Fen Raft Spider Survey
Pevensey Levels 2025

2. Introduction

Background

In June 2025, Natural England, in partnership with the British Arachnological Society (BAS) and the Cuckmere and Pevensey Water Level Management Board, commissioned Wil J. Heeney of Flauna Ecology to carry out a survey of the Fen Raft Spider across the Pevensey Levels. The Pevensey Levels holds the most extensive remnant population of this species in Britain, yet it had not been subject to a structured survey since 1990. This project represents the first comprehensive reassessment of the population in over 35 years.

The Fen Raft Spider

The Fen Raft Spider (Dolomedes plantarius) is one of Britain's largest spiders, with adults measuring up to 23 mm in body length. It is a semi-aquatic predator, hunting from the water surface by detecting the vibrations of prey — including invertebrates and occasionally small fish or tadpoles. Breeding takes place during summer, with females constructing a distinctive nursery web within emergent or floating vegetation above the water surface. The egg sac hatches as the web is constructed, and the female guards the spiderlings for approximately a week before they disperse.

The species can be confused with the closely related Raft Spider (Dolomedes fimbriatus), but habitat preference helps separate the two: the Fen Raft Spider favours lowland, neutral to alkaline wetlands, while the Raft Spider is found in upland acid bogs and lowland wet heaths. Confident identification requires microscopic examination or DNA analysis. The Pevensey Levels population, in typical D. plantarius habitat, was first correctly identified in 1988 (Kirby 1990). D. fimbriatus has not been recorded from the Pevensey Levels.

The Fen Raft Spider is one of Britain's rarest invertebrates. Only three natural populations are currently known: at Redgrave and Lopham Fen in Suffolk, at the Pevensey Levels in East Sussex, and at a disused canal near Swansea and nearby Crymlyn Bog in South Wales. Since 2010, four additional populations have been successfully established in the Norfolk Broads through a conservation translocation programme, and these have since expanded beyond their original release ditches.

Previous Surveys

The only previous broad-scale survey of D. plantarius on the Pevensey Levels was carried out in 1990 by Jones (1992), commissioned by English Nature (now Natural England). Approximately 90 km of ditch segments were surveyed, concentrated within the Pevensey Levels SSSI but including selected peripheral areas. Nursery webs were used as the primary detection method — the same approach used in 2025. The 1990 survey did not use a standardised random sampling design, and habitat data were recorded in detail only where spiders were present.

Aims of this Survey

The 2025 survey was designed to address four principal objectives: to survey approximately 50% of the estimated ditch network during the spider's breeding season (July–September); to compare findings directly with the 1990 survey; to produce both a written and an interactive digital report of results; and to develop a standardised digital recording form to support future volunteer monitoring of the species.

3. Methodology

Personnel

All fieldwork, data analysis, and reporting was carried out by Wil J. Heeney of Flauna Ecology, an ecologist specialising in invertebrate surveys.

Survey Design

The extent of the Pevensey Levels made it impractical to survey every ditch, so a stratified random sampling approach was used to ensure even, unbiased coverage. The survey area was divided into the eight sub-catchments used for drainage management across the Levels, and sampling was spread across all eight to avoid clustering effort in any one area. No segments were selected in the Ashbourne and Hugletts Stream sub-catchment due to access and parking constraints encountered during fieldwork.

No existing shapefile of the complete ditch network was available, so the entire network was mapped from scratch using aerial photography in QGIS — a process that involved digitising over 540 km of waterway and dividing it into individual ditch segments. Each segment was assigned a unique reference number with a prefix corresponding to its sub-catchment. The 1990 survey map was georeferenced and overlaid onto this network to identify previously surveyed segments as accurately as possible, ensuring they were all included in the 2025 survey design.

A reconnaissance survey of the Combe Haven area was also included. Combe Haven lies approximately 10 km west of the main Pevensey Levels, separated by the built-up area of Bexhill-on-Sea. The population was first recorded in June 2022 at TQ 761 103 and had not previously been subject to a baseline survey. It was treated as a separate, opportunistic component of the fieldwork.

Data Collection

Fieldwork was conducted between 5 July and 20 September 2025, covering 32 survey days. All data were collected in the field using the Coreo mobile app (Natural Apptitude), a platform designed for ecological data collection. Three custom survey forms were created: one for mapping ditch segments in the field, one for recording habitat information at each segment (including width, water depth, land use, and the presence and density of Water Soldier, Floating Pennywort, and NZ Pygmyweed), and one for recording nursery web samples.

When nursery webs were encountered, a standardised 4-metre sample was used: the surveyor counted all webs within the following 4 m, recording the total number of webs, how many were occupied, and how many were supported by Water Soldier. This process was repeated continuously along any length of ditch where webs were present. Georeferenced photographs were taken at each sampling point and at each segment midpoint.

