Predicting plant species distribution patterns using simple climatic parameters: A case study of Ecuadorian palms

A Geographic Information System (GIS) is used to model plant distributions. The model is based on annual mean temperature, humidity, and a georeferenced specimen data base and predicts potential and probable distribution for each taxa in the data base. Potential distribution range is suitable area determined by climatic constraints. Probable distribution range is here defined as the part of the potential range where a taxon is most likely to be found based on distance to existing collections. Gridded estimates of climatic parameters were calculated based on a Digital Elevation Model and data from ca 100 weather stations. The palm family in Ecuador was used to demonstrate the model. Potential and probable distribution ranges were calculated for each of the 127 Ecuadorian palm species. Maps of potential and probable species richness were calculated by overlaying and accumulating individual distribution maps. Potential species richness was highest in the Amazon lowland where it reached 60 species and lowest in the higher parts of the Andes and drier parts of the coastal plain. The probable species richness followed a similar pattern but was nowhere higher than 47 species. The ratio between potential species richness and probable species richness was used to find areas that are insufficiently sampled.