ASSESSMENT OF THE STATUS OF THE RED SQUIRREL IN CENTRAL WALES. Report Author: Anna Hobbs, with Introduction by Neil Matthew

Transcription

1 ASSESSMENT OF THE STATUS OF THE RED SQUIRREL IN CENTRAL WALES Mid Wales Red Squirrel Project Report 2005 A. M. Hobbs Date: September 2005 Report Author: Anna Hobbs, with Introduction by Neil Matthew Project Officer: Anna Hobbs Red squirrel in Lodgepole pine canopy in Fannog, Tywi Forest.

2 Contents Page Number 1 Summary 3 2 Introduction by Neil Matthew 4 3 Aims 6 4 Methods Survey Sites Trapping Methods and Trap Positions Processing Captured Squirrels 20 5 Results and Interpretations Red Squirrels Captured in the Project Area Red Squirrels Captured in Bryn Arau Duon Red Squirrel Genetic Analysis Grey Squirrels Captured in the Project Area and Results 27 from Blood Sample Analysis for Parapox Virus Antibodies 6 Discussion Red Squirrels in Mid Wales Red and Grey Squirrels in Project Survey Sites Project Methodology Habitat Management and Grey Squirrel Control for Red 30 Squirrel Conservation 7 Acknowledgements 33 8 References & Bibliography 34 9 Appendix I: Methods 36 Original Proposed Methods by Neil Matthew Appendix II: Sightings Record Appendix III: Trap Recording Forms Appendix IV: Red Squirrel Recording Forms Appendix V: Grey Squirrel Recording Forms Appendix VI: Bryn Arau Duon Red Squirrel Data Appendix VII: Summary of Press & Publicity for the Project 96 Press & Publicity Project Outputs, compiled by Lizzie Wilberforce Appendix VIII: Project photographs 98 2

3 1 Summary The principal aim of this project was to establish the presence of the red squirrel Sciurus vulgaris within Tywi, Irfon and Crychan North forests, situated in mid Wales. A programme of live capture trapping was conducted in 5 study sites within these forests. Red squirrels were found to be present in Tywi Forest, where a total of 14 individual red squirrels were captured within the study sites Fannog and Nant yr Hwch. A total of 31 red squirrel captures (including recaptures) were made in these sites. Hair follicle samples were taken from each red squirrel for genetic analysis. No red squirrels were captured in Irfon Forest or Crychan North Forest. Red squirrel hair samples collected during trapping sessions undertaken outside the project area in Bryn Arau Duon Forest were donated to the project for genetic analysis. A total of 16 individual red squirrels were captured in the forest. Results from genetic analysis of the red squirrel samples suggest that the red squirrels in mid Wales are of Welsh lineage. The level of genetic variation observed in the mid Wales red squirrels, along with the presence of a haplotype unique to mid Wales, identifies this population as important in the conservation of red squirrels in Wales. Grey squirrel S. carolinensis control was performed in all survey sites. A total of 43 grey squirrels were captured and humanely despatched during the course of the project. Blood samples taken from grey squirrels were analysed for parapox virus. A high proportion of the grey squirrels captured in the Irfon study site had been infected with this virus; in the other 4 sites, approximately 50% of grey squirrels captured had been infected. Recommendations for suitable habitat management and grey squirrel control for red squirrel conservation are given. 3

4 2 Introduction The red squirrel Sciurus vulgaris has become a priority species within the UK Biodiversity Action Plan on account of its widespread and long-term decline, which continues to accelerate. In Wales it has become confined mostly to a few sites in Anglesey, Gwynedd, Conwy, Denbighshire and central Wales, with only sporadic and often unconfirmed reports elsewhere. It is seriously threatened everywhere within this range. A number of surveys have been carried out in central Wales. The Countryside Council for Wales (CCW) commissioned a preliminary study in 1999 (Matthew & Bevan, 2000). This was mainly a review of available records and an initial assessment of red squirrel habitat. It showed that the red squirrel was still to be found in the 1990s in a number of large forests such as Brechfa, Crychan, Tywi and Irfon as well as in some smaller coniferous parcels north of Carmarthen. The survey also revealed that there had been a significant contraction of range since the 1980s and that this process seemed to be continuing even within forests such as Brechfa where there were fewer and less widespread records in the 1990s. A further survey was carried out by Dr Sarah Cartmel in the winter and spring of 2001 (Cartmel, 2002). Lines of hair tubes were set up in four forests (Tywi, Irfon, Crychan and Bryn Arau Duon) in the hope of recording red squirrel presence. Habitat assessments were also made for each forest. In the event few signs of red squirrel were discovered: at two locations only within Irfon forest. Grey squirrels (S. carolinensis) on the other hand were widespread and it was feared that they may have dominated the hair tubes and hence the survey could not be considered a true reflection of the abundance or pattern of distribution of the red squirrel within these forests. Management prescriptions for improving the forests for the red squirrel were given as was a recommendation to carry out a programme of grey squirrel control in areas that either had or showed potential to contain red squirrels. As the above survey proved that a red squirrel population continued to exist, it was decided that there was a need to review the suitability of the forests of central Wales for the species. This study was commissioned by CCW and funded jointly by them, Carmarthenshire Local Biodiversity Action Plan Partnership, and Ceredigion County Council and was carried out by Dr Sarah Cartmel (Cartmel, 2003). The review was undertaken in the light of forest design plans and the reality of long-term forestry operations over the next 50 years and concentrated on Tywi, Irfon and Crychan forests. It demonstrated that while suitable habitat would continue to exist it would become increasingly fragmented and if inappropriate broadleaf plantings were introduced the balance could be tipped in favour of grey squirrels. The study indicated what modifications to forestry plans and practice needed to be made to conserve red squirrel habitat in these areas. A red squirrel group was set up in 2002, largely under the auspices of the Carmarthenshire LBAP Partnership. A number of organisations and individuals are represented at meetings where a widespread enthusiasm to conserve the remaining red squirrels in the area has been evident. Red squirrel conservation has been actively pursued by woodland managers at Bryn Arau Duon since

