Tick hazard in the South Downs National Park (UK): species, distribution, key locations for future interventions, site density, habitats

South Downs National Park

The South Downs National Park (SDNP) covers 1,627 km² of the south-east of the British Isles, across Hampshire, West Sussex, and East Sussex. It encompasses two bioregions, a 140 km long chalk ridge called the South Downs, and a wooded lowland area called the Weald (from old English, meaning ‘forest’ (Hoad, 2003)). In total 23% is covered by woodland (SDNPA, 2019). Though much of the Park is subject to industrial agriculture, 4% remains species rich semi-natural chalk grassland, whilst 45% of its woodlands are legally classified as ancient (SDNPA, 2019) (i.e., wooded continuously since at least 1600 AD (Natural England & the Forestry Commission, 2022)). Both bioregions harbour biodiverse ecological communities of importance at national (e.g., 86 Special Areas of Scientific Interest (SSSIs)), and international scales (e.g., 13 Special Areas of Conservation (SACs)) (Crane & Williams, 2013). Portions of this text were previously published as part of a preprint: https://www.biorxiv.org/content/10.1101/2021.03.23.436533v1.full.

The Park is the most visited national park in the UK, with an estimated 39 million visitor days per year (NPUK, 2014), c. 120,000 people live and/or work within its borders, two million live within 5 km (SDNPA, 2020). UK national park designation does not indicate state or public ownership (unlike for example in the USA), and the great majority of land within national parks is privately owned (Butler, 2018). This is particularly so in the SDNP where 70% of the Park is farmed (SDNPA, 2019), and nearly a fifth (70,699 acres) is owned by just eight members of the gentry (two Dukes, three Viscounts, one Baron, and two Baronets) (Shrubsole, 2018, 2019). The public only have full statutory ‘Open Access’ to 4.4% of the Park (21,003 acres) (SDNPA, 2019). However, the area is crisscrossed by 3,218 km of public rights-of-way (TTC, 2018). Additionally, local authority owned country parks and land managed by the Forestry Commission (UK’s forest agency) bring a further 5% of the Park into public access (SDNPA, 2019), and some landowners also allow permissive paths. The South Downs Way, one of 15 UK national trails, is popular with walkers, cyclists, and horse-riders. In a 12-month period, 61,191 people passed one trail point (ESCC, 2016), locations close to carparks can be busier still (HCC, 2020).

Ecology and epidemiology of tick-borne diseases

Ticks (Ixodida) are second only to mosquitoes globally as vectors of human pathogens (Lawrie et al., 2004). Twenty species of ticks are native to Great Britain (Jameson & Medlock, 2011), 26 to northwestern Europe as a whole (Hillyard, 1996). Most are relatively host specific and primarily nidicolous (i.e., living in or near shelters used by their hosts), and therefore of minimal risk to humans (Gray, Estrada-Pena & Vial, 2014). In contrast to nidicolous species, some ticks feed on diverse host communities, climbing undergrowth or litter to attach to passing vertebrates, including humans. One of these, I. ricinus, is the most common tick species across temperate Europe, is known to feed on over 300 wild or domestic vertabrates (Földvári, 2016), and is the tick most often affecting humans and pets in the UK (Jameson & Medlock, 2011; Abdullah et al., 2016; Davies et al., 2017). As a three-host tick, it blood feeds once from a different individual host at each active stage of its development, but spends most of its average 4–6 year life-cycle off-host (developing, in diapuse, or questing for a new host) (Hillyard, 1996; Földvári, 2016). Off-host habitat suitability is strongly influenced by descication; if litter has low humidity and/or high tempreture for a protracted period survival is threatened (Hillyard, 1996). Thus, across its range it is mainly associated with woodland, but in the British Isles especially it also inhabits humid valley pastures and upland rough grasslands (Kahl & Gray, 2023).

Ixodes ricinus host selection is partly determined by the height to which a tick climbs vegetation in quest of a host. In Britain, its larvae and nymphs feed primarily on rodents and birds (e.g., wood mice (Apodemus sylvaticus), grey squirrel (Sciurus carolinensis), blackbird (Turdas merula), pheasant (Phasianus colchicus), red grouse (Lagopus lagopus scoticus)) but can also be found on larger animals. Adults, which tend to climb higher, mainly parasitise larger mammals (e.g., sheep (Ovis aries), deer (Capreolus capreolus, Cervus elaphus, and Dama dama), and cattle (Bos taurus) (Craine, Randolph & Nutall, 1995; Hillyard, 1996; Kirby et al., 2004; Lihou & Wall, 2022). At each feed, engorgement can take between 2 and 11 days depending on the life stage: larvae, 2–5 days; nymphs, 2–7; adult females, 6–11 (Balashov, 1972, as cited in Földvári, 2016). Transovarial transmission of some pathogens can sometimes cause larvae to hatch as infectious (Hauck et al., 2020). However, a nymph feeding on a human will also have had opportunity to become infected when feeding as larva, and an adult will have already had two blood meals, potentially from very different animals. Thus I. ricinus can act as a vector for the transmission of pathogens to humans from diverse taxa (Hillyard, 1996; Randolph, 2009b; Mannelli et al., 2012). For example, it is the UK’s most common Lyme disease vector (Jameson & Medlock, 2011; Medlock & Leach, 2015). Dobson, Taylor & Randolph (2011) found I. ricinus nymphal questing in recreational areas of Southern England peaked in most habitats in spring between April and June. During this period all life-stages are active, but I. ricinus questing then tends to reduce during the hottest part of English summers (to avoid desiccation), with a resurgence in the latter part of summer and early autumn (UKHSA, 2023; NHM, 2024). Questing continues through the remainder of the year, but at a significantly reduced level (UKHSA, 2023).

In three regions in England and Wales, patients consulted general practitioners about tick-bites at a rate of 54–204 per 100,000 inhabitants in 2011, 72.5% of respondents in Cumbria had removed ticks from patients 2011–13 (101/100,000 population) (Gillingham et al., 2020). This is only a partial glimpse of the full extent of bites. An estimated ⅓–⅔ of tick bites go unnoticed (Hofhuis et al., 2015), and this can be particularly so with bites by larvae and nymphs, which are smaller than adults. Even if noticed many do not seek medical advice. For example, a 2007 population survey in the Netherlands found a tick bite incidence of 7,198/100,000, c. 1.1 million bites were reported. This equates to approximately fifteen times the number of tick-bite related general practice consultations (Hofhuis et al., 2015). Lyme disease is the primary human tick-borne disease of concern in the UK. Cairns et al. (2019) used general practice data to estimate a 1-year Lyme disease incidence of 12/100,000 (cautious interpretation is warranted, 59% of these clinical diagnoses lacked documented laboratory confirmation). Besides the causative agent of Lyme disease, Borrelia burgdorferi s.l., other human pathogens have been detected in ticks and related hosts in the British Isles, including: spotted fever group rickettsia (Tijsse-Klasen et al., 2011, 2013), Borrelia miyamotoi (Hansford et al., 2015), tick-borne encephalitis virus (Holding, Dowall & Hewson, 2020), and Babesia venatorum (Gray et al., 2019; Weir et al., 2020). In 2019–20 the first presumed autochthonous human cases of tick-borne encephalitis and babesiosis were recorded in the UK (PHE, 2020a). Some of these recently detected health threats may result from emerging foci of imported pathogens. However, it is also possible that in addition to Lyme disease, there may be considerable levels of undiagnosed tick-borne infections affecting persons in the British Isles.

Public Health England have mapped UK tick distributions at 10 km² resolution by combining historical records (Pietzsch et al., 2005) with samples sent by the public, who most commonly found them on themselves, other human hosts, or their domestic pets, and to a lesser extent on livestock, and wild mammals and birds (Jameson & Medlock, 2011). UK general practice records of arthropod bites do not identify by species (Newitt et al., 2016), and though Hospital Episode Statistics (HES) have been used to map Lyme disease distribution (Cooper et al., 2017), HES records residential postcodes of patients, not where they were bitten. Thus, whilst HES is valuable to understanding disease burden, given UK tick-borne infections are very often linked to recreational exposure (Dobson, Taylor & Randolph, 2011), it’s use to map differing geographic tick hazard is limited. This is especially true in places such as the SDNP, with high numbers of regional, national, and international visitors.

Throughout this article we use ‘tick hazard’ to refer to tick species that routinely parasitise humans. Beyond simple presence/absence, tick density (‘an absolute measure describing the number of ticks per area unit’ (Kahl et al., 2002)) is the most reliable metric of site tick hazard (Ostfeld, 2011). However, knowledge of tick density is by its nature restricted to the relatively small number of places in Britain actively field-sampled. One of the most used density metrics is density of nymphs (DON) (Han et al., 2021). As Braks et al. (2016) discuss, though this is often reported as a risk metric, it is better understood as a hazard proxy. Given, as they state, a ‘hazard may not lead to a risk’ in the absence of someone being exposed to the hazard, risk at its most basic is ‘exposure times hazard’. An understanding of tick-bite risk at a particular site will then involve both ecological determinants of tick hazard and social determinants of exposure (e.g., visitor numbers, extent of use of protective clothing, etc.). In addition to DON, some field studies also report the metric density of ticks (DOT), which is the combined density of all life-stages of ticks collected (e.g., larvae, nymphs, adults) (for an example, see Han et al., 2021).

Tick hazard in the SDNP

The SDNP is a priority area for interventions that reduce tick-borne disease hazard whilst preserving ecosystem health. Prior to our study its downland section had been highlighted by the Health Protection Agency (part-precursor to Public Health England) as a ‘regional foci of Lyme borreliosis’ (HPA, 2012), whilst West Sussex was listed alongside the South Downs as one of 10 areas in England and Wales where Lyme disease infection was most frequent (HPA, 2010). Yet despite this and the Park’s very large visitor numbers, prior to our study no multi-site field sampling of tick hazard in the SDNP, or comparison of hazard between its key habitats, had been published. Elsewhere woodland has been linked to increased tick-borne disease hazard, specifically Lyme disease (Gray et al., 1998; Killilea et al., 2008; Takken, 2016; Van Gestel et al., 2021; Bourdin et al., 2023), though controversy remains over causal pathways (Levy, 2013). For example, research linking forest fragmentation to increased Lyme disease hazard (summarised best in Ostfeld (2011)) has been criticized by UK researchers (Randolph & Dobson, 2012). Sheep grazing supports vector populations in some UK grass uplands, and though not host competent for Borrelia burgdorferi s.l., sheep can support transmission cycles via tick co-feeding (Ogden, Nuttall & Randolph, 1997), and also host Babesia venatorum (Gray et al., 2019). However, compared to wildlife, the role of livestock in propagation of tick-borne diseases of human concern is under-researched (Stanek et al., 2012). Increased wildlife populations have been implicated elsewhere in rising incidence of tick-borne disease (e.g., Crimean-Congo haemorrhagic fever in Turkey (Randolph, 2009a); tick-borne encephalitis in East Europe (Randolph, 2009b)) setting up a potential conflict between biodiversity and human health. Given UK National Parks aim to enhance wildlife and encourage public enjoyment of the countryside (NPUK, 2017), such conflict would be problematic for the South Downs National Park Authority (SDNPA) and the local governments from which most of its members are drawn. However, its joint remit, bioregional framing, and coalition of stakeholder members makes it the ideal body to link and champion site-based and regional policies to reduce hazard, whilst avoiding or reducing conflict between public health and ecosystem health.

Aims

Our overall project (Tick-borne hazards in the SDNP and the potential for Planetary Health based interventions) includes (1) mapping and fieldwork to better understand tick hazard across the SDNP, including crucially at key potential locations for future interventions, (2) a systematic review of proposed interventions to reduce site hazard of the most common tick-borne disease in Britain, Lyme disease, with a focus on those actions not expected to negatively affect ecosystem health. Here, we report on our mapping and fieldwork, information on our systematic review can be found in Middleton, Cooper & Rott (2016).

Study objectives:

• identify and describe potential key locations for future interventions;

• map distribution of tick hazard across the SDNP;

• determine tick hazard (species and density) at potential intervention sites; and

• analyse habitat associations with tick hazard in the SDNP.

Sites selection for drag-sampling and potential future interventions

Five sites were selected: three prospectively, and two responsively after submission of ticks obtained by deerstalkers from sentinel deer (Fig. 1). The three prospectively chosen sites were located one in each of the SDNP’s three counties. We took this approach so as to sample from along the National Park’s length, and because one of our project’s primary audiences is county authorities which manage countryside sites within the SDNP with high numbers of recreational visitors (e.g., UK accredited country parks as defined by NE and DEFRA (2014)). These authorities are key to implementing potential interventions to reduce tick-borne disease risks in the SDNP as they elect governing members to SDNPA (responsible for strategic action across SDNP), and directly manage downland and woodland sites with high visitor numbers where interventions could be trialled.

Prospectively selected

Of the three counties within SDNP’s borders, two county councils managed country parks in the National Park: Hampshire County Council (SDNP’s western section), and East Sussex County Council (SDNP’s eastern section). West Sussex County Council (SDNP’s central section) does not perform this function within the National Park. The SDNP’s ranger service was consulted about which Hampshire County Council and East Sussex County Council sites had the highest visitor numbers (subsequently confirmed by councils themselves).

