PARENT SESSION
Oral Session # 13: Statistical Ecology.
Presiding: R Stevens
Monday, August 4. 8:00 AM to 11:30 AM, SITCC Meeting Room 205.
Quantifying phenotypic plasticity in ecological experiments.
McCoy, Michael *,1, Osenberg, Craig1, Bolker, Benjamin1, Miner, Benjamin1, Vonesh, James1, 1 University of Florida, Gainesville, Florida, USA
ABSTRACT- Phenotypic plasticity can change community dynamics by altering the nature of species interactions. Therefore, it is important to quantify phenotypic change and incorporate its effect into studies of community ecology. To compare magnitudes of plasticity across taxa and to understand the adaptive significance of these responses, we require unbiased estimates of changes in morphology induced by environmental factors. Because morphology changes with overall size, analytic procedures must control for allometric scaling of specific traits with body size. Many studies of phenotypic plasticity use a technique called shearing to correct for size. Shearing, like all methods of size correction, is based on the assumption that groups being compared share a common allometric pattern. In shearing, this common allometry is usually characterized by the first principal component (PC1) of the pooled data (across all treatment groups). Residuals from ordinary least squares regression of each trait against PC1 are then compared among groups. Shearing assumes that (1) PC1 obtained from pooled data correctly describes the within-group scaling relationships; and (2) the allometric relationships (covariance matrices) of the groups are identical. Morphological data may violate both of these assumptions; in particular, inducing phenotypic changes in experimental organisms often changes allometric patterns. Nonetheless, most studies of phenotypic plasticity in amphibians and other vertebrates have applied shearing to compare phenotypes without acknowledging or testing these important assumptions. In this study, we discuss potential biases imposed by violations of these assumptions and discuss Common Principal Components (CPC) and Burnaby′s back-projection method as an alternative for the study of phenotypic plasticity. We use simulated data and results from an experiment testing for predator-induced plasticity in pinewoods treefrog (Hyla femoralis) to compare the two approaches.
Key words: shearing, predation, phenotypic plasticity, tadpole