Rainfall partitioning in relation to forest structure in differently managed montane forest stands in Central Sulawesi, Indonesia

Management activities alter the structure of many tropical forest stands which can be expected to influence the magnitude of canopy water fluxes. The objectives of this study were to determine throughfall, stemflow and rainfall interception in differently managed forest stands, and to relate the observed pattern of rainfall partitioning to stand structural characteristics. The study was conducted in a lower montane rainforest region (800–1140 m asl) in Central Sulawesi, Indonesia. Stands of four management types (natural forest, forest subject to small-diameter timber extraction, forest subject to selective logging of large-diameter timber, and cacao agroforest under trees remaining from the natural forest) were analyzed with three replicates per use type. The tree basal area decreased from the natural forest (52.5 m ha ) to the agroforest (19.4 m ha ) which was paralleled by a reduction in mean tree height (trees 10 cm dbh) from 21.3 to 17.5 m. The estimated leaf area index (LAI), as derived from hemispherical photos, averaged 6.2 m m 2 in the natural forest, 5.3 in forests with small timber extraction, 5.0 in forests with large timber extraction, and 5.3 in the agroforest. The annual gross precipitation close to our different study plots varied locally between 2437 and 3424 mm during the time of the study. Throughfall was measured on 15–17 rain days per plot with a median of 70% of gross precipitation over all the natural forest plots, 79 and 80% in forest with small and large timber extraction respectively, and 81% in the agroforest. Stemflow was less than 1% in all studied use types. Thus, rainfall interception was highest in the natural forest where 30% (median) of the gross precipitation was re-evaporated back into the atmosphere, and much lower in the three other use types (18–20%). Variability in structure and rainfall partitioning was high even within the same forest use types, thus further analysis focused on gradual changes rather than categories. In the 12 stands, LAI alone did not correlate significantly with the pattern of rainfall partitioning, the throughfall percentage increased significantly with decreasing tree height (r = 0.63). In a multiple linear regression with tree height and LAI as influencing factors, 81% of the variation in throughfall percentage is explained. A possible reason for this tree height-LAI-throughfall relationship is that under the conditions prevailing in our study region the canopy may not completely dry up between subsequent rainfall events. Therefore, the actual water storage at the start of a rainfall event would be below its potential. We hypothesize that tall trees increase the vertical distribution of foliage and other canopy components contributing to the canopy water storage, resulting in a higher canopy roughness and a more effective energy exchange with the atmosphere. This would consequently lead to an increased re-evaporation of intercepted water, larger available water storage and, thus, a reduced throughfall in stands with tall trees. # 2006 Elsevier B.V. All rights reserved.