Premium
Spatio‐temporal patterns of nutrient concentrations and export in a north‐eastern European lowland catchment
Author(s) -
Mourad D. S. J.,
van der Perk M.
Publication year - 2009
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.7309
Subject(s) - streams , environmental science , hydrology (agriculture) , land cover , agricultural land , drainage basin , land use , nutrient , sampling (signal processing) , buffer zone , physical geography , ecology , geography , geology , biology , computer network , geotechnical engineering , cartography , filter (signal processing) , computer science , computer vision
In 2002–2004 we undertook six sampling campaigns during representative hydrological stages in a 901 km 2 Estonian lowland catchment to quantify the spatial and seasonal variability of in‐stream dissolved inorganic nitrogen (DIN) and dissolved reactive phosphorus (DRP) concentrations and to identify the influence of land cover and landscape structure. Using a synoptic approach we mapped concentrations in all stream orders. Using linear regression, the relations between the share of agricultural land and log‐transformed in‐stream concentrations were explored. Both the share of agricultural land in the entire ‘area of influence’ upstream from a sampling location, as well as the share in a 150‐m buffer around the stream were used as linear regression input variables. Log‐transformed DIN and DRP concentration variability was highest for lower order streams, while it averaged out in higher order streams during all seasons. Between‐season variation in export can mainly be attributed to discharge variation. In extremely dry periods, there are no significant relations between land cover/structure and in‐stream ln(DIN) concentrations and only weak relations for ln(DRP) concentrations. In other seasons, the share of agricultural land in the upstream area can explain concentrations in higher order streams better than in lower order streams. The prediction of ln(DIN) concentrations in lower order streams can be improved by using the share of agricultural land in a 150‐m buffer as an input variable. This indicates that hydrological connectivity must be taken into account for lower order streams, while land cover shares are enough to explain concentrations for higher order streams. Copyright © 2009 John Wiley & Sons, Ltd.