Species diversity (SD) and functional diversity (FD) have previously been reported to be associated in a variety of ways, and identifying the relationship between these two parameters is important for our understanding the relationship between biodiversity and ecosystem functioning. Since ecosystem functioning is mainly influenced by plant functional attributes, it is widely held that functional diversity exerts a greater influence upon ecosystem functioning than does species diversity. However, it remains unclear which basic component of functional diversity (i.e., species evenness or trait dissimilarity among species) contributes more to functional diversity and thus determines or alters the SD-FD relationship. In this study, a field manipulation experiment was conducted in an alpine Kobresia humilis meadow at the Haibei Research Station of the Chinese Academy of Sciences from 2007 to 2012. The experiment employed a split-plot design, in which vegetation was clipped to one of two lengths (above ground stubble =1 cm or 3 cm) or left unclipped. Subplots were treated with fertilizer (urea 7.5 g m^-2 a^-1 + ammonium phosphate 1.8 g m^-2 a^-1) or remained unfertilized. Data analysis methods, e.g. general linear model univariate ANOVA, multivariate analysis, binary linear regression and curve fitting, were employed to analyze the effects of these treatments on trait dissimilarity among species (d_ij, 13 traits of 33 species), community species evenness (Pielou evenness index, J) and community species diversity (Shannon-Wiener index,H'), and functional diversity (Rao index, FD_Rao; Functional dispersion, FDis) (13 traits of 33 species), as well as potential relationships among these measures. 1) SD (H') was significantly positively correlated with FD (FD_Rao, FDis), and J was positively correlated with SD (H') and FD (FD_Rao, FDis), whereas d_ij (dissimilarity of 13 traits among 33 species) was negatively correlated with SD (H') and FD (FD_Rao, FDis). J was more strongly correlated with SD (H') than d_ij, and was also correlated with FD (FD_Rao, FDis). Thus, it can be concluded that species evenness is a major driving factor of both variation in FD and the positive correlation between SD and FD. 2) Curve fitting revealed that the increase in FD (FD_Rao, FDis) leveled off as SD (H') increased, indicating that a certain degree of functional redundancy occurred in the community and that this increased with species diversity. As a result, species diversity (H') failed to reach the limit that would result in a non-significant correlation between species diversity and functional diversity. 3) The effect of FD (FD_Rao, FDis) on aboveground net primary productivity (ANPP) of the alpine meadow ecosystem was greater than that of SD (H'). Binary linear regression showed that the effect of SD (H') on ANPP was negligible compared with that of FD (FD_Rao,FDis).It should be noted that the loss of functional traits or a reduction in the number of trait space dimensions, such as results from the decline or loss of species diversity, would inevitably reduce the functional diversity of community, indicating a complementary relationship between species diversity and functional diversity. Thus, future studies on the relationship between biodiversity and ecosystem functioning should focus simultaneously on the effects of both species diversity and functional diversity on ecosystem functioning.