Llanelli Naturalists Newsletter - Summer 1993 - No 55
N.R. Matthews

1.    Introduction

In February 1993, I was contracted by the Countryside Council for Wales to' carry out a survey of dormice Muscardinus avellanarius in S.E. Carmarthenshire. In the previous six months or so a number of "new" dormouse sites had come to light in this part of the vice-county where characteristically chewed hazel nutshells had been found in several localities both on the coal measures, the limestone ridge and the old red sandstone to the north. Four of these sites were the result of fieldwork by Andrew Lucas, Assistant District Officer, Carmarthen and Llanelli Districts. It was decided that I, under the direction of Andrew, should continue searching for dormice in this area, concentrating on the coal measures which lie for the most part south and east of a line from Kidwelly to Tumble. It has been found by Paul Bright and others that dormice show a preference for areas of ancient woodland, so such areas became the focus of this survey. To help in this, recourse was had to the Nature Conservancy Council's Inventory of Ancient Woodland in Carmarthenshire. However, apart from this we thought it would be a good idea to stop at any locality which appeared to have a sufficient amount of fruiting hazels to make a search worthwhile. The time of year was particularly advantageous in that respect and the prominence of hazel catkins made it easy to identify places in which to search. We thought that there was probably no point in spending much more than half to one hour at each site.

Fieldwork was to occupy no more than six whole days. it was suggested that : should concentrate my searches in and around the valleys of Gwendraeth Fawr and Afon Morlais and also around Llanelli. I was to make a sample collection of hazel nutshells in each of the search sites, whether I considered any of them to be the work of dormice or not. My identification could then be verified by Andrew, or in difficult cases by other authorities. The results of my researches were to be written up in this report.

2.    Background to the Survey. A Brief History of the Dormouse in Carmarthenshire.

I am here using information from the Biological Records Centre and information provided by I.K. Morgan and others. A full list of dormouse sites (i.e. all those known to me) is given in Appendix 2. Until recently there have been only scanty records of the dormouse in Carmarthenshire. T.W. Barker in his "A Handbook to the Natural History of Carmarthenshire" published in 1906 mentions it as being "rather local in the county" and that he had found "a colony" in one of the fields around his home at Oaklands, Cwmffrwd, near Carmarthen. John Brunker, a naturalist active in the first half of this century, wrote around 1950 a manuscript list of the mammals of Llanegwad Parish (SN52), stating that "the dormouse is rare, only one record of its presence being noted.-

After this, there is no written record of the dormouse until 1978 when the animal was recorded by Elaine Hurrell (apparently a sighting) in woodland alongside the upper Afon Tywi, near Rhandirmwyn. In 1979 M.H. Williams reported finding dormice in a bird box alongside the Upper Lliedi Reservoir. In a recent conversation he told me that he had found two.torpid dormice in a box late in the year, perhaps December, while he was in the process of cleaning out the boxes in readiness for the next breeding season. He said that no more dormice have been seen here since then. Around 1980, hibernating dormice were found as a hedge was being laid near Capel Dewi. From the same year there is a record of dormice nearby at Tir Philip. Also in the same year there is a record for Llanarthne but I do not have the details for this. All these three sites are in or adjacent to the mid-Tywi valley. The next record (1983) comes from the mouth of the Tywi where dormice were reported in coastal scrub at St. Ishmaels. All these sites therefore, bar one, seem to be connected with the Tywi Valley. Over the past few years I have myself randomly searched for signs of dormice in woodlands in or near the Tywi valley around Llandeilo (eg Tregib Woods) but with, so far, no success.

The recent "spate- of dormouse records began in 1990 when the animal was found by Keith and Janet Crowden at their home near Llannon. Their cat had brought in a dead dormouse which they put in their freezer so I was able to look at it when I visited in the spring. In 1991 they found a dead dormouse under their hedge. Between whiles the cat had brought in another dormouse. In company with Richard Rees, at that time mammal recorder for Carmarthenshire, they found feeding remains in the form of hazelnut shells along their hedge. In the autumn 1991 I visited the Crowdens to have a closer look at the site. In addition to finding nutshells along the hedge bordering their paddock - we also found them in the hedge alongside the road in front of their house and later on I found more along a minor road which joins this one nearby. The intriguing thing about this site was that dormice were apparently living in hedgerows, there being no broadleaved woodland of any size in the vicinity. The hedgerows themselves, though overgrown were otherwise quite ordinary with hazel only being a minor component, the bulk of the hedge being of willow, hawthorn, ash, elder etc., with a small area of dense scrub, mostly blackthorn and willow. I wrote a brief description of my findings for the Dyfed Wildlife Trust's files.

The following year (1992) Janet Crowden went on to find nutshells in a wood (not apparently ancient) near Pontyberem, which was part of an opencast site proposed by British Coal. At the same time Terry Wells had a "close encounter" at his home near Four Roads when a dormouse jumped onto his garden gate: this in an area of hazel coppice and scrub on the limestone. Later in the year Andrew Lucas discovered four more sites in or near areas of ancient woodland on the Carboniferous Limestone and Old Red Sandstone.

The above forms the total of my knowledge of the distribution of the dormouse in Carmarthenshire before beginning the present survey. It is interesting that most of the earlier records (pre-1990) relate to the Tywi valley and that subsequent records come from the Old Red Sandstone, limestone and Coal Measure areas to the south.

3. Methodology of the Survey

The area of survey lay almost entirely within the area covered by the OS 1:25,000 Pathfinder series map of the two ten kilometre squares SN40/50 (Swansea, Gower and surrounding area, sheet 159) and used this map throughout the period of my fieldwork. Blocks of ancient woodland within this area were identified by using the NCC Inventory of Ancient Woodland (Provisional) for Carmarthenshire. See Map 1. My aim was to visit as many of these woodlands as possible within the allotted time, and also to look at a few other sites which I thought might be profitable. Six whole days were allowed for fieldwork but because of other commitments, making it sometimes difficult to spend a whole day, the survey was in fact undertaken over a period of nine days. At each survey site a sample of nutshells was collected and put in a plastic bag with a label giving the location with grid reference, the site code (using letters of the alphabet), the date and the number of any dormouse nuts found (all the dormouse nuts found at a site were included in the sample). At the end of the survey all the sample bags were delivered to Andrew Lucas. Some of the nutshells were difficult to identify definitely as the work of dormice, even with the use of a hand lens, and those I described as "possible dormouse nuts".

The time spent at each site varied. At the beginning of the survey I would sometimes, having found dormouse nuts after ten or 15 minutes, spend,another half an hour or more looking for more specimens. There was some merit in trying to obtain a good numerical sample, 6 or 7 nuts, but I had to admit that there was no formula that I knew of that equated frequency/numbers of dormouse nuts with an estimate of the abundance of the animal, hence this meant that as the survey progressed I felt that it would be better to leave after finding one or two dormouse nuts so that I could move onto another site. The main object was to get as big a geographical spread of records as possible. In places where it took more than 3/4 of an hour or so to find the first dormouse nut I would leave soon afterwards - as long as the nut was a good specimen this was sufficient for the purposes of this survey. At sites where no dormouse nuts were found a maximum of one and a half hours was spent searching.

At each site I recorded my findings and observations about the habitat etc by using a mini-cassette recorder. I visited a total of 21 sites and a map showing their distribution described alphabetically in order of survey is given (Map 2).

4.    Results of the Survey

Twenty-one search sites were visited and a list of these in order of survey with a summary of my findings is given in Appendix 1. Map 2 shows their distribution. Fuller details for each site are given here. Grid references for these sites are also given in the appendices.

Site A (SN576088)

Ancient woodland alongside the Afon Gwili. A narrow flood plain on either side of the river and lying between steep wooded banks. Abundant hazel Corylus  avellana coppice growing on the flood plain with other trees such as alder Alnus glutinosa, sycamore Acer pseudoplatanus, hawthorn Crataequs monocyna. Open ground contained bluebells Endymion non-scriptus and wood anemones Anemone nemorosa, but there were also patches of bramble Rubus fructicosa and bracken Pteridium aquilinum, quite thick in places. Hazels, younger and more open on the west side of the river, but with the coppice on the opposite side was closer together, taller and probably of c.20 years growth. The wooded banks behind the flood plain contained mature oaks Ouercus sp., birch Betula sp., a few ash Fraxinus excelsior with an understorey of hazel and holly Ilex aquifolium, some honeysuckle Lonicera periclymenum. Trees on each side of the river joined at the canopy, providing aerial pathways for dormice.

A lot of nutshells were found or, each side of the river. 40 minutes were spent searching on the west side of the river and 2 dormouse nuts were found, the first after 5 minutes and the second half an hour later. 45 minutes were spent on the east side but no dormouse nuts found. Total search time: 1 hour 25 min.

Site B (SN577080)

Ancient woodland, sandwiched between the A48 and the Afon Gwili and about one kilometre below Site A. A coniferous plantation lies to the south and I confined my search between this and where the line of pylons traverses the wood and river to the north. A narrow flood plain separates the river from the sloping woodland on the west side. This wood contains ash, oak, birch, willow Salix sp., sycamore and hazel. The growth is quite young and the structure fairly vertical with little distinction between the tree and understorey layers. There is also some honeysuckle here. Most hazels occur near the river bank, the area between these and the wooded slope behind being infilled with brambles. Aerial pathways exist between the river bank and the wooded slope. The hazels alongside the river are mostly quite large and of c.20. years growth. Underneath these big hazels the ground is open and there are bluebells, wood anemones, snowdrops Galanthus nivalis and violets Viola sp. The river is lined on both banks with large mature oaks and ashes, where there is alit: some holly and honeysuckle. On the east side of the river is improved pasture. Underneath the pylons, trees have been cut down resulting in shrubby growth of mixed species, perhaps attractive to dormice and allowing access further up the valley.

The first dormouse nut was found under riverside nazels after 10 minutes. A total of 50 minutes was spent in searching resulting in a total of 13 dormouse nuts being found, most of them towards the end of this period.

Site C (SN572082)

Ancient woodland less than .1/2 kilometre from Site B and divided from it by the dual carriageway of the A48. It occupies a deeply incised dingle containing a tributary stream of the Afon Gwili. At the east end of the wood and on the 'north side of the stream the wood is quite scrubby. Predominantly the wood is full of very tall trees which because of their height and vertical form are not very well interconnected with the vigorous understorey. The structure of the woodland "improves" further up the dingle where there are more lateral branches coming out of the large, mature oaks and ashes. Other tree species include sycamore and birch. There is a good understorey of hazel, holly, hawthorn with abundant honeysuckle and bramble.

It proved difficult to find nuts because of either shading of the hazels or the hazels overhanging undergrowth and bramble. The first dormouse nut was found after 40 minutes under hazels at the northern edge of the wood against the adjacent field. A total of 5 or 6 dormouse nuts was found in one hour's searching.

N.B. This site is a very short distance from site B and the dormice at both sites at one time probably formed the same population. Now, because of the dual carriageway, they are presumably separated.

Site D (SN531075)

Ancient woodland alongside Afon Morlais, near "The Mill" above Pont Morlais. This site proved different from what I had expected as the wooded hillside on the east side of the river had been replanted some time ago with conifers - not apparent from the map given in the "Inventory of Ancient Woodland". However, there is a strip of broadleaved trees along the river and an area of broadleaved woodland on the west side which appeared to have a good structure with a healthy understorey including hazel. It was too late in the day to take detailed notes of species components of this woodland.

I searched under hazels growing on the east bank close to the conifers and found the first dormouse nut after 5 or 10 minutes. Half an hour was spent searching and a total of 5, possibly 6, dormouse nuts were found.

Site E (SN546052)

Ancient woodland alongside the Afon Morlais and nearly 3 kilometres below Site D. At this point a minor road crosses the Morlais; to the north of the bridge the woodland seemed to be mostly conifers but with deciduous trees and shrubs near the river. On the south side and on the east bank of the river is a small block of deciduous woodland. Here, there is a small flood plain between the river and a wooded bank on the east side. The trees on this bank are mature oak, beech Tagus sylvatica, sycamore with elder Sambucus nigra, guelder rose Viburnum opulus and hazel with a ground flora of bluebells. The floodplain contained substantial hazels, large willows, alder and birch. The ground here is quite damp in places and flora included ivy Hedera helix, opposite-leaved golden saxifrage Chrysosplenium oppositifolium, wood anemone and pignut Conopodium magus. There is also honeysuckle and blackthorn Prunus spinosa. Aerial pathways are available both across the river and across the road. At the south end of this block there appears to be a bit of a gap between it and the riverside trees which themselves are gappy at this point; however it seems to connect with a well-wooded dingle coming in from the north.

The first dormouse nut was found after 5 minutes and after one hours search there was a total of 2, possibly 6 dormouse nuts - all found under hazels in the floodplain.

Site F (SN554035)

Riverside trees along the Afon Morlais, close to a block of ancient woodland  and a little over 2 kilometres below Site E. The large block of ancient woodland has all be replanted with conifers, but the strip that has been left i.e. the trees lining the river on both sides, and also areas of broadleaves and small wooded tributary valleys on the east side, are all probably "ancient". The search site is along the west bank at a point which is separated by an area of open ground from the forestry to the west. However, this "open" ground is full of scrub, thick bramble, bracken and some hazels. The river itself gives the appearance of being well-wooded with, on both

14 banks, large mature oaks, ash, sycamore, beech, holly, alder, birch, hazel. The hazels are big specimens and gave a big crop of nuts.

The first dormouse nut was found after 5 minutes and in half an hour a total of 6, possibly 7, were found.

Site G (SN567027)

A block of broadleaved woodland near Maesydderwen to the east of Llangennech. The northern part of a wood that seems to have been bisected by the bypass around Llangennech. An attractive wood with an "English" character, formerly part of the Llangennech Park Estate. The wood on gently sloping ground is entirely planted (ie not ancient woodland), perhaps in the early years of the nineteenth century - some 460,000 trees were planted by the Estate in c1804 (Davies & Miller, 1944). Mostly of tall, mature oaks with a few big ashes, sycamore. A good hazel understorey and also hawthorn, blackthorn. Areas of open ground had a flora of bluebells, violets while other areas were covered with bramble scrub. The habitat appeared to be good for dormice but isolated - no hedgerows running out of it.

No dormouse nuts were found in an hours searching. Site H (SN571053)

Ancient woodland alongside the Afon Gwili and between the road from Llannon to Hendy and the M4. 3-4 kilometres below Site B. Woodland on both sides of the river, quite scrubby at the southern end, but further up are quite large oaks, ash, sycamore, also alder, willow, holly, birch, apple Malus sp. Good understorey of hazel, hawthorn and a lot of ivy, bramble and honeysuckle. Ground flora includes celandine Ranunculus ficaria and kingcup Caltha palustris. Aerial pathways exist - over the river and up the valley.

Excellent habitat but difficult to find good nut trees until towards the end of a search lasting 3/4 hour. The only dormouse nut found was one by accident before starting the search, as I crawled underneath a fallen tree.

Site I (SN528091)

This site nearly 2 kilometres above site D was selected to try and determine the extent of dormouse distribution along the Afon Morlais. It is near the headwaters of the river which at this point is no more than a stream running through rough and fairly wet pasture. The river here is fringed with scrubby growth - mostly willow, alder, hazel, hawthorn, oak (mature but stunted?), rowan Sorbus aucuparia, holly, also guelder rose, bramble, ivy, honeysuckle. Ground flora includes fern and wood sorrel Oxalis acetosella.

The first dormouse nut was found after 10 minutes and a total of 2 in 1/2 hours searching.

Site J (SN515103)

This site was chosen as I was making my way towards Site K. Scrubby woodland occurs on each side of an old disused mineral railway line which ran from Llanelli Docks to Cynheidre Colliery and on to Tumble and beyond. Just to the north of the bridge that carries the Pontyberem to Llannon road over this line there is a wooded bank on the east side of the line. The wood is primarily of ash, oak, hawthorn, blackthorn, hazel with bramble.

One dormouse nut was found after 10 minutes. No more were found in a total search time of 20 minutes.

Site K    (SN513099)

Along the mineral line, less than 1/2 kilometre south of Site J. The line goes over a steep dingle which is well-wooded as are the embankments. In the dingle itself wet woodland of alder, birch and ash occurs with large oaks on the drier ground on the embankments. There is a good understorey of hazel, holly, ivy, bramble and abundant honeysuckle. Ground flora includes bluebell, wood anemone, cow wheat Melampyrum oratense, celandine, kingcup. Further down the valley the wet woodland runs into an area of ancient woodland which I didn't investigate but which seems to offer good habitat. A thick woodland of oak and ash with a good understorey.

found the first dormouse nut after 5 minutes, under a hazel on the embankment. There were heavy-cropping hazels down in the dingle but no more nuts found until returning to the top of the embankment I found one more as : followed an old hazel hedge. A total of 2 nuts in 3/4 hour search.

