Quantifying changes and trends of NO₃ concentrations and concentration-discharge relationships in a complex, heavily managed, drought-sensitive river system

Highlights

• Legacy-N led to stream NO₃ concentrations showing chemostatic behavior.

• Export pattern closely linked to export regime of stream NO₃ concentrations.

• Future droughts will lead to a decrease in stream NO₃ concentrations.

• Droughts will result in stream NO₃ concentrations showing a more positive correlation with streamflow.

• Droughts will result in stream NO₃ concentrations showing more chemodynamic behavior.

Abstract

Stream nitrate nitrogen (NO₃) concentrations and concentration-discharge (C-Q) relationships are key indicators of water quality. However, their long-term variability and response to climate change in large-scale catchments are still poorly understood. Here, we report a synoptic survey and long-term observations (2000–2021) in the Spree catchment (10,105 km²), Germany, to identify the spatial–temporal patterns in stream NO₃ concentrations and C-Q relationships, as well as the underlying environmental drivers. We found that in the context of gradual reduction of nitrate inputs, the stream NO₃ concentrations had a decreasing trend, but showed different magnitudes at different spatial scales (-0.01–0.48 mg L^-1 yr⁻¹). In the upstream parts of the catchment - with a high proportion of farmland - high levels of stream NO₃ concentration remained with a risk of eutrophication due to the large nitrogen legacy. Especially in winter, stream NO₃ concentrations were much higher due to the groundwater export with high NO₃ concentrations, a decreased dilution effect of rainfall and vegetation uptake became clear. Consequently, the large nitrogen legacy in the upper catchment resulted in NO₃ concentrations not significantly changing with streamflow and showing chemostatic behavior. This was different from NO₃ dynamics in the mid- and lower-catchment areas, which were positively related to streamflow, showing a chemodynamic behavior. Further, we found the export patterns (i.e. enrichment vs. dilution) of stream NO₃ concentrations strongly correlated with the export regimes (i.e. chemostatic vs. chemodynamic), which were also affected by drought conditions. This is probably due to the decrease in stream depth which would enhance benthic removal leading to a decrease in stream NO₃ concentrations. Reduced hydrological connectivity leads to higher spatial heterogeneity of residual NO₃ in soil and drainage water, which will result in chemodynamic behavior of stream NO₃ concentrations. After rewetting, the export of these high NO₃ concentration in soils led to elevated stream NO₃ concentrations, resulting in a positive correlation with streamflow. Our work revealed significant heterogeneity in stream NO₃ concentrations and C-Q relationships at different spatial–temporal scales in large catchments. Critically, current stream NO₃ concentrations and C-Q relationships are likely to respond strongly to future drought, leading to challenges for future land and water management.

Introduction

Over the past century, dramatic increases in global agricultural land conversion and urbanization, along with widespread fertilizer use and population growth has led to the deterioration of stream water quality worldwide (Basu et al., 2022, Liu et al., 2020a, Schilling et al., 2016). A particular problem is nitrate nitrogen (NO₃) pollution due to its high mobility and impact on water eutrophication (Liu et al., 2022a, Wu et al., 2022). Given that streamflow is the main transport pathway of NO₃, clarifying the stream NO₃ concentration-discharge (C-Q) relationships can improve understanding of the biogeochemical processes governing NO₃ and thus help to prevent and control stream NO₃ pollution in catchments (Musolff et al., 2017, Thompson et al., 2011).

In response to increasing stream NO₃ pollution, stringent NO₃ management measures to reduce input loads have been adopted in a wide range of catchments (Van Meter et al., 2018, Wu et al., 2021, Yang et al., 2022). However, the response of stream NO₃ concentrations and C-Q relationships to such management measures can have long time lags due to long water residence times. In addition, mainly due to geographic variations in climate and landscape characteristics, inconsistent results have been reported in terms of dominant controlling factors (Ebeling et al., 2021, Guo et al., 2022, Lintern et al., 2021, Liu et al., 2022a, Liu et al., 2022b, Zhi and Li, 2020). Moreover, the response of stream NO₃ concentrations and C-Q relationships to climate change is still largely unknown. These factors limit an improved understanding of the driving mechanisms of stream NO₃ in complex landscape catchments as well as projection of likely future trends.

