Local and landscape influences on turbidity in urban streams: a global approach using citizen scientists

Miguel-Chinchilla, L, Heasley, E, Loiselle, S and Thornhill, I (2019) 'Local and landscape influences on turbidity in urban streams: a global approach using citizen scientists.' Freshwater Science, 38 (2). pp. 303-320. ISSN 2161-9549

[img]
Preview
Text
11994.pdf - Published Version
Repository Terms Apply.

Download (1MB) | Preview
Official URL: https://doi.org/10.1086/703460

Abstract

The ecological degradation of urban rivers and streams has been termed the ‘urban stream syndrome’ and attributed to increased catchment urbanization. Limiting future degradation requires an understanding of the drivers of reduced water quality at both catchment and site scales. The goal of this study was to identify the probable drivers of turbidity in river ecosystems in highly urbanized areas, under the premise that turbidity does not respond consistently to urbanization. Catchment-scale data were compiled from remotely-sensed datasets, whereas local-scale data were collected by citizen scientists as part of the global FreshWater Watch program. The local-scale data included nearly 2600 coincident measurements of turbidity and observations of other local characteristics taken with a common method between March 2013 and June 2016 across 127 unique locations in 6 major population centers: Vancouver (Canada), Sao Paulo (Brazil), Curitiba (Brazil), Buenos Aires (Argentina), Hong Kong SAR (China), and Guangzhou-Foshan (China). Catchment- and site-scale information were modeled with Boosted Regression Trees (BRT) to identify likely drivers of increased turbidity both across the entire dataset and within individual cities. Urbanization was not consistently associated with turbidity. The global BRT model explained 60% of the variation in turbidity, and key predictors were catchment area, % of the catchment as grassland, rainfall, Gross Domestic Product and % of the catchment as artificial surfaces. City-specific BRT models explained 35–67% of the variation in turbidity. Key predictors varied between cities and were often different than those observed at the global scale. Local-scale data collected by citizen scientists were less predictive of turbidity than catchment-scale factors and explained ~12% of the observed global variability in turbidity. Factors such as riverbank vegetation and the presence of point pollution sources explained some of the variation in turbidity, indicating their management could help mitigate elevated turbidity and sediment load in some urban rivers. Through this high-resolution, site-scale information, we highlight how community-sourced data may add value to freshwater monitoring programs.

Item Type: Article
Keywords: NTU, turbidity, land use, catchment, water quality, urbanization, urban stream syndrome, boosted regression trees
Subjects: G Geography. Anthropology. Recreation > GE Environmental Sciences
Q Science > QH Natural history
Divisions: School of Sciences
Identification Number: https://doi.org/10.1086/703460
Date Deposited: 11 Mar 2019 15:36
Last Modified: 15 Aug 2021 09:52
URI / Page ID: https://researchspace.bathspa.ac.uk/id/eprint/11994
Request a change to this item or report an issue Request a change to this item or report an issue
Update item (repository staff only) Update item (repository staff only)