Ecological and Syntaxonomic Analysis of Pinus halepensis Mill. in the Iberian Peninsula and Balearic Islands

: This work aims to review the natural communities of Pinus halepensis in Spain. The methodology consisted of subjecting 400 phytosociological relev é s to georeferencing and statistical, biogeographical, and bioclimatic treatment. We analyse the communities of Pinus halepensis on the Iberian Peninsula and Balearic Islands. Five syntaxa with association rank are described in several works and included in the alliances Rhamno-Quercion and Oleo-Ceratonion . Ephedro-Pinetum halepensis was initially proposed as a community by Torres et al. and subsequently raised to the rank of association by Rivas-Mart í nez et al. In this work, we have separated the plant communities dominated by Pinus halepensis , which was previously included in other syntaxa, and as a result, we propose four new associations and a new alliance for the Iberian Peninsula: ass. Bupleuro rigidi-Pinetum halepensis ; ass. Ephedro nebrodensis-Pinetum halepensis ; ass. Rhamno nsis ; ass. Rhamno laderoi-Pinetum halepensis; all. Rhamno lycioidis-Pinion halepensis . In view of the fact that some of the communities have been published as edaphoxerophilous and climatophilous, we suggest separating the climatophilous from the edaphoxerophilous character in the diagnosis of the communities, and have therefore recently proposed the ombroedaphoxeric index I oex (I oex = P p − e/T p × CR), which considers positive precipitation P p , positive temperature T p , residual evapotranspiration (e), and water retention capacity CR (0.25, 0.50, 0.75). In conclusion, we propose the associations mentioned above, which will allow the implementation of a reforestation treatment in accordance with the natural environment.


Introduction
The genus Pinus L. is widely distributed on the Iberian Peninsula in the form of autochthonous and introduced species [1]. Flora Ibérica lists seven species of pine, of which five can be considered autochthonous, and two are here introduced [1]. The autochthonous species are Pinus pinaster Aiton, Pinus nigra Arnold subsp. salzmannii (Dumal) Franco, Pinus nigra Arnold var. latisquama (Willk.) Heywood, Pinus sylvestris L., Pinus uncinata Ramond Land 2022, 11, 369 3 of 31 the determining factors for its distribution are temperature, soil, and ombroclimate. It is located in thermo-and mesomediterranean thermotypes and ombrotypes that range from the semiarid to the subhumid, where it coexists with kermes oak and mastic woodlands [23].
Because of the controversial status of Pinus halepensis communities in Spain in terms of their autochthonous or introduced character, since we previously demonstrated [9] that Pinus halepensis is an autochthonous species, as recognized by Rivas-Martínez et al. [18], and since this species is widely used in reforestation in Spain, we set ourselves the objective of reviewing the natural communities of this species. We maintain the hypothesis that the ombrotype, with an influence of the substrate, is the cause of the ambivalent character of Pinus halepensis, which may have an edaphoxerophilous or climatophilous character, as demonstrated by applying the ombroedaphoxeric index.
All vegetation relevés were collected following the Braun-Blanquet methodology [26], establishing the concept of minimum area for sampling this type of tree formation. The method basically consists of sampling homogeneous plots of vegetation from the physiognomic point of view and estimating the dominance/abundance of each taxon based on the Braun-Blanquet indices. The sampling area of each relevé or plot was 300 m 2 , largely coinciding with that proposed in the Braun-Blanquet methodology, discarding any relevés whose areas were not between 250-350 m 2 . Other vegetation relevés were discarded because they were repopulations or forestry crops.
The first step was the georeferencing and implementation in a Geographic Information System (GIS), which subsequently allows each vegetation relevé to be associated with other topographic data such as slope, orientation, and biogeographic or bioclimatic information. This establishes a first filter with which to discard any relevés from reforestation, or in which the surrounding vegetation does not correspond to a forest of Pinus halepensis.
The selected relevés were georeferenced with a maximum error of 100 m 2 in 100 × 100 squares. After the database had been created, the geographic variables for orientation and slope for each relevé were obtained from a digital terrain model of the Iberian Peninsula with a resolution of 25 m. The climate variables for maximum, minimum, and mean monthly temperature and mean monthly precipitation were obtained for each sampling point from 4987 meteorological stations distributed around the Iberian Peninsula and North Africa by means of neural network-assisted interpolation using the neural software in the Decision Tools package.
With this preliminary information, a database was created with the geographic, climate, and bioclimatic variables for each georeferenced relevé.
Before conducting any statistical or ordination analysis, the climatic, geographic, and bioclimatic data were normalised according to the following formula: where X is the value of the variable, µ is the mean of the variable, and σ is the standard deviation of the variable. A factor analysis was done after normalising the data in Table 1. Factor analysis is one of the series of multi-variable analytical methods for studying the relations of interdependence that occur between a set of variables or individuals. The criterion for choosing the bioclimatic variables was selected after rotating the table of factors with the Land 2022, 11, 369 4 of 31 VARIMAX algorithm, which transforms the initial factor matrix into a rotated factor matrix to make it easier to interpret. This procedure was followed only for the bioclimatic variables. The bioclimatic variables selected correlated with the factors of over 90%. Table 1. Correlation between the climatic and bioclimatic variables with the two first rotated factors. (MAX = Mean maximum temperature of each month, MED = Mean temperature of each month, MIN = Mean minimum temperature of each month, PE = Thornthwaite annual potential evapotranspiration index, PE = Thornthwaite annual potential evapotranspiration index for February, PRE = Precipitation, Psw = Precipitation of the coldest semester of the year, Pw = Precipitation of the winter quarter, T = Mean annual temperature, Tmax = Mean annual maximum temperature, Tpw = positive temperature of the coldest four-month period, Tpw1 = positive temperature of the coldest month, Tpw2 = positive temperature of the two coldest months, Ts = mean summer temperature, Ts2 = mean temperature of the two warmest summer months, Itc = compensated thermicity index.

