Taxonomy, Ecology and Distribution of Juniperus oxycedrus L. Group in the Mediterranean Basin Using Bioclimatic, Phytochemical and Morphometric Approaches, with Special Reference to the Iberian Peninsula

: Several studies have been conducted in the past to clarify various aspects of species in the genus Juniperus L. One critical group is Juniperus oxycedrus L., especially from the taxonomical point of view. For this reason, we have studied the ecology, taxonomy and distribution of the taxa in the J. oxycedrus group. From an ecological and distribution standpoint, in this work we use the ombroedaphoxeric index (Ioex) to explain the presence of Juniperus populations in ombrotypes that are not optimum for these taxa. The controversy over the taxonomy of J. oxycedrus subsp. badia (H. Gay) Debeaux and J. oxycedrus subsp. lagunae (Pau ex C. Vicioso) Rivas Mart. is clariﬁed, and it is accepted as a valid name, J. oxycedrus subsp. badia . The phytochemical differences in essential oils (EO) are addressed and their similarities analyzed; greater similarities are observed between oxycedrus and badia , and between navicularis Gand. and macrocarpa (Sm.) Ball. species. The phytochemical, molecular and distribution differences allow J. oxycedrus subsp. macrocarpa (Sm.) Ball and J. navicularis Gand. to be maintained as species. The results obtained make it possible to establish the rank to which the taxa belong and allow clear discrimination between species in groups that are difﬁcult to interpret. Ecological, bioclimatic, phytochemical and morphometric similarities allow us to subordinate the subsp. macrocarpa to the species J. navicularis .


Introduction
In this work, we update the state of knowledge of the ecology, taxonomy and distribution of the taxa of the Juniperus oxycedrus group, based on our study and on research by various authors which has, occasionally, been a source of controversy.
Juniperus oxycedrus L. has its distribution in the Mediterranean region, from eastern Portugal to northern Iran, including Spain, the Balearic Islands, France, Corsica, Italy, Sicily, Sardinia, Northern Africa [1], Albania, Croatia, Montenegro, Serbia, Northern Macedonia, Bulgaria, Greece, Crete and Karpathos, Cyprus, Turkey, Israel and Jordan, the Crimea and Lebanon [2,3]. According to Amaral Franco [2], this species has three clearly differentiated subspecies for the Iberian Peninsula: J. oxycedrus L. subsp. oxycedrus, J. oxycedrus subsp. badia (H. Gay) Debeaux and J. oxycedrus subsp. macrocarpa (Sm.) Ball. The subspecies This study, in agreement with others [9,[17][18][19][20], also confirms that in order to effectively discriminate among species in groups that are difficult to interpret, morphometric, phytochemical and bioclimatic approaches are very useful to clarify the rank to which the taxa belong. The bioclimatic approach is generally also very important for biodiversity conservation, according to [21][22][23][24][25][26][27]. As a hypothesis, we maintain that there is a close relation between bioclimatology and ecology and the morphology and phytochemistry of the taxon, and hence with the plant community to which the taxon belongs. The morphometric analyses of leaves, galbuli and essential oils reveal no more than genetic expressions, which justifies their use in taxonomy, as we maintained previously with the fractal analysis in Quercus leaves [19].
The aim of this work is to clarify the taxonomy, ecology and distribution of the taxa in the J. oxycedrus group, which form the communities included in the Habitats Directive 92/43 EEC such as the "Arborescent matorral with Juniperus spp." (5210), in the western Mediterranean.

Sampling Design
We studied the J. oxycedrus group and collected specimen and phytosociological samples following previous studies [28,29], inventories of ours and [12,30]. We have used 288 own samples of phytosociological and morphometric character, for the study of leaves and fruits. The work was carried out in the southern half of the Iberian peninsula, to determine the habitat in which these taxa occur. We examined the differences in the ecology, distribution and taxonomy of the taxa in the J. oxycedrus group by analyzing their morphological, ecological and phytochemical differences. With these characters, we studied the resemblance between the taxa using a similarity analysis based on the data from Spain (sierra Pandera) [31], Greece and Turkey [32][33][34] and from SW Portugal (Figure 2), with This study, in agreement with others [9,[17][18][19][20], also confirms that in order to effectively discriminate among species in groups that are difficult to interpret, morphometric, phytochemical and bioclimatic approaches are very useful to clarify the rank to which the taxa belong. The bioclimatic approach is generally also very important for biodiversity conservation, according to [21][22][23][24][25][26][27]. As a hypothesis, we maintain that there is a close relation between bioclimatology and ecology and the morphology and phytochemistry of the taxon, and hence with the plant community to which the taxon belongs. The morphometric analyses of leaves, galbuli and essential oils reveal no more than genetic expressions, which justifies their use in taxonomy, as we maintained previously with the fractal analysis in Quercus leaves [19].
The aim of this work is to clarify the taxonomy, ecology and distribution of the taxa in the J. oxycedrus group, which form the communities included in the Habitats Directive 92/43 EEC such as the "Arborescent matorral with Juniperus spp." (5210), in the western Mediterranean.

Sampling Design
We studied the J. oxycedrus group and collected specimen and phytosociological samples following previous studies [28,29], inventories of ours and [12,30]. We have used 288 own samples of phytosociological and morphometric character, for the study of leaves and fruits. The work was carried out in the southern half of the Iberian peninsula, to determine the habitat in which these taxa occur. We examined the differences in the ecology, distribution and taxonomy of the taxa in the J. oxycedrus group by analyzing their morphological, ecological and phytochemical differences. With these characters, we studied the resemblance between the taxa using a similarity analysis based on the data from Spain (sierra Pandera) [31], Greece and Turkey [32][33][34] and from SW Portugal ( Figure 2), with data extracted from the works of these authors. They used leaves and fruits as analysis material, of the studied localities, and a bioclimatic analysis of said places. data extracted from the works of these authors. They used leaves and fruits as analysis material, of the studied localities, and a bioclimatic analysis of said places.

Bioclimatic Analysis
A bioclimatic study was carried out to explain the presence of Juniperus species in different rocky substrates. To understand the presence of communities of Juniperus in sites dominated by species in the genus Quercus, we applied Thornthwaite's index, ETPmonthly = 16(10.T/I) a , to calculate the potential evapotranspiration and the residual evapotranspiration (e) as e = 0.2ETP by [35]. With these data, we used the ombroedaphoxeric index (Ioex) proposed by [12], for this we have used 9 weather stations, in a comparative analysis with the ombro-climatic index Io proposed by [36], which justifies the presence of micro-woodlands of Juniperus and Pinus.
The ombro-thermic index and ombro-climatic index conceptually are synonymous. The word "ombrotype" expresses a category related to the precipitation rate. These three terms are related to the precipitation factor (climate). However, in the calculation of the ombroedaphoxeric index, a second parameter is already involved; soil.
We used phytosociological and bioclimatic studies from different localities (Almadén-Minas, Cabezas Rubias, Aracena, Santiago Pontones, Vadillo Castril, Grazalema, Montoro, Pozoblanco, Villanueva del Arzobispo) [37] to establish the ecology and distribution in Spanish places. For the rest of the Mediterranean, we use data from different authors [12,20,36,38] and the inclusion of Juniperus communities in plant associations and in their corresponding habitats [12], as all phytosociological studies specify the ecology and distribution of taxa and plant communities.

