Floristic quality assessment of one natural and three restored wetland complexes in North Dakota, USA

Abstract

Floristic quality assessment is potentially an important tool for conservation efforts in the northern Great Plains of North America, but it has received little rigorous evaluation. Floristic quality assessments rely on coefficients assigned to each plant species of a region’s flora based on the conservatism of each species relative to others in the region. These “coefficients of conservatism” (C values) are assigned by a panel of experts familiar with a region’s flora. The floristic quality assessment method has faced some criticism due to the subjective nature of these assignments. To evaluate the effect of this subjectivity on floristic quality assessments, we performed separate evaluations of the native plant communities in a natural wetland complex and three restored wetland complexes. In our first assessment, we used C values assigned “subjectively” by the Northern Great Plains Floristic Quality Assessment Panel. We then performed an independent assessment using the observed distributions of species among a group of wetlands that ranged from highly disturbed to largely undisturbed (data-generated C values). Using the panel-assigned C values, mean C values ( C ¯ "> C ¯ C¯ ) of the restored wetlands rarely exceeded 3.4 and never exceeded 3.9, with the highest values occurring in the oldest restored complex; all but two wetlands in the natural wetland complex had a C ¯ "> C ¯ C¯ greater than 3.9. Floristic quality indices (FQI) for the restored wetlands rarely exceeded 22 and usually reached maximums closer to 19, with higher values occurring again in the oldest restored complex; only two wetlands in the natural complex had an FQI less than 22. We observed that 95% confidence limits for species richness and percent natives overlapped greatly among wetland complexes, whereas confidence limits for both C ¯ "> C ¯ C¯ and FQI overlapped little. C ¯ "> C ¯ C¯ and FQI values were consistently greater when we used the datagenerated C values than when we used the panel-assigned C values; nonetheless, conclusions reached based on these two independent assessment techniques were virtually identical. Our results are consistent with the opinion that coefficients assigned subjectively by expert botanists familiar with a region’s flora provide adequate information to perform accurate floristic quality assessments.