Site L (SN490092)

Ancient woodland near Pont-Henri and alongside the Afon Hafren, a tributary of the Gwendraeth Fawr. A scrubby woodland in parts, in other areas a rather leggy, vertical woodland but with a lot of understorey with interconnections between it and the tree layer. Undergrazed by animals. Willow, alder, ash, holly, birch. Large oaks along the stream. Hawthorn, a lot of bramble and honeysuckle. Ground flora includes wood anemone.

It wasn't easy to find good hazels because of shading, or overhanging brambles, grass etc. No dormouse nuts found in an hour's searching.

Site M (5N474097)

Ancient woodland near Pont-Henri and alongside the Aber Lledle, a tributary of the Gwendraeth Fawr. Most of the woodland on the west side of the stream has been planted with conifers, a larch Larix sp., plantation. The stream itself is fringed with broadleaves and there is a block of broadleaves on the east bank, thinning out for a while up the valley before again enlarging on the east side. Woodland predominantly of large oaks, sycamores, holly, willows, with an understorey of hazel, also bramble and honeysuckle.

Initially looked on the east side of the stream, where there were some heavy-cropping hazels at the top of the bank near the wood-edge - but found no nuts. Then looked under hazels at the bottom of valley and on the west bank found 2 dormouse nuts almost immediately. Total search time about 25 minutes.

Site N (5N488076)

Ancient woodland near Cynheidre and on the steep banks of a valley containing a stream flowing westwards which joins the Afon Hafren, south of Pont-Henri. A thick woodland and vertical in structure, less so on the southside of the steam where more lateral growth affords better pathways for dormice. Main trees are oak, ash, sycamore, alder and willow with a quite vigorous understorey of hazel, holly. A lot of honeysuckle, bramble. Ground flora includes bluebell and wood anemone.

Probably because of shading it was difficult to find good cropping hazels. I found no dormouse nuts after a search of 3/4 hour.

Site 0 (SN4S2080)

Ancient woodland at Coed y Glyn, near Glyn fach. Woodland borders a stream which' is another tributary of Gwendraeth Fawr. A nice wood with a good structure and healthy understorey, ungrazed.

I wasn't able to find many fruiting hazels. I looked for c.1/4 hour before moving on. No dormouse nuts were found.

I moved on from here to another ancient woodland nearby to the west (Cwm Clydach). It occupies a steep dingle on the east side of the minor road going north from Pont-newydd. A stream, another tributary of Gwendraeth Fawr runs in the bottom of a narrow gorge. The wood contains oak, ash, holly. There is a lot of honeysuckle and some hazel but I was unable to find any nuts.

Site P (SN478086)

Ancient woodland alongside a stream that joins the Afon Hafren nearby which in turn shortly joins the Gwendraeth Fawr. This is the same stream and the other end of the woodland as at Site N. The woodland contains old mineral workings. Scrubby woodland on hillside to the north probably secondary, consisting of oak, birch, holly and'hazel. Between here and the stream is probably ancient woodland; apart from some ashes and oaks it mainly consists of tall scrub, hawthorn and hazel. The hazels and hawthorns are very high -c.30', and the hazels heavy croppers. A lot of ivy, honeysuckle, bramble and patches of bracken. Good aerial pathways available.

I found the first dormouse nut after 10 minutes and in a search lasting 25 minutes I found a total of 2. I found a honeysuckle that could have been stripped by dormice.

Site 0    (SN435072)

Ancient woodland near Morfa Bach and alongside a stream which is another tributary of Gwendraeth Fawn. An attractive wood containing large mature oaks, ash many of them ivy-covered. Understorey almost entirely of hazel with some holly. Mostly open ground which by its flora indicates slightly base-rich conditions; mostly bluebell, also wood anemone, yellow pimpernel Lysimachia  nemorum, harts-tongue fern Asolenium scolopendrium, pignut, dog's mercury Mercurialis perennis. Patches of bramble and some honeysuckle.

Aerial pathways connect the wood to the hedge on the other side of the road.

I searched under hazels both in the wood and along the hedge for c.3/4 tour but didn't find many nuts - and no signs of dormice.

Site R (SN435053)

A small area of scrubby woodland bordering another (much smaller) Afon Morlais and a dismantled railway line north of Llandyry. This woodland may extend some distance down the railway line but I confined my search to the corner on the west side of the B4317 near the bridge over the stream. There are mature oaks (ivy-covered) to the north of the railway line, otherwise there are alders by the stream and some willows. Quite a lot of hazel and thick bramble especially along railway line. Also some Japanese knotweed Polygonum cuspidatum.

I searched under a few big hazels but there was a lot of undergrowth like ivy and brambles. However, there was one good cropping hazel above bare ground near the bridge and here I found one dormouse nut after 1/2 hour when I moved on to the next site.

Site S (SN42-02-)

Ancient woodland at Coed y Marchog and behind Coed farm. Habitat looks excellent for dormice with mature oak and ash. Quite an Open canopy, under which smaller sycamores growing up and a vigorous understorey of hazel, hawthorn, elder, holly. There are also scattered conifers, larch and pine Pinus sp., and some elms Ulmus sp., (in flower). There is a thick undergrowth of rampant brambles and bracken. Ground flora includes bluebells. I looked under large hazels at the bottom and especially towards the top of the wood which covers this western flank of Mynydd Penbre. Although these hazels were large and covered in catkins I didn't find many fresh nuts underneath - they were mostly old specimens. There were good hazels with fresher nuts behind -Coed" but I found no dormouse nuts after 3/4 to 1 hour searching. A large woodland with seemingly very suitable habitat, so a disappointing result.

Site T (SN481026)

Ancient woodland alongside the Afon Cwm-mawr near the hamlet of Cwmbach. Woodland on the north side of the road from Llanelli to Trimsaran. Woodland to the south of the stream and sloping down from the road is of a fairly vertical structure, consisting of oak, birch, ash, holly, hazel with bramble and honeysuckle. The best hazels I found on the other side of the stream where there is a grove of large hazels, associated trees being alder and oak and near to thickets of bramble also hawthorn.

The ground underneath these hazels was quite bare and there were hundreds of nuts but didn't find a dormouse nut till after 40 minutes searching, and I left shortly afterwards.

There are a number of old (mostly filled in) mine adits in this woodland.

I had also looked at some ancient woodland in the vicinity of Stradey Castle, especially on the west side of the road going to Pwll. The habitat looked suitable for dormice with very mature oak, beech, chestnut Castanea sativa, sycamore and a good understorey of hazel, holly, bramble. The structure here was better than on the other side of the road, towards Stradey Castle, where the trees were more vertical, towering to a great height. However, I couldn't find any fruiting hazels, so I didn't stay.

Site U (SN514039)

Woodland on the west side of Cwm Lliedi Reservoir. This woodland looks perfect for dormice. Predominantly of mature oaks, with some birch and rowan, holly, willow and alder, probably of uneven age giving rise to gaps in the canopy which no doubt favour the very vigorous understorey of mostly hazel with abundant honeysuckle. The oak trees are very twisty with many lateral branches going off and connecting well with the understorey. Tangles of bramble are everywhere, the ground mostly covered with ivy which also covers many of the trees.

I confined my search to a small patch of hazels at the western end. I found 2 nuts after 2'minutes and a total of 10 in 20 minutes.

I noted the proximity of this wood to a mineral railway line (disused), on the other side of the public footpath at the western edge of the wood. This is the same railway line as at Site J and K. It could form a useful corridor for wildlife, including dormice. Fairly severe hedge-cutting and laying was going on in fields just to the north of this wood.

Of the 21 sites visited and from which a sample of hazel nutshells was taker., 15 proved "positive" of which 10 were in areas of ancient woodland.

It proved comparatively easy to find signs of dormice: at 11 of the 15 sites the characteristically opened hazel nuts were found within 10 minutes.

5. Discussion

From a situation a short time ago when very little was known about the occurrence of the dormouse in Carmarthenshire it now appears as a result of this survey that at least the south-eastern part of the vice-county could be something of a "stronghold" for this species. Signs of dormice were mostly found in river valleys, primarily because these are where ancient woodlands for the most part occur but they might also have additional value in providing shelter and corridors for the animals to move along. Certain valleys seem to have particularly strong populations, especially the Afon Morlais where signs were found over approximately 7km length of river, and also the Afon Gwili, Gwendraeth Fawr and Afon Lliedi. The occurrence of dormice in the valleys of the Morlais and the Gwili helps perhaps to account for their presence at Llwynteg, Llannon (more or less midway between these two valleys): their living here in sub-optimal habitat could reflect a dispersal from strong populations occupying better habitats in these valleys. The populations along the Afon Morlais are not now probably as great as they once were, much if not most of the ancient woodland having been converted to forestry plantations.

To obtain a better picture of dormice in this area, further surveys should be undertaken. Firstly, those areas of ancient woodland which I looked at and though appearing suitable for dormice failed to show signs, should be looked at again as I might easily have missed evidence owing to the time constraints of this survey. As for the "positive" sites, surveys should be carried out to see how well connected they are (by aerial routeways in the form of tree lines and hedgerows) and whether any sites have become isolated. Very few of the sites which I found to contain dormice were big enough to support a viable population on their own: Paul Bright estimates that in this respect a woodland needs to be at least 20 hectares in extent, hence good "corridors" are vital in this area where most woodlands are quite small. Further surveys of other blocks of ancient woodland and also, perhaps just as important, areas away from ancient woodland, would also help to build up a better picture. None of these surveys would help In estimating the size of a population in any area for this dormouse nest-box projects should be initiated in certain suitable localities.

As for the conservation of the dorMouse in this part of the country, opencast  coal-mining probably poses the biggest single threat. Dormice are likely to occur anywhere in the coalfield and where any site is proposed for open casting surveys to establish the presence of the animal should be undertaken as a matter of course. Not only would opencasting completely destroy dormouse habitat it could also destroy the routeways used by the animals, leading to isolation and hence extinction in certain areas. Much of the Gwendraeth Fawr valley has been proposed by British Coal as sites for opencasting and dormice seem to be particularly widespread here.

For reasons already mentioned it is necessary that hedgerows in the area are conserved and hopefully managed in a way suitable for the dormouse. New hedgerows of course would be beneficial, especially if species-rich. It is not clear how gaps in hedgerows affect dormice. Paul Bright maintains that dormice are extremely reluctant to come to the ground and that they will make a wide detour to avoid a gap. However, I believe that most of his studies have taken place in areas with strong resident populations and not in hedgerow situations with animals moving along or dispersing. Doubtless dormice cross gaps along gates but here, to introduce a detail, wooden gates could be better than metal ones, as dormice don't like putting their feet on cold metal: (pers. comm. John Messenger).

Apart from hedgerows, other corridors which are important for dormice include riverside trees: these must be conserved as not only are they very often large mature trees providing good feeding for dormice and nesting sites in tree hollows, they also appear vital as routeways in many places. Sometimes these trees are mismanaged (i.e. cut down) by landowners, otherwise pernaps the main threat comes from river-management schemes by NFU, where sometimes trees are felled or pollarded, but the rivers in this area are perhaps not likely to attract such schemes.

The several old disused mineral railway lines in this area could be important as corridors, as many are edged by rank overgrown hedges and themselves provide good habitat.

There is considerable scope for improving some broadleaved woodlands for dormice. Some woods could be carefully thinned to reduce shading and hence promote the flowering and fruiting of various tree species. Coppicing has long-- been thought to be beneficial to dormice but it has to be done very carefully, preferably in small blocks on a long rotation.

Finally, the public profile of the dormouse could be raised locally. By presenting it to the public more people would become aware of its special needs. Various events could be promoted as part of National Dormouse Week, perhaps in the Llanelli area: the woodland to the west of Cwm Lliedi Reservoir would be excellent in this respect, an ideal habitat containing abundant signs of dormice and adjacent to a popular place of recreation near to Llanelli.

Similar surveys to this could be carried out in other parts of Carmarthenshire. Although I'm sure many other sites will be found I feel it is unlikely that the dormouse will be found at such a density as in this area. This would raise other interesting questions to answer: what limits the distribution of the dormouse - availability of habitat or factors of climate?

6,    References

Bright, P. & Morris, P. (1989). A Practical Guide to Dormouse Conservation. Mammal Society.

Bright, P., Mitchell, P. & Morris, P. (Feb. 1992). Dormouse Distribution in Herefordshire: survey methods and effects of wood age, size and isolation.    English Nature.

Davies, B.L. & Miller, H. (1944). in The Land of Britain - The Report of the Land Utlilisation Survey of Britain, Ed. L.D. Stamp: Part 39 Carmarthenshire. London.

Lister, J. & Whitbread, A. (Jan. 1988). Carmarthenshire Inventory of Ancient Woodlands (Provisional). NCC.

