General enquiries on this form should be made to:Defra, Science Directorate, Management Support and Finance Team,Telephone No. 020 7238 1612E-mail: research.competitions@defra.gsi.gov.uk

SID 5 Research Project Final Report

SID 5 (Rev. 3/06) Page 1 of 28

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code WQ0202

2. Project title

Piloting a common framework for targeting and assessing the efficacy of User Manual sediment mitigation options     

3. Contractororganisation(s)

ADAS UK Ltd.                         

54. Total Defra project costs £ 35,500(agreed fixed price)

5. Project: start date................ 15 November 2008

end date................. 31 March 2009

SID 5 (Rev. 3/06) Page 2 of 28

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They

should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the

intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

Sediment pressures and impacts mean that sediment mitigation represents a priority management issue in many parts of England and Wales. Effective sediment management requires reliable information on the principal sources of the problem and on the efficacy of the mitigation options being deployed. Sediment source identification has traditionally been difficult, using conventional indirect procedures, due to logistical constraints, problems of representativeness and the costs involved. However, sediment source fingerprinting techniques have proved to offer a reliable direct alternative. The fingerprinting approach avoids many of the logistical problems and uncertainties associated with conventional sediment sourcing procedures by providing a reliable means of linking sediment directly back to its sources. Although to date, sediment fingerprinting has been used to provide information on generic sediment sources (e.g. grassland, arable land, channel banks) this pilot research project sought to refine existing procedures to improve the resolution of the sediment source evidence base. As an example of the refined sourcing procedure, high resolution sediment sources were investigated during this pilot project in a predominantly grassland catchment in Cumbria, England. More specifically, the proposed work sought to apportion sediment loss from grassland between poached gateways, poached cattle tracks and wider areas of trampling and degradation across pasture fields. The proposed work was designed as a complement to similar high resolution sediment sourcing currently being undertaken in an arable landscape in Norfolk, England, by the same contractor. The results for the Biglands Bog ECSFDI priority catchment suggested that over the study period, respective catchment scale sediment contributions from wider areas of hoofing damage, poached cattle tracks and poached gateways within pasture fields were 46±1%, 28±1% and 1±1%.Whilst high resolution sediment source apportionment is needed to target mitigation options, including those offered in the Capital Grant Scheme of the England Catchment Sensitive Farming Delivery Initiative (ECSFDI), an improved evidence base is also required in relation to the efficacy of available sediment mitigation options, including those listed in the Defra User Manual. Accordingly, the sourcing and tracing framework used to provide high resolution sediment source data was also deployed in a pilot study to assemble information on the efficacy of two sediment mitigation options: 6 m riparian buffer strips and channel bank fencing.A field experiment was undertaken at Repton, Derbyshire in order to assess the efficacy of a 6 m riparian buffer feature for controlling sediment delivery from arable land to neighbouring watercourses. The experiment was based on labelling two adjacent arable fields with dual signature tracers (magnetic and fluorescent) and inserting high strength magnets in neighbouring watercourses for trapping tracer grains breaching the buffer. Over the duration of the study (December 2008 – March 2009, inclusive) the efficacy of the 6 m riparian buffer at the experimental site was 100%. Magnets installed at intermediate locations between the tracer injection zones and the watercourse, including at the upslope margin of the buffer, demonstrated that soil mobilisation and delivery had occurred. These results should, however, be interpreted in the context of the extremely mature and tussocky grass characterising the buffer at the field trial site. The study to assess the efficacy of river bank fencing as a sediment mitigation option encompassed the Rivers Camel, Fal, Lynher, Plym, Tamar and Tavy in the SW of England. Work focused on assessing the efficacy of bank fencing for reducing bank erosion sediment pressures on salmonid spawning gravels. Respective sediment inputs to salmonid spawning gravels from eroding channel banks during the pre and post remediation study periods were computed at 97±1% vs 69±1%, 94±1% vs 91±1%, 12±1% vs 10±1%, 92±1% vs 34±1%, 31±1% vs 16±1% and 90±1% vs 66±1%.

SID 5 (Rev. 3/06) Page 3 of 28

Sediment source apportionment estimates for only the Rivers Fal and Plym were statistically significant between pre and post remediation surveys. The outputs on high resolution sediment sourcing and assessing sediment mitigation method efficacy provided proof of concept in relation to the utility of the proposed framework for providing urgently required sediment policy support.

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with

details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer).

Scientific objectives

To pilot the application of a common framework for targeting and assessing the efficacy of Defra User Manual sediment mitigation methods by:1. Piloting the application of a novel high resolution sediment sourcing procedure in a grassland dominated

agricultural catchment2. Piloting the use of the same common framework to assess the efficacy of 6 m riaprian buffer strips as a

sediment mitigation option3. Piloting the use of the same common framework to assess the efficacy of channel bank fencing as a

sediment mitigation option

Methodology

Establishing catchment scale sediment sourcesReliable information on catchment suspended sediment sources is required for a variety of reasons. For example,

such information is an essential prerequisite for assisting the design and implementation of targeted abatement strategies for controlling sediment and associated diffuse pollution problems (United States Environmental Protection Agency, 1999; Collins et al., 2001). Equally, improved datasets on sediment sources are needed to assist the interpretation of catchment suspended sediment budgets and downstream water quality response over time to management decisions (Dedkov and Moszherin, 1992; Reid and Dunne, 1996; Walling, 1999; Walling et al., 2001). It is therefore important to document catchment sediment sources.

Existing approaches to assembling information on catchment sediment sources comprise two key categories (Collins and Walling, 2004). The indirect approach to sediment source assessment is founded on the use of a number of techniques to measure sediment mobilisation in situ. But, on account of being developed to assess soil erosion rather than sediment sources per se, these methods take no explicit account of the substantial uncertainties in linking potential catchment sediment sources to the river channel. Areas of significant erosion will not represent sediment sources, unless there is clear connectivity with watercourses thereby permitting eroding areas to contribute to downstream sediment fluxes. Consequently, sediment sources can only be inferred on the basis of erosion data, unless the linkages between erosion, sediment transport, deposition and sediment flux can be readily quantified. Given the sediment delivery problem (Walling, 1983) and the uncertainties associated with sediment transfers from land to water, information on erosion must be interpreted in conjunction with an understanding of the remaining components of the sediment delivery system for the purpose of providing meaningful data on sediment provenance.

Indirect assessment of sediment sources can be undertaken using a range of techniques. For example, numerous studies have employed surveying based on profilometers (McCool et al., 1981; Shakesby, 1993), erosion pins (Haigh, 1977; Bull et al., 1995; Lawler et al., 1997; Couper et al., 2002), cross-profiling (Steegen et al., 2000; Springer et al., 2001) and GPS (Malet et al., 2002). Alternatively, both terrestrial and aerial photogrammetry has been used to monitor a range of sediment sources including eroding channel banks (Barker et al., 1997) and gullying (Nachtergaele and Poesen, 1999). In other cases, either bounded (Thomas et al., 1981; Vacca et al., 2000) or unbounded (Mutchler et al., 1988; Evans, 1995; Megahan et al., 2001) erosion plots have been deployed to obtain data. Geomorphological mapping has also provided an indirect means of elucidating sediment sources (Boardman, 1990; Lao and Coote, 1993; Hasholt and Hansen, 1993). It is important to note, however, that the deployment of these traditional methods for documenting catchment suspended sediment sources is

SID 5 (Rev. 3/06) Page 4 of 28

frequently constrained by a number of important problems, including the representativeness of data collection, logistical considerations and the costs involved (Loughran and Campbell, 1995; Collins and Walling, 2004).

Due to such problems, the fingerprinting approach has attracted increasing attention as a reliable alternative direct means of assembling information on catchment sediment sources. Sediment source fingerprinting is founded upon the link between the geochemical properties of suspended sediment and those if its sources. Assuming potential sediment sources can be readily distinguished on the basis of their constituent properties or ‘fingerprints’, the provenance of the sediment can be established using a comparison of its properties with those of the individual potential sources. Conventional sediment source fingerprinting was used as the basis for the novel framework piloted during this research project.

The discrimination of individual potential sediment sources using the fingerprinting approach has traditionally involved a wide range of fingerprint properties (Walling and Collins, 2000; Collins and Walling, 2002, 2004). The choice of properties has, to some extent, typically reflected access to the necessary laboratory analytical equipment as well as previous experience. Some studies have used mineral-magnetism to identify sediment sources on account that mineral-magnetic measurements are simple, cheap, rapid and non-destructive (Walling et al., 1979; Caitcheon, 1993; Slattery et al., 1995; Walden et al., 1997; Caitchen, 1998; Lees, 1999). Mineral-magnetic measurements provide a basis for discriminating catchment sediment sources because the non-directional magnetic properties of individual sources such as topsoil and subsoil can differ due to their inherent iron mineralogy and granulometry. Alternatively, some source fingerprinting investigations have used mineralogy or colour as a means of distinguishing potential sediment sources in catchments with heterogeneous geology and pedology (Wall and Wilding, 1976; Wall et al., 1978; Wood, 1978; Grimshaw and Lewin, 1980; Garrad and Hey, 1989; Woodward et al., 1992; Peart, 1993). In other cases, sediment geochemistry (Lewin and Wolfenden, 1978; Jones et al., 1991), environmental radionuclides (Walling and Woodward, 1992; He and Owens, 1995; Wallbrink et al., 1998), organic constituents (Peck, 1973; Brown, 1985; Hasholt, 1988; Oldfield and Clark, 1990; Peart, 1995), stable isotopic properties (Salomans, 1975; Douglas et al., 1995) or particle size measurements (Fenn and Gomez, 1989; Stone and Saunderson, 1992; Kurashige and Fusejima, 1997; Hillier, 2001) have been used to discriminate individual sediment sources.

