CONVERGENCE AND DIVERGENCE OF OLD-FIELD VEGETATION AFTER 11 YR OF NITROGEN ADDITION

Eleven years of data from experimental nitrogen gradients in three old fields and in native savanna are used to test the hypothesis that similarity of resource availability should influence similarity of plant species composition, Vegetation became less similar on plots that received different amounts of nitrogen, The difference in vegetation between plots was a function of both the difference in added nitrogen and the absolute levels of added nitrogen, Changes in similarity on plots that received the same amount of nitrogen were influenced by initial species composition, the amount of added nitrogen, and the rate of colonization by species not initially present on plots, Replicates of many treatments initially became less similar, only to increase in similarity after 3-6 yr. Data are generally consistent with predictions made in 1988, and clearly illustrate the importance of long-term responses to experimental manipulations.