Search

The correlation between primary producer diversity and the productivity (or biomass) of ecosystems is one of the most important and broadly studied relationship in ecology. Ecologists have discussed this relationship from two fundamentally different perspectives. Historically, productivity has been viewed as driver of species diversity. Recently, many studies have demonstrated that diversity can also control, rather than responds to, the production of biomass. These contrasting points of view have led to the debate about whether species diversity is the cause or the consequence of community productivity. Multivariate productivity-diversity hypothesis has been put forward to reconciling this debate. This hypothesis state that: (i) the environmental factors are the direct driver of species that can coexist within an area; (ii) the biomass of the area is directly influenced by the environmental factors that limit production; (iii) the environmental factors indirectly influenced biomass of the area via influencing the species number to coexist within the area that affects how efficiently environmental resources are converted into biomass. To date, empirical support for this hypothesis is scarce, especially for structurally complex terrestrial ecosystems. In this study, we modified multivariate productivity- diversity hypothesis to accommodate complex vertical structure of forest ecosystems, and used structural equation modeling and data from a large evergreen broad-leaved forest dynamic plot (24 hm2 in area) in subtropical China to understand the causal relationships among environmental factors, species richness, dominance and aboveground biomass. Trees were grouped into tow functional groups (overstorey and understorey) in order to improve the ability to detecting the diversity effect. The final model explained 30% and 58% of the variation in aboveground biomass of overstorey and understorey, respectively. Dominance was the most important factor in explaining the variation of aboveground biomass of overstorey, while both species richness and environment factors were the most important factors in explaining the variation of aboveground biomass of understorey. Consistent with the predictions of multivariate productivity-diversity hypothesis, (i) aboveground biomass were directly influenced by the environmental factors in both overstorey and understorey; (ii) environmental factors directly influence tree dominance and species richness of overstorey, but only directly influence species richness of understorey; (iii) environmental factors indirectly influence aboveground biomass of overstorey by influencing dominant, while indirectly influence aboveground biomass of understorey by influencing species richness. Interestingly, species richness was only important to explaining variation in aboveground biomass of understorey. This may be attributed to the low light condition in understorey, which may alleviative competition among trees of understorey. To our knowledge, this study provides the first empirical evidence supporting the multivariate productivity-diversity hypothesis in forest ecosystem thus contributes to extending the generality of this framework. Our result also highlights both dominant species and higher species richness is required to maximizing forest biomass.

The aims of this study were to investigate the diurnal and seasonal variations in soil respiration in different broad-leaved forest types with different degrees of human disturbance, to identify the major abiotic and biotic factors affecting the daily and seasonal patterns of soil respiration, and to evaluate the effects of different forest management types and human disturbance on soil respiration in subtropical forests. We selected four different 1 hm2 forests with different degrees of human disturbance in Gutianshan, Zhejiang Province: An old-growth forest (OF, almost no human disturbance); two secondary forests (SF1, a secondary forest regrown after clear cutting about 50 years ago, and SF2, a secondary forest regrown after clear cutting about 50 years ago and selective cutting about 20 years ago); and a plantation forest (AF, planted about 20 years ago). We measured the diurnal and seasonal changes in soil respiration rates in each plot every month from May 2011 to July 2012. We also measured soil temperature and water content of the surface soil layer during sampling. The annual means of daily accumulated soil carbon release were 1.48, 1.48, 1.51, and 0.87 g C m-2 d-1 in OF, SF1, SF2, and AF plots, respectively. The soil respiration rate in AF was significantly lower than those in the other three forest types. On the diurnal scale, there were no significant variations in soil respiration rates among the four forest types. Our results indicated that the rate of soil respiration measured any time during the day can be used to predict daily accumulated soil carbon release. Significant positive relationships were found between soil respiration and surface soil temperature ( R 2=0.88-0.94) for all four forest types. There were no significant relationships between soil respiration and soil water content. The Q10 value of OF was significantly higher than those of the other three forest types. The results showed that the soil respiration rate and its sensitivity to temperature were significantly decreased in the plantation with the most intense human disturbance (AF). The soil respiration rates and Q10 values of both SF forests were similar to those of OF forest. Soil temperature was the most important driving factor for soil respiration. Together, our results implied that different degrees of human disturbance had different effects on soil respiration in these forests. This information is important to estimate the role of human interference in regional carbon cycles. Also, the finding that there was no significant variation in the diurnal soil respiration rates is very important for guiding further research on soil respiration in related regions.

With the full survey data for a 24-ha subtropical evergreen broad-leaved forest dynamics plot, we evaluated spatial variation in forest structure characteristics (basal area and aboveground biomass), and calculated the minimal sample size and total sampling area necessary to estimate the forest structure characteristics within 20% (±10%) of the observed values with 95% probability for particular quadrat sizes by using a computer program that is designed to simulate the sampling process by allowing different sized quadrats to be randomly located within the sampling region. We found that (1) based on the 600 20 m×20 m subplots, basal area and aboveground biomass displayed a high degree of variation, with respective coefficients of variation of 27% and 31%; (2) based on the computer simulation analysis, the variability of basal area and aboveground biomass decreased with increasing quadrat size. The number of quadrats required to achieve the specified degree of precision dropped sharply with the increase of quadrat size. However, the total sampling area increased with increasing quadrat size, suggesting that using several small quadrats across the sampling area is more efficient than using fewer larger quadrats. Results of this study are valuable for evaluating the reliability of previous research and may assist researchers in designing effective sampling strategies for future field surveys, particularly in subtropical evergreen broad-leaved forests in China.