Due to the frequent need to distinguish several potential sediment sources, it is now widely accepted that the quest for a single diagnostic property is inappropriate on account of the problem of spurious source-sediment matches (Collins and Walling, 2002). In consequence, most recent source fingerprinting studies have used so-called ‘composite fingerprints’ comprising a range of different diagnostic properties (Collins and Walling 2002, 2004). Composite fingerprints comprise individual properties influenced by differing environmental controls and which thereby improve source discrimination by affording a substantial degree of independence. Such fingerprints can represent several diagnostic properties from either a particular property subset e.g. several radiometric (He and Owens, 1995), mineral-magnetic (Oldfield and Clark, 1990) or geochemical (Collins and Walling, 2002) properties, or a combination of geochemical, radiometric and organic constituents (Walling et al., 1993; Collins et al., 2001; Collins and Walling, 2002). In order to satisfy dimensionality, the number of fingerprint properties should exceed the number of potential sediment sources being discriminated (Foster and Lees, 2000; Collins and Walling, 2004).

Sediment source fingerprinting assumes that the selected fingerprint properties are readily transported and deposited in association with suspended sediment and that selective erosion and sediment delivery processes do not transform the properties (via enrichment, depletion, dilution) beyond what can be corrected for using appropriate procedures. Composite fingerprints should be identified using statistical verification (Collins et al., 1997a, 2000; Collins and Walling, 2002). Many investigations using the fingerprinting approach have used a simple qualitative comparison between the fingerprint properties of different potential sources and sediment samples as a means of elucidating sediment provenance (Peart, 1993; Walling and Kane, 1984; Walling and Amos 1999). But, in order to provide more useful quantitative information on sediment contributions from individual sources, composite fingerprints are now generally used in conjunction with a multivariate numerical mixing model (Walling et al., 1993; Collins et al., 1997a, 2001; Wallbrink et al., 2003; Krause et al., 2003; Motha et al., 2004). Sediment mixing models can be based on linear programming (Yu and Oldfield, 1989, 1993; Caitcheon, 1993, 1998) or optimisation algorithms (Collins et al., 1997a; Walling et al., 1999; Owens et al., 2000; Walling, 2005).

Application of the sediment fingerprinting approach to document catchment suspended sediment sources necessitates collection of representative samples of individual potential sediment sources. The latter can be defined in a variety of ways. In some investigations, especially those in large-scale river drainage basins, it has proved most meaningful to investigate the spatial provenance of suspended sediment sources, defined in terms of individual tributary sub-catchments (Collins et al., 1996; Walling et al., 1999; Collins et al., 2009) or discrete geological zones (Collins et al., 1998; Walling et al., 1999; Owens et al., 2000; Bottrill et al., 2000). In smaller catchments, it is commonly more appropriate to characterise sediment provenance in terms of individual source types comprising either surface and subsurface categories (Peart and Walling, 1986, 1988) or surface soils supporting different land use and eroding channel banks (Collins et al., 1997b; Walling et al., 1999; Collins et al., 2000; Russell et al., 2001; Krause et al., 2003; Motha et al., 2004; Walling and Collins, 2005; Walling, 2005; Collins et al., 2009). Sediment source fingerprinting affords a convenient basis for investigating spatial provenance and source type in an integrated manner (Walling and Woodward, 1995; Collins et al., 1997b, 2009). As well as documenting contemporary suspended sediment sources, the fingerprinting approach provides a unique means of reconstructing longer-term sediment provenance and thus for examining linkages between soil erosion patterns and land use change (Collins et al., 1997c; Owens et al., 1999) or the occurrence of extreme flood events (Collins et al., 1997d). The approach has recently been used to examine the contribution of channel bed sediment remobilisation to suspended sediment flux at the outlets of lowland groundwater-fed catchments in the UK (Collins and Walling, 2006).

In tandem with the adoption of statistical and numerical data processing techniques for sediment fingerprinting, other important developments are associated with the use of various corrections and weightings during sediment source ascription. Since the properties of soil and sediment samples are strongly controlled by particle size composition and organic matter content (Horowitz, 1991), it is necessary to correct for differences in these characteristics. The selectivity of sediment delivery processes means that sediment samples are typically enriched in fines and organic matter content relative to the individual contributing source areas of the catchment. Approaches to correct for contrasts in particle size and organic matter content have varied in complexity. The most basic approach has been to restrict laboratory analyses to only the <63 µm (<0.063 mm) fraction of source and sediment samples, thereby ensuring a focus upon the dominant size class of suspended sediment

SID 5 (Rev. 3/06) Page 5 of 28

(Motha et al., 2002). Given that the composition of the <63 µm fraction tends to differ between samples, fingerprint property concentrations measured on this size fraction have been corrected using specific surface area information (Collins et al., 1997a, 1998; Gruszowski et al., 2003). Specific surface area provides a useful surrogate for particle size and is governed by the entire composition of a given size fraction. More complex approaches to correcting for particle size composition have been based on detailed information on the precise relationship between grain size composition and the concentrations of individual fingerprint properties (He and Walling, 1996; Russell et al., 2001). These approaches avoid the assumption that there is a consistent linear relationship between concentration and particle size composition for all properties. But, the identification of correction factors for individual properties requires substantial investment in laboratory resources. Alternatively, some researchers have adjusted the fingerprint property concentration data for source materials using information on the grain size characteristics of sediment and the concentration information for different size fractions of the source samples. Under these circumstances, source material fingerprint property concentrations are adjusted to reflect the same particle size composition as that measured for sediment (Slattery et al., 1995; Motha et al., 2002). Less attention has been directed towards correcting for contrasts in the organic matter content of samples. Organic matter content adjustments typically rely upon a simple ratio between the organic carbon content of source material and sediment samples (Collins et al., 1997a, 1998), or the adjustment of source material fingerprint property concentrations to reflect a similar organic matter content to that measured for sediment (Motha et al., 2002). Correcting for organic matter content frequently reduces the errors associated with numerical sediment source ascription (e.g. Walling et al., 2003), although the risk of double correction, in tandem with the use of a particle size correction factor should be carefully explored during each fingerprinting study.

Due to the need to take explicit account of the natural variability of source material properties, uncertainty testing is now incorporated into the quantitative source apportionment procedure, using a selection of Bayesian statistics and Monte Carlo routines (Rowan et al., 2000; Small et al., 2002; Motha et al., 2004; Douglas et al., 2003; Collins and Walling, 2007 a, b; Collins et al., 2009).

Piloting the application of a novel high resolution sediment sourcing procedure in a grassland dominated agricultural catchment

The novel high resolution sediment sourcing procedure was tested during an investigation of the primary sources of the sediment problem reported in the Biglands Bog ECSFDI priority catchment, Cumbria, northern western England (Figure 1). This study catchment is part of ECSFDI priority catchment number 19 (River Waver and Biglands Bog). More specifically, the pilot test aimed to apportion the primary sources of the sediment responsible for the siltation of the Biglands Bog SSSI. Biglands Bog SSSI is located 5 km north of Wigton, Cumbria and the site consists of acidic mire, bog and a eutrophic rich fen. The novel sediment sourcing framework comprised a combination of conventional fingerprinting and a dual signature particle tracking method. The former was deployed to apportion sediment collected from Biglands Bog SSSI between generic catchment sediment sources, whereas the latter was used to elucidate sediment loss from poached gateways, poached cattle tracks and wider areas of hoofing damaged across grass fields.

Source material and sediment samplingCollection of representative source material samples in the two sub-catchments (Bampton Beck and Aikton Beck)

comprising the Biglands Bog study area was completed in December 2008 and was stratified to encompass five primary potential sediment sources. These source types comprised pasture topsoils, cultivated topsoils, damaged road verges, channel bank/subsurface sources (including ditches, gullies and incised tracks cutting into the subsoil and regolith) and the STW located in the study area. Samples retrieved from agricultural topsoils and damaged road verges comprised surface scrapes (0-2 cm) susceptible to mobilisation by water erosion and subsequent delivery towards the river channel system. Channel bank/subsurface source sampling targeted actively eroding bank sections and gully systems or tracks. All samples were retrieved using a non-metallic trowel, which was repeatedly cleaned to avoid inter-sample contamination. Pedological variation was taken into account as a means of helping to ensure the representativeness of the corresponding fingerprint property datasets. Each source material sample comprised a composite of smaller scrapes collected at an individual site in order to increase the representativeness of the individual samples and of the over-arching sampling strategy. Channel bank samples comprised material from the full vertical extent of the bank profile. The sampling of material originating from the STW was undertaken in the channel system immediately adjacent to the outfall in order to provide general characterisation of sediment released from this point source discharge. Whereas previous source apportionment studies have sampled final effluent from STWs (Collins et al., 2009), this approach was not deemed appropriate, since the STW in the Biglands Bog study catchment comprised a series of reed beds for trapping particulates and cleansing final discharge. It was assumed that sampling adjacent to the outfall would characterise both final effluent and by-pass flow. A summary of the source material sampling exercise is provided in Table 1.

Table 1: Summary information on the source material samplingSub-catchment Source type

Pasture topsoils

Cultivated topsoils

Damaged road verges

Channel banks/subsurface sources

STWs

Bampton Beck 15 15 15 15 1Aikton Beck 8 8 8 8 -

SID 5 (Rev. 3/06) Page 6 of 28

In order to examine the relative significance of pasture or cultivated topsoils, damaged road verges, channel banks/subsurface sources and the STW as sediment sources over recent time, surface sediment samples were collected from the Biglands Bog SSSI at the outlet of the study area. This sampling exercise targeted sediment deposits at the outfall of each tributary into the SSSI. Comparison of catchment source material and Biglands Bog sediment samples permitted examination of sediment provenance during flood events representative of recent time (ca. 2-5 years). Table 2 summarises the surface sediment sampling exercise.

Figure 1: The Biglands Bog study area

SID 5 (Rev. 3/06) Page 7 of 28

Table 2: Summary information on the surface sediment samplingSub-catchment No. of surface sediment samplesBampton Beck 8Aikton Beck 8

Labelling high resolution sources in grass fields and capturing mobilised tracer grainsThe deployment of the dual signature tracking method necessitated the selection of appropriate monitoring sites.

Two representative sites, one in Aikton Beck at 54°52’11N, 003°07’31W or NGR NY 27879 53351 and one in Bampton Beck at 54°52’57N, 003°07’59W or NGR NY 27416 54817 were selected on the basis of field walking with the Catchment Sensitive Farming Officer (CSFO). These sites were judged to be representative of the high sediment mobilisation risk configurations in local grass fields comprising clusters of poached gateways or cattle tracks and wider hoofing damage, with obvious connectivity to neighbouring watercourses. The former sites was selected to be characteristic of those grass fields exhibiting less severe poaching problems, whereas the latter site was used to represent fields with more severe hoofing damage. The dual signature tracking method is based on labelling target areas with synthetic magnetic tracer grains (White, 1998; Black et al., 2007). Target areas are labelled with unique fluorescent signatures in order to assist the apportionment of inputs to neighbouring river channels (Figures 2-3). The synthetic tracers were carefully manufactured to resemble the typical particle size composition and density of the target areas identified in the grass fields (see Table 3). Dual signature tracers were applied using a road salt spreader to ensure an even distribution of tracer grains over deployment areas. Tracer seeding was undertaken on days with calm weather conditions to avoid significant aeolian redistribution. A yellow tracer was used to label areas of wider hoofing damage, whereas respective pink and blue tracers were deployed to seed poached cattle tracks and gateways.

In order to capture the tracer grains mobilised from the target areas in the grass fields by water erosion, 11,000 gauss bar magnets were placed in adjacent river channels (Figure 4). The magnets were protected by thin plastic sheaths and end caps to prevent magnetic material becoming permanently fixed to the actual magnet surfaces. During the recovery of the magnet samples in situ, the sheaths were carefully removed from the magnetic bars and placed into sample bags. A large plastic tray was used to capture any tracer grains dislodged during this recovery process. New sheaths were placed over all magnets prior to departure from the sampling sites.

Table 3: The typical grain size composition and density of the target areas used for the high resolution particle tracking work in the Biglands Bog study areaSub-catchment Density (kg m3) Size range () % sand % silt %clayBampton Beck 2129 -1.5 3.5 11.3 83.2 5.5Aikton Beck 2077 -1.1 3.5 13.6 81.5 4.9

Figure 2: An area of wider hoofing damage freshly seeded with a dual signature tracer

Figure 3: A poached gateway freshly seeded with a dual signature tracer

Figure 4: Bar magnet showing tracer grains captured from neighbouring seeded area

Laboratory analysesAll soil source material and Biglands Bog surface sediment samples were returned to the laboratory, oven-dried at

40oC, manually disaggregated using a pestle and mortar and dry sieved using a 63 µm mesh. Sieving times were standardised as much as possible. Screening facilitated comparison of source material and sediment samples. The STW by-pass flow sample was de-watered using settling, freeze-dried and sieved through a 63 µm mesh.

Based on the findings reported by Collins and Walling (2002), a range of potential fingerprint properties was selected for analysis in order to support the identification of powerful composite fingerprints responding to differing environmental controls. Available laboratory facilities were, however, an important consideration. A total of 47 properties were included in the analytical programme. Concentrations of Al, As, Ba, Bi, Cd, Ce, Co, Cr, Cs, Cu, Dy, Er, Eu, Fe, Ga, Gd, Ge, Hf, Ho, In, K, La, Li, Mg, Mn, Mo, Na, Nd, Ni, Pb, Pd, Pr, Rb, Sb, Sc, Sm, Sn, Sr, Tb, Ti, Tl, U, V, Y, Yb, Zn and Zr were determined using ICP-MS, post direct digestion with nitric and hydrochloric acid (Allen, 1989). The absolute grain size composition of all samples was measured using a Micromeritics laser diffraction granulometer following pre-treatment with hydrogen peroxide to remove organics, chemical dispersion with sodium hexametaphosphate and exposure to ultrasound. Particle size analysis assumed spherical particles in the estimation of specific surface area. C and N content was measured directly by pyrolysis using an automatic C/N analyser (Walling and Collins, 2000).

Upon return to the laboratory, all magnet samples were washed through a 500 µm sieve to remove any large native magnetic or vegetation debris. The <500 µm fraction was repeatedly exposed to an 11,000 gauss magnet until no further material was retrieved. The efficiency of the tracer grain separation procedure was verified using three replicate samples spiked with a known mass (1 g, 0.1 g and 0.01 g) of synthetic tracer. This test suggested an overall tracer grain recovery efficiency of 97.3±1.9%. High resolution microscope analysis was used to distinguish the tracers with unique fluorescence attached to the individual magnets. During the microscope analysis, a sub-sample of the magnetic grains recovered from the magnets was placed on a microscope slide with a small volume of distilled water. A minimum of 300 discrete particles were selected from each sample, dried and weighed, in order to assess the relative proportions of the individual fluorescent tracers applied to the target areas in the grass fields. All microscope analysis was undertaken using a Zeiss fluorescent microscope fitted with an excitation filter set.

Discrimination of generic sediment sourcesThe two-stage statistical procedure proposed by Collins et al. (1997a) was employed to test the ability of the

fingerprint properties to discriminate between the individual source types. Stage one was based on the use of the Kruskal-Wallis H-test to examine the ability of individual constituents to distinguish pasture or cultivated topsoils, damaged road verges, channel bank/subsurface and the STW source samples in an unequivocal manner. Deployment of the Kruskal-Wallis H-test is founded on the logical assumption that the selection of robust composite fingerprints requires confirmation of the

power of individual constituents to discriminate the source samples under scrutiny. The Kruskal-Wallis H-test is the non-parametric equivalent of analysis of variance and provides a distribution-free procedure for examining contrasts between sample sets. It has a power efficiency of ca. 95.5%, thereby rendering it suitable for testing in conjunction with relatively small sample sets (Hammond and McCullagh, 1978). Greater inter-group contrasts generate larger test statistics and where these exceed the critical value, Ho (i.e. the null hypothesis stating that measurements of the fingerprint property exhibit no significant differences between the source type categories) is rejected. The Kruskal-Wallis H-test is applied to the values of a specific property for the source material dataset as a whole. Consequently, a statistically significant output is suggestive of source inter-category contrasts, rather than confirming differences between all possible pairs of source categories (Fowler and Cohen, 1990). Stage one of the procedure provides a basis for eliminating redundant fingerprint properties.

The results of the Kruskal-Walis H-test for discriminating the source types in each sub-catchment of the Biglands Bog study area are presented in Tables 4-5. In the case of the Bampton Beck sub-catchment, a total of 46 individual properties passed the Kruskal-Wallis H-test, yielding test statistics in excess of the critical value of 9.49. Sc represented the weakest of the properties failing the first stage of the tracer selection process, yielding H- and p-values of 6.642 and 0.156, respectively (Table 4). The corresponding results for the Aikton Beck sub-catchment are illustrated in Table 5. Twelve individual properties (Cd, Cu, Hf, Mg, Pb, Pd, Rb, Sb, Sc, U, V and Zn) failed the selection process by failing to generate statistically significant test results. Of the 35 properties passing the test for the Aikton Beck sub-catchment, Al, Ce, In and La generated critical p-values of 0.000 (Table 5). All fingerprint properties passing the Kruskal-Wallis H-test survived the elimination process and entered stage two of the statistical verification.