East Sussex County Council’s site with the highest annual visits (visitor details, and population and proximities of nearby towns, Table 1), was the 280 ha Seven Sisters Country Park (sevensisters.org.uk) in the SDNP’s eastern section. It is composed of chalk grassland, saltmarsh, shingle seashore, woodland, and a meandering river. Conservation designations include SSSI, Area of Outstanding Natural Beauty (AONB), Heritage Coast, and Marine Conservation Area. Transects (Figs. 2A–2F) consisted of sheep grazed chalk downland and woodland, primarily beech (Fagus sylvatica) and sycamore (Acer pseudoplatanus) (Table S1). Hampshire County Council’s site with the most visitors (Table 1) was the 564 ha Queen Elizabeth Country Park (https://www.hants.gov.uk/thingstodo/countryparks/qecp) in the SDNP’s western section. It consists of downland and wooded hills, designations include: SSSI, National Nature Reserve, Special Area for Conservation, Scheduled Ancient Monuments. All transects (Figs. 2G–2L) were in woodland (beech, conifer, or hazel (Corylus avellana)). Some had sparse undergrowth with dense beech or conifer litter, others nettle patches (Urtica dioica) or bramble thickets (Rubus fruticosus agg.) (Table S2). Given West Sussex County Council do not manage an appropriate site for sampling, a third site was chosen at SDNP’s centre which represented a sizeable wealden woodland owned by a key Park stakeholder (Sussex Wildlife Trust). The Mens is a 166 ha nature reserve of largely unmanaged wealden forest (sussexwildlifetrust.org.uk/visit/the-mens) (visitor details, Table 1). The site is especially rich in plants, saproxylic invertebrates, and fungi (c. 600 species). The sampled transects (Figs. 3A–3F) followed footpath borders tufted with grass, and cut across ground with sparse undergrowth under high canopies of predominantly beech, and sections with dense waist-high brambles (Table S3).

Table 1:

Visitor information for sites drag-sampled.

Site	Nearest large populations	South Downs Way	Annual visitors	Carpark spaces (carparks, total m2)
Queen Elizabeth Country Park	Portsmouth: 19 km, est. 2019 pop. 229,851 (ONS, 2020).	Yes	202,559 vehicle entries March 2019–April 2020 recorded by number plate recognition system (HCC, 2020).	401 (8, 9,947)
Seven Sisters Country Park	Eastbourne: 8 km, est. 2019 pop. 114,809 (ONS, 2020). Brighton and Hove: 24 km, est. 2019 pop. 244,917 (ONS, 2020).	Yes	52,124 used visitor centre Jan–Dec 2019 (T. Wallace, 2020, personal communication), but only a minority of trips expected to include this.	292* (3, 6,542)
Ditchling Beacon Nature Reserve	Brighton and Hove: 4 km, est. 2019 pop. 244,917 (ONS, 2020).	Yes	Not recorded by land manager.	44* (2, 1,107)
Cowdray Estate	Chichester: 9 km, 2011 pop. 26,795 (ONS, 2015).	Yes	Not recorded by land manager.	10* (1, 315)
The Mens	Horsham: 18 km, 2011 pop. 49,000 (HDC, 2016).	No	Not recorded by land manager.	6* (1, 175)

DOI: 10.7717/peerj.17483/table-1

Note:

* As only Queen Elizabeth Country Park (QECP) had individual marked-out car spaces in its carparks, space per m² ratios from them were used to estimate capacity of each carpark at the other sites. In each case, the ratio chosen was from the QECP carpark closest in square metres to the unmarked site carpark (Article S1).

Tick collection from deer

Deerstalkers were recruited through the British Deer Society (bds.org.uk) newsletter and website, sent kits, and asked to collect ticks from deer culled for reasons unrelated to this project. Participants were instructed to inspect the whole animal, collect every visible tick, place them in pre-coded 1 ml cryovials (pre-filled with 0.5 ml 70% ethanol) and return by post (safety measures, Article S1). On receipt cryovials were deposited in a laboratory fridge (approx. 5 °C), and after identification transferred to a freezer (approx. −20 °C). Deerstalkers recorded: habitat type; deer species; six-figure grid reference (using ‘OS Locate’ (Ordnance Survey, London, 2014)); body sites ticks found at; and whether ticks were attached or not.

Tick collection by drag-sampling

Sites were sampled April to November inclusive. To collect questing ticks, sampling was not carried out when air temperature was <7 °C 50 cm above the ground or when vegetation was wet from recent rain/dew, as per James et al. (2013). Four sites were sampled in both 2015 and 2016, with an additional site sampled in 2016. At each, six 1 m × 50 m transects were sampled as per Dobson, Taylor & Randolph (2011). The first two transects chosen were those suspected to have the highest potential exposure of humans to ticks (e.g., vegetation alongside a footpath). Where sites included grassland and woodland, one chosen transect was selected from each. All others were selected using dice and a random number table. To reduce spurious conclusions from single sampling, each transect was planned to be sampled twice yearly, for 2 years. Over the 2 years of fieldwork each of the four sites analysed for difference were sampled during a 58-day spring period (beginning 23^rd April and ending by close of UK astronomical spring around 21^st June), and a 59-day period in the second half of summer and very early autumn (starting 2^nd August, ending 29^th September). This sampling was not conducted in between as early to mid-summer includes July which is generally the hottest UK month (MO, 2024), and known to see less I. ricinus questing. Additionally, the spring period included UK May public holidays, whilst the second half of summer and very early autumn period included most of the UK school summer holiday. Thus, these two seasonal sampling periods are of particular importance for establishing UK tick hazard due to known peaks in both tick questing and people using the countryside for recreation. At one private site used for game shooting, it was not possible to visit twice in 2016 due to a requirement to be accompanied by a deerstalker with restricted availability, and this also delayed some sampling.

