THE GENECOLOGY OF LAWN WEEDS

Abstract

In order to investigate the hypothesis that disruptive selection is acting on Poa annua L. populations within a mosaic environment of bowling greens and flower beds, two experimental approaches were employed. (1) Studies of samples of seedlings, raised from the seed present in soil and surface debris from the two bowling greens and one flower bed, established that in each site the seed population was more variable than the adult. This result is compatible with the notion of disruptive selection. (2) An experiment was set up to examine the response of the two growth forms to clipping and to estimate the effectiveness of clipping as a selective force in the bowling green environment. In the control (unclipped treatment), the erect and prostrate variants differed both in total production and in their relative energy allocation to both vegetative and floral reproduction. These differences have been discussed in relation to r and K selection strategies. Differential response to clipping was observed; the groups in descending order of relative fitness consisted of prostrate individuals from the two bowling green populations > prostrate individuals from the flowerbed populations > erect individuals from flower bed populations. Clipping appears to have two effects on the erect forms resulting in: (1) a greater percentage decrease in total dry matter production in the erect compared with the prostrate forms, and (2) a greater percentage of the dry weight produced being removed by the cut in the erect forms compared with the prostrate forms.Coefficients of selection acting against erect plants under a regime of clipping ranged from 0.53 to 0.68. A selection coefficient of 0.77 was obtained for prostrate plants in the absence of clipping. In the light of the management of the lawns and flower beds, evidence for selection and gene flow is discussed.