In accordance with stage two of the procedure proposed by Collins et al. (1997a), multivariate Discriminant Function Analysis (DFA) was used to test the ability of the properties passing the Kruskal-Wallis H-test to discriminate the source material samples into the correct categories. DFA estimates discriminant function coefficients indicative of the explanatory power of fingerprint properties. A multivariate stepwise selection algorithm, based on the minimisation of Wilks’ lambda, was used to identify the optimum (i.e. smallest) combination of properties, or composite fingerprint, for discriminating the source samples collected from a given sub-catchment. During the stepwise selection procedure, properties satisfying two principal test criteria, i.e. the partial F ratio and tolerance level, are entered in order of their explanatory power. Default values were used for both the partial F ratio (1.0) and tolerance level (0.001). As a means of avoiding the preferential selection of individual properties for inclusion in the final composite fingerprint, all parameters were assigned the default inclusion level (1.0). Stepwise selection ceases when all source material samples are classified correctly, or when none of the remaining constituents available for inclusion in the composite signature improve sample discrimination.

The results of applying DFA to the source material sample datasets collected for the Biglands Bog study sub-catchments are shown in Tables 6-7. In the case of the Bampton Beck sub-catchment, a total of ten properties were selected for the optimum composite fingerprint which discriminated 100% of the source type samples correctly (Table 6). For the Aikton Beck sub-catchment, a total of six properties were included in the optimum composite fingerprint, correctly distinguishing 96.9% of the source type samples (Table 7). Each of the composite signatures presented in Tables 6-7 comprised a combination of trace and heavy metals, confirming that fingerprints including a range of different constituents are likely to afford the most powerful discrimination (Collins and Walling, 2002).

Table 4: The output from applying the Kruskal-Wallis H-test to the source material fingerprint property dataset for the Bampton Beck sub-catchment of the Biglands Bog study areaFingerprint property H-value P-valueAl 26.197 0.000As 13.176 0.010Ba 29.067 0.000Bi 40.915 0.000Cd 16.038 0.003Ce 34.852 0.000Co 21.733 0.000Cr 17.297 0.002Cs 14.938 0.005Cu 30.761 0.000Dy 21.040 0.000Er 17.836 0.001Eu 21.802 0.000Fe 28.439 0.000Ga 27.629 0.000Gd 22.092 0.000Ge 23.900 0.000Hf 28.352 0.000Ho 19.617 0.001In 41.762 0.000K 41.899 0.000La 21.349 0.000Li 34.292 0.000Mg 22.734 0.000Mn 9.576 0.048Mo 30.864 0.000Na 37.791 0.000Nd 26.385 0.000Ni 13.580 0.009Pb 22.698 0.000Pd 23.670 0.000Pr 29.036 0.000Rb 20.599 0.000Sb 30.126 0.000Sc 6.642 0.156*Sm 24.933 0.000Sn 28.011 0.000Sr 28.898 0.000Tb 21.931 0.000Ti 29.632 0.000Tl 11.862 0.018U 23.316 0.000V 10.417 0.034Y 24.992 0.000Yb 19.329 0.001Zn 29.374 0.000Zr 29.401 0.000critical value @ 95.5% confidence = 9.49; * = not statistically significant at p <0.05

Table 5: The output from applying the Kruskal-Wallis H-test to the source material fingerprint property dataset for the Aikton Beck sub-catchment of the Biglands Bog study areaFingerprint property H-value P-valueAl 17.859 0.000As 8.815 0.032Ba 10.179 0.017Bi 5.020 0.170Cd 6.682 0.083Ce 19.804 0.000Co 11.429 0.010Cr 10.162 0.017Cs 16.597 0.001Cu 5.065 0.167Dy 14.702 0.002Er 14.315 0.003Eu 14.142 0.003Fe 14.656 0.002Ga 8.784 0.032Gd 16.077 0.001Ge 10.628 0.014Hf 7.634 0.054Ho 13.869 0.003Ln 14.841 0.002K 25.060 0.000La 13.452 0.004Li 18.651 0.000Mg 6.247 0.100Mn 8.815 0.032Mo 8.560 0.036Na 8.696 0.034Nd 15.807 0.001Ni 11.142 0.011Pb 6.597 0.086Pd 5.741 0.125*Pr 17.043 0.001Rb 7.264 0.064Sb 1.503 0.682*Sc 7.443 0.059*Sm 16.545 0.001Sn 13.875 0.003Sr 11.849 0.008Tb 14.563 0.002Ti 8.977 0.030Tl 8.696 0.030U 5.099 0.165*V 3.304 0.347*Y 14.253 0.003Yb 12.571 0.006Zn 6.253 0.100*Zr 10.534 0.015critical value @ 95.5% confidence = 7.82; * = not statistically significant at p <0.05

Table 6: The optimal composite fingerprint for discriminating individual sediment source types in the Bampton Beck sub-catchment of the Biglands Bog study area

Step Fingerprint property selected

Cumulative % source type

samples classified correctly

Wilks’ lambda

% source type samples

classified correctly

Tracer discriminatory

weighting

1 Sb 47.7 0.163 47.7 1.42 Ti 52.3 0.068 46.2 1.43 Ge 70.8 0.032 38.5 1.14 Zn 76.9 0.018 41.5 1.25 Al 80.0 0.013 58.5 1.76 Pd 84.6 0.009 33.8 1.07 Y 98.5 0.003 50.8 1.58 Nd 100 0.001 44.6 1.39 Zr 100 0.001 50.8 1.510 Er 100 0.001 40.0 1.2

Table 7: The optimal composite fingerprint for discriminating individual sediment source types in the Aikton Beck sub-catchment of the Biglands bog study area

Step Fingerprint property selected

Cumulative % source type

samples classified correctly

Wilks’ lambda

% source type samples

classified correctly

Tracer discriminatory

weighting

1 Ce 56.3 0.418 56.3 1.62 Ba 81.3 0.184 43.8 1.33 Zr 87.5 0.105 50.0 1.54 Cr 84.4 0.058 34.4 1.05 Tl 93.8 0.033 37.5 1.16 Sn 96.9 0.017 62.5 1.8

Generic sediment source ascriptionA new and revised version of the multivariate mixing model described by Collins et al. (1997a) was used to apportion

generic sediment sources in the Biglands Bog study catchment. This model is founded on the assumption that the concentrations of the properties comprising the composite fingerprint, measured in sediment samples collected from the Biglands Bog SSSI, represent the product of the corresponding concentrations in the original sources and the relative inputs contributed by those sources. Potential sources are represented in the mixing model using the mean concentrations of fingerprint properties. Use of the mean concentration value to represent a particular source can be justified since the sediment collected from the catchment outlet will inevitably represent a mixture of material mobilised and delivered from numerous locations upstream. As a result, the collection of representative source material samples from a range of locations throughout the catchment and the use of these samples to derive mean fingerprint property concentrations can be assumed to be analogous to natural sediment mixing during the sediment delivery process. The use of mean fingerprint property values is therefore physically meaningful.

Two linear boundary conditions are imposed on the mixing model iterations to ensure that the relative contributions () from the individual sediment sources are non-negative (Equation 1) and that these contributions sum to unity (Equation

2):

(1)

(2)

The original mixing model (Collins et al., 1997a) optimises estimates of the relative contributions from the potential sediment sources by minimising the sum of squares of the weighted relative errors, viz.:

(3)

where: = concentration of fingerprint property in Biglands Bog surface sediment sample; = the optimised percentage

contribution from source category ; = mean concentration of fingerprint property in source category ; = particle

size correction factor for source category ; = organic matter content correction factor for source category ; = tracer

specific weighting; = number of fingerprint properties comprising the optimum composite fingerprint; = number of sediment source categories.

The revised mixing model algorithm (Collins et al., 2009) also optimises estimates of the relative contributions from the potential sediment sources by minimising the sum of squares of the weighted relative errors, but includes revised weightings, viz.:

(4)

where: = concentration of fingerprint property in Biglands Bog surface sediment sample; = the optimised percentage

contribution from source category ; = mean concentration of fingerprint property in source category ; = particle

size correction factor for source category ; = organic matter content correction factor for source category ; =

weighting representing the spatial variation of fingerprint property in source category ; = tracer discriminatory

weighting; = number of fingerprint properties comprising the optimum composite fingerprint; = number of sediment source categories.

The particle size correction factor is included in the sediment mixing model since it is widely understood that grain size exerts an important influence on element concentrations in soil and sediment samples (Horowitz and Elrick, 1987; Horowitz, 1991; Stamoulis et al., 1996; Queralt et al., 1999). In consequence, the fingerprint properties of source material and sediment samples cannot be directly compared, even after sieving, unless a correction factor is utilised. Due to particle size selectivity during sediment transportation from source to river channel, the typical sediment sample is enriched in fines compared to the corresponding samples collected to represent the individual sources. In order to calculate a particle size correction factor, specific surface area (m2 g-1) is used as a surrogate measure for grain size composition because it exerts a key control on element concentrations (Horowitz, 1991). During the application of the mixing model for generic source type apportionment, the ratio of the mean specific surface area of the Biglands Bog surface sediment samples to the corresponding mean value for each individual source type was used. Although this approach assumes a linear relationship between fingerprint property concentration and specific surface area, it provides a pragmatic means of addressing the need to take explicit account of particle size selectivity.

The mixing model algorithm also includes an organic matter content correction since the latter also influences element concentrations in soil and sediment samples. This correction is calculated in the same manner as the equivalent for particle size, but using information on organic carbon content. Because the influence of particle size and organic matter content on element concentrations can be closely related, the combined use of the correction factors was carefully examined in order to ensure that the over-correction of the source material datasets was avoided. Sensitivity tests confirmed that the combined use of the particle size and organic matter content correction factors was appropriate for all applications of the revised sediment mixing model concerned with the apportionment of contemporary sediment sources in the Biglands Bog ECSFDI priority catchment.

A weighting to reflect the spatial variation of individual tracers in each source was incorporated in the revised mixing model. This new weighting was included to ensure that the fingerprint property values for a particular source characterised by the smallest standard deviation exerted the greatest influence upon the optimised solutions. It is logical that as the standard deviation of the fingerprint property values increases, the uncertainty associated with the source ascription also increases. The weighting was calculated using the inverse of the root of the variance associated with each fingerprint property measured for each source. The spatial variation weighting provided a means of representing the compound affect of a number of sources of uncertainty, including the variance of the tracer datasets for specific sources and the differing levels of precision associated with laboratory measurements of those tracers. Sensitivity tests during previous work, demonstrated that inclusion of the spatial variation weighting in the revised sediment mixing model resulted in the average range of the relative contributions generated for each source type being 6% narrower (Collins et al., 2009).

The revised mixing model algorithm also incorporated a weighting to reflect tracer discriminatory power (Equation 4). This weighting was based on information on the discriminatory efficiency of each individual tracer included in any given composite fingerprint provided by the results of the DFA (Tables 5-6). Sensitivity analysis during previous work has supported the inclusion of a tracer specific weighting (Collins et al., 2009).

Where possible, it can be advantageous and meaningful to incorporate informative priors into numerical mass balance modelling. A review of sediment sources in the UK by Walling and Collins (2005) and more recent work (Collins et al., 2009) suggested that typical channel bank contributions rarely exceed 50%. The only two catchments of those reviewed, where more than 50% of the suspended sediment load originated from bank loss, were the Aire (55%) and the Worm Brook (55%). On this basis, the upper boundary constraint for the bank erosion/subsurface source contribution in the revised numerical mass balance sediment mixing model was set at 0.5.

The uncertainty of the optimised results obtained using the revised sediment mixing model was investigated using a Monte Carlo framework. A total of four permutations were applied during the uncertainty analysis:

random sampling based on the conventional approach of using the mean and standard deviation of each fingerprint property for each source type and sub-catchment sediment outfall to generate cumulative Normal distributions using a random number generator

random sampling based on a permutation of the conventional approach involving the use of robust statistics to generate cumulative Normal distributions for each fingerprint property for each source type and sub-catchment sediment outfall

stratified sampling of the simulated Normal distributions for each fingerprint property for each source type and sub-catchment sediment outfall based on the conventional approach

stratified sampling of the simulated Normal distributions for each fingerprint property for each source type and sub-catchment sediment outfall based on a permutation of the conventional approach involving the use of robust statistics

A non-negativity constraint was implemented during the generation of the simulated Normal distributions using either the conventional or alternative approach. The four permutations of uncertainty analysis incorporated the uncertainty associated with both source material and sediment sampling (cf. Collins et al., 2009). Nearly all previous sediment fingerprinting studies have focused exclusively upon the uncertainty of sampling the source material fingerprint property means (e.g. Collins and Walling, 2007a,b). It is clearly important to take explicit account of the corresponding uncertainty associated with sampling sediment fingerprint property concentrations. Regardless of the permutation of the uncertainty analysis, the set of linear equations pertaining to the optimum composite fingerprint for each sub-catchment was repeatedly solved 5000 times as a means of generating 95% confidence limits for the mean contributions from individual sources.

The robustness of the optimised mixing model solutions was interrogated using the relative mean error (RME) or goodness of fit, viz.:

(5)

The RME calculated for the mixing model runs associated with apportioning the sources of surface sediment samples collected from Biglands Bog at the outfall of each sub-catchment is presented in Table 8. These estimates confirmed that the revised mixing model is capable of simulating meaningful surface sediment mixtures for the Biglands Bog SSSI.

Table 8: Relative mean error (RME) associated with the optimised mixing model solutions for each sub-catchment of the Biglands Bog study area

Sub-catchment RME %1

RME %2

RME %3

RME %4

Bampton Beck 5.6 5.1 6.2 5.0Aikton Beck 17.7 12.4 12.5 12.11 = random sampling: conventional approach2 = random sampling: robust statistics3 = stratified sampling: conventional approach4 = stratified sampling: robust statistics

Sediment source apportionment data for the Biglands Bog study areaFigure 5 shows the mixing model output for generic sediment source apportionment in the Bampton Beck sub-

catchment. Use of robust statistics to scale the simulated fingerprint property distributions during uncertainty analysis clearly helped to constrain the ranges in contributions from the pasture and damaged road verge categories regardless of whether a random or stratified sampling approach was adopted. For example, relative inputs from eroding pasture topsoils were computed at 0-100% using the random and stratified approaches coupled with conventional scaling, compared with respective ranges of 52-84% and 61-84% generated using random or stratified sampling and parameter scaling using robust statistics. The opposite effect was observed for cultivated topsoils and channel banks/subsurface sources in that the use of robust statistics to scale parameters during uncertainty analysis increased the ranges in relative contributions. On the basis of the information in Figure 5 and Table 8, the mixing model output generated using stratified sampling and robust statistics parameter scaling was used to estimate the typical generic sediment source contributions from Bampton Beck (Figure 7). These inputs were estimated at 73±1% (pasture topsoils), 17±1% (cultivated topsoils), 1±1% (damaged road verges), 8±1% (channel banks and subsurface sources) and 1±1% (the STW).

The mixing model results for generic sediment source apportionment in the Aikton Beck sub-catchment of the Biglands Bog study area are presented in Figure 6. In this case, the four approaches to sampling and scaling the simulated fingerprint property distributions resulted in more similar probability density functions (pdfs). As a result, the predicted ranges in relative contributions from eroding pasture topsoils were 0-99% for each permutation. Similarly, the predicted relative inputs from damaged road verges in this sub-catchment were 0-99% for all permutations apart from the uncertainty analysis comprising stratified sampling and a robust statistics approach to parameter scaling (0-91%). On the basis of the information in Figure 6 and Table 8, the mixing model output generated using stratified sampling and a robust statistics approach to parameter scaling was used to estimate the typical generic sediment source contributions from Aikton Beck (Figure 7). These inputs were estimated at 77±1% (pasture topsoils), 1±1% (cultivated topsoils), 12±1% (damaged road verges) and 10±1% (channel banks and subsurface sources). Synthesizing the typical sediment source apportionment data for the Bampton Beck and Aikton Beck sub-catchments, the overall generic sediment source inputs to the Biglands Bog SSSI from the entire upstream catchment were computed at 75±1% (pasture topsoils), 9±1% (cultivated topsoils), 6±1% (damaged road verges), 9±1% (channel banks and subsurface sources) and 1±1% (the STW).

High resolution tracking suggested that the relative losses of tracer grains from the three high risk components of grassland fields were of the order of 79% (wider areas of hoofing damage), 20% (poached cattle tracks) and 1% (poached gateways) in the Bampton Beck sub-catchment. Tracking results for the Aikton Beck sub-catchment suggested that the relative losses from the three high risk components of grassland fields were of the order of 44% (wider areas of hoofing damage), 54% (poached cattle tracks) and 2% (poached gateways). These results were judged to be consistent with field observations in that wider areas of grassland in the Aikton Beck sub-catchment were less damaged than those corresponding areas in Bampton Beck. The results of synthesizing the average relative proportions of fluorescent tracer particles captured from the high risk pasture areas by the magnets in Bampton and Aikton Becks with the overall mean generic source fingerprinting estimates are presented in Figure 8. Over the duration of the study period, wider areas of hoofing damage across pasture fields contributed 46±1% of the total sediment delivered to Biglands Bog, compared to 28±1% from poached cattle tracks and 1±1% from poached gateways. These results should be viewed as tentative given the spatial extrapolation to sub-catchment scale on the basis of the tracking work at two target sites.

Figure 5: Ranges in the relative contributions from the individual generic sediment sources in the Bampton Beck sub-catchment computed using the four permutations of uncertainty analysis

Figure 6: Ranges in the relative contributions from the individual generic sediment sources in the Aikton Beck sub-catchment computed using the four permutations of uncertainty analysis

Figure 7: Mean relative contributions from the individual generic sediment sources in the Bampton Beck and Aikton Beck sub-catchments of the Biglands Bog study area

0

10

20

30

40

50

wider areas of hoofing damage poached cattle tracks poached gateways

High resolution grassland sediment source

% c

ontri

butio

n

Figure 8: Tentative relative contributions from each high resolution grassland sediment source to sediment pressures on the Biglands Bog SSSI

Assessing the efficacy of 6 m riparian buffer strips as a sediment mitigation optionThe CAP health check is likely to require EU Member States to revise the use of Axis 2 agri-environment measures

in order to provide a greater focus on key challenges including improved protection of aquatic habitats and biodiversity. Potential changes might include the increased use of riparian buffers to protect aquatic ecosystems. Within the context of the CAP health check, policy teams in England are considering means of simplifying and improving existing Cross Compliance requirements for soil protection (via the soil protection review) and of targeting Environmental Stewardship options to help retain the benefits previously provided by soil, nutrient and crop protection plans. Such efforts are addressing the challenge from the EU to introduce into GAEC (Good Agricultural and Environmental Condition) rules requirements that retain the environmental benefits of set-aside and which address priority water management issues. There is therefore a need to continue expanding the evidence base for using riparian buffers to mitigate some of the environmental benefits lost due to the abolition of set-aside from January 1st 2009. An improved evidence base is required to determine whether the recommendations of Sir Don Curry’s ‘High Level Group’, concerning making available an unspecified % of land for environmental benefit under cross compliance, is feasible.

At present, land managers in England claiming Single Payment Scheme (SPS) must adhere to cross compliance requirements to maintain land in Good Agricultural and Environmental Condition (GAEC). To comply with GAEC 14 for watercourses, land managers must not: cultivate or apply fertilisers or pesticides to land within 1 m of the top of the bank of a watercourse or field ditch and, in

addition, must: take all reasonable steps to maintain a green cover on land within 1 m of the top of the bank of a watercourse or field ditch.

The 1 m buffer is designed to protect sensitive field boundaries and their associated habitats. But, it is unlikely that the current cross compliance 1 m wide buffer strip is able to provide any significant filtering benefit for pollutants. Equally, the potential habitat within a 1 m strip may be of limited value due to the risks of herbicide spray drift impacts.

Given the above context, this pilot project used the proposed sourcing and tracing framework to investigate the efficacy of 6 m riparian buffers for controlling sediment loss from arable land to neighbouring watercourses. Accordingly, a field trial site (see Figure 9) was selected at Repton, Derbyshire, England (52°50’20N, 001°31’59W or SK 31595 27078) following visits to a number of farms. The experimental site in Repton was selected to be representative of arable land on medium and calcareous soils (sandy clay loams, clay loams, silty clay loams) and with moderate slopes of 3-7 degrees (cf. Defra 2005 soil erodibility typology). Medium and calcareous soils currently account for ca. 71% of arable land on moderate slopes across England (see Figure 10).

Figure 9: Arable land at the experimental site at Repton, Derbyshire

Figure 10: Soil classes for arable land on moderate slopes (3-7 degrees) across England

Setting up the riparian buffer experimental site at Repton, DerbyshireFollowing the collection of local soil samples for matching grain size distribution and density characteristics, the dual

signature tracer was applied to arable land at the experimental site in Repton, using a road salt spreader (Figure 1). The manufactured tracer was spread evenly across the slope of two neighbouring arable fields, both with 6 m riparian buffers at the bottom of the slope (Figures 12 and 13).

Figure 11: Labelling the arable land at Repton, Derbyshire with a dual signature tracer

Figure 12: Photograph of the dual signature tracer running across the arable land at Repton, Derbyshire (with the 6 m riparian buffer strip evident at the bottom of the slope)

Figure 13: The 6 m riparian buffer at the experimental site at Repton, Derbyshire

Following the application of the dual signature tracer across the arable fields, a total of 16 11,000 gauss bar magnets were inserted in the neighbouring watercourses at the base of the slope and at intermediate locations between the tracer injection zones and the outer (i.e. watercourse) margin of the buffer features. This permitted assessment of the transport pathways of the tracer mobilised from the injection zones. Buffer performance was assessed using the difference between the

mass of tracer grains captured by the high strength magnets in the watercourse and those at the field margin of the buffer features. The deployment of the tracers at the buffer experimental site covered the period December 2008 – March 2009.

The efficacy of the 6 m riparian buffer at the experimental site at Repton, DerbyshireThe high strength magnets at the experimental site at Repton, Derbyshire were visited monthly between the period

December 2008 and March 2009, inclusive. The plastic sheaths covering the magnets were all recovered and replaced on each visit. Throughout this study period, no dual signature tracer grains were collected by the magnets in the watercourses neighbouring the fields with the injection zones. This suggested that the 6 m riparian buffer at the experimental site was characterised by a sediment trapping efficacy of 100%. Those magnets inserted at intermediate locations including those at the upslope margin of the buffer feature, did trap tracer grains, demonstrating that soil mobilisation and delivery did occur during the monitoring period. It is important to note that the 6 m riparian buffer at the experimental site at Repton Derbyshire was characterised by extremely mature and tall tussocky grass (see Figure 13) which could be expected to provide an extremely efficient buffer for trapping mobilised sediment (cf. Schmitt et al., 1999). The findings of the field trial were consistent with those reported in the recent ADAS review of buffer performance (Figure 14) which suggested an efficacy range of 58-95% for 6 m buffers. Since the riparian buffer at the study site was characterised by extremely mature and dense vegetation cover, the results should be taken as representative only for similar buffers. Buffer performance is highly site-specific (Dorioz et al., 2006).

0

20

40

60

80

100

0 10 20 30 40 50 60 70

Buffer Width (m)

Sedi

men

t rem

oval

effi

cien

cy (%

)

Figure 14: Riparian buffer performance for sediment on the basis of width (vertical lines represent 1 m, 3 m and 6 m options)

Assessing the efficacy of channel bank fencing as a sediment mitigation optionThe siltation of spawning gravels has been increasingly identified as a key factor contributing to the declining success

of salmonid fisheries throughout England and Wales (Olsson and Persson, 1988). Spawning gravel siltation reduces the survival-to-emergence of salmonids, by virtue of its detrimental impact on the nests, or redds, which play a critical role in salmonid reproduction. The presence of elevated quantities of fine matrix sediment adversely affects two critical properties of spawning gravels, namely permeability and porosity. Permeability controls the rate of supply of dissolved oxygen and the rate of removal of carbon dioxide and metabolic waste. Egg-to-hatching success is strongly influenced by these fluxes, which decline sharply with the excessive accumulation of fine interstitial sediment (Iwamoto et al., 1978; Turnpenny and Williams, 1980). Porosity, exerts an important influence on the intra-gravel movement and eventual emergence of the newly hatched fry or alevins, because the excessive accumulation of fines can block the interstitial pathways, smother the surface of the gravel and cause concretion of the spawning substrate (Phillips et al., 1975; Crisp, 1993).

Remedial measures currently employed to counter the detrimental impact of river bed siltation upon salmonid spawning are generally reach-based and include artificial re-stocking or egg box installation schemes and gravel substrate cleaning or restoration. Varying success rates have been reported for such remedial measures. Gravel cleaning or restoration commonly involves the removal of aquatic macrophyte beds and fine sediment deposits by tractor rotovating or raking and either high-powered jet or pump washing (Reeves et al., 1991; Shackle et al., 1999). Secondary siltation is, however, frequently experienced due to channel margin disturbances and transfer of sediment downstream and such methods are of limited value where increased sediment inputs to river channels are a recurrent problem (Carling, 1984). Alternatively, channel narrowing works have been employed to increase flow velocity and thereby reduce siltation within reaches important for salmonid spawning (Acornley and Sear, 1999), and, in some instances, clean gravel has been artificially introduced into stream channels as a means of restoring benthic habitat quality. Such remedial measures are frequently short-lived and their widespread application is often constrained by high cost, logistical problems and the need to avoid disturbance of fragile habitats.

In response to the shortcomings of conventional remedial measures for dealing with the problem of spawning gravel siltation, it is now increasingly recognised that a programme of prevention rather than cure is required. Such sediment control

programmes will in general require a catchment-wide perspective, since there is a need to reduce the supply of fine sediment, which commonly represents the underlying cause of the problem. The development of effective control or management strategies is, however, heavily dependent upon identification of the main source or sources of the sediment causing the siltation, since there is a need to reduce the supply of sediment from those sources. For many catchments, changes in agricultural practices have been identified as the underlying cause of spawning gravel siltation. In particular, increases in the area of arable cultivation and, more specifically, the shift from spring to autumn sowing of cereals which renders bare rolled soils susceptible to erosion by winter rains, have been cited as causing increased sediment inputs to the river system. Other factors related to land use change and changing land management have been implicated as causing increased connectivity within the sediment delivery system, thereby increasing the efficiency of sediment transfer. These include the decline and lack of maintenance of water meadows, which can act as sediment traps, and field enlargement, which results in the removal of hedgerows and other barriers to slope-channel connectivity. Equally, increased livestock stocking densities have been implicated as causing poaching and increased surface runoff and erosion in pasture areas (Foster and Walling, 1994; Collins et al., 1997a; Russell et al., 2001) as well as the trampling and poaching of channel banks. The latter can increase bank erosion and thus sediment inputs to the river system, and can also cause channel widening, with associated reduction in flow velocity and therefore greater potential for sediment deposition.

