Premium
Cryptic wetlands: integrating hidden wetlands in regression models of the export of dissolved organic carbon from forested landscapes
Author(s) -
Creed I. F.,
Sanford S. E.,
Beall F. D.,
Molot L. A.,
Dillon P. J.
Publication year - 2003
Publication title -
hydrological processes
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.222
H-Index - 161
eISSN - 1099-1085
pISSN - 0885-6087
DOI - 10.1002/hyp.1357
Subject(s) - wetland , dissolved organic carbon , watershed , hydrology (agriculture) , environmental science , bedrock , ecosystem , ecology , geology , geomorphology , biology , geotechnical engineering , machine learning , computer science
This study examines the relationship between wetlands hidden beneath the forest canopy (‘cryptic wetlands’) and dissolved organic carbon (DOC) export to streams and lakes in forested ecosystems. In the Turkey Lakes Watershed (TLW), located in the Algoma Highlands of central Ontario, Canada, there is substantial natural variation in average annual DOC export (kgC ha −1 year −1 ), ranging from 11·4 to 31·5 kgC ha −1 year −1 in catchments with no apparent wetlands. We hypothesized that the natural variation in DOC export was related to cryptic wetlands. Cryptic wetlands were derived manually from geographic coordinates that were surveyed with a differential global positioning system, and automatically from identification of topographic depressions and flat slopes (<1·5° ) within a digital elevation model (DEM) in a geographic information system. For the TLW catchments, which are characterized by shallow soils over bedrock, a significant correlation ( r 2 ≥ 0·9, p < 0·001) between manual and automated methods was observed for scales up to 50 m when a light detection and ranging DEM was used for the topographic analysis. Regression models indicated that cryptic wetlands (%) explained the majority of the natural variation in DOC export (kgC ha −1 year −1 ), with r 2 = 0·88 ( p < 0·001) for the model based on the manually derived wetlands and r 2 = 0·85 ( p < 0·001) for the model based on the automatically derived wetlands. The strength and significance of the automatically derived wetlands (%) versus DOC export (kgC ha −1 year −1 ) regression model diminished when other sources of DEMs were used. This study emphasizes the importance of including cryptic wetlands in predictive models of DOC export, particularly in catchments where the topography includes depressions and flat areas but no apparent wetlands. Copyright © 2003 John Wiley & Sons, Ltd.