FACTORS INFLUENCING SPECIES COMPOSITION IN TROPICAL LOWLAND RAIN-FOREST - DOES SOIL MATTER

Authors
Citation
P. Sollins, FACTORS INFLUENCING SPECIES COMPOSITION IN TROPICAL LOWLAND RAIN-FOREST - DOES SOIL MATTER, Ecology, 79(1), 1998, pp. 23-30
Citations number
46
Categorie Soggetti
Ecology
Journal title
ISSN journal
00129658
Volume
79
Issue
1
Year of publication
1998
Pages
23 - 30
Database
ISI
SICI code
0012-9658(1998)79:1<23:FISCIT>2.0.ZU;2-4
Abstract
The soil properties most likely to influence species composition in lo wland order of importance: P availability, Al toxicity, drainage, wate r-holding capacity, and availability of K, Ca, and Mg. A total of 18 s tudies were located in which species occurrence was studied in relatio n to such soil properties. Several of these report clear trends with s oil physical properties, mainly drainage as affected by topographic po sition. Only three offer evidence for correlations with soil chemical properties. In all three, the study area spanned soils of very widely differing age and thus soil fertility. Failure to find correlations wi th chemical properties may be due to: lack of range in soil fertility across the sites studied. failure of soil test methods to measure avai lability of nutrients to plants, or temporal and spatial variability i n soil properties.Existing soil classification systems do not provide enough information on any particular soil to help establish relations between its chemical properties and plant distribution. Nonetheless, s oils at ecological study sites must be classified if their nature and properties are to be made clear to others working worldwide. Tradition al, largely subjective, soil-mapping ;; techniques may reduce sampling needs by allowing stratification of soil and plant sampling by broad soil types. The mapping must be done, however, at a scale similar to t he patchiness of the plant community (usually finer than 1:5000), some thing done in only 3 of the 18 studies located. Correlative studies; a re ?only, the ;first step in understanding causal relations between so il properties and plant species distribution. Next, nutritional, drain age, and water requirements must be established for individual species . Then field experiments must be set up to establish cause and effect.