Spatial averaging sustains increased productivity at high ecosystem connectivity

Background/Question/Methods

Different rates of dispersal have distinctive impacts on species richness and productivity which are important aspects of ecosystem form and functioning. Studies have shown that these relationships are typically non-monotonic, with maxima at intermediate dispersal rates. The decrease of productivity at high dispersal rates is mainly attributed to the decrease of species richness, being however moderated by the effect of spatial averaging. So far, experimental studies on the relationship between productivity and dispersal have not addressed this effect. Here, we experimentally tested spatial averaging effects on phytoplankton assemblages using a gradient of environmental conditions: Grazer-free water from the Aegean Sea was inoculated into three treatments that differed in the frequency of nutrient pulses (every 1 day, 3 days, and 6 days) but not in the pulsing magnitude.  After a 3-month period of self-organization, the fragmented assemblages characterized by varied pulsing frequencies reached steady-state forming assemblages that differed in composition, structure and biomass. At the end of the self-organization process, assemblages across the three treatments were re-connected to metacommunities employing 3 levels of dispersal (low, intermediate, high). The initial isolated assemblages were used as controls, in order to compare the effect of dispersal on productivity.

Results/Conclusions

In agreement with theory and previous experimental studies, species richness in the metacommunity (i.e. the 3 treatments connected at a given dispersal rate) declined at high dispersal due to the exclusion of less competitive species. Productivity in the metacommunity was higher than the isolated assemblages (i.e. at zero dispersal) and remained high as dispersal increased. This is in contrast with theory that predicts a decrease in productivity at high dispersal rates. In our experiments, this high productivity is attributed to the dominance of a species that was best adapted to the average environmental conditions (i.e. the 3-day pulsed assemblages) in the initial isolated assemblages. During fragmentation, this species went nearly extinct from the 1 and 6 day pulsed assemblages probably because it was unfit to compete in isolation at these pulsing frequencies. Our results indicate that at high connectivity the effect of spatial averaging by a highly productive species, best adapted to intermediate environmental conditions, can mask the effect of decreased species richness, thereby maintaining high metacommunity productivity.