Nutrient export from catchments on forested landscapes reveals complex nonstationary and stationary climate signals

Abstract

[1] Headwater catchment hydrology and biogeochemistry are influenced by climate, including linear trends (nonstationary signals) and climate oscillations (stationary signals). We used an analytical framework to detect nonstationary and stationary signals in yearly time series of nutrient export [dissolved organic carbon (DOC), dissolved organic nitrogen (DON), nitrate (NO 3 À -N), and total dissolved phosphorus (TDP)] in forested headwater catchments with differential water loading and water storage potential at the Turkey Lakes Watershed in Ontario, Canada. We tested the hypotheses that (1) climate has nonstationary and stationary effects on nutrient export, the combination of which explains most of the variation in nutrient export; (2) more metabolically active nutrients (e.g., DON, NO 3 À -N, and TDP) are more sensitive to these signals; and (3) catchments with relatively low water loading and water storage capacity are more sensitive to these signals. Both nonstationary and stationary signals were identified, and the combination of both explained the majority of the variation in nutrient export data. More variation was explained in more labile nutrients (DON, NO 3 À -N, and TDP), which were also more sensitive to climate signals. The catchment with low-water storage potential and low water loading was most sensitive to nonstationary and stationary climatic oscillations, suggesting that these hydrologic features are characteristic of the most effective sentinels of climate change. The observed complex links between climate change, climatic oscillations, and water nutrient fluxes in headwater catchments suggest that climate may have considerable influence on the productivity and biodiversity of surface waters, in addition to other drivers such as atmospheric pollution.Citation: Mengistu, S. G., C. G. Quick, and I. F. Creed (2013), Nutrient export from catchments on forested landscapes reveals complex nonstationary and stationary climate signals, Water Resour.

Spatiotemporal Analysis of Dissolved Organic Carbon and Nitrate in Waters of a Forested Catchment Using Wavelet Analysis

“…Previous observations of the dynamics of DOC and NO 3 − (Bernal et al, 2005;Heffernan and Cohen, 2010;Halliday et al, 2012;Lutz et al, 2012;Worrall et al, 2015), however, have shown that their temporal dynamics may vary (Taylor and Townsend, 2010;Mengistu et al, 2013;Thomas et al, 2014). Therefore, a tool is required to investigate these different non-stationary dynamics.…”

“…Wavelet transform coherence (WTC) analysis reveals correlations between two time series that are timescale specific. In the past, there were only a few studies that included DOC and NO 3 − (Kang and Lin, 2007;Rusjan and Mikoš, 2010) or water quality time series in general (Mengistu et al, 2013) in wavelet analysis. Generally, a spatial component was not included.…”

The river as a chemostat: fresh perspectives on dissolved organic matter flowing down the river continuum

“…At short time scales (e.g., hydrologic events), terrestrial-aquatic transport within a catchment, which is often expressed through hysteretic patterns between runoff and DOC concentrations or DOM composition, can differ dramatically across the longitudinal continuum as a result of differences in source contributions and mixing (Pacific et al 2010). At longer time scales (e.g., years to decades), DOM export is influenced by climatic cycles (e.g., El Nino-Southern Oscillation) and trends (e.g., warming) (Mengistu et al 2013).…”

“…Additionally, riverine DOM is the product of many complex processes that produce, consume, and modify organic matter. These processes change over daily to decadal time scales and over space scales ranging from low-order streams to high-order rivers (Mengistu et al 2013). Unraveling these complex processes is key to improving our understanding of rivers and the many ecosystem services they provide.…”

The application of dynamic self-organised multilayer network inspired by the Immune Algorithm for weather signals forecast

“…Hence, the statistical properties (e.g. mean, variance and autocorrelation) of the data are stable with respect to time [25], such as climate oscillations [26]. In mathematics, stationary can be defined as follows, when the distribution of (xt 1 , .…”