The biogeographical regionalization of the study area was defined within Geographical Information Systems (GIS) framework using ArcMap 10.1 software (ESRI 2012), where a vector file of the biogeographical regionalization of the study area was created by converting the already published map by Good (1974) into a polygon vector shapefile to enhance the quality of future analyses. The process was performed in 3 steps: 1) the original map of Good (1974) was scanned and converted into TIFF format, then it was georeferenced using a previously georeferenced shapefile; 2) polygons defining the boundaries and the geographic locations of the sub-regions across the study area were created; and 3) attributes and information related to each of the defined provinces were added and linked to the shapefile (Fig. 1).

Delimitation of the Mediterranean region in this study was mainly assessed according to the system of Good (1974). It is necessary to keep in mind that regions are separated from each other not by lines, but by belts inhabited by taxa belonging to the contiguous regions. According to Vargas (2020), there is principally a consensus for recognizing five vegetation belts in the Mediterranean region: (1) thermo-Mediterranean vegetation belt (coastal scrub: maquis/garrigue/phrygana) and conifer-oak woodlands on inland hills; (2) meso-Mediterranean (sclerophyll woodland on plains and piedmont); (3) supra-Mediterranean (oak and pine woodlands up to the altitudinal timberline); (4) oro-Mediterranean (high-altitude bush communities); and (5) cryoro-Mediterranean (alpine communities of herbs). In fact, only the thermo-Mediterranean vegetation belt is present in Egypt.

The chorionomic boundary of the Mediterranean region coincides with the boundary of the typical Mediterranean forests of Quercus ilex L. and the primary area of Olea europaea L. subsp. europaea cultivation. The Mediterranean region can be subdivided into: North-western sub-region, North African sub-region, Eastern sub-region (levant) and Balkans sub-region (Fig. 1).

Regarding Egypt, the Mediterranean territory extends for about 970 km from Sallum, on the Egyptian-Libyan border, to Rafah, on the Egyptian-Palestinian border, with an average width 15-20 km in north-south direction and a limited area of ca. 16,500 km². The Mediterranean phytogeographical region was divided into three main sub-sectors: 1) western a.k.a. the Mareotis, with mean annual rainfall between 220-150 mm and stretching for 550 km between Sallum and Alexandria; 2) middle a.k.a. Deltaic, stretching for 180 km between Alexandria and Port Said, with width of about 12 km; and 3) east a.k.a. Sinaitic, with an annual rainfall between 150-100 mm and extending for 220 km between Port Said and Rafah (Zahran & al. 1985; Zahran & Willis 2008).

A preliminary list of Mediterranean endemics in Egypt was prepared based on Täckholm (1974), Greuter & al. (1984-1989), Boulos (1999- 2005), Greuter & von Raab-Straube (2008), Ahmed (2009), Shaltout & al. (2010), and El-Khalafy (2018).

Authentication of the taxa was assessed depending on floras of countries of Mediterranean basin such as Quézel & Santa (1962-1963), Tutin & al. (1964-1980), Davis (1965-1985), Zohary (1966, 1987), Mouterde (1966-1984), Beck-Mannagetta (1967- 1983), Franco (1971-1983), Zangheri (1976), Haslam & al. (1977), Feinbrun-Dothan (1978, 1986), Jafri & El-Gadi (1977-1988), Duvigneaud (1979), Pottier-Alapetite (1979-1981), Meikle (1977-1985), Guinochet & Vilmorin (1973-1984), Smythies (1984-1986), Castroviejo & al. (1986-2021), Davis & al. (1988), and Güner & al. (2000).

The following databases were also consulted to check the recorded plants: Hassler (2004-2022), Flowers in Israel (2005-2022), Euro+Med (2006), Danin & Fragman-Sapir (2016), Med-Checklist (2016), Dimopoulos & al. (2020), eflora Maghreb (2021), African Plant Database (2022), Chikhali (2022), EPPO (2022), Flora of Cyprus (2022), Flora of Libya (2022), Flora of Turkey (2022), GBIF (2022), JSTOR (2022), POWO (2022), VicFlora (2022), and WCSP (2022). Data were also collected from field visits and collection research including the following herbaria: Real Jardín Botánico, Madrid (MA), Tanta University (TANE), Alexandria University (ALEX), Cairo University (CAI), Assiut University (ASTU), Agricultural Research Center (CAIM), Desert Research Center (CAIH), National Research Centre (CAIRC), and National Registry for Egyptian Herbaria (2022). Information was also compiled from available literature (papers, books, M.Sc. and Ph.D. theses, and scientific reports).

The recorded taxa were arranged alphabetically according to Angiosperm Phylogeny Group system (APG IV: Byng & al. 2016). The accepted names followed International Plant Names Index (IPNI 2022), World Flora Online (WFO 2022), World Checklist of Selected Plant Families (WCSP 2022), and Plants of the World Online (POWO 2022).

General characteristics of the recorded plants such as habitats, geographical distribution, growth and life forms, sex forms, flowering period and dispersal types were determined from the following publications: Täckholm & Drar (1950-1969), Zohary (1966, 1987), Täckholm (1974), Jafri & El-Gadi (1977-1988), Feinbrun-Dothan (1978, 1986), Boulos (1999-2005, 2009), El-Hadidi & Hosni (2000), Seif El-Nasr & Bidak (2006a, 2006b), Ahmed (2009), Shaltout & al. (2010), El-Khalafy (2018), Bedair (2020), Mushtaq & al. (2020), Bedair & al. (2020), Ghosh & al. (2021), Bedair & al. (2022 a & b), Ghosh & al. (2022), Abdelsalam & al. (2022) and Shaltout & Bedair (2022). The dispersal type of the collected diaspores of many of the recorded taxa was assessed according to Dansereau & Lems (1957) (Table 1). The main assigned sex forms were bisexual (i.e. hermaphrodite) having both male and female sex organs in the flower, or unisexual, either monoecious (each plant has both male and female flowers) or dioecious (the male and female blossoms appear on separate plants), and polygamous.

Life forms follow the system of Raunkiaer (1937). This system which is designed for the northern temperate zone and based on the location of renewal buds was coded and identified as follows:

Phytogeographical sectors in this study were assigned according to Zahran & al. (1985). Besides, the global distribution of the recorded taxa (i.e. floristic regions) was assigned following the system of Good (1974), which divided the globe into six kingdoms, three subkingdoms and thirty nine floristic regions. Distribution of the recorded taxa in countries of Mediterranean was also checked in some databases such as Hassler (2004-2022), Euro+Med (2006), Med-Checklist (2016), Flora of Libya (2022), POWO (2022), WCSP (2022), GBIF (2022), and JSTOR (2022).

Rarity forms of the recorded taxa were assessed depending on the national geographical range, habitat specificity and local abundance according to Rabinowitz (1981). The eight rarity forms are coded and explained in Table 2.

A preliminary list of 275 Mediterranean endemic taxa in Egypt was prepared. Boulos (1999-2005) recorded 275 Mediterranean endemic taxa in the Egyptian flora. Of these, 106 (38.5%) Mediterranean endemics were recorded in the western Mediterranean region of Egypt by Ahmed (2009). Indeed, the present study has reduced this number to be 57 (20.7%) Mediterranean endemic taxa in Egypt (Appendices 1 and 2). Taxonomic authorities are ommitted from the names of taxa mentioned in the text below; they are included in the taxon lists of Appendices 1 and 2.

There were 57 Mediterranean endemic taxa, including species, subspecies and varieties, recorded in this study belonging to 46 genera and 22 families. The most represented genera were Allium and Fumaria (4 taxa each); Bellevalia (3 taxa); Muscari, Centaurea and Limonium (2 taxa each). Gymnosperms were not represented by any taxon, while angiosperms were represented by 2 clades (Table 3).

Four monocot families were recorded (Amaryllidaceae, Asparagaceae, Poaceae and Posidoniaceae), with 10 genera and 17 species, while eudicots were the most represented with 18 families. The highly represented families were Asteraceae (10 taxa), Fabaceae and Asparagaceae (8 taxa each), and Amaryllidaceae (5 taxa) (Fig. 2). On the other hand, 12 families were represented by only 1 taxon (Posidoniaceae, Convolvulaceae, Euphorbiaceae, Resedaceae, Apiaceae, Caprifoliaceae, Caryophyllaceae, Rubiaceae, Santalaceae, Solanaceae and Scrophulariaceae) (Appendix 1). The diversity of the major taxonomic groups indicated that the eudicots were the most diverse in terms of families, genera, species, subspecies and varieties, whereas monocots were the least. The ratio of genus to family (G/F) had a maximum value in monocots (2.5) and a minimum in eudicots (2). The ratio of species to genus (S/G) had a maximum in monocots (1.7) and a minimum in eudicots (1.1), while the subspecies to species ratio was very low in all the major taxonomic groups (Table 3).

Fourteen major habitats harbour the Mediterranean endemics in Egypt: 10 were natural habitats and 4 were anthropogenic habitats. The most represented habitats were the coastal dunes (17 taxa = 29.8%), followed by non-saline depressions (16 taxa = 28.1%), whereas alluvial soils (only one taxon: Bellevalia warburgii) and waste lands (2 taxa = 3.5%) were the least represented habitats (Fig. 3).

National geographic distribution.-The Mareotis (west) subsector was the richest of the three subsectors with 46 taxa (= 80.7%; e.g. Allium mareoticum, Muscari albiflorum and Scilla peruviana), followed by the Sinaitic (east) subsector with 15 taxa (26.3%; e.g., Bellevalia warburgii, Coronilla repanda, and Vicia sinaica), whereas, the Deltaic (middle) subsector had the least Mediterranean endemics with only five taxa (8.8%).

Global geographical distribution.-All the recorded taxa belonged to the Mediterranean region in the Boreal kingdom, which is divided into four sub-regions: North African sub-region (52 taxa = 91.2%), Eastern (levant) sub-region (36 taxa = 63.2%), Balkans sub-region (21 taxa = 36.8%) and North-western sub-region (16 taxa = 28.1%). The Mediterranean endemics in Egypt were recorded in 21 Mediterranean basin countries out of 25 (Table 4). The greatest number of taxa was recorded in Libya (29 taxa = 50.9%), followed by Palestine (24 taxa = 42.1%), then Syria and Lebanon (22 taxa each = 38.6%), while the lowest number (one taxon = 1.8%) was reported in San Marino, Slovenia and Bosnia and Herzegovina. On the other hand, no taxa were recorded in Malta, Monaco, and Holy See (Vatican City) (Fig. 4). Interestingly, 47 Mediterranean endemics (82.5%) were recorded in Egypt (excluding Sinai Peninsula that belongs to North African sub-region), and 15 (26.3%) in Sinai Peninsula belonging to Eastern sub-region.

Furthermore, there were 11 steno-endemics recorded only in Egypt (19.3% of the total Mediterranean endemics). They were Allium mareoticum, Anthemis microsperma, Bellevalia salah-eidii, Echinops taeckholmianus, Fumaria microstachys, Limonium sinuatum subsp. romanum, Muscari albiflorum, Pancratium arabicum, Thesium humile var. maritima, Muscari salah-eidii and Vicia sinaica. In addition, nine near-endemic taxa were recorded in Egypt and Libya (Allium barthianum, Allium blomfieldianum, Ebenus armitagei, Euphorbia parvula, Helianthemum crassifolium subsp. sphaerocalyx, Herniaria cyrenaica, Valantia columella, Valerianella petrovitchii and Verbascum letourneuxii), while two taxa (Lycium schweinfurthii var. aschersonii and Linaria joppensis) were recorded in Egypt and Palestine. Besides, Thymbra capitata, Posidonia oceanica and Lotus cytisoides were widely distributed in the Mediterranean basin. Regarding the life span, the perennials dominated the annuals.

Regarding the growth forms of the recorded taxa, the highest percentage of them were perennials (31 taxa = 54.4% of the total taxa), while annuals were represented by 26 taxa (45.6%). The life form determination indicated that therophytes (26 taxa = 45.6%) constituted the highest percentage of Mediterranean endemics, followed by geophyte-helophytes (13 taxa = 22.8%), hemicryptophytes (nine taxa = 15.8), and chamaephytes (five taxa = 8.8%). Posidonia oceanica was the only represented hydrophyte, while phanerophytes were represented by Ebenus armitagei, Helichrysum orientale and Lycium schweinfurthii var. aschersonii (Fig. 5). The sex form of Mediterranean endemics was arranged ascendingly as follows: monoecious (three taxa = 5.3%) and hermaphrodite (54 taxa = 94.7%).

Analysis of the flowering time indicated a gradual increase in the frequency of the flowered taxa from December (two taxa = 3.5%) till reaching a maximum in April (48 taxa = 84.2%) and March (41 taxa = 71.9%) forming positively skewed curve. Generally, the period from March to May was characterized by the highest flowering peak, while the period from September to December was characterized by the lowest (Fig. 6). In relation to life forms therophytes had maximum values followed by hemicryptophytes and geophyte-helophytes. It’s obvious that all the recorded life forms had maximum values in March and April. Surprisingly, only phanerophytes stay flowered all the year (Fig. 6).

Determination of dispersal types indicated that the ballochores (21 taxa = 36.8% of the total plant taxa) were the most represented dispersal type, followed by the pogonochores (11 taxa = 19.3%), then the sarcochores (nine taxa = 15.8%) and microsclerochores (six taxa = 10.5%) (Fig. 7). On the other hand, cyclochores, desmochores and sporochores were the least represented dispersal types (only one taxon each = 1.8%).

The relation between the number of Mediterranean endemics and the rarity forms indicated that SNN was the most represented (60 taxa = 92.3%), followed by SNA (four taxa = 6.2%). On the contrary, SWN was the least represented by only one taxon (Verbascum letourneuxii).

In the present study, 218 taxa were excluded from the preliminary list of Mediterranean Egyptian endemism. For example, Tripodion tetraphyllum was recorded in the Netherlands (JSTOR 2022) and inner Anatolia (Hassler 2004-2022). Although Leiotulus alexandrinus is a Mediterranean element, it was collected along Gulf of Suez, eastern desert, Egypt (Abd El-Ghani & al. 2017; Hassler 2004-2022). Therefore, it was excluded from Mediterranean endemics list in this study. Although the endemic Brassica deserti was recorded by Kamel & al. (2008) only in the Mediterranean coast of Sinai, it was indicated by El-Khalafy & al. (2021) in Gebel Igma, southern Sinai. Hence, it was excluded from the present study. The same status is noticed for Convolvulus palaestinus that was collected from southern Sinai by El-Husseini & al. (2008).

Notably, some Mediterranean elements were excluded for their extension to adjacent regions. For example, Aegilops longissima, Astragalus trimestris, Ononis natrix subsp. stenophylla and Trigonella maritima extend into adjacent territories of the Saharo-Arabian region as mentioned by Feinbrun-Dothan (1978, 1986) and Zohary (1966, 1987). Even though Trisetaria koelerioides was recorded in Suadi Arabia in POWO (2022) and Hassler (2004-2022), it is recorded as Mediterranean endemic in the present study. It does not exist in Chaudhary (1999-2001). In addition, identification of T. koelerioides in the literature could have been mistaken with Trisetaria chaudharyana H.Scholz (Eflora of the Kingdom of Saudi Arabia 2020).

