The Ecology of Morphology: The Ecometrics of Locomotion and Macroenvironment in North American Snakes

  • A. Michelle Lawing
  • Jason J. Head
  • P. David Polly
Chapter
Part of the Springer Earth System Sciences book series (SPRINGEREARTH)

Abstract

Morphological traits that have a functional relationship with the environment can be used to study relationships between organisms and environments through time and across space. Dynamics of the trait-environment complex can be studied with ecometrics in individuals, in populations, and in communities. We explored how closely correlated three skeletal traits are with substrate use, and thus macrohabitat, among communities of snakes with the goal of better understanding how climate and macrovegetation might affect snake assemblages. Substrate use explained a large part of the variance in mean length-to-width ratio of vertebrae (R2 = 0.66), PC1 of vertebral shape of a mid trunk vertebra (R2 = 0.46), and relative tail length (R2 = 0.71). Furthermore, mean relative tail length in snake assemblages across North America is strongly associated with ecoregions and vegetation cover (R2 = 0.65 and 0.47, respectively). The close relationship with macrovegetation makes relative tail length a useful tool for predicting how snake assemblages will change as climates and biomes change across space or through time. This “ecometric” approach provides a medium-scale link between data collected from ecological studies over decades to data assembled from the fossil record over thousands, tens of thousands, or even millions of years. We show how historical vegetation changes between the early twentieth and twenty-first centuries at five preserves in North America resulted in ecometric changes that parallel the geographic distribution of relative tail length in snake communities across North America.

Keywords

Climate Community morphology Ecometrics Ecomorphology Geographic variation Paleoenvironment Serpentes Taxon-free Vertebrae 

7.1 Introduction

Assessing the impact of climate and habitat change on local communities is often hampered by lack of long-term historical baselines from which to judge changes in habitats and species (Willis et al. 2005; Tingley and Beissinger 2009). To truly understand the dynamics of changing climates, landscapes, and biotas an understanding of patterns is needed at local, landscape, regional and continental geographic scales, as well as seasonal, annual, decadal, millennial, and geological time scales (McGill 2010; Wiens and Bachelet 2010). In this paper we present an ecometric analysis – a broad-scale, community-based functional trait analysis applicable to patterns in modern ecological systems as well as paleontological systems – with the goal of linking patterns on broad geographical scales with patterns relevant to local conservation efforts. Our focus is on North American snakes, a group whose ectothermic physiology links them closely with climate (Webb and Shine 1998; Head et al. 2009), are known to have had pronounced geographical responses to Quaternary climate cycles and that are expected to have large responses to climate change over the next century (Lawing and Polly 2011), shows global signs of population decline (Reading et al. 2010), and whose ecology and conservation are of intense interest (Mullin and Seigel 2009). In our study we use ecometrics as a tool for middle-scale analysis linking continental-scale range and biome data with local historical surveys of changing land cover and species conservation. The same kind of analysis has the potential for quantifying temporal changes using the fossil record and to compare rates and magnitudes of faunal changes observed over the past 50–100 years with changes that occurred with climate cycles during the Quaternary.

Ecomorphology is the study of trait variation within and between species as it relates to climate and environment (Ricklefs and Miles 1994). An analysis of ecomorphological traits at the community level is important for integrating spatial and temporal data to understand biological responses to climate change (Eronen et al. 2010a). Community-wide trait variation that relates to climate and environment is termed community or faunal ecomorphology (Polly 2010). The signature of within species ecomorphology varies greatly between species and is not consistently reliable for inferring changes in the environment (Polly 2003; Barnosky et al. 2004); however, variation in community ecomorphology – which captures the average response of all species in the community – should be a stronger indicator of environment, because it reflects the combination of adaptive responses, dynamic geographic range shifts, and extinction (Polly 2010 and references within).

The quantitative community ecomorphology approach, also known as “ecometrics”, models the relationship between trait variation and environment in a taxon-free, trait-based, quantitative and testable manner (reviewed in Eronen et al. (2010a)). The taxon-free nature of ecometrics allows for a systems approach to address the relationship between community structure and environment without having to focus on a limited species group. Even ‘taxon-free’ ecometric traits, like all morphological traits, have phylogenetic histories as well as environmental correlations (Little et al. 2010), but phylogenetic and environmental correlations are not mutually exclusive (Caumul and Polly 2005). The phylogenetic acquisition of traits may often result from adaptation to new environmental or climatic conditions that today make up the geographic matrix of modern environments and thus are intimately related to contemporary correlations between trait and environment. Environmental controls on the geographic distribution of taxa may result in non-random phylogenetic structure in communities, either mixing unrelated species with different traits or favoring single clades with uniform traits (Kraft et al. 2007; Emerson and Gillespie 2008). Removing the effects of phylogeny will contribute to the identification of the variation in a trait that is not influenced by relatedness, but in cases where environmental sorting is based on traits that are shared phylogenetically (e.g., Jablonski 2008) the uncritical removal of phylogenetic effects risks removing the environmental effects that one is hoping to measure (Kozak et al. 2005; Helmus et al. 2007). Measuring traits instead of species occurrences in a community facilitates more robust quantitative comparisons spatially and temporally (Damuth et al. 1992; Webb et al. 2010). In addition, ecometric models developed with data from the ecological present can be validated with the fossil record and, contrarily, ecometric indices developed with data from the geological past can be validated with the ecological present.

Ecometric tools have been developed for plant and mammal communities. For example, the proportion of untoothed leaf margins in a plant community is positively correlated with mean annual temperature (Wolfe 1979, but see Little et al. 2010), the average tooth crown height in a mammalian herbivore community is a proxy for aridity (Janis and Fortelius 1988; Fortelius et al. 2002) and the average ankle posture in a mammal carnivore community is highly associated with its ecological province, vegetation cover, and mean annual temperature (Polly 2010). Paleontologists can use these community ecomorphologic indices in fossil assemblages as a tool for reconstructing paleoenvironment (Polly et al. 2011).

Herpetofauna have the potential to offer precise estimates of paleoclimate and paleoenvironment. As ectotherms whose internal body temperatures are regulated primarily by environmental temperatures, amphibian and reptile diversity, distributions, and ecology are ultimately limited by climatic parameters. For example, species richness in amphibians is explained primarily by the amount of energy and water available in the environment and in reptiles is explained mostly by the amount of atmospheric energy (Rodríguez et al. 2005).

The dependency of extant herpetofaunas on climate has been previously used to infer paleoenvironments from fossils. Blain et al. (2010) use an amphibian and reptile assemblage to infer the climate and environment of the earliest West European hominins by using mutual climate ranges of the extant local herpetofauna and habitat weighting. Böhme et al. (2006) developed a method categorizing ecophysiological groups within herpetological communities that is highly associated with annual precipitation. Both of these methods require fossil taxonomic identification and assume conserved ecological restraints and roles of species through time.

An ecometric tool for reconstructing paleoclimate with snakes has been developed for body size. The maximum size an ectotherm can reach is limited by the ambient temperature in the environment and mass-specific metabolic rate (Makarieva et al. 2005). Head et al. (2009) used this relationship to determine mean annual paleotemperature of the Paleocene Neotropics based on comparisons between body size and climate in living anacondas and giant fossil boid snakes. This method of paleothermometry has been extended to other fossil reptiles (Head 2010).

Additional metrics of snake anatomy have been correlated to habit, and may represent useful ecometric tools. Lindell (1994) examined vertebral counts and body size in modern colubroid snakes and found that the number of vertebrae relative to body size predicted locomotory habits and constriction behavior. Guyer and Donnelly (1990) categorized an assemblage of tropical snakes into four morphological groups that are associated with habitat preference. Johnson (1955) recognized differences in length/width ratios between locomotory habits, but could not separate them from phylogenetic effects. Martins et al. (2001) showed body size, tail length, and stoutness is associated with macrohabitat in Neotropical pitvipers. Wiens et al. (2006) examined body/tail length ratios in snakes and other limbless squamates and found a general correspondence between relative tail length and fossorial vs. surface dwelling habits.

Here we explore three morphological traits in snakes as a potential proxy for paleoenvironmental reconstruction with a two pronged approach, ultimately developing a taxon-free ecometric. First, we investigate the relationship between three traits and substrate use. The traits are vertebral length-to-width ratio, vertebral shape and relative tail length. Substrate use is related to locomotion and is defined here categorically as arboreal, semiarboreal, terrestrial, fossorial, semifossorial, aquatic, and semiaquatic. Second, we develop relative tail length into an ecometric using North American snake communities. We show the strong relationship between the average relative tail lengths in a snake community and ecological province, vegetation cover, and mean annual temperature.

