Small terrestrial mammals of Albania: distribution and diversity (Mammalia, Eulipotyphla, Rodentia)

Abstract In this paper new records are reported for 23 species of small terrestrial mammals (STM) of Albania collected during the field work campaigns organised in the framework of the project “Strengthening capacity in National Nature Protection – preparation for Natura 2000 network” (NaturAL) in Albania during the summer and autumn of 2016 and 2017 Data on small mammals were primarily collected through Sherman live-trapping campaigns in six high priority protected areas of Albania: Korab-Koritnik, Bredhi i Hotovës, Tomorri, Llogara-Karaburun, Divjakë-Karavasta, Liqeni i Shkodrës (Skadar lake), Lëpushë-Vermosh. Other data were obtained by analysis of owl pellets or by direct observation of individuals (dead or alive) in the field. For 21 species Erinaceus roumanicus, Neomys anomalus, Crocidura suaveolens, Crocidura leucodon, Suncus etruscus, Talpa stankovici/caeca, Myocastor coypus, Sciurus vulgaris, Glis glis, Dryomys nitedula, Muscardinus avellanarius, Microtus levis/arvalis, Microtus subterraneus, Microtus thomasi, Microtus felteni, Myodes glareolus, Apodemus sylvaticus, Apodemus flavicollis, Apodemus epimelas, Mus musculus, and Mus macedonicus additional records are provide and their distributions reviewed, while the presence of two new species of shrews (Sorex araneus and Sorex minutus) for Albania is reported for the first time. A comprehensive review of the published and unpublished distribution records of STM species of the country is made, together with an updated checklist and distribution maps of the species.


Introduction
Although the first records of small terrestrial mammals (STM) from Albania date back to Miller (1912), subsequent studies were published only after 1945 and were largely based on very limited records (Puzanov et al. 1955, Rosický and Gjini 1960, Koçi 1961, Bajrami and Serezi 1981, Vangjeli 1987, Anděra 1991, Kryštufek 1994, Prigioni 1996, Macholán and Vohralík 1997, Macholán et al. 2003. More comprehensive studies, published in Albanian language (Bego 1997(Bego , 2001(Bego , 2003, remained largely unnoticed by mammalogists outside the country. Consequently, by the end of the 20 th century, the mammal fauna of Albania was the least known in Europe (Prigioni 1996, in contrast with other terrestrial vertebrates (amphibians, reptiles and birds) that were better known and recently evaluated (Bego and Koni 1999, Zeneli et al. 2014. In 2008, Bego et al. reported a synopsis of the small mammals of Albania (Eulipotyphla, Rodentia), outlining both new records and previously published data. Twentyfour species (seven eulipotyphla and 17 rodents) have been reported for the country. Nine species (Neomys anomalus, Crocidura leucodon, Talpa stankovici, Dryomys nitedula, Muscardinus avellanarius, Micromys minutus, Mus macedonicus, Myodes glareolus, and Microtus thomasi) were recorded in Albania for the first time. Furthermore, a probable presence of a further eleven species was anticipated. Few years later, Paspali et al. (2013a, b) provided supplementary data on small mammals from the southern region of Albania. In 2014, Bego et al. reported the first record of Spalax (Nannospalax) leucodon in Albania. Ciechanowski and Sachanowicz (2014) contributed further knowledge on the distribution and habitat usage of the fat dormouse (Glis glis), by providing data on 14 new localities distributed mostly in the northern half of Albania. The number of known small mammal species for Albania has increased recently, with three new species of Cricetids (Dinaromys bogdanovi, Chionomys nivalis and Microtus subterraneus) recorded by  and .
The aim of this paper is to provide the most recent records for 23 small mammal species found in protected areas visited within the framework of the "Strengthening capacity in National Nature Protection -preparation for Natura 2000 network" (NaturAL) project, with focus on the new records for the two new shrew species for Albania, and ultimately, to summarise the current distribution of species belonging to the orders Eulipotyphla and Rodentia in the country. Simultaneously, this paper represents the most updated checklist and distribution of the small terrestrial mammals of Albania.

