Can traits be used to predict the susceptibility of aquatic insects to a low flow disturbance?

Predicting the response of ecological communities to disturbance is a major challenge in ecology.  Traits are increasingly advocated as a method for increasing predictive capabilities.  We explored whether aquatic insect traits could predict the susceptibility of aquatic insects to a low flow disturbance.  We lowered flow in three Connecticut streams by experimentally diverting water over a 100-m stream reach in the summers of 2005, 2006, and 2007.  We sampled aquatic insects under normal and low flow conditions, grouped insects according to trait characteristics following Poff et al. 2006, and analyzed density shifts in trait groups as a result of lowered flow.  We focused on traits that are not highly phylogenetically linked and found that lowered flow resulted in a decreased density of aquatic insects that are univoltine, commonly occur in drift, have a very low or low crawling rate, are small or medium in size, and prefer erosional habitats.  Traits may be useful for mechanistically linking the individual responses of stream organisms to lowered flow to stream community composition following a low flow disturbance.