Tuesday, June 5, 2007
491

Evaluating the Efficacy of Water Quality Sampling Strategies: Blending Sensor Technology, Robotics, and Geospatial Analysis Reveals Important Reach-Scale (10-100m) Environmental Heterogeneity in Rivers

Eric Steimle1, Andrew Casper1, Barnali Dixon1, James Gore1, Robyn Conmy2, and Paula Coble2. (1) Environmental Science & Policy Program, University of South Florida St. Petersburg, 140 6th Ave. South, DAV 124, St. Petersburg, IL 33701, (2) College of Marine Science, University of South Florida Tampa, Tampa, FL 33701

In order to monitor, model, and manage the impacts various point and non-point source pollution or changes in watershed dynamics and land use, we must distinguish natural patterns of variation from anthropogenic. There can be substantial variability within a study area, thus a key consideration is choosing an appropriate spatial scale for sampling. Researchers are often limited to sampling designs developed for evaluating statistical measures of central tendency; those replicated at one site or, alternatively, unreplicated across multiple locations, then extrapolated to whole river reaches. This is also a problem in shallow waters that cannot be sampled from either boat or shore. Thus in addition to continuous data in a GIS format we also employed an inexpensive, portable surface ROV to compare/contrast the channel with near shore habitats where diversity, productivity, and environmental heterogeneity are greater. Specifically we collected and plotted continuous measures of fluorescence, CDOM, temperature, turbidity, and conductivity across an urban-forested transition in a watershed (the Hillsborough River, FL). These are displayed as a 2D kriged surfaces instead of the mean and variability of point samples. Point sampling and subsequent extrapolation of measures of means and standard deviations tended to mask the actual patchy spatial mosaic. 


Web Page: www.stpt.usf.edu/coas/espg/emat/default.asp