Variable
Factor For the biogeographical location of the relevés, we follow Rivas Martínez et al. [42] at the province, sector, and biogeographical district level, and prepare the indices Ic, Io, It/Itc for the bioclimatic analysis [43].
An ordination analysis is applied to the selected variables (biogeography, bioclimatology, topography, plant composition), using the statistical packages CAP (Community Analysis Package III) and Past (PAleontological STatistics) to prepare a hierarchical classification dendrogram of the relevés. This cluster dendrogram was created based on the Kendall distance (which is not a parametric test and is therefore robust to the distribution of the data), and the classification was performed with the complete linkage clustering method. A multivariate analysis was then applied using DCA (Detrended Correspondence Analysis) to classify the relevés. The different alliances were separated with a comparative floristic analysis. We consulted Rivas Martínez et al. [18] to choose species used as characteristic of the alliance for the Iberian Peninsula.
To determine whether there are bioclimatic differences between the different Pinus halepensis communities, the values of the bioclimatic indices previously selected from the VARIMAX analysis were statistically analysed by comparing them (Table 1). An exploratory data analysis and the Shapiro-Wilks normality test were carried out beforehand. Due to the nature of targeted sampling, our data do not follow a normal distribution, so nonparametric analyses were applied. In this case, the Kruskal-Wallis test was applied to determine the existence of any bioclimatic differences between the different associations,    Five syntaxa with association rank are described in several works and included in the alliances Rhamno-Quercion and Oleo-Ceratonion. Ephedro-Pinetum halepensis was initially proposed as a community by ourselves and subsequently raised to association rank by Rivas Martínez et al. [18]. It was described as the head of the semiarid-dry climatophilous and edaphoxerophilous Accitano-Baztetan vegetation series. These are open formations with a climatophilous character that develop in a lower mesomediterranean thermoclimate in the Accitano-Baztetan biogeographical unit (Guadiana Menor valley), Betic biogeographical province, with an optimum in the semiarid ombroclimate on a substrate of poor gypsumrich loam. The semiarid character and the abundance of gypsum do not allow for the development of Quercus rotundifolia, so the climax corresponds to the pinewood. As shown in Table 2, bioclimatic differences can be established between the different associations, with statistically significant differences between the values of the bioclimatic indices analysed for each of the proposed associations. This type of woodland should not, therefore, be considered as edaphoxerophilous when it is located on loams unless it develops on soils with a high gypsum content, as gypsum has the property of retaining water (trapped water), which is not useful for the plant. In these circumstances, the semiarid territory behaves as an arid territory [45]. The floristic composition of this pinewood includes Ephedra fragilis, Juniperus oxycedrus, Rhamnus lycioides, Asparagus horridus, Pistacia lentiscus, and occasionally Juniperus phoenicea.

