Date of Graduation

5-2018

Document Type

Thesis

Degree Name

Master of Science in Crop, Soil & Environmental Sciences (MS)

Degree Level

Graduate

Department

Crop, Soil & Environmental Sciences

Advisor/Mentor

J. Thad Scott

Committee Member

Kristofor R. Brye

Second Committee Member

William R. Green

Third Committee Member

Lisa S. Wood

Keywords

Chlorophyll-a, Modeling, Phosphorus-loading, Reservoir, Secchi, Vollenweider

Abstract

Reservoir limnology theory predicts that phytoplankton biomass (PB) is greatest in riverine-transition zones and least in lacustrine zones leading to an inverse pattern in water clarity. These theoretical patterns were utilized to create a statistical model of chlorophyll-a (Chl-a), an indicator of PB, and Secchi transparency (ST), an indicator of Chl-a, in Beaver Lake, Arkansas, a 12,800-ha reservoir, in order to hindcast historical conditions. Sampling for Chl-a, ST, and photic depth occurred semimonthly at 12 locations along a 78-km transect from the river inflow to the dam during the 2015 growing season. The ratio of Chl-a and ST measured at each site to the Chl-a and ST measured at the dam (FracDAMChl-a and FracDAMST, respectively) were computed for each sampling date, and regression models were developed to predict FracDAMChl-a and FracDAMST as a function of distance from reservoir inflow. The models were used to estimate Chl-a (r2 = 0.83, p = 0.0003) and ST (r² = 0.98, p < 0.0001) at any location in the lake in years where spatially explicit data were not available. United States Geological Survey (USGS) monitoring data collected at the dam, and three additionally overlapping sites, from 2001 to 2015 were used to develop and test the hindcast models. Residuals of the modeled-measured USGS data suggested that variation in hydrology across years created predictable interannual variation in the spatial patterns in Chl-a and ST across the riverine-lacustrine continuum. Whole-lake averages of Chl-a and ST were related to whole-lake total phosphorus (TP) for modeled, measured, and target data sets between 2001-2015 for the purposes of estimating Vollenweider P loads. The most important finding of this study revealed that average of modeled and measured P loading required reductions of 18.9% and 33.3% to meet the newly adopted Beaver Lake Chl-a and ST standards, respectively.

Share

COinS