7.2 Methods

Three datasets were used in this study to examine the ecometric potential of vertebral length-to-width ratio, vertebral shape and relative tail length in snakes (Fig. 7.1). The dataset with the strongest environmental correlation was then further explored as an ecometric index using communities of North American snakes. Each dataset was derived from available trait information collected by the authors. In all analyses, we compare potential ecometric variables with environmental substrate use. Substrate use was categorized for each species as arboreal, semiarboreal, aquatic, semiaquatic, fossorial, semifossorial, or terrestrial. Due to the low sample size in some groups, we combined arboreal and semiarboreal, aquatic and semiaquatic, and fossorial and semifossorial. We note that lumping these locomotor types combines snakes with very different anatomical specializations and we discuss potential effects of this grouping on the results. The dataset associated with each trait is described below. To determine whether the traits are reliable predictors of substrate use we used Analysis of Variance (ANOVA) to test the mean ratio (corrected for evolutionary relatedness) against substrate category as the factor. To correct for relatedness, we calculated the ANOVA test statistic and then obtained a null distribution of test statistics by simulating data along a phylogenetic tree using the R package GEIGER (Garland et al. 1993). The phylogenetic hypotheses that we considered are discussed below in Sect. 7.2.5. Analyses were performed in the R Statistical Programming Language.
Fig. 7.1

Three traits explored for snake ecometrics. (a) Mean vertebral length-to-width ratio (2/1), averaged for all vertebrae within an individual; (b) Vertebral shape represented by principal components of Procrustes superimposed landmarks (n = 23); (c) Relative tail length (4/3)

7.2.1 Mean Vertebral Length-to-Width Ratio

Twenty-nine snake species were selected from across the globe to represent a taxonomically and ecologically diverse sample (see Supplementary Appendix Table 1, available from IU Scholarworks at http://hdl.handle.net/2022/14288). Length and width measurements were collected for all the vertebrae of one specimen from each of the species. These measurements were then averaged across each specimen and the mean length to mean width ratio (heretofore, mean length-to-width ratio) was tested as an ecometric trait.

7.2.2 Vertebral Shape

A mid-trunk vertebra was selected from 60 snake species and photographed in anterior view. Snake species were selected to represent a diverse taxonomic and ecological sample as well as for the availability of disarticulated skeletal material (see Supplementary Appendix Table 2, available from IU Scholarworks at http://hdl.handle.net/2022/14288). Figure 7.1b shows the landmark scheme we used to represent vertebral shape. After digitizing the landmarks, we removed the effects of location, position and size by translation, rotation and scaling of the landmarks using Procrustes superimposition (Sneath 1967; Gower 1975; Rohlf and Slice 1990). Due to the low sample size in some groups, we combined arboreal and semiarboreal, aquatic and semiaquatic, and fossorial and semifossorial.

To determine if variation in substrate use is explained by vertebral shape, we performed a MANOVA on the principal components (PCs) of Procrustes superimposed landmarks with substrate use as the factor and in a phylogenetic context (the use of PCs follow Dryden and Mardia 1998). We also performed ANOVA with each PC to further investigate each uncorrelated shape variable and its relationship with substrate use. To perform MANOVA and ANOVA, we calculated the test statistic (Wilks’ lambda and F, respectively) and then corrected for relatedness, using the method described above.

7.2.3 Relative Tail Length

The length of each species’ tail relative to its total body length was obtained from literature for 588 North American snake species. Snake species were selected to represent the full North American snake fauna (see Supplementary Appendix Table 3, available from IU Scholarworks at http://hdl.handle.net/2022/14288). Species missing relative tail lengths were a result of (1) broken or damaged tails in their species description; (2) tail length not reported in the species description; or (3) taxonomic changes make it unclear as to which species the reported tail lengths are associated with. Relative tail lengths were obtained by taking the percentage of the length of the tail to the total body length.

In many cases, tail lengths were not reported, but the scale counts for the ventrals and subcaudals were reported. To maximize the number of species used to represent the North American snake fauna, we also calculated relative tail length by taking the percentage of the max scale count for subcaudals to the max scale count for ventrals plus subcaudals. Both tail lengths and scale counts were reported for 160 snakes and the relative tail length calculated from both measurements are highly correlated (R2 = 0.89) with the slope and intercept of a fitted linear regression suggesting the usage is fairly interchangeable (slope = 1.07, intercept = 0.026). Scale counts were reported in the literature more often than tail lengths, so when both were reported we used scale counts to calculate relative tail lengths.

Snake tails are sexually dimorphic (King 2008), and tail lengths were reported for both males and females of some species. Sex-specific characteristics are not well studied in snake skeletons, so for application of this ecomorphological trait in the fossil record we grouped males and females into the same analysis. We averaged the male and female tail lengths in order to obtain an overall estimate. However, repeating the analysis with only males or females produces slightly different quantitative results, but did not change the significance or produce qualitatively different results.

7.2.4 Ecometric Analysis

Because it was the trait most closely correlated with substrate use, we further explored relative tail length as an ecometric index. Relative tail length has an added advantage for use in ecometrics because it is reported often in species accounts, making data compilation tractable for large numbers of species. Geographic distributions of species with relative tail lengths were obtained from species entries in the Catalogue of American Reptiles and Amphibians. Ranges were digitized in ArcMap 9.2. Only 494 of the 588 species had geographic distributions that were known well enough to be included in our analysis. Our data are geographically biased in that species distributions in the United States and Canada are almost completely documented, but those in Mexico, Central America, and the Caribbean less so. Nevertheless, reasonable data are available for most species with widespread distributions in these regions and we do not believe that this bias has a substantial effect on our results, except that species-level diversity may be undercounted in some of these areas. We sampled our data using a grid of equidistant points spaced 50 kmapart across the whole of North America (following Polly (2010). Grid point coordinates are available for download at http://mypage.iu.edu/~pdpolly/Data.html). This scale is similar to the geographic mixing that is likely to be present in fossil assemblages, therefore this level of sampling is appropriate for establishing patterns that can be measured both in the fossil record and the modern world (Fortelius et al. 2002). The species present at each grid point were recorded and the mean and standard deviation of their relative tail length was calculated.

We tested the association between the geographic distribution of the mean relative tail length and species number, elevation (Row and Hastings 1994), mean annual temperature (Willmott and Legates 1998), annual precipitation (Willmott and Legates 1998), local vegetation cover (Matthews 1983; Matthews 1984), and ecoregion province (Bailey 1998, 2005). See Fig. 7.2 modified from Polly (2010) for maps of these environmental variables sampled using the same 50 km grid that we used for the species assemblages. We calculated Pearson’s product moment correlation (R) for continuous variables and ANOVA-derived R2 for categorical variables. We squared species number, took the square root of elevation, and the natural log of precipitation to transform variables for normality. A randomization and bootstrap procedure was performed to calculate significance and confidence intervals for the Pearson’s product moment correlation and ANOVA (following Polly (2010)).
Fig. 7.2

Environmental variables in 50 km grid point spacing modified from Fig. 7.2 of Polly (2010). (a) Elevation is divided into 10 categories using Jenk’s natural breaks algorithm. Each category is colored to represent terrain height from the lowest elevation colored as dark green (1–173 m) to the highest elevation colored as white (2596–3660 m); (b) Mean annual temperature is divided into 10 categories using Jenk’s natural breaks algorithm. Each category is colored to represent temperature from the lowest temperature colored dark blue (−19.9°C to −12.5°C) to the highest temperature colored dark red (21.4–28.6°C); (c) Annual precipitation is divided into 10 categories using Jenk’s natural breaks algorithm. Each category is colored to represent precipitation from the lowest precipitation colored light yellow (49.1–257.4 mm) to the highest precipitation colored dark blue (2988.8–5239 mm); (d) Matthews’ vegetation cover is represented with a categorical color scheme, with some categories amalgamated as described in the text. The vegetation approximately ordered from densest vegetation colored dark green (tropical evergreen forest and subtropical evergreen seasonal broad-leaved forest) to sparsest colored bright red (ice). Letters on the map indicate the conservation sites shown in Fig. 7.8; (e) Bailey’s ecoregion provinces are represented with an arbitrary categorical color scheme

Finally, we summarized the relationship between the relative tail length and macrovegetation by creating an “ecometric space” for macrovegetation types. To do this we calculated the mean and standard deviation of relative tail length for each of the 50 km snake community assemblages and recorded the macrovegetation type at each point. We then binned the means and standard deviations into a 25 × 25 cell bivariate histogram, tallying the frequency of different macrovegetation types in each cell. The most frequent vegetation type associated with the points in the bin was taken to be the “expected” vegetation given a particular community mean and standard deviation. Macrovegetation types were modified from the Matthews vegetation cover dataset described above, but modified by amalgamating similar vegetation types when they could not be distinguished by ecometric analysis: cold-deciduous forest and cold-deciduous woodland were amalgamated into “cold-deciduous forest”; tall, medium and short grasslands were amalgamated into a single “grassland” category; meadow and forb formations were amalgamated into “meadow”; evergreen needleleaved or microphyllous shrubland/thicked and evergreen needleleaved woodland were amalgamated into “needleleaved evergreen forest”; eight categories of drought-deciduous, sclerophyllous, and xeromorphic woodlands, thickets and shrublands were amalgamated into a single category “scrub”; and tropical evergreen rainforest, tropical/subtropical evergreen needle-leaved forest, and tropical/subtropical evergreen broad-leaved forest were amalgamated into “tropical forest”. Note that the correlations between relative tail length and vegetation above were performed using Matthews’ original categories.

