OOS 51-6
Do community-mean trait values reflect optimal strategies? Insights from Puerto Rican forests

Wednesday, August 12, 2015: 3:20 PM
327, Baltimore Convention Center
Robert Muscarella, Section for Ecoinformatics and Biodiversity, Aarhus University, Aarhus, Denmark
Maria Uriarte, Ecology, Evolution, and Environmental Biology, Columbia University, New York, NY

Trait-environment relationships provide a physiological basis for understanding species distributions and community variation along abiotic gradients.  Theory predicts that community-weighted mean traits (CWMs) reflect optimal strategies under particular abiotic conditions (the CWM-optimality hypothesis).  However, the typically high amount of local trait variation among co-occurring species suggests niche partitioning at the local scale, highlighting the need to reconcile these two perspectives.  We tested the hypothesis that habitat suitability is negatively related to a species’ deviation from the local CWM for 173 tree species in Puerto Rico.  To do so, we measured three key tree functional traits (wood density, leaf mass per area, and maximum height) for each species and quantified CWM values for each trait in twelve 0.25-ha plots along a marked precipitation gradient (ca. 900-2,200 mm yr-1).   For each species in each plot, we calculated delta CWM as the absolute difference between the species-mean trait value and the local CWM.  We used a large database of species occurrence records to generate species-specific ecological niche models, and we used these to estimate habitat suitability for each species in each census plot.  We considered significant negative relationships between delta CWM and habitat suitability as support for the CWM-optimality hypothesis.


CWM values for all three traits varied in ways consistent with our expectations.  Specifically, CWM values of wood density and LMA decreased, and maximum height increased, with greater mean annual precipitation.  For a majority of species, we found significant negative relationships between estimated habitat suitability and delta CWM across plots, providing broad support for the CWM-optimality hypothesis.  Additionally, relationships between species traits and the conditions in their most suitable habitats mirrored the observed shifts in CWM values across the precipitation gradient.  We also identified, however, more species than randomly expected with positive relationships between habitat suitability and delta CWM.  This result reflects the success of alternative life history strategies that are key to maintaining local functional diversity.  Our study demonstrates that while community-level variation of key traits is strongly related species distributions along environmental gradients for a majority of species, successful alternative trait combinations are critical for the maintenance of local functional diversity.  Future work quantifying physiological and demographic variation along abiotic gradients will help to clarify the mechanisms underlying these patterns.