OR/13/031 General discussion and conclusions

Tye, A M, Hurst, M D, and Barkwith, A K A P. 2013. Nene phosphate in sediment investigation — Environment Agency project ref: 30258. (Water Framework Directive). British Geological Survey Internal Report, OR/13/031.

Sediment phosphorus chemistry

The results of the sediment TP and OEP chemistry and its dynamics conform to previous research in similar fluvial settings. However, the expectation that sediment texture in the sampled cores would become finer as distance from the headwaters was not met, as it was obvious that spatial and temporal influences as well as the depositional environment were large determinants on sediment texture. Organic matter content of the sediment was generally found to increase with distance from the headwaters, and results suggested that this was strongly dependent on clay content and not particle size. Bedrock geology appeared to be the biggest influence on TP concentration, with the depositional environment in which the various rocks were laid down controlling the associations of P containing minerals. This was demonstrated by the strong correlations to Mn and Fe. However, there were few geochemical associations between OEP and the measured reactive surfaces that are generally considered to be important for sorption, thus suggesting that OEP concentration in the sediment was more a function of the local depositional environment. OEP was also found to be a small proportion of TP, typically being <5%. The importance of depositional environment was demonstrated by the analysis with depth (core_D) from each water body. No consistent pattern being found throughout the length of the river analysed. Pinay et al. (2002) and Fisher et al. (2004) both emphasise the importance of flooding in particular for resetting the sediment structure and texture both within the channel and floodplain.

Concentrations of OEP in the sediment were found to be as high as 100 mg kg^-1 and these levels of phosphate P will obviously have the potential to increase macrophyte growth. Rooted aquatic plants have the potential to derive almost all their P requirements from bio-available sediment P reserves (Mainstone & Parr, 2002), although Pelton et al. (1998) suggested that the relative contribution of root uptake to macrophyte P demand varied on the SRP concentration in the overlying water. The increased growth of aquatic plants where there are high levels of bio-available P can lead to other problems. Along with greater plant growth trapping more sediment, Mainstone & Parr (2002) suggest that extra P (i) increases re-growth after plant management, (ii) the species community structure can be altered, favouring species with high growth rates and (iii) root depth is reduced, potentially increasing the plants susceptibility to being ripped out at high river flows and associated sediment remobilization.

As the amount of OEP is higher in the sediments than is recommended for most agricultural soils, this may suggest that the sediments have sorbed additional phosphate from non-agricultural sources, once in the channel. Indeed, results from Table 6 examining the potential for further sorption of SRP using EPC_0Sat suggest that the sediments have the potential to adsorb significantly more SRP. Sediments with such high potentially bio-available phosphate may pose problems in the future with issues associated with climate and river management. In particular, the management of phosphates associated with Sewage and Waste Water may present challenges. The limited analysis of PO₄-P:B ratios in the river waters suggests that the PO₄-P present is largely associated with the output of sewage and wastewater treatment. The benefit of examining the ‘Effective Phosphorus Concentration’ is that it has demonstrated that the sediment is likely to absorb, rather than desorb P. However, this depends largely on the concentration of PO₄-P in the river water and consideration that the EPC₀ is a function of sediment and water chemistry and therefore can be expected to vary slightly as river conditions change and sediment properties change with time (Stutter & Lunsdon, 2008). Whilst water PO₄-P concentrations were always greater than the EPC₀ when we sampled, examination of data supplied by the EA for PO₄ concentrations at Oundle give a range of between 0.05 and 0.6 mg L^-1. Dividing these values by three to obtain PO₄-P concentrations, suggest that on occasion the water PO₄-P concentrations may be lower than the EPC₀ calculated for water body 6 (0.05 mg L^-1). Thus although these events don’t appear to last long it may be possible that at some points in the yearly cycle, the sediments may desorb PO₄-P. However, generally water PO₄ concentrations are around 0.2 mg L^-1 PO₄-P which means they are greater than the EPC₀ and sediments will favour the sorption of PO₄-P. Calculations suggest that the sediment can take up about 10% of the PO₄-P that occurs in the 10 cm water depth above the sediment over the distance of a kilometre.

It would appear that a major decrease in PO₄ water concentration was achieved for the River Nene in 1998, where values were reduced from 2–3 mg L^-1 to 0.1–0.2 mg L^-1. Thus, if further improvements were to be made to the STW output, the concentration of SRP in the water could become lower than the EPC₀ more often leading to desorption of PO₄-P. Evidence from the EPC₀ calculations suggest that initially water bodies 5 and 6 would be most vulnerable to decreasing PO₄-P concentrations in river water as these have the highest EPC₀ values.

