Abstract Detail

Ecology

Merritt, Benjamin [1], Yadav, Sunita [1], Culley, Theresa [1], Witsell, Theo [2], Kephart, Susan [3].

Can ecological niche models be used to differentiate taxa? A test case using the wild hyacinth (Camassia spp.) in the eastern United States.

Species are deeply rooted within their environments such that it is possible to use biotic and abiotic variables to describe habitat suitability and potentially even differentiate species based on unique habitat types. The plant genus Camassia is a taxonomically complex group that provides an excellent opportunity to test the use of ecological niche modeling in differentiating closely related species. This genus consists of a number of taxa found across North America that inhabit diverse environments, ranging from open prairies to wet fields to shaded forests. Using a GIS-based approach, we used ecological niche models to construct distribution maps of the eastern taxa: C. angusta, C. scilloides, and a putative new Camassia taxon found in southern glades of the eastern United States. A one kilometer resolution DEM was used in concert with 19 WorldClim climatic variables, as well as slope, aspect, solar radiation, land-use land-cover, and the number of growing degree days to identify habitat suitability using field-collected occurrences in MaxEnt. Predictors found to be uncorrelated and useful in minimizing variance inflation were used to develop species-specific habitat suitability maps, narrowing the list to 13 abiotic variables. Ensembles of small models (ESM) were used to account for the limited number of verified presence records for each taxon. These ESMs were built using all bivariate combinations of the 13 abiotic variables. A weighted average of the ESMs was used to develop the final species distribution map for each taxon. Here we present preliminary findings of model comparisons that were used to differentiate species based on an ecological niche modeling approach. Our next steps include identifying additional variables that may increase model performance and subsequently validate the developed models. Ultimately, our goal is to determine whether ecological niche models are consistent with the current taxonomic classification of these taxa as based on morphological and genetic analyses.

1 - University Of Cincinnati, Department Of Biological Sciences, 614 Rieveschl Hall, Cincinnati, OH, 45221-0006, USA
2 - Arkansas National Heritage Commission, 323 Center St., Suite 1500, Little Rock, AK, 72201, USA
3 - WILLAMETTE UNIVERSITY, Department Of Biology, SALEM, OR, 97301, USA

Keywords:
GIS
MaxEnt
ecological niche modeling
ensembles of small models (ESM).

Presentation Type: Poster
Session: P, Ecology Section Posters
Location: Exhibit Hall/Savannah International Trade and Convention Center
Date: Monday, August 1st, 2016
Time: 5:30 PM This poster will be presented at 5:30 pm. The Poster Session runs from 5:30 pm to 7:00 pm. Posters with odd poster numbers are presented at 5:30 pm, and posters with even poster numbers are presented at 6:15 pm.
Number: PEC021
Abstract ID:565
Candidate for Awards:Ecological Section Best Graduate Student Poster