7.2.5 Phylogenies

We explored several phylogenetic hypotheses for the comparative analysis. For each trait, different species were included in the analysis and a relevant phylogenetic hypothesis was assembled. The assembled phylogenies were first built to be as fully resolved as possible (i.e. to the species level). However, not all species were represented in these phylogenies so we also ran an analysis where the phylogenies were only resolved to the genus level and the relationships of the species within them were treated as unresolved polytomies. The second analysis produced qualitatively similar results as the first, so we present results only from the fully resolved phylogenetic hypotheses.

We compiled phylogenetic relationships from several sources as follows. Higher interrelationships follow a morphological phylogeny compiled by Wilson et al. (2010). Changing the relationship of the dwarf boas, Tropidophiinae, and the Mexican sunbeam snake Loxocemus, to follow the molecular phylogeny slightly changes the quantitative results of this study. It does not change the significance of the relationships determined here nor does is change the qualitative results or implications of this study. Other relationships within the phylogenies follow Pyron et al. (2011), Sanders et al. (2010), Adalsteinsson et al. (2009), Burbrink (2005), Lawson et al. (2005), Slowinski and Lawson (2002), and Wilcox et al. (2002). Most within-Colubroidea relationships follow Pyron et al. (2011) with the exception of the following listed families and genera. Within Crotalinae relationships follow Castoe and Parkinson (2006) and Jadin et al. (2010). The placement of Enulius, Cristantophis, and Urotheca follow Crother (1999). The placement of species within Tantilla, Ficimia, and Gyalopion follow Holm (2008). Rhinobothryum’s placement is according to Vidal et al. (2000). Newick formatted phylogenies can be found in our Supplementary Appendix (available at http://mypage.iu.edu/~pdpolly/Data.html and from IU Scholarworks).

7.3 Results

7.3.1 Three Ecomorphological Traits

Mean vertebral length-to-width ratio and relative tail length were both significantly correlated with substrate use, both with and without respect to phylogenetic relationships (Table 7.1; Fig. 7.3). Substrate use explained 66% of the variance in mean vertebral length-to-width ratio and 71% of the variance in relative tail length. Vertebral shape represented by all of its PCs does not have a significant portion of its variation explained by substrate use; the first PC alone does, however, have a significant portion of its variation (46%) explained by substrate use (Table 7.1; Fig. 7.3). The shape described in PC1 captures 41.5% of the total variation in shape between species. No other principal component alone has a significant portion of its variation explained by substrate use. Most of the geometric shape variation was, therefore, not explained by substrate use (see below).
Table 7.1

ANOVA and MANOVA results for vertebral mean length-to-width, vertebral shape, and relative tail length. The null distribution for the phylogenetic correction was built from simulating new dataset of the dependent variable on a phylogenetic tree with a Brownian motion model of evolution. The test statistics (degrees of freedom- d.f.; F; and the associated probability that the correlation departs from zero – P) were compared from the original data and the simulated null distribution. Asterisk denotes significance at alpha equals 0.05. Adjusted R2 is the amount of explained variance by substrate use in the model (ranges from 0 to 1)

 

Error d.f.

F-value

P

R2

Vertebral length-to-width

    

Substrate use

25

4.3736

0.0132*

0.66

Phylogenetic correction

  

0.0210*

 

Vertebral shape

    

All Principal components

    

Substrate use

56

1.1534

0.3008

0.00

Phylogenetic correction

  

0.3237

 

First Principal component

    

Substrate use

56

21.684

1.90e–9*

0.46

Phylogenetic correction

  

9.99e–4*

 

Relative tail length

    

Substrate use

327

22.615

2.2e–16*

0.71

Phylogenetic correction

  

9.99e–4*

 
Fig. 7.3

Box plots of three potential ecometrics grouped by substrate use. The thick line in the center of each box represents the median, the box represents two quartiles (50% of the data), and the ticks off the box represent the range of the data. (a) Mean length-to-width ratio (n = 29); (b) PC1 of vertebral shape (n = 60); (c) Relative tail lengths (n = 588)

For the mean length-to-width ratio, the variances in the arboreal and fossorial substrate use categories do not overlap with each other or with the terrestrial and aquatic categories (Fig. 7.3a). The arboreal category has a range of ratios close to 1 (equal length and width of vertebrae) and the others have vertebrae with greater widths than lengths on average. Terrestrial and aquatic substrate categories overlap with each other, possibly due to the grouping of semi-aquatic and aquatic snakes substrate use categories. We repeated the box plots and ANOVA with the original substrate use categories and found aquatic and terrestrial overlap only in their lowest (aquatic) and highest (terrestrial) quartiles (data not shown). Including substrate use categories with very low sample sizes decreases the power of the ANOVA to detect differences in the factor and may decrease the explanatory variance; however, with this conservative approach the substrate use category still explains a significant amount of the variation (R2 = 0.41; p < 0.001).

The relationship between substrate use and PC1 of vertebral shape is mainly driven by the fossorial category, although its variation does have some overlap with the other three substrate use categories (Fig. 7.3b). Figure 7.4 shows variation in the first PC represented by the landmark constellation of a hypothetical vertebra at two positive standard deviations away from the mean and arrow tips at two negative standard deviations away from the mean. PC1 of vertebral shape is mainly characterized by the dorso-ventral flattening of the vertebra, achieved primarily through the shortening of the neural spine (landmark 6). The aquatic, arboreal, and terrestrial categories overlap extensively, but not because of the grouping of substrate categories. Box plots with the original substrate use categories show extensive overlap in all categories, except fossorial and semifossorial (data not shown). An ANOVA with the original substrate use categories is in close agreement with the grouped substrate use categories (R2 = 0.46 and R2 = 0.45, respectively).
Fig. 7.4

Variation in the first PC of vertebral shape represented by the landmark scheme of a hypothetical vertebra at two positive standard deviations away from the mean (dark circles) pointing to two negative standard deviations away from the mean (arrow tips). Each landmark is arbitrarily numbered

Like mean length-to-width ratio, the relationship between substrate use and relative tail length is strongly driven by the fossorial category (Fig. 7.3c). The terrestrial snakes overlap with most of the variation in all substrate use categories. The arboreal and semiarboreal categories have the longest tails (represented by the medians) followed closely by aquatic and semiaquatic. As expected, the fossorial snakes have the shortest tails.

We further explored relative tail length as an ecometric index because it explains the substrate use category better than the other two potential ecometrics (71%) and it will be used for the remaining analyses. In addition, relative tail length is widely available in the literature for most North American snake taxa.

7.3.2 The Geography of Morphology and Its Ecometric Implications

The geographic distribution of North American snakes covers approximately 57% of the North American continent (5,496 of 9,699 grid points are occupied by at least one species). Each North American grid point represents a potentially different community composition where communities range from 1 to 42 species, have a median of 11 species, a mean of 13.24 species, and a standard deviation of 10.29 species (Fig. 7.5a). Figure. 7.6 shows the ordering of relative tail lengths in the 494 species with associated geographic distributions. Generally, the shortest tailed snakes are fossorial and the longest tailed snakes are arboreal.
Fig. 7.5

Species diversity and relative tail length ecometrics. (a) Number of snake species within a 50 km grid community as used to calculate the relative tail length ecometric. Color scheme ranges from light blue (1–4 species) to dark red (38–42 species); (b) Relative tail lengths averaged within each 50 km grid community. Color scheme ranges from orange (longest tail averages) to dark blue (shortest tail averages); (c) Standard deviations of relative tail lengths. Darker orange represents the highest standard deviations where the lightest orange represents little or no standard deviation (sample size here is usually one). The red points are four representative locations shown in Fig. 7.7

Fig. 7.6

Rank order plot of relative tail lengths in North American snakes. Short tail lengths generally correspond to fossorial lifestyles and the longest tail lengths generally correspond to arboreal lifestyles. The taxon axis is labeled every tenth species to accommodate many species labels and to indicate the type of species in that relative tail length vicinity. Illustrated taxa from left to right are sidewinder, copperhead, garter, and common green

The geographic distribution of mean relative tail lengths is mapped in Fig. 7.5b. The mean relative tail lengths are grouped into 10 bins using Jenks natural breaks algorithm (Jenks 1977). The color scheme ranges from dark blue, representing the mean shortest tailed snakes, to dark red, representing mean longest tailed snakes. The mean relative tail lengths range from 0.11 to 0.40 and their mean is 0.28. Standard deviations for mean relative tail lengths for each 50 km community are mapped in Fig. 7.5c. Darker colors represent higher standard deviations within snake 50 km communities. Latitude generally corresponds to the distribution of the standard deviations.

