John M Clapp v The Secretary of State for Environment, Food and Rural Affairs

Background

1. This appeal concerns land at Hartnoll Farm in Tiverton, Devon.

2. Regulation 4(2) of the Nitrate Pollution Prevention Regulations 2015 (“the Regulations”) requires the Secretary of State to monitor the nitrate concentration in freshwaters over a prescribed period. The nitrate must be measured, at representative locations, in order to identify water that is affected by pollution (or could be if the controls provided by the Regulations are not applied). By Regulation 4(5), the Secretary of State must then identify land which drains into those waters and that contributes to its pollution. If necessary, such land may then be designated as a “nitrate vulnerable zone” (“NVZ”).

3. For the purposes of the Regulations ‘pollution’: means the discharge, directly or indirectly, of nitrogen compounds from agricultural sources into the aquatic environment, the results of which are such as to cause hazards to human health, harm to living resources and to aquatic ecosystems, damage to amenities or interference with other legitimate uses of water (Council Directive 91/676/EEC).

4. Groundwaters are polluted if they contain a concentration of nitrates greater than 50 mg/l (or could do so if the controls in the Regulations were not to apply there), as provided in Annex I to the aforementioned Council Directive. Laboratories usually report nitrate results as the mass of nitrogen only. When reported in this way, the critical value is 11.3 mg/l.

5. The Regulations provide that every four years the Secretary of State must, where necessary, revise or add to the areas of land to the designation of NVZs. Before doing so, the Secretary of State must publish the proposals and send written notice under Regulation 5(3)(b) to the owner or occupier of a “relevant holding”.

6. The Regulations define “a relevant holding” as land and any associated buildings used for growing crops in soil, or rearing livestock for agricultural purposes, that fall wholly or partly in an NVZ. The occupier of a relevant holding must comply with NVZ-specific rules concerning the use of nitrogen fertilisers and the storage of manure.

7. Before the Secretary of State revises or adds to the designation of NVZs, regulation 5 requires the proposals to be publicised and written notice sent to anyone appearing to be the owner or occupier of a relevant holding. Regulation 6 then affords such an owner or occupier a right of appeal to the Tribunal. The only permitted grounds of appeal are that the relevant holding (or any part of it): (a) does not drain into water which the Secretary of State proposes to identify, or to continue to identify, as polluted [or] (b) drains into water which the Secretary of State should not identify, or should not continue to identify, as polluted. The Secretary of State refers to these as Type A and Type B appeals, respectively. The Respondent’s decision

8. A written notice dated 12 May 2025 (“the Notice”) was served on the Appellant, as the owner/occupier of land identified as falling within NVZ G18. The prefix ‘G’ stands for groundwater. The Notice states that the designated areas largely remain the same as they were in 2021-2024 . The appeal

9. On 6 June 2025, the Appellant appealed against the Notice on the basis that the water should not be identified as polluted, i.e. a Type B appeal. The grounds of appeal pose a series of questions on why designation has occurred.

10. The Respondent resists the appeal in a response filed pursuant to Rule 23 of the Tribunal Procedure (First-tier Tribunal) (General Regulatory Chamber) Rules 2009. The hearing

11. The Tribunal received a revised hearing bundle consisting of some 340 pages.

12. The Appellant, John Clapp, appeared at the hearing to confirm that his unsigned and undated statement was correct and true to the best of his knowledge and belief. He was represented at the hearing by his agricultural consultants, Janice Radford and Douglas Green, both of whom submitted witness statements.

13. Throughout the proceedings, the Environment Agency (“EA”) has been authorised by the Department of Environment, Food & Rural Affairs (“DEFRA”) to act on behalf of the Respondent. The bundle included witness statements from Natalie Kieboom, Senior Specialist in the EA’s research team specialising in hydrogeology. Ms Kieboom attended the hearing to give evidence on behalf of the Respondent.

14. The Tribunal has considered all the information placed before it and made the findings below on the balance of probabilities. In doing so, we do not identify every piece of evidence relied upon, it being unnecessary and impractical to do so. Nor do we attempt to explain technical data and evidence. Rather, we have taken the approach of explaining our conclusions on the main issue in dispute with sufficient reasons for the parties to understand why they have won or lost. Main issue

15. The main issue is whether the Secretary of State was wrong to continue to identify the water into which the land drains as polluted.

