PS 98-170
Classification and mapping of upland prairie and characterization of Dakota skipper habitat on the South Dakota Prairie Coteau

Friday, August 14, 2015
Exhibit Hall, Baltimore Convention Center
Diane Narem, Natural Resource Management, South Dakota State University, Brookings, SD
Lan Xu, Natural Resource Management, South Dakota State University, SD
Gary E. Larson, Natural Resource Management, South Dakota State University, Brookings, SD
Dave Ode, 3South Dakota Game Fish & Parks Department
Background/Question/Methods

Located in the tallgrass prairie ecoregion, the South Dakota Prairie Coteau is a unique geologic formation that retains a large amount of native tallgrass prairie and provides refuge for numerous species of concern including the federally threatened Dakota skipper (Hesperia dacotae) butterfly. To develop a targeted conservation plan for the Dakota skipper, upland prairie was classified and mapped within a 15 by 15 mile (58,275 ha) region of the SD Prairie Coteau, Dakota skipper habitat was characterized and potential habitat was identified. Upland prairie communities were sampled using the relevé sampling protocol, classified using multivariate analysis and mapped in ArcGIS according to associated ecological sites. Characterization of Dakota skipper habitat was conducted using 50 m transects subjectively placed in high quality dry-mesic prairie at sites inhabited by the Dakota skipper and at sites where it had been extirpated. Cover by species using modified Daubenmire classes was estimated in six 1-m2 quadrats placed along transects, alternating every 10 m, and flowering stems were counted in a two-meter belt along each transect. Transects were used as experimental units by averaging quadrat data per transect and ordinated using Nonmetric Multidimensional Scaling (NMDS). Individual species and functional groups were compared using Mann-Whitney U tests.

Results/Conclusions

Multivariate analysis of the relevé plot data found that axis one was driven by a moisture gradient and explained 55% of the variation, while management regime and phenological variation both influenced axis two which explained 22% of the variation. Classification analysis using flexible beta = -0.25 and Sorenson distance produced three plant communities: Upland Dry Prairie, Dry-Mesic Prairie and Mesic Prairie. On the final map Upland Dry Prairie, Dry-Mesic Prairie and Mesic Prairie covered 13537.93 ha, 31049.06 ha and 3167.98 ha of land respectively. A potential habitat layer was created in ArcGIS by intersecting Upland Dry Prairie and Dry-Mesic Prairie classifications with an undisturbed grassland layer. The condition and management of the grasslands in this layer need to be assessed to determine actual habitat suitability. The NMDS of the transect data found no clear pattern between vegetation composition at inhabited and extirpated sites, indicating that other factors such as management and critical minimum size also play an important role in the extirpation of local populations. The classification and mapping of upland communities and the potential habitat layer will provide a guide to future Dakota skipper surveys and aid in developing a recovery plan.