Mean relative tail length is significantly correlated with species number (Fig. 7.5a), elevation (Fig. 7.2a), mean annual temperature (Fig. 7.2b), and annual precipitation (Fig. 7.2c). However, elevation and annual precipitation explain little of the variance in mean relative tail length (7 and 12%, respectively; Table 7.2). The categorical variables vegetation cover (Fig. 7.2d) and ecological province (Fig. 7.2e) explain a high amount of the variation in relative tail length (47 and 65%, respectively; Table 7.2).
Table 7.2

The relationship between mean relative tail length and climate and environment. Pearson’s product moment correlation coefficient (R) is reported for the continuous variables along with their 95% confidence intervals and significance cut-off. The significant cut-off is the maximum value R can have when there is no correlation. The amount of explained variance (R2) and its 95% confidence interval is reported for categorical variables

 

R

(95% CI)

Significance cut-off

R2

(95% CI)

Number of species

−0.58

(−0.56– − 0.60)

0.043

0.34

(0.32–0.36)

Elevation

−0.26

(−0.24– − 0.28)

0.044

0.07

(0.06–0.08)

Mean annual temperature

−0.47

(−0.45– − 0.49)

0.042

0.22

(0.20–0.24)

Annual precipitation

0.34

(0.32–0.36)

0.044

0.12

(0.10–0.13)

Vegetation cover

   

0.47

(0.45–0.49)

Ecological province

   

0.65

(0.64–0.67)

7.4 Discussion

7.4.1 Three Ecomorphological Traits

Of the three potential ecometric traits explored here, relative tail length has the closest association with substrate use. Many species descriptions report total body length and either tail lengths or scales counts from which we can derive tail lengths, making this variable attractive to develop into an ecometric. In the fossil record, total body skeletons of snakes are rarely preserved, making it difficult to estimate relative tail length for fossil taxa. A more strategic ecometric for the fossil record might therefore include the first PC of vertebral shape from the anterior view of one midtrunk vertebra, as explored in this study. The necessity of photographs of disarticulated skeletonized specimens en masse for a majority of the North American snake fauna prevented us from further developing this ecometric along with relative tail lengths. Future efforts in developing snake ecometrics should consider vertebral shape.

Phylogenetic relatedness has been shown to contribute to snake community structure (Franca et al. 2008), so we wanted to assess the phylogenetic correlation of the morphological traits to better understand the history of their association with substrate use. In all cases, controlling for phylogeny did not change the significance of the test (Table 7.1). However, this is not to say that the traits have not arisen phylogenetically, nor is it to say that substrate use in snakes is not phylogenetically structured; rather our results simply highlight that these morphological traits have a strong convergence across clades with respect to substrate use and that phylogeny does not complicate the interpretation of the environmental signal derived from the community-level averages in the traits.

7.4.2 The Geography of Morphology and Its Ecometric Implications

Mean relative tail length has a strong geographic structure and varies with climate and the environment. The shortest-tailed communities are distributed in the desert southwest and in central Mexico, where the longest-tailed communities are distributed in Mexican coastal regions, Central American, the Caribbean and at the northern extent of North American snake distributions in Canada. High mean relative tail length in the northern parts of North America is partially an artifact of extremely low species number. Throughout most of the northernmost region the snake communities consist only of Thamnophis sirtalis, which has a relative tail length of 0.33. Determining whether the lack of fossorial snakes at high latitudes is due to ecometric sorting or to coincidental physiological specialization for cold-resistance in a relatively long-tailed species (cf. Fig. 7.6) is beyond the scope of our data. Generally, within-community variation in relative tail length is correlated with latitude (the further south, the higher the standard deviation within a grid point), with exception of the northwestern and western steppe and mountain provinces.

Fischer (1960) suggested that species diversity increases with decreasing latitude, citing that species numbers in Canada, the United States and Mexico are 22, 126, and 293, respectively. Although species number increases by country, diversity on a smaller scale may show a different signal due to size of geographic distributions. We mapped the number of species within North American snake communities with a spacing of 50 km points across North America (Fig. 7.5a). Within the United States and Canada, species number increases with latitude, but farther south, species number is higher in coastal regions and lower inland suggesting that proximity to the moist Gulf of Mexico is as important a driver as latitudinal temperature gradients. The species diversity we present reflects our 50 km sampling and does not necessarily reflect the exact species diversity at any given field locality; however our data are likely to be a good proxy for local community diversity observed in the field over areas of several square kilometers.

As examples, the community distribution of ecometric variation among local snake taxa are compared directly with the local climate at four points from different regions of North America to illustrate the association between climate, environment and ecometric (Fig. 7.5a). The within-community mean relative tail lengths are shown in Fig. 7.7. Point (a) is in the Canadian province Ontario, close to the northern extent of all snake distributions in North America. This location, which is part of Bailey’s Warm Continental Division, has mixed deciduous and coniferous forest vegetation cover, warm humid summers and cold winters. Two snakes occur at this sampling point with a mean relative tail length higher than average and a low standard deviation. Point (b) is in the state of Missouri in the United States. It is classified in the Prairie Division of Bailey’s ecoregion classification, has forests with steppes and prairies, has cold winters and has hot and humid summers. Twenty-five species occurred at this sampling point that have a mid-range mean and standard deviation for relative tail length. Point (c) is in the state of Arizona in the United States. This point is part of the Tropical/Subtropical Desert Division of Bailey’s ecoregion classification; it has mainly deserts on sand, mild winters, hot summers and is arid year around. We sampled 19 species in this snake community, which has a very low mean relative tail length with a mid ranging standard deviation. Point (d) is in the department of Olancho in Honduras. This point in Olancho is categorized as Bailey’s Savanna Mountains and it is covered with forest steppe; it has a consistently hot temperature year around and has highly seasonal wet and dry periods. We sampled 11 species in this snake community, which has a higher than average mean relative tail length and a high standard deviation.
Fig. 7.7

Four representative locations showing the association between species community composition, relative tail lengths within that community, climate and ecoregion. (a) A representative cold-climate fauna. Ontario, Canada has two species for the representative grid point with 0.3 mean relative tail length; (b) A representative temperate climate fauna. Missouri, USA has 25 species for the representative grid point with 0.27 mean relative tail length; (c) A representative hot-weather climate fauna. Arizona, USA has 19 species for the representative grid point with 0.18 mean relative tail length; (d) A representative seasonally wet tropical climate fauna. Olancho, Honduras has 11 species for the representative grid point with 0.3 mean relative tail length. The right side panels depict mean monthly temperatures (red line with brown fill) and precipitations (blue line and fill). Also reported in each graph are Bailey’s ecoregion division and province, latitude, longitude, mean annual temperature and precipitation (see Fig. 7.5a for geographic locations). In (b) and (c), taxa are labeled every other species to accommodate many labels and to indicate the type of species in the vicinity of relative tail lengths

Mean relative tail length is significantly correlated with species number, elevation, mean annual temperature, and annual precipitation. In addition, a high amount of the variation in mean relative tail length is explained by vegetation cover (47%) and ecological province (65%). Interestingly, these results are similar in magnitude to the ecometric related to average posture in mammalian carnivores. Polly (2010) found vegetation cover explained 49% and ecological province explained 70% of the variation in average posture. Low postures correspond to plantigrade stance and high posture corresponds to digitigrade stance. Comparing the geographic distribution of average posture (Fig. 7 in Polly 2010) with the geographic distribution of relative tail length (Fig. 7.5), the shortest tailed snake communities are distributed with the highest postured mammalian carnivore communities.

Olalla-Tarraga et al. (2006) mapped the geographic distribution of average snake body size in North America and Canada and showed that mean annual temperature, potential evapotranspiration, and elevation are the best descriptors of the distribution of average snake body size. Their map of average body sizes showed that the largest average body sizes are distributed in northwestern North America, specifically in the Rocky Mountain, Columbia Plateau, Basin and Range, and Colorado Plateau Geological Provinces. Those authors note that the geographic area with the largest average body sizes also has the highest variation. Those regions correspond roughly to the areas of highest variation in relative tail length on our maps (Fig. 7.5c). We found higher standard deviations in Mexico and Central America, but could not compare these areas with Olalla-Tarraga et al.’s results because their analysis was limited to the United States and Canada. These southern regions of North America have a very complex topography compared to the United States and Canada and foster complex and variable microhabitats. This complexity probably contributes to the high variability of average body sizes and relative tail lengths.

