The relationship of total and per-gram rankings in competitive effect to the natural abundance of herbaceous perennials

1 Using a field experiment and a garden experiment, I estimated the rankings in total and per-gram competitive effect of non-woody perennial old-field species. 2 Total competitive effects were defined as the relative reduction in growth of a target from no-neighbour to with-neighbour conditions. Per-gram competitive effects were defined as the per-unit relative reduction in target growth among increasing neighbour densities, and were determined from the shape of a nonlinear curve fit through a distribution of normalized target performance against neighbour mass. 3 In both experiments, mean total competitive effect differed significantly among species, indicating a strong competitive hierarchy. In the garden experiment only species at opposite ends of the ranking differed significantly in per-gram competitive effect, resulting in a weaker competitive hierarchy based on this measure. 4 Nonetheless, rankings of per-gram competitive effect were more strongly correlated with rank in abundance than were rankings of total competitive effect. 5 Per-gram competitive effect may be more predictive of natural abundance than total competitive effect for at least two reasons. The effects of neighbour abundance on targets are nonlinear, and unlike total effects, per-gram estimates of competitive effect may therefore indicate how competition changes over time with changing neighbour densities. Also, if higher per-gram competitive effect reflects higher per-unit nutrient uptake rates, it would probably be advantageous to a species throughout the individual's life span, rather than only when the individual is larger than its surrounding neighbours.