Graduation Semester and Year




Document Type


Degree Name

Doctor of Philosophy in Quantitative Biology



First Advisor

Andre Pires Da Silva


I examined various facets of morphological variation and evolution in the speckled rattlesnake (Crotalus mitchellii) species group. I performed three separate, but related, studies pertaining to the following topics: body size of putative ancestral source species and patterns of snake occupancy on islands in the Baja California region, the evolution and prevalence of gigantism and dwarfism among insular populations of speckled rattlesnakes, and the climatic correlates of scalation patterns in speckled rattlesnakes.I examined the influence of body size of mainland source species on insular distribution patterns by using a species index that weights isolation of islands while simultaneously penalizing island area. The Isolation/Area Index (I/A Index) can then be incorporated in regression analyses with maximum body size of putative mainland derivative species as the predictor variable. I applied this nonphylogenetic test to insular snake populations derived from 27 mainland species that occur on 39 islands off the coast of Baja California and detected a significant relationship between the I/A Index and maximum body size of mainland populations, providing evidence that snake body size affects dispersal ability. This relationship was especially clear among heavy-bodied snakes (e.g., rattlesnakes). In addition, the influence of a potential confounding relationship, the prediction that small islands would support smaller-bodied snake species, was ruled out in that small snake species were notably absent from small islands. In light of our findings I interpret this relationship to result from the increased colonization ability of larger snakes; thus larger snakes would be more likely to encounter and become established on smaller islands, regardless of their ability to persist through ecological time.Populations of speckled rattlesnakes are known to occupy multiple islands off the coast of Baja California, Mexico. Two of these populations (Angel de la Guarda and El Muerto) have been recognized as subspecies based primarily on size divergence from putative mainland ancestral populations; however, a survey of morphological variation from other islands inhabited by these snakes has not previously been reported. I examined variation in body size, size-adjusted multivariate phenotype, and island physical parameters among two mainland clades and 12 island populations and demonstrate through partial set correlations that speckled rattlesnakes tend to dwarf as an inverse function of island area on islands smaller than approximately 20 km2. Body size was not significantly related to measures of island isolation. Furthermore, divergence in size-adjusted multivariate phenotype was associated with divergence in body size rather than any of the island physical parameters. These results suggest that overall morphological divergence in insular speckled rattlesnakes with respect to putative mainland source clades resulted primarily from directional selection for body size evolution. With one exception (Isla Partida Sur), island populations were more similar in phenotype to their geographically most proximate mainland clade. Morphological divergence in this dwarfed population is best explained by convergence towards a similar morphospace with other dwarfed populations.The arrangement and counts of epidermal scales (scalation) has been fundamental to the systematics of various vertebrate groups; however, relatively little is known of their developmental integration or ecological determinants. I incorporated a randomization test to evaluate the multivariate correlation, and hence integration, of a set of 12 head scale counts within the speckled rattlesnake (Crotalus mitchellii) species group. I also evaluated the evolutionary correlation of head scalation and bioclimatic variables that may exert selection on scale counts using phylogenetic canonical correlation analysis (PCCA) and multiple regression. The various standardized scale counts were highly intercorrelated across species providing further evidence that the various standardized head scale counts are not independent characters, but comprise a developmental module. Total number of head scales was positively correlated with precipitation and inversely correlated with temperature, a finding consistent with comparisons of other squamate reptiles. The interdependence of head scale counts may reduce independent evolution of scale traits, but likely increases the probability of adaptive covariance across common selective gradients.


Biology | Life Sciences


Degree granted by The University of Texas at Arlington

Included in

Biology Commons