Filtering processes in the assembly of plant communities: Are species presence and abundance driven by the same traits?

Question: Is the response of plant traits to environment at the community level similar when considering species abundance and when considering species presence only? Location: Mountain grasslands, central Argentina. Methods: We used data from 57 floristic samples, ordinated through DCCA along moisture and grazing gradients combined with trait values from 85 species (plant height, leaf area, leaf thickness leaf toughness and SLA). For each sample, we calculated the weighted average (considering species abundance) and the simple average (considering only species presence). Through multiple regressions we analysed how each average (dependent variable) responded to moisture and grazing (DCCA scores along Axes 1 and 2, respectively, as independent variables). Results: Weighted averages of all traits were significantly associated to both gradients, while simple averages did not always respond. In some cases the responses followed similar but weaker trends than the responses of weighted averages, but in other cases these responses were qualitatively different. Traits more associated with size (plant height, leaf area, leaf thickness) responded more consistently (similar trends for both averages) to grazing than to moisture, while traits more associated with plant resource acquisition (SLA, leaf toughness) responded more consistently to moisture than to grazing. Conclusion: The trait values and combinations which determine the probability of species presence are not necessary the same as those which determine their probability of becoming abundant. To understand community assembly rules, both species presence and species abundance should be taken into account as the result of different, although closely linked, filtering processes.