Changes in the functional consequences of biodiversity loss along a latitudinal gradient

Background/Question/Methods

Most studies evaluating the effects of biodiversity loss have involved experiments conducted at a single location. Here, we evaluate how changes in producer diversity affect producer biomass along a latitudinal (spanning ~2.5° of latitude) and thermal  (a 5°C range in mean air and water temperatures) gradient to quantify how environmental context mediates the functional consequences of biodiversity loss.  Initial observations at our 9 sites – 3 sites in each of 3 regions spanning 500 km of the Gulf of Maine coastline from Boston, USA, to the Canadian border – suggested that whereas seaweed cover (a good surrogate for biomass) increased with seaweed diversity, this pattern was driven primarily by the more northern sites in Maine. We evaluated whether diversity was actually influencing seaweed cover by removing seaweed species from plots to create 5 experimental treatments: monocultures of each of the 3 most abundant seaweed species, polycultures containing all three of those species, and un-manipulated controls. Experiments were established during the summer of 2010 and have been maintained as press manipulations since then.

Results/Conclusions

After one year of experimental species removals we observed overyielding at sites in Maine, with strongest effects in the middle of the latitudinal gradient. At those Maine locations, diverse polyculture assemblages were characterized by higher seaweed cover than even the best-performing single-species treatments, despite the removal of an equivalent amount of algal biomass from polyculture plots. However, we found no evidence for overyielding at our 3 sites in Massachusetts. A key difference between our Massachusetts sites and our sites in Maine is that Massachusetts locations experience much greater thermal stress during the summer. Low-tide intertidal air temperatures were consistently warmer at our southernmost sites, with high temperatures occasionally exceeding 40°C near Boston. Further north, temperatures were much more moderate, never exceeding 30°C at sites in the northernmost region. Thus, the consequences of biodiversity loss are context-dependent. Our data suggest that the effect of diversity in enhancing seaweed cover may be unimodally related to temperature stress, with strongest effects in areas where stresses are severe enough that interspecific facilitation enhances producer biomass but not so severe that they limit the effects of diversity.