The present study comprises quality-controlled discharge and water quality data from 40 stations in South and Southwest Germany. Discharge Q (m³ s⁻¹) and water quality data C (mg L⁻¹) are delivered by various federal agencies, covering catchments located in Rhineland Palatinate (RLP), Baden-Württemberg (BW), and Bavaria (BY; LfU Bayern, 2022; LfU RLP, 2022; LUBW, 2022). Subsequently, the catchments include different regions with contrasting climate and catchment characteristics (Tables S1–S2 in the Supplement). The study focuses on mid-mountain catchments in Eifel, Hunsrück, Palatinate Mountains, Black Forest, Swabian Alb, and Upper Palatinate Forest. Additionally, catchments are also located in Upper Rhine Lowlands, Württembergisches Unterland, and Franconia (Fig. 1). All regions vary in climate, geology, land cover, and altitude. Catchment altitude varies between 147.5 and 835 m a.s.l. Average yearly precipitation ranges from 527 mm yr⁻¹ up to 1596 mm yr⁻¹ (DWD, 2022). According to the Köppen and Geiger climate classification, the climate is humid continental (Dfb) in Bavaria and most parts of Baden-Wurttemberg (Beck et al., 2018). The influence of ocean climate (Cfb) increases in the Northwest of Baden-Wurttemberg and Rhineland Palatinate (Beck et al., 2018). Streams reveal nival and pluvial regimes. The mean discharge spans three orders of magnitude and varies from 0.41 to 49.73 m³ s⁻¹.

The regional geology is heterogeneous with crystalline rocks in the Palatinate Mountains and Black Forest. Formations of carbonate rocks are located in Swabian Alb, Alps and Franconia, respectively. Further rock formations are Hunsrück schist in the Hunsrück, sandstone, and slate in the Eifel (BGR, 2006). Catchment sizes vary between 49 (Schutter) to 5434 km² (Naab). However, the focus is on mid-scale catchments (< 1000 km²), which account for over 80 % of all catchments (Fig. 1). Arable land, pastures, or forests predominantly cover the majority of these catchments.

Discharge and water quality samples of public agencies typically are not sampled at the same locations, resulting in data selection criteria (see below) a.

Assuming that the general behaviour of discharge generation and solute mobilization is not changing abruptly within one catchment, a small distance between the respective gauges was an essential criterion for selecting streams. For that reason, only close stations (max. distance ca. 8 km) of discharge and water quality were accepted as one site in this study.

Further, alteration in SEM caused by varying humidity level, seasonal variation and changing climatic conditions across decades were determined by splitting the data set into various observation periods: a.

Whole observation period, comprising all measured solute concentrations.

The analysis of in-stream solute concentrations reflects the static signatures and nutrient stress levels of anthropogenically impacted catchments, with trends possibly influenced by human activity and climatically changing conditions (Table 2). Mean nutrient concentrations ranged from 0.74–9.37 mg L⁻¹ for NO₃-N, 0.04–0.34 mg L⁻¹ for NH₄-N, 0.06–0.56 mg L⁻¹ for TP, 0.02–0.27 mg L⁻¹ for SRP, and 1.9–7.78 mg L⁻¹ for TOC. Trend analysis across the full observation period for each catchment revealed significant decreases in mean concentrations in 57.5 % (NH₄-N), 72.5 % (NO₃-N), 60.5 % (TP), and 67.5 % (SRP) of catchments, as confirmed by ANCOVA (p<0.05). Geogenic solutes occurred at higher concentrations (Ca2+: 13–180.9 mg L⁻¹; Mg2+: 2.57–55.12 mg L⁻¹) and showed greater absolute variability, but with lower interannual variation compared to nutrients (Table 2).

Long-term slope b and CV_C / CV_Q metrics in southern Germany revealed explicit solute export mechanisms (SEM) for nutrients and geogenic solutes, reflecting differences in mobilization, transport, and transformation processes influenced by seasonality, humidity, and climatically changing conditions (Table S4). Nutrients such as NH₄-N, SRP, and TP exhibited more pronounced chemodynamic behaviour than conservative solutes like Ca2+ and Mg2+ (Fig. 3). When exhibiting chemodynamic behaviour, NH₄-N showed dominant enrichment pattern (b>0), SRP tended toward dilution dynamics, and TP displayed strong enrichment behaviour (b>0.6) at elevated CV_C / CV_Q ratios (>1). Under condition of high enrichment behaviour (b>1), TOC showed chemodynamic behaviour indicative of discharge-decoupled processes, whereas b<1 generally reflected both chemostatic and chemodynamic patterns. In contrast, Ca2+ and Mg2+ typically displayed dilution dynamics, while NO₃-N combined dilution and enrichment pattern but retained an overall chemostatic signature.

