Publication AbstractClose Publication Abstract

Predicting Habitat for Eurasian Watermilfoil with Mahalanobis Distance Methods

Prince Czarnecki, J. M., Madsen, J. D., Shaw, D. R., & Brooks, C. P. (2010). Predicting Habitat for Eurasian Watermilfoil with Mahalanobis Distance Methods. Weed Science Society of America Abstracts. Denver, CO.

Abstract

A number of modeling techniques have been developed that predict habitat suitability based on species presence (e.g., Mahalanobis distance, Maximum Entropy, etc.). While presence-only models have limitations, presence data is often the only data available for regional to large-scale research. Additionally, sampling methods can often force the use of presence-only methods because a lack of data in specific areas cannot be treated as absence of a species. In this study, Mahalanobis distance was used to characterize and predict habitat for the invasive aquatic macrophyte, Eurasian watermilfoil (Myriophyllum spicatum). Mahalanobis methods were applied within a GIS framework. Mahalanobis is ideal for characterizing and predicting potential habitat as long as the range of variance for all predictor variables for the set of occupied habitats includes the range present in available habitats, and that only a single optimum exists. This technique is robust even when data are highly spatially autocorrelated or when covariance is high between predictors. Using data from Minnesota waterways and lakes, habitat factors (or predictors) were identified for model input. These included: distance from water access point, presence of bass, total alkalinity, Carlson TSI, acreage of waterbody, and Secchi depth. Data were obtained from the Minnesota Department of Natural Resources and several researchers at the University of Minnesota. Preliminary model iterations indicate that predictor selection requires more attention. Evaluation of validity used established methods shows that current models have low reliability and limited applicability. Future work will use step-regression and additional measures to increase model accuracy yielding successful validation.


Geosystems Research Institute  •  Contact GRI
Modified: December 2, 2020  •  WebMaster  •  Intranet