| Species diversity commonly used in previous ecological studies to characterize the changes in ecosystem function, and the relationship between the two objects had a long-term debate, the focus of debate was whether changes in ecosystem function was the direct effect of the species diversity? In many past studies, species diversity was equivalent to the functional diversity, and species richness was used to represent species diversity without taking into account the functional properties of the different species. This fact implied the assumption that different groups on the effect of ecosystem function is equal, which is bound to make the species-based biodiversity-ecosystem function relationship studies have significant limitations. With the in-depth study of ecosystem functions, people realized that the traits-based functional diversity was more effective than species diversity to characterize the changes in the functioning of ecosystems. Traits-based plant functional type, was in response to environmental changes, revealled the changes in vegetation dynamics and ecosystem function. Under the specific environmental conditions, the species would exhibit the trait-convergence assembly pattern(TCAP), which led to the niche overlap, but this pattern could not be lasting stability, interspecific competition would also lead to discrete traits and the trait-divergence assembly pattern(TDAP). If we could determine the TCAP and TDAP, we would be able to filter out the optimal trait subsets and plant functional types, and revealled the impact of environmental changes and interspecific competition on the changes in plant functional traits. A field manipulation experiment was conducted in alpine meadow at the Haibei Research Station of the Chinese Academy of Sciences from2007to2011. In the experiment, a split-plot design was adopted. Three clipping levels (stubbled1cm,3cm and unclipped) of clipping treatment were used on the whole plot and subplots were treated with or without fertilizer and watering. For25main species and12plant traits in communities, the recursive algorithm and multivariate analysis were implemented to search for optimal trait subsets and plant response types (PRTs), which could response to experimental treatments, and to identify plant effect types (PETs) impacting the aboveground net primary productivity of community. The results showed that:(1) A recursive algorithm to filter12plant traits, results show that:Under each of four resource available conditions:i.e. without fertilizer and watering (NFNW), fertilizer (F), watering (W) and fertilizer plus watering (FW), the optimal response trait subsets were different, i.e. leaf margin-plant height-leaves weight-specific leaf area, life cycle-plant height-leaves weight-specific leaf area, life cycle-chlorophyll content-leaf surface-plant height-leaves weight-specific leaf area, and organ of vegetative propagation-leaf margin-plant height, respectively. In the four optimal trait subsets, plant height occured four times, the frequency was100%, leaves weight and specific leaf area occured three times, the frequency was75%, while the remaining five traits appeared about1-2times, frequency≤50%, therefore, plant height, leaves weight and specific leaf area were more sensitive responses to treatments than others.(2) The analysis of variance for4qualitative traits in the optimal trait subsets, results showed that:clipping treatment had a significant effect on the leaf margin, life cycle and organ of vegetative propagation. and had a high significant impact on the leaf surface characteristics. With the increase of clipping, the leaf margin showed the transition from the crack edge to the entire margin; leaf surface characteristics in unclipped community showed the glabrous type was more than the hairy type, while in the moderate and severe clipping the hairy type was the majority; With the increase of clipping, rhizome plants and no-rhizome plants increased, straight-stem plants were replaced by low plants with the height reduction gradually. Fertilization treatment had a significant effect on the leaf margin, life cycle and organ of vegetative propagation; watering treatment had no significant effect on the four kinds of qualitative traits; the interaction of fertilization and clipping had a high significant impact on leaf margin, life cycle and leaves surface characteristics; watering and clipping treatment interactions had a significant effect on the leaf margin, and had a weak effect on the life cycle; fertilization and watering treatment interactions had a weak effect on the leaf margins; interaction of fertilizing, watering and clipping treatment had a significant effect on life cycle and had a weak effect on the leaf surface characteristics.(3) The analysis of variance for4quantitative traits in the optimal trait subsets, results showed that:clipping treatment had a significant impact on chlorophyll content, and had a high significant impact on height, leaf dry weight and specific leaf area; fertilization treatment increased height significantly, watering treatment significantly increased the content of chlorophyll; clipping and fertilizing treatment interactions only affected height highly significantly. Clipping treatment had a significant impact on all optimal traits, and showed some trends with clipping gradient, it indicated that the optimal traits by the recursive algorithm showed significant correlation with the clipping gradient, which proved the algorithm to study the effectiveness of plant traits with environmental variables relationship.(4) Using a recursive algorithm, we get the TCAP, TDAP and the maximum congruence in the optimal traits subsets along the clipping gradient, the results showed that:on clipping gradient, TCAP is greater than TDAP in the optimal trait subsets under the four kinds of fertilization and watering treatment combinations communities, but in NFNW communities, TCAP or TDAP were lower. In the four communities, trait redundancy size was FW>F>NFNW>W. In NFNW communities, mainly of plant traits for clipping treatment showed convergence response, and there was no significant divergence change. F, W and FW communities showed strong convergence and divergence, which also reflected the different role of water and fertilizer. That was because of traits convergence highest after fertilization and no causing the divergent response; watering treatment not only caused traits convergence, but also led to the highest divergence, FW communities showed medium convergence and divergence. That water treatment directionally affect the life cycle, plant height, leaf dry weight and specific leaf, fertilization treatment directionally influence the life cycle, leaf chlorophyll content, leaf surface structure, plant height, leaf dry weight and specific leaf area. Fertilizing and watering were the main reason to cause TCAP and TDAP in functional response traits along clipping disturbance gradients, which had a relation with clipping, soil nutrients and water resources interactions.(5) Under the conditions of four types of resource, we found14optimal PRTs among20PRTs, The optimal PRTs account for70%of all PFTs. Clipping increaser type, clipping decreaser type and clipping unimodal type were the optimal PRTs.(6) Under the conditions of four types of resource, we found4PETs among20PRTs, the optimal PETs account20%of all PFTs. PETs account for28.5%of the optimal PRTs, PETs partly overlapped with PRTs. This means that80%of the response function type or71.5%of the optimal response function type can only be used to characterize the changes in community characteristics, and can not reveal the changes in aboveground net primary productivity.(7) In NFNW, F, W and FW community, the PRTs and PETs had the same composition in traits, but their trait status value, the number of contained species and species composition were different; in different communities, traits composition, their trait status value, the number of contained species and species composition were different in PRTs and PETs. Despite of these differences, plant functional type in the same type of response pattern still had some common characteristics, such as the main feature of increased functional response pattern was small plant, lower leaf dry weight, larger leaf area; reduced functional response pattern showed tall plants, high leaf dry weight, low specific leaf area; unimodal response pattern was characterized by blade margin entire, short or medium height, medium leaf dry weight, low specific leaf area.(8) Stepwise regression analysis showed:in F and FW communities, there was a negative correlation between PETs and functional diversity; in W and NFNW communities, PETs was independent of functional diversity. There was resource-dependent relationship between PETs and functional diversity.These results indicate that the response of vegetation caused by clipping disturbance and functional changes of ecosystem could be accurately predicted by measuring plant functional traits which is usually relatively easy. However, the optimal trait subset, PRTs and PETs could be different accordingly when the heterogeneity of resources is taken into account. The relation of PETs and functional diversity was also different. PETs partly overlapped with PRTs. This study has shown that plant functional traits link environmental change with vegetation response. Based on plant functional traits and plant functional types, a larger-scale change in ecosystem functions and response to different kinds of land use or/and management regime can also be predicted. |
PDF Full Text Request