Our study area is Tuscany (Italy). Our relevés spanned from the Selva Pisana in the north until the Laguna di Orbetello in the south (Fig. 1). Our relevés were placed in Regional Protected Areas, partly coinciding with SACs belonging to the Natura 2000 network. The macrobioclimate is Mediterranean with upper mesomediterranean thermotype and ombrotypes ranging from upper humid to upper dry (Pesaresi et al. 2017).

The target vegetation types included the following habitats: 1310 - Salicornia and other annuals colonizing mud and sand, 1410 - Mediterranean salt meadows (Juncetalia maritimi), 1420 - Mediterranean and thermo-Atlantic halophilous scrubs (Sarcocornietea fruticosi), and 1510 - Mediterranean salt steppes (Limonietalia), the latter being a priority habitat type of conservation interest.

We gathered a dataset comprising a total of 418 relevés (Suppl. material 1, Table S1), of which 157 were carried out by the authors between 2018-2021, for the purposes of monitoring activities following Art. 17 of the Habitats Directive. Additionally, data from 261 plots were derived from published scientific literature. The distribution of the relevés is illustrated in Fig. 1. Data concerning the localities of the sampled sites and the correspondence between data and references are reported in Suppl. material 2, Table S2.

The position of our own plots (each of 4 m²) was selected by stratified random sampling in each of the SACs. The locations of the SACs were indicated in the EU habitat maps provided by the HaSCITu (Habitats in the Sites of Conservation Interest in Tuscany) program of the Tuscan Regional Administration (Regione Toscana 2018). After surveying the available scientific literature, only the relevés that were taken between 2000 and today and an area < 50 m² were retained for our analysis.

Species occurring less than 4 times across all relevés were deleted, resulting in a final table consisting of 418 relevés and 84 (out of the originally 152) species. The Braun-Blanquet cover-abundance scale was transformed according to the ordinal scale proposed by van der Maarel (1979) and Noest et al. (1989): (r = 1; + = 2; 1 = 3; 2 = 5; 3 = 7; 4 = 8; 5 = 9) to have a linear scale and ensure comparability of data from different sources taken using different cover-abundance scales.

The relevés-species matrix (Suppl. material 1, Table S1) was analysed performing a LDA (Linear Discriminant Analysis) in the software PAST 3.14. LDA tests the assignment of each relevé to a given habitat, which in our case was either indicated in the original literature, or, in case of our own data, assigned during the data sampling. The diagnostic species of each habitat were statistically calculated by computing the “phi coefficient of association” based on presence/absence data (Chytrý et al. 2002). The significance of this fidelity coefficient was calculated according to a Fisher’s exact test. We considered a species as diagnostic when phi > 20 and p < 0.01 (Douda et al. 2016). For each habitat, we reported species that had a phi coefficient > 20 and/or a cover coefficient (sum of the central cover values divided by the number of relevés in the table x 100, Géhu 2006) > 50, together with their frequencies. All these values were calculated on our dataset and are therefore valid only at the local scale. Thus, the here identified groups of species, hereafter “local assembly of typical species”, will be useful for the recognition of the saline habitats in the wild and for monitoring their conservation status in Tuscany. Also, only for comparison, we added the species reported by Biondi and Blasi (2009; sub: “combinazione floristica di riferimento”), European Environment Agency (2021; sub: “typical species”), European Commission (1996; 2013), Angelini et al. (2016, 2018; sub: “specie tipiche”) and Regione Toscana (Regione Toscana 2018; sub: “specie indicatrici”).

Species taxonomy was retrieved from Acta Plantarum (Acta-Plantarum 2021); the names of syntaxa above the association level follow Mucina et al. (2016). The delimitation of habitats follows Biondi and Blasi (2009), European Commission (2013), Angelini et al. (2016) and the Tuscany Region website (Regione Toscana 2018).

