Modeling masses of midges: Following the path of aquatic insects from a lake onto land

Background/Question/Methods

Mývatn (“lake of midges”) is a large, eutrophic lake in northern Iceland from which tens of tons of midges (Diptera: Chironomidae) emerge annually and move over land to mate, in the process transporting large amounts of nitrogen that stimulates terrestrial primary production. However, determining the mass and spatial extent of this aquatic insect deposition on the terrestrial landscape surrounding the lake is difficult.  We developed a model to describe midge emergence from and deposition around Mývatn and to quantify the expected rates of nitrogen flux into the lakeside terrestrial ecosystem.  For four consecutive summers (2008-2011) we measured aquatic insect emergence from the lake and relative densities of the two dominant species of midge at various locations around and away from the lake edge.  From these data we generated spatial models of varying complexity, incorporating factors such as wind direction and fetch, that predict midge deposition over tens of square kilometers adjacent to Mývatn. 

Results/Conclusions

Our models predict total annual midge emergence from the lake to range from 30 – 50 Mg dry mass per year.  Deposition patterns follow a distance-decay pattern in which deposition peaks at 40m from the lake, and trails off to near zero by 400m for Tanytarsus species and 700m for the much larger Chironimus species. When spatial patterns of midge distribution and wind direction are taken into account, terrestrial midge deposition in high midge years such as 2009 may exceed 2 kg m^-2 yr^-1 in areas 20-50m from the lake in which wind direction is predominantly onshore. This delivers approximately 180g m^-2 N yr^-1 (1800 kg N ha^-1 yr^-1) over the same area.  Average deposition rates in high midge years range between 1.0-1.4 kg m^-2 yr^-1 at 20-50m. All of our models predict terrestrial midge deposition to represent significant inputs of N over large areas.  The effects of this transfer include significant shifts in vegetation and arthropod assemblages over spatial midge gradients. Our unique approach to sampling and modeling could be used in a wide array of aquatic systems to characterize the magnitude of previously unappreciated aquatic material fluxes to land.