7.4.3 Macrovegetation and Community Ecometrics

Because of the close correlation between the distribution of relative tail length in local snake assemblages and macrovegetation cover, an expectation exists for the kinds of vegetation expected given a particular mean and standard deviation and vice versa. The most frequent vegetation associated with means and standard deviations of relative tail lengths in our data is shown in Fig. 7.8 as an “ecometric space”. Tropical vegetation is associated with snake assemblages in which the tail is relatively long (as it is in arboreal snakes) and the variation among species is relatively high. Cold deciduous forests (forests where leaves are lost due to cold winter temperatures rather than drought) are associated with assemblages that have relatively long tails but little variation among guild members. Grasslands are associated with assemblages whose average tail length is medium with little variation among species. Scrublands have assemblages with comparatively short tails, but some local assemblages have low variance and some high variance. Needle leaved forests are not readily distinguished from cold-deciduous forests or scrublands based on tail ecometrics, nor are meadowlands.
Fig. 7.8

Most common macrovegetation types for combinations of mean and standard deviation of relative tail length in North American snake assemblages. Historical and contemporary ecometric values are plotted for the snake communities at five conservation areas. Locations of the conservation areas are shown in Fig. 7.2d

7.5 Case Studies in Conservation Ecometrics

In the ecometric approach, functional traits provide a common denominator for comparing biotic-environmental dynamics across spatial and temporal scales, potentially integrating patterns observed in the paleontological record with those observed through ecological and conservation studies (Eronen et al. 2010a; Polly et al. 2011). The strategy of ecometrics is to simplify the modern world to make it as coarse as the fossil record, sampling only at spatial scales similar to those from which the faunas of paleontological sites are drawn (50 km grids and simple presence-absence occurrence data in the case of this study), and using environmental data that are no more specific than can be estimated for fossil sites. While this scale of analysis has been shown to be effective for paleontological data (Fortelius et al. 2002; Eronen et al. 2010b) and for continental scale analysis of vertebrate faunas (Polly 2010; Eronen et al. 2010c), it remains to be shown whether ecometrics have the precision necessary to be meaningful on the spatial and temporal scales of ecological and conservation studies. If they do, then ecometrics has the potential to provide much deeper historical baselines for the modern world (Willis et al. 2005) and new avenues for modeling biotic responses to changing climates and environments (Maurer 1999). We evaluated the potential of ecometrics to be applied to conservation-scale problems using five recently surveyed snake communities that have historical baseline data from the first half of the twentieth century.

7.5.1 Theodore Roosevelt National Park, North Dakota

Theodore Roosevelt National Park (TRNP) is a 28,509 ha reserve in the Little Missouri Badlands of North Dakota that has been a US National Wildlife Refuge since 1946 (Fig. 7.2d). The park is located in the drainage of the Little Missouri, with a mixture of riparian vegetation and wetlands set within a region that is naturally mixed-grass prairie. A study of the changes in the reptile and amphibian faunas of TRNP between 1920 and 2002 was recently published by Hossack et al. (2005), who found that the herpetofauna in general had changed very little during that time, and the snake fauna had not changed at all (Table 7.3). Hossack et al. (2005) concluded that the prairie habitats at TRNP have been minimally disturbed compared to other areas of the Great Plains, thus explaining the apparent stability of the herpetofauna. We calculated the ecometrics of the TRNP species, which have a mean relative tail length of 0.243 and standard deviation of 0.094, a combination most frequently associated with grassland or meadow in our study (Fig. 7.8). Even though the ecometric signal does not correspond to some local habitats at TRNP, such as wetlands and prairie, it does place the fauna in the correct regional habitat: grassland. The habitat TRNP was relatively unchanged over the period of study, as was the snake fauna, which means that our estimate of the ecometric change is technically consistent with field observations, though this particular example obviously does not test the ecometric approach very rigorously.
Table 7.3

Ecometric changes in snake assemblages from Theodore Roosevelt National Park, North Dakota 1954–2002. The dates of the historical baseline and the most recent survey are given, along with species lists, relative tail length for each species, and the status of each species today. The historical and current mean and standard deviation of relative tail length are reported

Species

Relative tail length

Status

Coluber constrictor

0.365

Persists

Crotalus viridus

0.113

Persists

Heterodon nasicus

0.223

Persists

Pituophis catenifer

0.22

Persists

Thamnophis radix

0.296

Persists

 

Mean

STD

1954

0.243

0.094

2002

0.243

0.094

7.5.2 Jackson-Pulaski Fish and Wildlife Refuge, Indiana

Brodman et al. (2002) described changes to the herpetofauna at natural reserves in the Kankakee River drainage of northwestern Indiana between 1931 and 1994. They resurveyed the Jackson-Pulaski, Willow Slough, and LaSalle Fish and Wildlife Areas (Fig. 7.2d). Before 1800, the area was covered by marsh, wetlands, and dry prairie (Lindsey et al. 1965), but by the time of the first herpetological surveys the area had been largely drained for agricultural use. Since then the area has seen the return of closed-canopy oak woodlands and a general homogenization of the mosaic of habitats that were part of the highly altered early twentieth century landscape. The snake fauna declined from 13 species in the 1930s to 10 in the 1990s, producing an ecometric shift in mean relative tail length from 0.295 to 0.300 and a drop in standard deviation from 0.080 to 0.076 (Table 7.4). While the ecometric difference seems small, it is substantial enough to be detected in our continent-scale ecometric analysis, but our vegetation categories are too coarse to distinguish the habitat changes that occurred (Fig. 7.2d). Finer scale vegetation data would improve the power of an ecometric analysis in this particular example, but such resolution is probably too fine for making comparisons with data derived from the paleontological and geological records.
Table 7.4

Ecometric changes in snake assemblages from Jackson-Pulaski Fish and Wildlife Refuge, Indiana 1930s–2002. The dates of the historical baseline and the most recent survey are given, along with species lists, relative tail length for each species, and the status of each species today. The historical and current mean and standard deviation of relative tail length are reported

Species

Relative tail length

Status

Coluber constrictor

0.365

Persists

Heterodon platirhinos

0.26

Persists

Lampropeltis triangulum

0.195

Decline

Nerodia sipedon

0.323

Persists

Opheodrys vernalis

0.38

Absent

Pantherophis vulpina

0.221

Persists

Pituophis catenifer

0.22

Persists

Sistrurus catenatus

0.159

Absent

Storeria dekayi

0.291

Persists

Thamnophis proximus

0.379

Persists

Thamnophis radix

0.296

Absent

Thamnophis sauritus

0.42

Persists

Thamnophis sirtalis

0.33

Decline

 

Mean

STD

1930s

0.295

0.08

2002

0.3

0.076

7.5.3 Fort Riley Military Reservation, Kansas

Fort Riley is located in the Flint Hills of east central Kansas (Fig. 7.2d). This area, like Theodore Roosevelt National Park, is naturally grassland, but is more highly impacted than TRNP. Busby and Parmelee (1996) reported a decline in snake species at Fort Riley from 22 in 1930 to 17 in 1996, corresponding to an ecometric shift from a mean relative tail length of 0.250 to 0.257 and standard deviation from 0.064 to 0.062 (Table 7.5). The change at Fort Riley is also small but detectable in our continental-scale ecometric analysis (Fig. 7.8). The ecometrics of both the 1930s and 1990s snake faunas indicate grassland as the most likely macrovegetation of the area, which is consistent with observed conditions in both the 1930s and 1990s: the area retains native vegetation as the dominant land cover, even though the river valleys have been altered by agriculture and settlement and there are more trees and herbaceous vegetation than in the 1930s.
Table 7.5

Ecometric changes in snake assemblages from Fort Riley Military Reservation, Kansas 1930–1996. The dates of the historical baseline and the most recent survey are given, along with species lists, relative tail length for each species, and the status of each species today. The historical and current mean and standard deviation of relative tail length are reported

Species

Relative tail length

Status

Agkistrodon contortrix

0.247

Persists

Carphophis vermis

0.187

Absent

Coluber constrictor

0.365

Persists

Crotalus horridus

0.12

Absent

Diadophis punctatus

0.23

Persists

Heterodon nasicus

0.223

Persists

Heterodon platirhinos

0.26

Absent

Lampropeltis calligaster

0.185

Persists

Lampropeltis getula

0.174

Persists

Lampropeltis triangulum

0.195

Persists

Nerodia sipedon

0.323

Persists

Pantherophis emoryi

0.235

Persists

Pantherophis obsoleta

0.246

Persists

Pituophis catenifer

0.22

Persists

Regina grahamii

0.265

Absent

Storeria dekayi

0.291

Persists

Tantilla gracilis

0.272

Persists

Tantilla nigriceps

0.257

Persists

Thamnophis proximus

0.379

Persists

Thamnophis radix

0.296

Absent

Thamnophis sirtalis

0.33

Persists

Tropidoclonion lineatum

0.199

Persists

 

