An algorithm for the identification of indicator taxonomic units and their use in analyses of ecosystem state.

Abstract

Biological community structure can be used as an ecological state descriptor, and the sensitivity of some taxonomic groups or biological entities to environmental conditions allows for their use as ecological state indicators. This work describes an algorithm developed for the identification of such taxonomic units when comparing environments or ecosystems under different anthropic impacts. Based solely on presence or absence information in a database, the algorithm identifies indicator taxonomic units for each environment, estimates the belonging of any additional samples to a given environment, approximates the ecological niche of any taxonomic unit based on two or more selected environmental factors, and analyzes the frequency of any taxonomic unit in a selected combination of the environmental factors chosen. By using the approximation to the ecological niche of the taxonomic units present, given a new sample, the physicochemical parameters of the environment it was taken can be estimated by the units present in the sample. These analyses can be performed simultaneously for two or more taxonomic units. This work provides a description of how the mathematical method was developed. Based on this methodology, a freely downloadable R package for easy use was developed, (Ecoindicators, DOI: https://github.com/lsaravia/EcoIndicators). One of the advantages of this method, and the R-package mentioned is that it can be used for any ecosystem for which there is a suitable biological dataset associated with environmental factors. In addition, both this mathematical procedure and the package referred to, can be tailored by other researchers to fit their own needs.