Habitat 1310: The vegetation types reported for this habitat relate to several associations (at least 13 vegetation types, depending on the different interpretations of the syntaxonomic level), of three phytosociological classes: Crypsietea aculeatae, Therosalicornietea and Saginetea maritimae. Four types described for the coast of Tuscany were dominated by Salicornia perennans Willd. (sub: S. patula Duval-Jouve) and/or S. procumbens Sm. (sub: S. emerici Duval Jouve and S. dolichostachya Moss) from northern (Bertacchi et al. 2021), central (Viciani and Lombardi 2001) and southern areas (Andreucci and Castelli 2001; Andreucci 2004). Other associations referred to the class Saginetea maritimae, order Frankenietalia pulverulentae, reported for Orbetello (Andreucci 2004), Pianosa (Foggi et al. 2008), Gorgona (Viciani et al. 2011) and Giannutri (Foggi et al. 2011). Only one association relative to the class Crypsietea aculeatae was reported: Crypsietum aculeatae (Bojko 1932) Wenzl 1934 (Viciani and Lombardi 2001; Bertacchi et al. 2021). Several phytocoenoses referred to this class were reported also for Corsica (Paradis and Lorenzoni 1994). However, they were included in the habitat 3170 (Mediterranean temporary ponds) by Bensettiti et al. (2002), pointing towards the need for further investigations. Most vegetation types of the habitat 1310 need an update from the nomenclatural point of view due to the several novelties in the taxonomy of the Salicornia L. genus.

Habitat 1410: For this habitat, nine vegetation types were reported for Tuscany. All these vegetation types can be referred to one class and one order: Juncetea maritimi and Juncetalia maritimi, and further to three alliances: Limonion etrusci, Juncion maritimi, and Plantaginion crassifoliae. These findings confirm a high homogeneity of this habitat along the coast of Tuscany, even though the floristic composition is not always useful to separate it from the habitat 1420 (see also Fig. 2 where a little overlap with 1420 was found). However, the two habitats differed in their physiognomic configuration: 1410 is dominated by rushes and sedges, while 1420 is dominated by succulent shrubs or subshrubs. A succulent subshrub like Limbarda crithmoides subsp. longifolia, is generally reported as “typical” for the habitat 1410, which is confirmed by our results (phi = 48). Furthermore, Limbarda crithmoides subsp. longifolia was chosen as characteristic of the association Limbardo crithmoidis-Limonietum etrusci (Viciani et al. 2012). Mucina et al. (2016) assigned the alliance Limonion etrusci, for which the association is the nomenclatural type, to the order Juncetalia maritimi. Accordingly, Limbarda crithmoides subsp. longifolia should be selected as a suitable “typical species” for the habitat 1410. In the habitat 1410, two invasive alien species were found: Symphyotrichum squamatum (phi = 21.1) and Xanthium italicum (phi = 20.5), indicating an unfavourable state of conservation of this habitat in the studied region.

Habitat 1420: According to our results, this is a highly homogeneous habitat within one class, one order (Salicornietea fruticosae and Salicornietalia fruticosae, respectively) and two alliances: Salicornion fruticosae and Arthocnemion glauci. Only three associations were found.

Habitat 1510: According to our results (Fig. 2) and from a strictly phytosociological point of view, we propose that the communities dominated by Limonium narbonense, should be included in the order Juncetalia maritimi instead of Limonietalia (as reported by Viciani et al. 2012). Our proposal is based on the absence of Limonietalia species, except for Limonium narbonense. The same is true for the alliance Limonion etrusci (Viciani et al. 2012), that according to Mucina et al. (2016) was classified to the order Juncetalia maritimi. The coenoses with Limonium etruscum, previously referred to 1510 by Viciani et al. (2012), were here included in the habitat 1410.

Syntaxonomical scheme for high ranks, according to Mucina et al. (2016). Reference to habitats is indicated in brackets.

