OPS 1-15 - Classification of Pawnee National Grasslands: Challenges of fitting locally-geographic plot-based classifications into the USNVC

Monday, August 7, 2017
Exhibit Hall, Oregon Convention Center
Michael H. Schiebout, Biology, Union University, Jackson, TN, Scott B. Franklin, Biological Sciences, University of Northern Colorado, Greeley, CO, Jozef Shibik, Institute of Botany, Slovak Academy of Sciences and Don Hazlett, Rocky Mountain Research Station, U.S. Forest Service
Background/Question/Methods

As part of the “U.S. National Vegetation Classification: Advancing the Description & Management of the Nation’s Ecosystems” organized poster session, this work presents an example of fitting a local classification into the US National Vegetation Classification (USNVC). The Vegetation Subcommittee of the Federal Geographic Data Committee developed a standard for vegetation classification in the United States. This classification of the Pawnee National Grassland follows that standard as well as international standards. A posiplot method was used to locate plots, first based on ecological land types and subsequently on visual vegetation communities. A total of 102 plots were located based on homogeneity of vegetation and site characteristics. Plot sampling followed the Carolina Vegetation Survey method, which uses ten 10X10 modules (0.1 ha) to assess presence and canopy cover of all taxa. In an attempt to compare PNG plots with those already classified, we retrieved an additional 64 plots from four other regional research projects within VegBank. We classified data using a hierarchical cluster analysis using Sorenson distance measure and -0.25 Flexile Beta group linkage method: data were square-root transformed. We determined the appropriate number of groups using OptimClass. We compared diagnostic species of our classified groups to USNVC published concepts. 

 Results/Conclusions

Analysis of the Pawnee National Grassland only plots identified two major groups with arid Bouteloua gracilis-Buchloe dactyloides Grassland Association being separated from mesic vegetation communities. These two major groups could be future divided into either 4 or 10 community clusters. The ten cluster dendrogram essentially splits the two major groups into four groups each. Although, caution should be used in interpreting the mesic clusters because so few plots were sampled; total plots = 20, each group ranged from 3-7 plots. Regional analyses clearly separated more mesic communities from mixed grass and short grass steppe. In addition, the Bouteloua gracilis-Buchloe dactyloides Great Plains Shortgrass Prairie Macrogroup was separated from the Hesperostipa comata-Pascopyrum smithii-Festuca hallii Grassland Macrogroup to the north and east. The striking difference of flora affirms the classification on the Pawnee sites as shortgrass steppe and negates the occurrence of mixed grass communities in the Pawnee National Grasslands. Our substantial data characterize the Bouteloua gracilis-Buchloe dactyloides Grassland Association dominating the PNG. Because these data are from a limited area within the entire range of the Association, we simply document the characteristics typical for the PNG. In addition, we offer characteristics of three subassociations helpful for local management.