Ass. Arbuto-Pinetum halepensis
Arbuto-Pinetum halepensis has been described as a climatophilous and edaphoxerophilous community with an Alcañizano-Gandesan distribution, and is considered a vicariant of Pistacio-Pinetum halepensis. This association is described as calcicolous, calcodolomitic, and clayey materials in the semiarid-dry (the mean of Io index is 1.97) lower mesomediterranean bioclimate (the Itc index mean is 285.9). The Connover-Iman test establishes that there are bioclimatic differences between these two settlements for the variables MAX-JANUARY (p-value = 0.0002), PREC-JANUARY (p-value < 0.0001), PREC-DECEMBER (p-value < 0.0001) and Ic (p-value < 0.0001). Arbutus unedo is a species with a subhumid-humid and acidophilus optimum that penetrates in decarbonated clayey iron oxide-rich soils. The presence of the acidophilus elements A. unedo and Viburnum tinus gives this association a neutrophilous aspect. This species is not usually found in semiarid-dry environments, although it can occur on some humid rocky sites, allowing it to be potentially diagnosed as an edaphoxerophilous association. According to Rivas-Martínez et al. [18], Arbuto-Pinetum halepensis is Alcañizano-Gandesan, which corresponds to the easternmost territories of the Bardenero-Monegrino sector (Central Iberian Mediterranean province). Due to their floristic composition and distribution, we add to the relevés the author uses to establish the association (Table 75.7.26, Itinera Geobotanica 18:2, page 427) the relevés of Álvaréz de la Campa Fayos [24] which he included in Querco cocciferae-Lentiscetum Br.-Bl., Font Quer, G. Br.-Bl., Frey, Jansen & Moor 1936; and those of Rovira i López [40], which were adscribed to the association Rhamno lycioidis-Quercetum cocciferae Br.-Bl. and Bolòs 1954, but which have a predominance of P. halepensis, collected in the Valenciano-Tarraconensean sector (Catalano-Provenzal-Balearic province). These two associations present notable bioclimatic differences.

Ass. Ephedro nebrodensis-Pinetum halepensis
The Bardenero, Monegrino, and Somontano biogeographical sectors belong to the Central Iberian Mediterranean province [42][43][44][45], which include two groups of relevés of Braun-Blanquet & Bolòs [26]. The first group is relevés located in very warm environments with an ombrotype ranging from the lower dry to the upper semiarid in areas of d'Escatrón-Caspe-Candasmos, in calcareous semi-arid territories with gypsum, which were adscribed to Rhamneto-Cocciferetum pistacietosum lentisci. These samples did not correspond to a kermes oak-mastic woodland, as P. halepensis is the dominant element ( EnPh24 to EnPh40, holotypus rel. EnPh32) for the climatophilous character. This association is distributed in the Zaragozano Estepario and Belchitano Hijarensean districts in the Bardenero-Monegrino sector, but may occasionally extend into the Bilbilitano-Serrano Cucalonensean district in the Northern Oroiberian sector. Bioclimatically it is very similar to Arbuto-Pinetum halepensis. However, there are significant bioclimatic differences between the rest of the associations in variables related to the distribution of precipitation, maximum, average, and minimum monthly temperatures and in the values of the bioclimatic indices for PE, PEs, Ic, Io, and Itc, as can be seen in Table S1 (Supplementary Material).

Ass. Bupleuro rigidi-Pinetum halepensis
The relevés in the second group collected by Braun-Blanquet & Bolòs [26] are located at higher altitudes than the previous group, between 600-700 m. These are more continentalised and rainier environments on calcareous substrates in the Cincovillés, Zaragozano Estepario, and Somontano Aragonés districts in the Somontano sector; and in the Monegrino district in the Bardenero Monegrino sector, occasionally extending to the Alcañizano district, with an upper semi-arid to lower dry ombrotype, and in more continentalised environments than the previous association. This can be deduced from phytosociological table no. 45 of Braun-Blanquet & Bolòs [26], in which the authors include Rhamneto-Cocciferetum subass. cocciferetosum and subass. caricetosum humilis. In both cases, there is a predominance of Pinus halepensis, along with other species such as Quercus coccifera, Rubia peregrina, Rhamnus alaternus, Thymelaea tinctoria, Artostaphylos uva-ursi, Quercus faginea, and Q. rotundifolia; these last three taxa establish the more humid and colder character of the territory. Based on the biogeographical, ecological, and floristic differences, we propose as edaphoxerophilous the association Bupleuro rigidi-Pinetum halepensis (Braun-Blanquet & Bolos 1957) ass. nova (Table A2, rel. BrPh10 to BrPh52, holotypus rel. BrPh15). This association is characterised bioclimatically by its development in environments with a mean Io of 2.11 (σ = 0.43), a mean Ic of 18.55 (σ = 0.74), and a mean Itc of 262 (σ = 20.85). The bioclimatic differences between this association and the rest of the associations can be seen in Table S2 (Supplementary Material).

Ass. Ceratonio siliquae-Pinetum halepensis
The pinewoods of Pistacio-Pinetum halepensis have been described for the semiarid-dry (Io mean = 2.05) lower thermo-mesomediterranean bioclimate (Itc mean = 365.06) in Valencian territories, as a climatophilous and edaphoxerophilous community acting as secondary woodlands derived from Rubio longifoliae-Quercetum rotundifoliae [18]. It is reasonable for this pine-mastic woodland, which has previously been described for the semiarid and dry ombroclimate, to act as the head of the climatophilous series in semiarid environments, but not in dry environments where it is located in rocky areas on skeletal soils or sites denuded of soil. The holm-oak woodland of Rubio-Quercetum rotundifoliae occupies deep soils in the dry ombroclimate. All the relevés used in the description of this association by Rivas-Martínez et al. [18], Table 75.6.20 of Itinera Geobotánica 18(2), page 458, together with the relevés published by Pérez Badia [36] and Molina Cantos et al. [35] and included in the association Rhamno lycioidis-Quercetum cocciferae, belong to the association Pistacio lentisci-Pinetum halepensis, whose name was proposed by Rivas-Martínez et al. [18]. Unfortunately, this name had already been used by De Marco et al. [46], whose typification corresponds to relevé 17 in Table IV, Table 75.6.20. 2011); the relevés studied are in the Setabensean sector in the Allorano Cofrentino, Huertano Valenciano Turiano, Alcoyano Dianense, Yeclano Villenensean districts (Valenciana-Provenzal-Balearic province), and occasionally in the Alicantino Murciano sector in the Murciano Almeriensean province; this association is clearly differentiated from its northern vicariant Arbuto unedonis-Pinetum halepensis due to its richness in thermophilous species such as Osyris quadripartita, Chamaerops humilis and Arisarum vulgare. This association develops from the thermo-to mesomediterranean in an upper semiarid to lower dry ombrotype, although it may extend to the dry subhumid and may therefore contain elements of the subhumid, including Arbutus unedo, Pistacia terebinthus, Buxus sempervirens, and Certonia siliqua. This type of pine forest is therefore edaphoxerophilous.

Ass. Querco cocciferae-Pinetum halepensis
The association Querco cocciferae-Pinetum halepensis described by Rivas-Martínez et al. [18], Table 75.7.20 Itinera Geobotanica 18(2), page 460, has been given for semiarid-dry (Io mean = 2.09) mesomediterranean territories (Itc mean = 302.93) and is distributed in territories in the Manchego sector; it was previously proposed by Loisel [47] in the French Provence, but not typified. This association has a more continental character (Ic mean = 17.54) than Pistacio lentisci-Pinetum halepensis (Ic mean = 15.93) from which the thermophilous elements have disappeared, and includes the pine forests of the Manchuela Conquense, with some Setabensean influence and included by Rodríguez Rojo et al. [39] in the kermes oak woodlands of Rhamno lycioidis-Quercetum cocciferae. The relevés studied by us are located in the upper semiarid-lower dry ombrotype of the Serrano Espuñense and Jumillano Hellinense district (Manchego sector), and the Allorano Cofrentino district in the Setabensean sector.
According to its authors, the association Querco cocciferae-Pinetum halepensis can act as a primary and secondary woodland, the latter obtained from the burning of the holm-oak woodland Asparago acutifolii-Quercetum rotundifoliae. It is only natural that the secondary woodland should be edaphoxerophilous and not climatophilous, because in response to the burning of the holm-oak woodland and the loss of soil, it expands and seeks to occupy the biotope of the holm-oak formation.

