Estimation of Plant Density Using Line Transects

The line-transect method for the estimation of the percentage ground cover by different species in a stand is well established in theory and practice as giving a level of precision in the estimate for a given effort which compares very favourably with other methods. This is especially so when marked aggregation of species occurs. The use of transect data for auxiliary measures indicative of plant numbers for each species has been examined by Bauer (1943), using artificial populations. Starting with known numbers of circles of particular sizes and colours he computed the expected relative number of each type of circle contacting the transect (numerical abundance) and also the relative proportions of transects contacted by the various types (frequency). The estimated measures showed good agreement with experimental data obtained by random transects. The deficiency of 'numerical abundance' as a measure of abundance of a species is obvious, as it is determined both by density, that is the number of plants per unit area, and by the frequency distribution of diameters. Bauer does not consider the inverse problem of estimating density from transect data. The probability of a circle of a particular diameter being cut by a transect at random is proportional to an area surrounding the transect with boundary at radius distance from the transect line and its terminal points, and the expected number of contacts for all plants of the same diameter is proportional to this area and the number of plants in the class. This weighting of the frequencies in different diameter classes by a factor nearly proportional to the diameter is a source of difficulty in considering the inverse problem of estimating density from chords of intersected plants when the frequency distribution of diameters of the sampled population is not known. However, the possibility of measuring density from transect data is a matter of some interest to workers concerned with open grassland and low shrub communities and an examination of the point seems warranted.