Mean

STD

1930

0.25

0.064

1996

0.257

0.062

7.5.4 University of Kansas Natural History Reservation

A more interesting example is found at the University of Kansas Natural History Reservation at Lawrence, Kansas (Fig. 7.2d). Lawrence is situated in the ecotone between the eastern deciduous forest biome and the grasslands of the North American plains (Fig. 7.2d). In 1947 when the reserve was established the area had been largely cleared for agriculture and was otherwise covered with tall-grass prairie (Fitch 2006); since then the agricultural areas reverted first to a grass-weed mixture, with the subsequent invasion of forest habitats because the area was protected from burning. Fitch (2006) reported a major change in the herpetofauna of the reserve between 1947 and 2006, over which time four of the original 10 species had disappeared and another four had declined notably in abundance (Table 7.6). The ecometric mean of the 1947 snake fauna was 0.251 and its standard deviation was 0.076, values firmly associated with grassland macrovegetation in our “ecometric space” (Fig. 7.8). The 2006 fauna had a higher mean (0.269) and a lower standard deviation (0.054), a shift that moved the reserve from area of ecometric space associated with grasslands to one associated with deciduous forest, mirroring the actual vegetative changes at the reserve (Fig. 7.8). In this case where the area experienced a real biome shift, changes in the fauna and changes in the vegetation were correctly, albeit coarsely, modeled by our ecometric analysis.
Table 7.6

Ecometric changes in snake assemblages from University of Kansas Natural History Reservation 1947–2006. The dates of the historical baseline and the most recent survey are given, along with species lists, relative tail length for each species, and the status of each species today. The historical and current mean and standard deviation of relative tail length are reported

Species

Relative tail length

Status

Coluber constrictor

0.365

Absent

Croatalus horridus

0.12

Absent

Diadophis punctatus

0.23

Decline

Lampropeltis calligaster

0.185

Absent

Lampropeltis triangulum

0.195

Decline

Nerodia sipedon

0.323

Persists

Pantherophis obsoletus

0.246

Decline

Pituophis catenifer

0.22

Absent

Storeria dekayi

0.291

Decline

Thamnophis sirtalis

0.33

Persists

 

Mean

STD

1947

0.25

0.076

2006

0.27

0.054

7.5.5 Aguascalientes, Mexico

A different kind of example of a major ecometric shift is provided by the herpetofauna of Aguascalientes in central Mexico, where a resurvey in 2004 found a total of 12 snake species, whereas a 1958 survey of the same area found only 3 (Table 7.7; Sigala Rodríguez and Greene 2009). Sigala and Green attributed the increased number of species to better surveying rather than real changes in the community composition (some real changes may have occurred but they were impossible to document because of lack of a good historical baseline). Aguascalientes, an area of 5,600 km2, is substantially larger than our other examples and covers a wider range of habitats, including tropical deciduous forests, grasslands, shrub forests, and agricultural lands. The ecometric mean and standard deviation of the under-sampled snake fauna of 1958 were 0.189 and 0.118 respectively, outside the range of our data (Fig. 7.8). The resurveyed fauna falls ecometrically within the tropical forest macrovegetation, which is largely consistent with the actual vegetation cover of Aguascalientes, more so since a substantial proportion of the species occur in the forested areas (Sigala Rodríguez and Greene 2009). The ecometrics of the 1958 fauna, which are inconsistent with any extant North American snake assemblage, serve as an indicator that the fauna was poorly sampled.
Table 7.7

Ecometric changes in snake assemblages from Aguascalientes, Mexico 1958–2004. The dates of the historical baseline and the most recent survey are given, along with species lists, relative tail length for each species, and the status of each species today. The historical and current mean and standard deviation of relative tail length are reported

Species

Relative tail length

Status

Coluber bilineatus

0.42

New

Coluber mentovarius

0.368

New

Coluber taeniatus

0.407

New

Conopsis nasus

0.197

New

Crotalus lepidus

0.128

Persists

Crotalus molossus

0.113

Persists

Hypsiglena torquata

0.214

New

Oxybelis aeneus

0.491

New

Pituophis deppei

0.228

New

Salvadora bairdi

0.508

New

Senticolis triaspis

0.284

New

Tantilla bocourti

0.24

New

Tantilla wilcoxi

0.296

New

Thamnophis melanogaster

0.314

New

Thamophis eques

0.325

Persist

 

Mean

STD

1958

0.189

0.118

2004

0.302

0.12

7.6 Conclusions

Like body mass (Olalla-Tarraga et al. 2006; Head et al. 2009; Head 2010), tooth hypsodonty (Janis and Fortelius 1988; Fortelius et al. 2002), and mammalian digitigrady (Polly 2010), we found that mean relative tail length, a trait-based proxy for locomotor specialization, is geographically sorted among snake communities at a continental scale. The sorting of this ecometric trait is most strongly associated with ecological provinces and vegetation cover. We showed that some North American environments, such as those found in the Warm Continental Division (Fig. 7.7a), appear to favor little variation among the species in local communities, probably because the substrates available in this region are comparatively homogenous; other environments, such as Savanna Mountains (Fig. 7.7d), favor more variation among community members, probably because of locally heterogeneous substrate types. The shortest mean relative tail lengths are distributed in the desert southwest, suggesting sparse vegetation and aridity associated with unconsolidated soils and sediments facilitate burrowing habits.

The strong association between relative tail length and environment in North American snake communities open the possibility of using the ecometric approach as a bridge between broad-scale studies, such as continental or global analyses of the dynamics between changing climates, vegetation, and faunas (e.g., Prentice et al. 1992; Scheiter and Higgins 2009) or paleontological analyses of those dynamics through time (e.g., Mosbrugger et al. 2005; Fortelius et al. 2002; Williams et al. 2011). With the knowledge of the mean relative tail length of a North American fossil snake community, one can estimate with reasonable confidence the macrovegetation cover and ecological province associated with that community at a resolution that is commensurate with the best proxy data from the paleontological and geological record, making ecometrics an additional tool along with related tools like species distribution modeling (e.g., Maguire and Stigall 2009; Myers and Lieberman 2010; Varela et al. 2011; Svenning et al. 2011) and faunal-based distribution modeling of paleoclimate (e.g., Polly and Eronen 2011).

The resolution of our ecometric data were just fine enough to detect major changes on the scales relevant to conservation and ecology. In the five examples we reviewed, ecometric changes to all the snake faunas over intervals of 50–100 years were detectable, though the associated changes in vegetation were only detectable in cases where the local area shifted from one biome to another. The ecometric method as developed by us is not fine-scaled enough model changes in local abundance or the threat to individual species, but it is capable of measuring faunal-wide changes related to the extirpations, range expansions, and extinctions that accompany regional or continental changes in climate or land use. Ecometrics also appears to be useful as a crosscheck on whether local faunas have been fully sampled by comparing observed values to trait baselines derived from continental scale analyses.

The trait-based ecometric approach is a promising mid-level analysis that bridges the scale of paleontology, with its rich historical record of climate and faunal change, and the scale of ecology and conservation. By applying ecometrics to paleontological data, the potential exists for measuring the rates of climatic and biotic change that result in major extinctions or massive ecological reorganizations, which can then be compared to historical changes in ecological data to gauge the threat to local and regional ecosystems, communities, and species.

Notes

Acknowledgments

Matthew Rowe, Laura Scheiber, and Susan Spencer at the William R. Adams Zooarchaeology Lab, Indiana University, Ron Richards at the Indiana State Museum, Eileen Westwig at the American Museum of Natural History, Phil Myers at the University of Michigan, Kevin DeQueiroz and George Zug at the Smithsonian Institution, Kevin Seymour at the Royal Ontarioi Museum, Colin McCarthy and David Gower at the Natural History Museum, London, Christopher J. Bell at the University of Texas at Austin, Heidi Price-Thomas at Queen Mary, University of London, Carl Franklin and Jonathan Campbell at University of Texas at Arlington and Bill Stanley and Harold Voris at the Field Museum of Natural History provided specimens in their care. Christopher J. Bell, Jussi Eronen, Mikael Fortelius, Robert Guralnick, Anne Hereford, Steve Le Comber, Norman Macleod, Jesse Meik, and Eric Smith discussed or assisted with parts of this work. This work was supported by Indiana University and a grant from the US National Science Foundation (EAR-0843935) and is a contribution to the Integrated Climate Change Biology programme (iCCB) of the International Union of Biological Sciences (IUBS). Early data collection was supported by a NSF Biological Informatics Postdoctoral Fellowship to JJH (NSF 98–162, 0204082).