Ass. Junipero turbinatae-Pinetum halepensis
The association JPh Junipero turbinatae-Pinetum halepensis represents the microforests of Pinus halepensis var. ceciliae with Juniperus turbinate and has a Balearic distribution that is typical of semiarid-dry (Io mean = 2.04) thermomediterranean environments (Itc mean = 388.71) on calco-dolomitic materials. This group is floristically differentiated from the other associations, and the formations act as edaphoxerophilous in areas with a dry ombroclimate. The strong drying winds (anemogenous character) produce an excessive loss of water that does not allow any other type of climax. However, in inland areas without strong winds and with a semiarid ombroclimate, these woodlands of P. halepensis var. ceciliae must be considered climatophilous. This association was described by Rivas-Martínez et al. [18], Table 75.6.18, Itinera Geobotánica 18(2), page 449, using the type relevé of the island of Majorca and two relevés for Menorca. We have included the relevés published by Rivas-Martínez et al. [18] and Bolòs & Molinier [25] for this syntaxon. We expand the distribution of this association with our observations to the islands of Ibiza and Formentera.

Ass. Rhamno angustifoliae-Pinetum halepensis
The pine forests sampled by us and extracted from the SIVIM [41] of the Serrano Mariense, Serrano Estanciano, and Serrano Bastitano districts in the Hoyano Accitano Bastitano sector (Betic province), located in the mesomediterranean with an ombrotype ranging from the upper semiarid to the upper dry on calcareous substrates at altitudes between 900-1600 m, constitute a new association that is floristically differentiated from its neighbours Ephedro fragilis-Pinetum halepensis with a semi-arid character in the Hoyano Accitano-Bastitano sector, the Manchegan and continental Querco cocciferae-Pinetum halepensis, and Rhamno almeriensis-Pinetum halepensis, with a thermomediterranean, Gadorensean and western Almeriensean distribution. The floristic composition of these pine forests comprises P. halepensis, Pistacia lentiscus, Quercus coccifera, Rhamnus alaternus subsp parvifolia, R. infectoria, R. myrtifolia, R. lycioides and R. oleoides subsp. angustifolia. We propose the association Rhamno angustifoliae-Pinetum halepensis ass. nova (Table A3 rel. RanPh044 to RanPh045, Holotypus rel. RanPh030). This association has an edaphoxerophilous character and an optimum in the dry ombrotype (Io mean = 2.98, σ = 0.48) with some typical species of this ombroclimate, such as Pistacia terebinthus, Quercus rotundifolia, and Crataegus monogyna. The bioclimatic characterisation of these Aleppo pine groves is based on the fact that they are established on sites with a mean Io of 2.98, a mean Itc of 317.97 (σ = 64.1), a PE of 768.8 (σ = 59.24), and a mean Ic of 16.79 (σ = 1.89). Table S3 (Supplementary Material) shows the bioclimatic differences between this association and the other associations studied.
This association is dominant in the thermo-(max Itc = 434.03) and lower mesomediterranean (Itc mean = 317.97). It may occasionally extend to the upper mesomediterranean (min Itc = 238.14), where it contacts Rhamno myrtifolii-Juniperetum phoeniceae. Its ombrotype ranges from the lower dry (min Io = 1.97) and the lower subhumid (Io max = 3.74) with the average Io = 2.97 (upper dry). Table S4 (Supplementary Material) shows the bioclimatic differences between this association and the other associations studied.
Finally, Rhamno lycioidis-Pinetum halepensis (RlPh) has been described by Torres et al. [9] as Subbetic edaphoxerophilous on rocky limestone and limestone-dolomitic crests with a mesomediterranean thermoclimate (Itc mean = 264.85) and a dry-subhumid ombroclimate (Io mean = 3.97), owing to the water loss caused by their rocky nature. The I oex is therefore semiarid-dry. All the associations studied have floristic, ecological, and biogeographical differences between them; however, from the bioclimatic point of view, they range from the semiarid to the subhumid.

