^{1}

^{*}

^{2}

Relationships between biodiversity and ecosystem functioning are vital to understand ecosystem properties, and have frequently been addressed in small-scale studies. However, interactions and changes differ at large scales, and should be similarly evaluated to monitor biodiversity and ecosystem functional alterations. In the present study, Mainland China was divided into 241 quadrats of 2° latitude by 2° longitude. Ecosystem function was comprehensively assessed using three indicator variables in each quadrat, primary productivity, bird species richness, and relative humidity. Relationships between each ecosystem function variable were regressed against seed plant species richness. All three indicators exhibited the same change model, a linear model when plant species richness was under 5.000 species, and a hump-back/quadratic model when seed plant richness was over 5.000 species, with an increase in seed plant species richness at a larger scale.

Ecosystem function is the capacity for natural processes and components integral to those processes to provide goods and services, which ultimately meet human needs, either directly or indirectly (

A persistent debate continues regarding the relevance of studies that underlie the conclusions explaining biodiversity and functioning in mature natural ecosystems. Most experimental studies of constructed species assemblages determined that increasing diversity contributes to greater biomass production (

In the present study, we explored the relationship between plant ecosystem biodiversity, and three potential ecosystem function indicators (primary productivity, bird species richness, and annual mean relative humidity) at a large scale in Mainland China.

Mainland China was divided into 241 quadrats (including Hainan Island as one quadrat), each 2° latitude by 2° longitude (approximately 36 450 km^{2}). If a quadrat had an area less than half a standard quadrat on national boundaries, it would be merged into its neighboring quadrat (

where

where ^{2},

Regression analysis was chosen to assess the relationships between plant species richness and ecosystem function indicators, and analyzed using the SPSS statistical package (SPSS Inc., Chicago, Illinois, USA). AIC (Akaike’s Information Criteria) was applied to compare different regression models as follows:

Linear and quadratic models were compared in an analysis of the following relationships: ecosystem primary productivity, bird species richness, and relative humidity, each against plant species richness. Results indicated quadratic models were more robust (with smaller AIC) when all quadrat data were used, but linear models were better (with smaller AIC) when the three highest quadrats were excluded from the analysis (^{2} = 0.6671, P < 0.001, n = 241 - ^{-2} at 3500 species (^{2} = 0.2121, P < 0.001, n = 241 -

Results depicted a close relationship between relative humidity and the number of seed plant species in mainland China (R^{2} = 0.4727, P < 0.001, n = 241 -

Primary productivity is the fundamental indicator of ecosystem function, and is at the base of ecosystem function (

The linear and hump-back relationships have been supported by several empirical studies, but showed variability at the small scale. Indeed, positive and negative curve patterns of biodiversity and ecosystem functions have previously been reported (

Moreover, productivity is one indicator frequently assessed for ecosystem function in empirical studies. It is a viable parameter because productivity is the primary result of ecosystem function (

The study was financially supported by the National Natural Science Foundation of China (Grants No. 31170494 and No. 30870399).

Regression analysis between primary productivity of ecosystem and species richness of seed plant in the mainland of China. (A): quadratic model; (B): linear model.

Regression analysis between species richness of bird and species richness of seed plant in the mainland of China. (A): quadratic model; (B): linear model.

Regression analysis between annual mean relative humidity of ecosystem and species richness of seed plant in the mainland of China. (A): quadratic model; (B): linear model.

Comparison of regression results of linear and quadratic models in the analysis of relations between primary productivity of ecosystem, species richness of bird and relative humidity with species richness of seed plant with and without the three highest quadrats in the mainland of China. (x): refers to species richness of seed plant; (Y): refers to primary productivity of ecosystem, species richness of bird and relative humidity respectively; (AIC): 2k - 2 ln(L), where k is the number of parameters in the regression model, and L is the maximized value of the likelihood function for the regression model.

Variables | The 3 highest quadrats | Model | Regression equation | ^{2} |
P | AIC | F | P |
---|---|---|---|---|---|---|---|---|

Primary productivity | With | Linear model | 0.4995 | < 0.001 | 12.225 | 8.982 | 0.000 | |

Quadratic model | ^{2} + 0.9998 |
0.667 | < 0.001 | 12.065 | 10.437 | 0.000 | ||

Without | Linear model | 0.5515 | < 0.001 | 11.405 | 8.78 | 0.000 | ||

Quadratic model | ^{2} + 1.1049 |
0.669 | < 0.001 | 12.032 | 10.302 | 0.000 | ||

Species richness of bird | With | Linear model | 0.1911 | < 0.001 | 10.563 | 8.523 | 0.000 | |

Quadratic model | ^{2} + 0.0657 |
0.2121 | < 0.001 | 9.887 | 8.987 | 0.000 | ||

Without | Linear model | 0.1507 | < 0.001 | 9.538 | 8.736 | 0.000 | ||

Quadratic model | ^{2} + 0.0934 |
0.2046 | < 0.001 | 10.203 | 9.02 | 0.000 | ||

Relative humidity | With | Linear model | 0.3453 | < 0.001 | 8.683 | 9.868 | 0.000 | |

Quadratic model | ^{2} + 0.0198 |
0.4727 | < 0.001 | 8.657 | 7.986 | 0.000 | ||

Without | Linear model | 0.3819 | < 0.001 | 8.23 | 9.431 | 0.000 | ||

Quadratic model | ^{2} + 0.0219 |
0.4709 | < 0.001 | 8.509 | 7.505 | 0.000 |