To improve chances of picking up disease signals in planned follow-up research (usually only a minority of ticks at any site are infected (Vollmer et al., 2011)), extras were acquired by drag-sampling between transects, and at follow up visits where possible. Tick sampling techniques differ in efficacy and are affected by habitat/vegetation type (Dantas-Torres et al., 2013). To reduce bias ticks were collected simultaneously along transects using a blanket drag, heel flags, and chaps (all made from wool, Fig. 4). Wools were examined after each transect, ticks placed individually in 70% ethanol filled micro-centrifuge tubes, deposited same day in a laboratory freezer (approx. −20 °C). At each transect, at each sampling, photos were taken along with field notes including: ticks collected; date/time; weather; habitat; visitors observed; dominant vegetation; vegetation height; main litter constituents; relative humidity and temperature (both at 50 cm and in litter, measured with a Fisher Scientific Traceable Hygrometer). Locations were recorded (10-figure OS grid references, bearings) using ‘OS Locate’ and ‘OS Mapfinder’ (Ordnance Survey, London, 2014) on a Samsung Galaxy Note II phone (Researcher safety and inter-site contamination control, Article S1).

Tick identification

Identification was conducted in-lab with a hand lens (Hilkinson Ruper x20 15 mm achromatic), and where necessary a dissecting microscope (Leica EZ4; Wetzlar, Germany). A species key was used (Hillyard, 1996), and identification aided by reference to Baker (1999) and Beati, Needham & Klompen (2016). Species and life stage was recorded, adult ticks sexed. Larvae were not fully keyed as clearing for slide mounting would have reduced the sample pool available for future pathogen detection. However, each larva was inspected for characters which identified them to genus, and indicated likely species. If nymphs/adults of more than one species were identified at any site, 10% of larvae from that site (to a maximum of 50) would have been slide mounted and keyed. A limitation of many similar studies has been not enabling retrospective evaluation of species identification (Estrada-Peña et al., 2013). To corroborate identification, voucher specimens (including all life stages/sexes) have been deposited into the University of Oxford Museum of Natural History Arachnida Collection (Accession Lot No. ENT-OUMNH-2024-009).

Mapping distribution of tick hazard

Using ArcMap 10.7 (ESRI, Redlands, CA, USA) sites where ticks had been submitted by deerstalkers or drag-sampled by JM were mapped, indicated by points at 100 m² resolution. In addition, to map recorded presence of tick hazard at 10 km² resolution by species, data from the following sources were compared and combined as layers: (1) the most recent published Public Health England/Health Protection Agency tick maps for England and Wales (Cull et al., 2018; PHE, 2016; HPA, 2013a, 2013b, 2013c, 2013d) (2) National Biodiversity Network Atlas (NBN, 2020) (which includes historic data 1890 onwards, and into which Public Health England now submits tick records (PHE, 2020b)), (3) a single site drag-sample in 2014 by Layzell et al. (2018), (4) point locations of ticks submitted from culled deer or collected by drag-sampling in this project 2015–16, (5) point locations drag-sampled for Haemaphysalis punctata by Public Health England and the Animal and Plant Health Agency, primarily 2015–18 (Medlock et al., 2018), (6) records from pan-species surveying at Sussex Wildlife Trust reserves, mainly 2016–17 (previously unpublished data collected by Trust volunteers and staff and supplied directly to the authors for inclusion in this study). Digital basemaps were obtained from OS OpenData (2020) and Natural England (2020). (Layer generation detailed in Article S1).

Site visitors

In addition to receiving guidance from rangers and local authorities on which sites they managed in the SDNP had the highest amount of visitors, land managers were asked for annual site visitor numbers when available. Population sizes of nearby towns were taken from reports of the UK Office for National Statistics (ONS). All five drag-sampled sites were rural, and most visitors annually could be expected to arrive by private vehicles. As Weitowitz et al. (2019) demonstrated in their nationwide survey of parking provision at UK nature conservation sites, car park capacity is broadly predictive of differences in visitor numbers. Thus, to indicate relative difference in number of expected site visitors (and thus exposure to site tick hazard), all car parks (15 in all) at each of the five sites were measured on 25 December 2023. Car space per m² ratios from Queen Elizabeth Country Park’s eight carparks (the only ones with marked-out individual car spaces, Article S1) were then used to estimate number of carpark spaces at the other sites (Table 1). No changes to the carparks had been made in the period between the tick survey and carpark measurements.

