Strategies for conserving or enhancing plant species richness in agroecosystems have generally focused on either expanding the land area in semi-natural non-crop habitat fragments or enhancing the diversity within arable crop fields through changes in within-field management practices, for instance by transitioning to organic methods. This dichotomy has been dubbed the “land sparing vs. land sharing” debate. Although complete patterns in plant diversity are often only apparent at landscape scales, most available data on plant diversity in agroecosystems has been collected at plot scales, preventing an accurate assessment of the relative benefits of land sparing vs. land sharing approaches. In this study, we shed light on the debate using data from an intensive survey of the vascular flora of Lancaster, PA, a major agricultural county in the Mid Atlantic region. Data were collected in summer 2008 in arable field, pasture, woodlot, and grassland habitat types (16,000m² sampled per habitat type) across a 225km² region of farmland. We used this data to produce a species area curve (SAC) for each habitat type and then combined extrapolations from the curves with estimates of among-habitat type community overlap to generate estimates of landscape scale species richness. We then modified these “baseline” estimates by adjusting parameters in the SACs to compare potential effects of land sharing and land sparing conservation practices on plant species richness in agricultural landscapes.

Results/Conclusions

The relative species richness of different habitat types strongly contrasted the relative land area of each type in the region, with 89 total found in arable fields (71% of total land area), 227 in grasslands (2.4%), 160 in pastures (8.1%), and 209 in woodlots (4.9%). For the baseline scenario, our model estimates that a 1km radius landscape (314ha) in this region would contain 576 plant species. Adjusting the SAC for the arable fields to reflect improving all fields to the upper 95 percentile of observed richness provided an increase to 603 species. However, converting just 2ha of arable fields into each of the three non-crop habitats also increased richness to 600 species. These results indicate that the overwhelming majority of plant species richness occurs in small fragments of non-crop habitat, and that sweeping changes in within arable field diversity may produce changes in plant species richness equivalent to a very modest conversion of cropland to non-crop habitats.