Determining Habitat Constraints on Invasive Species in a Tidal Freshwater Ecosystem using Remote Sensing Data and Hydrodynamic Model Outputs
Abstract
The Sacramento - San Joaquin Delta in California is one of the most invaded estuaries in the world. It is a tidal ecosystem kept fresh throughout the year in order to transport drinking and irrigation water to central and southern California. Both floating and submerged invasive aquatic species commonly found in lakes, reservoirs and river floodplains have made their way to this ecosystem before 1950. Today they occupy at least a third of the waterways by area in the Delta. Delta waterways can be categorized into three kinds of habitat, 1) channels: shallower at the edges with slower water velocities and of varying depth in the center with faster velocities, 2) flooded islands: remnants of old subsided islands now permanently flooded with broken levees surrounding them and restricting tidal action creating lake-like environments, and 3) slow shallows: shallow regions between remnant marsh islands that are adjacent to channels but have slower water velocities.
Spectroscopic airborne data was acquired over the Delta in 2015 and mapped to community level for submerged aquatic vegetation, and genus level for floating aquatic vegetation, using multiple data reduction and enhancement techniques and the random forest classifier. To determine the kind of habitat where submerged and floating invasives are likely to colonize, we categorized the entire network of waterways in the Delta into three habitat types: channel, flooded island, and slow shallows. Cross-sections were generated at fixed distances along the length of the channels and overlaid on the class maps produced from the spectroscopy data. Patterns of invasion in the 3 different habitats for both floating and submerged vegetation were quantified. We found that floating and submerged communities showed distinctly different patterns of colonization along the length and width of the waterways in the 3 habitats. Finally, we used hydrodynamic model outputs characterizing depth, salinity, and velocity in terms of means, standard deviation, maximas and minimas to understand the physical constraints that were influencing the observed pattern of colonization.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.B53A..04K
- Keywords:
-
- 0414 Biogeochemical cycles;
- processes;
- and modeling;
- BIOGEOSCIENCES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1622 Earth system modeling;
- GLOBAL CHANGE