Climate forecasts for future decades suggest that dryer conditions are more likely to occur in the summer in eastern England (Murphy et al. 2009). This could produce low flow conditions in the Nene. However, assuming that the water PO₄-P is largely derived from sewage waters, it is likely that concentrations would increase as sewage waters may constitute a greater proportion of the river causing less dilution of P (Neal et al. 2010). Jarvie et al (2006) also suggests also that as a guide, where river water SRP concentration exceeds EPC₀, release of SRP from the sediment to the water column during periods of low flow (i.e. times of greatest eutrophication risk) is low. With respect to flooding conditions, it would generally be considered that dilution of PO₄-P may occur which could lower river PO₄-P concentrations to below those of the EPC₀. However the survey of Nene waters carried out in this research showed concentrations of PO₄-P still greater than EPC0 values, even after a prolonged period of high stream flow. Neal et al. (2010) suggest that river SRP concentrations could be maintained during flooding periods due to overflow of sewage facilities such as septic tanks.

Sediment dynamics within the river system

As large quantities of phosphate within the river channel is primarily attached to sediment, the erosion, deposition and transport of the sediment within the channel has been shown to be of great importance to where phosphate accumulates, and then its potential interactions with the river water. The CPD model has demonstrated that there is low sediment input into the Nene, largely due to the catchment topography. Thus, the dynamics of sediment input and transport can be seen as being largely driven by both low sediment supply as well as precipitation.

The survey of the sediment during the sampling program demonstrated that the sediment observed was largely based in little inlets of the main river channel, in backwaters and around locks, these being areas where water currents are slowest. The striking outcome of this survey was how little sediment was found. Sampling was carried out after a long period of sustained high water flow. It is recognised that high flow can act to scour sediment from the channel. In addition the removal of plants, which normally act as sediment traps (Cotton et al. 2006) will probably allow greater erosion of existing sediment deposits during the high flow periods. This was observed particularly around Water body 4 (Willy Watt Marina) but also in Water bodies 5 and 6. Brierley et al. (1989) report a similar occurrence of large scale plant removal through scouring after abnormally high flows during the winter of 1976/77.

Without previous knowledge of sediment conditions before the wet autumn and winter of 2012/13 it is impossible to assess the extent of this sediment scouring in the river channel of the Nene. However, there is a strong indication, based on the survey undertaken, that these large and prolonged flooding events can largely reset the sediment system by flushing. Trimmer et al. (2012) suggest that major flooding events can change the sediment structure and distribution in the river channel and this can have major impacts on the residence times of nutrients within the catchment by (i) depositing sediment back on the floodplains and (ii) washing sediment out towards the coast. If extensive sediment flushing of the river channel is linked to periods of extended high water flow (e.g. one in 50 year events), then calculations for the accumulation of sediment in the river channel can be started at the end of these events. The CDP model and literature calculations give a potential range of sediment erosion/accumulation for each water body covering baseline to typical catchment erosion rates under typical yearly weather patterns. Thus, depending on the weather in any one year somewhere between 1000–10000 t yr^-1 of sediment could be expected to depart water body 6. The CDP model suggests, based on morphology of the catchment that sediment erosion is greatest in water body 5 followed by significant deposition in water body 6. With phosphate being attached to the sediment, phosphate deposition and erosion in the channel will typically follow sediment transport patterns.

Implications of results for river management

The results of this work demonstrate that the gradual, long-term reduction of SRP concentrations in the River Nene will only be achieved by balancing a range of management strategies. The central points from this work to be considered when developing these strategies are:

• After the recent floods (Sept 2012 to Jan 2013) only a small quantity of sediment remained in the main channel of the Nene.
• Undertaking major de-silting operations would seem inappropriate at the present time (summer 2013) because there is insufficient silt for this to have any substantive impact. Targeted de-silting could disrupt the remaining aquatic plants which require silt for growth. Our analyses showed that whilst there are large concentrations of OEP in river sediment, this sediment still acts as an SRP sink, and will continue to take up SRP from the river water.
• The remaining silt is the substrate for macrophyte growth and these plants will also absorb some of the SRP from river water, in addition to bio-available OEP in the river sediment. Harvesting these plants would remove some of the phosphate (contained within the biomass) but also may have detrimental effects on biodiversity (described in Sediment phosphorus chemistry and Sediment dynamics within the river system).
• There was a correlation between dissolved SRP and boron in the water bodies of the Nene which suggests that sewage treatment works have an influence on river SRP concentrations.
• Any further decrease of SRP from sewage works will need to be undertaken in recognition of the EPC₀ of sediments. To a large degree the EPC₀ is related to the geochemical properties of the sediment as determined by geology (e.g. clay type, oxides concentration) and sediment architecture as defined by depositional environment and is likely to be in similar to those determined in this work. Decreasing river water SRP concentrations below the EPC₀ by STW treatment will potentially allow SRP to desorb from the sediment initially, so there may be a potential time-lag before the benefits of STW treatment are seen as the system moves towards a new equilibrium.

Further work

There are a number of aspects where further understanding of the interactions between sediment and SRP in the River Nene may require further research to improve the evidence-base where management interventions aim to improve ecological status. These are as follows:

• Now that river conditions have returned to more typical flow conditions (June 2013), it would be helpful to quantify the importance of STW source in contributing to SRP in river and sediment. It may be possible to do so using the strong relationship between SRP and boron concentrations given that data from surveys by the British Geological Survey show that across the wider catchment there are few bedrock sources of boron. Confirming the dominant source of SRP (is it STW derived) would be a major stepping stone to developing management plans to deliver improved ‘Ecological Status’.
• The major flushing of sediment that appears to occur during periods of flooding and could be a fundamental process that removes sediment associated P from the upper six water bodies of the Nene. An improved understanding of both the long-term (decadal) and seasonally-related annual cycle of erosion, deposition, storage and transport of sediment within the river channel would provide fundamental knowledge related to the outcomes of management interventions. In particular understanding the frequency of high river flow conditions required for these natural flushing events to occur would be beneficial.
• The extent to which the results of this study are representative of the long term state of the river Nene system following the recent wet winter remains unclear. Future modelling efforts could focus on simulating sediment transport for the period leading up to and including the recent wet winter of 2012/2013 for comparison to the existing modelled periods to identify whether the recent winter was likely to have resulted in exceptional levels of sediment transport, or whether the model would predict that the hypothesised ‘flushing events’ occur regularly/frequently.
• The efficiency of bedload and suspended load sediment transport are governed by typical values from the literature. Sediment transport estimations are also highly sensitive to the imposed grain size distribution. Better parameterisation of sediment transport in the river Nene could be achieved through monitoring the distribution of turbidity and water discharge at stations along the catchment (e.g. water body outlets) in order to quantify the spatio-temporal distribution of suspended sediment load which could be used to calibrate the model.
• The small, natural erosion rates across the catchment (as determined by the CPD model) suggest that we require a better understanding of the magnitude of human- influenced point sources of sediment input (e.g. land drains). In addition, understanding the proportion of sediment derived from point sources that are subsequently stored in the channel or lost (transported) from the system could improve river silt management. Comprehensive measurements on P speciation (particulate, dissolved) and fluxes of these drainage inputs would be of great value in understanding their importance for P dynamics (EPC₀) of this sediment which subsequently enters the main channel, with implications for SRP concentrations in the main channel.
• The Representative Soil Sampling Scheme (RSSS) of England and Wales (Baxter et al. 2006) showed that since 1971, a broad decrease in total P in agricultural soils has taken place, especially in the east of England. This would suggest that in future soil eroded into the river will probably contain less total P than in recent decades. However, identification of the volume and P dynamics of possible legacy P sources could be important. Although this study encompassed the main channel of the Nene, it did examine sediment depths and P dynamics in the headwaters/backwaters, where boat navigation was not practicable. In a complex, bifurcating river system such as the Nene, backwater areas are substantial, typically with slower water flows and longer sediment residence times. Phosphorus stored in sediment in drainage ditches throughout the catchment may also provide potential legacy issues. The influence of buffer strips may also need to be investigated to assess whether greater SRP is generated in these and later lost to the rivers through drains (Roberts et al. 2012; Stutter et al. 2009).
• Any further changes in the management of PO₄-P from STW should be monitored in relation to the EPC₀ so that sediment does not become a source of PO₄-P to the water in the future.
• Resolving many of these issues could be achieved by undertaking a more detailed study of sediment input, transport and SRP interactions in a representative sub-catchment or water body where significant STW inputs occur.