PARENT SESSION

Systematics/Zoogeography 2 -- Session Chair: Robert Anderson-- Nelson Hall East, Goodwin Forum

BOOTSTRAPPING ANALYSES IN HETEROMYS ANOMALUS (RODENTIA: HETEROMYIDAE): A MODERN APPROACH TO ASSESSING NONGEOGRAPHIC VARIATION. Robert P. Anderson¹ and Norman A. Slade². ¹ Dept. of Biology, City College of the City University of New York, New York, NY, USA; ² Natural History Museum, and Dept. of Ecology & Evolutionary Biology, University of Kansas, Lawrence, KS, USA.

ABSTRACT- Most mammals follow allometric growth, with corporal proportions changing during ontogeny. Explicit modeling of allometric patterns is necessary to characterize differentiation in size and shape effectively in multispecies studies of morphological evolution. Under the generalized equation of allometry, multivariate statistical analyses can be used to characterize allometric growth in a single-species sample of various ages (by the first principal component of the covariance matrix of log-transformed linear measurements). Although such analyses are often conducted, few studies compare the observed empirical pattern with the null model of isometry (no changes in proportion) in a statistical framework. Here, we employ bootstrapping analyses to examine growth in the Caribbean spiny pocket mouse Heteromys anomalus. We used eleven linear measurements taken on the skulls of 91 specimens from an insular population (Bush Bush Forest in Trinidad) and 35 from a continental population (Latal in Venezuela). For Bush Bush Forest, we first examined males and females separately. Both sexes showed a strongly allometric growth pattern, with measurements reflecting the rostral region growing more than expected under isometry, and those relating to the neurocranium growing less than under isometry. The growth vectors for males and females could not be distinguished statistically. Hence, the sexes change proportions in the same way as they grow (their growth vectors are parallel in multi-log space). Analyses of all specimens (males and females pooled) from each of the two sites (Bush Bush Forest and Latal) yielded the same allometric patterns. Populations from those two sites differed significantly from isometry, but the analysis did not detect a difference between their respective growth trajectories. Thus, they share an evolutionarily conserved pattern of allometric growth. Future studies can now use these techniques to determine whether congeneric species exhibit the same ontogenetic trajectory as H. anomalus. If so, that growth vector could then be removed from the samples to test for differences in shape that are independent of the effects of allometric growth.

KEY WORDS: bootstrapping, morphometrics, allometry, growth