Once seeded, half the plots within each treatment are left alone to assemble without further intervention. The other half of the plots are resown with any species that fail to establish, or that disappear after having established initially. The first treatment broadly corresponds to a theoretical recipe of letting excessively species-rich systems collapse to stable end points (Tregonning and Roberts, 1979; Roberts and Tregonning, 1981), the second to methods of allowing both invasion and local extinction during assembly (Post and Pimm, 1983; Drake, 1990a).
During the first 4 years, plot diversity varied considerably in the 12- and 16-species treatments, but remained steadier in the 4- and 8-species treatments (Figure 1A). Lower diversity in the fourth year for the higher diversity treatments was due to declining numbers of weeds in these plots (Figure 1B). Plant communities of
healthy prairie are over 99% perennial species, whereas annuals constituted most of the cover within the community assembly plots. By the second year, 1995, however, the percentage of cover by annuals was already decreasing as treatment diversity increased. In the fourth year, weedy cover in the most diverse treatment was ~1/10 that of the least diverse treatment (7 vs. 65%). Even field bindweed (Convolvulus arvensis), a pernicious perennial weed, declined as treatment diversity increased (data not shown). Of the four major plant guilds, percentage cover by perennial C3 grasses and composites increased significantly with diversity treatment, and legumes showed a trend in this direction (Figures 2 and 3). Percentage cover by perennial grasses and composites is on a steady upward trajectory, with overall cover by perennial composites, in particular, increasing more than 12-fold from 1994 to 1997.
Was this article helpful?