期刊
BIOTROPICA
卷 41, 期 1, 页码 16-26出版社
WILEY
DOI: 10.1111/j.1744-7429.2008.00451.x
关键词
arboreal palms; canopy trees; Central America; dispersal assembly; environmental factors; niche; tropical rain forest
类别
资金
- NSF-IGERT [0114304]
- Organization of American States
Studies of tropical rain forest beta-diversity debate environmental determinism versus dispersal limitation as principal mechanisms underlying floristic variation. We examined the relationship between soil characteristics, terrain, climate variation, and rain forest composition across a 3000 km(2) area in northeastern Costa Rica. Canopy tree and arboreal palm species abundance and soils were measured from 127 0.25-ha plots across Caribbean lowlands and foothills. Plot elevation, slope, temperature, and precipitation variation were taken from digital grids. Ordination of forest data yielded three floristic groups with strong affinities to foothills and differing lowland environments. Variation in floristics, soil texture, and climate conditions showed parallel patterns of significantly positive spatial autocorrelation up to 13 km and significantly negative correlation beyond 40 km. Partial Mantel tests resulted in a significant correlation between floristic distance and terrain, climate and soil textural variables controlling the effect of geographical distance. Separate comparisons for palm species showed significant correlation with Mg and Ca concentrations among other soil factors. Arboreal palm species demonstrated a stronger relationship with soil factors than did canopy trees. Correlation between floristic data and geographical distance, related to seed dispersal or unmeasured variables, was not significant after controlling for soil characteristics and elevation. Canopy trees and palms showed differing relationships to soil and other environmental factors, but lend greater support for a niche-assembly hypothesis than to a major role for dispersal limitation in determining species turnover for this landscape.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据