To reduce the risk of sampling any sub-catchment only during a peak or trough in breeding activity — the spider can produce two or occasionally three broods across the season — visits to segments within each sub-catchment were spread across the full survey period rather than concentrated into a single window.

The Interactive Report

To make the survey findings accessible to a non-specialist audience, the results were also compiled into this interactive digital report. Built as a self-contained HTML file using the open-source Leaflet.js mapping library, it requires no internet connection or specialist software. All data layers — spanning 1990 and 2025 nursery web records, habitat data, comparison categories, and Combe Haven — can be explored via the interactive map, with clickable popups providing segment-level detail including survey outcomes, habitat scores, and georeferenced photographs from the 2025 survey.

Statistical Analysis

All statistical analyses were carried out in R. Associations between nursery web presence and habitat variables were tested using logistic regression, Fisher's exact test, and the Mann-Whitney U test as appropriate. The significance of population change between 1990 and 2025 at directly comparable sites was assessed using McNemar's test for paired presence/absence data. Spearman's rank correlation was used to examine the relationship between Water Soldier density and the number of nursery webs per sample. A significance threshold of p < 0.05 was applied throughout.

4. Results

Survey Coverage

The 2025 survey selected 1,166 segments (256.4 km) for survey across the eight sub-catchments, representing 46.3% of the total mapped network of 540.8 km. Of these, 289 (24.8%) could not be assessed due to access restrictions, predominantly in fields with cattle, giving a realised survey of 877 segments (193.5 km). This compares to 763 segments (165.1 km; 30.3%) surveyed in 1990.

Nursery Web Distribution in 2025

Nursery webs of the Fen Raft Spider were recorded in 80 ditch segments across the survey area in 2025, compared to 246 segments in 1990. However, direct comparison of these raw totals is not straightforward: the two surveys differed in coverage, methods, timing, and in the degree to which access restrictions affected sampling. These limitations are discussed fully in Section 5.

A total of 243 point records were made, comprising 237 nursery web sample locations. Across these samples, 580 nursery webs were recorded, of which 268 (46.2%) were occupied. Of all webs recorded, 402 (69.3%) were supported by Water Soldier. Nursery web occurrence was spatially concentrated, with ten ditch segments each containing ten or more webs.

Comparison with the 1990 Survey

Direct comparison between the two surveys was possible for 376 segments surveyed in both years and accessible in 2025. Of these, nursery webs were present in both years at 22 segments (5.9%), in 1990 only at 76 segments (20.2%), and in 2025 only at 22 segments (5.9%). This net reduction was statistically significant (McNemar's test, p < 0.0001), with nursery webs no longer recorded in 76 segments and newly recorded in 22, a ratio of 3.5:1. However, this figure must be treated as a minimum: a further 62 segments with 1990 nursery web records could not be accessed in 2025. The majority of these lie within the Pevensey Levels NNR, where habitat management is favourable and Water Soldier is present — population persistence here is considered likely.

Nursery webs were additionally recorded at 36 segments not included in the 1990 survey, distributed across previously unsampled areas of the network.

Water Soldier and Habitat

Water Soldier (Stratiotes aloides) was by far the strongest predictor of nursery web presence. Ditches with Water Soldier were over eight times more likely to have nursery webs recorded (42.9% of Water Soldier segments, compared to 8.2% without; Fisher's exact test, OR = 8.38, p < 0.0001). Where nursery webs were present, the number of webs per sample was strongly linked to Water Soldier density (Spearman ρ = 0.72, p < 0.0001): samples in Water Soldier ditches contained on average three times as many webs as those without. Nearly all webs (96.5%) in Water Soldier segments were constructed directly over the plant, consistent with active selection of it as a web substrate.

Of the invasive non-native species recorded, neither New Zealand Pygmyweed (Crassula helmsii) nor Floating Pennywort (Hydrocotyle ranunculoides) showed a negative association with nursery web presence at current levels of occurrence. A positive association was found with Floating Pennywort (OR = 2.73, p < 0.0001), most likely reflecting the large surface area of emergent structure it creates. Dense Floating Pennywort growth may reduce the detectability of nursery webs, and monitoring of Floating Pennywort and NZ Pygmyweed coverage is recommended.

Combe Haven — Outlying Population

A reconnaissance visit to the Combe Haven area, approximately 10 km west of the main Pevensey Levels, was made on 20–21 July 2025. A total of 43 ditch segments were surveyed across 7.86 km of waterway, covering the majority of the accessible ditch network at the site. Survey conditions were substantially more challenging than on the main Levels: a high proportion of segments were heavily overgrown, wooded, or reedy, with 12 of 43 segments inaccessible. Water Soldier was absent throughout.

The presence of the Fen Raft Spider was reconfirmed in one segment (c003, TQ 761 103), where a single occupied nursery web was recorded. The population was first discovered at this location in June 2022. A dedicated, repeat survey is recommended to better characterise the extent and status of this population.

5. Discussion

Water Soldier as Critical Habitat

The strength of the association between Water Soldier and nursery web presence, long recognised but here quantified for the first time, is a key finding of this survey, and has direct implications for ditch management across the Pevensey Levels. The plant functions not just as an indicator of suitable conditions but as an active resource: spiders build their nursery webs directly over it, and the number of webs per sample increases markedly with Water Soldier abundance. Conservation management should therefore prioritise maintaining existing Water Soldier stands, particularly in the ditches where it is currently established.

It is important to note that Water Soldier is not a prerequisite for spider survival — populations persist without it at Redgrave and Lopham Fen and at Crymlyn Bog in Wales, and Norfolk Broads translocated populations have expanded well beyond their original Water Soldier ditches, as in Broadland more generally. Nonetheless, on the Pevensey Levels, ditches with Water Soldier support substantially higher nursery web densities, and their value to the spider should be explicitly recognised in any ditch management plans.

Interpreting the Survey Comparison

The raw comparison between surveys — 80 ditches in 2025 versus 246 in 1990 — cannot be interpreted as evidence of population loss. Significant differences between the two surveys in methods, coverage, and access restrictions all complicate direct comparison. The statistically significant net reduction at directly comparable accessible sites is real, but almost certainly overstates the true extent of change for several reasons.

Most importantly, the 62 inaccessible segments within the NNR remain unassessed, and populations there are considered likely to have persisted given favourable management conditions. SWT monitoring in 2012 and 2017 confirmed that this core area continued to support population hotspots relatively recently. Additionally, the single-visit design means that a low-density ditch visited during a trough between broods could produce a false negative result. Nursery webs were also recorded at 36 segments not included in the 1990 survey, suggesting the population occupies a broader area than the baseline data indicated.

Of the 76 segments where nursery webs were no longer recorded, field notes recorded reed infilling or scrub encroachment at around a third of sites — habitat degradation that is a plausible mechanism for genuine local losses. The remaining two-thirds were surveyed without notable habitat concerns; whether absence at these sites reflects genuine extinction, missed detection, or seasonal timing cannot be determined from single-visit data.

Non-Native Plants — No Current Concern

Neither NZ Pygmyweed nor Floating Pennywort appears to be limiting the spider population at current levels of occurrence. A significant positive association was found with Floating Pennywort, most likely reflecting the large surface area of emergent structure it creates, though it appears much less suitable than Water Soldier. Dense Floating Pennywort growth may reduce the detectability of nursery webs. Floating Pennywort coverage is already monitored by the Water Level Management Board and Environment Agency. Water Soldier, while technically non-native to Sussex, is Nationally Scarce and has never been managed as a problem species on the Levels; given its value to the spider, this approach should continue.

Digital Recording for Future Monitoring

This survey demonstrated the value of digital data collection using the Coreo app, which provided automatic GPS location capture, real-time data review, and georeferenced photographs. A key outcome of the project is the proposal to develop a bespoke Fen Raft Spider recording form on the iRecord platform to support future volunteer-led monitoring. Discussions with the Biological Records Centre regarding the development of this form are ongoing.

Combe Haven

The single confirmed record at Combe Haven reconfirms the presence of a breeding population, first discovered in June 2022. A dedicated repeat survey to determine the extent of the population and form the baseline for future monitoring is recommended. The site differs substantially from the Pevensey Levels — ditches are heavily overgrown, shaded, and lacking Water Soldier — and the ditch network appears to have received little active management. Priority management recommendations include clearance of overhanging scrub and carr woodland, rotational cutting of dense reedbed, and sympathetic grazing management. On the Pevensey Levels, cattle poaching of ditch margins creates tussocky vegetation and small water-filled depressions that provide good habitat; similar management at Combe Haven could deliver comparable benefits. A dedicated repeat survey, ideally boat-based during peak breeding season (late July to mid-August), is recommended to characterise the population properly.

Climate Change and Long-term Viability

The Pevensey Levels is a low-lying coastal grazing marsh, much of it below sea level, and faces long-term risks from sea level rise, saline intrusion, and increasing summer drought. Survey notes from 2025 recorded several segments as dry or with substantially reduced water levels. Fen Raft Spiders ideally require a sufficiently high water table to support their water surface hunting behaviour; they can tolerate lowered water levels, but these conditions are likely to result in progressive population decline rather than catastrophic failure. These pressures reinforce the case for expanding the translocation programme to establish additional self-sustaining populations at geographically distinct, climatically resilient sites.