References

  1. Adalsteinsson SA, Branch WA, Trape S, Vitt LJ, Hedges SB (2009) Molecular phylogeny, classification, and biogeography of snakes of the family Leptotyphlopidae (Reptilia, Squamata). Zootaxa 2244:1–50Google Scholar
  2. Bailey RG (1998) Ecoregions map of North America. US Forest Serv Misc Publ 1548:1–10Google Scholar
  3. Bailey RG (2005) Identifying ecoregion boundaries. Environ Manage 34:S14–S16CrossRefGoogle Scholar
  4. Barnosky AD, Kaplan MH, Carrasco MA (2004) Assessing the effect of Middle Pleistocene climate change on Marmota populations from the Pit Locality. In: Barnosky AD (ed) Biodiversity response to climate change in the Middle Pleistocene. University of California Press, BerkeleyGoogle Scholar
  5. Blain H-A, Bailon S, Cuenca-Bescós G, Bennàsar M, Rofes J, López-García JM, Huguet R, Arsuaga JL, Castro JMBd, Carbonell E (2010) Climate and environment of the earliest West European hominins inferred from amphibian and squamate reptile assemblages: Sima del Elefante Lower Red Unit, Atapuerca, Spain. Quaternary Sci Rev 29:3034–3044CrossRefGoogle Scholar
  6. Böhme M, Ilg A, Ossig A, Küchenhoff H (2006) New method to estimate paleoprecipitation using fossil amphibians and reptiles and the middle and late Miocene precipitation gradients in Europe. Geology 34:425–428CrossRefGoogle Scholar
  7. Brodman R, Cortwright S, Resetar A (2002) Historical changes of reptiles and amphibians of northwest Indiana fish and wildlife properties. Am Midl Nat 147:135–144CrossRefGoogle Scholar
  8. Burbrink FT (2005) Inferring the phylogenetic position of Boa constrictor among the Boinae. Mol Phylogenet Evol 34:167–180CrossRefGoogle Scholar
  9. Busby WH, Parmelee JR (1996) Historical changes in a herpetofaunal assemblage in the Flint Hills of Kansas. Am Midl Nat 135:81–91CrossRefGoogle Scholar
  10. Castoe TA, Parkinson CL (2006) Bayesian mixed models and the phylogeny of pitvipers (Viperidae: Serpentes). Mol Phylogenet Evol 39:91–110CrossRefGoogle Scholar
  11. Caumul R, Polly PD (2005) Phylogenetic and environmental components of morphological variation: skull, mandible, and molar shape in marmots (Marmota, Rodentia). Evolution 59:2460–2472Google Scholar
  12. Crother BI (1999) Phylogenetic relationships among West Indian Xenodontine snakes (Serpentes; Colubridae) with comments on the phylogeny of some mainland Xenodontines. Contemporary Herpetology 2:1–4Google Scholar
  13. Damuth JD, Jablonski D, Harris RM, Potts R, Stucky RK, Sues HD, Weishampel DB (1992) Taxon-free characterization of animal communities. In: Beherensmeyer AK, Damuth JD, diMichele WA, Potts R, Sues HD, Wing SL (eds) Terrestrial ecosystems through time: evolutionary paleoecology of terrestrial plants and animals. University of Chicago Press, ChicagoGoogle Scholar
  14. Dryden IL, Mardia KV (1998) Statistical Shape Analysis. John Wiley and Sons, New YorkGoogle Scholar
  15. Emerson BC, Gillespie RG (2008) Phylogenetic analysis of community assembly and structure over space and time. Trends Ecol Evol 23:619–630CrossRefGoogle Scholar
  16. Eronen JT, Polly PD, Fred M, Damuth J, Frank DC, Mosbrugger V, Scheidegger C, Stenseth NC, Fortelius M (2010a) Ecometrics: the traits that bind the past and present together. Integr Zool 5:88–101CrossRefGoogle Scholar
  17. Eronen JT, Puolamäki K, Liu L, Lintulaakso K, Damuth J, Janis C, Fortelius M (2010b) Precipitation and large herbivorous mammals I: estimates from present-day communities. Evol Ecol Res 12:217–233Google Scholar
  18. Eronen JT, Puolamäki K, Liu L, Lintulaakso K, Damuth J, Janis C, Fortelius M (2010c) Precipitation and large herbivorous mammals II: applications to fossil data. Evol Ecol Res 12:235–248Google Scholar
  19. Fischer AG (1960) Latitudinal variations in organic diversity. Evolution 14:64–81CrossRefGoogle Scholar
  20. Fitch HS (2006) Collapse of a fauna: reptiles and turtles of the University of Kansas natural history reservation. J Kans Herpetol 17:10–13Google Scholar
  21. Fortelius M, Eronen J, Jernvall J, Liu LP, Pushkina D, Rinne J, Tesakov A, Vislobokova I, Zhang ZQ, Zhou LP (2002) Fossil mammals resolve regional patterns of Eurasian climate change over 20 million years. Evol Ecol Res 4:1005–1016Google Scholar
  22. Franca FGR, Mesquita DO, Nogueira CC, Araujo AFB (2008) Phylogeny and ecology determine morphological structure in a snake assemblage in the Central Brazilian Cerrado. Copeia 1:23–38CrossRefGoogle Scholar
  23. Garland T, Dickerman AW, Janis CM, Jones JA (1993) Phylogenetic analysis of covariance by computer simulation. Syst Biol 42:265–292Google Scholar
  24. Gower JC (1975) Generalized Procrustes analysis. Psychometrika 40:33–51CrossRefGoogle Scholar
  25. Guyer C, Donnelly MA (1990) Length–Mass relationships among an assemblage of tropical snakes in Costa Rica. J Trop Ecol 6:65–76CrossRefGoogle Scholar
  26. Head JJ (2010) Climatic inferences from extant and fossil reptiles: toward a metabolic paleothermometer. AGU fall meeting abstracts, vol #B51F-0412. Smithsonian/NASA Astrophysics Data SystemGoogle Scholar
  27. Head JJ, Bloch JI, Hastings AK, Bourque JR, Cadena EA, Herrera FA, Polly PD, Jaramillo CA (2009) Giant boid snake from the Palaeocene neotropics reveals hotter past equatorial temperatures. Nature 457:715–718CrossRefGoogle Scholar
  28. Helmus MR, Savage K, Diebel MW, Maxted JT, Ives AR (2007) Separating the determinants of phylogenetic community structure. Ecol Lett 10:917–925CrossRefGoogle Scholar
  29. Holm PA (2008) Phylogenetic biology of the burrowing snake tribe Sonorini (Colubridae). University of Arizona, TusconGoogle Scholar
  30. Hossack BR, Corn PS, Pilliod DS (2005) Lack of significant changes in the herpetofauna of Theodore Roosevelt National Park, North Dakota, since the 1920s. Am Midl Nat 154:423–432CrossRefGoogle Scholar
  31. Jablonski D (2008) Species selection: theory and data. Annu Rev Ecol Evol Syst 39:501–524CrossRefGoogle Scholar
  32. Jadin RC, Gutberlet RL, Smith EN (2010) Phylogeny, evolutionary morphology, and hemipenis descriptions of the Middle American jumping pitvipers (Serpentes: Crotalinae: Atropoides). J Zool Syst Evol Res 48:360–365CrossRefGoogle Scholar
  33. Janis CM, Fortelius M (1988) On the means whereby mammals achieve increased functional durability of their dentitions, with special reference to limiting factors. Biol Rev Camb Philos Soc 63:197–230CrossRefGoogle Scholar
  34. Jenks GF (1977) Optimal data classification for choropleth maps. University of Kansas Department of Geography Occasional Papers 2:1–24Google Scholar
  35. Johnson RG (1955) The adaptive and phylogenetic significance of vertebral form in snakes. Evolution 9:367–388CrossRefGoogle Scholar
  36. King RB (2008) Sexual dimorphism in snake tail length: sexual selection, natural selection, or morphological constraint? Biol J Linn Soc 38:133–154CrossRefGoogle Scholar
  37. Kozak KH, Larson A, Bonett RM, Harmon LJ (2005) Phylogenetic analysis of ecomorphological divergence, community structure, and diversification rates in dusky salamanders (Plethodontidae: Desmognathus). Evolution 59:2000–2016Google Scholar
  38. Kraft NJB, Cornwell WK, Webb CO, Ackerly DD (2007) Trait evolution, community assembly, and the phylogenetic structure of ecological communities. Am Nat 170:271–283CrossRefGoogle Scholar
  39. Lawing AM, Polly PD (2011) Pleistocene climate, phylogeny, and climate envelope models: an integrative approach to better understand species’ response to climate change. PLoS One 16:e28554CrossRefGoogle Scholar
  40. Lawson R, Slowinski JB, Crother BI, Burbrink FT (2005) Phylogeny of the Colubroidea (Serpentes): new evidence from mitochondrial and nuclear genes. Mol Phylogenet Evol 37:581–601CrossRefGoogle Scholar
  41. Lindell LE (1994) The evolution of vertebral number and body size in snakes. Funct Ecol 8:708–719CrossRefGoogle Scholar
  42. Lindsey AA, Crankshaw WB, Qadir SA (1965) Soil relations and distribution map of the vegetation of presettlement Indiana. Bot Gazette 126:155–163CrossRefGoogle Scholar
  43. Little SA, Kembel SW, Wilf P (2010) Paleotemperature proxies from leaf fossils reinterpreted in light of evolutionary history. PLoS One 5:e15161CrossRefGoogle Scholar