Materials and methods
Data on small mammals were primarily collected through Sherman live-trapping campaigns organised in summer and autumn 2016 and 2017 in various high priority protected areas of Albania selected by the NaturAL project: Korab-Koritnik, Bredhi i Hotovës, Mali i Tomorrit, Divjakë-Karavasta, Shkodra lake (Skadar lake), Llogara-Karaburun and Lëpushë-Vermosh. A set of 100 Sherman live traps has been set for 3-4 nights in each of the surveyed protected areas. A mixture of peanut butter and cereals was used as bait. Sometimes, pieces of apple were added inside the trap as additional bait. Captured individuals were handled in plastic bags. After being measured (weight), species and sex determined, including reproduction activity (sexually active or not active, pregnant, lactation etc.), the captured individuals were released. Only animals found dead (very few) inside the traps and few individuals of shrews species recorded for the first time in Albania were taken for collection. Other data were obtained by analysis of owl pellets (20 STM specimens), mostly Tyto alba pellets, collected in three sites (two monasteries in Divjakë-Karavasta area, and one in Llogara-Karaburun, Cave of Duk Gjonit), supplemented by direct and indirect (footprints, feeding signs, faeces, burrows, carrions, etc.) observations (28 records) in all surveyed protected areas (Fig. 1). Specimens retrieved from owl pellets were determined at species level, with the help of a Stereomicroscope. The voucher specimens of few individuals found dead inside the traps, including the two new red-toothed shrew species, were morphometrically measured [head and body length (HB), tail (T), hind foot (HF), weight (W)], and sex and sexual activity were determined. Cranial and dentition examinations were conducted for better identification of both red-toothed shrew species. Examined specimens were deposited at the Department of Biology, Tirana University.
The taxonomic identification of our material was based on morphology (Wilson and Reeder 2005). Although such an approach provides reliable results for the great majority of the taxa, in some cases (e.g., the genus Talpa and Microtus arvalis-levis group) our results should be confirmed through karyological and DNA studies. Therefore, it is likely that in some areas, such as those in the northern part of Albania, the number of reported species may be underestimated.
The current knowledge on small mammal species distribution in Albania, based on published data (Bego et al. 2008, Rogozi et al. 2012, Paspali et al. 2013a, b, Bego et al. 2014, Ciechanowski and Sachanowic 2014 and new records provided in this paper is presented in the Supplementary Information. The distribution maps of the STM species on 10x10 km grid cells were built based on distribution data of published sources and new records collected by this study (Fig. 2). All previous and recent records on STMs are included in the BioNNA database (http://www.bionna.al/) recently established in Albania (Pacifici et al. 2018). Based on previous and recent records a comprehensive checklist of small mammals of Albania was produced as Supplementary Information of this paper. Chorotype and global distribution range of the species are according to IUCN (2017). Distribution records of the species are in WGS84, decimal degrees. ArcGis was used for producing the altitudinal distribution of the records and the distribution maps of the species. Coordinates have been converted to UTM projection to produce the maps. R was used to derive the altitudinal distribution of the species.

Study area
Albania is located on the eastern coasts of the Adriatic and Ionian seas. In spite of its small surface area (28,748 km 2 ), the country is topographically highly diverse. The Albanian mountains (the highest peak is Mt. Korab, 2753 m a.s.l.), which are physiographically part of Dinarides and Hellenides mountain ranges, rise very steeply from the coastal strip and cover 70 % of the country. As a consequence of Alpine orogenesis, these mountains are characterised by complex folding and faulting, although outcrops of more ancient volcanic and metamorphic rocks are common throughout. The northern mountains, in particular, are singled out by the dominance of carbonate rocks. Thick layers of Late Tertiary or Quaternary sediment cover the 20-30 km wide strip of the coastal lowlands.
Although Albania falls within the temperate zone, topographic heterogeneity and the combination of Mediterranean and continental influences ensure great climatic and vegetation diversity ). The coastal region has a Mediterranean climate with dry, hot summers and mild, wet winters. Further east the mountains act as a barrier to the penetration of warm air masses. The climate of the mountainous regions is Alpine, with strong shifts in precipitation and temperature depending on height and exposure. Differences in temperature between the coast and inland regions are by far most marked in winter, whilst that in precipitation are stronger in summer. Predominant vegetation along the coast is evergreen maquis, which is replaced by oak woodland further inland and finally by beech forests and grasslands at higher altitudes. Forests, which are frequently degraded, cover c. 40 % of the whole surface, while meadows/pastures and arable land cover, respectively, 15 % and 26 % of Albania. Mean human density is 115 inhabitants per km 2 , but the mountainous regions are scarcely populated (Kabo 1990, Bego and Koni 1999, Zeneli et al. 2014. The main protected areas of Albania are national parks, nature parks, and strict nature reserves. There are 799 protected areas in Albania covering a surface of 460.060.9 ha, or 16 % of the whole national territory. These include two Nature Reserves, 14 National parks, one Marine park, 750 Natural monuments, 22 Habitat/Species Management Areas, nine Protected landscapes, four managed resources areas, one biosphere reserve, three World Heritage sites, and four Ramsar wetlands. The most notable protected areas in Albania are the national parks, having a total area of 210,501.4 ha, or roughly 6.9% of the territory (Zeneli et al. 2014).