Analysis

As well as site vector presence/absence, for sites drag-sampled both years tick hazard was assessed as (1) questing density of ticks, all life stages (DOT), and (2) questing density of nymphs (DON). These were calculated as means of totals of four site samplings: six 1 m × 50 m transects, sampled twice yearly for 2 years. Additionally, DOTs and DONs were calculated for each of the two seasonal sampling periods. To determine significance of difference of tick (all life stages) and nymph counts between the four sites sampled in both years, Kruskal-Wallis Tests with Bonferroni corrected post-hoc Dunn’s Tests were carried out on counts from all individual transect samplings (i.e., 90 50 m × 1 m drag-samplings). Further Kruskal-Wallis Tests with Bonferroni corrected post-hoc Dunn’s Tests examined habitat types determining tick (all life stages) and nymph hazard. Statistical analysis was carried out in Minitab17 (Minitab Inc, State College, PA, USA). Drag-sampling results from the additional site sampled in 2016 are reported as presence/absence, and in distribution mapping, but were not included otherwise in analysis (justification of analysis, Article S1).

Extent of tick hazard across the SDNP

Distribution and species

Ticks collected by drag-sampling or submitted by deerstalkers confirmed presence across much of SDNP (Fig. 1). Ticks were present in both of its characteristic habitats: sheep grazed downland, and wealden woods. Ticks were found at all sites drag-sampled (though not on all transects, Figs. 2 and 3; Tables S1–S4). All ticks submitted from deer were Ixodes ricinus (Fig. 5A), also the only species collected at three of the four sites drag-sampled in both 2015 and 2016 (Queen Elizabeth Country Park, The Mens, Cowdray Estate). The nationally rare H. punctata (Fig. 5B) was the sole tick collected from the remaining site sampled in both years (Seven Sisters Country Park), and was also found at Ditchling Beacon Nature Reserve in 2016.

Figure 6 maps distribution of tick records at 10 km² resolution. The first report for I. ricinus is from 1964, 33/37 of the Park’s grid squares have had at least one record (often multiple) in the last 15 years (Fig. 6A). In contrast, Ixodes hexagonus (the UK’s second most common Lyme disease vector (Jameson & Medlock, 2011; Medlock & Leach, 2015) has been recorded far less, most squares where presence has been recorded represent historic records only (Fig. 6B). The earliest H. punctata report is from 1920, but all related grid squares have had recorded presence in the last decade, mostly in its known foci in the far east of the Park (Fig. 6C). Locations included from recent drag-sampling by Public Health England and Animal and Plant Health Agency suggests it has spread westwards somewhat, and this observation by Medlock et al. (2018) is confirmed by drag-sampling in our study at Ditchling Beacon Nature Reserve which extends its known range further still, as does a previously unpublished isolated recording by Sussex Wildlife Trust 44 km further west. A second rare species in the UK, Dermacentor reticulatus, was recorded by Sussex Wildlife Trust at one of its West Sussex reserves in 2004 (Fig. 6D). To our knowledge D. reticulatus has not otherwise been recorded in the SDNP or its constituent counties; this record was not previously included in National Biodiversity Network Atlas or Public Health England/Health Protection Agency published mapping.

Ticks collected from deer

Eighty-seven ticks were submitted (Table 2) obtained from 14 deer at 12 locations (Fig. 1). All but one were adult ticks, and 82% of those were females. Most males were attached in mating; the majority of females were engorged (73%, n = 71, two not recorded by collectors), as was the sole nymph. The commonest habitat ticks were collected off deer at was ‘wood’ (59%, n = 87). Most of the remaining were from deer shot at mixed edge-habitats involving wood (28%, n = 87), e.g. ‘wood-heath’. Most hosts were fallow deer (Dama dama) (10, 71%), followed by roe (Capreolus capreolus) (3, 21%) (n = 14, one host species not recorded by collector). Tick burden was 1–35 per deer. The commonest attachment sites were abdomen and sternum, and posterior and frontal axillae (Fig. 7).

Table 2:

Ticks submitted by deerstalkers.

n = 87		Tick (%)
Tick species
Ixodes ricinus		87 (100)
Tick life stage and sex
Adult females		71 (82)
Adult males		15 (17)
Nymph		1 (1)
Larvae		0 (0)
Female engorgement (n = 71)
Females engorged*		52 (73)
Females not engorged		19 (27)
Not recorded		2 (3)
Habitat
Wood		51 (59)
Wood-chalk grassland		15 (17)
Wood-pasture		7 (8)
Pasture		6 (7)
Wood-arable		3 (3)
Wood-heath		3 (3)
Other		2 (2)
Host species
Fallow deer, Dama dama (n = 10)		45 (52)
Roe deer, Capreolus capreolus (n = 3)		38 (44)
Not recorded (host n = 1)		4 (5)
Range of ticks per host		1–35

DOI: 10.7717/peerj.17483/table-2

Notes:

Deer culled for reasons unrelated to this project. Percentages rounded to whole numbers. Engorged status covers adult females and nymphs only, as adult males do not engorge.

* in addition the single nymph was engorged.

