No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Using recently developed statistical techniques, the structure of a chalk grassland community on the Kent North Downs was investigated with special reference to Brachvpodium pinnatum. This species and two others, Bromus erectus and Festuca rubra were, the main constituents of the sward and attention was therefore confined to these three. Pattern analyses were used to detect and measure scales of non-randomness in the spatial distributions of these species. Covariance analyses clearly indicated that mosaics of alternating patches of Bromus and Brachypodium were present at several different scales, and that Festuca tended to accompany Brachypodium. Trial applications of nutrients to the sward indicated that the soil levels of boron, nitrogen and iron were limiting the growth of the vegetation as a whole and pattern analyses of these variables indicated that all were contagiously distributed. A controlled experiment yielded results which suggested that while nitrogen applications favoured equally the growth of all three species, boron applications affected their growth differentially, especially favouring that of Brachypodium. Covariance analyses of contiguous quadrat data confirmed that in the field the non-random distributions of the species reflected the heterogeneities of the boron and nitrogen levels in the soil. They revealed also that the nutrient level heterogeneities controlled the Bromus and Brachypodium distributions through their differential effects on the competitive abilities of the two species which enabled Brachypodium to exclude Bromus from the high boron level areas and vice-versa. The effect on Festuca was more direct, however, growth being better at one nutrient level than at another, regardless of how that level affected the associated species. The effects of nutrient applications on patterns over a period of time were also examined, but except in the case of nitrogen applications, the results obtained were not sufficiently clear to further confirm the conclusions reached above.