PS 24-32
The effect of dispersal along a stress gradient in micro-algae communities

Tuesday, August 11, 2015
Exhibit Hall, Baltimore Convention Center
Jonathan De Raedt, Applied Ecology and Environmental Biology Department, Ghent University, Gent, Belgium
Jan M. Baert, Applied Ecology and Environmental Biology Department, Ghent University, Gent, Belgium
Colin R. Janssen, Applied Ecology and Environmental Biology Department, Ghent University, Gent, Belgium
Frederik De Laender, Biology Department, University of Namur, Belgium

The fast decline of biodiversity resulted in a growing concern about the effects of biodiversity loss on ecosystem functioning. The majority of the biodiversity experiments have found a positive relationship between diversity and ecosystem stability and performance. However, most of these experiments have been performed in closed communities in which species composition is the result of random species extinctions. In reality, final community composition is the result of a species’ specific responses to dispersal, environmental conditions and species interactions. In order to evaluate the effects of dispersal and toxic stress (i.e. environmental conditions), 5 different communities of 4 marine diatom species (Bacillariophycaceae) were exposed to three levels of toxic stress (0, 25 and 250 ppb atrazine) and three levels of dispersal (no, low and high). Each treatment was replicated 3 times, resulting in 135 communities. Dispersal was performed by adding a fixed volume of 4 different species to the community once (low) or twice (high) a week from a species pool of 12 species.


Dispersal had a negative effect on biovolume in the communities (p<0.001). However, at high stress, there was a positive interaction effect between atrazine and dispersal on biovolume (p<0.001). This positive interaction effect was larger than the negative effect of dispersal. Hence, interactions between dispersal and the toxicant by far compensated for dispersal-induced biovolume loss. Dispersal had a negative effect on evenness in communities (p<0.001). However, the mechanism causing this negative effect was different for no and low than for high stress levels. At no and low stress levels, newly arriving species barely contributed to biovolume production. Indeed, community composition at the end of the experiment was dominated (average 94%) by species initially present in the community. Thus, newly arriving species were not able to colonize and grow, because of high biovolume of resident species and high competition. At high stress levels, this dominance of resident species decreased (p<0.001). Only species which were tolerant for the toxicant were able to grow. We conclude that the effect of dispersal changed along a toxic stress gradient.