae | −5.510 | ||
| Boidae | −5.072 | ||
| Elapidae | −5.459 | ||
| Homalopsidae | −5.324 | ||
| Lamprophiidae | −5.552 | ||
| Pythonidae | −5.177 | ||
| Viperidae | −5.110 | ||
| Scolecophidia | −5.642 | ||
| Viviparous | 0.069 | ||
| Cathemeral | −0.030 | ||
| Nocturnal | −0.057 | ||
| Aquatic | 0.163 | ||
| Arboreal | −0.312 | ||
| Burrower | 0 | ||
| Terrestrial-arboreal | −0.068 |
Viviparous colubrids were heavier, for their lengths, than oviparous colubrids (intercept difference = 0.143, t = 2.341, P = 0.02), and diurnal species were heavier than nocturnal species (intercept difference = 0.089, t = 2.346, P = 0.02). Differences concerning venomousness (intercept difference = 0.04, t = 1.250, P = 0.21) and microhabitat use (aquatic species being heavier than others and arboreal species being lighter than others, F 1,4 = 14.435, P ≪ 0.001) were qualitatively similar to those found in the previous analysis. This model explained 89.9% of the variation in colubrid mass.
Phylogenetic model
The phylogenetic model included mass–SVL data for 241 species (Appendix S2). The maximum likelihood value of λ (0.557) was significantly different from both zero and unity (P ≪ 0.001). The best model for mass included SVL (slope = 2.677, SE = 0.074), mode of reproduction (viviparous species heavier than oviparous species; intercept difference = 0.118, P = 0.015) and microhabitat use, with arboreal species lighter than others (P ≪ 0.001 in all cases) and burrowing and aquatic species heavier than terrestrial (P = 0.018 and 0.05, respectively) and terrestrial–arboreal (P = 0.01 and 0.03, respectively) species. Activity time was not correlated with mass, although the tendency was similar to that of the nonphylogenetic model (i.e. nocturnal species were lighter than diurnal species). The model accounted for 86.9% of the variation in snake mass. There were no interactions between SVL and other factors. Parameter estimates of the best phylogenetic model are shown in Table 4.
Table 4.
Parameter estimates for phylogenetic model. Slope and intercept values for oviparous species in different microhabitats. The value of 0.118 should be added to the intercept values for the calculation of the intercept value of viviparous species in each microhabitat (for example, the intercept value of a viviparous arboreal species is −5.699 + 0.118)
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.677 | ||
| Arboreal | −5.699 | ||
| Aquatic | −5.326 | ||
| Burrower | −5.329 | ||
| Terrestrial–arboreal | −4.488 | ||
| Terrestrial | −5.451 | ||
| Viviparous | 0.118 |
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.677 | ||
| Arboreal | −5.699 | ||
| Aquatic | −5.326 | ||
| Burrower | −5.329 | ||
| Terrestrial–arboreal | −4.488 | ||
| Terrestrial | −5.451 | ||
| Viviparous | 0.118 |
Table 4.
Parameter estimates for phylogenetic model. Slope and intercept values for oviparous species in different microhabitats. The value of 0.118 should be added to the intercept values for the calculation of the intercept value of viviparous species in each microhabitat (for example, the intercept value of a viviparous arboreal species is −5.699 + 0.118)
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.677 | ||
| Arboreal | −5.699 | ||
| Aquatic | −5.326 | ||
| Burrower | −5.329 | ||
| Terrestrial–arboreal | −4.488 | ||
| Terrestrial | −5.451 | ||
| Viviparous | 0.118 |
| Intercept | Intercept difference | ||
|---|---|---|---|
| Slope | 2.677 | ||
| Arboreal | −5.699 | ||
| Aquatic | −5.326 | ||
| Burrower | −5.329 | ||
| Terrestrial–arboreal | −4.488 | ||
| Terrestrial | −5.451 | ||
| Viviparous | 0.118 |
Predictive ability of the equations
We calculated mass for 302 species. Our calculated mass did not differ significantly from the real mass (paired t-test, t = 0.466, P = 0.641, see Appendix S4 for detailed results), whereas the mass calculated using Pough's equation did (paired t-test, t = −13.052, P ≪ 0.001). The mass predicted by our equations deviated from the actual mass by 13.53%, on average, vs. a mean deviation of 19.83% with Pough's equation.
Discussion
SVL and TL are the two most common measures for snake body size, and both are used in ecological and evolutionary studies of snake body size. Snake mass is seldom reported in herpetological, ecological and evolutionary studies. This is despite its probability of being a better predictor for body size and a measure which allows comparisons among taxa with different body shapes (Hedges, 1985; Meiri, 2010).
Here, we developed clade and family allometric equations for the calculation of mass from SVL and TL, and found that the former is generally a better predictor of mass (Table 2). This is somewhat paradoxical, because TL measures more of the animal than does SVL. Among legless squamates, snakes have relatively short tails (the 'short-tailed burrowing morph'; Wiens, Brandley & Reeder, 2006; Sites, Reeder & Wiens, 2011). Nonetheless, relative tail length varies among species. Tail lengths can vary greatly between individuals of the same species (for example, male snakes have longer tails than females; Kaufman & Gibbons, 1975), between species inhabiting different microhabitats (Guyer & Donnelly, 1990; Martins et al., 2001) and between lineages. Thus, we attribute the differences in the predictive ability to a high variation in tail length between species with similar SVLs. Blindsnakes, a lineage of burrowing snakes, have distinctively short tails ( Wiens et al., 2006; O'Shea, 2007), and this explains the similar predictive ability of SVL and TL in this lineage. Colubrids, however, have long tails (on average, ∼30% of SVL in our data), and their tail length/SVL ratio differs greatly between species. The SVLs of colubrids active in different microhabitats are different (F 4,116 = 9.08, P ≪ 0.001), but those of terrestrial (N = 64, mean SVL of 488 mm) and arboreal (N = 14, mean SVL of 653 mm) colubrids are quite similar [Tukey's honestly significant difference (HSD) test, P = 0.18], but the former have significantly shorter tails (127 mm vs. 270.6 mm, 26% vs. 41% of SVL; Tukey's HSD test, P ≪ 0.001).
The differences in body shape between families are substantial (note the R 2 gains when we correct for family, Table 2). This suggests that the equations of Pough (1980) may be too general because they do not consider phylogenetically and ecologically related body shape differences between lineages. Furthermore, the 95% confidence interval of the slope found for all snakes in our study (for both SVL and TL, Table 1) is entirely below three (indicating that long snakes are more slender than short snakes) and does not incorporate the slope of 3.02 of Pough. It is thus not surprising that the masses calculated by the application of Pough's equations to our SVL data were significantly different from the actual masses of these snakes (Appendix S4).
Mass and hence body shape are influenced by factors other than length. Phylogeny, natural history and ecology are undoubtedly important drivers of shape variation ( França et al., 2008). In addition to phylogeny, we found that the mode of reproduction and microhabitat use and, to a lesser extent, diel activity play an important role in the relationship between mass and length. Our specific parameter estimates (Table 3) enable the calculation of the mass of a snake with known traits. For example, an oviparous, diurnal, terrestrial colubrid (e.g. Liophisreginae, SVL = 438.5 mm; Appendix S1) is expected to weigh approximately 34.6 g (log mass = 2.668 × log SVL – 5.510), whereas a viviparous, diurnal, terrestrial viperid of roughly the same length (e.g. Crotalus pricei, SVL = 421.3 mm; Appendix S1) is predicted to weigh about 91.4 g (log mass = 2.668 × log SVL – 5.110 + 0.069). An oviparous, burrower, nocturnal lamprophiid species of similar SVL, such as Atractaspis irregularis (SVL = 405 mm, Appendix S1), is predicted to weigh just 22.2 g (log mass = 2.668 × log SVL – 5.552 – 0.057).
