The Princeton Guide to Ecology

The study of ecology frequently draws on comparative observations and experiments that rely on the similarities and differences among species and the correlations among species traits and the environment. In such studies, consideration of the phylogenetic relationships among species provides valuable information for statistical inference and an understanding of evolutionary history underlying presentday ecological patterns. From a statistical perspective, related species do not necessarily provide independent data points for hypothesis tests, due to inheritance of shared characteristics from common ancestors. This similarity can be addressed through a variety of statistical techniques, including the widely used method of phylogenetic independent contrasts. Independent contrasts play a particularly valuable role in the analysis of trait and trait–environment correlations and may point toward alternative interpretations of comparative data. In community ecology, measures of the phylogenetic clustering or spacing of co-occurring species provide a useful tool to test alternative processes underlying community assembly. Co-occurrence of close relatives most likely reflects ecological filtering, in which related species with similar traits share the ability to tolerate local conditions. The reverse pattern of phylogenetic spacing of cooccurring species may reflect a variety of processes, and additional observations of species traits in relation to environment and interacting taxa will be necessary to address underlying processes. Use of comparative methods has increased dramatically with the rapid growth in phylogenetic information and computing power and will continue to play an important role in ecological research.

See figure 1 for illustrations of main terms.

branch lengths. These may indicate either the number of inferred character changes or a measure of relative or absolute time along any particular branch connecting two nodes. If the molecular data underlying a phylogeny do not violate a molecular clock, a single rate may be imposed such that branch lengths will represent relative time, and contemporaneous taxa will be placed at the same distance from the root (i.e., the same age). If a molecular clock is violated, rate-smoothing methods have been developed to obtain the best-supported estimate of relative time. Fossils and biogeographic or paleoecological information may then be used to calibrate these branch lengths and convert them to units of absolute time. Rate-smoothing and calibration methods are fraught with difficulty, and branch lengths should be treated with caution. (Note that branch lengths may also be set arbitrarily for convenience when one is drawing trees, in which case they have no intrinsic biological meaning.)

character states. Phylogenetic trees are reconstructed based on analysis of a matrix of characters, where each character can take on one of two or more states (binary or multistate, respectively) for each taxon in the group. Phylogenies can be reconstructed from molecular and/or morphological data, although the former are now much more common. Analyses that include morphological data are advantageous as they make it possible to incorporate taxa or fossils for which molecular data are not available.

lineage. This refers to a single line of ancestor–descendant relationship, connecting nodes within a phylogeny.

most recent common ancestor (MRCA). The MRCA is the most recent node that is shared by any two taxa in a tree.

-117-