Temporal analyses revealed significant changes in SEM across seasons (NH₄-N and SRP) and across decades (NH₄-N, SRP, TP, and TOC). NH₄-N exhibited the strongest seasonal variability, with pronounced enrichment pattern in summer and reduced enrichment dynamics in winter. SRP typically tend towards non-significant dilution behaviour during summer and enrichment behaviour in winter, highlighting its temporal variability (Fig. 4). Over the recent decade, however, dilution processes weakened for SRP, as indicated by a positive Δb, particularly in catchments that had displayed dilution SEM in earlier decades (Fig. 5). For both NH₄-N and SRP, variability in Δb increased with higher CV_C / CV_Q ratios.

In contrast, TP and TOC showed less pronounced seasonal changes. However in period 2, characterized by warmer, drier summers and sporadic intense precipitation events (KLIWA, 2021), TP and TOC shifted toward enrichment pattern indicated by a positive Δb. TP especially exhibited increasing slope b (Δb), predominantly in catchments with prior dilution dynamics. For both solutes, catchments with chemodynamic signatures predominantly shifted toward stronger enrichment behaviour (Fig. 5).

NO₃-N, Ca2+, and Mg2+ showed no significant responses to seasonal variability, humidity, or the climatic differences between period 1 and period 2 (Fig. 4). Changes in Δb for these solutes were minor (-0.25 to +0.25), with only slight, non-significant trends toward enrichment for NO₃-N and dilution for Ca2+ and Mg2+ (Table 4). Overall, solutes with chemostatic signatures (NO₃-N, Ca2+, Mg2+) displayed low temporal variability compared to reactive nutrients (NH₄-N, SRP), which are influenced by biogeochemical processes. Additionally, no SEM changes were observed in response to variations between wet and dry years (Fig. 4).

SEM of nutrients are influenced by distinct explanatory controls. The impact of these controls is potentially influenced by changing climatic conditions. Pearson correlation coefficients (ranging from 0.33 to 0.75, p<0.05 corrected by Benjamini-Hochberg procedure) were used to assess relationships between SEM (slope b) and explanatory controls for the decade before and after 1 January 2012. In addition, statistical differences between correlations for period 1 and period 2 were formally evaluated using analysis of covariance (ANCOVA). The obtained Pearson correlation coefficients indicate both positive and negative correlations (Table 3).Shallow-sourced nutrients are primarily influenced by near-surface environmental conditions, particularly climate (e.g. temperature and precipitation), as well as soil moisture and the decomposition of organic material on the ground and in the upper soil layers. In contrast, geogenic solutes such as Ca2+ and Mg2+ are predominantly controlled by geological weathering processes. In addition to climatic influences, land cover, soil moisture conditions and geological characteristics affect shallow-sourced solutes.

For NH₄-N, soil moisture (-) shows negative correlations with slope b, indicating dilution behaviour in wet catchments and enrichment behaviour in dry catchments. During period 2, the impact of these controls intensified, leading to a stronger regulation of c–Q relationships by soil moisture. SRP is primarily influenced by the climatic factor evapotranspiration (++) and the geological factor clay rock (-), with both controls becoming more pronounced in the recent decade (period 2). High evapotranspiration enhances enrichment SEM, while high clay rock proportions in catchments promote dilution behaviour. TP, which includes both particle-bound and soluble reactive phosphorus, is influenced by a broader range of factors, including drought index (++), precipitation (++), evapotranspiration (++), evaporative index (-), aridity index (-), soil moisture (++), carbonate rock (+), and clay rock (-). While the influence of climatic controls (precipitation, evapotranspiration, dMI) and soil moisture has increased, geological controls have become less influencing in regulating SEM. Further, TOC, which is primarily governed by surface processes, is exclusively influenced by evapotranspiration (++), with its regulatory impact on SEM diminishing in period 2.

Unlike the more reactive solutes, NO₃-N exhibits chemostatic behaviour but remains influenced by multiple controls, including arable land (+), urban area (-), drought index (--), precipitation (-), and evapotranspiration (-). Arable land, a diffuse source, supports enrichment behaviour, whereas urban areas, acting as potential point sources, promote dilution dynamics. Wet catchments (characterized by high soil moisture, high precipitation, and humidity inferring drought index values) are more likely to exhibit dilution SEM. Over time, the regulatory influence of explanatory controls on NO₃-N export dynamics has remained stable, indicating a robust system. As expected, the SEM of geogenic solutes (Ca2+ and Mg2+) are largely unaffected by external controls. For Mg²⁺ the proportion of sandstone correlates positively with slope b, whereas for Ca²⁺, no significant regulatory controls on SEM were observed. Notably, in the recent period, the effect of sandstone proportion on Mg2+ export dynamics remained unchanged.

