Oral contraceptives are being developed to manage the impacts of the grey squirrel. To be effective, contraceptives will need to be deployed at a landscape scale and will require a delivery system that is practical and economically viable. Understanding grey squirrel feeding behaviour is important so that delivery methods can be designed to ensure enough target individuals receive an effective contraceptive dose at a time of year that will ensure they will remain infertile throughout peak times of breeding. The main aims of this study were to assess how sex, season, squirrel density and bait point density influenced; 1) the probability of a squirrel visiting a feeder and 2) the amount of bait consumed from feeders. Field trials were conducted on six woodland populations of squirrels in three seasons, with four days of bait deployment via purpose-designed squirrel-specific bait hoppers with integrated PIT-tag readers.

Trials were conducted in five independent woods that were located at least 3.5 km apart and within 25 km of York, England (53.96^oN, -1.09^oE; Figure 1). Populations of grey squirrels were sampled from woods HW and SC in winter 2017/2018, woods SC and PW in summer 2022 and woods GE and BW in spring 2023. Autumn was not trialed due to reduced squirrel breeding activity and high natural food availability, which could potentially deter squirrels from using feeders. To provide two suitable replicates for each season, woods were selected that were of a similar size (between 7 and 8 ha) and structure (mature broadleaf or mature broadleaf conifer mixed tree species), that had high levels of grey squirrel activity reported by landowners and established squirrel populations, with minimal squirrel control conducted within the previous 12 months. At the end of each trial, the squirrel population was trapped and removed from each wood as part of other related studies. Therefore, wood could not be re-used in trials that were within 12 months of each other. The squirrel population at Wood SC was removed in January 2018 and the wood was recolonized by squirrels, confirmed by the trap results in this study, so was reused in Summer 2022. Thus, it was considered that six independent populations of squirrels were assessed. 

Grey squirrel trapping and sampling methods were approved by the joint Animal and Plant Health Agency (APHA) and Fera Science Ltd Animal Welfare and Ethical Review Body (AWERB). All animal studies were conducted in accordance with the UK Guidance on the operation of the Animals (Scientific Procedures) Act 1986. Field trials were conducted from December 2017 to February 2018 in woods HW and SC, from June to August 2022 in woods PW and SC, and from April to June 2023 in woods BW and GE. Single-catch squirrel cage traps (density = 3/ha, n = 21-24) were deployed across the whole area of each wood on 1-meter-high wooden stands and pre-baited with a combination of maize, peanuts, a small number of whole hazelnuts and approximately 2 grams of 100% hazelnut paste (Bulk^TM, UK) every 2 to 4 days for between 3 and 13 days, dictated by human resource availability. For animal welfare and health and safety reasons, traps were not set if heavy rain, high winds (> 30 mph) or high (> 30^oc) or low (< 2^oc) temperatures were forecast. Each trap was partly covered with a waterproof sheet to provide animals with shelter. Traps were either set early in the morning and checked in the afternoon or set late afternoon and checked the following morning.

Trapped squirrels were anaesthetised on site using isoflurane via a mask and a PIT-tag (Identichip^®, York, UK) implanted subcutaneously in the scruff of the neck. Squirrels were then sexed, weighed (to the nearest gram) and recorded as adult or sub-adult (< 450 g, slim, soft pelage, body lacking muscle, head large in proportion to body and female nipples or male testes not visible). Females were assessed for evidence of recent breeding (extended or lactating nipples and palpable foetuses). Hair was clipped from the tail for visual identification. Once recovered from anaesthesia, squirrels were released under a Natural England licence in the location at which they had been trapped. Trapping was conducted for 3 to 4 days, completed within 18 days, and then the traps removed.

In spring and summer, within two days of PIT-tagging, hoppers were deployed at each wood, while in winter, hoppers were deployed within four weeks of PIT-tagging. Hoppers were deployed evenly across the whole area of each wood, with locations guided by a 1 ha grid generated in ArcGIS (version 10.7.1) overlayed onto a satellite map using the ArcGIS Collector mobile phone application. Locations were, however, adjusted according to accessibility; for example, steep slopes or thick vegetation were avoided. Each hopper was fixed to a 1-metre-high wooden stand to reduce non-target access. In winter and summer, hoppers were deployed at a density of 3/ha (n = 21-24). Squirrel hopper use data from these studies suggested that lowering the hopper density to 2/ha could be more cost-effective for bait delivery in terms of field hours and bait quantity required. This was tested in spring 2023, using a Before-After-Control-Impact (BACI) study design. At woods GE and BW, hoppers were deployed at a density of 2/ha (N = 14) and baited for four days, then increased to 3/ha (N = 21) at BW and maintained at 2/ha at GE for an additional four days each. Each hopper was then baited daily with 40 g of hazelnut paste mixed with the bait marker dye Rhodamine B (Merck Life Sciences UK Limited, Dorset, UK) at a concentration of 0.18% for four consecutive trial days in winter and summer and eight consecutive trials days in spring. The bait was manually weighed in and out of the hoppers each day, and the hopper entrance and stand checked for evidence of bait spillage, easily identified due to the pink colouration of the bait. Any spillage found was weighed.Hoppers recorded the date and time at which an individual PIT-tagged squirrel was detected and the date and time at which the bait door was opened, through the engagement and disengagement of a magnetic door switch. In the summer and spring trials, hoppers were also fitted with a strain gauge, which recorded the weight of the bait taken for each squirrel visit.

Within 4 days of hopper deployment in spring and summer, and within 21 days of hopper deployment in winter, squirrel live-capture cage traps were installed in each wood at a density of 2-3 traps/ha. Traps were secured to the same 1-metre-high wooden stands used in the hopper trials. Trapping was conducted using the same protocol as the trapping for the PIT-tagging, though for this session, traps were set and checked at least once every 24 hours where possible. If weather conditions did not permit trapping overnight, traps were set early morning and checked late afternoon the same day. Trapping was conducted within a four-week period, typically Monday to Friday, for a minimum of 5 days and for a maximum of 14 days, or until squirrel capture rates were reduced to an average of less than one per day over three consecutive days. Squirrels that were trapped were humanely dispatched using a UK Home Office-approved (Schedule 1) method, by a trained and competent person, and the PIT-tag ID, sex and tail clip (if present) were recorded and at least 20 hairs were taken from the flank, placed in a plastic sample bag, to be later analysed using a Leica DMLB ultra-violet microscope (Leica Microsystems UK Ltd) at ×4 magnification for the presence or absence of RB fluorescence.