5 DNA analysis on a red squirrel killed on the road near the Sugar Loaf (near Glyn Saer, Crychan North Forest) revealed a previously unknown haplotype and demonstrated a kinship with populations on Anglesey and a more remote similarity with animals on the Isle of Wight and the Channel Islands (as opposed to those in northern England and Scotland). In the interests of maintaining genetic health and diversity, it is desirable to maintain as many populations as possible. In the future it may be possible to exchange animals that are surplus to an area (isolation and fragmentation preventing a natural expansion outwards) in order to improve the strength of each population. This gives added importance to the conservation of the species in central Wales. Following Sarah Cartmel s habitat review (Cartmel, 2003) discussions have taken place with forestry managers, both public and private, about how to adapt forest operations to promote red squirrel conservation. On all sides a willingness to adopt her suggestions, where this is possible, has been shown. Discussions took place on the proposal to carry out a trapping programme in order to gain more information about the occurrence of red squirrels and to control greys. It was hoped that this project would result in much better information than could be gained by the hair tube survey. The area of focus has been Tywi, Irfon and Crychan North forests where it is considered that the successful conservation of the red squirrel could be best achieved. It is felt that it is less realistic to defend the species within Brechfa forest, Crychan South and other smaller woodlands. Tywi, Irfon and Crychan North forests are large, covering several thousand hectares, are fairly isolated and have only a limited component of broadleaved species within them. There is therefore a better than usual possibility of defending these areas against grey squirrels (particularly in Tywi). This core area lies within the three counties of Carmarthenshire, Ceredigion and Powys and this pilot project was conducted with more or less equal effort in all three. It is hoped that this successful trapping programme will be continued into the future subsumed perhaps within the landscapescale project proposed by CCW East Area. Neil Matthew CCW Conservation Officer Carmarthenshire 5

6 3 Aims The principal aim of this project was to establish the presence of the red squirrel within Tywi, Irfon and Crychan North forests, situated in mid Wales. This aim was to be achieved by a programme of live trapping within these areas. It was hoped that data collected would provide an indication of red squirrel distribution and population size. DNA samples were to be collected from each individual red squirrel captured for subsequent genetic analysis, with the aim of assessing: (i) level of genetic variation in mid Wales relative to other Welsh localities and (ii) relationship of these squirrels to other Welsh, UK and European populations. A secondary objective was to control grey squirrels in order to reduce the competitive pressure from the red squirrel and to improve the chances of capture. Blood samples were to be taken from despatched grey squirrels for subsequent analysis for parapox virus antibodies, with the aim of gaining information on the occurrence of this virus in grey squirrels within the study areas. Additional project aims included notifying landowners of the presence of the red squirrel on their land, and advising on grey squirrel control and habitat management for red squirrels; and publicising and raising awareness of the project and the decline of the red squirrel (see Appendix VII). 6

7 4 Methods The project commenced 16/8/04 and ended 15/9/05. The first month involved setting up the practical side of the project. This included conducting map-work and field surveys of the sites identified for trapping, in order to target habitat areas most suitable for red squirrels. There were two main trapping periods; survey period 1, which commenced 20/9/04 and ended 29/3/04; and survey period 2, which commenced 4/4/05 and ended 15/8/05. The last month of the project involved writing up project reports. 4.1 Survey Sites Survey sites identified in the original proposed methods (see Appendix I) were revised in order to focus on the most suitable locations for capture of red squirrels. Surveying was conducted in coniferous plantations in three areas, one in each of the three counties of Carmarthenshire, Ceredigion and Powys. The areas under investigation were subdivided into 5 survey sites (shown in Figures 1 & 2): Carmarthenshire 1. Glyn Saer Ceredigion 2. Dalarwen Powys 3. Nant yr Hwch 4. Irfon 5. Fannog Glyn Saer forms part of Crychan North Forest; Dalarwen and Nant yr Hwch lie within the Tywi Forest; and the Irfon site forms part of the greater Irfon Forest block. The total combined area of these survey sites is approximately 3500 ha. Separate red squirrel survey work was undertaken outside the project area in Bryn Arau Duon, a forest block within Carmarthenshire, to the west of the project sites. Suitable habitat in which to conduct trapping sessions was identified from map-work and preliminary field surveys for each of the 5 survey sites. Habitat identified comprised one or more of the following tree species: favoured red squirrel tree species Norway spruce Picea abies (Lurz et al., 1995; Cartmel, pers. comm.) and Scots pine Pinus sylvestris (Gurnell et al., 1997; Cartmel, 2002); Japanese larch Larix decidua, Lodgepole pine Pinus contorta, and Douglas fir Pseudotsuga menziesii, the seeds of which are also eaten by red squirrels. Further work identified plantations of the above species that were of, or above, the age when the trees start to produce good seed crops upon which squirrels feed. Areas containing broadleaf, especially those comprising species that produce large seeds, e.g. oak Quercus spp., beech Fagus sylvatica and hazel Corylus avellana, in which grey squirrels have a competitive advantage over red squirrels were avoided. Although it has been shown that Sitka spruce Picea sitchensis is generally not favoured by red squirrels, it is occasionally utilized during mast years (Cartmel, pers. 7

