THE CASE FOR EXTENSIVE RATHER THAN INTENSIVE SAMPLING IN GENECOLOGY

S ummary This paper is primarily concerned with the technique of genecological investigation. The data refers to thirteen taxa sampled at the rate of up to five individuals per site, and in up to eighty sites, all within the Agrostis‐Festuca grassland community. Most of the fifty‐four species‐characters recorded are significantly different between populations using the within‐population variance as error, but on further investigation little of this is found to have genecological significance: it is believed to be due to genotype reduplication in sampling. Basing analyses on population means and either employing the analysis of variance of grouped data or correlation with a habitat factor, few of the characters reach significance and for the most part these characters seem to be indubitable examples of genecological differentiation. The genecological conclusions were that within the local Agrostis‐Festuca grassland the more leached sites carry the earlier populations of Cerastium vulgatum and Poa trivialis and the longer‐leaved or taller vegetative plants of Carex caryophyllea and Poa trivialis . There was some evidence in Galium spp. that apparently genecological differentiation could result from introgression. Interspecific correlation of genetical characters was investigated but no significance interpretable genecologically was found in this study. It seems likely that the species showing differentiation in the flushed to leached series of sites are differentiated in relation to different underlying environmental factors. The place of extensive methods in genecology is discussed and possible developments indicated.