Phosphorus buffering in streams by benthic sediments : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

The loss of phosphorus (P) to aquatic ecosystems accelerates eutrophication – a problem felt worldwide. Central to any effort to monitor and mitigate the effect of P in stream is knowing how inputs of P, whether point or diffuse, map to P transport downstream. However, the stream itself possesses several mechanisms to attenuate P inputs thus blurring the connection between P inputs and P availability in-stream. For example, various stream biota and geochemical processes may remove P from or even release P to the water column. In particular, benthic stream sediments have the capacity to sorb P to their surfaces which may later desorb back into solution. This P sorption means benthic sediments can behave much like a buffer for P: a transient store of P which may offset changes to P concentrations in the stream. The thesis of this work is that the benthic sediment-P buffer is a predominant control on P availability at baseflow in streams. In its five studies, I investigate sediment-P sorption in detail but also examine multiple alternative pathways for stream P retention. Special attention is given to the sediment equilibrium phosphate concentration at net zero sorption (EPC0), which is the dissolved reactive P (DRP) concentration towards which sediments buffer DRP concentrations in the solution (i.e., sediment porewater and, via hyporheic exchange, the water column) through sorption. A systematic review and meta-analysis of the EPC0 in streams at baseflow – covering 45 studies and 466 stream sites across the globe – found wide variability in the disparity between in-stream DRP concentrations and sediment EPC0 (termed as a potential for P exchange). This contrasts with previous views that P in sediments and streamwater is typically in an ‘equilibrium’. Further, this potential for P exchange was moderated by sediment and stream characteristics, including sorption affinity, pH, and sediment exchangeable P concentrations. For example, fine benthic sediments are often highly sorptive but may also restrict hydrological exchange between the water column and the hyporheic zone, leading to wider disparities. Methodological factors were also influential (e.g., choice of solution, sediment pre-treatment, equilibration time), indicating a need for research on unified EPC0 methodology. The second study established that drying sediments prior to analyses (either air- or freeze-drying) biases sediment P fractionation measurements and inflates the variance of EPC0. Such drying techniques may lyse microbial biomass P, alter organic P availability, and age metal oxides responsible for sorption, thus complicating the natural sediment P chemistry. Instead, analyzing stream sediments fresh (wet) is recommended. The third study surveyed a variety of stream waters and sediments from catchments draining three distinct lithologies (alluvium, sedimentary, and volcanic basic) to assess the likelihood of various geochemical controls on stream DRP concentrations. Geochemical equilibria in the water-column indicated no significant potential for the (co-)precipitation of minerals that could sequester P (e.g., calcite). However, the sediments stored large amounts of P in labile and redox-sensitive forms: indeed, this labile P correlated with stream DRP concentrations but only for streams with likely sufficient hyporheic exchange. Catchment geology and redox cycling in stream influence sediment reactivity for P and so are a major source of between-stream variability in DRP concentrations. The fourth study focused on a confounding factor when interpreting EPC0: is P sorption or microbial P cycling responsible for sediment P buffering? Unlike some previous work, this study found a minimal role for sediment microbes to alter sediment EPC0 values even with replete labile carbon and nitrogen sources available. Further, sediments sterilized via γ-irradiation did increase in EPC0, but this increase was attributed to lysis of the microbial biomass – an overlooked P stock in streams. The study highlights that the sediment P buffer, while largely abiotic in nature, may subsidize microbial P demand in sediment biofilms, thus influencing stream ecological function. The last study examined P uptake at the stream reach scale. Considering two contrasting but predominant controls on stream P uptake – periphyton P demand and sediment P sorption – a natural way to separate the two processes was to measure P uptake under light and dark conditions. Stream gross primary productivity (driven by periphyton) was high as expected for this open-canopy stream. However, this did not translate to an increase in stream P uptake when compared to dark conditions. Sediments were highly sorptive and their relatively low EPC0 suggested a potential for P removal throughout the experiment. Thus the sediment P buffer was likely most responsible for the measured stream P uptake although different stream conditions (e.g., greater nitrogen availability) could increase periphyton’s relevance and should be studied further. In summary, the benthic sediment P buffer can contribute to the regulation of P availability in many streams. Sediments may attenuate P inputs, thus dampening DRP variation at baseflow and prolonging the legacy of past P inputs in the catchment. Stream hydrology (e.g., hyporheic exchange), geochemistry (e.g. pH), and biota (e.g., sediment microbial P demand) are among the chief external factors that may moderate or interact with the sediment P buffer and deserve further study. Predicting P availability in streams remains a major challenge, but understanding the sediment P buffer will greatly improve our ability to prevent stream eutrophication.