Bioclimatic Analysis
A bioclimatic study was carried out to explain the presence of Juniperus species in different rocky substrates. To understand the presence of communities of Juniperus in sites dominated by species in the genus Quercus, we applied Thornthwaite's index, ETP monthly = 16(10.T/I) a , to calculate the potential evapotranspiration and the residual evapotranspiration (e) as e = 0.2ETP by [35]. With these data, we used the ombroedaphoxeric index (Ioex) proposed by [12], for this we have used 9 weather stations, in a comparative analysis with the ombro-climatic index I o proposed by [36], which justifies the presence of micro-woodlands of Juniperus and Pinus.
The ombro-thermic index and ombro-climatic index conceptually are synonymous. The word "ombrotype" expresses a category related to the precipitation rate. These three terms are related to the precipitation factor (climate). However, in the calculation of the ombroedaphoxeric index, a second parameter is already involved; soil.
We used phytosociological and bioclimatic studies from different localities (Almadén-Minas, Cabezas Rubias, Aracena, Santiago Pontones, Vadillo Castril, Grazalema, Montoro, Pozoblanco, Villanueva del Arzobispo) [37] to establish the ecology and distribution in Spanish places. For the rest of the Mediterranean, we use data from different authors [12,20,36,38] and the inclusion of Juniperus communities in plant associations and in their corresponding habitats [12], as all phytosociological studies specify the ecology and distribution of taxa and plant communities.
Sites behave differently in response to the general climate, the type of substrate and the topography of the terrain. For this reason, areas on rocky crests-although they may be located in rainy environments and surrounded by climactic forests-behave differently from the areas around them. In these circumstances, islands evolve which may contain edaphoseries, minoriseries and permaseries [39]. All plant communities growing on rocky crests and steeply sloping areas with extreme gradients, among others, are very significantly influenced by the soil, which conditions their existence. All places have a particular type of substrate and an orography that determines whether they have a greater or lesser capacity to retain water. In ideal situations, with good soil texture and structure and with no slopes, the water retention capacity (RC) can be assumed to be at maximum (100%). Otherwise, losses occur due to runoff and drainage, causing the RC to vary. Water is also lost through evapotranspiration (ETP). However, as plants have the capacity to self-regulate their losses, it can be accepted that the residual evapotranspiration e = 0.2ETP, the value of e marks the ability of plants to withstand drought. There are therefore two parameters (e and RC) implicated in the development of a vegetation, which is essentially conditioned by rainfall. The ombro-climatic index I o, therefore, does not explain the presence of plant communities influenced by the substrate and, for this reason, we use the ombroedaphoxeric index (Ioex) [12] to explain the presence of communities of Juniperus and Pinus for territories with a thermotype ranging from the thermo-to the supra-Mediterranean.
The Ioex index is obtained by subtracting the residual evapotranspiration (e) from the positive precipitation (P p ), the value obtained is divided by the positive temperature (T p ) and the result is multiplied by the value taken by RC (0.25, 0.50, 0.75); with the value obtained, the Rivas-Martínez table [40] is applied to obtain the ombric horizons: where P p = Positive precipitation (mm); T p = Positive temperature ( • C) [24]; e = residual evapotranspiration whose value is 0.2 ETP [36]; RC = retention capacity in parts per unit, whose values may be 0.25, 0.50, 0.75 and 1; Ioex is the summer ombro-thermal index, calculating Ioex1, Ioex2 and Ioex3 (in function of CR = 0.25, 0.5, and 0,75, respectively); the most representative is Ioex2, which marks the arid character of the locality.

Phytochemical Analysis
For the phytochemical data, we follow the works of [31] and [32] where they describe the material used from Greece and Spain (Jaén herbarium). An Excel table was prepared with the phytochemical data for each taxon and the similarity dendrogram Jaccard was obtained to determine the degree of relation between the taxa. Pearson is used for correlation analysis. The 19 essential oils used are as follows: alpha-pinene, limonene, beta-pinene, myrcene, p-cymene, beta-phellandrene, manoyl oxide, germacrene D, alpha-campholenal, beta-bourbenene, Sesquiterpenes /only for 1 species, gamma-terpinene, terpinen 4-ol, alpha-phellandrene, limonene + beta-phellandrene, alpha-terpinene, terpinolene, alphaterpinol, Sesquiterpenes only for 1 species.
The number of samples used corresponds to the phytochemical studies of Salido et al. and Adams [32,41]. For statistics, the software PAlaeontological STatistics (PAST) and Community Analysis Package III were used.

Morphometric Analysis
For the morphometric analysis, we used leaf length and width and galbuli size as characters. For this more than 100 measurements in millimeters (mm) were made for each of the taxa, either from herbarium material or from our own herbalizations, by consulting and reviewing the material from the following herbaria: Herbarium Jaén, Herbarium Algarbe and the Extremadura Centre for Technological Scientific Research (HSS herbarium) (see Supplementary Material), and bibliographic references taken from the Anthos database [42], Flora Ibérica [43], and the database of the flora of Alicante [44]. For the authorship of the taxa we have followed [4,[45][46][47][48][49][50][51]:  Table 1 shows the values of Pp (Positive Precipitation), Tp (Positive Temperature) and Io (Ombro-climatic Index) according to the criterion established by Rivas-Martínez. Thornthwaite's formula, ETP monthly = 16(10.T/I) a , is applied to obtain the value of ETP (potential evapotranspiration), where T is the mean monthly temperature, I is the annual heat index and a parameter that depends on the values taken by I. [12].

Bioclimatic Analysis
Applying the formula I oex for the assumptions that RC is 0.25, 0.50, 0.75, we obtain three values of which the most representative is I oex2 . Table 1 shows the equivalence values so that although the territorial bioclimate allows the existence of climactic forests in wild areas with RC = 50%, the humid ombrotype becomes dry or subhumid depending on whether the value of RC = 25% or 50%. The subhumid becomes dry and the dry becomes semiarid or arid. Therefore, in areas with Io > 8 we obtain I oex2 values of 3.86 and 4.94, which is the equivalent to subhumid. This explains the fact that, in rocky areas, there is an edaphoxerophilous community of Quercus faginea s.l. or Abies pinsapo Boiss., as occurs in Grazalema (Cadiz); or that a value of I oex1 = 2.47 is obtained in the case that RC = 25%. In this situation, there is a presence of an edaphoxerophilous community of Quercus rotundifolia in Cazorla (Jaén) and in Grazalema (Cadiz). When the underlying ombrotype is subhumid, the equivalence value of I oex2 corresponds to dry; and a situation that initially has a dry I o gives semiarid and even arid values of I oex2 , if we start from a horizon that is lower than dry. This does not allow the development of Quercus tree species but does allow the development of the Juniperus and Pinus genera. The value of I oex is affected by climate change, as evidenced by Del Río et al. [38], who report a heterogeneous trend in terms of annual rainfall redistribution, with a decline in most of the mountainous areas of Grazalema, Ronda, Cazorla, Segura, Sierra Nevada and a large part of the Sierra Morena. However, these authors have detected an increase in rainfall on the Andalusian coast and particularly in Almería. This affects the forest stands and, in conjunction with human activity [50], favors a redistribution of the current forests with a decline in Quercus woodlands and an increase in the micro-woodlands of Juniperus.  Table 2 shows that the alpha-pinene is common to the four taxa and p-cymene to J. oxycedrus subsp. oxycedrus, J. oxycedrus subsp. macrocarpa and J. navicularis (Table 2). This is justification for including them all in the J. oxycedrus group. Moreover, there are certain morphological and ecological similarities between J. oxycedrus subsp. oxycedrus and J. oxycedrus subsp. badia and between J. oxycedrus subsp. macrocarpa and J. navicularis, according to the presence of certain oils. The component myrcene is exclusive to the subspecies oxycedrus, whereas the exclusive compounds in the subspecies badia are: germacrene D, alpha-camphelenal, beta bourbenene and nine sesquiterpenes. The subspecies macrocarpa does not have any of its own essential oils as it shares gamma-terpinene and terpinen 4-ol with J. navicularis, while the exclusive chemical compounds in J. navicularis are alphaphellandrene, alpha-terpinene, terpinolene, alpha terpinol and ten sesquiterpenes. The species J. navicularis presents different sesquiterpenes from those of J. oxycedrus subsp. badia.  As specified by [31] and [32], there is a clear phytochemical differentiation between the three subspecies of J. oxycedrus (J. oxycedrus subsp. oxycedrus, subsp. badia and subsp. macrocarpa). Subsequently [8] clarifies the phytochemical differences between J. oxycedrus and J. deltoides, a species described by this author [7] in Greek territories. This taxon has a low content in alpha-pinene and a high content in limonene, which, among other morphological differences, justifies the rank of species for this taxon. There are also substantial differences in the essential oils of J. navicularis Gand. compared to the rest [33]. The Pearson correlation analysis ( Table 3) gives values of r = 1 or near 1 for group G2 (J. oxycedrus subsp. macrocarpa and J. navicularis). In the case of group G1 (J. oxycedrus subsp. oxycedrus and J. oxycedrus subsp. badia) the values are close or equal to 1; however, the relationship between both groups is low owing to the phytochemical differences between them. In Jaccard's similarity analysis in Figure 3, two clearly differentiated groups are established (G1 and G2), which is due to the concentrations and different chemical components they present. macrocarpa). Subsequently [8] clarifies the phytochemical differences between J. oxycedrus and J. deltoides, a species described by this author [7] in Greek territories. This taxon has a low content in alpha-pinene and a high content in limonene, which, among other morphological differences, justifies the rank of species for this taxon. There are also substantial differences in the essential oils of J. navicularis Gand. compared to the rest [33].