Discussion
We have already mentioned that in some cases, the pinewood is described as a secondary woodland derived from Quercus rotundifolia forest owing to soil loss and the expansion of the genus Pinus, as occurs with the genus Juniperus [49][50][51][52]. This situation has led us to propose the separation of the climatophilous from the edaphoxerophilous character in the diagnosis of the communities [53]. We have, therefore, recently proposed the ombroedaphoxeric index I oex = P p − e/T p × CR [54], which takes into account positive precipitation P p , positive temperature T p , residual evapotranspiration (e), and water retention capacity CR (0.25, 0.50, 0.75). The application of this index serves to differentiate the edaphoxerophilous associations from the climatophilous associations in Pinus halepensis.
The community of Ephedra fragilis and Pinus halepensis was described by Torres et al. [9] for the Guadiana Menor valley (Accitano-Baztetan sector), and raised to the association rank by Rivas-Martínez et al. [18]. Ephedro-Pinetum halepensis is described as climatophilous in semiarid environments on gypsum loams and gypsum [55][56][57]; whereas Rhamno-Pinetum halepensis has been described as edaphoxerophilous for the Subbetic sector on limestone and limestone dolomites and in dry-subhumid environments, and the association Junipero turbinatae-Pinetum halepensis as edaphoxerophilous in semiarid-dry environments on calcodolomitic materials. Arbuto-Pinetum halepensis has been described as semiarid-dry, which appears doubtful as the species Arbutus unedo and Viburnum tinus included in the table cannot thrive in dry environments without soil compensation. In other cases, the pinewood is described as a secondary woodland derived from Quercus rotundifolia woodland owing to soil loss and the expansion of the genus Pinus, as occurs with Juniperus [48][49][50]53].
Following the ecological, floristic, bioclimatic, and biogeographical criteria for the relevés from the associations published previously and the new relevés, and proposing to separate the climatophilous from the edaphoxerophilous communities due to their different ecology, flora, and catenal contacts, we maintain the previously published associations with all the relevés belonging to and extracted from the SIVIM with a predominance of P. halepensis that have been adscribed to some of the erroneously published associations, and the relevés published by authors and included in different syntaxa [26,31,35,36], and we propose four new associations. However, in his studies on the Sierra de Cazorla, Gómez Mercado [34] assigns the Aleppo pine forests with Juniperus phoenicea to the association Rhamno lycioidis-Pinetum halepensis. We use this information to make an updated proposal on the syntaxonomy and distribution of the Aleppo pine communities on the Iberian Peninsula ( Figure 3). halepensis that have been adscribed to some of the erroneously published associations, and the relevés published by authors and included in different syntaxa [26,31,[35][36], and we propose four new associations. However, in his studies on the Sierra de Cazorla, Gómez Mercado [34] assigns the Aleppo pine forests with Juniperus phoenicea to the association Rhamno lycioidis-Pinetum halepensis. We use this information to make an updated proposal on the syntaxonomy and distribution of the Aleppo pine communities on the Iberian Peninsula ( Figure 3). Several associations described by Rivas-Martínez et al. [18] have been proposed as climatophilous and edaphoxerophilous in semiarid-dry ombrotypes. When these pinewoods occur in a semiarid ombrotype, they are understood to be climatophilous, but act as edaphoxerophilous in a dry ombrotype. Although this appears to be true, there is some uncertainty, as the same syntaxon cannot have two different behaviours and maintain its floristic composition and catenal contacts.
This situation leads us to recommend the separation of the climatophilous from the edaphoxerophilous character in the diagnosis of the communities, and we have therefore recently proposed the ombroedaphoxeric index I oex , which considers positive temperature T p , residual evapotranspiration (e), and water retention capacity (CR) (0.25, 0.50, 0.75). Pinus halepensis has its bioclimatic optimum in semi-arid environments, where it acts as a climatophile, whereas in dry and lower sub-humid environments it can occupy rocky environments, where it acts as an edaphoxerophile; this behaviour is explained based on the ombroedaphoxeric index [54].
Rivas-Martínez et al. [18] includes several associations of Pinus halepensis in the alliances Oleo-Ceratonion and Rhamno-Quercion cocciferae within the order Pistacio lentisci-Rhamnetalia alaterni. Biondi et al. [58] later describe the order Pinetalia halepensis with a thermo-and mesomediterranean character and the alliance Pistacio lentisci-Pinion halepensis. These authors report that these forests are dominated by Pinus halepensis and have a centraleastern Mediterranean distribution. This alliance is also again present in the north of Algeria, according to Rachid Meddour et al. [59]. Pesaresi et al. [19] have recently proposed several alliances for this order: Rosmarino officinalis-Pinion halepensis for the territories of Corsica, Sardinia, Sicily, and the Italo-Tyrrhenian coasts, and Sarcopoterio sipinosi-Pinion halepensis in Greek territories. Alkanno baeoticae-Pinion halepensis has been proposed by Mucina et al. [60] for serpentines on the Greek island of Euboea. The alliance Thymo vulgaris-Pinion halepensis has been described for the northern areas of the Italo-Tyrrhenian province and the Occitanian-Provencal sector in the Catalonian-Balearic-Provencal province, in the upper subhumid mesomediterranean thermotype. Its authors have proposed the association Cisto albidi-Pinetum halepensis as a holotypus for this alliance, with a subacidophilous mesomediterranean character in the Maritime Alps in Liguria.
The Iberian forests of P. halepensis have a different floristic and biogeographical composition from the forests described previously; however, they are similar from the physiognomical and ecological points of view. In our opinion, Pinetalia halepensis described for the central-eastern Mediterranean is also present in the western Mediterranean, although with differences between the eastern and western Mediterranean. We cannot include the associations analysed on the Iberian Peninsula in any of the alliances described for the order Pinetalia halepensis due to their major biogeographical, ecological, and floristic differences. We therefore propose the new alliance Rhamno lycioidis-Pinion halepensis all. nova hoc loco. The order Pinetalia halepensis on the Iberian Peninsula occupies semiarid bioclimatic environments where the forests have a climatophilous character, whereas it assumes an edaphoxerophilous character in dry and subhumid bioclimatic areas.
Holotypus  (Table 3). Finally, a synthetic table was prepared with all the associations studied (Table 4).

