Abstract Detail

Biodiversity Informatics & Herbarium Digitization

Kunkel, David [1], Fishbein, Mark [2].

Improving Ecological Niche Modeling: Comparing Methods of Spatial and Climatic Rarefaction on Geographically Restricted and Widespread Species of Asclepias (Apocynaceae).

Ecological niche modeling, an approach for predicting potential species occurrence based on niche components, is a tool that has been commonly used to study the ecological niche of a species and how that corresponds to geographic space. However, previous studies have shown that these models are often sensitive to not only the quality of the data that is being used to infer the niche, but also the spatial distribution of that data. The method of rarefaction has been developed to act as a way of solving this problem of spatial autocorrelation and has become increasingly popular as a way of preventing model-overfitting. Rarefaction was originally developed in the context of geographic space, but new developments have suggested that applying it as a climatic filter may work considerably better. However, it remains unclear how both of these methods perform in the context of geographically restricted versus widespread species. Here, I use two Asclepias species, A. hypoleuca, a geographically restricted species, and A. asperula, a widespread species, as a model to compare the performance of geographic and climatic rarefaction at reducing the tendency of ecological niche models to overfit in rare and common species. I hypothesized that for rare species, the climatic method of rarefaction would have a higher performance, while for common species geographic rarefaction would perform better. This work will inform ecological niche modeling strategies that should be used depending on the context of the species being modelled.

1 - 208 S Duncan St, Apt 3, Stillwater, OK, 74074, United States
2 - Oklahoma State University, Dept Of Plant Biology, Ecology & Evolution, 301 Physical Science, Stillwater, OK, 74078, United States

Ecological niche model
data filtering
spatial data.

Presentation Type: Poster
Number: PBI005
Abstract ID:431
Candidate for Awards:None

Copyright © 2000-2022, Botanical Society of America. All rights reserved