Co-occurring tree species show contrasting sensitivity to ENSO-related droughts in planted dipterocarp forests

This study explores the diversity in sensitivity to drought of moist tropical forest tree species. Yearly tree growth records collected over a ten-year period in two one-hectare 70-year-old damar agroforest plots in Sumatra are analysed. These agroforests are mixed tree plantations, dominated by Shorea javanica K. & V., a dipterocarp tree cultivated and tapped for its commercially valuable resin (damar). Many indigenous fruit tree species grow in these agroforests, as well as timber tree species originating from the nearby natural forest. During the census period the multi-species stands were subjected to three El Nino Southern Oscillation (ENSO)-related droughts (1994, 1997 and 2002). At the tree community level, these droughts were associated with a marked decrease in radial stem growth. Multilevel modelling was used to explore the relative contribution of species, tree size and individual tree characteristics to the observed response to drought. All tree species appeared to be sensitive to drought but the amplitude of the response varied significantly across species. Predicted species mean decrease in stem radial growth rate on drought years (i.e. years with 6 months with less than 50 mm/month rainfall) ranged from less than 5% to more than 80%. Shared species were ranked consistently between plots indicating that the results were robust. Stem diameter significantly affected tree sensitivity to drought in two species only, but in opposite ways: in S. javanica, larger trees appeared to be less sensitive while the opposite was true for Lansium domesticum, an abundant fruit tree. Individual tree sensitivity to drought contributed significantly albeit to a small extent to the overall response to drought. This individual tree effect did not show any pattern of spatial correlation and hence could not be related to topographic features. It is likely to reflect the individual's unique history and genotype. © 2009 Elsevier B.V. All rights reserved.