The relationship between snake microhabitat use and body shape is well known (Pough & Groves, 1983; Guyer & Donnelly, 1990; see above). In agreement with former studies (Lillywhite & Henderson, 1993; Martins et al., 2001; França et al., 2008), we found that arboreal species are relatively lighter for their lengths than terrestrial or aquatic species. This shift in body form from short-tailed, heavy bodied snakes to long-tailed slender snakes may reflect the need of arboreal snakes to move on light branches and to manoeuvre across vegetation gaps (Lillywhite & Henderson, 1993; Martins et al., 2001). The long tail of arboreal forms may be an adaptation to greater manoeuvring ability among branches (Lillywhite & Henderson, 1993). Such a long tail is not required for terrestrial locomotion, and terrestrial species probably require the extra musculature for increased locomotion and prey handling efficiency. The (all aquatic) Homalopsid snakes have long and narrow forebodies and bulky posteriors (Murphy, 2007). Sea snake bodies (Elapidae: Hydrophiinae) are high for their sizes (Brischoux & Shine, 2011). This shape probably helps aquatic species to stabilize their bodies in the water and to facilitate locomotion in water (Murphy, 2007; Brischoux & Shine, 2011). We suggest that the robustness of aquatic species may also contribute to heat retention in a medium with high levels of heat conductivity. This assumption is supported by the findings of Spotilaa et al. (1973) that, the heavier a reptile, the less its body temperature fluctuates (whereas they found that length had a minor effect on body temperature).
Contrary to our prediction, we found no difference between venomous and nonvenomous snakes. Although this result may indeed reflect reality, we assume that it may well result from different clade-specific body shapes. Viperids and elapids are all venomous, but viperids are heavier than elapids of similar lengths (the viperid slope is steeper, and the two family lines intersect at 22.5 mm, way below the minimum length of snakes; Table 1, Fig. 1). We attribute this to their different foraging strategies. Most elapids are active hunters, whereas most viperids are ambush predators (Greene, 1997). Foraging mode may well be an important factor determining snake shape, with active foragers lighter than ambush predators (Greene, 1997; Ford & Hampton, 2009), perhaps in order to reduce the costs of locomotion. Furthermore, differences in body size and in mass–length relationship may also result from selection on other traits, such as on gape size, or reflect adaptations to hunt different prey types ( Martins et al., 2001; Pyron & Burbrink, 2009).
Nocturnal snakes are generally lighter than diurnal species, contradicting our hypothesis (and mirroring the pattern found for lizards; Meiri, 2010). We predicted that nocturnal species would be heavier than diurnal species for the purpose of better retaining heat. This result, however, may indicate that the importance of heat absorption is higher than the gain of heat retention. However, because we have no data on the thermal environment to which the species are exposed (e.g. temperature, solar radiation), this conclusion should be taken with caution.
In congruence with our hypothesis, we found that viviparous species are heavier than oviparous species (Fig. 2). We attribute this to the differences in the development stage of the eggs and the period of time needed to retain the eggs. Being elongated, snakes have limited abdominal space to hold their litter or eggs ( Bonnet et al., 2000). Oviparous species usually retain their eggs for less than 50% of the embryonic development (Shine, 1983; DeMarco, 1993), and eggs absorb water and become larger and heavier after they are laid (Qualls & Shine, 1995). Viviparous species, however, retain their eggs until full development, and may thus require more abdominal space for their developing young.
Length is considered by some authors (e.g. Boback, 2003; Boback & Guyer, 2003) to be a better proxy than mass for snake size, but length data alone may fail to provide any significant information about the species biology or life history characters (Vitt, 1987). Others argue that the high correlation between the two makes them equally valuable (Kaufman & Gibbons, 1975; Guyer & Donnelly, 1990); however, as shown here, these correlations are influenced by phylogeny, ecology and life history traits. We suggest, however, that mass, and not length, is a superior proxy for size in most ecological and evolutionary contexts regarding snake body size, especially when comparing taxa that differ in shape. If weights are estimated from length using specific equations, such as those developed here (Tables 1 and 3), rather than from actual body masses, body condition, reproductive condition and feeding will not affect mass estimates. We show here that different snake clades have different allometries for the relationship between mass and length. These allometries reflect the differences in body shape that we attribute to phylogeny as well as to ecological and life history traits. We believe that the use of length as a measure for comparison between clades or between ecological and life history traits may poorly reflect reality. For example, oviparous (N = 214) and viviparous (N = 104) snakes have similar SVLs (529 and 526 mm, respectively), and thus similar masses, when mass is predicted by Pough's equation (104 g vs. 106 g, t = 0.085, P = 0.932). Viviparous species, however, are heavier than oviparous species, both when we compare their actual masses (103.5 g vs. 52.4 g) and the masses calculated using our equations (96.8 g vs. 56.1 g; calculated mass: t = –2.711, P = 0.007; actual mass: t = –3.619, P < 0.0001). This result is congruent with the tendency found (Fig. 2) for viviparous species to be heavier than oviparous species of the same SVL.
Although mass is an extremely important measure of snake biology, mass data are still seldom reported. We thus urge researchers to publish mass data as much as possible, especially for the less studied clades (e.g. Uropeltidae, Tropidophiidae). Using mass rather than length in ecological and evolutionary studies is likely to prove highly valuable, and increase our ability to arrive at meaningful conclusions across lineages.
Acknowledgements
We thank Erez Maza for great help with collection of the data from the Tel Aviv University Natural History Museum. Erez also helped us, together with Barak Levi, to measure live snakes. We gratefully thank Marinus Hoogmoed, Tiffany Doan, Jossehan Galúcio da Frota, Gleomar Fabiano Maschio, Alessandro Costa Menks Dale Nimmo and Otavio Marques for providing us with invaluable data. Three anonymous reviewers contributed helpful comments on an earlier version of this paper.
References
Amarello
M
, Novak EM Taylor EN Schuett GW Repp RA Rosen PC Hardy DL
2010
.
Potential environmental influence on variation in body size and sexual size dimorphism among Arizona populations of the western diamond-backed rattlesnake (Crotalus atrox)
.
Journal of Arid Environments
74
:
1443
–
1449
.
Ashton
KG
, Feldman CR
2003
.
Bergmann's rule in nonavian reptiles: turtles follow it, lizards and snakes reverse it
.
Evolution
57
:
1151
–
1163
.
Aubret
F
, Shine R
2009
.
The origin of evolutionary innovations: locomotor consequences of tail shape in aquatic snakes
.
Functional Ecology
22
:
317
–
322
.
Boback
SM
2003
.
Body size evolution in snakes: evidence from island population
.
Copeia
2003
:
81
–
94
.
Boback
SM
, Guyer C
2003
.
Empirical evidence for an optimal body size in snakes
.
Evolution
3
:
195
–
211
.
Bonfim
FS
, Diniz-Filho JAF Bastos RP
1998
.
Spatial patterns and the macroecology of South American viperid snakes
.
Revista Brasileira de Biologia
58
:
97
–
103
.
Bonnet
X
, Naulleau G Shine R Lourdais O
2000
.
Reproductive versus ecological advantages to larger body size in female snakes, Vipera aspis
.
Oikos
89
:
509
–
518
.
Brischoux
F
, Shine R
2011
.
Morphological adaptations to marine life in snakes
.
Journal of Morphology
272
:
566
–
572
.
Bronikowski
A
, Vleck D
2010
.
Metabolism, body size and life span: a case study in evolutionarily divergent populations of the garter snake (Thamnophis elegans)
.
Integrative and Comparative Biology
50
:
880
–
887
.