8 comm.). As fieldwork progressed it became apparent that it was an exceptional mast year with abundant squirrel feeding signs on Sitka cones in most of the survey sites (pers. obs.). It was not possible to distinguish between red and grey feeding signs on cones. Therefore, traps were generally erected in Sitka spruce where squirrel feeding signs were found if the Sitka occurred near other, more favoured red squirrel tree species. Red squirrel sightings records (recent sightings given in Appendix II; sightings records up to 2000 in Matthew & Bevan, 2000) were used to further increase the chances of red squirrel capture, by concentrating efforts on locations where most recent sightings had occurred in each site. During survey period 1, trapping sessions were conducted in Fannog, Nant yr Hwch and Irfon. In survey period 2, trapping sessions were conducted in Irfon, Glyn Saer, Dalarwen, Fannog and Nant yr Hwch. 8

9 Figure 1: Map Showing Fannog, Nant yr Hwch, Irfon and Dalarwen Survey Sites. 9

10 Figure 2: Map Showing Glyn Saer Survey Site. 10

11 4.2 Trapping Methods and Trap Positions The original outlined trapping methods are given in Appendix I. This was revised prior to commencement of the trapping programme and the revised version given below. Further revisions were made during survey period 1 and prior to the start of survey period 2. These additional changes are also detailed below. Survey Period 1 Placement of Individual Traps and Pre-baiting Further habitat field surveys were performed within identified blocks of suitable red squirrel habitat for each survey site prior to erecting traps. The purpose of these surveys was to identify trap sites within areas where trees were bearing cones and where signs of squirrel presence (e.g. dreys and squirrel feeding signs on cones) were located. Once such locations had been found, individual trees were identified in which to place the traps. Physical aspects of the habitat were also considered when choosing trap trees, e.g. trees chosen with good arboreal links to surrounding canopy, or connections between other trees or across streams via fallen trees. Traps were to be positioned at a density of approximately 1 per 5 ha. Traps employed were modified mink traps; a live-single-capture cage trap attached to a wooden platform, fitted with a wooden box in which squirrels could take refuge after being captured (see photograph 1, Appendix VIII). Each trap was positioned at eye-level within a tree and tied onto the branches using garden wire, so that the side of the trap sat against the tree trunk (in order to minimize exposure of visiting, or trapped squirrels). If the location could be visited by the general public, the trap was positioned higher up the tree. For each trap erected, a unique trap position number was assigned and following data recorded for each trap position: location; species and age of tree trap placed into, and whether tree coning; area of stand and indication of tree density; species in stand and approximate percentage composition; ages of stand trees and whether coning; signs of squirrel presence near trap tree. Data was recorded on Trap Recording Forms (see Appendix III). The ground was pre-baited in the vicinity of each trap with whole maize, and the traps pre-baited with a mixture of whole maize, peanuts and sunflower seeds (in 5:3:1 ratio) and left unset and open. Traps remained in pre-bait condition for approximately 5 days. Traps were visited every 2-3 days, to replenish pre-bait as required. The ground was only pre-baited on the first 2 occasions, after which, traps only were pre-baited. During visits, trap and ground pre-bait was examined for signs of squirrel feeding on the maize (kernels removed from the maize leaving a characteristic v-shape) and the date upon which first such feeding signs seen, recorded in Trap Recording Forms. Squirrel feeding signs on cones near each trap were similarly recorded. Pre-bait period dates were recorded for each trap in Trap Recording Forms. Setting Traps Traps were baited daily with a mixture of mostly peanuts and a few hazelnuts, and set at approximately first light and left for 3-4 hours before checking. Traps were checked twice a day, and left unset over night. Wherever possible, traps were viewed from a distance, in order to reduce disturbance of visiting squirrels. Each trap remained in position until, after a capture, there were approximately 2 consecutive days with no captures; or until the trap had been in position and set for a 11