  44. Maguire KC, Stigall AL (2009) Using ecological niche modeling for quantitative biogegoraphic analysis: a case study of Miocene and Pliocene Equinae in the Great Plains. Paleobiology 35:587–611CrossRefGoogle Scholar
  45. Makarieva AM, Gorshkov VG, Li BL (2005) Gigantism, temperature and metabolic rate in terrestrial poikilotherms. Proc R Soc B Biol Sci 272:2325–2328CrossRefGoogle Scholar
  46. 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). J Zool 254:529–538CrossRefGoogle Scholar
  47. Matthews E (1983) Global vegetation and land use: new high-resolution data bases for climate studies. J Clim Appl Meteorol 22:474–487CrossRefGoogle Scholar
  48. Matthews E (1984) Prescription of land-surface boundary conditions in GISS GCM II: a simple method based on high-resolution vegetation datasets NASA TM-86096. National Aeronautics and Space Administration, Washington, DCGoogle Scholar
  49. Maurer BA (1999) Untangling ecological complexity: the macroscopic perspective. University of Chicago Press, ChicagoGoogle Scholar
  50. McGill B (2010) Matters of scale. Science 328:575CrossRefGoogle Scholar
  51. Mosbrugger V, Utescher T, Dilcher DL (2005) Cenozoic continental climatic evolution of Central Europe. Proc Natl Acad Sci USA 102:14964–14969CrossRefGoogle Scholar
  52. Mullin SJ, Seigel RA (2009) Sankes: ecology and conservation. Cornell University Press, IthacaGoogle Scholar
  53. Myers CE, Lieberman BS (2010) Sharks that pass in the night: using geographical information systems to investigate competition in the Cretaceous Western Interior Seaway. Proc R Soc B Biol Sci 278:681–689CrossRefGoogle Scholar
  54. Olalla-Tarraga MA, Rodriguez MA, Hawkins BA (2006) Broad-scale patterns of body size in squamate reptiles of Europe and North America. J Biogeogr 33:781–793CrossRefGoogle Scholar
  55. Polly PD (2003) Paleophylogeography: the tempo of geographic differentiation in marmots (Marmota). J Mammal 84:369–384CrossRefGoogle Scholar
  56. Polly PD (2010) Tiptoeing through the trophics: geographic variation in carnivoran locomotor ecomorphology in relation to environment. In: Goswami A, Friscia A (eds) Carnivoran evolution: new views on phylogeny, form, and function. Cambridge studies in morphology and molecules: new paradigms in evolutionary biology. Cambridge University Press, CambridgeGoogle Scholar
  57. Polly PD, Eronen JT (2011) Mammal associations in the Pleistocene of Britain: implications of ecological niche modelling and a method for reconstrucing palaeoclimate. In: Ashton N, Lewis SG, Stringer C (eds) The ancient human occupation of Britain. Elsevier, New YorkGoogle Scholar
  58. Polly PD, Eronen JT, Fred M, Dietl GP, Mosbrugger V, Scheidegger C, Frank DC, Damuth J, Stenseth NC, Fortelius M (2011) History matters: ecometrics and integrative climate change biology. Proc R Soc B Biol Sci 278:1131–1140CrossRefGoogle Scholar
  59. Prentice IC, Cramer W, Harrison SP, Leemans R, Monserud RA, Solomon AM (1992) A global biome model based on palent physiology and dominance, soil properties and climate. J Biogeogr 19:117–134CrossRefGoogle Scholar
  60. Pyron RA, Burbrink FT, Colli GR, Montes de Oca AN, 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. Mol Phylogenet Evol 58:329–342CrossRefGoogle Scholar
  61. Reading CJ, Luiselli LM, Akani GC, Bonet X, Amori G, Ballouard JM, Filippi E, Naulleau G, Pearson D, Rugiero L (2010) Are snake populations in widespread decline? Biol Lett 6:777–780CrossRefGoogle Scholar
  62. Ricklefs RE, Miles DB (1994) Ecological and evolutionary inferences form morphology: an ecological perspective. In: Wainwright PC, Reilly SM (eds) Ecological morphology: integrative organismal biology. University of Chicago Press, ChicagoGoogle Scholar
  63. Rodríguez MA, Belmontes JA, Hawkins BA (2005) Energy, water and large-scale patterns of reptile and amphibian species richness in Europe. Acta Oecol 28:65–70CrossRefGoogle Scholar
  64. Rohlf FJ, Slice D (1990) Extentions of the Procrustes method for the optimal superimposition of landmarks. Syst Zool 39:40–59CrossRefGoogle Scholar
  65. Row LW, Hastings DA (1994) TerrainBase worldwide digital terrain data (release 1.0). National Oceanic and Atmospheric Administration, National Geophysical Data Center, BoulderGoogle Scholar
  66. Sanders KL, Mumpuni AH, Head JJ, Gower DJ (2010) Phylogeny and divergence times of filesnakes (Acrochordus): inferences from morphology, fossils and three molecular loci. Mol Phylogenet Evol 56:857–867CrossRefGoogle Scholar
  67. Scheiter S, Higgins SI (2009) Impacts of climate change on the vegetation of Africa: an adptive dynamic vegetation modelling approach. Global Change Biol 15:2224–2246CrossRefGoogle Scholar
  68. Sigala Rodríguez JJ, Greene HW (2009) Landscape change and conservation priorities: Mexican herpetofaunal perspectives at local and regional scales. Revista Mexicana de Biodiversidad 80:231–240Google Scholar
  69. Slowinski JB, Lawson R (2002) Snake phylogeny: evidence from nuclear and mitochondrial genes. Mol Phylogenet Evol 24:194–202CrossRefGoogle Scholar
  70. Sneath PH (1967) Trend-surface analysis of transformation grids. J Zool 151:65–122CrossRefGoogle Scholar
  71. Svenning J-C, Flojgaard C, Marske KA, Nógues Bravo D, Normand S (2011) Applications of species distribution modeling to paleobiology. Quaternary Sci Rev 30:2930–2947CrossRefGoogle Scholar
  72. Tingley MW, Beissinger SR (2009) Detecting range shifts from historical species occurrences: new perspectives on old data. Trends Ecol Evol 24:625–633CrossRefGoogle Scholar
  73. Varela S, Lobo JM, Hortal J (2011) Using species distribution models in paleobiogeography: a matter of data, predictors and concepts. Palaeogeogr Palaeoclimat Palaeoecol 310:451–463CrossRefGoogle Scholar
  74. Vidal N, Kindl SG, Wong A, Hedges SB (2000) Phylogenetic relationships of xenodontine snakes inferred from 12S and 16S ribosomal RNA sequences. Mol Phylogenet Evol 14:389–402CrossRefGoogle Scholar
  75. Webb JK, Shine R (1998) Using thermal ecology to predict retreat-site selection by an endangered snake species. Biol Conserv 86:233–242CrossRefGoogle Scholar
  76. Webb CT, Hoeting JA, Ames GM, Pyne MI, Poff NL (2010) A structure and dynamic framework to advance traits-based theory and prediction in ecology. Ecol Lett 13:267–283CrossRefGoogle Scholar
  77. Wiens JA, Bachelet D (2010) Matching the multiple scales of conservation with the multiple scales of climate change. Conserv Biol 24:51–62CrossRefGoogle Scholar
  78. 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 60:123–141Google Scholar
  79. Wilcox TP, Zwickl DJ, Heath TA, Hillis DM (2002) Phylogenetic relationships of the dwarf boas and a comparison of Bayesian and bootstrap measures of phylogenetic support. Mol Phylogenet Evol 25:361–371CrossRefGoogle Scholar
  80. Williams JW, Blois JL, Shuman BN (2011) Extrinsic and instrinsic forcing of abrupt ecological change: case studies from the late Quaternary. J Ecol 99:664–677CrossRefGoogle Scholar
  81. Willis KJ, Gillson L, Brncic TM, Figueroa-Rangel BL (2005) Providing baselines for biodiversity measurement. Trends Ecol Evol 20:107–108CrossRefGoogle Scholar
  82. Willmott KM, Legates DR (1998) Global air temperature and precipitation: regridded monthly and annual climatologies (version 2.01). Center for Climatic Research, University of Delaware, NewarkGoogle Scholar
  83. Wilson JA, Dhananjay MM, Peters SE, Head JJ (2010) Predation upon hatchling dinosaurs by a new snake from the late cretaceous of India. PLoS Biol 8:e1000322CrossRefGoogle Scholar
  84. Wolfe JA (1979) Temperature parameters of humid to mesic forests of Eastern Asia and relation to forests of other regions in the Northern Hemisphere and Australasia. US Geol Survey Prof Pap 1106:1–37Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • A. Michelle Lawing
    • 1
  • Jason J. Head
    • 2
  • P. David Polly
    • 3
  1. 1.Department of Geological Sciences and BiologyIndiana UniversityBloomingtonUSA
  2. 2.Department of Earth and Atmospheric SciencesUniversity of Nebraska-LincolnLincolnUSA
  3. 3.Department of Geological SciencesIndiana UniversityBloomingtonUSA

Personalised recommendations