16. At the start of the hearing, the parties agreed this was the main issue to be decided. The Appellant’s case

17. The Appellant’s grounds of appeal are summarised below.

18. Hartnoll Farm lies between tributaries that feed into either the River Exe or River Culme. It is assumed that the issue with water quality lies with the River Culme because of the number of farms lying to the north of Hartnoll Farm and also along the River Exe that are not within the NVZ. [The Tribunal notes that these comments on water quality were made in the mistaken belief that it was a surface water NVZ whereas it is in fact a groundwater NVZ.]

19. From the maps and data available, the Appellant queries the logic of the designation.

20. The outcome sought is for the boundary of the NVZ area to be moved to exclude Hartnoll Farm. Failing that, the Appellant seeks the provision of data and information to explain and justify why the farm is within the NVZ. John Clapp’s evidence

21. In a witness statement, Mr Clapp expresses the belief that his business must have significantly reduced its contribution to groundwater nitrogen over the past 16 years. The business has always been on the very edge of the NVZ area. Since 2009 Mr Clapp has adopted and invested in many management changes that will have reduced nitrate leaching from his farm. He considers there are some serious questions regarding NVZ areas and how they are managed and regulated. He is angry and frustrated at the restrictions in farming practices within the NVZ and the disconnect with what water companies and other industries are permitted to do. Janice Radford’s evidence

22. Ms Radford explains in her witness statement that over the last 30 years she has been involved in producing nutrient management plans for livestock farmers. Her research reveals a total of 17.7 million fewer kilograms of manure nitrogen produced per annum from a reduction in livestock numbers in Devon since 2009. Whereas nitrogen production from humans is likely to have increased by 8.23 million kilograms over the same period. Artificial fertiliser use has also declined significantly. Given these changes, Ms Radford questions whether agriculture is 50-60% of the problem and if onerous constraints should be placed on agriculture if they only cause 17% of the problem. [The Tribunal notes that the reference to 17% derives from the European Court of Justice ruling in Case C-221/03 , Commission v Kingdom of Belgium at [86], where agricultural contributions were 17% of the total nitrogen in one river and 19% in another. The Court pointed out that “ although minor, those contributions are by no means insignificant ” when designating NVZs.]

23. Ms Radford is critical of the 2023 NVZ designation because it did not attempt to exclude monitoring points affected by urban runoff or sewage effluent, potentially skewing results against agriculture. Ms Radford considers that the Review relied on questionable datasets, such as the Crop Map of England (“CROME”) instead of the DEFRA agricultural census data interpreted via ADAS, with possible misclassifications. There was very little data in terms of water quality results. No consideration was given to the impact of winter cover crops or recent reductions in fertiliser use.

24. Having been supplied with the water quality results within the G18 NVZ, Ms Radford comments that only 20 water tests were done in 2025 to cover the whole 727km 2 area. Historically, there were 22,000 samples taken annually to determine NVZ areas in England according to NVZ review document 2023. Ms Radford considers that this undermines confidence in NVZ boundary decisions. The selective discontinuation of low-nitrate sampling points may skew average results. As groundwater testing is limited to November to March, only worse case nitrate levels are captured.

25. Despite great efforts by the Appellant and many farmers to reduce nitrate leaching, these efforts do not appear to be reflected in the groundwater nitrate levels. Ms Radford suspects that the improvements in farming practices are being masked by pollution coming from human sources.

26. It is suggested that farmers are being unfairly targeted compared with other contributors. Douglas Green’s evidence

27. Mr Green has been an agricultural consultant for over 40 years. He manages a team of qualified consultants providing nutrient management and slurry management advice to farmers.

28. Mr Green refers to the average reading from the sample sites provided for this NVZ area being 9.7 mg/l as nitrogen. It has been below the limit of 11.3 mg/l as nitrogen for the last 8 years. This is without taking into consideration the 4 sample sites that have been removed from the dataset, which were all low and would have reduced the 2025 average to just 8.4 mg/l as nitrogen.

29. Only 6 of the 20 tests taken were at or above 11.3 mg/l as nitrogen, so designation of this large area relies on just 6 water tests in a whole year. The water tests are also all taken during November to March when the groundwater nitrate levels are likely to be at their highest. These tests are likely to show the worst-case scenario.