Brown
JH
, Gillooly JF Allen AP Savage VM West GB
2004
.
Toward a metabolic theory of ecology
.
Ecology
85
:
1771
–
1789
.
Calder
WA
1984
.
Size, function and life history
.
Cambridge, MA
:
Harvard University Press
.
Cope
ED
1887
.
The origin of the fittest
.
New York
:
Appleton
.
Cox
CL
, Boback SM Guyer C
2011
.
Spatial dynamics of body size frequency distributions for North American squamates
.
Evolutionary Biology
38
:
453
–
464
.
DeMarco
V
1993
.
Estimating egg retention times in sceloporine lizards
.
Journal of Herpetology
27
:
453
–
458
.
Ford
NB
, Hampton PM
2009
.
Ontogenetic and sexual differences in diet in an actively foraging snake, Thamnophis proximus
.
Canadian Journal of Zoology
87
:
254
–
261
.
Foster
JB
1964
.
Evolution of mammals on islands
.
Nature
202
:
234
–
235
.
França
FGR
, Mesquita DO Nogueira CC Araujo AFB
2008
.
Phylogeny and ecology determine morphological structure in a snake assemblage in the central Brazilian Cerrado
.
Copeia
2008
:
23
–
28
.
Fry
BG
, Casewell NR Wüster W Vidal N Young B Jacksone TNW
2012
.
The structural and functional diversification of the Toxicofera reptile venom system
.
Toxicon
60
:
434
–
448
.
Glazier
DS
2009
.
Ontogenetic body-mass scaling of resting metabolic rate covaries with species-specific metabolic level and body size in spiders and snakes
.
Comparative Biochemistry and Physiology Part A: Molecular and Integrative Physiology
153
:
403
–
407
.
Greene
HW
1997
.
Snakes – the evolution of mystery in nature
.
Berkeley, CA
:
University of California Press
.
Griffiths
D
2012
.
Body size distributions in North American freshwater fish: large-scale factors
.
Global Ecology and Biogeography
21
:
383
–
392
.
Guyer
C
, Donnelly MA
1990
.
Length–mass relationships among an assemblage of tropical snakes in Costa Rica
.
Journal of Tropical Ecology
6
:
65
–
76
.
Heatwole
HF
, Taylor J
1987
.
Thermal ecology
.
Ecology of reptiles
.
Chipping Norton
:
Surrey Beatty & Sons
,
21
–
96
.
Hedges
SB
1985
.
The influence of size and phylogeny on life history variation in reptiles: a response to Stearns
.
The American Naturalist
126
:
258
–
260
.
Hedges
SB
2008
.
At the lower size limit in snakes: two new species of threadsnakes (Squamata: Leptotyphlopidae: Leptotyphlops) from the Lesser Antilles
.
Zootaxa
1841
:
1
–
30
.
Hutchinson
GE
, MacArthur RH
1959
.
A theoretical ecological model of size distributions among species of animals
.
American Naturalist
93
:
117
–
125
.
James
FC
1970
.
Geographic size variation in birds and its relationship to climate
.
Ecology
51
:
365
–
390
.
Kardong
KV
1979
.
'Protovipers' and the evolution of snake fangs
.
Evolution
33
:
433
–
443
.
Kaufman
GA
, Gibbons JW
1975
.
Weight–length relationships in thirteen species of snakes in the southeastern United States
.
Herpetologica
31
:
31
–
37
.
Kleiber
M
1947
.
Body size and metabolic rate
.
Physiological Review
27
:
511
–
541
.
Lee
MSY
, Hugall AF Lawson R Scanlon JD
2007
.
Phylogeny of snakes (Serpentes): combining morphological and molecular data in likelihood, Bayesian and parsimony analyses
.
Systematics and Biodiversity
5
:
371
–
389
.
Lillywhite
HB
1987
.
Temperature, energetic, and physiological ecology
. In: Seigal RA Collins JT Novak SS
Snakes: ecology and evolutionary biology
.