Characteristic Species AuPh CsPh JtPh BrPh EnPh RlaPh RaPh RlPh EfPh QcPh RanPh
Pinus halepensis Mill.   [19] and state that its distribution agrees with the Mesomediterranean basophilic scrub of Rosmarinetalia. Evidently, the area occupied by this order is much larger than that occupied by this alliance, so their distribution areas do not agree. They also state that the distribution of this alliance partially overlaps that of Pistacio lentisci-Pinion halepenis-which is synonymized by Rosmarino officinalis-Pinion halepensis-based on a larger dataset than used by the authors in the description. We cannot agree with this, as it does not consider the phytosociological nomenclature code (Def X and VI) [41]. The correct name is Rosmarino officinalis-Pinion halepensis, an alliance described for the biogeographical territories of the Italo-Tyrrhenian province and the easternmost Occitanian-Provencal sector of the Catalonian-Balearic-Provencal province, which according to these authors ranges from the upper thermo-to the upper meso thermotype.
According to Bonari et al. [63], the alliance Pistacio lentisci-Pinion halepensis is thermo-Mediterranean and is distributed from continental Greece to eastern Spain, and probably also in some areas of northwestern Africa, thus increasing the distribution area of this alliance. This does not coincide with the distribution given by these authors in their map (Figure 3, p. 10), and they do not include any Iberian association in this alliance.
However, the authors describe this alliance for the central-eastern Mediterranean, with an upper thermo-to upper meso thermotype. The original description was based on diagnostic species belonging to other phytosociological classes and is maintained by Bonari et al.; these species have very different ecological niches from that of Pinus halepensis and the taxa generally belong to Quercetea ilicis and Rosmarinetea officinalis [18].
A comprehensive debate is therefore required because of this controversy. For our part, the floristic, biogeographical, and bioclimatic differences allow us to propose a new alliance, which due to the large number of floristic elements in Quercetea ilicis, we prefer to keep in the order Pinetalia halepensis, class Quercetea ilicis.

Conclusions
This research confirms the autochthonous character of Pinus halepensis in Spain, and by accepting the associations already described and the new natural associations, we open the door to future repopulations in the bioclimatic environments typical of this species.
The analysis of the 11 associations in the study reveals certain similarities with common floristic elements. Although many associations have been described as climatophilous and edaphoxerophilous, if, in addition to the ombroclimatic factor, the soil factor is considered and the ombroedaphoxeric index is applied, sites with dry-subhumid ombroclimate index values that allow the development of Quercus rotundifolia woodlands cannot be considered as biotopes for P. halepensis unless they are located on rocks where the water loss prevents the holm oak woodland from thriving. This will then lead to the installation of the pinewood, which can be considered edaphoxerophilous. However, P. halepensis can thrive when the value of the ombroclimatic index is semiarid, and Q. rotundifolia is unable to do so when it forms a climatophilous microforest. As a result of the diagnosis of the communities, we propose to separate the climatophilous from the edaphoxerophilous communities.