Although the effects of agricultural land use change and intensification outlined above have been widely implicated as the cause of increased siltation of salmonid spawning gravels, there have to date been few attempts to establish the main source or sources of fine interstitial sediment recovered from silted gravels. There is an important need for such information to inform ongoing debates surrounding the causes of spawning gravel siltation. Equally, such information is an essential prequisite to the development of effective control strategies. If the dominant source of such sediment is erosion of channel banks, attention should focus on controlling bank erosion, for example by excluding livestock from stream channels. Information on the provenance of the fine sediment responsible for siltation of spawning gravels is essential to guide the development of effective control strategies, which target control measures to the sources of the sediment involved (Scrivener and Brownlea, 1989; O’Connor, 1998; Shackle et al., 1999; Heaney et al., 2001).

Given the above context, a collaborative project was undertaken in 2000-2001, involving a reconnaissance survey of the provenance of fine-grained interstitial sediment retrieved from the spawning gravels of 18 important salmonid rivers in England and Wales (Walling et al., 2003). This project was based on the use of the fingerprinting approach to obtain information on the relative importance of surface and subsurface sources for samples of interstitial fine sediment recovered from salmonid spawning gravels during an extensive sampling programme conducted by the EA. The outputs from the project suggested that channel banks represented the dominant source of the fine-grained sediment damaging salmonid redds across SW England and accordingly, major bank fencing schemes were initiated. It was therefore deemed useful to revisit the study rivers across the SW of England and to deploy a repeat sediment source fingerprinting exercise as a means of assessing the efficacy of the stream bank fencing for protecting salmonid spawning areas.

The bank fencing study rivers and collection of source material and interstitial sediment samplesThe repeat sediment sourcing exercise to assess the efficacy of river bank fencing schemes was undertaken on six

rivers across SW England, using those sampling sites assessed during the pre-remediation study (Figure 15). These rivers were selected to represent a gradient of bank fencing implementation. Whereas the original study included the Rivers Fowey and Yealm, a reconnaissance survey of the presence of bank fencing by the West Country Rivers Trust, as part of this project, suggested that insufficient new fencing had been implemented since the original study to cause an effect. The catchment source material samples collected during the pre-remediation survey were used in the current project and the sourcing work aimed to apportion fine sediment degrading salmonid redds between surface (moorland, rough pasture, woodland/forest, improved pasture, cultivated combined) and channel bank/subsurface sources. Table 9 summarises the source material sampling exercise.

Representative samples of interstitial fine sediment were recovered from salmonid spawning gravels within each study river during the period November 2008-March 2009. The interstitial sediment samples during the initial study were collected over the 1999-2000 salmonid spawning season. Sample collection was based on the use of retrievable basket samplers (cf. Collins et al., 2008), which were inserted into artificial redds constructed in spawning gravels at the representative locations. Two aspects of the design, installation and deployment of the sampling baskets merit particular attention. First, the baskets were installed into artificial redds using a procedure that mimicked as closely as possible the action of a female salmon in cutting a redd. The process of redd construction caused much of the existing fine matrix sediment to be removed by winnowing and the baskets installed in the artificial redds were filled with representative clean framework gravel (>6.4 mm) prior to emplacement. The fine sediment recovered from the basket samplers following their extraction from the river bed at a later date, thus provided a sample of the fine sediment for use in source fingerprinting. Secondly, the specially designed sampling baskets incorporated an outer sleeve, which could be collapsed around the base of the basket during emplacement and raised around the sampling basket prior to removal from the bed gravel. Raising of the sleeve prevented loss through winnowing of the fines that had collected in the trap over the period of deployment. Table 10 provides summary information on the spawning gravel sampling sites employed for each study river.

Following removal from the river bed, the gravel and finer matrix sediment contained within the sampling basket were separated in situ using a 500 µm sieve, in order to separate the fine sediment that had accumulated within the basket after emplacement. The <500 µm fraction was subsequently wet sieved through a 125 µm sieve, and the <125 µm fraction was recovered by sedimentation and freeze-dried. In this study, attention has focussed on the <125 µm fraction, since existing research has demonstrated its adverse effects on spawning gravel permeability and porosity. Although the initial study sampled interstitial sediment at Crowford on the River Tamar, no samples were collected at this sampling site during the repeat sourcing survey, since no habitat suitable for salmonid spawning could be identified.

Figure 15: The study rivers and sampling sites used to assess the efficacy of stream bank fencing

Table 9: Summary information on the source material sampling exerciseCatchment Sediment source

SurfaceMoorland Rough

pasture Woodland/

forestImproved pasture

Cultivated Channel bank /

subsurfaceCamel - 8 11 10 10 10Fal - - 10 10 10 10Lynher - - 9 8 9 8Plym 8 - 7 7 - 8Tamar - - 11 11 9 12Tavy 8 - 8 9 - 9

Total = 230

Table 10: The location and number of interstitial sediment samples collected from each study catchmentCatchment Tributary Sampling site NGR No. of samplesCamel Allen Trehanick SX066791 2

main stem Kenningstock SX096807 2Fal main stem Tregony SW922450 2

main stem Golden Mill SW929468 2Lynher Deans Brook Villaton SX382623 2

main stem Bathpool SX286749 2Plym Meavy Clearbrook SX526665 2

main stem Bickleigh SX526618 2Tamar Inny Penpont Finches

BridgeSX260815 2

Lyd Foxcombe SX477874 2Sydenham SX429838 2

Ottery Canworthy Water SX229917 2main stem Crowford SX289994 0

Tavy main stem Brookmill SX477733 2main stem Iron Bridge SX511786 2Walkham Grenofen SX488709 2

Apportioning sediment inputs to salmonid spawning gravelsThe composite fingerprint used to apportion the interstitial sediment degrading salmonid spawning gravels between

surface and channel bank/subsurface sources comprised a combination of radiometric (137Cs, unsupported 210Pb, 226Ra) and organic (C, N) constituents. Fallout radionuclides (i.e. 137Cs and unsupported 210Pb) have consistently proved useful in sediment fingerprinting investigations, because their behaviour is essentially independent of geology and soil type and they provide an effective means of discriminating surface and channel banks / subsurface sources (Collins and Walling, 2004). Because of their origin as fallout, these radionuclides are restricted to the surface horizons of soils and their presence provides a good indicator of surface-derived material. Sediment mobilised from sources deeper in the soil or regolith, for examples from gullies and ditches, or from channel banks will be characterised by low or zero concentrations of 137Cs and unsupported 210Pb, because of limited or negligible exposure to fallout. However, topsoil from cultivated areas is typically characterised by a lower radionuclide content than uncultivated areas, because tillage mixes fallout within the plough layer thereby reducing surface concentrations. Radium-226, a radionuclide produced by the natural decay of 238U in situ, was also employed as a fingerprint property, because its measurement was necessary to determine unsupported 210Pb concentrations and since it has also proved a useful fingerprint property in previous source fingerprinting investigations. Measurements of the 137Cs, unsupported 210Pb and 226Ra content of source materials and interstitial fine sediment samples were undertaken by gamma spectrometry using high-resolution n-type HPGe detectors, in accordance with the procedures described by Joshi (1987) and Walling and Collins (2000). The source material radiometric data were decayed for comparison with the interstitial sediment samples collected during the repeat survey.

Organic constituents offer considerable potential for distinguishing surface and channel bank/subsurface sources, since they are commonly preferentially associated with surface soil horizons, rather than the underlying parent material (Peart, 1993). Contrasts in the concentrations of organic constituents in surface soils, due, for example, to the effects of repeated tillage lowering the organic mater content of cultivated soils compared to pasture and woodland areas, mean that such properties can also afford a means of discriminating surface soils under different land use. In this study, simple measurements of the organic carbon (C) and nitrogen (N) content of the source materials were employed as fingerprint properties. Concentrations of organic C and N were measured by pyrolysis using a CE Instruments NA2500 automatic elemental analyser. Since it was not possible to discriminate between surface sources under different land uses, a mean value of the fingerprint property measured for the individual surface sources was used.

On the assumption that the concentrations of the selected fingerprint properties in any given sample of interstitial fine sediment directly reflect the corresponding concentrations in the original source materials and the relative proportions of sediment contributed by these sources, the multivariate sediment mixing model represented by Equation 3 was used to determine the provenance of the samples of interstitial fine sediment. This mixing model incorporated corrections for particle size and organic matter selectivity, as well as tracer specific weightings based on the precision of laboratory analyses.

