Journal of Tropical Ecology

Do current stem size distributions predict future population changes? An empirical test of intraspecific patterns in tropical trees at two spatial scales

Kenneth J. Feeley a1c1, Stuart J. Davies a2, Md. Nur Supardi Noor a3, Abdul Rahman Kassim a3 and Sylvester Tan a4
a1 Center for Tropical Forest Science, Arnold Arboretum Asia Program, Harvard University Herbaria, 22 Divinity Avenue, Cambridge MA 02138, USA
a2 Center for Tropical Forest Science, Smithsonian Tropical Research Institute, Unit 0948 APO AA 34002, Panama
a3 Forest Environment Division, Forest Research Institute Malaysia, Kepong, Kuala Lumpur 52109 Malaysia
a4 Forest Research Centre, Sarawak Forest Corporation, Jalan Datuk Amar Kalong Ningkan, 93250 Kuching, Sarawak, East Malaysia

Article author query
feeley kj   [PubMed][Google Scholar] 
davies sj   [PubMed][Google Scholar] 
noor mn   [PubMed][Google Scholar] 
kassim ar   [PubMed][Google Scholar] 
tan s   [PubMed][Google Scholar] 


It is critical to understand the responses of tropical tree species to ongoing anthropogenic disturbances. Given the longevity of large trees and the scarcity of appropriately long-term demographic data, standing size distributions are a potential tool for predicting species' responses to disturbances and resultant changes in population structure. Here we test the utility of several different measures of size distribution for predicting subsequent population changes at the intraspecific level using demographic records from two subsampled 50-ha tree plots in Malaysia (Pasoh and Lambir). Most measures of size distribution failed to successfully predict population change better than random; however, the ‘coefficient of skewness’ (a measure of the relative proportion of small vs. large stems in a population) was able to correctly predict the direction of population change for approximately three-quarters of species at both sites. At Pasoh, the magnitude of this relationship decreased with adult stature and rate of turnover, but was unrelated to sapling growth rates at either site. Finally, using data for species common at both forests, we found that size distributions were generally uninformative of subsequent differences in population change between sites (only median dbh correctly predicted the direction of change for more species than random). Based on these results we conclude that some measures of intraspecific differences in size distribution are potentially informative of population trends within forests but have limited utility across broader spatial scales.

(Accepted November 20 2006)

Key Words: coefficient of skewness; forest dynamics; Malaysia; population dynamics; Weibull distribution.

c1 Corresponding author. Email: