Biodiversity effects on community functioning shift along a toxicity gradient
In the last two decades numerous studies have demonstrated that biodiversity affects community/ecosystem functioning by both enhancing overall system performance and temporal stability. Yet, so far, little attention has been paid to how environmental conditions can modulate these biodiversity effects on ecosystem functioning. The performance enhancing effect has classically been partitioned into two effects: the selection and the complementarity effect. These effects express to what extent community performance is related to either the selection for high performing species or the effects of niche complementarity and species interactions, respectively. Here we have investigated how the performance enhancing effect, and its two constituents, on community functioning (biomass production) shift along a gradient of toxic stress. We cultured North Sea diatoms (Bacillariophyceae) in microcosms in a full factorial design of 5 diversity levels (1, 2, 4, 6 and 8 species) and 3 atrazine concentrations (0, 25 and 250 ppb) for 4 weeks, after which the microcosms could recover in atrazine-free growth medium for 3 more weeks. Within each diversity level 10 random species combinations were drawn from the species pool, except for diversity levels 1 and 8 where only 8 and 1 combinations were possible, with 3 replicates at each concentration (=351 microcosms).
Toxic stress modulated the performance enhancing effect both during and after exposure. During exposure, community biomass was significantly affected by the number of species and by the atrazine concentration (p<0.001). Biodiversity effects became more important with increasing toxic stress. At the highest atrazine treatment, biodiversity significantly enhanced biomass yield (p<0.001). The importance of the selection and complementarity effects shifted along the toxicity gradient: for diversity levels of 4 species onward, the sampling effect increased with increasing stress (p<0.005) while the complementarity effect remained unaffected.
After 3 weeks of recovery, community biomass was only related to the number of species (p<0.05). However, both the sampling effect and the complementarity effect were clearly related to the community’s exposure history: from diversity levels of 4 species onward the sampling effect significantly increased (p<0.001) and the complementarity effect significantly decreased (p<0.001) with increasing past atrazine concentrations.
These results demonstrate that toxic stress alters biodiversity effects on ecosystem functioning. At higher stress levels the selection for more tolerant species appears to be an increasingly dominating driver for biodiversity effects and its influence remains even when exposure has ceased. This demonstrates that diversity is important under high-stress conditions because it allows for selection.