Multiple traits and multifarious environments: integrated divergence of morphology and life history
Data files
Dec 16, 2019 version files 134.67 KB
-
Individual_data.xlsx
-
PLS-SEM_data.xlsx
Abstract
Understanding complex responses of multiple character suites (e.g., behaviour, life history, morphology) to multifarious environments is a challenging task. Here we use a multivariate approach (partial least squares structural equation modelling) to disentangle drivers (i.e., predation, resource availability, and population demographics) of phenotypic divergence among populations of Bahamas mosquitofish (Gambusia hubbsi) inhabiting blue holes. We further employ a two-block partial least squares analysis in a novel approach to uncovering integrated and independent aspects of divergence in correlated character suites. Results suggest that phenotypic divergence mainly resulted from differences in predation regimes, with population demography and resource availability also influencing particular aspects of divergence. We uncovered statistically significant covariation of life histories and morphologies, and revealed that phenotypic divergence between predation regimes involved both integrated and independent responses. For instance, female life histories diverged mostly independently of morphology, although some morphological shifts (abdominal distension) resulted from changes in fecundity. In contrast, males showed strong morphological divergence independent of life history, but much of their life-history shifts reflected joint morphological changes (lean weight and body shape). Our study illustrates the utility of gathering disparate data types from multiple populations/species to understand the causes and nature of phenotypic divergence in the wild.
Usage notes
This dataset consists of two separate files.
The first file (individual data.xlsx) has all the individual level data for both male and female Gambusia hubbsi used in the 2B-PLS analyses and in the MANCOVAs for life histories and morphologies (centroid size and relative warps 1-18), where we either included or didn’t include 2B-PLS vectors to test for integrated/independent responses of the character suites. For each fish it also provides details on the population (i.e, identity of the blue hole), whether it was high-predation or low-predation.
The second file (PLS-SEM data.xlsx) has all population-specific data for all 14 populations used in the PLS-SEM analyses (both the original variables used to construct the latent variables, and the latent variable scores themselves).