12 maximum of 10 days (12-14 days including weekend(s) when traps left in pre-bait state, baited with a mixture of maize and peanuts). When the 10-day limit was reached for the majority of traps placed in one of the 5 survey sites, all traps were generally moved to the next site. In some circumstances, certain traps were left in position after the 10 day cut-off, e.g. if no capture had been made but there were still squirrel feeding signs near the trap. If it was not feasible to check individual traps remaining in the previous survey site every day (for example, due to the next study site being some distance away), traps were left in pre-bait inbetween days checked. When surveying in Nant yr Hwch, a second trapping session in that site was undertaken before surveying the subsequent study site. Following each site trapping session, all traps utilized were cleaned thoroughly, disinfected and left to dry, before employing in the subsequent survey site. This was performed in order to reduce risk of spreading disease, especially of parapox virus. Traps were also disinfected immediately after the capture of a grey, before re-setting, during trapping sessions. Dates traps set were recorded in Trap Recording Forms. The above procedures were repeated at each of the study sites in turn. Following the first trapping session in Fannog, the following revisions were made to the methods: The majority of traps were to be positioned at approximately 100 m intervals in line transects or clusters within the suitable stands. Traps were left in pre-bait condition for 2 weeks (i.e. 14 days). Where grey squirrels were at relatively high densities and, therefore, the risk of parapox virus being spread to reds within the same area increased, the ground only was pre-baited for the first pre-bait week and traps pre-baited in the second pre-bait week. Survey Period 2 Revisions introduced following survey period 1: Following survey period 1, further traps were acquired. This allowed 49 traps to be left in position more permanently, to increase the time available for squirrels to find and utilize the traps and thus, allow for the more cautious behaviour of red squirrels. Traps were positioned within all 5 study sites. Certain survey sites were combined and trapped concurrently; whilst one group of traps was set, remaining traps were placed into pre-bait and vice versa. Traps were left set overnight and checked late in the day during the breeding season to minimize the probability of leaving captured red squirrels (especially females with dependent young) in traps for a lengthened period over night. Trap bait was replenished during the first visit of the day, conducted in the morning. During survey period 2, pre-bait maize ceased to be examined for signs of squirrel feeding. This followed results from survey period 1 where squirrel captures were not made despite feeding signs on maize. It was proposed that other small mammals create similar feeding signs, therefore, these signs cannot always be taken as an indication that squirrels are feeding at traps. Figures 3 to 9 show the positions of traps during survey periods 1 and 2, along with the assigned trap position number. 12

13 Figure 3: Map Showing Trap Positions in Fannog During Survey Period 1. 13

14 Figure 4: Map Showing Trap Positions in Nant yr Hwch During Survey Period 1. 14

15 Figure 5: Map Showing Trap Positions in Irfon During Survey Period 1. 15

16 Figure 6: Map Showing Trap Positions in Fannog and Nant yr Hwch During Survey Period 2. 16

17 Figure 7: Map Showing Trap Positions in Irfon During Survey Period 2. 17

18 Figure 8: Map Showing Trap Positions in Dalarwen During Survey Period 2. 18

19 Figure 9: Map Showing Trap Positions in Glyn Saer During Survey Period 2. 19

20 4.3 Processing Captured Squirrels Red Squirrels Captured red squirrels were transferred from the trap to a hessian sack placed over the end of the trap, and then manoeuvred into a wire mesh handling cone (see photographs 2 & 3, Appendix VIII) to hold them securely. Squirrels were weighed and examined in order to determine sex and age-class (adult or juvenile), and to comment upon external appearance and condition, including signs of disease and presence of parasites. Data were recorded in Red Squirrel Recording Forms (see Appendix IV). For each individual red squirrel captured, a unique I.D. code was assigned and a sample of hair follicles collected for subsequent DNA analysis. (Approximately hairs were plucked from the tail, ensuring follicles still attached. Each sample was placed in a separate 2 ml phial labelled with the squirrel I.D. code and containing 70% ethanol, ensuring that follicles were at the base of the tube). The right-hand ear of the squirrel was tattooed with a unique number before releasing the animal at the same location as captured. Following the initial 5 red squirrel captures, each subsequent red squirrel caught was given a tail fur clip to enable a recapture to be identified without unnecessary handling. Hair samples were stored upright in a freezer before sending to Wildlife DNA Services (Bangor) for genetic analysis. Red squirrel sightings made during trapping sessions were also recorded on Red Squirrel Recording Forms, and included in the sightings record. Hair samples collected from red squirrels captured in Bryn Arau Duon were donated to the project by Huw Denman. These were sent with project samples for genetic analysis. Greys Grey squirrels were transferred from the traps to a hessian sack (different to that used to handle red squirrels), and as the law necessitates (Wildlife & Countryside Act, 1981), humanely despatched. Each squirrel was assigned an individual I.D. code. If a blood sample was to be taken for subsequent analysis for parapox virus antibodies, this was performed immediately after dispatching the squirrel (1 ml of blood collected into a 2 ml phial labelled with the squirrel I.D. code; obtained either directly from the heart using a medical needle and syringe, or from the aorta after opening the abdomen down the ventral midline with a scalpel blade. Separate scalpel blades were used on each occasion to avoid contamination.) Each squirrel was weighed and examined to determine sex, age-class and condition, including presence of parasites or signs of disease. Data were recorded in Grey Squirrel Recording Forms (see Appendix V) Blood samples collected were allowed to clot and were stored in a fridge before sending to Moredun Research Institute (Penicuik) for analysis. It was initially proposed that 5 samples of blood were collected from each of the study sites. However, samples were taken as routine if time permitted. 20

21 5 Results and Interpretations Appendix III gives the raw data recorded in Trap Recording Forms for each site and trap position, including the number of red and grey squirrel captures made at each trap location. Table 1: Summary of the Total Number of Days Traps Set and the Number of Red and Grey Squirrels Captured at Each Trap Position During Survey Periods 1 and 2 (Red Squirrel Figures Include Recaptures). Trap Position Number & Survey Site Total Number of Days Trap Set Number of Red Squirrel Captures Number of Grey Squirrel Captures Survey Period 1 Survey Period 2 Survey Period 1 Survey Period 2 Survey Period 1 Survey Period 2 Fannog Nant yr Hwch