Therosalicornietea Tx. in Tx. et Oberd. 1958 (habitat 1310)

Therosalicornietalia Pignatti 1952

Therosalicornion Br.-Bl. 1933

Saginetea maritimae Westhoff et al. 1962 (habitat 1310)

Saginetalia maritimae Westhoff et al. 1962

Saginion maritimae Westhoff et al. 1962

Frankenietalia pulverulentae Rivas-Mart. ex Castroviejo et Porta 1976

Frankenion pulverulentae Rivas-Mart. ex Castroviejo et Porta 1976

Crypsietea aculeatae Vicherek 1973 (habitat 1310)

Crypsietalia aculeatae Vicherek 1973

Cypero-Spergularion salinae Slavnic 1948

Juncetea maritimi Br.-Bl. in Br.-Bl. et al. 1952 (habitat 1410)

Juncetalia maritimi Br.-Bl. ex Horvatić 1934

Juncion maritimi Br.-Bl. ex Horvatić 1934

Plantaginion crassifoliae Br.-Bl. in Br.-Bl. et al. 1952

Limonion etrusci Viciani et al. 2012

Limonium narbonense community (habitat 1510)

Salicornietea fruticosae Br.-Bl. et Tx. ex A. Bolòs y Vayreda et O. de Bolòs in A. Bolòs y Vayreda 1950 (habitat 1420)

Salicornietalia fruticosae Br.-Bl. 1933

Salicornion fruticosae Br.-Bl. 1933

Arthrocnemion glauci Rivas-Mart. et Costa M. 1984

Results of the phi analysis are shown in Tab. 1. In Tabs 2–5, we report the “local assembly of typical species” according to the above-mentioned rules.

Habitat 1310: Among the species found in this habitat (N = 82), the “local assembly of typical species” included 12 species with phi > 20 and 2 species with a cover coefficient > 50. This result might be related to the heterogeneity of this habitat. Among the species with phi > 10, we found Soda inermis (Tab. 2) with a very low coefficient value. According to our results, many of the species reported as “typical” in Biondi and Blasi (2009), European Environment Agency (2021) and Angelini et al. (2016) cannot be considered for Tuscany. Our results suggest that species, previously considered as typical for the studied habitats in regional sources, cannot be applied at the local level in Tuscany. We found one invasive alien species: Erigeron canadensis with phi = 15.8 and a cover coefficient > 50, that, at least locally, indicated a low conservation status of habitat 1310.

Habitat 1410: The habitat 1410 was rich in species (N = 91). This finding is also reflected by the large number of species that meet the above defined criteria of the “local assembly of typical species”. Among both, the own relevés and the relevés derived from literature allocated to 1410, 13 species had a phi > 20, and two species reached a phi > 10 and a cover coefficient > 50 (Tab. 3).

Habitat 1420: The “local assembly of typical species” for the habitat 1420 included very few species. Only three of them (Salicornia perennis, Arthrocaulon macrostachyum, Halimione portulacoides) showed medium-high phi-values (phi > 30) but a high coefficient value (> 1500). These species were succulent (low-)shrubs characterizing the habitat also from a physiognomic point of view (Tab. 4). We found two more species with a coefficient value > 50: Puccinellia festuciformis, generally used as characteristic species of several associations referred to this habitat, and Triglochin barrellieri.

Habitat 1510: As already stated, the relevés allocated to the habitat 1510 were found in the area delimited by the relevés of the habitat 1410. However, the constant presence of Limonium narbonense with very high phi-values (phi = 69.7) and a high coefficient value (> 6000), pointed towards the attribution of habitat 1510, defined as rich in perennial species of the genus Limonium (Biondi and Blasi 2009) (Tab. 5). The habitat 1410 is linked to humid depressions and is dominated by species of rushes belonging to the genus Juncus (Tab. 4). Juncus species are generally absent at sites located at higher topographic levels, where the habitat 1510 can be found. In this topographic position, the sites remain generally dry during the whole year and are intensively grazed by cattle. The peculiarities of habitat 1410 call for conservation practices markedly different from the current once. Therefore, to help local authorities in identifying and monitoring the implemented conservation measures, habitat 1410 needs to be addressed separately.

In the habitat 1510, Phragmites australis was found among the species with a phi > 10 (25.4) and with a coefficient value > 50. However, P. australis cannot be included in the “local assembly of typical species” because it was also found in other communities, with higher values of phi and/or coefficient value. Here the species must be considered as a negative presence.