Significant differences in mean nutrient and geogenic solutes concentrations were observed across catchments. Reactive solutes (NH₄-N, SRP, TP) exhibited low mean concentrations, consistent with previous studies (Ebeling et al., 2021; Musolff et al., 2015). Low NH₄-N concentrations reflect its rapid turnover, while phosphoric species are quickly taken up by plants, limiting their availability (Birgand et al., 2007; Martí and Sabater, 1996). By contrast, NO₃-N concentrations averaged 4.04 mg L⁻¹, reflecting agricultural inputs and persistent legacy effects (Aubert et al., 2013). Nutrient concentrations in Germany were historically elevated but have declined since 1990 due to reduced fertilizer use. However, NO₃-N shows a slower decrease, which can be explained by legacy effects compared to NH₄-N, TP, and SRP (Basu et al., 2010). Declines in NH₄-N, TP, and SRP concentrations can also be attributed to prolonged droughts associated with the climate of period 2, as extended dry summer periods lead to lower water levels and reduced mean concentrations (Outram et al., 2016; Van Loon et al., 2019). In contrast, intense high-flow events increase nutrient export, but biweekly to monthly sampling likely misses these peaks, and hence, leads to potential underestimation of total nutrient loads. Nevertheless, the sampling frequency remained unchanged across both study periods. Thus, the observed changes in SEM are more likely explained by hydroclimatic shifts associated with changing climatic conditions rather than by sampling limitations. Still, further high-resolution monitoring would be valuably to better capture event-driven dynamics, as individual events can exhibit slope b values that deviate distinctively from the seasonal behaviour (Knapp et al., 2020; Winter et al., 2024).

TOC, is less reactive and more persistent, with carbon-rich landscape types (e.g. wetlands or riparian zones) serving as sources, visible in higher and more variable mean concentrations. However, interannual variability in TOC concentrations is less pronounced than for other nutrients, suggesting lower influence from anthropogenic actions. The mean concentrations of geogenic solutes (Ca2+ and Mg2+) are consistent with previous reports, and their spatial variability reflect the heterogeneous geology of the catchment regions (Musolff et al., 2015). These solutes, primarily governed by geological processes, showed only slight increases in mean concentration for some catchments (Ca²⁺: 15 %, Mg²⁺ 12.5 %). The persistence of the lithosphere and weathering processes contributed to low interannual variability. However, changing climatic conditions from period 1 to period 2, including warmer temperatures and lower groundwater levels, could promote higher residence times and enhanced weathering rates. This may lead to slight increase in solute concentrations, but further confirmation is needed (Li et al., 2022; Musolff et al., 2015).

Changes in slope b over time reveal shifts in solute mobilization, transportation, and transformation processes, depending on seasonality, humidity levels, and responses to climatically different periods. SEM cluster into three groups: (1) seasonal-dynamic solutes, (2) long-term dynamic solutes, and (3) long-term stable solutes (Fig. 3).

Seasonal-dynamic and long-term dynamic solutes exhibit increased enrichment and reduced dilution dynamics under the influence of changing climate. Catchment characteristics, including land cover, geology, pedosphere, and climate, significantly shape SEM in both period 1 and period 2, as confirmed by Pearson correlation coefficients (p<0.05).

For seasonal-dynamic NH₄-N, humidity conditions expressed by the drought index (dMI) regulate SEM, driving dilution dynamics under wet conditions. Rapid turnover rates (e.g., uptake, transformation) deplete NH₄-N sources as discharge rises and hydrological connectivity increases (Birgand et al., 2007; Martí and Sabater, 1996; Pohle et al., 2021). In drier catchments, drought reduces hydrological connectivity and limits solute transport particularly during prolonged heatwaves, enhancing accumulation, and intensifying enrichment dynamics. Warmer climatic conditions in period 2 amplifies the effects of drought (increasing impact of dMI on SEM), explained by extended transport limitations while increasing solute accumulation, ultimately intensifying enrichment dynamics.

For SRP, evapotranspiration and clay-rich soils are key controls. Low evapotranspiration sustains hydrological connectivity in shallow soil, linking terrestrial SRP sources to streams and promoting dilution via source depletion. Clay soils act as buffers, hindering SRP mobilization through adsorption-desorption processes (Basu et al., 2011; Thompson et al., 2011). In period 2, this buffering effect has intensified, highlighting rapid SRP leaching in areas lacking clay minerals. At the same time, higher evapotranspiration increasingly drives SRP accumulation and reduced hydrological connectivity in period 2, thereby weakening dilution behaviour in response to the altering climatic conditions.