Connectivity, Expansion and Recovery

The 36 segments with nursery webs in previously unsampled areas indicate that the population is not confined to historically documented locations and has the capacity to occupy suitable habitat more broadly across the Levels. However, the spatial concentration of breeding activity into a smaller number of high-density hotspots raises questions about connectivity between occupied segments. Improving the habitat quality of intervening segments — through water level management, ditch clearance, and the encouragement of Water Soldier — would increase network connectivity. The development of eDNA monitoring tools for D. plantarius, identified as a priority in Natural England's Threatened Spiders Recovery Actions, would substantially improve the ability to assess population connectivity and detect presence in difficult-to-survey habitats. Development of eDNA methodology for the species is already in progress at the University of Nottingham.

6. Recommendations

For Natural England and the British Arachnological Society

Resurvey the inaccessible NNR segments. The 62 segments with 1990 nursery web records that could not be accessed in 2025 lie predominantly within the Pevensey Levels NNR. Systematic monitoring should be resumed within the SWT reserve, with a view to initiating a monitoring scheme for the Levels as a whole — these segments represent the most significant gap in the current dataset.

Follow up the 76 segments where nursery webs were present only in 1990. A targeted resurvey would help distinguish genuine population loss from detection failure. Sites with documented reed or scrub encroachment should be prioritised for habitat intervention.

Investigate eDNA monitoring. Environmental DNA methodology, identified as a priority in Natural England's Threatened Spiders Recovery Actions, would substantially improve the ability to detect presence in difficult-to-survey habitats — such as those at Combe Haven and within inaccessible NNR segments. Development work is already in progress at the University of Nottingham.

Expand the translocation programme nationally. The risks posed by summer drought and coastal squeeze reinforce the case for establishing additional self-sustaining populations at geographically distinct, climatically resilient sites beyond the current range.

For Land Managers and Farmers

Manage ditch vegetation to maintain open water. Reed encroachment and bankside scrub development were recorded as factors at around a third of segments where nursery webs were present only in 1990. Ditch management — including removal of scrub from southern and eastern bankside aspects, and removal of reed rhizomes when reprofiling — can restore habitat suitability at degraded sites and improve connectivity across the network.

Protect Water Soldier stands. Water Soldier is the single most important habitat feature for nursery web density on the Pevensey Levels. Where it is established, it should be maintained and protected from mechanical ditch clearance during the breeding season.

Maintain adequate water levels during summer. Fen Raft Spiders require a sufficiently high water table to support surface hunting and the construction of nursery webs above a water surface. Water level management during July–September, as set out in the Water Level Management Plan, should continue to take the spider's requirements into account.

Enable survey access where possible. Nearly a quarter of the 2025 survey area could not be fully assessed due to access restrictions, primarily in cattle fields. Where health and safety considerations allow, temporary stock movements or prior notification during the survey window would substantially improve the completeness of future monitoring.

For Future Volunteer Monitoring

Adopt standardised digital recording via iRecord. A bespoke Fen Raft Spider recording form is proposed for the iRecord platform, managed by the Biological Records Centre. This would allow volunteers to submit georeferenced records directly into national biodiversity databases, eliminating transcription errors and enabling near-real-time monitoring of survey progress. Development of this form is underway and discussions with BRC are ongoing.

For Combe Haven

Commission a dedicated repeat survey. Ideally boat-based and timed for peak breeding season (late July to mid-August), this would allow a far greater proportion of the ditch network to be assessed and provide the first reliable characterisation of the population.

Prioritise ditch restoration. Clearance of overhanging scrub and carr woodland, rotational emergent vegetation management, and sympathetic grazing management would all improve habitat suitability and increase the detectability of nursery webs during future surveys.

7. References

Bristowe, W.S. 1958. The World of Spiders. Collins, London.

Duffey, E. 1958. Dolomedes plantarius Clerck, a spider new to Britain, found in the upper Waveney Valley. Transactions of the Norfolk and Norwich Naturalists' Society 18: 1–5.

Jones, E. 1992. The status of Dolomedes plantarius (Clerck) on Pevensey Levels, East Sussex in August 1991. English Nature, Peterborough. [unpublished report]

Kirby, P. 1990. Dolomedes plantarius in East Sussex. Newsletter of the British Arachnological Society 58: 18.

Natural Apptitude. 2026. Coreo [mobile application]. Available at: https://coreo.io

QGIS Development Team. 2026. QGIS Geographic Information System. Open Source Geospatial Foundation. Available at: https://www.qgis.org

R Core Team. 2025. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Available at: https://www.r-project.org

Smith, H. 2020. The Fen Raft Spider: from unknown past to uncertain future. British Wildlife 32, pp. 98–109.

Smith, H., Clarke, D., Heaver, D., Hughes, I., Pearce-Kelly, P. & Sainsbury, T. 2013. Translocation and augmentation of the fen raft spider populations in the UK. In: Soorae, P.S. (ed.) Global reintroduction perspectives: 2013. IUCN/SSC Re-introduction Specialist Group, Abu Dhabi. pp. xiv + 282.