30. Mr Green expresses grave concerns about the validity of sampling used to justify an NVZ. Three of the sampling points are close to the Appellants farm. Only one has a nitrate concentration above 11.3 mg/l as nitrogen. The number of sampling sites in the NVZ area is also very low with less than 20 currently for the entire NVZ area of 727 km 2 . The location is decided by the EA on a risk basis and are not randomly assigned. Mr Green maintains that results are skewed from the low number of sampling sites. He does not believe the results to be scientifically rigorous. Location of the sampling points are in clusters, and most are very remote from the Tiverton area where Mr Clapp’s farm is located. It is not proven that the elevated levels of nitrates result from farming activity. The Respondent’s response

31. The Respondent maintains that the appeal does not constitute significant new information or evidence to remove the appealed land from the NVZ designation.

32. As this is a groundwater NVZ, the proximity of the appealed land to surface water is not the main consideration. The Respondent states that it used the most recently available water quality monitoring data and land use data to assess whether any significant changes have occurred to affect the existing designations.

33. Water quality sampling data was provided from the vicinity of the Appellant’s farm. Two of the samples were from groundwater boreholes. Although the sampling results were similar to previous years, at approximately 12 mg/l as nitrogen and 5 mg/l as nitrogen, an NVZ cannot be de-designated based on individual monitoring results alone. Changes in groundwater chemistry occur very slowly, and improvements in water quality take time to become apparent. The Method and Data Sheet documents detail the amount of information, statistical analysis and complexity of the designation process.

34. For groundwater NVZs, the boundary extent is set by geological, geographical or hydrological boundaries and an understanding of the behaviour of groundwater and nitrates. In this instance, the Respondent considers there is no justification to amend the NVZ boundary.

35. The Appellant has considered discharges of sewage effluent to surface water rather than groundwater in their evidence. The Respondent would not generally expect surface water quality to affect groundwater quality for reasons given in Ms Kieboom’s evidence. Natalie Kieboom’s evidence

36. Ms Kieboom states that water quality monitoring data from within the NVZ continues to show levels at or above the 50mg/l nitrate concentration threshold (equivalent to 11.3mg/l when measured “as nitrogen”) .

37. Following a full review of all the groundwater monitoring locations, no evidence has been found that the monitoring points are affected by discharges of sewage effluent. Ms Kieboom would not expect surface water quality to affect groundwater quality as groundwater flows in the direction of the hydraulic gradient from high to low head. It discharges to the low points in the system, which are surface watercourses, lakes and the sea. So, groundwater generally flows into surface water, rather than vice versa. Where groundwater and surface water are in hydraulic continuity, as the generally are, surface water will not flow into groundwater because to do so it would have to flow against the hydraulic gradien t.

38. The groundwater in this location is unlikely to receive much water from overlying watercourses. As such, the groundwater is not considered to be affected by any sewage entering the surface waters.

39. In answer to the Appellant’s points, Ms Kieboom explained at the hearing that the specific reason for the reduction in water sampling in 2023 was that it had taken time after the global Covid-19 pandemic to get sampling back up and running. There had been reductions in sampling over time since 2008 anyway due to funding cuts and laboratory capacity.

40. More recently, the EA has lost access to some boreholes on private land due to new safety requirements for the electrical certification of pumps in boreholes. In addition, three boreholes became unavailable to the EA in this NVZ as the farms supplied by those boreholes “went on to mains connection”. Nitrate readings had historically been lower in those three boreholes. The EA had not deliberately omitted them from sampling. The three boreholes were simply no longer available.

41. In this NVZ, groundwater samples were taken twice a year between November to March because priority is given in Devon to collecting samples for bathing water between May and September. It was explained that the same team of people was responsible for both types of sampling.

42. In reply to the Appellant’s criticism that sampling had been undertaken at the worst time of year, Ms Kieboom did not necessarily agree that nitrate levels would be lower at other times of year. It varies on the time taken to pass through the unsaturated zone and reach the water table and other variables such as the depth of boreholes. Nevertheless, Ms Kieboom acknowledged that winter is “definitely the higher risk time” for nitrate leaching. When asked if there could be seasonal skew in the figures, Ms Kieboom replied that not all boreholes have a seasonal variation. She accepted that it cannot be known either way as there is no summer data.

43. Ms Kieboom confirmed that, with the agreement of DEFRA, the EA had not applied the published methodology. The assessment by WSP Ltd dated 2023 was the process for this NVZ cycle and the information that was looked at.

44. Following full review over three cycles since 2008, with changes of less than 1% in the water quality sampling data, there had not been significant changes to warrant de-designation. Findings of fact

45. This groundwater NVZ has been designated since 2008. It covers an area of over 700 km 2 .

46. The methodology applied in the original designation is not disputed by the Appellant.

47. The EA confirms that the land to which the disputed notice applies is associated with the address to which the notice has been sent. The Appellant did not dispute this.

