Three approaches for analysing the dynamics ofthe horizontal layout of tillers in grass patches. Application totall fescue swards surveyed over two growing seasons

Abstract

Abstract The layout of the tillers was surveyed throughout two consecutive growing seasons in almost pure tall fescue swards. These swards were periodically cut and differed in age (2 or 6 years at the beginning of the survey) and nitrogen application level (150 or 350 kg N·ha –1 ·yr –1 ). Ten permanent quadrats in each of the four sward types were observed four times a year, after each cut. At each observation time, each quadrat was divided into a 10 × 15 grid of 2 × 2 cm cells and the living tillers counted in each. Three approaches for analysing this survey were compared. The means of the successive tiller numbers in each cell and their variability displayed a steady background of stand structure with stable sparse and stable dense areas. The young N-rich swards have similar proportions to these zones, while the old N-poor swards have larger stable sparse areas compared to stable dense areas. Calculating the differences between two successive observations cannot easily be used to locate the disappearance and the appearance of tillers, as they can change position. Nevertheless, between consecutive observations, the young N-rich swards seemed to be less conservative than the others. The principal component analysis gives a global analysis of the dynamics. In all sward types, the steady background accounted for about 50 % of variation with the main changes being between years, which accounted for about 20 % of variation. The results bring into question the relative importance of local fertility and the processes of clonal growth in the dynamics of a grass patch.