The efficacy of bank fencing for reducing sediment pressures on salmonid spawning gravels from channel banks

Figure 16 compares the results of the sediment source apportionment modelling for the pre and post remediation surveys. Source apportionment was based on 1000 repeat iterations of the mixing model. In the case of the River Camel, the estimated sediment inputs from surface and channel bank sources during the pre remediation survey were predicted to range between 0-100% with the corresponding respective means computed at 3±1% and 97±1%. The mixing model iterations for the post remediation study suggested ranges of 0-100%, but with accompanying mean contributions of 31±1% from surface sources and 69±1% for channel bank/subsurface sources. For the River Fal, the source fingerprinting exercise computed respective ranges of 0-10% and 90-100% for surface and channel bank sources and corresponding means of 6±1% and 94±1% for the pre remediation survey. The corresponding respective estimates for the post remediation study were computed at 0-76% and 24-100%, with associated means of 9±1% and 91±1%. Turning to the River Lynher, channel bank/subsurface sources were computed to contribute a mean of 12±1% (range 0-78%) of the fine sediment degrading salmonid spawning gravels during the pre remediation period, compared to 10±1% (range 0-33%) post remediation. For the initial survey, fine-grained sediment originating from eroding channel banks was estimated to contribute a mean of 92±1% (range 0-100%) of the siltation problem for salmonid spawning gravels in the River Plym. The corresponding contribution for the second survey was estimated to range between 0-100%, with a mean of 34±1%. In the case of the River Tamar, the estimated sediment inputs from surface and channel bank sources during the pre remediation study were computed to range between 0-100%, with respective means of 69±1% and 31±1%. Mixing model solutions for the post remediation survey suggested the same ranges, but with respective mean contributions of 84±1% and 16±1%. The source fingerprinting exercise suggested that surface and channel bank/subsurface sources in the River Tavy contributed a mean of 10±1% and 90±1% of the spawning gravel siltation problem during the pre remediation period, compared to 34±1% and 66±1% post the implementation of additional river bank fencing.

The two-sample Kolmogorov-Smirnov test was used to confirm statistically significant differences between the probability density functions (pdf’s) computed for the pre and post remediation surveys. This test was selected since it is sensitive to differences in the location and shape of predicted frequency distributions. Use of the Kolmogorov-Smirnov test ensured that the entire set of mixing model solutions for the pre and post remediation periods was taken into account for each study river, as opposed to the estimated means alone. It is important to take account of the uncertainty in the ranges of the contributions predicted for each sediment source type. Table 11 presents the results of the analysis. The Z statistic is a product of the combined sample size and the largest absolute difference between the two probability density functions being compared. A significance value of <0.05 indicates that the two pdf’s are significantly different. The results of the Kolmogorov-Smirnov test suggested that the pdf’s computed for channel bank contributions pre and post remediation were statistically significant for only the Rivers Fal and Plym. Corresponding statistical output for the remaining study rivers was not significant at the 95% level of confidence, although the results for the River Camel were almost significant (Table 11). Differences between the pdf’s representing predicted channel bank contributions to the siltation of salmonid spawning gravels for pre and post fencing remediation were smallest for the Rivers Lynher and Tavy (Table 11).

Figure 16: Ranges in the relative contributions from eroding channel banks for the pre and post remediation surveys

Table 11: The results of the Kolmogorov-Smirnov test comparing the probability density functions computed for pre and post remediation using the mixing model to apportion interstitial fine-grained sediment between surface and channel bank sources

Interstitial sediment sourceCatchment Surface Channel bank/subsurface

Z sig. Z sig.Camel 1.342 0.055 1.342 0.055Fal 1.565 0.015 1.565 0.015Lynher 0.671 0.759 0.671 0.759Plym 2.012 0.001 2.012 0.001Tamar 1.118 0.164 1.118 0.164Tavy 0.447 0.988 0.447 0.988

Implications of the findings

a) High resolution sediment sourcing in the Biglands Bog ECSFDI priority catchmentThe pilot project in the Biglands Bog ECSFDI priority catchment successfully demonstrated the utility of the proposed

novel framework for apportioning sediment loss from grass fields between poached gateways, poached cattle tracks and wider areas of more general hoofing damage. Given the limited number of tracking sites, the findings should be viewed as tentative. Outputs from the high resolution sourcing framework will be catchment-specific. In the case of the study catchment, the findings suggested that the mitigation of sediment loss from grass fields should primarily target wider areas of hoofing damage and poached cattle tracks. Sediment mobilisation and delivery to watercourses from poached gateways requires less attention. All outputs from the novel sourcing framework should be interpreted in the context of the study period and the rainfall and runoff patterns as well as the farm management decisions observed.

b) Assessing the efficacy of 6 m riparian buffer strips as a sediment mitigation optionThe use of the novel framework suggested that the efficacy of a 6 m riparian buffer for reducing sediment loss from

arable land (medium soils and slopes of 3-7 degrees) to the neighbouring watercourse was 100%. This result should, however, be interpreted within the context of the pilot study encompassing a single winter and the mature vegetation cover of the study buffer. Riparian buffer performance is site-specific and typically limited in duration for many pollutants. Climate change could increase the risk of buffer breaching on account of higher rainfall intensities and concentrated flow across arable fields.

c) Assessing the efficacy of channel bank fencing as a sediment mitigation option (commentary from the Salmon and Trout Association)

The pre and post impact study did suggest statistically significant reductions in bank erosion contributions to the artificial spawning redds in two of the six study rivers. Mixing model output for all of the study rivers suggested a reduction in the mean contribution from eroding channel banks to the degradation of salmonid spawning gravels. The Salmon and Trout Association (S&TA) believe these findings highlight the utility of bank side fencing projects in helping to protect valuable salmonid spawning habitats. But, fencing schemes are required at catchment scale and should be well-maintained. During the sampling period it was noted that some of the bank side fencing schemes were very fragmented, in a poor state of repair or had no buffer strip between the fencing and the riverbank, which is likely to have influenced the results. In some cases, cattle drinking bays had been included in the fencing schemes, thereby focusing trampling and increasing bank erosion and degradation. Equally, it is important to note that interstitial sediment was collected over a single spawning season and that the sampling period coincided with heavy flooding arising from snow melt in the south west study region, which may have increased the levels of bank erosion sediment inputs due to higher magnitude river flows and resulting channel scour. Since the findings suggested that a positive effect can be detected at catchment scale, the S&TA believes that bank fencing schemes should continue to be implemented as long as satisfactory resource can be directed towards maintenance and more careful consideration is extended to the installation of river access drinking bays. The latter will be particularly important with respect to additional pollutants such as faecal indicator organisms. Inclusion of river access drinking bays in bank fencing schemes should, where possible, be avoided to maximise the potential reductions in sediment pressures derived from eroding channel margins.

Potential future workThe proposed novel sourcing and tracing framework could be applied in future work in the following ways:

a) To characterise sediment loss from generic and higher resolution sediment sources for representative farm types. Such work would provide invaluable support for policy scenario modelling and for the localised targeting of mitigation options by CSFOs. Representative farm types could be classified on the basis of Robust Farm Types (MAFF, 1997) (e.g. Dairy, Cereals, Cattle and Sheep in lowland areas, Cattle and Sheep in Less Favoured Areas and Mixed). Fieldwork could combine Robust Farm Types with the Defra erosion risk typology (Defra, 2005) to permit extrapolation to national scale.

b) To assess the efficacy of sediment mitigation options to continue improving the evidence base. To date, strategic policy support for sediment (e.g. Collins et al., 2007; Collins and Anthony, 2008a,b) has used expert judgement on method efficacy to inform modelling frameworks and projections. Reliable data on the efficacy of the sediment mitigation options listed in the Defra User Manual (Cuttle et al., 2007) is urgently required for the range of environmental conditions across England and Wales. The novel sourcing and tracing methodology could be applied in pre- and post-remediation surveys to assemble such datasets. A major issue for ECSFDI is the demonstration of environmental improvement. To date, much effort has focused upon monitoring downstream water quality. But, given the attenuation of catchment response to mitigation planning that could be expected due to intermediate processes of storage and remobilisation (Collins and McGonigle, 2008), it would be more

appropriate to monitor environmental response at the source end of the catchment sediment budget. Pre- and post-remediation source apportionment studies using the proposed sourcing framework could be undertaken to assemble an evidence base on the environmental outcomes of the ECSFDI in relation to controlling sediment loss from those portions of river basins actually impacting on instream ecology. The use of plot scale studies to test method efficacy generates problems in upscaling the findings to catchment scale, as required by management programmes driven by the EU WFD, and fails to provide information on the efficacy of options for reducing sediment impacts in watercourses as opposed to in-field losses on catchment hillslopes. It is important to link mitigation option efficacy to sediment pressures in river channels and ecological condition. Coupling understanding of pollution pressures, potential for mitigation and habitat status is demanded by the WFD. Improved data on the efficacy of sediment mitigation methods for helping to reduce adverse instream sediment impacts on freshwater ecology would be invaluable for helping to engage catchment stakeholders and for providing the evidence base to policy teams that management options have the potential to close the ‘gap’ between current or future projected and compliant losses. Given the need to couple an improved evidence base on sediment mitigation option efficacy and ecological status, biological monitoring and experimentation could be usefully synthesized with the proposed sourcing and tracing framework. Such work could be undertaken as part of the new Defra/EA Demonstration Catchment project.

Dissemination of the findingsThe preliminary findings of the high resolution sourcing work in the Biglands Bog ECSFDI priority catchment were

presented at the European Geosciences Union (EGU) meeting in Vienna, April 2009. The final results will be used by the CSFO to engage stakeholders via the ECSFDI. The contractor intends to publish the results of the high resolution sediment source apportionment in an appropriate international Journal.

The results of the assessment of stream bank fencing will be disseminated by the Salmon and Trout Association, Wild Trout Trust and ECSFDI. In addition, the contractor intends to publish the results of the bank fencing efficacy work in an appropriate international Journal.

References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.

The references cited in this final report are listed in Appendix A.