Ecological Informatics by Friedrich Recknagel & William K. Michener

Author:Friedrich Recknagel & William K. Michener
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

11.3.2 Modeling the Relationship Among Watershed Physiography, Land Use Patterns, and Phosphorus Loading

One of the emerging imperatives of eutrophication management is the advancement of our understanding of the relationships among land use, agricultural activities, hydrological processes, and water quality (Wellen et al. 2015). Prior to the watershed modeling exercise, Kim et al. (2016) implemented Self-Organizing Maps (SOM) to gain insights into the physiographical features and land-use patterns in the Bay of Quinte watershed, and to subsequently associate them with the phosphorus non-point source loading. In this application, eighteen classification variables were used, such as the landscape slope, saturated soil hydraulic conductivity, soil bulk density, and areal fractions for different land use types (lakes, ponds, alvars, bogs, coniferous swamps, deciduous swamps, fens, marshes, deciduous forests, coniferous forests, cutovers, mining areas, urban lands, pastures, and croplands) in 73 gauged and 137 ungauged subwatersheds. Thus, a total of 210 spatial units were distributed on 2-dimensional hexagonal maps, and then clustered in different groups according to their similarities.

Based on the spatial heterogeneity of these classification variables, SOM delineated six spatial clusters in the Bay of Quinte watershed with fairly distinct land-use patterns (Fig. 11.11). Coniferous and deciduous coverage along with pastures and croplands dominate the landscape in cluster 1. Different types of wetlands, such as fen (≈10%), coniferous swamp (≈8%), and alvar (≈0.4%) have also their highest areal fraction values in the same cluster. In cluster 2, the average landscape slope is steep and the soil bulk density is high. The areal fractions of forests as well as mining and logging sites are also high. In cluster 3, most of the subwatersheds are located in the vicinity of the Bay of Quinte, where crops occupy ≈75% of the area. Not surprisingly, the annual TP yield per area and average TP concentrations are the highest (528 kg km−2 year−1 and 103 μg L−1) in these same regions. In cluster 4, soil hydraulic conductivity is significantly higher, deciduous swamp are more abundant relative to the rest of the watershed, cropland coverage is the second highest (≈41%), and thus the net TP export is high. In cluster 5, urban land represents ≈74% of the land-use coverage and net TP export and yield are the second highest (3.72 tonnes year−1 and 209 kg km−2 year−1), which is further accentuated by the increased point source loading (2.44 tonnes year−1). In cluster 6, pasture and cropland approximately correspond to 60% of the area, and these subwatersheds are mainly located adjacent to the Bay of Quinte.

Fig. 11.11Map of Bay of Quinte watershed: (a) land use types, and (b) classification based on artificial neural networks and associated phosphorus export per subwatershed [Reproduced from Kim et al. (2016)]

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.