48. DEFRA’s published methodology from December 2016 is titled “Implementation of the Nitrate Pollution Prevention Regulations 2015 in England - Method for designating Nitrate Vulnerable Zones for groundwaters.” (“the Methodology”). The Respondent did not apply the Methodology for the 2024 review cycle.

49. The WSP Ltd report does not present any data specific to land use in NVZ G18. Consideration and Conclusions

50. In this Type B appeal the Appellant must demonstrate that the Secretary of State should not identify, or should not continue to identify, the relevant groundwater as polluted.

51. Ms Kieboom acknowledged that the EA did not, with the agreement of DEFRA, apply its own Methodology. Instead, the approach taken was to look at changes over the last three cycles of NVZ review rather than following the more robust measures within the Methodology. Nothing within the statutory framework prescribes the procedure to be followed by the Respondent in discharging the duties under Regulation 4. That being so, the EA was entitled to use appropriate statistical methodologies in order to draw a reasonable conclusion. The basis of that conclusion is challenged by the Appellant.

52. On a procedural note, whilst an alternative approach can be taken in discharging the Respondent’s duties, any such alternative should be clearly set out. It does not suffice in our view for the Respondent to simply take lesser steps than those in its Methodology only to argue that the Appellant has no evidence to contradict its decision. Of course, an Appellant will not have had the benefit of undertaking its own groundwater body-wide sampling and invariably will not have the same expert analysis available as the EA.

53. With the benefit of specialist members sitting on the panel, the Tribunal has been able to question the evidence at the hearing to adjudicate on this appeal.

54. The crux of the matter is whether the Respondent’s decision was ‘wrong’ at the time it was taken ( Waltham Forest LBC v Hussain & Ors [2023] EWCA (Civ) 733 ). The Court of Appeal in Waltham Forest elaborated: “64. “Wrong”, as Upper Tribunal Judge Cooke explained in Marshall v Waltham Forest LBC [2020] UKUT 35 (LC) , [2020] 1 WLR 3187 at [61]–[62], means in this context that the appellate tribunal disagrees with the original decision despite having accorded it the deference (or “special weight”) appropriate to a decision involving the exercise of judgment by the body tasked by Parliament with the primary responsibility for making licensing decisions. It does not mean “wrong in law”. Put simply, the question that the FTT must address is, does the Tribunal consider that the authority should have decided the application differently?”

55. In this regulatory context, the same principle applies with the decision being that of the Secretary of State (who has delegated certain functions to the EA).

56. The thrust of the Respondent’s case is that water quality monitoring data within the NVZ continues to show levels at or above the 50mg/l nitrate concentration threshold. In addition, Ms Kieboom states: “ The wider NVZ review of groundwater boreholes within the NVZ showed some elevated nitrogen concentrations and upward trends, indicating a risk of groundwater pollution. The analysis of monitored concentrations together with nitrogen loads underpins the recommendation to retain the designation.” Hence, retention of the designation is based upon: (a) monitored concentrations and (b) nitrogen loads.

57. Nitrate measurements in groundwater were taken at 26 locations across the area of the NVZ catchment. The conclusion reached by the EA is that there are some elevated values and increasing trends. Notably, the WSP Ltd report considers the position nationally and it does not specifically mention NVZ G18.

58. The amount of groundwater sampling has depreciated significantly over the years. There are now also fewer sampling points. The timing of samples may or may not have skewed results. It cannot be known either way as the EA did not pursue further testing to eliminate any seasonal skew for the boreholes tested within this NVZ.

59. The 2017 designation note sets out the predicted concentrations for 2027 using the statistical and modelling methods described in the Methodology. The modelling results predicted some marked increase in concentrations in some locations. There is no clear evidence to suggest that concentrations are rising towards these high levels. Conversely, the actual measured values for many of the locations are stable.

60. The Tribunal recognises that groundwater flow is complex and a delayed "slug" of polluted water may be migrating though the unsaturated zone given the sandstone-dominated geology. It is unlikely that no rising trend would be apparent this close to the 2027 if the modelling were accurate. It is probable that the modelling is overestimating the nitrate loading. It is unfortunate that this important line of evidence was not updated for the 2025 designation round.

