A Review of the Occurrence of Bats (Chiroptera) on Islands in the North East Atlantic and on North Sea Installations

The bats recorded from Iceland, the Faroe Islands, the Shetland Islands, the Orkney Islands, and North Sea installations are reviewed to the end of 2012. In total 12 species have been positively identified, while a considerable proportion of all records are sightings of unidentified bats. Eight of the species are European in origin and four originate from the New World. The largest number of species (8) has been recorded in Iceland, but the greatest number of individuals (180) has been found in Orkney. The bat invasion on the Faroe Islands in 2010 is without precedence, when 70 observations of a minimum of 45 individuals were noted. Most bat observations in the study area occurred in the autumn, with fewer in the spring. Most observations were of single animals, but there were also sightings of up to 12 individuals. There has been a marked increase in bat records in the past three decades. We discuss whether this is a real increase, or due to improved communications, increased public awareness, increased shipping, changes in weather patterns and/or the effects of climate change. All factors appear to be involved.


INTRODUCTION
The order Chiroptera is widespread through most parts of the World. The greatest proportion of families, genera and species are found in the tropics and sub-tropics, with a lower proportion in temperate zones. Within temperate zones, the number of species decreases with increasing latitude. In temperate zones, the insectivorous bats are restricted to 3 those areas that are capable of providing good supplies of food to sustain them through the spring and summer breeding periods and enable them to build up adequate fat reserves in the late summer and autumn ready for their retreat into hibernation through the winter (Yalden and Morris, 1975;Hill and Smith, 1984;Rolland et al., 2014).
Migratory behaviour and vagrancy are particularly well-known in birds and bats (Ahlén, 1997;Hutterer et al., 2005;Ahlén et al., 2009). Their ability to fly makes it easier for them than for most other animals to disperse outside their normal range. Bats employ a variety of strategies to make best use of the available habitats. Some species occupy relatively small territories, travelling only short distances in search of food and between breeding roosts and winter hibernacula, while other species migrate for medium to long distances, in order to exploit resource-rich areas.
The North East Atlantic islands, comprising Iceland with its fluctuation between subarctic winters and temperate summers and the Faroe, Shetland and Orkney islands with temperate but cold winters and cool summers, with one exception do not have resident populations of bats. Nor are the islands considered to be on normal migration routes, although stragglers do reach them on occasion. The species involved and the frequency of these occurrences undoubtedly depends on the distance from their normal range, but a number of other factors, such as population numbers, vagrancy, and unintentional transport by humans, are also important. In the long run, global environmental changes such as climate change may also modify traditional distribution patterns.
The present study examines records of bats from the North East Atlantic islands of Iceland and the Faroe Islands to the end of December 2012, including those reported in previous publications (Koopman and Gudmundsson, 1966;Petersen, 1993Petersen, , 1994Baagøe and Bloch, 1994). Comparisons are made with published data on bats recorded from the Shetland 4 Islands and the Orkney Islands, off the north coast of Scotland, as well as oil and gas installations in the North Sea.
Information is provided on all of the species recorded from the study areas; the locations of the records are mapped and the species composition, distribution and seasonal occurrence analysed. The possible reasons for the marked increase in the occurrences of bats in these study areas, especially the unusual invasion of the Faroe Islands in 2010, are discussed.

MATERIALS AND METHODS
This study summarizes all available information to the end of 2012, based on specimens of bats preserved mainly in the collections of the Náttúrufraeðistofnun Íslands The specimens preserved in the collections of IINH and FMNH have been acquired opportunistically over many years by interested members of the public. Specimens were identified by the use of literature available at the time of acquisition and by consultation with staff in other European museums containing international collections of bats. Photographs are available as proof for some bat sightings, although the species could not be established in most cases. Sound recordings using a Pettersson Ultrasound detector D100 have also been used in the Faroe Islands. This detector type may only be used to locate flying bats but not for 5 species identification. In Iceland and the Faroes, sightings were assessed by interviews and requests for photographs because of the problems that exist due to the general public confusing bats with insects and birds. Insects which are especially likely to cause confusion are the larger moths (Jensen and Sivertsen, 2010), while small petrels (seabirds) have an erratic flight similar to that of bats and may be easily misidentified at a distance.
Records of bats from the Orkney and Shetland Islands, and from North Sea installations were obtained by literature searches. Records of bats from the Orkney Islands have been particularly well documented over the last two decades, as a result of observations by members of the very active Orkney Field Club (Booth, 1992-Booth, 2012. Information for the Shetland Islands was initially derived from the website of the Shetland Biological Records Centre (http://www.nature-shetland.co.uk/brc/bats.htm) and (http://www.natureshetland.co.uk/naturelatest/archives/), and for North Sea installations from Swift (2004); where possible these records have been traced back to the original references (Appendices 3 -

5).
Coordinates for localities on Iceland were determined using Google Earth (http://earth.google.com), while the website http://www.findlatitudeandlongitude.com was used for localities on the Faroe Islands. Coordinates for localities on the Shetland and Orkney Islands were principally located by reference to the gazetteer (Ordnance Survey, 1999) supplemented by Google Earth (http://earth.google.com). Coordinates for North Sea Installations not included in the original references were determined using maps of North Sea oil and gas fields (http://www.acorn-ps.com/web/page/oilgas/nsfields/snsmap.htm and http://tools.decc.gov.uk/en/content/cms/tools/quad_maps/quad_maps.aspx). Distribution maps were prepared using DMAP (http://www.dmap.co.uk).

6
To the end of December 2012, 12 species of bats all belonging to the family Vespertilionidae have been recorded from the area of study: eight species from Iceland, six species from the Shetland Islands, and five species each from the Faroe Islands, the Orkney Islands and North Sea installations and vessels. The results summarised in Table 1 show that the greatest number of records is 180 for Orkney, 96 for the Faroes, 58 for Shetland, with totals of 38 and 30 respectively for Iceland and North Sea installations. Information on each of these species is provided below.

The records
All the records for North Sea installations and the Shetland Islands were of single animals. In Iceland 31 of the records (94%) were of single animals, with one instance of two animals and another of six. This differs from the Faroe Islands, where 74 records (77%) were of single animals, 21 of two (10%), three (5%), or four (6%) animals. In the Orkney Islands four of the species were recorded as single animals, but there were many multiple sightings of Pipistrellus pipistrellus ranging from three to 12 animals.
A considerable proportion of the records are of unidentified bats. Less than half of the Icelandic and Orkney records fall into this category but as many as two-thirds of the Faroese records. This is explained by the variable nature of the records, for while only a third of the bat records from Iceland were sightings, the majority (76%) in the Faroe Islands were sightings. Just over half of the Icelandic records are of bats which were found alive, while this category was 16% for the Faroe Islands. For the Orkney Islands, although only 8% of the records are of identified individuals, 39% of the records are of sightings identified to species with the use of a bat detector and only 11% are sightings. In contrast all of the published records for North Sea installations are identified to species.

Species accounts
Eptesicus fuscus (Palisot de Beauvois, 1796) 7 Occurs from South Canada to Colombia and North Brazil; Alaska, Greater and Lesser Antilles, Bahamas (Simmons, 2005). There are two records from England of imported specimens (Hutson, 2008a).
A single individual was found alive in March 2000 in a store in Reykjavík, Iceland that imports fruits and live plants from different parts of the World. The bat was believed to have arrived in a container of bananas from South America. This species is very similar to and may be conspecific with E. serotinus (Simmons, 2005). The specimen was identified on the basis of its smaller size and probable origin.
Eptesicus nilssonii (Keyserling and Blasius, 1839) Occurs in central and eastern Europe (Dietz et al., 2009). It is considered to be the most common and widespread species of bat in Norway and also in Sweden, where it has been recorded north of the Arctic Circle in both countries (Syvertsen et al., 1995;Ahlén, 2011;Frafjord, 2013). It occurs above 64° N in northern Finland (Siivonen and Wermundsen, 2008). There are very few records in England (Hutson, 2008b).
Single individuals have been recorded respectively in August 1993 from the "Dundee Explorer" oil platform in the North Sea (Speakman et al., 1995) and from the Faroes in May 1994 (Fig. 2).
Eptesicus serotinus (Schreber, 1774) Occurs throughout Europe as far north as 55° N (Dietz et al., 2009) and also in southern Sweden (Ahlén, 2011). In England it is found mainly in the southeast, but also in central and northwest England and Wales (Entwistle et al., 2001;Hutson, 2008b).
There is a single record from Whalsay, Shetland in October 1991 (Fig. 3) and a recently deceased specimen was found in timber storage on the Faroes in April 2007 (Fig. 2) and may have arrived by ship.
Lasiurus cinereus (Palisot de Beauvois, 1796) 8 Occurs in Canada, throughout the USA, Mexico, Guatemala, most of South America, Bermuda and the Galapagos Islands (Simmons, 2005). This species has a broad range to the maximum latitude of 64° 20' N in the USA and Canada (Pierson, 1998). However this latitude corresponds to the extralimital record on Southampton Island mentioned by Hill and Yalden (1990), who give the summer limit far lower at latitude c 53° N in southern Canada, whereas in winter the limit is up to 38° N in the USA. Cryan (2003) provided information on distribution patterns throughout the year in North America.

Myotis lucifugus (LeConte, 1831)
This species has a broad range to a maximum latitude of 66° 34' N in the USA and Canada (Pierson, 1998), occurring from Labrador and Newfoundland in Canada to Alaska, southern California, northern Arizona and northern New Mexico in the USA (Simmons, 2005). There is a single record from England of an imported specimen (Hutson, 2008a).
A single immature specimen was recorded from Reykjavik, Iceland in August 1944 (Guðmundsson 1944;Ryberg 1947;Hayman, 1959;Koopman and Gudmundsson, 1966;Jakobsson 1967;Petersen, 1994). Another specimen was found August 1981 on board a ship arriving in Reykjavik harbour from the east coast of the USA (Fig. 5).
Myotis septentrionalis (Trouessart, 1897) Found in Eastern United States and Canada west to British Columbia, E. Montana, E Wyoming and south to Alabama, Georgia and Florida Panhandle (Simmons, 2005). This species has a broad range in Canada, intermediate in the USA to the maximum latitude of 61° 25' N (Pierson, 1998). There are no reports of this species from the British Isles. Petersen (1994) recorded two specimens as Myotis keeni septentrionalis from Iceland in August 1981 and August 1993. Myotis keeni and M. septentrionalis are currently regarded as distinct species (Simmons, 2005). The 1981 specimen was found at the same time as another North American species, Myotis lucifugus, and four other bats that were discarded before they could be identified. All specimens were found on board a ship that came from the east coast of the USA, so the identification of this specimen is consistent with M. septentrionalis, whereas Myotis keeni (Merriam, 1895) is distributed on the west coast of the USA (Alaska Panhandle to West Washington (Simmons, 2005). The other Icelandic individual flew into a house and died, but could have arrived by ship at the nearby Hafnarfjörður harbour.
Nyctalus leisleri (Kuhl, 1817) Occurs throughout western Europe to the Urals, Caucasus and Turkey to nearly 57° N, however there are no records from Denmark and Scandinavia, except for southern Sweden (information from Simmons, 2005;Dietz et al., 2009;Ahlén, 2011). Although rare, this species occurs throughout the British Isles, mainly in central and southern counties of England, also in south western Scotland, a few records from northeast Scotland, and it is widespread and common in Ireland (Haddow and Herman, 2000;Entwistle et al., 2001;Sheil et al., 2008).
Single individuals have been recorded respectively from the Faroes in July 1984 (Baagøe and Bloch, 1994) (Fig. 2), Iceland, where it was captured alive in May 2000 (Fig. 5), and on an oil rig in 2002 (Swift, 2004). There are three records from Shetland in July 1968, August 1978, October 1996(Corbet, 1970; Shetland Biological Records Centre) (Fig. 3). The reference to a record from Orkney (Swift, 2004) is probably in mistake for one of those from Shetland.
Nyctalus noctula (Schreber, 1774) The species is widespread in Europe to the Urals and the Caucasus (Dietz et al., 2009). It has been recorded rarely in southern Norway (Syvertsen et al., 1995) and has an uneven distribution in central and southern Sweden (Ahlén, 2011). The species is widespread in England, Wales and southwest Scotland but is absent from Ireland (Haddow and Herman, 2000;Entwistle et al., 2001;Mackie and Racey, 2008).
A few bats of this species have been recorded from all study areas except Faroes. A pregnant female was recorded from Orkney in June 1976 (Racey, 1977), others in September 1978 and October 1988 (Booth, 1979(Booth, , 1989  Pipistrellus nathusii (Keyserling and Blasius, 1839) Occurs in western Europe, from southern Scandinavia southwards to France and northern Spain and eastwards to the Urals and Caucasus (Dietz et al., 2009). It has been recorded in Norway (Syvertsen et al., 1995;Swenson et al., 2010) and is found in central and southern Sweden as far north as 61° 21' N (Ahlén, 2011). Occurs in southern and eastern Scotland, most of England and Wales, and eastern and central Ireland (Russ et al., 2001;Russ, 2008). This is the only species of pipistrelle reliably recorded from North Sea installations, Shetland, Faroes and Iceland, where it is also the most commonly recorded species of bat (Table 1; Figs. 2-3, 5-6). It has also been recorded on the Orkneys (Fig. 4), but here it is less frequently encountered than is P. pipistrellus (7% of the total P. pipistrellus and P. nathusii records and only 3% of the total records of bats from the Orkneys). The earliest record of this species dates from Shetland in November 1940 (Herman, 1992), the other records from Shetland cover the period from 1987 to 1998, with most records during the autumn and winter months (Russ et al., 2001). Similarly, most of the 17 records from the Faroes were in the autumn and winter months and, of the 20 individuals recorded from North Sea rigs in the period from 1988 to 2001, most were in September (Speakman et al., 1991(Speakman et al., , 1993Russ et al., 2001). There were six records from Iceland dating from 1971 to 2012, most in autumn and winter, with one in May. Of the records from Orkney from 1995 to 2010, two were in June and three in September. Four of these were reliably identified specimens, the fifth a sighting confirmed by use of a bat detector (Booth, , 2003. It is possible that some of the earlier unidentified Pipistrellus records in all study areas are also representatives of this species. Pipistrellus pipistrellus (Schreber, 1774) Widespread in Europe, more common than P. pygmaeus in central Europe, but rare or absent in the Netherlands and Scandinavia (Jones and Racey, 2008;Ahlén, 2011). Details of distribution remain uncertain on mainland Europe. Occurs throughout the British Isles (Barlow and Jones, 1999;Jones and Racey, 2008).
Reliable records of P. pipistrellus, based on professionally identified specimens and the use of bat detectors, have only been obtained from the Orkney Islands (Table 1 and Fig.   4), where it is by far the most commonly recorded species (75% of the total records of Pipistrellus and 39% of the records of bats from the Orkneys). There are 71 reliably identified records of this species dating from the earliest record in September 1908 (Spence, 1909) while the majority of the records cover the period from 1992 to 2011. Over this period the earliest sighting occurred in April, the latest in November, with the majority of sightings occurring from May to September.
Although the Orkney Islands may be considered marginal for the existence of resident bats, there is evidence of summer roosts of P. pipistrellus from Melsetter, Hoy from 1994 to 2002 (Holmes, 1994;Holmes et al., 1995;Booth, -2003; from the Finstown area of Mainland from 2000 to 2010 (Booth, 2001; from Binscarth Wood, Mainland from 2007 to 2008 ; and there were sightings during the summer months from Evie, Mainland and on Hoy during 2010   (Fig. 7). Although a summer roost was apparently well established at Melsetter on Hoy for nine years, there have been no subsequent reports from this locality .
There are four records of this species from Orkney: one in February 1931 (Marwick, 1931), another in October 1948 (Booth, 1986), one observed over a period of eight days in August 1987 (Booth, 1989) and a sighting in September 2006   (Fig. 4). One was recorded from Shetland in December 1947 (Venables and Venables, 1955), another in 1983 and a third in March 1987 (Shetland Biological Records Centre) (Fig. 3).
It occurs in scattered coastal localities in southern Norway as far north as c. 61° 56' N 13 (Syvertsen et al., 1995) and also has an uneven distribution in southern Sweden (Ahlén, 2011). Occasional records from the British Isles are considered to be of vagrants and migrants .
There have been 17 records of this species from all study areas except Orkney (Table   1; Figs. 2-3, 5-6). The first was from Whalsay, Shetland in March 1927 (Ritchie, 1927), with a further six records in November of 1981, 1984, August 2003, June 2009and June 2011Racey et al., 2008). An adult male was recorded in June 1965 from a North Sea oil rig (Stansfield, 1966)

The invasion in 2010
The bat invasion in the Faroe Islands in 2010 has no precedence. It began on September 12 th with two bats caught alive but which later died; one was Vespertilio murinus, the second record for the Faroe Islands, and the other Pipistrellus nathusii, the 14 th record for the islands. After the first two bats were discovered, the general public was asked for information on the local radio. During the next seven weeks (until November 2  During the invasion period, two unidentified bats were recorded in Iceland and three in Shetland, which may signify a connection with the phenomenon in the Faroes. None were 14 recorded on North Sea installations at this time and it is unlikely that the two records of P. pipistrellus in Orkney during this period were relevant to the situation in the Faroes.

Seasonal data
The seasonal occurrence of the bats found in all study areas is shown in Figs. 9-10.
The majority of P. nathusii records peak in September but are also high in October, with a moderate number of records through the autumn and winter but few, if any in July and August (Fig. 9). This contrasts markedly with the seasonal occurrence of P. pipistrellus in which high numbers have been recorded from May to September, with a sharp reduction in October and November and no further records for the remainder of the winter. For all other species of bats (Fig. 10), a similar pattern may be seen to that of P. nathusii, with the majority of records occurring in September and October.

Frequency data
Records on North Sea installations peaked in 1990-1999 (Table 2). In contrast, there has been a measurable increase in records for the other study areas, in which records in

DISCUSSION
Monitoring of all wildlife may be considered the duty of any independent nation and the incursion of any non-native species is a matter of interest for a variety of reasons. Bats are of particular interest as they have been recorded as being very rare and irregular vagrants on 15 the north east Atlantic islands and North Sea installations over the course of many years. The first mention of bats in Iceland is a second-hand report in Pennant (1784). Since this reference does not include a Latin name, identification of the species would be impossible were it not for the reference to an earlier work (Pennant, 1781) in which "411 Common [bat]" clearly refers to Vespertilio murinus. Subsequent authors such as Preyer (1862) chose to interpret Pennant's common bat as Pipistrellus pipistrellus, while Krzanowski (1977) suggested that Lasiurus cinereus was a more likely candidate. However while some species may be more likely than others, such speculation about the species involved is impractical, since Pennant was merely reporting another person's observations. Further north and west, bats have also occasionally been observed on Greenland, and are understood to have been blown over from North America according to Møbjerg and Rosing (2001). These authors state that the bats have both an East Greenlandic name "Avangiarlik" (the one which resembles a lemming) and in West Greenland, "Imangertaq" (the one which has been silenced). This indicates that bats have been recorded on both sides of the huge Greenland ice-shield, although individual records with locality and date do not seem to be available.

Movements
The bats recorded from the study areas may be broadly divided into sedentary or migratory species as follows. The tendency to vagrancy and migratory behaviour, which greatly varies between bat species, no doubt has much bearing on whether bats reach the islands without assistance.

Sedentary bats
All three species of Eptesicus, two of the three species of Pipistrellus, Plecotus auritus and Myotis septentrionalis are regarded as largely sedentary species of bats. Both the European species of Eptesicus, E. nilsonii and E. serotinus have small home ranges, moving only short distances between roosts, but with occasional longer distance dispersal flights in Europe (Hutterer et al., 2005;Hutson 2008b). There are, however indications that E. serotinus may be spreading slowly northwards (Dietz et al., 2009). Eptesicus fuscus is a more or less sedentary species according to Hutson (2008a). Plecotus auritus is considered to be a relatively sedentary species (Hutterer et al., 2005;Baagøe, 2007;Entwistle and Swift, 2008). Corbet (1970) commented that a specimen found in the autumn on a lightship about 48 km east of Norfolk, UK may have originated from the continent, as this was coincident with a massive fall of birds from Scandinavia. Although recorded rarely in the study areas and then over a long period of time, there are more records of this species than of the other sedentary species. The New World M. septentrionalis is a short-range migrant reported to travel a distance of up to 56 km between the summer habitat and the winter hibernacula (Caceres and Barclay, 2000). There are very few records from the study areas of any of these sedentary species (Table 1) and only one individual of a sedentary species, E. nilsonii has been recorded from an oil rig. Most of these are believed to be vagrants, whose presence may be attributed to inadvertent transportation by humans.
Most populations of P. pipistrellus in central Europe are sedentary, travelling only short distances between winter and summer roosts, although this species may be more migratory in eastern Europe (Hutterer et al., 2005;Jones and Racey, 2008). The presence of P. pipistrellus on Orkney is a special case, as it is the only species with apparent breeding roosts so far North in Britain. The results show that summer roosts have only been recorded on the largest of the Orkney Islands, namely Mainland and Hoy, with no reports of this species on any of the northern or eastern islands. The locations of summer roosts have changed over time, sometimes with some overlap, when two roosts co-existed for several years. This suggests that there may be sufficient bats for the formation of splinter groups, which may or may not be sustainable depending on conditions. The reasons for these shifts from one location to another may be due to a number of factors, such as weather conditions, adequate food supplies, not only for the spring and summer months for breeding and raising the young but also for building up sufficient fat reserves for the long hibernation period.
While the status of this species on the Orkneys may be accepted as that of a resident, it should also be recognised that the populations may be vulnerable, as they are at the northern boundary of their distributional range.
There is little available information on Pipistrellus pygmaeus (Leach, 1825), because most predates the taxonomic separation from P. pipistrellus, although Dietz et al. (2009) suggested that at least some short-range migration occurs in P. pygmaeus. Details of the distribution of this species remain uncertain but it is widespread in Europe and southern Scandinavia. It occurs throughout the British Isles but is apparently less common than P. pipistrellus in northern mainland Scotland (Barlow and Jones, 1999;Jones and Racey, 2008).
To date this species has not been recorded from any of the study areas but could conceivably be included amongst the records of unidentified Pipistrellus. There is one equivocal reference to a bat detector recording in Orkney of a pipistrelle calling in the 50 -55 kHz range which was attributed to P. pipistrellus  although calls above 52 kHz are in the range of P. pygmaeus. In this study it has been included as an unspecified Pipistrellus.

Migratory species
Nyctalus leisleri is considered to be a long-distance migrant, showing regular seasonal NE to SW movements between the summer and winter habitat in Europe (Hutterer et al., 2005;Sheil et al., 2008). It seems unlikely that migration was a factor in most of the records in the study areas which, apart from two of the records for Shetland in October and December, are mainly in the summer. However strong winds may have affected dispersal in some instances (Corbet, 1970).
In contrast Nyctalus noctula is not considered to be such a long range migrant in all areas; it is migratory in eastern Europe, but probably less so in western Europe (Strelkov, 1969;Hutterer et al., 2005) and populations are not known to migrate from England in the winter . A few individuals have been recorded migrating from southern Sweden in the autumn (Ahlén, 1997) however, while the species is migratory within Sweden, the extent to which it migrates beyond this country is unknown. Information on migratory activity in Scandinavia has been recorded by Baagøe (2007) and Ahlén et al. (2009). Records from the Orkneys, Shetlands and North Sea installations are regarded as vagrants from Europe  and perhaps some may originate from Scandinavia. The single specimen from Iceland was found on board a freighter, arriving from Europe, and was undoubtedly ship-assisted.
Pipistrellus nathusii is known to be a seasonal long-distance migrant in Europe (Strelkov, 1969;Pētersons, 2004;Hutterer et al., 2005). Despite the fact that it is regarded as a rare species in Sweden, Ahlén (1997) reported dispersal and migratory behaviour in this species in late August and early September, with bats accumulating in coastal areas in numbers ranging from less than ten to more than two hundred individuals. Some of these bats were observed to fly out over the open sea in a southern or south-westerly direction and specimens marked in Sweden have been recorded in Germany and Belgium. Of the seven specimens recorded from Iceland, three were found amongst timber imported from Germany and Holland, so were probably ship assisted. However it is not inconceivable that the specimen recorded in May 2003 on the island of Heimaey could have been a storm-borne migrant.
Pipistrellus nathusii was first recorded from the British Isles by Stebbings (1970) and for the next two decades it was considered to occur only as an occasional vagrant in Britain.
With an increase in the number of records in the late 1980s, it was subsequently considered to be a winter migrant (Speakman et al., 1991;Hutson, 1993). More recently the discovery of three summer maternity colonies in England and Northern Ireland (Hutson, 1997;Russ et al., 1998Russ et al., , 2001 means that this species in Britain and Ireland is now considered to consist of both a resident breeding population present throughout the year, and a migratory population present during autumn, winter and spring (Russ, 2008). The increasing number of records in 20 the study areas over the same period mirrors the situation in mainland Britain. Similarly the seasonal peaks for sightings and records of this species are in September and October, coinciding with autumn migratory behaviour so, although in these areas the records are more likely to be of vagrants, it is conceivable that some could be migrants.
Vespertilio murinus is a long-distance migrant normally spending the summer in northern, central and eastern Europe and flying generally in a south-westerly direction to southern Europe to hibernate (Hutterer et al., 2005;Stebbings et al., 2007). In southern Sweden a few individuals were recorded in the autumn leaving land and flying out over the sea (Ahlén, 1997). There has been an increase in records in the British Isles since 1980, predominantly in the autumn and spring suggesting that migrants are sometimes deflected from continental Europe . This is also possible for records from Shetland and North Sea installations, and maybe also for records from the Faroes. It is possible that the specimen recorded from NE Iceland (Fig. 5) in October 2004, could be a storm-assisted migrant.
Lasiurus cinereus is a fast flying, highly migratory species that has been recorded in Orkney and Iceland in autumn and winter (Stebbings, 1986;Stebbings et al., 2007). Three of the specimen records from Iceland were from small, remote locations on the southern coast of Iceland, so ship-assisted passage was considered unlikely (Koopman and Gudmundsson, 1966), while the fourth specimen was from Heimaey Island, Vestmannaeyjar Islands off the southern coast of Iceland. Koopman and Gudmundsson (1966) also mentioned that meteorological records of storm tracks just prior to the October records of 1957 and 1964 indicated winds suitable to carry bats from eastern Canada to south western Iceland. This was further substantiated by a meteorologist (Jakobsson, 1967).
Although Wolley (1849Wolley ( , 1850 was of the opinion that the Orkney record was of a ship-assisted individual, Hill and Yalden (1990) and Yalden (2008) pointed out that at least 21 two American land-birds were also recorded in western Europe in the autumn of 1847.
Furthermore, there were 93 records of American land-birds and over a thousand records of American waders in the British Isles between 1958 and 1972, many of these occurring in October (Hill and Yalden, 1990). It seems highly likely therefore that the Orkney and Iceland records are of bats storm-carried whilst migrating, although, in the case of the Orkney record, at least partial ship-assisted passage may not be excluded.
Myotis lucifugus is a medium-range migrant with movements of more than 300 km to hibernation sites according to Hutson (2008a). The record in August could be of a stray migrant, although Koopman and Gudmundsson (1966) concluded that it was more probable that this immature bat was ship-assisted to Reykjavik harbour.

Frequency of occurrence
The apparent increase in the occurrence of bats, raises the following questions that need to be addressed.

(a) Is there a real increase in records?
The results show that there has been a measurable increase in records for four of the five study areas (Iceland, Faroes, Shetland and Orkney). There has always been an interest in the news of unusual sightings of animals, witness the early records of bats in Shetland in 1774 (Low, 1879), 1815 (Laing, 1815), and 1904 (Tulloch, 1904), in Orkney in 1847 (Wolley, 1849) and 1908 (Spence, 1909), and in Iceland in the 18 th century (Pennant, 1784), circa 1817 (Faber, unpublished) and1936 (Petersen, 1994). There are no similar early records in the Faroese literature and indeed Degerbøl (1940) makes no mention of them.
It appears that bats have become more and more common over the last decades. For example amateur naturalists have lived on Nólsoy in the Faroe Islands for the last 150 years, yet the first bats were not observed on the island until 1987 and had increased to 25 records in the period from 2000 to 2009. The Orkney records are also interesting, in that the numbers of bats reported have increased dramatically from the few in the 1970s and 1980s to the much greater reports in the 1990s and 2000s. This increase may be due in part to greater awareness and involvement of local naturalists and others inspired by a very keen recorder (Booth, 1979(Booth, , 1980(Booth, , 1989(Booth, , 1992(Booth, -2012, but it is also likely to be a true reflection of the situation since the Orkney Field Club was also flourishing in previous years with publications dating from 1968. Researchers at IINH have taken an interest in recording bats in Iceland, since at least 1940, probably from as early as 1889 when the Icelandic Natural History Society and its natural history museum was established. However the records remained steadily at two per decade from the 1960s to the 1990s, with a marked increase from 2000 to 2012. (d) Is there an increase in shipping traffic?
Shipping traffic into Iceland, especially the use of containers to transport all manner of goods has increased considerably over the last few decades. Air traffic has also increased 23 and is an important source of goods transport for all study areas. At least ten of the records from Iceland are believed to have been associated with shipping, and one possibly with air transport. In the Faroe Islands only one bat is presumed to have been imported with timber and there are no other Faroese observations to suggest that bats have arrived with either aircraft or ship.
(e) Is the apparent increase due to changes in weather patterns and the effects of climate change?
Unusual weather patterns have been invoked in the past to account for the occurrence of vagrants, such as that of the specimen of L. cinereus in Orkney, which coincided with that of North American land birds in western Europe. The relationship between bat occurrences and weather patterns was identified in Iceland in the 1960s (Koopman and Gudmundsson, 1966;Jakobsson, 1967). The 2010 invasion of the Faroes also included birds from the USA, Europe and the Far East, and European Lepidoptera. During the same period, a number of European Lepidoptera reached south east Iceland, borne there on a warm south-easterly wind (Erling Ólafsson, personal communication). It is difficult to reach any conclusions concerning the possibility of changes in environmental factors affecting the situation in the Faroes at this time, for although September and October were amongst the seven months that were warmer than normal that year (Cappelen, 2011) weather maps for the Faroes showed no unusual wind conditions at that time (http://www.dmi.dk/faeroeerne/arkiver/vejrarkiv/).
It is important to distinguish the influence of occasional unusual weather patterns, from an overall trend in climate change. The evidence about long term changes in weather patterns linked to global environmental change is still fragmentary but nevertheless strongly indicated, and evidence to demonstrate the ecological response of species to climate change is accruing in many taxa (Thomas and Lennon, 1999;Parmesan et al., 1999;Visser and Holleman, 2001;Walther et al., 2002;Parmesan and Yohe, 2003;Sherwin et al., 2013). As 24 far as the study areas are concerned, Cappelen (2011) reported that the decade from 2001-2010 was the warmest in the Faroes since measurements began in 1890.
The status of Pipistrellus nathusii in the British Isles has changed over more than four decades from vagrant to that of a resident and winter migrant. This may represent a range extension which could conceivably be linked to climate change, or may just be a question of mistaken identity in the past and an increase since the mid-1980s of skilled recorders (Russ et al., 2001). The increase in frequency of recording of P. nathusii in the study areas however, suggests that migrating bats are more often in areas where they are more prone to being swept off-course. In a recent study, this species has been used to demonstrate range expansion linked to climate change in the UK (Lundy et al., 2010). These authors mapped the change in habitat suitability for this species from 1980 to 2000 based on UK Meteorological Office data and modelled the projected change in habitat suitability from 2020 to 2080. The presence of resident P. pipistrellus on Orkney may also be regarded as a range extension, implying that overall conditions have improved sufficiently to support breeding and overwintering bats.
Other recent studies that examined the predicted impact of climate change on bats include Rebelo et al. (2010), Prydatko et al. (2011) and Sherwin et al. (2013). Rebelo et al. (2010) modelled the potential distribution of bat diversity for three biogeographic groups: temperate, boreal and Mediterranean. Four of the ten species in their temperate group, included species reported in the current study: P. nathusii, N. leisleri, E. serotinus and P. auritus, while N. noctula was one of four species in their boreal group.
The measurable increased frequency of the occurrence of bats in Iceland, the Faroes, Shetland and Orkney suggests that all the above factors are important but it is difficult to distinguish between them. The importance of different factors appears to be variable depending on the study area. Overall the European migratory species are probably more likely to reach Shetland and Orkney and North Sea installations. The North American bat 25 species seem more prone to appear in Iceland than in the other study areas. However although further away from sources of bats, whether in Europe or North America, migratory bats may well reach the Faroe Islands and Iceland as do vagrant birds (Pétursson and Skarphéðinsson, 1978) and insects (Kaaber et al., 1994;Jensen, 2001).