Affordable Access

Access to the full text

Development and spatial application of a submerged aquatic vegetation model for Cootes Paradise Marsh, Ontario, Canada

Authors
  • Tang, Rex W. K.1
  • Doka, Susan E.1
  • Midwood, Jonathan D.1
  • Gardner Costa, Jesse M.1
  • 1 Central and Arctic Region, Fisheries and Oceans Canada, 867 Lakeshore Road, Burlington, ON, L7S 1A1, Canada , Burlington (Canada)
Type
Published Article
Journal
Aquatic Sciences
Publisher
Springer International Publishing
Publication Date
Nov 06, 2020
Volume
83
Issue
1
Identifiers
DOI: 10.1007/s00027-020-00760-w
Source
Springer Nature
Keywords
License
Yellow

Abstract

Cootes Paradise Marsh (CP) is an urban wetland and is part of the Hamilton Harbour Area of Concern (AOC). Anthropogenic stressors have degraded the system’s water quality. Submerged aquatic vegetation (SAV) provides critical fish habitat, and its recovery is crucial to this AOC’s delisting efforts. We developed predictive models to recommend water clarity (Secchi depth) targets that can potentially achieve a minimum SAV presence of 230 ha in CP, using macrophyte monitoring data that have been collected since 1996 by the Royal Botanical Gardens (RBG). A random forest approach was used for modelling SAV presence and SAV % cover. The final model for predicting presence of SAV consisted of Secchi depth, west wind fetch, and water level; the model had high accuracy (accuracy = 0.88, kappa = 0.77). For predicting SAV cover, the final model consisted of water depth, Secchi depth, percent slope, average fetch, water level, and substrate type; it had moderate accuracy (σ2explained = 0.66, root mean square error = 26.09, and weighted absolute percentage error = 58.96). Both models were then applied spatially using a digital elevation model to predict areas of CP where SAV would likely occur under different water level and water clarity scenarios. We recommend a delisting Secchi depth target of greater than 0.75 m to achieve the maximum potential of SAV areal extent under different water level scenarios.

Report this publication

Statistics

Seen <100 times