The indigenous vascular flora of the forest domain of Anela (Sardinia, Italy).

The importance of mountains for plant diversity and richness is underestimated, particularly when transition zones between different bioclimates are present along altitudinal gradients. Here we present the first floristic data for a mountain area in the island of Sardinia (Italy), which exhibits Mediterranean bioclimates at the bottom and temperate bioclimate at the top. We discovered a very high floristic richness, despite the fact that the number of endemic taxa is not high and the number of exclusive taxa is very low. Many of the detected taxa are at their range periphery and/or ecological margin. We conclude that climate transition zones in Mediterranean mountains and especially on islands are key areas regarding plant biodiversity and should be better investigated and protected.


Introduction
Mountains are a critical landscape and ecosystem; they not only provide water for the lowlands but are a source of well-being and inspiration for numerous people (Korner 2004). The green 'coat' of the world's mountains is composed of specialised biota, all nested in a great variety of microhabitats. Mountains biota are determined by a series of climatically different life zones over short elevational distances (Rahbek 1995, Korner 2000, Hemp 2002, Korner and Paulsen 2004, which often result in areas of high biodiversity of high conservation interest (Korner 2004). However, those areas are also under high threat regarding climate change, as it is expected that they experience drastic changes (Inouye 2008).
Mountain biodiversity can be studied at a multitude of scales in space, time and function (Molau 2004). Even though species richness is usually the focal component in nature conservation, genetic diversity within species is equally important. The smallscale distribution of species in the tropical Andes, as exemplified by the plant genera Calceolaria (Calceolariaceae) and Bartsia (Orobanchaceae), contrasts against the situation in high-latitude mountains, e.g. the Scandes, where species have wide ranges and many are circumpolar (Molau 2004). Several studies on alpine plants, based on molecular data, show that the intraspecific genetic diversity tends to increase with latitude, a situation brought about by glaciation cycles permitting repeated contractionexpansion episodes of species' distributions (Abbott et al. 2000, Abbott and Brochmann 2003, Gamache et al. 2003, Holderegger and Abbott 2003, Abbott and Comes 2004. In tropical mountains, species distributions are geographically much narrower, often as a result of relatively recent, local speciation (Deshpande et al. 2001, Friar et al. 2001, Tremetsberger et al. 2003a, 2003b. Thus, the classical decrease of genetic diversity observed from the equator toward the pole can eventually be blurred for mountain species. Actually, repeated contractionexpansion of species ranges has led rear edge populations to maintain some genetic diversity, therefore counterbalancing the effect of peripheral isolation . Conjointly, the high genetic differentiation between populations underlines the conservation relevance of those populations. Mediterranean mountains represent an interesting case, because they often have a relic temperate-like bioclimate at their top (with no or little summer drought) in a context characterised by severe water deficit for at least two consecutive months at lower altitudes. Mediterranean mountains can therefore be considered as climatic islands, where plant diversity patterns are influenced by different factors (or in different ways) with respect to temperate and boreal mountains (Winkler et al. 2016). Furthermore, climatic and land-use changes have different effects on Mediterranean vs Boreal-Temperate mountains of Europe, being detrimental for the floral richness of the first and increasing the species richness of the second (Pauli et al. 2012). Considering that expected climatic trend is an increasing of temperature and a decrease of precipitation (mainly during spring) in Mediterranean mountains, whereas non-Mediterranean European mountains will not experience a reduction of annual and spring precipitation (Bravo et al. 2008), the urgency rises to monitor those mountains at the transition between Temperate and Mediterranean bioclimates. Moreover, before the middle of the century, the expected climatic changes will provoke the disappearance or strong reduction of a suitable habitat in the summit area, where most of the endemic and/or rare species are located (Benito et al. 2011). The most endangered habitats and species are those linked to water availability like streams, wet meadows and temporary ponds (Ghosn et al. 2010, Pérez-Luque et al. 2015. On islands, threats to mountain floras are even more acute compared to mainlands, because narrower spatial scales of habitats and the usually lower mountain altitudes (Vogiatzakis et al. 2016), led some species to have a relic distribution , Mayol et al. 2015, Fazan et al. 2017. Historical climatic fluctuations and associated ecological constraints are the basis of the fragmented distribution of Boreal-Temperate species on Mediterranean mountains (Mayol et al. 2015, Iszkulo et al. 2016) and determine the presence of plant refugia, climatically stable areas that constitute key areas for the long-term persistence of species and genetic diversity, especially at present and future decades given the threat posed by the extensive environmental change processes operating in the Mediterranean region. These refugia, including large Mediterranean islands, represent 'phylogeographical hotspots'; that is, significant reservoirs of unique genetic diversity favourable to the evolutionary processes of Mediterranean plants (Médail and Diadema 2009).
The island of Sardinia, the second largest in the whole Mediterranean basin, was already known to have a prevalent Mediterranean bioclimate, with a temperate bioclimate in the two main massifs of the island, the Gennargentu (centre-eastern Sardinian, maximum elevation 1834 m a.s.l.) and the Limbara (north-eastern Sardinia, maximum elevation 1359 m a.s.l.) (Arrigoni 1968). Recent detailed bioclimate analysis (Canu et al. 2015) also showed that the only mountain chain of the island named Marghine-Goceano (located between the Limbara and the Gennargentu massifs, maximum elevation at Mt. Rasu 1259 m a.s.l.) is characterised by a temperate bioclimate (in the sub-Mediterranean variant) along the ranges summit. Although the mountain floras of the Gennargentu and Limbara are well documented (Veri andBruno 1974, Arrigoni andCamarda 2015), floristic information about the Marghine-Goceano range is lacking (Valsecchi and Corrias 1966).
This paper goes some way to fill this knowledge gap by reporting on the indigenous flora of a forest domain located in the middle of the Marghine-Goceano range. Our aim was to provide a checklist of the flora located in this area to enable future characterisation of the biotic environment of this mountain area of Sardinia. This data will also allow the identification of target species to monitor and understand climate changes in the particular context of Mediterranean islands.

Study area
The forest domain of Anela is a public property since 1886, at present managed by the Sardinian regional agency Forestas (Fig. 1). The domain covers 1280 hectares of which 1200 ha fall in the municipality of Anela, 55 ha in that of Bultei (to the east) and 25 ha in that of Bono (to the west). The lowest altitude is about 600 m a.s.l. in locality Badu Edras whereas the summit point is at Punta Masiedda 1158 m a.s.l. The Figure 1. The study area, Forest of Anela and its location in Sardinia (red rectangle on the inset map). Colours on the map represent different isobioclimates (derived from Canu et al. 2015). In the domain, we can recognise five different isobioclimates: Violet: upper mesotemperate (subMediterranean), lower humid, weak semi-continental; blue: lower supraMediterranean, lower humid, weak semi-continental; orange: upper mesoMediterranean, lower humid, weak semi-continental; lilac: upper mesoMediterranean, upper subhumid, weak semi-continental; pink: upper mesoMediterranean, lower subhumid, weak semi-continental. Thick black lines represent domain limits; thin black lines represent altitude intervals of 100 m. geographic coordinates of the forestry station headquarter are 40°27'14"N; 9°01'36"E. At present, the vegetation cover is mainly characterised by coppices and mature shrubs linked to sub-Mediterranean woods Glechomo sardoae-Quercetum congestae and Saniculo europaeae-Quercetum ilicis above 800 m a.s.l. and meso-Mediterranean Loncomelo pyrenaici-Quercetum ichnusae and Galio scabri-Quercetum ilicis below 800 m a.s.l., as described by Bacchetta et al. (2009) (Sechi and Falchi 2013). It should be noted that a large fire destroyed 800 hectares of the domain on 31 July 1945, so the wooded area decreased from 72.4% in 1910 to less than 20% in the 50s (Sechi and Falchi 2013).
In the ambit of the Sardinian-Corsican biogeographic province (as defined by Bacchetta et al. 2012), the study area falls in the Goceano-Logudorese sector .
The geology of the study area comprises Palaeozoic granites and schists (Madrau 2013). The impermeable nature of these substrates has created a substantial aquifer evident by the presence of 39 springs (half perennial and half seasonal) in the study area (Farris 2013b).
Bioclimate analysis of 1971-2000 data (Canu et al. 2015) showed that 96.9% of the area falls in the Mediterranean Pluviseasonal Oceanic bioclimate, whereas 3.1% in the Temperate Oceanic bioclimate (submediterranean variant). A total of 64.6% of the area is included in the meso-Mediterranean thermotype, 32.3% in the supra-Mediterranean and 3.1% in the meso-Temperate.
Thermo-pluviometric data of the period 1951-1985 showed annual mean temperature of 11.2 °C and annual mean rainfall of 1040 mm; after the year 2000 temperatures did not vary significantly, whereas a reduction of ca. 30% in the annual rainfall was recorded. Late spring and summer rainfall (May-August) decreased even more (more than 50%, see Farris 2013a).
The study area is entirely included in the Natura 2000 site of community importance ITB 011102 ''Catena del Marghine e Goceano'', extended on 14,984 ha and is also nominated as a Protection Oasis for wildlife "Foresta Anela", managed by the Province of Sassari.

Floristic research
Floristic research started in the year 2000 and was intensified in the years 2012-17 with regular monthly sampling. Each month, we made one day excursions, which covered three altitudinal ranges (< 800 m a.s.l.; 800-1000 m a.s.l.; > 1000 m a.s.l. on the third). For each excursion, we tried to visit as many habitats as possible in order to capture the highest environmental heterogeneity. Collected plants were stored at the Herbarium SS, where we also searched for specimens collected in previous decades (if present, they are reported in the floristic list).
Plant names were derived from the Euro+Med PlantBase (Euro+Med 2006-2018), except for: a) families not already included in this database for which we referred to the Checklists of Italian Flora (Conti et al. 2005;Bartolucci et al. 2018), APG IV (APG 2016); b) the family Orchidaceae (for which we follow GIROS (2016)); c) the genus Orobanche, for which we follow Domina and Arrigoni (2007); d) the genus Dianthus, for which Bacchetta et al. (2010) is followed; e) and the species Struthiopteris spicant which we use in preference to Blechnum spicant (Gasper et al. 2016); f ) for endemics, we also consulted Arrigoni et al. (1976Arrigoni et al. ( -1991 and Peruzzi et al. (2014). The Italian floras (Pignatti 1982(Pignatti , 2017(Pignatti -2018 and the Sardinian flora (Arrigoni 2006(Arrigoni -2015 were also consulted. When other relevant literature was followed, it is specified in the text. Plant authorities and names were further verified using 'The Plant List', 'The World Checklist of Selected Plant Families' and 'The International Plant Names Index' (IPNI). Herbarium acronyms follow Thiers (2018).
The taxonomic circumscription of orders and families, as well as their sequence in the list was derived from Smith et al. (2006) for Pteridophytes; and APG III (APG 2009), APG IV (APG 2016) and Haston et al. (2009) for Angiosperms. Within each family, genera, species and subspecies are listed in alphabetical order. Species and subspecies are numbered progressively.
For each taxon we report: Progressive number Scientific name (with authority) Biological type, Chorologic type Abundance (locality(ies) of collection is(are) specified only for uncommon or range restricted taxa): Habitat Notes (eventual) Biological types are in accordance to Raunkiaer (1934) and were verified on the collected samples and also in Pignatti (1982Pignatti ( , 2017Pignatti ( -2018; chorologic types were determined following maps reported in the Euro+Med PlantBase (Euro+Med 2006 and again verified in Pignatti (1982Pignatti ( , 2017Pignatti ( -2018  Turanian.
Here we consider as endemics sensu stricto all taxa limited to the Corsican-Sardinian biogeographic province (sensu Bacchetta et al. 2012), therefore including the Tuscan Archipelago. Other taxa are considered endemic sensu lato, which includes those present in western Mediterranean islands and continental areas -Calabria in Europe, Kabylies in Africa -as far as the Miocene part of the Hercynian chain (Hercynian endemics sensu Mansion et al. 2008). Finally, other endemics sensu lato are 'administrative endemics', i.e. taxa confined within Italian national borders (Peruzzi et al. 2014 Sicily.
Abundance is expressed on the basis of the following criteria: RR range restricted: taxa present in only one locality within the study area or covering a surface not exceeding 1 hectare, i.e. Mentha requienii subsp. requienii; U uncommon: taxa found in 2-5 localities within the study area, or covering a surface not exceeding 1 km 2 , i.e. Arisarum vulgare; L localised: taxa present in 6 or more localities within the study area, but covering less than 2.5 km 2 , i.e. Agrostis capillaris; C common: taxa covering more than 2.5 km 2 , i.e. Quercus ilex. Asplenium foreziense Magnier H ros, NW-Medit.-Mont. U (Badu Edras): Shady rocks and cliffs Notes: the taxon has been excluded from the Sardinian flora by Marchetti (2004) and Bartolucci et al. (2018), but confirmed by Arrigoni (2006Arrigoni ( -2015 Pignatti (1982), Conti et al. (2005), Arrigoni (2006Arrigoni ( -2015, Pignatti (2017Pignatti ( -2018, Bartolucci et al. (2018) Bagella and Urbani (2006), this is the valid name for Malus domestica Borkh. (nom. illeg.), also reported in the Euro+Med Plant-Base. Yet Galasso et al. (2018) call a taxon Malus domestica, considering it as a non-native species, while Camarda and Valsecchi (2008), Arrigoni (2006Arrigoni ( -2015 and Pignatti (2017Pignatti ( -2018 still call it M. dasyphylla. Finally, Bartolucci et al. (2018) report the taxon M. sylvestris in Sardinia. Malus pumila is reported as a synonym of M. domestica by Galasso et al. (2018), it is excluded from the Sardinian flora by Arrigoni (2006Arrigoni ( -2015, finally, it was not mentioned by Camarda and Valsecchi (2008 (Mossa et al. 1998(Mossa et al. , 1999. Until the various treatments are resolved, we prefer to treat this variation as a complex (or Arrigoni (2006Arrigoni ( -2015, but later confirmed by Mereu (2012) (Conti et al. 2005 consider the subsp. arundanum (Boiss.) Nyman as present in Sardinia, whereas, the Euro+Med PlantBase considers subsp. arundanum absent from the island (and the whole Italian peninsula) and that, instead, subsp. orientale is present. Our specimens fit well with the diagnostic characters of subsp. orientale as described by Bothmer (1967

Ecological and biogeographical analysis of the indigenous flora of Anela
Here we assess the presence in the forest domain of Anela of 391 taxa, belonging to 32 orders and 74 families.
Hemicryptophytes dominate within the Boreal-Temperate and the Eurasian components; annual species prevail within the widespread and the Mediterranean-Atlantic groups. The Mediterranean component hosts similar percentages of annuals and hemicryptophytes (Fig. 2).
The endemic component of the flora of Anela is dominated by those of the Sardinian-Corsican biogeographic province (sensu Bacchetta et al. 2012) accounting for 28 taxa (endemics sensu stricto, 7.4%), of which 19 taxa are Sardinian-Corsican (42.2% of the endemic component), followed by Sardinian entities (5, 11.1%) and those present on Sardinia, Corsica and the Tuscan Archipelago (4, 8.9%). Tyrrhenian or Hercynian endemics (those present in Sardinia, Corsica, Tuscan Archipelago, the Balearic and Hyeres Islands and Sicily) account 12 (26.7%) and, finally, 11.1% is constituted by 5 entities with larger ranges including some continental areas (Sardinia and northern Africa or Sardinia and Italy).  On the basis of our criteria, 241 taxa (61.6%) can be considered common at the local level, 113 (28.9%) are localised, 23 (5.9%) are uncommon, 9 (2.3%) are range restricted and 5 (1.3%) are locally extinct in the last 50 years. Common taxa are the dominant category in all the geographic groups, whereas range restricted taxa are found only in the widespread, Boreal-Temperate and the Mediterranean groups (Fig. 3).

Biogeographical description of the mountain
Our research discovered a high species density at the study area (30.6 taxa km -2 ), that is one of the highest ever documented in the Sardinian mountain floras (Table 1). Even if there is a clear inverse relationship between the area investigated and species' density, we should note that, for areas having a comparable surface (~ 10 km 2 ), the floristic density recorded at our study area is second only to the Mt. Gonare complex (Camarda 1984a(Camarda , 1984b. It is noteworthy that the summit area of Sardinia (> 1500 m a.s.l.), having a surface of 16.8 km 2 , hosts "only" 214 taxa of which 66 are considered endemics (Arrigoni and Camarda 2015). So we can argue that areas at the edge between the Mediterranean and the temperate bioclimates, like Foresta Demaniale Anela and Mt. Gonare, host floristic components from both the two bioclimatic -biogeographic regions, having therefore more abundant floras than areas located in coastal or summit zones.
The hemicryptophytes/therophytes (H/T) ratio, as previously noted by Arrigoni and Camarda (2015), underlines the co-presence of two main elements, the perennial and the annual herbs, having very different life-cycles and summing 71.1% of our flora.The H/T ratio, that in Sardinia peaks at 2.5 at the summit of Gennargentu (Arrigoni and Camarda 2015), but decreases to 0.74 as the regional average, is at Anela 1.03. Limestone mountains like Mt. Albo, with a karst geology and consequently a pronounced summer drought, have a H/T ratio even lower than the regional average, whereas mountain complexes with impermeable substrates (plutonic, volcanic, metamorphic) approaching 1000 m a.s.l. have a H/T ratio ~ 1 gradually increasing with elevation (Table 1). This means that at 1000 m a.s.l., the co-presence of two large groups of non-woody plants, having an annual or perennial life cycle, has been detected: the annuals have a greater prevalence at lower altitudes, the perennials at higher altitudes and their ratio ~ 1 at 1000 m a.s.l. underlines the transition character of this altimetric level in Sardinia. Important differences with the regional (Sardinian) value (Pignatti 1995) have also been detected for the Mediterranean floristic component, particularly the steno-Mediterranean taxa having a 28.9% regional percentage and 19.7% at the Anela forest domain; contrarily, the euri-Mediterranean component has 16.1% regional average but increases to 24% at our study area, the same percentage (24.3%) reached by the sum of the Boreal-Temperate and the Eurasian floristic components. Whereas lower altitude floras have a dominant steno-Mediterranean component and the floras at the summit of Mediterranean mountains show the prevalence of southern-European and Mediterranean orophytes and narrow endemics Arrigoni and Camarda 2015), our flora is a good example of transition areas, having the 80% of taxa quite equally distributed amongst steno-Mediterranean, euri-Mediterranean, Boreal-Temperate and Eurasian and the endemic contingents. High species density, H/T ratio ~ 1, balance amongst different chorologic groups and endemic percentage ~ 10% can be considered characteristic features of mountain areas at the transition between the Mediterranean and the temperate bioclimates.
The composition of the flora of the Forest Domain of Anela is also peculiar because it is one of the few examples, not only in Sardinia but in the whole Mediterranean area, with no native Gymnosperms. Junipers (Juniperus phoenicea subsp. turbinata (Guss.) Nym. and J. oxycedrus subsp. macrocarpa (Sibth. & Sm.) Neilr.) in NW Sardinia are mainly confined in coastal areas (Farris et al. 2017), but Yew (Taxus baccata L.) and Prikly Juniper (Juniperus oxycedrus L. subsp. oxycedrus) are usually present in high hills and mountains. However junipers are not present in NW Sardinia inland areas (Farris et al. 2017), but the Yew is occurring in all the massifs and mountain ranges, including the two forest domains bordering Anela, the Fiorentini Forest Domain to the east (municipality of Bultei) and the Mt. Pisanu Forest Domain to the west (municipality of Bono, see Farris and Filigheddu 2008). The total absence of Gymnosperms in the native flora of the Anela forest domain is therefore surprising, most probably anomalous and it seems likely to be linked to the management history of the area rather than a natural pattern (Sechi and Falchi 2013).
Despite the fact that in 2004 (last forest census) 90.4% of the domain area was covered by forest or shrub communities (Sechi and Falchi 2013), it is striking that the 45% of the detected taxa were linked mainly to herbaceous habitats (annual and perennial grasslands, dry and wet pastures and meadows), already described for their peculiar and original floristic composition (Farris et al. 2013). Traditional grazing, particularly ovine pastoralism characterised by low flock density and transhumance, has been proven to be beneficial for the plant biodiversity of Mediterranean silvo-pastoral systems, whereas abandonment is detrimental even at short temporal scales (Farris et al. 2010a). The forest domain of Anela is a typical case where ovine stocks had a dramatic decrease in a short period: between 1990 and 2007, a decrease from 0.77 sheep ha -1 to 0.13 sheep ha -1 has been recorded (-83%, Farris et al. 2010a), whereas wood and shrub communities linked to potential natural vegetation (sensu Farris et al. 2010b) are recovering very fast, following a trend common to all Italy (Falcucci et al. 2007) and particularly to Sardinia (Puddu et al. 2012).