New York
:
Macmillan Publishing Company
,
422
–
477
.
Lillywhite
HB
, Henderson RW
1993
.
Behavioral and functional ecology of arboreal snakes
. In: Seigel RA Collins JT
Snakes: ecology and behavior
.
New York
:
McGraw-Hill
,
1
–
48
.
Madsen
T
, Shine R
1993
.
Phenotypic plasticity in body sizes and sexual size dimorphism in European grass snakes
.
Evolution
47
:
321
–
325
.
Martins
M
, Araujo MS Sawaya RJ Nunes R
2001
.
Diversity and evolution of macrohabitat use, body size and morphology in a monophyletic group of Neotropical pitvipers (Bothrops)
.
Journal of Zoology
254
:
529
–
538
.
Meiri
S
2010
.
Length–weight allometries in lizards
.
Journal of Zoology
281
:
218
–
226
.
Murphy
JC
2007
.
Homalopsid snakes: evolution in the mud
.
Malabar, FL
:
Kreiger Publishing Company
.
Nagy
KA
2005
.
Field metabolic rate and body size
.
The Journal of Experimental Biology
208
:
1621
–
1625
.
O'Shea
M
2007
.
Boas and pythons of the world
.
London
:
New Holland Publishers
.
Olalla-Tárraga
MA
, Rodrıguez MA Hawkins BA
2006
.
Broad-scale patterns of body size in squamate reptiles of Europe and North America
.
Journal of Biogeography
33
:
781
–
793
.
Olden
JD
, Hogan ZS Zanden MJV
2007
.
Small fish, big fish, red fish, blue fish: size-biased extinction risk of the world's freshwater and marine fishes
.
Global Ecology and Biogeography
16
:
694
–
701
.
Olson
V
, Davies RG Orme CDL Thomas GH Meiri S Blackburn TM Gaston KJ Owens IPF Bennet PM
2009
.
Global biogeography and ecology of body size in birds
.
Ecology Letters
12
:
249
–
259
.
Pagel
M
1999
.
Inferring the historical patterns of biological evolution
.
Nature
401
:
877
–
884
.
Peters
HR
1983
.
The ecological implications of body size
.
New York
:
Cambridge University Press
.
Pincheira-Donoso
D
, Fox SF Scolaro JA Ibargüengoytía N Acosta JC Corbalán V Medina M Boretto J Villavicencio HJ Hodgson DJ
2011
.
Body size dimensions in lizard ecological and evolutionary research: exploring the predictive power of mass estimation equations in two Liolaemidae radiations
.
The Herpetological Journal
21
:
35
–
42
.
Pough
FH
1980
.
The advantage of ectothermy for tetrapods
.
The American Naturalist
115
:
92
–
112
.
Pough
FH
, Groves JD
1983
.
Specialization of the body form and food habits of snakes
.
American Zoologist
23
:
443
–
454
.
Pyron
RA
, Burbrink FT
2009
.
Body size as a primary determinant of ecomorphological diversification and the evolution of mimicry in the lampropeltinine snakes (Serpentes: Colubridae)
.
Journal of Evolutionary Biology
22
:
2057
–
2067
.
Pyron
RA
, Burbrink FT Colli GR de Oca ANM Vitt LJ Kuczynski CA Wiens JJ
2011
.
The phylogeny of advanced snakes (Colubroidea), with discovery of a new subfamily and comparison of support methods for likelihood trees
.
Molecular Phylogenetics and Evolution
58
:
329
–
342
.
Qualls
CP
, Shine R
1995
.
Maternal body-volume as a constraint on reproductive output in lizards: evidence from the evolution of viviparity
.
Oecologia
103
:
73
–
78
.
R Development Core Team
2011
.
R: a language and environment for statistical computing
.
Vienna
:
R Foundation for Statistical Computing
.
Rambaut
A
2010
.
