Comparative study plots

In 2009 and 2010, the joint Chinese-German-Swiss research project “BEF China” has established a large forest Biodiversity and Ecosystem Functioning (BEF) experiment in subtropical forests at Xingangshan (Jiangxi Province, China). In total, 566 plots were established at two sites, using different pools of a total of 42 native tree species and 10 shrub species with more than 300 000 planted saplings, covering about 50 ha. In a parallel observational approach, 27 Comparative Study Plots (CSPs) were set up in existing forests in an adjacent National Nature Reserve (Gutianshan, Zhejiang Province).


Figure 1: Location of the experimental sites with Comparative study plots showing in green area and main experimetal sites in red area.


Figure 2: Map of a CSPs, indicating plot usage and measurements for different subprojects (SP).

Experimental design:

Twenty-seven plots of 30 × 30 m area were deliberately selected to span factorial gradients in both tree species richness and successional age resulting from timber cutting by local communities. Average distance between plot pairs was ~3km. The closest pair was 40 m apart, followed by 165 m and 243 m for the next-closest pairs. For each plot, we determined tree species richness from the inventory data we recorded. Successional age was assigned to five age classes (<20, 20–40, 40–60, 60–80, or >80 years old) based on the age of the fifth-largest tree of each plot (determined from a stem core), because the precise date of the last logging event could generally not be determined. Our goal was to evenly cover the range in tree diversity and successional ages present at the site, although it was not possible to keep these two fixed, independent variables fully orthogonal to each other. In the further course of the study, two plots were lost due to (illegal) timber cutting. All analysis presented are therefore based on data from the remaining twenty-five plots.
We did not select plots randomly, because such a “sample survey” design would have resulted in a concentration of plots around mean tree species richness values, with a typically bell-shaped distribution. In sample surveys (and meta-analyses based on sample surveys), correlations between species richness and productivity are bi-directional relationships between two dependent variables. This problem can be alleviated by fixing one variable as independent variable at different levels that are similarly replicated, and then measuring the other variable as dependent variable. This approach is recommended e.g. in the classical statistical textbook by Snedecor & Cochran [15] who refer to this type of study as comparative study and rank it between sample surveys and designed experiments (with randomized treatments) with regard to the power to detect causal relationships between variables (Baruffol, et al. 2013).


Figure 3 Comparative study plots in Gutianshan natural nature reserve. Different colors represent different successional stages.

There are many papers showing the tree diversity effect on ecosystem function from this comparative study design, such as:

Bruelheide H, Bohnke M, Both S, et al. (2011). Community assembly during secondary forest succession in a chinese subtropical forest. Ecol Monogr 81:25-41.

Baruffol M, Schmid B, Bruelheide H, et al. (2013). Biodiversity promotes tree growth during succession in subtropical forest. Plos One 8.

Schuldt A, Baruffol M, Bohnke M, et al. (2010). Tree diversity promotes insect herbivory in subtropical forests of south-east china. J Ecol 98:917-926.

Schuldt A, Assmann T, Bruelheide H, et al. (2014). Functional and phylogenetic diversity of woody plants drive herbivory in a highly diverse forest. New Phytol 202:864-873.

Schuldt A, Bruelheide H, Durka W, et al. (2014). Tree diversity promotes functional dissimilarity and maintains functional richness despite species loss in predator assemblages. Oecologia 174:533-543.

Schuldt A, Bruelheide H, Härdtle W, et al. (2015). Early positive effects of tree species richness on herbivory in a large-scale forest biodiversity experiment influence tree growth. J Ecol 103:563-571.