Surrey Harvest Mouse Project Report for People’s Trust for Endangered Species David Williams CMIEEM January 2015 Surrey Wildlife Trust, School Lane, Pirbright, Woking, Surrey GU24 0JN 1 The Harvest Mouse (Micromys minutus) project began in 2011 as a result of a successful application for funding to People’s Trust for Endangered Species. Surrey Wildlife Trust also part funded the project. Although populations are reported to have declined by 71% (Sargent et al 1997) very little successful survey work has been undertaken in Surrey. Lack of survey information about Harvest mice numbers is probably due to the ineffectiveness of survey and trapping techniques (Riordan et al 2009; Poulton and Turner 2009 and Meek 2011). Nest searches can be overestimated with Harvest mice building several nests for each brood (R.C.Trout and S.Harris, Mammals of Great Britain 2008). The project used several types of surveys and different trapping methods. The aims were to enhance our knowledge and status of the species (which would inform management plans to ensure sensitively managed habitats) and raise public awareness of the conservation status of the Harvest mouse, using them as a flagship species to promote landscape scale conservation of floodplain grazing. These aims would be achieved by trapping, nest surveys, owl pellet dissection, positioning hair tubes and camera traps. We trialled three types of small mammal live traps at two different heights (one being at floor level). Nest searches were undertaken and owl pellets were collected and dissected to search for Harvest mice remains. Hair tubes were trialled at various heights, along with trail cameras placed in the stork zone. Sites were chosen by using any one of three criteria: 1) Habitat looked suitable 2) Nests had been found 3) SWT held previous records. Volunteers were recruited before the project started and were invited to our first training day. Twenty-five attended. Attendees learnt about the ecology of the Harvest mouse, our trapping methods and owl pellet dissection. We also had live captive-bred Harvest mice from The British Wildlife Centre. The three types of trap that were used were the plastic trip-trap, the well tried and tested Longworth trap and the small sized Sherman trap. 2 All three were tested for trap efficiency by placing them in the Harvest mice enclosures at The British Wildlife Centre. The Longworth’s trip bar had been set to trip at the lightest possible weight, as the harvest mouse is the lightest of our small mammals. All three types of trap were tripped when an animal entered to eat the bait. Hair tubes were also trialled to find out the minimum size required to trap hairs from the small Harvest mouse. This proved difficult but eventually a hair was caught in a tube whilst it was in a cage. Hair tubes had to be 20mm internal diameter. Hairs for comparison were taken from these captive-bred animals. The project used 20 of each type of trap at the two different heights; these were placed close together to maintain as similar habitat as possible for each group, which was labelled as a station. This gave us 6 traps at each station and ten stations at each site. Thirty stakes were made so that the above-ground traps could be placed at as near to 600mm as possible. Stations were placed at 15-20 metres apart depending on the habitat. 3 Traps were set up on Monday evening and then checked and re-set and re-baited at 0630 the following morning; a further check followed at 1230 and again at 1830. This gave us three sampling sessions per day; this continued for 5 days, with the last check taking place on Saturday morning when the traps and posts were removed. Bait type was selected and used throughout the whole project. This consisted of a bird seed mixture, with castors added to ensure any Shrews had food for survival, and a small piece of apple to provide moisture for trapped animals. This proved difficult in Sherman traps as the seed would easily slide under the trip mechanism and stop the trap from closing. To avoid this, peanut butter was used with the seed and castors, mixed in to make a paste to stick to the back plate and avoid spillage under the trap door. 20 sites were trapped with 12 having 60 traps and 8 having 120 traps at different locations within the site. Two sites were re-visited (see appendix 1 for list of sites). It was hoped that we could run two different sites concurrently however resources were unavailable to enable us to do so. Volunteers could not be found to complete all the checks for a full week and an additional SWT staff member was unavailable. The first trapping session took place in August 2011. A site was chosen with recent records of Harvest mice nests. The site was an SWT reserve and 2 sites were used with a total of 120 traps. Although no target species were caught, it was a useful exercise to test our equipment. Some minor adjustments were made. Trapping continued on a further six sites before temperatures dropped (when trapping was considered to be a welfare issue). During this first year, Harvest mice were caught at 4 sites. Trapping was planned for the spring of 2012, but high temperatures made it a welfare problem especially with the traps on posts getting too hot. Trapping continued in September to November at another six sites, with Harvest mice caught at three. In 2013, only 5 sites were completed although 3 of the sites had 120 traps. This was due to 4 sickness and bad weather. Harvest mice were caught at two sites. It was decided to continue the project for another year to gain more meaningful data. In 2014, a further 4 sites, each with two sets of traps, were undertaken. Harvest mice were trapped at all of these sites (results table in appendix 1). Many of the trapping sessions were completed in darkness which meant working by torch light. A total of 65 volunteers were used during the project, with many becoming well trained in trapping methodology. 5 Results A total of 197 harvest mice were trapped; of these, 126 were new catches; 71 were re- traps, identified by fur clips. The range of totals per site (where Harvest mice were caught) was from 1 up to 46, with only 6 sites reaching double figures. Harvest mice were caught in all types of traps and at each of the two heights. Figures show that the most trappings by trap type and height were Longworths on the ground. 6 All of the new captures were weighed and sexed and all data collected. Each new Harvest mouse caught was fur clipped. Average weight was 6g, with several at 12g, which were pregnant or lactating females. Sex ratio was in favour of females at 57% . 7 Harvest mice were trapped at each session morning, midday and evening, confirming that they are active day and night. The largest numbers of Harvest mice were found at the morning checks. A total of 71 Harvest mice were re-trapped. 8 We trapped some harvest mice in late October that showed a clear moult line. Nest searches were made during each survey and also at other times. Nests were more easily found where Harvest mice were trapped, but also found at sites where trapping was unsuccessful. 9 Nest searches were also undertaken at training sessions with volunteers from other organisations: PTES, National Trust and Mammal Society. Habitats varied: heathland, grassland areas, reedbed, flood plain and River meadows. Reedbed site River meadow site 10 Camera traps were deployed with very little success. Only three sequences of a Harvest mouse were obtained, with many other images captured of reeds blowing, rain falling and also a Fox (Vulpes vulpes) investigating traps. CAMERA TRAPPING Owl pellets were collected but we had difficulty finding them close to our sites. Pellet dissection was part of the training days we held, so we were able to train volunteers how to identify small mammal remains, especially Harvest mice. None were found in pellets from our area, but pellets obtained from other areas outside Surrey had Harvest mice remains to help with identification skills. 11 Riverside, the most productive site in 2013, was heavily flooded in the 13/14 winter, with the water completely covering the vegetation. Trapping was tried during August without success, however in October we were very successful. Harvest mice had somehow survived the flood. There was higher ground nearby, so it is possible they fled to higher ground to survive, returning when the flooding subsided. Conclusions Hair tubes were unsuccessful and no hairs were obtained. These were trialled on captive Harvest mice, and even then, only on one occasion did we collect a hair. Camera traps gave us poor results despite some being aimed at nests. The speed of which a Harvest mouse can move through the stems is possibly too fast for cameras to trigger. Owl pellets can be a good source of information provided they can be found in the vicinity of the site surveyed. Nest searches were extremely useful in deciding presence or absence on a site, but could not be relied upon to indicate population numbers. Sherman traps had no place for bedding, causing welfare problems in cold weather and were difficult and time consuming to clean and re-bait. Harvest mice were captured, but not as many as in Longworth traps. Trip traps were successful, but had the problem of trapped animals sweating or getting cold. Like the Sherman traps, there was no place for bedding.