Recently, the NaturAL project has initiated the process of establishing the Natura 2000 network and strengthening the national capacity in biodiversity conservation. One of its aims is to improve the knowledge of and experience in biodiversity monitoring in protected areas, by investing in equipment and training the protected areas staff on methods and techniques for studying habitats, and plant and animal species. The NaturAL project is focused on five primary (Liqeni i Shkodrës/Skadar lake, Korab-Koritnik, Divjakë-Karavasta, Tomorr, Bredhi i Hotovës) and 5 secondary (Bunë-Velipojë, Dajti, Shebenik-Jabllanicë, Nartë-Vjosë, Llogara-Karaburun) protected areas, therefore the field campaigns (live trapping, searching for owl pellets and observations of live or dead animals and other species presence signs) were conducted in six priority protected areas and one proposed protected area (Lëpushë-Vermosh), while field observations on small terrestrial mammals were extended to Butrint national park and Pashtrik proposed protected area, situated north of Korab-Koritnik (Fig. 1).  Main habitats inside some of the visited protected areas are presented in the Figure 3. Pictures in the Fig. 3 show main sites where live-trapping campaigns were conducted, as well as one of the sites where owl pellets were collected.

Results
A total of at least 23 small mammal species have been recorded in all surveyed protected areas ( Table 1). The highest number of sampled species was reported in Korab-Koritnik (11-14 species), followed by Divjakë-Karavasta (9 species), Lëpushë-Vermosh (7-9 species), Bredhi i Hotovës (7 species), and Llogara-Karaburun (7 species). In Butrint park area no small mammal trapping campaign was organised, but this is the only site where the coypu (Myocastor coypus) is established in Albania.
The uncertain number of species in , Lëpushë-Vermosh (7-9), and Pashtrik (2-4) is explained by the fact that in this part of Albania we expect a possible sympatry of Talpa caeca, T. stankovici, and T. europaea, and a sympatry of Microtus arvalis and M. levis (mystacinus).
A full list of the known STM species of Albania (31 species) with their number of records and global distribution range is presented in Table 2. Literature records and new (unpublished) records are presented separately, indicating both the number of localities of species being recorded and number of specimens captured, observed, or retrieved from the owl pellets.
Our study presents the records for two red-toothed shrew species that are reported for the first time in Albania (Sorex araneus and S. minutus), and new records for the coypu (Myocastor coypus), an introduced species that is already established in Butrint area, southern Albania. Seven STM species were not recorded during our field investigations (Ch. nivalis, D. bogdanovi, S. leucodon, R. rattus, R. norvegicus, M. minutus, and M. spicilegus) due to the location of priority protected areas, location and altitude of the sites where live-trapping took place (maximum altitude 1450 m a.s.l.), as well as the limited sampling efforts (in total 33 localities, of which 13 trapping sites, 3 sampling sites for owl pellets and 17 observations for signs of presence of STMs (Table 2).
Although most of the STM species occurring in Albania have a wide global distribution range, being either Palaearctic or Cosmopolite, six species endemic to Balkans (Talpa stankovici, Apodemus epimelas, Microtus felteni, Microtus thomasi, Dinaromys bogdanovi, and Spalax leucodon) and two other species endemic to Europe (Talpa caeca and Mus spicilegus) have an important part of their global distribution range inside the territory of Albania. However, for most of these endemic species we still have limited data on their distribution and conservation status, except for Thomas' vole (Microtus thomasi), that is found in high density in main lowland agriculture fields and considered as pest species (Bego et al. 2008, Paspali et al. 2013b. By integrating literature occurrence records with those obtained by this study an altitudinal distribution of STM species and frequency of occurrence records by altitude in Albania was produced and shown in the Figure 4. The density plot indicates that most of the records have been found between 0-200 m a.s.l. and 1200-1400 m a.s.l., that both correspond to altitudes of more frequent sampling efforts (respectively, records retrieved from owl pellets in lower altitudes and records from trapping in higher altitudes). For more information see Suppl. material 2. The annotated list of 23 STM species reported in this study with synopsis of distributional records (previous and new records) and notes on habitats of species occurrence are described in the following part of the paper.   Four new records are added to the previous ones reported by Bego et al. (2008). All recent records consist of dead animals found on the roads. These records are shown on the map (Fig. Suppl. material 3: S-01). All records (previous and new ones) are spread over the entire Albania, from the sea level to the upper forest line (1500 m). In agreement with hedgehog preferences for lowlands and hills between 300 and 800 m a.s.l.  in Europe, all our recent records were below 800 m a.s.l (see Fig. 4). As a result of the expanding transport network, the number of road casualties for this species is increasing. Earlier papers refer to this species as Erinaceus europaeus (Puzanov andMitrushi 1955, Koçi 1961) or E. concolor (Bego 1997).
The common shrew is recorded in Albania for the first time, although its presence was anticipated by Bego et al. (2008). The habitat where most of the specimens were captured represents a mixed broadleaved and fir forest bordering wet meadows (see Fig. 5). The common shrew has a wide distribution in the Palaearctic, and is recorded from sea level to 2,500 m (Anděra 1999), but its distribution in Albania is expected to be limited in forest and pastures of high mountains of northern and north-eastern parts of the country, under the influence of continental climate (Fig. 4).
The species was found in sympatry with the common shrew, but in smaller numbers than the latter (Suppl. material 3: S03). It was trapped along the woodland edges confined with swamps and wet meadows in mountain mixed coniferous and broadleaved forests, typical habitats for the species (Hutterer 1990(Hutterer , 1999. Two dead individuals inside the trap were taken for collection. Likewise the common shrew, the presence of pigmy shrew was anticipated by Bego et al. (2008), and this species is added for the first time in the list of the known Albanian terrestrial mammals (see Figure 5).

Neomys anomalus (Cabrera, 1907)
Previous records. Gjirokastër: Sofratikë, Libohovë, Lazarat, Arshi Lengo, Serat e Mashkullorës, Ura e Kardhiqit: in total 100 specimens from the pellets during 2006-2011 (Bego et al. 2008, Paspali et al. 2013a Dibër: one dead individual along the Drini i Zi river, close to the bridge in proximity of Maqellarë (18 April 2013) (Paspali et al. 2013b)  The Mediterranean water shrew was reported for the first time in Albania in 2008, with records from the Drinos valley (Bego et al. 2008). Since then it has been regularly found in the owl pellets collected in several localities along the Drinos valley in the southern part of the country (Paspali et al. 2013a, b). On 18 April 2013, Bego found a dead specimen along the Drini i Zi river, close to the bridge in proximity of Maqellarë (Dibër) (Suppl. material 3: S04). The recent record from Lëpusha stream dated 29 July 2017 confirms that this species in the Balkans is more widespread than N. fodiens (e.g. Petrov 1992); consequently, its distribution in Albania is expected to be wide, regardless of the small number of records to date (Fig. 4).  (Bego et al. 2008, Paspali et al. 2013a.
The bicoloured shrew was reported for the first time in Albania in 2008 (Bego et al. 2008). Localities in Albania range from sea level up to 1100 m (Fig. 4). Although the records from the northern part of Albania are missing (Suppl. material 3: S05), it is believed that the bicoloured shrew is widespread in Albania. According to the literature, the bicoloured shrew is common around Skadar Lake in Montenegro (Petrov 1992). The lesser white-toothed shrew is the most widespread shrew in Albania and was found from the sea level up to 1200 m (Fig. 4). Interactions between the two Crocidura species are poorly understood. C. suaveolens was found on a larger number of localities than C. leucodon, however for both species, ca. 50% of records are from lowlands (< 400 m a.s.l.) (Suppl. material 3: S06)
All the previous and recent records on Etruscan shrews (n = 425) were from T. alba and other owls' pellets (Table 2 and Suppl. material 2). Albanian records are all from coastal lowlands, below 400 m a.s.l ( Fig. 4 and Suppl. material 3: S07). Given its occurrence in the coast of Montenegro (Petrov 1992), it is doubtlessly present also in northern Albania. In the Balkans this tiny shrew is similarly restricted to the coastal regions below 600 m a.s.l. (Vohralík and Sofianidou 2000) and is common in owl pellets from the north-eastern Adriatic coast (e.g. Lipej and Kryštufek 1991).

Talpa caeca Savi, 1822
Previous records. Tropojë: Tropoja (=Tropojë; Kryštufek, 1994) Three species of moles (T. europaea, T. caeca, and T. stankovici) are partly sympatric in the countries surrounding Albania . Although they are well defined by their diploid chromosomal numbers (Soldatović and Dunđerski 1972), cranial differences are slight and vary geographically. Materials from Albania were earlier reported as T. caeca (Bego 1997(Bego , 2001, but their re-examination revealed also the presence of T. stankovici (Kryštufek 1994). The Blind mole is known in Albania from a small number of localities at high altitude (1100-2000 m) (Fig. 4). Characteristic mole hills were also observed in other parts of the country. Particularly, those from the northern alpine pastures (Albanian Alps) very likely belonged to T. caeca. However, the presence of T. europaea in the Albanian Alps is not excluded (Suppl. material 3: S08).
Records from the coastal area of central Albania are the first evidence connecting the contiguous range of T. stankovici in Macedonia and Greece to an isolated occurrence from Ulcinj, Montenegro (Kryštufek 1994). Given its wide altitudinal occurrence in Albania (5-1550 m, Fig. 4) and Macedonia (up to 2200 m;Petrov, 1992), T. stankovici is putatively widespread to the south of the River Drin (Bego et al. 2008) (see Suppl. material 3: S09). The Eurasian red squirrel is widespread in forested regions throughout Albania, from the sea level to the tree line (1600-1800 m, Fig. 4). Most of our records are based on observations of animals and their characteristic signs (footprints, feeding signs) or vocalisations (Suppl. material 3: S10). Dark colouration was dominant. In 1960-1970 this animal was reported as a game species (Puzanov andMitrushi 1955, Koçi 1961).  (Table 2). Both previous and new records confirm that the fat dormouse is widespread in the forested regions of Albania (cf. Mitchell-Jones et al. 1999), including shrubland or almost barren landscapes with sparse bushes and vertical rocky walls or very steep slopes, man-made habitats (bunkers, tunnels, old mines) or natural (karstic caves) (Fig. 4). Glis glis uses a variety of natural and manmade habitats, abandoned military structures, that help in maintaining populations of the species under the heavy ongoing deforestation of Albania (Suppl. material 3: S11). In the 1960s it was considered as a game species (Puzanov and Mitrushi 1955). Previous and new records of this species are linked with old growing beech forest (Stravaj, Tomorr) and oak woodland (Goricë e Vogël). Albeit rare, the forest dormouse is widespread in the Balkans (Kryštufek and Vohralik 1994); consequently it is likely to be more widespread in the broadleaved forests of Albania (Fig. 4).  (Bego et al. 2008, Paspali et al. 2013a.
Although the common dormouse had already been collected in June 1914 (NMW, specimen from Vermosh, 1150 m), its occurrence in Albania had not been published until 2008 (Bego et al. 2008). Most of the previous and recent records of M. avellanarius were retrieved from Tyto alba pellets (60 specimens) (Table 2, Suppl. material 2). So far, only two specimens, one in mixed woodland forest on the hills near Vorë (Tiranë), and one in a mixed forest near Dardhë were collected through snap trapping (Bego et al. 2008). Muscardinus avellanarius seems to be common in Albania and its altitudinal range extends from the sea level up to 1494 m (Fig. 4), whilst in neighbouring countries (Montenegro, Macedonia), the common dormouse has been found only in mountain forests (Kryštufek and Petkovski 2003). Therefore, we suggest that the common dormouse is probably widespread in Albania (Suppl. material 3: S13).  (Bego et al. 2008, Paspali et al. 2013a. Tomorri mountain: 23 Sept 2015 (40.621°N, 20.177°E, 2045 m a.s.l.); one individual, female, sexually active captured through live trapping  In the Balkans, the taxonomy of field mice from the Sylvaemus group (sensu Musser et al. 1996) has not yet been completely agreed upon. A subspecies of this group, Apodemus sylvaticus stankovici, described for the Macedonian side of Mt. Korab has been considered to represent a form of A. flavicollis (Kryštufek and Stojanovski 1996) or a true species, A. stankovici (Petrov 1994). More recently, Bugarski-Stanojević et al. (2008) questioned the presence of A. sylvaticus in the Balkans, although the identity of specimens from Greece was ascertained by mitochondrial DNA cytochrome b sequences (Michaux and Filippucci 2005). Our identification is based on morphological characteristics that allow the distinction of A. sylvaticus and A. flavicollis in the western Balkans (Kryštufek and Stojanovski 1996). Apodemus sylvaticus is widespread in Albania from the sea level up to 2045 m a.s.l   (Fig. 4 and Suppl. material 3: S14). In many parts of its distribution range in Albania it is found in sympatry with A. flavicollis and A. epimelas (Bego et al. 2008).
Most western broad-toothed field mice were found in degraded woodland on rocky grounds. This is surprising because open rocky areas are its main habitat further north, in former Yugoslavia (Petrov 1992). Records are quite scarce for Albania; nevertheless we suggest this mouse to be widespread on karstic substrates. In Albania this mouse is present between 50 and 1100 m a.s.l. (Fig. 4), but it has been recorded up to 1600 m a.s.l. in neighbouring regions (Petrov 1992). Mus musculus has already been reported in Albania by Vangjeli (1987), Bego (1997), and Bego et al. (2008). Based on previous and recent records (Suppl. material 3: S17), M. musculus is probably widespread in rural and sub-rural areas of Albania, as a sinanthropic species, whilst free-living populations mostly belong to M. macedonicus (Macholán et al. 2003). Accordingly, in extensive owl-pellet material from the coastal sites of Xeng and Karavasta, collected between 2000 and 2008, we identified only 10 specimens of M. musculus (Bego et al. 2008). New record from Lëpusha (July, 2017) raises the distribution range for the commensal house mouse in Albania up to ca. 1400 m a.s.l. (see Fig. 4).  (Bego et al. 2008, Paspali et al. 2013a.
Mus macedonicus is the most common out-door Mus species in Albania (Table  2). Most our records are from the coastal lowlands of central and southern Albania (Fig. 4). In the districts of Lushnjë and Fier M. macedonicus is evidently sympatric with the morphologically and ecologically similar M. spicilegus. Based on owl pellet content the former is by far more abundant (483 vs. 88 specimens in the district of Lushnjë) (Bego et al. 2008). New record from Novosejë (Korab-Koritnik Park) in July, 2016 raises the upper limit of distribution range for the Macedonian mouse in Albania up to ca. 1400 m a.s.l. (see Suppl. material 3: S18).
Considering the ranges of the two sibling Microtus species in eastern Macedonia (Petrov 1992) and adjacent Greece (Sofianidou and Vohralík 1991), the materials collected in the southern Albania belong to M. levis, while those of northern Albania rather belong to M. arvalis. Recent chromosomal diversity in the genus Microtus at its southern distributional margin in Iran (Mahmoudi et al. 2014) suggested a new name for M. levis as Microtus mystacinus, and a possible sympatry of the latter with Microtus arvalis in the northern Albania (Suppl. material 3: S24). Karyotype evidence, however, is required to ascertain unambiguously their taxonomic identity. Vangjeli (1987) was the first to report M. levis (as M. epiroticus) for Albania, specifically for the Korça agricultural area.

New records. Korab-Koritnik:
The Balkan pine vole is a rare and little known Balkan endemic. It has been observed in only 13 localities (Kryštufek and Petkovski 2003), but its potential range might cover an area of ca. 40,000 km 2 in Albania, Serbia, Kosovo, Macedonia, and Greece (Shenbrot and Krassnov 2005). The first record of the species for Albania was reported by Anděra (1991). Our collection related to this species includes at least 31 specimens found in pellets, a pregnant female (with six embryos) captured in a small clearing inside mixed deciduous woodland in Pogradeci district (Bego et al. 2008), and recent record in Novosejë (27 July 2016), a male, sexually active, trapped again in a small clearing inside a mixed deciduous and coniferous woodland.
In Albania, Thomas' pine vole is the most common vole in the diet of Tyto alba and our collection includes 1835 specimens, of which 1760 retrieved from owl pellets and 75 specimens from live trapping. Although all our records are from the coastal lowlands, we assume that Thomas' pine vole is probably more widespread at low and medium altitudes, between 0-600 m (cf. Shenbrot and Krassnov 2005). In some parts of Myzeqe Field it is causing damages to perennial crops and is considered as a pest. Microtus subterraneus occurs primarily in Europe, where it occupies central regions from the Atlantic coast of France to European Russia, and the Balkan peninsula (Shenbrot and Krasnov 2005). There are records of this species from sea level to 2,300 m . Microtus subterraneus is known in Kosovo, Macedonia, and Montenegro (Kryštufek and Petkovski 2003). This species was reported for the first time in Albania by  in the Tomorri Mountain, where an individual of M. subterraneus was trapped in a grassland situated at the altitude of 2045 m a.s.l. on 23 Sept 2015. The new records of this species from Novoseja-Shishtaveci area, part of the Korab-Koritnik park, indicates that this vole species is probably more widespread in northern and eastern Albania.
This species is native to South America, and has been introduced to North America, Europe, Africa, and Asia. It is patchily distributed throughout its historic range in association with mesic habitats and water bodies. It is increasing in number, and is considered as a pest in parts of its range and has been the subject of eradication measures (Bertolino et al. 2005). This species has been introduced and is present in the surrounding countries (Montenegro, Kosovo, Macedonia, and Greece; Vohralík 1991, Kryštufek andPetkovski 2003). In Albania it has been introduced in the early 70's of the last century in the Patoku hunting reserve, but since then there has not been any evidence of its presence in the area. In recent years, the species has been reported in Butrint National Park and Ramsar site, where the species is well established (Suppl. material 3: S31).

Discussion
The high species diversity and endemism for mammals in the Balkans is indicated for the mountainous regions of the southern Dinarides and the Shara-Pindos Mountains, which encompass Albania Griffiths 2002, Kryštufek 2004). Albania shares the distribution range of six Balkan endemic STM species (Talpa stankovici, Apodemus epimelas, Microtus felteni, M. thomasi, Dinaromys bogdanovi and Spalax leucodon); therefore, Albania is expected to play an important role in the conservation of these endemics (Bego et al. 2008, Kryštufek et al. 2009). A B

D C
The new and previous records reported and analysed in this paper confirmed the species diversity of this part of the Balkans: 31 STM species are already reported in Albania, but the presence of another five species is anticipated (Sorex alpinus, Neomys fodiens, Talpa europaea, Microtus arvalis, Arvicola amphibius), due to the existence of these species in the countries surrounding Albania (Petrov 1992, Kryštufek and Vohralík 1992, Kryštufek and Petkovski 2003, Niethammer 1986, Vohralík and Sofianidou 1987, Sofianidou and Vohralík 1991. A notable progress in studying STM species of Albania was made over the recent years (see Table 2 and Suppl. material 2). Thus, Bego et al. (2008) (Table 2 and Suppl. material 2). Paspali et al. (2013a, b) (Fig. 4).
The list of shrews in Albania is amended with two red-toothed shrew species (Sorex araneus and S. minutus) whose first records of their presence in the country come from Novosejë-Shishtavec area, Kukës district, part of Korab-Koritnik Nature Park. Considering habitat requirements of these species and their distribution records in the neighbouring countries, we believe that both species are likely present in other parts of northern Albania (Anděra1999, Hutterer 1990, 1999, Petrov 1992, Kryštufek and Vohralík 1992, Kryštufek and Petkovski 2003. In this paper we also report additional records of M. subterraneus and M. felteni in north-eastern Albania (Novosejë-Shishtavec, Kukës district, as part of Korab-Koritnik Nature Park).
All four recently reported rodent species (D. bogdanovi, M. subterraneus, C. nivalis, and S. leucodon) were recorded in the mountainous habitats of northern, north-eastern and south-eastern Albania. In spite of the very low number of specimens captured to date, their distribution range in Albania is expected to be wider (Ondrias 1966, Kryštufek and Bužan 2008, Petrov 1992, Shenbrot and Krasnov 2005, Vohralík and Sofianidou 1987, Sofianidou and Vohralík 1991. Therefore, future field investigatioon should focus on these regions, so that to verify and confirm their distribution and conservation status in Albania.
Our sampling campaigns and field investigation in priority Protected Areas of Albania during 2016 and 2017 enriched our knowledge for several other small terrestrial mammals that are still poorly known, such as Suncus etruscus, Neomys anomalus, Dryomys nitedula, Muscardinus avellanarius, Apodemus epimelas, Microtus subterraneus, Microtus felteni, and Myodes glareolus. In spite of the progress, the knowledge on distribution range of dormice, especially for D. nitedula and M. avellanarius, remains poor in Albania, due to lack of proper methods and equipment to study them; the records for these two dormouse species come from live-trapping and owl pellets, while dormouse nest boxes, the most effective method, has not been practiced yet. However, the use of the recently purchased dormouse nest boxes by the trained local staff of Protected Areas under the NaturAL Project will bring more data on the dormice species in Albania in a near future.
The presence of the coypu, Myocastor coypus, an introduced rodent species to Europe but native to South America (Bertolino et al. 2005, IUCN 2017), is for the first time confirmed in the southern Albania, both in the Butrint ancient city and Bufi lake, part of Butrint National park area. The species has been repeatedly reported over the last four years (2013-2017) by rangers of Butrint National Park, including the evidence of reproduction, and this is a good indication that M. coypus is well established in the Butrint Park area. The possible interactions and impacts of this introduced species on native fauna of the Butrint area should be studied and monitored.
The sampling efforts done over recent years in Albania have not yet produced the evidence of presence for the water vole (Arvicola amphibius), alpine shrew (Sorex alpinus), and water shrew (Neomys fodiens) in the country. The water vole is recorded around two lakes shared between Albania and Montenegro (Lake Shkodra) and Macedonia (Lake Ohrid; Petrov 1992), but it was not found in the Albanian part of Shkodra (Skadar) Lake. Habitat degradation and high seasonal fluctuations of water level in Shkodra Lake might be the reasons for the absence of the species in the Albanian part of the lake. In any case, T. europaea and M. arvalis are presumably present in the mountains north of the River Drin (Yigit et al. 2016). However, further sampling efforts should be made to confirm this assumption.
The taxonomic identification of our material was based on morphology. Although such an approach provides reliable results for the great majority of the taxa listed above, in some cases (e.g., the genus Talpa and Microtus arvalis-levis group) our results need to be strengthened by karyological and DNA studies. Chromosomal variability is perhaps not indicative of cryptic speciation, however it reveals the phylogeographic structuring of a small-range endemism (Zima J. 2004, Mahmoudi et al. 2014). The Balkan Peninsula was one of the major refugia from European glaciation over the last two million years of climatic oscillations. Unsurprisingly, given the high topographic diversity of the region, phylogeographic studies based on molecular markers have revealed many microrefugia within the major refugium (e.g. Kryštufek et al. 2007). The role played by Albania as a refugium for small terrestrial mammals is still mostly unknown and should be further studied.
Finally, it must be underlined that there are major conservation issues associated with Albanian small terrestrial mammals. The number of extinctions strongly correlates with the number of endemisms, and the species which display both restricted ranges and low densities suffer the highest risk of extinction (Nott and Pimm 1997). All mammalian species endemic to the Balkans should receive the greatest attention with respect to conservation and research. Natura 2000 as an important instrument of nature conservation for EU member states can offer little support to mammalian endemic species of the Balkans. The inclusion of sites in the N2000 network is related to the fact that occurring species are included in the EU Habitats Directive, but Balkan endemic species often very important for conservation are still not present in the HD lists simply because they live in non-EU countries. Throughout the Balkans, and Albania in particular, the reality of conservation issues sharply contrasts with species requirements. The recent decline of the Mt. Galičica population of D. bogdanovi, a western Balkans endemism, is representative of a wider situation (Kryštufek and Bužan 2008). The low numbers of specimens for endemic and rare species such as D. bogdanovi, M. felteni, C. nivalis, and absence of other species such as Sorex alpinus, Neomys fodiens, A. amphibius, might be a strong signal of worsening conditions for these species due to anthropogenic stressors, including habitat degradation and destruction and climate change.

Conclusion
Knowledge on small mammals of Albania has been significantly improved over the last two decades, due to contributions from local and foreign researchers. The number of known species has increased (at present 31 species, of which nine species of Eulipotyphla and 22 species of Rodents), almost reaching the potential number of STM species in the country (36 species) according to literature review of species distribution in the countries around Albania (Bego et al. 2008). However, in spite of the progress made, the presence and distribution range of five other species (Talpa europaea, Sorex alpinus, Neomys fodiens, Arvicola amphibius, Microtus arvalis) in Albania remain to be verified in the coming years. Sampling efforts should be focused on the central and northern parts of Albania, where most of these species co-occur. Subalpine and alpine areas of these parts of Albania should be sampled more intensively, with the use of the proper sampling techniques and supplemented with karyological and DNA analysis (Zima 2004, Mahmoudi et al. 2014, Storalik et al. 2017, Storalik and Jablonski 2017 to better distinguish potential cryptic taxa.