Application of a dynamic 2D hydraulic model to predict riparian wetland community transitions

Abstract

In the midwestern United States, large scale changes in river channel hydraulics and floodplain inundation dynamics have resulted from regional agricultural expansion and increased precipitation intensity. Riparian wetlands, i.e. wetlands located within a river's floodplain and hydrologically connected to stream or river channels, are directly exposed to changing inundation conditions due to their location at the dynamic terrestrial-aquatic interface. However, the effects of large scale, dynamic hydraulic changes on local wetland community loss, gain and presence are poorly understood, in part due to difficulties quantifying local inundation conditions. In this research, a dynamic 2D hydraulic model was developed for Locust Creek, a small river in the Midwestern U.S., and simulated continuous inundation depth and duration at a 30 m × 30 m spatial resolution and one-minute time step over a 10-year period. Local maximum values of depth and duration were used to predict wetland community loss, gain and presence. Results were compared to observed riparian wetland vegetation community transitions from two observational surveys. For both bottomland hardwood forest and wet prairie, two wetland communities of importance in this region, local losses were well predicted but local gains were not, indicating inundation is a first order control on loss but other factors must be considered to identify locations of naturally occurring gains. Quantification of hydraulic stressors via easily measured or modeled metrics enhances resource managers' ability to adapt to the changing hydraulic conditions by informing monitoring, sampling, and rehabilitation activities needed to manage wetlands.