Pilot experiment

Can niche plasticity promote biodiversity–productivity relationships through increased complementarity?

Pascal A. Niklaus published a paper from the results of pilot experiment in Ecology, showing that richness-productivity relationships are promoted by interspecific
niche complementarity at early stages of stand development, and that this effect is enhanced by architectural plasticity.

In the first stage of the BEF-China project, the BEF-China consortium agreed on an additional short-term Pilot Experiment. This small-scale tree biodiversity experiment was designed in order to study early growth patterns and species interactions of young trees. The establishment of the Pilot Experiment gave the opportunity to investigate mechanisms of competition between species with varied functional traits and under different environmental treatments. This knowledge can be considered and used as baseline data for the interpretation of results gained in the Main Experiment.

The Pilot Experiment has been established on a former agricultural field near the experimental sites in Xingangshan on a gross area of 7900 m2. A total of 1598 plots (size of 1 m × 1 m) were established. Except for certain density treatments, each plot was planted with 16 tree individuals, reflecting a diversity gradient from 1, 2 or 4 species. 21 tree species were chosen according to the species composition in nearby Gutianshan National Nature Reserve. Tree species were assigned to 8, partly overlapping, species pools containing 4 species each. The species assignments were chosen to reflect certain characteristics in natural forests, i.e. early-successional, commercial versus native and late-successional species. In addition, three species pools were comprised with species collected by seed families in order to control for genetic diversity.

Experimental design

The basic design comprised pair-wise competition diallels and four-species mixtures within each species pool (Figure 1). In each pool, there were four monocultures, the six possible pair-wise mixtures and the four-species mixture, resulting in eleven communities. All experimental plots contained 16 individuals in an array of 4 x 4, and in mixed communities each species was represented by the same number of individuals at the edge and in the center (Figure 1). All treatment combinations were replicated four times, once in each of four blocks. The treatment combinations were randomly assigned to plots within blocks. This basic design was modified regarding the particular requirements of the different subprojects (see below). For example, the treatment using different soil nitrogen tracers to assess belowground complementarity used only monocultures and four-species mixtures yet a greater number of replicates to allow for sequential harvests.


Figure 1: Eleven different species compositions of one of the six pools in the Pilot Experiment. All possible monocultures and 2- and 4-species mixtures of the four species in the pool (here species A, B, C, D) were established.

Each experimental community consisting of 16 individuals was densely planted on an area of 1 m2 (= 1 plot). Planting distances between neighbors were 25 cm, except for additional density treatments of subproject 2 where neighbor distances were 15 cm and 25 cm (plots with 16 individuals) and 50 cm (= plots with only one individual). The dense planting at a distance of 25 cm was intended to induce competitive interactions already in the early stages of tree growth (i.e. within months to a few years). In the first block, individuals within the community were planted regularly (Figure 1), while in the three additional blocks the individuals were randomly distributed, randomizing first the four individuals of the inner square (= center), and second the twelve individuals in the outer square (= edge). Each random distribution was repeated within the block for the different treatments and for the different diversity levels.

The Pilot Experiment was established in March 2009 and finished in July 2011, depending on the subprojects involved, has been studied for 1, 2 or 3 growing seasons. The treatments included light vs. shade (Figure 2); shallow vs. deelp nitrogen labeling; high vs. low genetic diversity; density low vs. hight; fungicide vs. mycorrhizal inoculation; pesticide vs. control (with herbivory), etc.

Screen Shot 2015-08-14 at 12.12.37 PM.png

Figure 2 Light vs. shade plants in the plot.