PREDICTING COASTAL DEAD ZONE SIZE: A STATISTICAL APPROACH USING GIS

The number of coastal dead zones has grown exponentially since the 1960s, with large environmental and economic consequences. Dead zones occur in aquatic ecosystems experiencing hypoxic (low oxygen) or anoxic (no oxygen) conditions and are caused by nutrient over-enrichment. Watershed management techniques are sought to decrease dead zone sizes in order to mitigate or reverse environmental degradation. Dead zone (DZ) size is thought to be a function of natural and anthropogenic variables and may be predicted by land cover patterns and population characteristics. In this study, a spatial and statistical analysis of 56 dead zones in 15 watersheds within the United States was conducted to determine what variables best predict dead zone area. The 2006 National Land Cover Database was used to determine natural and anthropogenic characteristics for each watershed. Population density was found with 2010 U.S. Census data.

Dead zones are found in virtually every coastal watershed in the contiguous United States. DZ areas in the U.S. range from 0.1 km² to 22,000 km², with a mean and median of 1900 km² and 300 km². Stepwise regression analysis shows a positive relationship between agriculture and DZ area. For dead zones less than 800 km², population density is positively correlated with DZ area. No statistical relationships were found between natural land cover variables and DZ area.

These results highlight the relationship between nutrient over-enrichment from human activities and the occurrence of dead zones. Proposed management techniques to reduce the influx of nutrients into coastal ecosystems include modifying agricultural practices and constructing forests and wetlands to act as nutrient sinks. Based on this study, modifying agricultural practices may be more effective than constructing forests or wetlands in reducing dead zone sizes.