Phytochemical Analysis
The Pearson correlation analysis ( Table 3) gives values of r = 1 or near 1 for group G2 (J. oxycedrus subsp. macrocarpa and J. navicularis). In the case of group G1 (J. oxycedrus subsp. oxycedrus and J. oxycedrus subsp. badia) the values are close or equal to 1; however, the relationship between both groups is low owing to the phytochemical differences between them. In Jaccard's similarity analysis in Figure 3, two clearly differentiated groups are established (G1 and G2), which is due to the concentrations and different chemical components they present.

Morphometric Analysis
The main differences between J. oxycedrus subsp. oxycedrus and J. oxycedrus subsp. badia are based on their physiognomy and the size of their mature fruits [2]. Whereas the subspecies oxycedrus tends to occur as a shrub, the subspecies badia is a tree of considerable height and with a pyramidal form. The size of the mature galbuli in the subspecies oxycedrus does not generally exceed 1 cm, while in the subspecies badia it is over 1 cm. The leaves of subspecies oxycedrus have a width of 1-1.5 mm while those of subsp. badia are 1.2-2 mm. Coincidentally these subspecies are frequently found coexisting in similar biotopes, which has led to frequent confusion by several authors. The subsp. macrocarpa has galbuli of 1.2-1.5 cm with a purplish chestnut color when ripe; this is an erect tree, rarely prostrate, growing up to 3 m. In contrast J. navicularis has a similar ecology to subsp. macrocarpa but grows to a maximum height of 2 m; its leaves are 1-1.5 mm wide and its galbuli are between 0.7-1 cm (Table 4). Linnaeus did not describe J. oxycedrus in the first part of his work Species Plantarum, but merely noted the existing synonym [52]. Theoretically, the authorship of J. oxycedrus L. should be attributed to Clusius, who described it: "Nusquam autem majorem vidisse memini, quam supra Segoviam et Guadarrama, itinere Madritiano, ubi magnarum arborum interdum altitudinem". He also says of the fruit "fructum initio viridem... postremo, cum maturuit punicea colorís, Juniperi fructu multo majorem ut interdum avellanam magnitudine aequet" [53]. This description is specifically included in the work of [45]. Clusius' description, in which he states literally that it has a "thick pruinose red fruit", corresponds to Juniperus oxycedrus L. subsp. badia (H. Gay) Debeaux [Basion. Juniperus oxycedrus var. badia H. Gay in Assoc. Franç. Avancem. Sci. Compt. Rend. 1889:501 (1889)], included in [2]. The lack of typification of the specimen described by Clusius is justification for these facts and supports the valid acceptance of Juniperus oxycedrus subsp. badia (ICN article 41). Due to the high number of morphological and ecological differences and differences in essential oils between subsp. macrocarpa, subsp. oxycedrus and subsp. badia, and to its similarity with J. navicularis, the subspecies macrocarpa can be proposed to be subordinated to J. navicularis as J. navicularis Gand. subsp. macrocarpa (Sm.) Ball; however, the molecular study by [41] establishes major molecular differences between the four taxa J. oxycedrus subsp. oxycedrus, J. oxycedrus subsp. badia, J. oxycedrus subsp. macrocarpa and J. navicularis, and based on this he raises them to the rank of species. The phytochemical differences between groups G1 and G2 are very clear, so the species rank could be accepted for the four taxa. However, in the case of their distribution, the populations of J. oxycedrus subsp. macrocarpa and J. navicularis are separated despite having similar ecologies, while this is not the case of J. oxycedrus subsp. oxycedrus and subsp. badia. Therefore, it is logical to accept J. macrocarpa and J. navicularis as species and to maintain J. oxycedrus subsp. oxycedrus and J. oxycedrus L. subsp. badia (H. Gay) Debeaux with the rank of subspecies.
There are also some doubts about the presence of J. oxycedrus subsp. badia in the African continent. Some authors such as [47] do not recognize this taxon in northern Africa, although [54] [56,57], a new combination of Juniperus oxycedrus subsp. lagunae was formulated. This all serves to highlight the complexity of this taxon, whose distribution area is insufficiently known. However, its presence is very evident in the center and south of the Iberian Peninsula, where it grows in formations with a broad extension, generally on scree and biotopes with shallow soils. In these habitats, holm oaks (Quercus rotundifolia) cease to be dominant or simply cannot exist due to the absence of the necessary ecological and/or soil conditions for these taxa to develop [30] (Figure 4). oxycedrus subsp. oxycedrus and subsp. badia. Therefore, it is logical to accept J. macrocarpa and J. navicularis as species and to maintain J. oxycedrus subsp. oxycedrus and J. oxycedrus L. subsp. badia (H. Gay) Debeaux with the rank of subspecies.