Conservation issues of this Flora
Even if rarity is not always linked to threat (de Lange andNorton 1998, Bacchetta et al. 2012), it is an important feature to consider when setting conservation priorities within long lists of taxa (Bacchetta et al. 2012, Le Berre et al. 2018, as in the case of the flora of the Anela forest domain. Additionally, 14 out of 32 uncommon and range-restricted taxa found in this flora are linked to wet habitats: some belong to the Mediterranean and endemic contingents (Cerastium ligusticum subsp. palustre, Exaculum pusillum, Isoetes hystrix, Mentha suaveolens subsp. insularis, Mentha requienii subsp. requienii, Morisia monanthos, Oenanthe lisae), others to the Eurasian and Boreal-Temperate contingents (Struthiopteris spicant, Carex remota, Iris pseudacorus, Solanum dulcamara, Spiranthes spiralis). Those habitats are supposed to be highly vulnerable (Filipe et al. 2013), as changes in land use and modification of water balance (because of climate change or human use) are amongst the most important threats to wetlands. Moreover, little is known about the resilience of associated plant communities, a threat increased by the high spatial isolation of such places within a Mediterranean context. At the study site, we detected several species having a contraction of range or local extinctions caused by the capture of surface or underground water for human use, as for example Struthiopteris spicant, Cerastium ligusticum subsp. palustre, Mentha requienii subsp. requienii and the localized fern Osmunda regalis for which we documented a local decrease > 50% in the last 20 years. Other species had a decrease directly caused by drainage of temporary ponds (Exaculum pusillum, Isoetes hystrix, Morisia monanthos). Water management in a climatic changing scenario is and will increasingly be a key issue for the conservation of biodiversity in the Mediterranean basin (Casazza et al. 2014), a climatic change hotspot at the global scale (Giorgi 2006, Giorgi andLionello 2008), where wet habitats and the species linked are amongst the most threatened (Ghosn et al. 2010, Pérez-Luque et al. 2015. The 5 taxa, locally extinct, have no relationship with a particular habitat or human use from which they are (were) dependent for their survival in the area, with the exception of Chenopodium album whose disappearance could be explained with the abovementioned abandonment of pastoral activities, as it is a nitrophilous species. Their disappearance in the last decades, inferred from herbarium records, can be therefore a normal turnover in the composition of the local indigenous flora or an artifact derived from our sampling method (in the sense that these taxa are maybe still present in the area but we were not able to find them during our monthly sampling excursions).
Amongst the flora we inventoried, it is worth mentioning that several populations represent peripheral populations regarding the overall distribution of the taxa. First, a group of uncommon or range restricted species in the domain, are common plants in the Mediterranean bioclimate areas of Sardinia and sometimes in the whole basin. They are here confined to warm niches in the mountain area under study (Anemone hortensis, Arbutus unedo, Arisarum vulgare, Arum pictum, Celtis australis, Ficus carica, Ptilostemon casabonae), places relatively scattered through this mountain landscape. Oppositely, several Boreal-Temperate and Eurasian taxa confined in this sub-Mediterranean bioclimate island represent peripheral populations isolated sometimes by over 1000 km of their northern range. Those constitute rear edge populations  which may contain unique genetic variation, inherited from ancient species distribution and particular ecological conditions. These two contrasted situations have been highlighted several times within the Mediterranean flora (Lavergne et al. 2005(Lavergne et al. , 2006 and are characteristic of those climatic transition areas. These plants all share the characteristic of occurring as fragmented, disjunct and often highly isolated populations, which restrain gene flow with central population (Pironon et al. 2017) and enhance amongst-population differentiation . Thus, the relative isolation associated with potentially marginal ecological conditions highlight their evolutionary potential (Thompson 1999, Anacker andStrauss 2014), as it has recently been shown in Sardinia and Corsica for some marginal and peripheral populations of Cyclamen repandum . Additionally, these groups of taxa are often found in different macro-habitats which have very different links with human activities, therefore leading to different threats and management issues (Lavergne et al. 2006). Thus, conservation policies need to integrate such complex entities within their framework (Lesica and Allendorf 1995, Brunnell et al. 2004, Leppig and White 2006. Finally, those transition areas also contain numerous endemics, which render those places original and of high value for conservation.
Even if biodiversity hot-spots definition at multiple spatial scales is commonly based on the presence, density and distribution of endemic taxa (Myers et al. 2000, the data here presented support that other parameters should also be taken into account to more precisely define priority areas for conservation, as taxonomic complexity (Ennos et al. 2005) of floras and evolutionary potential of populations (Thompson et al. 2010), detected within continuous schemes of biodiversity monitoring (Marignani et al. 2014). This is particularly urgent in southern European mountains, whose biodiversity is threatened by both climate and land use changes (Bravo et al. 2008, Benito et al. 2011, Pauli et al. 2012, Vogiatzakis et al. 2016.