FigTree, version 1.3.1
. Edinburgh: Institute of Evolutionary Biology, University of Edinburgh.
Reed
RN
2003
.
Interspecific patterns of species richness, geographic range size, and body size among new world venomous snakes
.
Ecography
26
:
107
–
117
.
Schmidt-Nielsen
K
1984
.
Scaling. Why is animal size so important?
Cambridge
:
Cambridge University Press
.
Seigel
RA
, Ford NB
1987
.
Reproductive ecology
. In: Seigal RA Collins JT Novak SS
Snakes: ecology and evolutionary biology
.
New York
:
Macmillan Publishing Company
,
210
–
252
.
Seymor
RS
1987
.
Scaling of cardiovascular physiology in snakes
.
American Zoologist
27
:
97
–
109
.
Shine
R
1983
.
Reptilian reproductive modes: the oviparity–viviparity continuum
.
Herpetologica
39
:
1
–
8
.
Shine
R
1994a
.
Allometric patterns in the ecology of Australian snakes
.
Copeia
1994
:
851
–
867
.
Shine
R
1994b
.
Sexual size dimorphism in snakes revisited
.
Copeia
1994
:
326
–
346
.
Shine
R
, Harlow PS Keogh JS Boeadi
1998
.
The allometry of life-history traits: insights from a study of giant snakes (Python reticulatus)
.
Journal of Zoology
244
:
405
–
414
.
Sites
JW
, Reeder TW Wiens JJ
2011
.
Phylogenetic insights on evolutionary novelties in lizards and snakes: sex, birth, bodies, niches, and venom
.
Annual Review of Ecology, Evolution, and Systematics
42
:
227
–
244
.
Slip
DJ
, Shine R
1988
.
Thermoregulation of free ranging diamond pythons Morelia spilota (Serpentes, Boidae)
.
Copeia
1998
:
984
–
995
.
Smith
FA
, Lyons SK Ernest SKM Jones KE Kaufman DM Dayan T Marquet PA Brown JH Haskell JP
2003
.
Body mass of late Quaternary mammals
.
Ecology
84
:
3403
.
Spotila
JR
, Lommen PW Bakken GS Gates DM
1973
.
A mathematical model for body temperatures of large reptiles: implications for dinosaur ecology
.
The American Naturalist
107
:
391
–
404
.
Terribile
LC
, Diniz-Filho JAF Rodríguez MA Rangel TFLVB
2009
.
Ecological and evolutionary components of body size: geographic variation of venomous snakes at the global scale
.
Biological Journal of the Linnean Society
98
:
94
–
109
.
Vidal
N
, Delmas AS David P Cruaud C Couloux A Hedges SB
2007
.
The phylogeny and classification of caenophidian snakes inferred from seven nuclear protein coding genes
.
Comptes Rendus Biologie
330
:
182
–
187
.
Vidal
N
, Marin J Morini M Donnellan S Branch WR Thomas R Vences M Wynn A Cruaud C Hedges SB
2010
.
Blindsnake evolutionary tree reveals long history on Gondwana
.
Biology Letters
6
:
558
–
561
.
Vitt
LJ
1987
.
Communities
. In: Seigal RA Collins JT Novak SS
Snakes: ecology and evolutionary biology
.
New York
:
Macmillan Publishing Company
,
335
–
365
.
Wagenmakers
EJ
, Farrell S
2004
.
AIC model selection using Akaike weights
.
Psychonomic Bulletin & Review
11
:
192
–
196
.
Wiens
JJ
, Brandley MC Reeder TW
2006
.
Why does a trait evolve multiple times within a clade? Repeated evolution of snakelike body form in squamate reptiles
.
Evolution
60
:
123
–
141
.
© 2012 The Linnean Society of London
© 2012 The Linnean Society of London
Source: https://academic.oup.com/biolinnean/article/108/1/161/2415670
0 Response to "The Length of a Snake is It Continuous and Countable"
Post a Comment