Grassland fragmentation and its influence on woody plant cover in the southern Great Plains, USA

Context Woodland and agricultural expansion are major causes of grassland fragmentation. Fire and rainfall play important roles in maintaining grasslands, however, fire activity has been reduced in fragmented landscapes. Objectives Quantify the degree to which basic landscape fragmentation metrics could be used as drivers of woody cover potential. Methods Woody plant percent cover was calculated between 2004 and 2008 at > 2000 sites. At each site, we calculated these fragmentation metrics for grassland cover type (classified by the National Land Cover Database); # patches, landscape proportion, edge density, largest patch index, effective mesh size and patch cohesion index within 3 circular areas (10 km 2 , 360 km 2 and 3600 km 2 ) surrounding the sampling site. A quantile regression was performed to identify which metrics were useful at predicting the 25th, 50th, 75th or 95th quantile of woody cover distribution. Results Grassland proportion and edge density were significant predictors of the woody plant potential (75th and 95th quantile). Woody cover potential was positively associated with edge density suggesting that fragmented areas (i.e., areas with high number of edges) maintained higher woody cover, while grassland proportion was negatively associated with woody plant potential. Conclusion We propose that in addition to a lack of fire, fragmented landscapes may facilitate further woodland expansion by reducing natural land and restricting grasslands to smaller, less connected patches, which can maintain higher woody cover. Given current trends in woodland expansion, special attention should be given to areas that are found within a fragmented landscape and climatically prone to woodland expansion.