61. No updated land use data has been presented to the Tribunal. We note that the detailed land use and nitrate loading assessment (the ADAS Method) used in previous rounds was not available to the EA for this round. It was unclear why this critical element of the previously accepted methodology was not available. Although there is reference to data split by groundwater body in the WSP report (appendix c of the bundle), this was not provided. The Tribunal also notes the WSP Ltd report itself recognises several issues with the land use data applied (CROME) and that it is hard to meaningfully compare this to the previous reviews. Very little weight can be given to this data given the uncertainties set out in the Respondent’s own report.

62. The Appellant provided information noting a cattle reduction in the south-west of the country and indicated that similar patterns were seen for other livestock. This was not contested by the Respondent. The Appellant also provided data showing that fertilizer application has dropped within England and Wales with a marked drop from 2020 onwards. Again, this was not contested by the Respondent. The Tribunal notes that this is national level data only although we have no reason to suppose that the overall position would differ locally. Nevertheless, the question remains as to whether the groundwater in this NVZ remains polluted.

63. The EA states that additional waste generated by an increasing population and high tourist numbers, should be directed to and treated by local sewage treatment works. We find no reason to conclude that the Respondent was wrong in this regard. There are serious shortcomings in the approach taken by the EA in other respects. Whilst there is no requirement that pollution, or risk of pollution, be established with total certainty (even if that were possible or practicable), and whilst the EA’s assessment does not need to be comprehensive and free of all flaws, the applied method must still constitute a scientifically valid approach.

64. The Appellant has challenged the results from sampling with good cause. It is not sufficient to simply show a single sample above the standard and on that basis declare an entire waterbody polluted. We do not consider it a sound method of determining a representative concentration for a wide area to be compared against a standard. The Tribunal is not satisfied that one point of source of pollution is representative of wider water quality for such a large waterbody without variability being methodically assessed.

65. From the national review, it cannot be gleaned whether pollution levels have decreased in this NVZ area in line with the overall reduction nationally. However, the WSP Ltd report does show mean Total Inorganic Nitrogen (TIN) sources as below the standard for both 2016-2019 and 2022-2023 monitoring periods. Thus, on the EA’s own evidence the mean NVZ concentrations at least were below the standard.

66. Additionally, the groundwater quality monitoring data provided by the Respondent is analysed by the Appellant to also determine an average concentration for the waterbody, as referenced above. That average is 9.7 mg/l nitrate as N, or 8.4 mg/l nitrate as N when reincorporating excluded (but lower concentration) sampling locations. These values do not account for unknown but potential skew (i.e. sampling bias) towards higher (winter) concentrations as discussed. Nonetheless, both values fall well beneath the 11.3 mg/l nitrate as N threshold specified by the Regulations.

67. The Tribunal bears well in mind that the EA has expertise, as the body appointed by the Respondent for these purposes, to identify water as polluted. We give special weight to the professional judgement exercised by the EA that informed the Respondent’s decision. Despite according the EA that deference, its approach reveals significant cause for concern in the soundness of the conclusion reached in this case.

68. We find that it was wrong for the EA to conclude that the water is affected by pollution on the evidential basis relied upon in this case. Regulation 4(5)(a) also requires the Respondent to consider if the water could be affected by pollution if the controls required in NVZs were not applied to the land in question.

69. There is no updated predictive model provided by the EA to show that water could become polluted unless measures are taken through NVZ designation. Based on the Appellant’s uncontested evidence, it may be anticipated that changes to land use will gradually over time help to reduce groundwater pollution. As the Appellant says, the application of artificial fertiliser to agricultural land has reduced nationally due to the inhibitive high cost. The cost is a strong deterrent. The primary future risks as identified in this appeal arise from: (i) farmers increasing stocking numbers (i.e. more animals on the land), and (ii) spreading of sewage sludge.

70. Even if stocking numbers increased, farmers would still need nitrogen management plans in place. The spreading of sewage is regulated by The Reduction and Prevention of Agricultural Diffuse Pollution (England) Regulations 2018, also known as the ‘Farming Rules for Water,’ accompanied by the statutory guidance published by DEFRA titled ‘Enforcing the Farming Rules for Water’. Thus, the primary foreseeable risks are subject to controls notwithstanding designation.

71. There is nothing before us to give cause us to conclude that nitrate pollution levels could be exceeded if the controls in the Regulations are not applied in the area concerned, using the criteria in Annex I of Council Directive 91/676/EEC.

72. On the balance of probabilities, the Tribunal concludes that the Respondent was wrong, at the time the decision was taken, to designate the NVZ as polluted water. Accordingly, the appeal is upheld under Regulation 6(2)(b). Signed: Judge Saward Date: 2 February 2026