Ticks collected by drag-sampling

Drag-sampling four sites in both 2015 and 2016 collected 622 ticks (Table 3). Of these, 237 were along transects and are included in calculations of vector densities and analysed statistically. A total of 385 extras were stockpiled to aid future pathogen detection. Ticks were present at all four sites and the additional site sampled twice in 2016 (one adult only). Of ticks collected by drag-sampling at all sites (n = 623), most were nymphs (53.8%), followed by larvae (42.5%), and a small number of adults (3.7%, 13 males, 10 females (Tables S1–S4). 93.3% (222) of ticks gathered along transects (n = 238) had attached to woollen blankets, 6.7% (16) to woollen chaps, and none were attached to flags (Table 3). Figure 8 shows a breakdown of tick data from transects by collection month, and by tick life stage and species (all sites sampled in both 2015 and 2016, n = 237). Of these, H. punctata only made up a small number (1.7%, four; two nymphs and two adults, both collected in the spring sampling period), whilst I. ricinus made up the vast majority (233, 98.3%). Most of these I. ricinus were nymphs (64.8%, 120 from spring, 30 from the late summer and early autumn sampling period). In addition, 72 (30.9%) were larvae (spring, four; late summer early autumn, 68), and 10 (4.3%) were adults (spring, seven; late summer early autumn, three).

Table 3:

Ticks collected by site drag sampling.

			On-transect				Off-transect				Site totals
			Life stages and sex				Life stages and sex
	Species	Collected on	Larvae	Nymphs	Adults	Totals	Larvae	Nymphs	Adults	Totals*
Seven Sisters Country Park Six 50 m2 transects, each sampled four times (twice each in 2015 & 2016).	H. punctata	Blanket		2		4	65	8	1♀	74	78
		Chaps			2♀
		Flags
Queen Elizabeth Country Park Six 50 m2 transects, each sampled four times (twice each in 2015 & 2016).	I. ricinus	Blanket	37	39	1♀ 4♂	89	25	65	3♀ 1♂	94	183
		Chaps		8
		Flags
The Mens Six 50 m2 transects, each sampled four times (twice each in 2015 & 2016).	I. ricinus	Blanket	35	78	1♀ 3♂	121	100	104	2♀ 3♂	209	330
		Chaps		4
		Flags
Cowdray Estate Six 50 m2 transects, each sampled three times (twice in 2015, once in 2016).	I. ricinus	Blanket		21	1♂	23	3	5		8	31
		Chaps		1
		Flags
Ditchling Beacon Nature Reserve Six 50 m2 transects, each sampled two times (twice in 2016).	H. punctata	Blanket			1♂	1				0	1
		Chaps
		Flags
		Totals	72	153	13	238	193	182	10	385	623

DOI: 10.7717/peerj.17483/table-3

Note:

Results of individual samplings in Supplemental Material, Tables S1–S4.

* Off-transect ticks were those additionally collected opportunistically for later pathogen detection, and were not included in calculations of DON or DOT.

Tick hazard at potential sites for interventions

Figures 2 and 3 and Tables S1–S4 show the number of ticks obtained along transects at each sampling (the latter also provide collection dates and environmental data).

Habitat associations with tick hazard in the SDNP

Of the four sites drag-sampled in both years, sites with transects entirely in woodland (Queen Elizabeth Country Park, The Mens) had the highest tick hazards (Tables 3 and 4, S2 and S3). However, tick hazard was present at all sites, including on the grazed downland sections of the two sites (Seven Sisters Country Park; Cowdray Estate) which had transects in downland and woodland (Tables S1 and S4). The tick hazard present was not universal in wooded sections of sites; no ticks were found in the forested part of Seven Sisters Country Park (Table S1). A KWT performed of ticks collected (all life stages) on habitat coded transects showed average ranks differed significantly (<0.05) for at least one of the three coded habitat types (results, Table 4C). A Bonferroni corrected post-hoc Dunn’s Test (Table 4C; sign CIs and magnitudes/directions of differences, Fig. S3) showed there was not a statistically significant difference in the number of ticks (all life stages) between deciduous woodland and conifer woodland/planting. However, both habitats had significantly more ticks than downland. Similarly, a KWT showed average ranks differed significantly (<0.05) for nymphs (Table 4C), and a Bonferroni corrected post-hoc Dunn’s Test (Table 4C; sign CIs and magnitudes/directions of differences, Fig. S4) showed no significant difference between deciduous woodland and conifer woodland/planting. However, as with ticks (all life stages) both habitats had significantly more nymphs than downland.

Tick hazards, associated pathogens, and general management within the SDNP

Ixodes ricinus or H. punctata ticks were recorded at all sites drag-sampled, on some transects in high numbers (Figs. 2 and 3). Extent of tick hazard differed between sites. It was significantly higher in woodland compared to grazed downland, but was still present at all downland sites. Our mapping indicates tick hazard is widely distributed across the SDNP, as confirmed by the deerstalker submission of ticks from sites across the National Park.

