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

• DOI: 10.1101/2022.05.16.492087
• Publication Date: 2022-05-18T17:40:11Z
• Authors: Hernán de la Vega; Liliana Falco; Leonardo Saravia; Rosana Sandler; Andrés Duhour; Víctor N Velazco; Carlos Coviella
• Keywords: Unidades taxonómicas; Algoritmos; Matemáticas; Ecosistemas; Comunidad biológica; 15. Life on land; SUSTAINABILITY; 03 medical and health sciences; https://purl.org/becyt/ford/1.6; ANTHROPIC IMPACT; ECOLOGICAL INDICES; MATHEMATICAL ECOLOGY; Sustentabilidad; https://purl.org/becyt/ford/1; 0305 other medical science; Ecología matemática

AbstractBiological 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 a mathematical methodology developed for the identification of such taxonomic units when comparing environments or ecosystems under different anthropic impacts. Based on this methodology, a freely downloadable R package for easy use was developed, (Ecoindicators, DOI: 10.5281/zenodo.5772829). Based solely on presence or absence information, the method 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 returns the frequency of any taxonomic unit in a selected combination of environmental factors. By using the approximation to the ecological niche of species present, given a new sample, the physicochemical parameters can be estimated by the species present in the sample. These analyses can be performed simultaneously for two or more taxonomic units. This paper provides a mathematical description of how the mathematical method was developed. One of the advantages of this method, and the referred R-package 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. We also expect other developers to further improve it.