Waiting for the snow to melt…
Species distribution modeling is a powerful ecological tool that combines remotely sensed data with known presence locations to yield habitat suitability predictions that can be applied to a variety of hypotheses and conservation planning efforts, and can also be used on a smaller scale to examine individual species’ habitat preferences. Here, species distribution modeling techniques are applied to the yellow-bellied marmot, Marmota flaviventris, a semi- fossorial, facultatively social rodent that is found across high elevations in much of western North America. M. flaviventris is known to prefer microhabitat patches with characteristics related to predator detection and avoidance; good visibility and the number of large rocks are the best predictors of marmot presence and persistence. To examine the habitat preferences of M. flaviventris, Maximum Entropy was used with environmental variables including slope, visibility, elevation, local geology, and vegetation, as well as 75 presence locations, to produce and verify a model of yellow-bellied marmot habitat preference in the East River Valley surrounding Gothic, Colorado. Statistical verification demonstrated that the model is a significant predictor of marmot presence, and field validation of the model yielded moderate success (72.9% correct predictions, or 97/133 sites). The importance of visibility and slope to the model indicate that predator-related variables are indeed the most important predictors of marmot presence, although the number of large rocks, as well as other potentially relevant biotic factors, were not included in the model. While MaxEnt modeling is generally undertaken on a larger scale, this model’s success indicates that fine resolution modeling with the goal of field validation and examination of small-scale environmental variations is plausible, especially if environmental predictors are carefully selected and informed by a thorough consideration of the ecology of the focal species. Voss 3
Knowledge graph centered on Field validation of a distribution model for the y with 16 nodes and 59 connections. Top connected: Rocky Mountain Biological Laboratory, Gothic, Marmota, Yellow-bellied Marmot, Colorado East River Valley.
Items connected by shared entities, co-authorship, citations, or semantic similarity.
9 in Knowledge Commons, 38 external