Recommendations for key locations for future interventions in the SDNP

Recommendations for actions beyond the SDNP

This work is part of a regional intervention planning exercise, so findings are highly site specific limiting generalisation. However, this study may be useful as a model for intervention planning elsewhere, and the data is available for meta-analyses. Specifically, we would encourage National Park authorities in the UK and beyond to directly support similar surveillance and planning exercises in their territories as a foundation for taking action when tick hazard is present. We are cautious about making recommendations for specific control measures beyond the bioregions our study is grounded in. Nevertheless, some of the management suggestions we have provided may be appropriate elsewhere. For example, edge-mowing alongside the most heavily used paths in country parks, and protective acaracide-treated clothing for countryside workers. These have minimal cost implications and potentially high impacts in reducing exposure (de Groot, 2016; Verheyen & Ruyts, 2016), so adoption would seem reasonable based on existing (though limited) evidence. In contrast, the larger more landscape-scale interventions we have recommended for some of our drag-sampled sites (i.e., predator introduction; pasture spelling) could be expected to have more implications, and whilst there is good reason to think they may be effective at some sites (with some tick species) (Hillyard, 1996; Hofmeester et al., 2017), the evidence base is not yet strong. We would still encourage others to consider these measures if they are managing similar habitats to the SDNP (deciduous woodland with high masting, sheep grazed grasslands, etc.), but only as part of rigorous trials. This is our intention in the South Downs, and we are keen to be in contact with others doing similar projects so that we can build the evidence base together.

Strengths and weaknesses

One study strength was deerstalker involvement, whose submissions enabled responsive site selection and contributed to SDNP wide tick hazard mapping. Another strength was our use of density of ticks, rather than only density of nymphs. Starting in the USA with Lyme disease the latter is more common (Han et al., 2021). However, H. punctata presence in the Park illustrates the reduced appropriateness of reporting only density of nymphs for European work. Transovarial transmission of spotted fever group rickettsiae is established, Macaluso & Paddock (2014) so a hazard metric that excludes larvae (as density of nymphs does) will under-count vector density for spotted fever group rickettsiae and other pathogens (e.g., B. miyamotoi (Hansford et al., 2015)). Similarly, tick-borne encephalitis virus cycles require larvae-nymph co-feeding (PHE, 2019). The idea, based on US data, that transovarial transmission of B. burgdorferi s.l. is rare or non-existent (Ostfeld, 2011, p.43) is the main basis on which density of nymphs has been used for Lyme disease site hazard assessment. However, Van Duijvendijk et al. (2016) showed Borrelia afzelii, a common cause of Lyme disease in Europe which is not present in the US, can be transmitted by I. ricinus larvae. In a UK study by Hall et al. (2017), 0.7% were B. burgdorferi s.l. positive. A study weakness concerns number and months of repeat sampling. Firstly, our first four sites were intended to be drag-sampled twice per year, for 2 years. However, this was not possible for one site, which was sampled three times only, and with one sampling-day delayed (Fig. 8). Secondly, some species lifecycles (most notably, I. ricinus) cause seasonal differences in questing tick numbers/life-stage proportions. However, for logistical reasons the months’ individual site samplings took place in varied. This was mainly due to between-site distances and the need to avoid sampling during or after rain (which reduces tick questing and drag-sampling efficacy), which in both seasonal sampling periods is generally highly frequent and changeable spatially across the 140 km length of the Park. This reduces confidence in validity of site comparisons somewhat, though this is partially balanced by in-year and multiple-year repeat samplings. Site DON and DOT measures from both seasonal sampling periods accorded with rankings of overall DON and DOT per site, suggesting multi-year rankings were not overly skewed by sampling-date difference. Nevertheless, increased confidence in inter-site comparisons could be obtained in future studies by simultaneous sampling by multiple fieldworkers working across the sites on days suitable for cross-Park sampling. In this study, intervals between site samplings were not uniform, which introduces a risk that a depletion effect may differentially bias measures of hazard across sites. Future work would thus benefit from fixed uniform intervals, which could be enabled by fieldworkers working simultaneously across the Park.

We did not ourselves survey visitor numbers, and such data was not available from land managers for most sites. For this reason, though we do infer which sites are the priority for public health measures based partly on visitor related data (annual visitor totals, when available; relative site difference in carpark capacity; nearby conurbations), we could not quantify tick risk itself. To enable this, similar studies would benefit from parallel site surveys of tick densities (i.e., hazards) and visitor numbers (i.e., exposures), though this would require increased resource. GIS tick hazard mapping successfully brought together all publicly available, and some unpublished, tick records from SDNP. However, to our knowledge vouchers are unavailable for the historic and pan-species records included, leaving some uncertainty regarding correct species identification. In contrast a strength of our fieldwork is our vouchers.