Prioritization of beaver (Castor canadensis) reintroduction sites within semi-arid grassland rivers in the Great Plains
Date
2020
Authors
Kornse, Zachary, author
Wohl, Ellen, advisor
McGrath, Daniel, committee member
Morrison, Ryan, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
River restoration has become more of a concern with human influence on natural systems on the rise. Beaver provide a relatively inexpensive and natural opportunity to restore rivers to a pre-settlement state. Quantitative models can be used to better understand where beaver reintroduction should occur to maximize the odds of a reintroduced beaver population establishing an ideal habitat to thrive in. The Beaver Restoration Assessment Tool (BRAT) is a quantitative model that uses elevation, hydrology, and vegetation inputs to estimate the carrying capacity of beaver. The primary objective of this project is to develop baseline data that can inform river restoration of the Dale Creek watershed on Old Elk Ranch, a northern tributary to the Poudre River in Northern Colorado, although the methods used in this study are broadly applicable to other watersheds. This objective was addressed during two phases of work – the channel and riparian condition assessment during summer 2018 and the assessment of historic and contemporary potential beaver habitat using BRAT and field surveys during summer 2019. Through the utilization of BRAT, survey mapping, and remote sensing, I was further able to determine the best methods for estimating potential beaver population density, as well as the accuracy of the BRAT results relative to the field survey results. Remote sensing provides a unique opportunity to increase the accuracy of BRAT through image classification and analysis. A national vegetation dataset produced by LANDFIRE, with a spatial resolution of 10m, is not as spatially accurate as vegetation datasets derived though supervised classification of NAIP imagery with a 1m spatial resolution. Due to its coarser spatial resolution and being a product of generalized models and field data, LANDFIRE data missed important details in vegetation, such as riparian willows along valley bottoms and variabilities within the floodplain that are crucial for beaver survival. Furthermore, this increase in vegetation accuracy led to increasing the accuracy of BRAT predictions for beaver carrying capacity relative to ground-based mapping of past beaver occupation, allowing for better assessment of where beaver should be reintroduced. Historical BRAT estimates revealed that zero and 1st order channels had the highest carrying capacity of dams historically, accounting for 56% of reaches capable of the highest BRAT classification (15+ dams/km). According to the field survey and historical vs contemporary BRAT estimates, carrying capacity has been greatly reduced post human settlement, primarily in zero and 1st order channel valleys where cattle were introduced. The 2018 condition assessment confirmed that while riparian vegetation was healthiest in higher-order channels, zero and 1st order channels were heavily browsed by cattle, deer, and elk. Based on the distribution of riparian vegetation health, and carrying capacity estimates from BRAT and field surveys, stream restoration would be most effective on 0th and 1st order streams.
Description
Rights Access
Subject
beaver restoration assessment tool
image analysis
beaver reintroduction
river restoration
BRAT