22 Trap Position Number & Survey Site Number of Days Trap Set Number of Red Squirrel Captures Number of Grey Squirrel Captures Survey Period 1 Survey Period 2 Survey Period 1 Survey Period 2 Survey Period 1 Survey Period 2 Nant yr Hwch Irfon Dalarwen Glyn Saer Total Captures

23 Survey Period 1 No red squirrels were captured during survey period 1; conducted in Fannog, Nant yr Hwch and Irfon. One red squirrel sighting was made prior to the commencement of trapping on 6/9/04 at SN810516, on the border of Fannog and Nant yr Hwch. During survey period 1, 15 grey squirrels were captured; 3 in Fannog, 1 in Nant yr Hwch and 11 in Irfon. Survey Period 2 During survey period 2, 31 red squirrel captures were made (including recaptures); 10 in Fannog and 21 in Nant yr Hwch. Trap positions in which red squirrels were caught are shown in Figure 10. No red squirrels were caught in Irfon, Glyn Saer or Dalarwen. During survey period 2, 5 grey squirrels were captured in Fannog; 2 in Nant yr Hwch; 6 in Irfon; 10 in Glyn Saer; and 5 in Dalarwen. 5.1 Red Squirrels Captured in the Project Area Raw data for red squirrel captures are given in Appendix IV. A total of 14 individual red squirrels were captured. The number of recaptures totalled 16; 6 of the 14 individual red squirrel caught were recaptured between one and 4 times. It is uncertain whether a further capture, of the squirrel with ID code NYHRS5, was a recapture, or a new individual (it was necessary to release this animal before completing the examination for a tattoo and occurred prior to tail fur clipping being introduced). Table 2: Individual Red Squirrels Captured in the Project Area. Red Squirrel I.D. Code Sex (M/F) Age (Adult/ Juvenile) Weight (g) Squirrel of Breeding Weight (BW) if >290g/ Below BW (<BW) if 290g Site Nant yr Hwch (NYH)/ Fannog (F) Initial Capture Recapture(s) Haplotype Date Trap Site Date Trap code Position Position Number Number NYHRS1 F Adult 295 BW NYH ang1 NYHRS2 M Adult 285 <BW NYH wc3 FRS3 M Adult 275 <BW F wc3 NYHRS4 M Adult 310 BW NYH NYH NYH NYHRS6 M Adult 270 <BW NYH F F F F NYHRS7 *M Adult 280 <BW NYH wc3 NYHRS8 F Juv. 220 <BW NYH wc3 FRS9 M Adult 270 <BW F F wc3 NYHRS10 M Juv. 240 <BW NYH wc3 NYHRS11 F Adult 430 BW NYH NYH F F wc3 mw1 **No result available NYHRS12 M Adult 300 BW NYH NYH wc3 NYHRS13 M Juv. 220 <BW NYH wc3 FRS14 F Juv. 155 <BW F wc3 NYHRS15 F Adult 280 <BW F NYH wc3 * Sex uncertain. ** DNA extracted from hair follicles could not be analysed due to the quality of the DNA. N.B. Four of the 16 recaptured squirrels were released without handling and not identifiable by appearance. These recaptures do not, therefore, appear in Table 2. 23

24 Figure 10: Map Showing the Trap Locations at Which Red Squirrels Captured. 24

25 All red squirrels captured appeared to be in good condition, with no visible signs of disease. The majority of individuals captured had one or a few very small ticks; fleas were seen on a couple of individuals. There was a marked variation in fur colouration between individuals; ranging from a very dark back and tail to a ginger back and varying amounts of tail and ear tuft bleaching (see Appendix VIII). Of the 10 adults captured, 4 were of breeding weight being over 290g (Cartmel, pers. comm.); the other 6 individuals weighed between 270g and 285g. The fact that 4 juveniles were captured proves that red squirrels are breeding successfully within the area. In addition, the juveniles captured were products of at least two breeding attempts (NYHRS8, NYHRS10 and NYHRS13 were captured within 24 days of each other in 3 adjacent traps [106, 110 and 30] and were all similar in weight; FRS14 was noticeably younger). Evidence of another successful breeding attempt was gained from the captures of NYHRS11. This female was thought to be pregnant when first captured in Nant yr Hwch on 22/6/05. When recaptured 22 days later in Fannog, she was much reduced in size with enlarged mammary glands and nipples visible with fur loss around them; signs which occur when females suckle young (Gurnell, 1994). Ten individuals were captured in Nant yr Hwch and 6 individuals were captured in Fannog (2 squirrels were caught in both sites). Generally recaptures occurred in traps relatively close (within 500m) to the initial capture location. The exception was NYHRS11, who was recaptured 1.5km from original capture location. It is likely that this female moved location in order to produce and raise her young. 5.2 Red Squirrels Captured in Bryn Arau Duon Raw data of red squirrel captures in Bryn Arau Duon was provided by Huw Denman, and is given in Appendix VI and summarised in Table 3. A total of 20 red squirrel captures (including recaptures) were made in Bryn Arau Duon during trapping conducted in ; 2 red squirrels (BAD01 and BAD02) were captured in the first trapping session in September/October 2004; 18 were captured during the second session undertaken in July Captures of BAD01 and BAD02 occurred prior to tail fur clipping being introduced to mark individuals. It is not possible to ascertain whether or not either of these animals was recaptured in the second session when fur clipping was undertaken. One recapture was recorded; another squirrel escaped and it is not known whether this was a new individual or another recapture. Thus, the number of definite individuals caught in Bryn Arau Duon is 16 (BAD03-BAD18). Squirrels were not assigned to an age-class on capture. For the purpose here, those weighing 250g and above have been classed as adults and those below have been classed as juveniles (Cartmel, pers. comm.). Of the 16 known individuals, 5 of the adults were of breeding weight and the other 5 adults weighed between 250g and 290g. The presence of 6 juveniles proves that squirrels are breeding successfully in Bryn Arau Duon. 25

26 Table 3: Red Squirrels Captured in Bryn Arau Duon. Red Squirrel I.D. Code Sex (M/F) Age (Adult/ Juvenile) Weight (g) Squirrel of Breeding Weight (BW) if >290g/ Below BW (<BW) if 290g Initial Capture Recapture(s) Haplotype Date Trap Position Date Trap Position code BAD01 M *Adult 290 <BW SN ang1 BAD02 M *Juv. 240 <BW SN ang1 BAD03 M *Adult 250 <BW SN ** BAD04 M *Adult 310 BW SN ** BAD05 M *Adult 330 BW SN ** BAD06 M *Adult 270 <BW SN ** BAD07 M *Adult 340 BW SN ** BAD08 M *Juv. 180 <BW SN SN ** BAD09 M *Adult 280 <BW SN ** BAD10 M *Adult 300 BW SN ** BAD11 F *Juv. 170 <BW SN ** BAD12 F *Juv. 190 <BW SN ** BAD13 M *Juv. 240 <BW SN ** BAD14 M *Juv. 180 <BW SN ** BAD15 F *Juv. 160 <BW SN ** BAD16 F *Adult 270 <BW SN ** BAD17 M *Adult 290 <BW SN ang1 BAD18 M *Adult 310 BW SN ang1 * Squirrels not assigned to an age-class on capture and placed into age-classes by weight. ** Haplotype not determined due to insufficient sample. 5.3 Red Squirrel Genetic Analysis Results from genetic analysis were provided in a report from Wildlife DNA Services. The report gave the following: Of the 32 samples analysed (14 project samples and 18 samples collected in Bryn Arau Duon), DNA was successfully extracted and amplified from 17 samples (13 of project samples and 4 of the samples from Bryn Arau Duon). The resulting DNA sequences revealed 3 different mitochondrial DNA haplotypes. Mitochondrial DNA is inherited from the maternal line, and is useful in studies of population ancestry. The three haplotypes are coded: ang1, wc3 and mw1. Tables 2 and 3 show the haplotype code for the individual red squirrels. All three haplotypes were present in the project samples, and Bryn Arau Duon samples displayed the ang1 haplotype. All haplotypes had previously been observed in Wales; ang1 observed on Anglesey (Pentraeth population); wc3 in Clocaenog Forest (north Wales); and mw1 was previously recorded in mid Wales from the road-killed individual near the Sugar Loaf (Glyn Saer). However, their presence in a single locality had not been recorded before and the mw1 haplotype would appear to be unique to mid Wales. Assessment was made of the relationship of the mid Wales red squirrels to other Welsh, UK and European populations. Results show that the haplotypes ang1, wc3 and mw1 are very similar to each other, and that genetic distances among the three were smaller than their respective distances from the other haplotypes currently found in Wales (wc9, in Clocaenog) and from other known European haplotypes. This suggests that the red squirrels in mid Wales belong to a relatively narrow lineage that may include an ancestral Welsh population (Ogden, 26

27 et al., 2005), as opposed to comprising more widely mixed lineages. In other words, they are more closely related to each other than they are to other British and European populations, and they may have been less influenced than other British populations by the introduction of European red squirrels to the UK in the 19 th century. The number of haplotypes within this single population indicates that the mid Wales red squirrels are important as a reservoir of genetic diversity within Welsh red squirrels as a whole, and of clear importance in the conservation of red squirrels in Wales. 5.4 Grey Squirrels Captured in the Project Area and Results from Blood Sample Analysis for Parapox Virus Antibodies Raw data for grey squirrel captures within the project area are given in Appendix V. A total of 43 grey squirrels were captured over the course of the project. All grey squirrels captured were humanely destroyed. An indication of grey squirrel density may be gained from the results (see Table 1). If number of traps and days traps set in each site is taken into account, it is apparent that grey squirrels are at highest density in Irfon and Glyn Saer. They are at a lower density in Dalarwen and Fannog, and at a relatively low density in Nant yr Hwch. Table 4: Showing Results of Grey Squirrel Blood Sample Analysis for Parapox Virus Antibodies Survey Site Total Number of Grey Squirrels Captured Number of Grey Squirrels from Which Blood Sample Taken Number of Grey Squirrels +ve for Parapox Virus Antibodies Number of Grey Squirrels ve for Parapox Virus Antibodies Fannog Nant yr Hwch Irfon Glyn Saer Dalarwen Totals As sample sizes for each survey site are small, data cannot be analysed statistically, and results can only be taken as an indication of the extent of parapox virus within the grey squirrel population in the area. It would appear that a high proportion of the grey squirrels in Irfon have been infected with parapox virus. In the other sites, approximately half of the squirrels from which blood samples were taken are positive. 27

