Optimising multi-site sensor networks in lowland permeable catchments for comprehensive water quality monitoring and nitrogen mass balancing during baseflow conditions

High-frequency data are essential to elucidate intricate fluvial water quality dynamics, but current understanding is often limited by measurements taken solely at the catchment outlet. In densely populated and agriculturally intensive lowland permeable catchments, such as Chalk streams in southern England, the spatial heterogeneity of processes driving solute mobilisation, transport, and fate can only be unravelled through monitoring at high spatio-temporal resolutions. In this study, we deployed a network of in-situ sensors in the lower section (ca. 18 km third-order reach) of a chalk stream during baseflow conditions to address this limitation. We focused particularly on reach-scale processing of reactive nitrogen (N), using a mass-balance approach based on high-frequency measurements, to quantify the relative importance of different sources (springs, sewage effluent) and sinks (microbiological processes in the river). Continuous in-situ measurements revealed important event-based influences that grab sampling would likely fail to capture, such as rainfall disturbance of metabolic activity and polluted discharges from combined sewer overflows. Mass balances showed the majority of fluvial nitrate load originates in the first half (ca. 8 km) of the study reach, where it increased by a factor of 2.8 from 324 kg d^-1 to 914 kg d^-1. This was mainly attributed to a sewage treatment discharge (37% of accreted load), and chalk spring discharges (55%) carrying loads primarily from agricultural inputs. Nitrate assimilation (overall for the study reach 80 mg m^-2 d^-1) by autotrophs was estimated to be the main retention pathway but accounted for only 2.6% ± 0.6% of total loading to the stream reach. Despite the short study duration (5 weeks) and extreme low-flow conditions, we concluded that the river has limited capacity to attenuate gross N loads, causing detrimental ecological impacts to the downstream marine conservation area. Our findings underscore the management imperative of reducing catchment N loading to nitrate-enriched streams as the most effective way of controlling N exports and restoring the removal efficacy of natural stream ecosystems. Our novel mass-balance sensor-network approach is effective at quantifying the relative importance of solute sources and sinks in heterogenous catchments. But, to maximise the data value of multi-station networks, we recommend recording measurements over long durations and in unison with other sampling strategies e.g. tracer studies, terrestrial and ecological monitoring, sediment-core sampling, and longitudinal profiling.