Effects of coastal geographical characteristics on the abundance of submerged aquatic vegetation

Submerged aquatic vegetation (SAV) consists of a taxonomically diverse group of plants that lives entirely beneath the water surface. SAV provides habitat and supplies food for aquatic life, absorbs excess nutrients, and helps purify the water. Since the 1960s, SAV coverage has experienced dramatic decline worldwide due to serious deterioration of water quality in coastal ecosystems. In recent years, landscape analyses have been used to predict direct or indirect effects of geographic characteristics on aquatic organisms. In this study, we analyzed the effects of coastal geographical characteristics (such as watershed land use types, sub-estuary mouth width, shoreline fractal dimension, ratio of watershed area/sub-estuary area, high resolution average wave height and estuary sediment, etc.) on the spatial variation of SAV abundance based on the long-term dataset (1984--2009) from 99 sub-estuaries and their linked coastal sub-watersheds of the Chesapeake Bay in the United States. Coastal watershed land use is an important factor influencing inputs of nutrient and sediment to its associated sub-estuary, and thus indirectly affects SAV abundance. Our results showed that sub-estuaries with different dominant land covers in each salinity regime had significantly different variation in SAV abundance. Abundance of different SAV species varies spatially among different salinity regimes in shallow water area. In each salinity regime, SAV abundance in sub-estuaries declined with their linked coastal dominant watershed land use type in the following order: mixed-land > developed land or forested land > agricultural land in fresh water (TF) regime; mixed-land or agricultural land > developed land > forested land in oligohaline (OH) regime; forested land or mixed-land > developed land or agricultural land in mesohaline (MH) regime; mixed-land or forested land > developed land > agricultural land in polyhaline (PH) regime. Sub-estuaries with different salinity regimes for each dominant land cover also had significantly different levels of SAV abundance. In each dominant watershed land use type, SAV abundance in sub-estuaries declined with salinity regime in the following order: no significant difference among four salinity regimes for developed land; OH > MH > TF or PH for agricultural land; OH or PH > TF or MH for mixed-land; MH or PH > TF or OH for forested land. Relative weaker linear correlation was found between a single geographical characteristics and SAV abundan