28 6 Discussion 6.1 Red Squirrels in Mid Wales The principal project aim of establishing the presence of the red squirrel has been met for Tywi Forest, but not for Irfon Forest or Crychan North Forest. It is not possible to accurately assess red squirrel population size and distribution in mid Wales from the results obtained. In order to achieve this, it is suggested that further red squirrel survey work is conducted over an extended time period and area. New individual red squirrels were still being captured in the last trapping session in Fannog and Nant yr Hwch, suggesting that the number of animals in these survey areas is greater than the total number captured during the project. Furthermore, recent red squirrel sightings (see Appendix II) have occurred outside the project survey sites in Irfon and Tywi forests. This suggests that the distribution of the red squirrel within these forests is more extensive than the areas identified in which red squirrels were captured. The fact that 30 individual red squirrels were captured (combined totals from Fannog, Nant yr Hwch and Bryn Arau Duon) in a relatively short period of time, and the fact that they are breeding successfully, suggests that red squirrels in mid Wales persist at a higher level than previously thought from sightings records and results of the hair tube surveys in 2001 (Cartmel, 2002). Results from the genetic analysis of red squirrel samples suggest that red squirrels in mid Wales are of Welsh lineage, as opposed to more widely mixed lineages. The presence of a haplotype unique to mid Wales and a higher level of genetic diversity than previously recorded elsewhere in Wales mean that the mid Wales red squirrel population is of considerable importance in the conservation of red squirrels in Wales as a whole. 6.2 Red and Grey Squirrels in Project Survey Sites An indication of grey squirrel density was gained from grey squirrel capture results. It appears that the density of grey squirrels is relatively high within the areas surveyed in Irfon Forest and Crychan North Forest. It is possible that grey squirrels have reached a level within these survey areas where red squirrels are no longer able to compete for resources. Wider, more intensive survey work is required in both these forests to further investigate whether red squirrels are still present. Similarly, further work is required in Dalarwen where no red squirrel captures were made, but where grey squirrel density appeared to be lower. It appears that the density of grey squirrels in the combined survey sites Nant yr Hwch and Fannog is relatively low in comparison to densities in Glyn Saer and Irfon survey areas. The number of grey squirrels captured in Nant yr Hwch and Fannog was lower than the total number of individual red squirrels captured in these sites. As red squirrels were captured during the first trapping session of survey period 2, before any grey squirrels captured during that period had been removed, it is not thought that the capture of red squirrels was related to the reduction of competitive pressure from grey squirrels (via trapping) in this instance. From the results, it is feasible that grey squirrels dominated individual traps, and that subsequent grey capture and removal may have increased the capture of red squirrels in those traps; but this did not appear to be the case for the majority of traps. The results suggest that, at present, the red squirrel may have a competitive advantage over the grey squirrel within Nant yr Hwch and Fannog. Both areas have very little broadleaf 28

29 within the forest and are surrounded mainly by open hillside or reservoir. Thus, these areas provide habitat which is more suitable for red squirrels than grey squirrels, and the ingress of grey squirrels will be slowed by the open hillside and reservoir buffer zones. However, it is strongly recommended that a programme of grey squirrel control is undertaken in both Fannog and Nant yr Hwch, to ensure that a low grey squirrel density is maintained in efforts to conserve the red squirrel. It is also vitally important that future habitat management within Nant yr Hwch and Fannog does not tip the balance in favour of the grey squirrel and that suitable habitat management for red squirrels is incorporated into the forest design plans in order to conserve the red squirrel. Recommendations for habitat management and grey squirrel control are presented in section 6.4. Monitoring of identified populations of red squirrels should also be performed in order to identify changes within the populations and to assess the success of conservation work undertaken. Grey squirrels that had been infected with parapox virus were found to be present in each of the study areas. When this virus is transmitted to red squirrels it proves fatal. As grey squirrels are implicated in the transmission of parapox virus to red squirrels, these results add further to the importance of grey squirrel control in areas where red squirrels are present. 6.3 Project Methodology For a one year study, the intention of assessing the status of the red squirrel in mid Wales was ambitious. In reality, this would require a number of years extensive survey work to achieve. The initial proposed methodology given in Appendix I was largely unsuitable for the purpose of the project. The proposed density of traps (at 1 per 5ha) was too low (Cartmel, pers. comm.); the intention to employ 40 to 50 capture traps was not feasible, when considering the necessity to visit each trap 3 times a day; the timescale required in order to carry out work intended was greater than time available to the project; and pre-bait length and the length of time traps remained in situ were insufficient when performing survey work in areas where red squirrels have not previously encountered and become habituated to feeding from baited traps (Cartmel, pers. comm.). In addition, time allowances had not been made for initial setting up of the project, initial map and field surveys, habitat surveys prior to trapping periods, positioning and movement of traps, or for additional work required by the project. Based on the initial proposed methods, the areas calculated for trapping were much greater than feasible. The decision was taken to sample survey sites by concentrating efforts on smaller areas in which it was most likely that red squirrels would be captured, rather than attempting to cover larger areas with the associated risk of not conducting the survey thoroughly enough to achieve red squirrel capture for any one area. Of considerable importance was the time of year in which the survey was carried out. The second survey period was performed in what is considered the optimal time for red squirrel trapping; however, the first survey period was carried out during suboptimal time, when there is a low probability of red squirrel capture (due to reduced forage and movement of red squirrels and the abundance of natural food). It is recommended that the above points are taken into consideration when planning future red squirrel survey work. 29