TP SEM reflects mobilization through discharge and biogeochemical processes since TP occurs in both solid and soluble forms. Increased discharge activates preferential pathways, enhancing enrichment behaviour through sediment mobilization. TP SEM are influenced by drought index (dMI), precipitation, evapotranspiration, soil moisture, clay rock, and carbonate rock. Carbonate rock facilitates preferential flow, amplifying enrichment behaviour, while clay materials slow phosphorus mobilization (Basu et al., 2011; Thompson et al., 2011). Changing climatic conditions from period 1 to period 2 expands TP sources and transport limitation (control: evapotranspiration, soil moisture, drought index), intensifying mobilization during shortened wet periods in summer, as well as in autumn and winter when water table rises (control: precipitation). This enhances enrichment dynamics and weakens dilution patterns. In period 2, controls of climate, soil moisture and carbonate rock become stronger, whereas clay's impact on SEM declines.

For TOC, prolonged drought (control: evapotranspiration) leads to the accumulation of large TOC sources that are increasingly mobilized during shortened wet periods, thereby amplifying enrichment behaviour (Dupas et al., 2025).

SEM of NO₃-N is regulated by several controls but maintains a distinctive chemostatic behaviour representing a long-term stable solute. Wet conditions, explained by high precipitation, and humidity inferred drought index values, enhance dilution dynamics. In contrast, arable land and high evapotranspiration rates promote enrichment behaviour. Fertilizer use on arable land accumulates nitrogen in the soil creating diffuse sources. Meanwhile, evapotranspiration-induced dryness reduces constant drainage and prevents source depletion. When hydrological connectivity is restored (e.g., during rainfall and rising water table), accumulated nutrients are rapidly mobilized, resulting in pronounced concentration increases. Urban areas typically create dilution behaviour due to steady contributions from point sources (Aubert et al., 2013; Basu et al., 2010; Dupas et al., 2018; Musolff et al., 2015). In recent years, the effects of land cover (urban area and arable land) on SEM have lessened, suggesting reduced anthropogenic disturbance due to improved wastewater treatment and decreased fertilizer use. Minor changes, like lower hydrological connectivity and increased NO₃-N generation during prolonged heatwaves, slightly amplify enrichment behaviour and weaken dilution SEM. However, the increase in slope b remains primarily masked by anthropogenic NO₃-N inputs, which continue to dominate SEM.

Consequently, in addition to climatic factors, other drivers may have influenced the observed changes in SEM. In particular, legacy effects and land use changes over the study period may have contributed to long-term solute transport dynamics. Historical changes in fertilization regulations and land management practices may have resulted in persistent nutrient stores that continue to affect present-day solute export within individual catchments (Dupas et al, 2016; Li et al., 2024). Nevertheless, our analyses indicate that climatic parameters exhibit a distinctly stronger explanatory power for changes in solute export mechanisms between the two climatically different periods than land use parameters and their dynamics over time (Table 4). While urban areas and arable land show a significant influence on NO₃-N export, no significant influence was detected for NH₄-N, SRP, TP, or TOC, highlighting a comparatively weaker influence of land use parameters relative to climatic controls. Additional analyses of land use change demonstrate that substantial changes occurred between 2006 and 2018, which may have influenced changes in SEM (Kong et al., 2022). However, despite these land use alterations, no significant correlations were identified between changes in SEM and changes in land use composition (Table S5). This suggests that, although land use change and nutrient legacy effects might contribute concurrently, the observed changes in SEM are predominantly associated with climatic variables and should be interpreted in this broader context. However, the potential concurrent influence of land use change and nutrient legacy effects should be considered when interpreting changes in solute export mechanisms. Unlike seasonal-dynamic and long-term dynamic solutes, Ca2+ and Mg2+ are mainly governed by geology, particularly sandstone. For Mg²⁺, sandstone significantly hinders the production of Mg2+ ions and weakens dilution dynamics. The climatic conditions in period 2 may extends lithosphere residence time, slightly but not significantly enhancing dilution dynamics (Botter et al., 2020; Li et al., 2022). However, despite changing climatic conditions, Ca²⁺ and Mg²⁺ export dynamics have remained stable over the decades, with regulatory controls maintaining a consistent influence on Mg²⁺ SEM. Changing climatic conditions from period 1 to period 2 markedly shifts controls on seasonal-dynamic and long-term dynamic solutes, strengthening climatic and soil moisture effects that drive enrichment SEM, while weakening drivers of dilution pattern. This trend may increase solute concentration peaks, risking ecosystem and human health by surpassing water quality standards (Outram et al., 2016; Radach et al., 1990; Winter et al., 2020). Effective catchment management strategies to counter enrichment behaviour are essential. Climate impacts on long-term stable solutes remain minor, though slight increases in Ca2+ and Mg2+ dilution dynamics are assumed due to longer residence times.