||An important debate in ecology considers the relative importance of mean condition vs. variation around that mean for controlling ecosystem structure and function. Wetlands provide a useful setting for examining this question because hydrology acts as the primary exogenous ecosystem driver and may exhibit large variation or fluctuate only slightly around the mean. Most previous studies have compared wetland study sites which vary greatly in mean condition but also in water chemistry, morphology, timing of flooding and dispersal characteristics, limiting independent evaluation of variation. Floodplain wetlands along the Silver River in north Florida provide an ideal hydrologic setting for a controlled natural experiment; stage variation near the spring boil is small so hydrologic conditions deviate only slightly from the mean condition, whereas downstream conditions, which are controlled by the flood regime in the Ocklawaha are highly variable, deviating from the mean with great amplitude and frequency. Where water level variation changes longitudinally, hydroperiod varies laterally across the floodplain creating two orthogonal gradients, which offer the opportunity to isolate their effects independently.
Our objective is to evaluate the hypothesis that variation around the mean is as important as the mean condition in controlling ecosystem structure and function. This is accomplished by contrasting the effect of hydrologic mean condition and variation on attributes of long-term ecosystem self-organization, including net primary productivity, surface morphology/microtopography, organic matter accumulation, and forest community structure. We selected 10 study sites, pairing a short hydroperiod (10-30% inundation) and long hydroperiod (40-60% inundation) located at each of 5 transects along the Silver River gradient in water level variation. We are in the preliminary stages of analyzing our results, and seek to separate responses of ecosystem attributes to hydroperiod and water level variation.