30 6.4 Habitat Management and Grey Squirrel Control for Red Squirrel Conservation Habitat management for red squirrels should be based around two main aims: 1. Managing the woodland/forest habitat to ensure that red squirrel populations are maintained and increased, predominantly through increasing amounts of suitable red squirrel habitat and the seed production of suitable conifer species; 2. Reducing the impact of grey squirrels on red squirrel populations. Before habitat management can begin the following need to be identified and mapped: i. Core areas. These are areas where red squirrel populations are known to exist. These areas can be identified through appropriate survey (as detailed in this report). ii. Areas for future population expansion. These will be outside the core areas but should be managed with a view to creating suitable habitat for red squirrel populations to expand into. Having identified the above, the following issues should be considered when developing habitat management plans: i. Forest Design Plans (FDPs). These are an integral part of sustainable forest/woodland management. They enable the integration of red squirrel conservation objectives with other forestry/woodland objectives (such as timber production, recreation, etc) and are the first key stage of producing a management strategy. ii. Area. An area of forest containing suitable conifer species between 2000 and 5000 ha is the recommended size in which a viable population of red squirrels can be sustained. Within these areas rides and edges are important features, as trees along such edges tend to cone more heavily than those within a stand (Pepper & Patterson, 1998). iii. Species. Within extensive single species crops, in particular Sitka spruce, it is recommended that up to 20% of the crop should be made up of a mixture of species such as Norway spruce, Scots pine, Lodgepole pine, Larch or firs (Pepper and Patterson, 1998). Large-seeded broadleaf species should be kept to an absolute minimum. Smaller seeded broadleaves such as Rowan Sorbus acuparia, Birch Betula spp. and Willow Salix spp. are acceptable in small proportions (Cartmel, 2002). iv. Seed production. Planning should allow for suitable conifer species to maximise their seed production. This can be attained in a number of ways, such as retaining trees on south facing slopes; creating irregular coupe boundaries/forest edges; delayed felling of large-crowned trees; incorporation of Low Impact Silvicultural Systems, creating a diverse forest/woodland structure; restocking with suitable conifer species (Pepper & Patterson, 1998). v. Connectivity. Corridors of standing timber should be maintained throughout the forest to link suitable feeding and breeding areas. These corridors should be a minimum of 20 metres wide (Gurnell, 1994). They should contain a mix of suitable species and mature coning trees (Pepper and Patterson, 1998). vi. Age Structure. Ideally 50-60% of a forest should contain conifers species of seed producing age. A structure should be developed and maintained so that approximately 30% of trees are 0-15 years old, 30% of trees are years old and 40% are over 30 30

31 years old (Gurnell, 1994). Due to the fact that Norway spruce starts to cone at a relatively older age, the age structure for Norway spruce would be 30% trees 0-30 years old; 30% trees years old and 40% trees over 45 years old (Cartmel, 2002) vii. Shape. Long thin forest blocks are less suitable than more rounded blocks. The more round the block the less grey squirrel incursion occurs (Gurnell, 1994). This also allows for increased diversity in terms of species and age structure, thus reducing the need for red squirrels to travel too far to find suitable feeding and breeding sites. viii. Buffer zones. A buffer zone around the boundary of the area under consideration of between 1-3km wide is recommended, consisting of unsuitable habitat for grey squirrels such as conifer forest, open land (arable, moorland) or small-seeded broadleaved woodland (Pepper & Patterson, 1998) ix. Clear-felling. Clear-felling can still take place but thought must be given to the size and shape of a coupe. Corridors need to be maintained and should link areas for feeding. x. Thinning. Thinning should aim to increase seed production of suitable tree species (Pepper & Patterson, 1998) and remove unsuitable tree species such as large-seeded broadleaves that may occur within a crop. xi. Low Impact Silvicultural Systems (LISS). This method of silvicultural management could allow for more sympathetic management of large forest blocks for red squirrels. Group felling and restocking with conifer species suitable for red squirrels would add to the diversity of species and age (which is important for continuity of coning cycles). Reduction of Grey Squirrel Impact As previously discussed, grey squirrels have a detrimental impact on red squirrels in two main ways: i. They are thought to be a vector for the parapox virus which is fatal to red squirrels if they become infected. ii. Where the two species co-exist, grey squirrels appear to out-compete red squirrels for resources. Grey squirrel impact can be reduced by: i. creating and maintaining a buffer zone around a forest block to reduce incursions of grey squirrels into the forest block; and ii. removal of grey squirrels through trapping or shooting (the only legal methods of grey squirrel control in known red squirrel areas). Suitable buffer zones will not stop an incursion of grey squirrels, but will slow their progress enough for them to be controlled effectively within forest/woodland blocks. Shooting is not a particularly effective method of controlling grey squirrels, with less than 20% of a colonising or established population being removed (Pepper & Patterson, 1998). Trapping grey squirrels using live-capture traps is the most appropriate and effective means of control in known red squirrel areas. Trapping should be targeted within the core areas identified; areas that have been identified where grey squirrel incursion occurs; and any appropriate buffer zones. 31