Lucas, A. (1993). The Dormouse in Carmarthenshire - Filling in the Gaps. Llanelli Mats. Newel. Winter 1992-93: 30-31

Matthew, N.R. (1992). The Dormouse. Llanelli Mats. Newsl. Winter 1991-92:9-10 Dormouse Survey 1993    Appendix 1    Summary of search sites

Dormouse sites

Location    Grid Ref    Date    No of nuts, if any

Site A. Ty Isaf    SN576088    19.2.93    2 dormouse nuts

Site B. Plas Newydd Mill    5N577080    20.2.93    13 dormouse nuts

Site C. Nr. Ystlysycoed Uchaf    SN572082    20.2.93    5, poss 6 dormouse ruts

Site D. Nr. Mill, Pont Morlais    SN531075    21.2.93    5, poss 6 dormouse nuts

Site E. Nr. Gelliwernem Wood,  Afon Morlais    SN546052    22.2.93    2, poss 6 dormouse nuts

Site F. Nr. Troserch, Afon Morlais    SN554035    22.2.93    6, poss 7 dormouse nuts

Site G    Nr. Maesydderwen, Llangennech    SN567027    23.2.93    0 dormouse nuts

Site H. Nr. Geibren Farm, A. Gwili    SN571053    23.2.93    1 dormouse nut

Site I. Nr. Gellidoc Farm, A. Morlais    SN528091    26.2.93    2 dormouse nuts

Site J. Mineral line, n. of bridge    SN515103    26.2.93    1 dormouse nut

Site K. Mineral line & wooded dingle    SN513099    26.2.93    2 dormouse nuts

Site L    Woodland along Afon Hafren    SN490092    27.2.93    0 dormouse nuts

Site H  Woodland at Glynhir, Pont-Henri    SN474097    27.2.93    2 dormouse nuts

Site N    Valley woodland, Cynheidre    SN488016    27.2.93    0 dormouse nuts

Site 0    Coed y Glyn, Glyn fach    SN452080    27.2.93    0 dormouse nuts

Site P. Nr. Ynyshafren Farm, Pontyates    SN478086    28.2.93    2 dormouse nuts

Site Q    Morfa Bach Wood    SN435072    28.2.93    0 dormouse nuts

Site R. Nr. old railway line, Llandyry    SN435053    28.2.93    1 dormouse nut

Site S    Coed y Marchog & Coed    SN4202    28.2.93    0 dormouse nuts

site T. Cwm Mawr nr. Cencoed Isaf    SN481026    1.3.93    1 dormouse nut

Site U. Woodland west of Cwm Lliedi Reservoir    SN514039    1.3.93    10 dormouse nuts

Appendix 2    Dormouse Records for Carmarthenshire    March 1993

Date    Location    Grid Ref    Habitat Type    Type of Record    Recorder

c.1903    Cwmffrwd    SN425165    T.W. Barker.

1978    Nr. Rhandirmwyn    SN7644    Woodland    Sight    E. Murrell

1979    Upper Lliedi Res.    SN5004    Bird box in oak    2 hibernating    M.H. Willis: tree animals

c.1980    Capel Dewi    SN4820    Hedgerow    Hibernating    D.Jarman

1980    Tir Philip,    SN4818    per I.K.Morgan Capel Dewi

1980    Llanarthne    SN5320    per S.Seery

1983    St. Ishmaels    SN3608    Coastal scrub

1990-91    Llannon    SN551080    Hedgerows    Caught by cat.    J&K Crowden Nuts.

Sept 1992    Four Roads    SN445098    Hazel coppice    Sight    T. Wells and scrub.

Sept 1992    Pontyberem    SN512124    Woodland    Nuts    J. Crowden

Sept 1992    Pontyberem    SN505115    Back garden    Nuts    per L.Gande

Oct. 1992    Coed y Garn,    SN516145    Ash woodland    Nuts    A. Lucas Nr. Drefach    (ancient)

Nov.1992    Coed Gwempa,    SN436115    Oak woodland    Nuts    A.Lucas Nr Llandyfaelog    (ancient)

Dec.1992    Nr. Ferryside    SN377112    Hazel hedge    Nuts    A. Lucas

Dec.1992    Nr. Crwbin    SN468129    Ash woodland    Nuts    A.Lucas (ancient)

Feb-March 1993 Dormouse Survey by N.R. Matthew 15 sites - see Appendix 1

Mar. 1993    Nr. Tumble    SN527121    Ancient Wood    Nuts    E.Lewis

Mar. 1993    Nr. Pont morlais    SN5307/5407 -    J.Crowden