There are also some doubts about the presence of J. oxycedrus subsp. badia in the African continent. Some authors such as [47] do not recognize this taxon in northern Africa, although [54] [56,57], a new combination of Juniperus oxycedrus subsp. lagunae was formulated. This all serves to highlight the complexity of this taxon, whose distribution area is insufficiently known. However, its presence is very evident in the center and south of the Iberian Peninsula, where it grows in formations with a broad extension, generally on scree and biotopes with shallow soils. In these habitats, holm oaks (Quercus rotundifolia) cease to be dominant or simply cannot exist due to the absence of the necessary ecological and/or soil conditions for these taxa to develop [30] (Figure 4). The areas dominated by the species of juniper are currently undergoing a process of expansion due to the increase in areas of scree and rocky areas in general. This phenomenon causes an increase in dry areas and a decrease in rainy areas, thus expanding the potential areas that can act as a refuge for endemic species [58]. A good approach for plan- The areas dominated by the species of juniper are currently undergoing a process of expansion due to the increase in areas of scree and rocky areas in general. This phenomenon causes an increase in dry areas and a decrease in rainy areas, thus expanding the potential areas that can act as a refuge for endemic species [58]. A good approach for planning the re-naturalization of habitats with Juniperus species is to use autochthonous species after fire or other environmental events, which is also effective for preserving the local genetic resources and avoiding genetic pollution [26,59]. Another good approach is phytotoponym analysis in order to gain a better understanding of their potential areas of growth, as proposed by [60].
In regard to the bioclimatology of the Juniperus oxycedrus group, its thermotype ranges between the thermo-and meso-Mediterranean, occasionally reaching the supra-Mediterranean; the ombrotype is between arid and dry and, in rocky areas, may reach the subhumid. These special bioclimatic and ecological characteristics account for the distribution of these Juniperus taxa, and allow a correlation to be established with their phytochemistry, which explains the differences between the type and concentration of their essential oils.
The bioclimatic and ecological differences between the taxa can be seen from their dissimilar morphological and phytochemical features. They differ significantly in the length and width of the leaves and galbuli, and even more so in their phytochemistry. The similarity analysis connects J. oxycedrus subsp. macrocarpa with J. navicularis, two taxa that effectively have their ecological optimum in sandy thermo-Mediterranean areas; whereas J. oxycedrus subps. oxycedrus and J. oxycedrus subsp. badia show a greater degree of similarity and have their optimum in more continental meso-Mediterranean environments.
Vasic and Dubak [1] discuss the anatomical characters (epidermis) of J. oxycedrus at various altitudes in the mountains in Serbia. Although these authors report that the morphological characters may vary somewhat, this is not the case as differences in the essential oils have been observed in the present study. In studies on the biogeography and genetic relations of J. oxycedrus in the Mediterranean, including Macaronesian regions, Boratyński et al. [61] demonstrate the close relation between J. navicularis, J. oxycedrus subsp. macrocarpa and J. brevifolia (Seub.) Antoine de Azores, and describe in detail the genetic separation between Juniperus from the eastern Mediterranean and Juniperus from the western Mediterranean-which currently includes Macaronesia. As in our study, their morphological analysis confirmed the separation between J. navicularis and J. oxycedrus

Conclusions
The similarity study based on the quantity, type of essential oils, and different bioclimatic distribution could justify maintaining the rank of species for J. macrocarpa Sm. and J. navicularis Gand. Instead, the distribution and co-existence of the two remaining taxa, in spite of their phytochemical differences, do not warrant the rank of species and we thus maintain their rank as subspecies: J. oxycedrus L. subsp. oxycedrus and J. oxycedrus L. subsp. badia (H. Gay) Debeaux. Due to the high number of morphological and ecological differences and differences in essential oils between subsp. macrocarpa, subsp. oxycedrus and subsp. badia, and to the similarity of subsp. macrocarpa with J. navicularis, this subspecies can be proposed to be subordinated to J. navicularis as J. navicularis Gand. subsp. macrocarpa (Sm.) Ball. Being of special reference for the differentiation between the types of junipers, the morphometric character of leaves and galbulas, types and content in essential oils, and the exclusive thermo-Mediterranean character of J. navicularis and J. navicularis Gand. subsp. macrocarpa.