Using Community-Level Analyses to Identify Dietary Patterns for Species in Space and Time

ABSTRACT

Non-parametric, multivariate analyses commonly used to characterize spatial and temporal patterns in the taxonomic composition of ecological communities can likewise be useful when applied to data generated from the examination of gut contents. A suite of analyses including non-metric multidimensional scaling (NMS), multi-response permutation procedures (MRPP), and indicator species analysis (ISA) revealed interesting dietary shifts that occurred across individuals of emerald shiner (Notropis atherinoides) collected from among different habitats and two different time periods (pre-and post-impoundment of the upper Tombigbee River). These patterns were delineated from data that exhibited non-linear distribution and that incorporated measurements made at different scales, two factors that confound traditional parametric analyses. NMS ordination provided clear, graphic representation of the separation that occurred when N. atherinoides diets were viewed across space and time. MRPP analysis validated that the observed separation was statistically significant, and ISA complemented these other two analyses by identifying diet contributors that were indicative of each grouping.