EDAPHIC VARIATION AND THE MESOSCALE DISTRIBUTION OF TREE SPECIES IN ANEOTROPICAL RAIN-FOREST

Citation
Db. Clark et al., EDAPHIC VARIATION AND THE MESOSCALE DISTRIBUTION OF TREE SPECIES IN ANEOTROPICAL RAIN-FOREST, Journal of Ecology, 86(1), 1998, pp. 101-112
Citations number
43
Categorie Soggetti
Ecology
Journal title
ISSN journal
00220477
Volume
86
Issue
1
Year of publication
1998
Pages
101 - 112
Database
ISI
SICI code
0022-0477(1998)86:1<101:EVATMD>2.0.ZU;2-#
Abstract
1 Do local edaphic factors over short environmental gradients affect t he distribution and abundance of tree species in tropical rain forests ? We addressed this question by examining the responses of tree specie s to soil type, topographic position and slope angle in an upland old- growth tropical rain forest landscape in Costa Rica, Central America. 2 The study area covered 216 ha of non-swamp old growth and included r eplicated landscape units such as ridgetop to swale catenas, small wat ersheds, and alluvial terraces. An existing soils map was refined usin g additional sampling along a regularly spaced grid. Three soil units were defined: residual soils derived from in-place weathering of basal tic parent material; old alluvial terrace soils formed by precontempor ary flooding; and soils of stream valleys. A Geographic Information Sy stem was used to assign soil type to 2190 post-establishment individua ls of nine tree species in a longterm demographic study. Topographic p osition and slope angle were measured in the field. 3 Data from 433 re gularly spaced sample points were used to generate an expected distrib ution of edaphic variables, which was compared with the number of indi viduals of each species in each edaphic category. 4 Non-random distrib utions among soil types were found for seven of the nine species, with topographic positions for six species, and with slope angle classes f or four species. 5 The issue of what constitutes an independent sample of establishment was analysed by considering the old-growth character of the landscape and evidence for frequent dispersal among edaphic un its. A re-analysis of species' distributions using only individuals > 4cm diameter showed that results from the original analyses were robus t. 6 Soil type (residual vs. old alluvial) was not significantly relat ed to diameter growth. Three species showed significant differences in size class distributions between soil types with increasing diameter. 7 Tree species in tropical rain forests are frequently non-randomly d istributed along relatively short edaphic gradients on upland soils. F uture studies should increase the number of species and spatial scale analysed, incorporate better analyses of edaphic variables, and includ e experiments to identify the ecological processes that generate these non-random distributions.