A technique for measuring the density and complexity of understorey vegetation in tropical forests

The dense understorey regrowth after disturbance of tropical forests has been linked to serious declines in many understorey and terrestrial animal taxa. We devised a method of measuring changes in vegetation characteristics using standardized photographs of the forest understorey. We describe the method using a sample dataset of photographs taken at 28 stations in primary forest and forest that had been logged selectively eight years previously, at the Danum Valley Field Centre in lowland Borneo. The understorey photographs were digitized and eight measures of vegetation density and complexity generated for each image using FRAGSTATS. Parameters included vegetation area, total vegetation edge, and mean fractal dimension of vegetation patches. There were no significant correlations between the vegetation density measures and those representing vegetation complexity showing there to be two unrelated components to understorey variability. We used PCA to derive two principal axes from the eight vegetation parameters. There were no significant differences in our measures of understorey density or complexity between the primary and logged forest stations. There was, however, a greater range of understorey characteristics in primary forest than in logged forest indicating a loss of understorey heterogeneity following logging. Most apparent was a loss of very open and non-complex understories and this may affect some animals in a number of ways, including altering the abundance of their prey and altering their ability to forage successfully in logged forest. (C) 2002 Elsevier Science B.V. All rights reserved.