| Summary: | 碩士 === 東海大學 === 生命科學系 === 100 === Understanding mechanisms of species coexistence is one of the most important themes in community ecology. Spatial distributions of plants have been widely applied to explain the theories of species coexistence. Previous studies, however, have emphasized differences in species abundance. Only few studies have documented spatial patterns of plant functional traits and link trait distributions with spatial variability in resource acquisition among coexisting species. Two hypotheses, habitat filtering and limiting similarity hypotheses, are proposed to explain the spatial pattern of functional traits for coexisting species. In this study, I used a trait-based approach to investigate whether the habitat filtering process occurs in a spatially heterogeneous forest in Taiwan. Three predictions were examined in this study. First, there is spatial autocorrelation of leaf functional traits. Second, plant functional traits within a habitat are similar and differ from random combinations. Third, the distributions of functional traits are correlated with environmental factors. This study was carried out in the 25 hectare of the Lienhuachih Forest Dynamics Plot. For each woody species, three leaves of the six smallest individuals were collected. Six leaf functional traits were measured, including specific leaf area (SLA), leaf thickness (TH), leaf dry matter content (LDMC), leaf area (LA), carbon to nitrogen ratio (C/N) and nitrogen to phosphorus ratio (N/P). Global Moran’s I was used to detect spatial autocorrelation of leaf functional traits and permutation tests were applied determine if the distribution of functional traits deviate from randomness. From February 2009 to February 2011, 2396 leaves samples from 828 small individuals of 130 species were collected. The results indicated that the detectable spatial patterns exist for six leaf traits, and spatial autocorrelation occurs at spatial scales of 21.5 - 72.4 m. Such spatial patterns were significantly different from the random combinations. Finally, our GLS models (generalized least squares regressions) showed that leaf functional traits were significantly correlated with topographic and soil factors. Our analyses suggested the importance of environment filtering in the Lienhuachih forest. In the future, phylogenetic relationships among species may be added in the analysis to further explore mechanisms of species coexistence in the Lienhuachih forest.
|
|---|
