Abstract
The spatial pattern of plants in one generation may alter the competitive interactions between plants in two ways: (1) it may affect the survival or fecundity of competing individuals in that generation, or (2) it may influence the success of seedlings in the following generation. I performed field experiments, spanning two generations, that allowed me to measure how the spatial pattern of the grass Poa annua influences the performance of two species of dicotyledonous annual weeds (Capsella bursa—pastoris and Senecio vulgaris). Randomly distributed weeds were introduced into plots with either a random or a patchy distribution of Poa. I found that seedling numbers in the following generation were four to six times as high for Capsella and Senecio when Poa was patchily rather than randomly distributed. This was not due to differences in the fecundity and survival of first—generation Capsella and Senecio between the two Poa distributions. Instead, the improved success of the subsequent generation in the patchy distributions was due to increased emergence or survival of seedlings in areas with a low density of dead Poa adults. In a second experiment to investigate why seedling survival increased in the absence of dead Poa, I found that interference by dead grass blades was responsible for high seedling mortality.