Testing the stress gradient hypothesis at the community level: Effects of shrub facilitation accross a dune stress gradient

Background/Question/Methods

Many studies have found that facilitation between plants increases with environmental stress, in support of the stress-gradient hypothesis (SGH). While it has been hypothesized that increased facilitation at the individual level will have important consequences for the diversity, abundance and invasion of plant communities across environmental gradients, this has rarely been tested. We evaluated these hypotheses among native and exotic plant species in a coastal dune in California. Specifically, we compared plant species composition, richness, individual abundance and total cover in communities growing under a native shrub, Ericameria ericoides, with those in adjacent shrub-free areas across a natural 220 m gradient of soil texture and wind. We performed a rarefaction analysis to determine whether shrub patches contributed to overall species richness at this site and conducted a shrub-removal experiment and a transplant experiment with a common exotic grass, Bromus diandrus, to test the mechanisms by which shrubs affect other plants in this system.

Results/Conclusions

Plant species composition in patches under shrubs was distinct from patches outside of shrubs and rarefaction analysis showed that shrubs increased overall native species richness. The effect of shrubs on average plant size became more positive as stress increased along the gradient, while the effect on the number of individual plants became more negative. Shrub effects on plot richness became more negative across the gradient, but effects on rarefied richness showed the opposite pattern. Over two seasons, experimental removal of shrubs had only weak effects on aboveground biomass and species richness. In contrast, B. diandrus grew significantly larger in soil from underneath shrubs than in open dune soil, which shows that soil modification is an important mechanism of shrub facilitation in this system. Our finding that facilitation of individual plant size increased with stress is consistent with the SGH, but negative effects on abundance and plot richness are counter to our expectations. Drawing from this result we suggest that the effects of facilitation on community level properties such as abundance and species richness may depend on the life history of the plants involved and whether facilitation increases growth or survival.