The Correlation And Variation Of The Functional Traits In Hainan Tropical Cloud Forest

Tropical cloud forest is characterized by a frequent cloud cover, usually at the canopy level, which located in the humid tropics and high altitude forest. Compared with the lower altitude vegetation, tropical cloud forest shows an unique environment and generally the lower plant diversity. It is an interesting ecological question how the plants adapt to the environment in tropical cloud forest. Plant functional traits can objectively express the adaptability of plant to external environment, and also can help we know the variation and trade-off relationships of different function traits. It is useful for people to understand the ways and strategies of plant adaptating to the environment, it also can be used as the evidence to predictive species distribution and environmental changesResearch on the subject is the tropical cloud forest of Bawangling Natural Reserve, located in Haina, China. The species were identified and described according to their composition, and the function traits(leaf mass area, LMA; leaf draw weight, LDW; chlorophyll, Chl; leaf thickness,LTh, the hight of plant, H; wood density, WD) of the plants whose DBH is more than5cm. The characteristic of soil nutrients in tropical cloud forests were measured seriously. The relationship between species diversity and environmental factor was assessed, as well as the relationship between function traits across intraspecific, interspecific and community levels, and across5m X5m,10m x10m and20m x20m plot sizes, and the variation of function traits in across intraspecific, interspecific and community levels. And we analysed the relationships between the soil nutrients and function traits. The main results are as follows:1) There were9323individual plants, belonging to40families,70genera and109species in8400m2plots. The dominant families were Hamamelidaceae, Myrtaceae, Symplocaceae and Fagaceae, ea al. and the dominant species were Distylium racemosum, Syzygium buxifolium, Psychotria asiatica and Engelhardia roxburghiana et al. Both the relationships of species richness-area and species richness-abundance were logarithm curves, with the minimum sample area1,000m2and the minimum sample abundance1,000. The relationships between the threshold of DBH and species richness, abundance, Simpson index and Shannon-wiener index all exhibited inverted J-shaped curves. The maximum threshold of DBH is3cm when the surveyed community has95%species richness. Stepwise linear regression analysis showed that community abundance, species richness and bootstrap index had strong relationships with soil available nitrogen content and soil available phosphorus content.2) H, LA, LMA, Chl, LTh and WD examined were significantly positively related. The slopes of Type II regression with1or-1had remarkable difference. The slopes of LMA-Chl, and LMA-LTh relationships did not change along with the levels of within-species, among-species, and community, and the three plot, indicating that these relationships are scale-independent. Contrastingly, relationships between the other eight trait pairs changed across the scales, suggesting that these functional trait relationships are scale-dependent.3) The changes of explained range variance of individual, intraspecies, interspecies and community levels were respectively0.06-0.47,0.09-0.35,0.35-0.72,0-0.07. The variation was the minimum in the community level, the variation of the individual, intraspecies and interspecies levels were bigger. The stepwise regression analysis indicated that the functional traits had strong relationships with soil organic matter, nitrogen and phosphorus.