Previously reported effects of richness and evenness on grassland productivity and invasion resistance may result from changes in species interactions that are confounded with richness and evenness. Altering richness and evenness changes the extent to which individuals experience heterospecific interactions and affects the mechanisms behind productivity and invasion responses, but it is unclear at what scales grassland species typically interact to determine these responses. In May 2010 we seeded an experimental field study where we manipulated fine-scale plant arrangement but maintained coarser-scale diversity to address the question: Does initial seeding pattern affect mechanisms that maintain coarser-scale (plot) diversity, productivity, and invasion in grassland communities? Plots (16m2) were seeded with 15 native tallgrass prairie species and one non-native cool season grass. Species were seeded into a 1m2 monoculture in each plot, either as one large patch or divided into 4 – 0.25 m2, 16 – 0.0625 m2, or 64 – 0.0156 m2 monoculture patches. We test the hypotheses that at similar initial coarse-scale (plot) diversity, plots planted with smaller initial patches of seed will be more diverse, more productive, and less invaded than plots with larger initial patches due to decreased heterospecific interactions and negative effects of patch size on invasion resistance.
Results/Conclusions
By the end of the first growing season, plots seeded with smaller patches were more productive and had greater overyielding than plots seeded with larger patches. Productivity in the first year was determined by interactions at scales of less than 0.25 m2, as productivity was significantly higher in plots seeded with 0.0156 m2 and 0.0625 m2 patches than in plots seeded with 0.25 m2 and 1 m2 patches. There was an early effect of seeding patch size on invasion of Chenopodium album and by the end of the second growing season invaders were more abundant in plots seeded with 1 m2 patches. Results suggest that species pattern at sub-meter scales affects grassland productivity and invasion responses, an effect which should be considered when manipulating richness and evenness in biodiversity ecosystem-function studies.