Mapping environmental properties in data‐poor areas using species occurrences

Study proposes method for using indicator species to derive environmental data in areas where information is scarce

Data resources used via GBIF : 30,000 species occurrences
Maidenhair fern (Adiantum sp.) collected in Soplin, Perú by M. A. Ríos Paredes et al. Photo: Field Museum of Natural History (CC BY-NC 4.0)

Environmental data is commonly used to define ecological niches and to model species distributions. When environmental data is scarce, however, the presence of indicator species can be used to infer environmental conditions.

In this study, authors outline a method for generating environmental maps derived from plot data in three layers: 1) indicators species with known environmental properties, 2) species occurrences from e.g. GBIF, and 3) environmental data only.

From the first layer, the authors are able to derive an environmental optimum. In the next step, this is can be used to infer estimates of environmental properties at species-only plots. These estimates when interpolated with known enviromental plots generate a map of an area of interest that can be validated using a external dataset.

The authors employ the proposed framework to successfully produce a map of soil quality (as measured through cation concentration) for Amazonia using ferns and lycophytes as indicator species, leading to a 12-fold increase in enviromental data.

Original paper

Zuquim G, Stropp J, Moulatlet GM, Van doninck Jasper, Quesada CA, Figueiredo FOG, Costa FRC, Ruokolainen K and Tuomisto H (2019) Making the most of scarce data: Mapping soil gradients in data‐poor areas using species occurrence records. Methods in Ecology and Evolution. Wiley 10(6): 788–801. Available at: