The Use of Point-to-Plant Distances in the Study of the Pattern of Plant Populations

In a quantitative study of the pattern of a plant species in a given area, there are usually three facts which an investigator wishes to learn at the outset. These are: (1) the density, or number of individuals per unit area; (2) whether or not the individuals are dispersed at random; and (3) the degree of non-randomness if they are non-random. The measure of non-randomness must be such that in comparing two separate populations neither of which is random one can decide whether the departures from randomness in the two cases differ significantly from each other. David & Moore (1954) have proposed an index of clumping by means of which departures from randomness may be assessed and compared but their index is calculated from the results of quadrat sampling and may thus be greatly affected by the size of quadrat used (Greig-Smith 1957, Pielou 1957), In order to avoid this effect the use of di-tance measures has been suggested. Distances may be measured either from randomly chosen plants to their nearest neighbours, or from random points to their nearest plants. In addition, the density of the individuals must be found independently, either by quadrat sampling or by a complete count of all individuals in the area under study. Distances may also be measured to nth nearest neighbour for n > 1 (Thompson 1956) or to the neighbours in each of the several sectors (quadrants, sextants, etc.) into which a circle centred on the origin of measurement may be divided (Dice 1952, Clark & Evans 1954, Morisita 1954, 1957); but the effort expended in getting these results is hardly warranted in the preliminary study of a population and they will not be considered further here. There may thus be three distinct items in one's observational results: