We assessed change in the seasonal timing of insect emergence from tundra ponds near Barrow, Alaskaover a four-decade timespan, and explored factors that regulate this significant ecological phenomenon.The early-summer pulse of adult insects emerging from myriad tundra ponds on the Arctic Coastal Plainis an annual event historically coincident with resource demand by tundra-nesting avian consumers.Asymmetrical changes in the seasonal timing of prey availability and consumer needs may impact arcticbreedingshorebirds, eiders, and passerines. We have found evidence of change in the thermal behaviorof these arctic wetlands, along with a shift in the phenology of emerging pond insects. Relative to the1970s, tundra ponds at Barrow now thaw about one week earlier, average 2°C warmer, and show a 30%increase in growing degree days. Analyzing insect emergence data collected during 2009-2012 relativeto similar data from the 1970s, we document a significant shift in the timing of insect emergence overthis four-decade span. Composition of the invertebrate community has changed little over the decades,with the same species of chironomid midges dominating numbers and biomass in these ponds. Treatingmidge emergence as a ‘time-to-event phenomenon’, we used proportional hazard models to test fordifferences in the seasonal timing of adult insect emergence between our historic and recent data sets.For all species pooled, we found a significant increase in the daily likelihood of adult emergence during2009-2011 relative to the 1970s, reflecting a shift toward emergence at earlier dates in the recent years.Analyzing individual species, we detected significant shifts toward earlier seasonal emergence in recentyears for 14 out of 16 abundant taxa. We explored mechanisms potentially regulating emergence timingat the species level, toward development of models for predicting insect response to further climatechange in this landscape. Date of pond thaw each spring explains much of the variation in overallemergence timing, and individual midge species show fairly consistent positions in the temporalsequence within any pond or year. Examining pre-emergence development of midge populationsrelative to hourly pond temperatures, we found that models based on an exponential rate responsewere more suitable than traditional “thermal-time” models that assume a linear response ofdevelopmental rate to temperature. We identify key biological and environmental variables needed topredict how the seasonal timing of insect emergence may respond to further warming of this arctichabitat.
