Reservoir eutrophication dynamics in semi-arid regions : assessing the role of excessive nutrient loading and temporal changes in climate forcing -

Abstract

Eutrophication of freshwater bodies is a major concern that has been hampering water management efforts globally over the last half century. Eutrophication defines the deterioration of water quality due to excessive nutrient inputs, leading to high algae concentrations and increased presence of cyanobacterial species. While eutrophication processes in temperate zone lakes and reservoirs are well studied, eutrophication dynamics in semi-arid water bodies remains poorly understood, largely due to the lack of data. Remote sensing has the potential to assess changes in eutrophication status over space and time in these infrequently monitored systems. One of the major limitations of remote sensing of inland water bodies is their optical complexity, which requires the development of regionalized empirical models. In this study, empirical Landsat-based algorithms were developed for Qaraoun Reservoir in Lebanon in an effort to measure chlorophyll-a, total suspended solids, and secchi disk depth from spectral reflectances. Developed empirical algorithms were then used to hindcast changes in water eutrophication status, water temperature and reservoir volume across the Landsat image record (1984-2015). Results showed that the reservoir has historically existed in a eutrophic state, but that eutrophication has significantly increased in the most recent decade. Furthermore, the results showed a statistically significant increasing trend in August surface water temperatures, with a net increase of 1.2 oC over the past three decades. Field physio-chemical and biological data collected between July 2013 and October 2015 were also used to understand the driving forces responsible for the marked increase in cyanobacteria biovolume in the reservoir. A Structural Equation Model (SEM) was developed to link cyanobacterial biovolume to the physio-chemical conditions in Qaraoun. The model showed that, given its hypereutrophic status, the cyanobacterial response to increased nutrient concentrations was comparatively lower than the resp