Species Diversity-functional Diversity Based On Disturbance Factors, Disturbance Intensity, And Functional Traits

At present, the global biodiversity is decreasing rapidly. The loss of biodiversity will no doubt cause significant impact on the function and service of ecosystem. Studying the relationship between species diversity (SD) and functional diversity (FD) plays a basic role in analyzing the relationship between biodiversity and the function of ecosystem. However, there still exist arguments about the relationship between species diversity and functional diversity. Some people think they are positive related. Some people think they are negative related, and others think their relationship conforms to a S-shaped curve. The functional traits are the bases for calculating functional diversity, as well as the bridge of combining the relationship of species diversity and functional diversity. Therefore, choosing of functional traits, such as the type and number of functional traits, is no doubt to have important effects on revealing the relationship between species diversity and functional diversity. However, this issue has long been neglected, and the effects of the change of the number of functional traits on species diversity and functional diversity remain to be verified.For this reason, a field manipulation experiment was conducted, on Kobresia humilis meadow from 2007 to 2013. In the experiment, a split-plot design was adopted. Three clipping level (stubbled 1 cm,3cm and unclipped) of clipping treatment were used on the whole plot and subplots were treated with or without fertilizing. We chose six plant functional traits (life cycle, plant inclination, growth form, main modes of reproduction, plant height and leaf dry weight per plant) which were sensitive to clipping and fertilizing treatment. The Rao index was used to measure functional diversity, and the species richness index and the Simpson index were used to measure species diversity. Single functional trait and six functional traits were used to calculate FD. At the same time, randomly chose the combination of one, two, three, four and five functional traits were used to calculate FD,. The FDi(i=1,2,3,4,5) index indicated that the functional diversity index were calculated based on different number of functional traits. The relationships of SD-FD were analyzed at different combination of clipping and fertilizing treatment. Besides, a correlation analysis were conducted on the relationship between the value of some particular single traits and their FD, aiming at providing a basis for choosing effective functional traits for predicting the change of the ecosystem function. The main results of the research are as follows:(1) The responses of SR and D with regard to clipping were different. The SR incresed after clipping treatment, and D did not vary obviously. Whereas, SR and D were decreased after fertilizing treatment. Considering the effects of clipping increased the SR only appeared in the fertilizing treatment, so the reason of difference between SR and D was that SR took only species richness into account, whereas D was a combined statistical quantity of species richness and evenness. The decrease of SR and D after fertilizing was due to that the competition for light between the species which were favor of fertilizer reduced the survival of those short species which were weak in the competition.(2) The decrease of FD6 after clipping showed that the functional traits of plants tend generally to be of the same variation trend. No matter in the fertilizing community or no fertilizing community, the FD6 of six functional traits did not vary obviously. Reason was that the Rao index had two compositions, which were called species evenness and trait dissimilarity among species. If the species evenness decreased while the trait dissimilarity among species increased, and the degree of the two compositions varied equally, the value of FD was constant.(3) From the results of variance analysis for the single functional trait, it can be concluded that the clipping treatment decreased FDLC?FDMR and FDLDW but did not significantly affect FDGF? FDPI?FDPH.Fertilizing treatment decreased FDLC and FDMR but increased FDPI. Besides, fertilizing did not affect any value of FD calculated by single functional trait. Therefore, FD6 decreased after clipping treatment. After fertilizing treatment, FD6 did not change significantly because of three different responses of FD which calculated by single functional trait.(4) The relationship of D-FD6 all presented positive linear correlation but SR-FD6 all presented no correlation. Clipping treatment only reduced the intercept of D-FD6 and it did not change the slope of the D-FD6; however, fertilizing treatment not only reduced the slope but also increased the intercept. Therefore, the effects of fertilizing on the D-FD6 relationship were more significant than clipping treatment. By comparison, the relationship between D (SR) and FD which calculated by single functional trait was more changeable. The reason for this was that the FD presented different responses to clipping and fertilizing treatment. Similarly, because of clipping treatment only changed the intercept of one functional trait's SR-FD, while fertilizing treatment changed the slope and intercept of three functional traits' SR-FD. In the same way, the effects of fertilizing on relationship of SR and FD which calculated by single functional trait were more significant than clipping treatment.(5) The number of functional trait had a significant impact on the R2 of the D-FD relationship. The R2 of the D-FD relationship would increase nearly two times when the number of functional traits increased one. The number of functional traits(x) and R2 conformed with exponential function:y=aebx, and it would not be affected by clipping and fertilizing treatment. When the R2 is a constant, the relationship of b and the number of functional trait (x) is a negative correlation. The function also indicates that the number of functional trait (x) and b presents a positive effect on R2 of D-FD relationship. Under the constraint conditions, the number of functional trait which makes the R2 reaches maximum value is the optimal number of functional trait for studying the SD-FD relationship. So, it will be of great significance for revealing the D-FD relationship in the future to establish the relationship between the number of traits and R2.(6) No matter under what kind of treatment combinations of clipping and fertilizing, for the same six functional traits, the relationship of D-FD presented a positive correlation, while SR-FD presented no relevant. The different pattern of D-FD relationship and SR-FD relationship indicated that experiment treatment and functional trait were not the main reason for this. The result mainly attributed to the two species diversity index of D and SR. As SR considers only the number of species and does not consider the abundance of species, it is not an ideal indicator for predicting the variation of FD.