Cyanobacteria and associated harmful algal blooms cause significant social, economic, and environmental impacts. Cyanobacteria synthesize hepatotoxins, neurotoxins, and dermatotoxins, affecting the health of humans and other species. The Cyanobacteria Assessment Network (CyAN) aims to inform the public and researchers about cyanobacteria by processing remote sensing imagery to provide estimates of freshwater cyanobacteria concentrations. The MEdium Resolution Imagery Spectrometer (MERIS) CyAN raster product provides cyanobacteria index (CI) values for over 1,800 lakes in the continental U.S. We use spatial analysis techniques to assess CyAN data and examine the influence of stream network connectivity on cyanobacteria blooms. Time series analyses (r2, MAE, ARIMA) are used to compare weekly maximum cyanobacteria concentrations for each lake from 2008-2011. Bloom characteristics are evaluated for each lake in terms of timing, duration, magnitude, spatial autocorrelation, and rate of change. Semivariogram sensitivity is used to examine the similarity of bloom patterns in different lakes based on direct proximity (overland distance) versus flowline proximity (stream distance). Finally, we use StreamCat data and multivariate correlation to assess the relationship between watershed characteristics and blooms.
Results/Conclusions
We evaluated weekly maximum cyanobacteria concentrations provided by the CyAN MERIS product in 136 Florida lakes for the year 2010. Lakes within the study area vary in size from 1.5 km2 to >1,000 km2. The majority of lakes are moderate in size with 60% having surface areas <10km2 and only six with surface areas greater than 100 km2 (Lake Istokpoga, Lake Seminole, Lake Apopka, Lake Kissimmee, Lake George, and Lake Okeechobee). These larger lakes show a moderate positive correlation between lake size and maximum cyanobacteria concentration (r2= 0.58). Lakes <100 km2 have a weaker positive correlation between surface area and maximum bloom concentration (r2 = 0.17). In many lakes, seasonal trends showed more variable cyanobacteria concentrations and surface area extent during the warmer, wetter summer months. Initial spatial autocorrelation metrics indicate that for lakes with active cyanobacteria blooms, the blooms are highly clustered (global Moran’s I = 0.68, p-value ~0). CyAN data support research efforts and accelerate our scientific understanding of cyanobacteria bloom patterns in freshwater. These data provide opportunities to better predict bloom behavior to protect human health and the environment.