Cooperative Fish and Wildlife Research Units Program: Mississippi
Education, Research and Technical Assistance for Managing Our Natural Resources


Ramirez-Reyes C., G. Street, F.J. Vilella, D.T. Jones-Farrand, M.S. Wiggers, and K.O. Evans. 2021. Ensemble species distribution model identifies survey opportunities for at-risk bearded beaksedge (Rhynchospora crinipes) in the southeastern United States. Natural Areas Journal 41(1):55-63.

Abstract

Locating additional occurrences of at-risk species can inform assessments of their status and conservation needs (including potential legal protections). The perennial bearded beaksedge (Rhynchospora crinipes) has been found from Mississippi to North Carolina, but known occurrences are limited. Because of the species’ apparent rarity, a model to identify areas with a high likelihood of locating additional occurrences will allow conservationists to effectively prioritize and allocate scarce surveying resources. We used known occurrence records, a suite of environmental datasets, and four species distribution modeling techniques (generalized additive, GAM; maximum entropy, MaxEnt; generalized boosted, GBM; and weighted ensemble) to generate maps to inform surveys for R. crinipes. The ensemble approach improved predictive performance (AUC-PR = 0.95) compared to other techniques (individual model AUC-PR ranged from 0.7 to 0.8). We also obtained quantitative insights on the species’ habitat relationships, including the likelihood of R. crinipes’s presence near Atlantic white cedar (Chamaecyparis thyoides) habitat and floodplains, which is consistent with prior field observations. The ensemble model indicated that 3.6% of the study area could be suitable habitat, but only 0.38% had high suitability. Small stream riparian habitats and Atlantic swamp forests in Alabama, Florida, and Georgia had the highest proportion of suitable areas. Prioritizing surveys in areas with model-indicated high habitat suitability is expected to reveal additional R. crinipes occurrences. We suggest surveying efforts for other at-risk species may benefit from using an ensemble modeling approach to identify and prioritize survey areas and improve ecological knowledge of these species.