Climate is a major driver of N biogeochemical processes, especially over longer-time scales, and both climate and consequent biogeochemical processes have undergone profound changes in many areas over the past half century (Guo et al., 2022, Liu et al., 2023, Yang et al., 2022). Currently, climate warming is intensifying and increasing droughts will potentially reduce streamflow and stream depths over extensive areas (Davidson and Janssens, 2006, Wu et al., 2022). In shallower streams, the relative contact area of the water volume with the sediment surface, and therefore exposure to denitrification at the benthic interface is enhanced. Thus, the majority of N removal usually occurs upstream with shallower stream depths (Alexander et al., 2000, Liu et al., 2022a). For example, studies in the Mississippi River catchment have shown that a rapid decline in the average first-order rate of N loss with increasing stream depth - from 0.45 days⁻¹ to 0.005 days⁻¹ (Alexander et al., 2000). Reduced streamflow can even result in streams becoming intermittent, with disconnected flow pathways affecting the hydrological transport of NO₃ (Wu et al., 2021). These changes will affect stream NO₃ concentrations and C-Q relationships, and there is an urgent need to consider climate change in predictions of stream NO₃ concentrations and C-Q relationships as well as the effects of current and future management measures.

Catchments have diverse landscape characteristics and anthropogenic pressures that influence NO₃ export-retention and hydrological transport processes (Goyette et al., 2019, Liu et al., 2021, Liu et al., 2020a, Wu et al., 2022). Stream NO₃ concentrations and C-Q relationships differ at different spatial scales due to variations in landscape composition and configuration (Ebeling et al., 2021, Liu et al., 2020a, Zhi and Li, 2020). Given that many landscape ecosystem functions (e.g. release and retention of NO₃ and water) also change at seasonal and interannual scales (Guo et al., 2022, Liu et al., 2022a), stream NO₃ concentrations and C-Q relationships vary temporally. Previously, the percentage of agricultural land was considered to be the main driver of increased stream NO₃ concentrations due to high N fertilizer inputs, which led to an enrichment and usually chemostatic behavior of stream NO₃ concentrations (Basu et al., 2010, Thompson et al., 2011). Recently, the impact of agriculture on stream NO₃ concentrations is gradually decreasing due to the reduction of N inputs (Basu et al., 2022, Krupa et al., 2011, Van Meter et al., 2018). Also, the enrichment or dilution patterns of nutrients depending on streamflow variations in agricultural catchments have been inconsistently reported (Miller et al., 2017, Zhi and Li, 2020), and little is known about the effects of export regimes (C-Q relationships; i.e. chemostatic vs. chemodynamic). Meanwhile, understanding of long-term changing stream NO₃ concentrations and C-Q relationships is still limited for larger, heavily managed catchments with complex mosaics of landscape characteristics and human interventions but is particularly important for informing management of stream NO₃ in the future.

To address these research gaps, we investigated the dynamics of stream NO₃ concentrations and C-Q relationships in the large, complex and heavily managed catchment of the River Spree and 67 of its sub-catchments in Germany, with different land use, soil, climate and topographic characteristics, and which is increasingly drought prone. The study used both intensive synoptic surveys and long-term observations (2000–2021). We hypothesized that stream NO₃ concentrations and C-Q relationships would differ according to landscape spatial heterogeneity and changes in environmental conditions over long-term time scales. We aimed to

• Identify spatial–temporal patterns of NO₃ concentrations in a large, complex and heavily managed catchment;

• Investigate C-Q relationships at different spatial scales and over longer time periods.

• Quantify the influence of different environmental controls on NO₃ dynamics and understand the impact of recent intensifying drought conditions on stream NO₃ concentrations and C-Q relationships.