Functional structure of an arid steppe plant community reveals similarities with Grime's C-S-R theory

Questions: (1) How do community-weighted mean (CWM) trait values of 23 functional traits measured on 34 plant species vary along a gradient of aridity under grazed and ungrazed conditions in an arid steppe? (2) How does variation in our CWM trait values differ from those of more mesic grasslands? Location: Eastern Morocco. Methods: We measured relative abundance and functional traits along a short aridity gradient over two consecutive years at five heavily grazed sites, each with an exclosure preventing grazing. We analysed the relationship between aridity, grazing, and the expression of CWM trait values using ordination methods and a fourth-corner analysis. Results: Unconstrained and constrained ordinations identified three distinct suites of temporally consistent functional traits that co-varied with aridity and grazing, and the fourth-corner analysis identified a number of significant but weak trait-environment associations. Grazing selected for short, fast-growing annual species with high SLA, high pastoral value and low seed mass, while aridity selected for species possessing succulent leaves with high δ 13C leaf content, spines, low LDMC and short stature, although the relative importance of precipitation and grazing changed between years. Conclusions: Although distinct from more mesic grasslands, our study sites exhibited patterns of trait correlations that were similar to the worldwide leaf economics spectrum. These correlation patterns represented three groups that were reminiscent of Grime's C-S-R model. Direct ordinations supported this interpretation. Temporal variation in our results was due in part to precipitation fluctuations. Our results also indicated selection for a grazing avoidance strategy under heavy grazing. Integrating plant functional traits in conservation and management of arid ecosystems represents a novel and challenging task to ensure more sustainable use of these lands. © 2011 International Association for Vegetation Science.