Indeed, Heliotropium hirsutissimum was recorded in Suadi Arabia in Hassler (2004-2022), but Chaudhary (1985) elucidated that most reports of this species from Suadi Arabia could possibly be a misidentification of H. arbainense Fresen. Consequently, it can be considered as an east Mediterranean endemic. The endemic Anthemis microsperma was recorded by Täckholm (1974) in only the western Mediterranean region, but Boulos (1999 -2005)recorded it from northern Sinai. There are no authentically identified collected specimens supporting its occurrence in Sinai. It might have been observed there only one time and then disappeared or was misidentified with another similar taxa. The same case holds for Anchusa undulata subsp. hybrida that was recorded in Nile region in Boulos (1999-2005) in error.

As a result of climate change, the Mediterranean region’s borders with the Euro-Siberian region to the north (Europe) and the Saharan region to the south (Africa) have been fluctuating over millions of years (Vargas 2020). However, the northern and southern boundaries of the region are overall relatively decipherable, and disagreement among authors appears only in relation to comparatively small areas (mainly the Iberian and Balkan Peninsulas). In contrast, the establishment of both the western and eastern boundaries particularly is not generally agreed upon (Takhtajan 1986).

On the national scale, along the years, there have been some problems associated with the delimitation of phytogeographical regions, especially the Mediterranean floristic region in Egypt. Muschler (1912), Täckholm & Drar (1950-1969), Hassib (1951) and Täckholm (1956, 1974) demonstrated that the Mediterranean region in Egypt comprises two subregions: the Western Mediterranean coastal region (which stretches from Sallum to Alexandria, including Rosetta) and the Eastern Mediterranean coastal region (which stretches from Port Said to Rafah). Further, Zohary (1973) clarified that Mediterranean coast of Egypt harboured vegetation of the “Batha” type as well as the order Ballotelia undulatae belonging to the class Quercetia calliprini of the Eu-Mediterranean vegetation.

Zohary (1973) and Takhtajan (1986) proposed that there were two gaps in the Mediterranean territory along the North African sub-region: first gap in northern Egypt from Omayed to Rafah, and the second one in Libya along the Great Syrte between Tripoli and Benghazi. Hence the Saharo-Arabian vegetation advances quite close to the coast. Wickens (1976) restricted the North African gap to be only in Sinai (between Port Said and Rafah).

It is sad to report that Egypt had already lost over one hundred kilometers of its Mediterranean territory due to climate change, human activities, especially tourist summer resorts. As a result, it is now difficult to find any trace of the natural vegetation reported earlier in this region (El-Hadidi & Hosni 2000; Halmy 2012; Halmy & al. 2015a, 2015b; Halmy 2019; Halmy & al. 2019). Nevertheless, the current available data provide enough evidence for extending the Mediterranean territory in Egypt from Sallum, on the Egyptian-Libyan border, to Rafah, on the Egyptian-Palestinian border, with 3 subsectors: western (from Sallum to Alexandria), middle (between Alexandria and Port Said) and eastern (between Port Said and Rafah) as elucidated by Zahran & al. (1985). The Mareotis (western) subsector of Mediterranean coastal belt in Egypt is by far the richest part in Mediterranean endemics because it receives a higher amount of rainfall (Zahran 2010). One other study, to our knowledge, has shown that about 60% of the Mediterranean taxa in Egypt do exist in the Mareotis subsector (El-Hadidi & Hosni 2000).

The Mediterranean coastal land of Egypt is characterized by outstanding physiographic variations which leads to the differentiation of several types of habitats. The predominance of coastal dunes vegetation found in this study agrees to some extent with Batanouny (1999) and Ahmed & al. (2015) who stated that the majority of the Mediterranean coast in Egypt is bordered by sand dunes which represent a landscape with special characteristics and features. In fact, there are different types of dunes: (1) the mobile dunes that are characterized by loose, mobile sand and specialized plant species that are usually sand binders, which can tolerate the burial of their shoot systems in sand and the exposure of their roots. These plants are capable of producing adventitious roots from the buried vegetative part (such as Hyoseris radiata subsp. graeca, Centaurea pumilio and Pancratium arabicum); and (2) the consolidated dunes, characterized by phreatophytes that send deep roots which penetrate into the dunes to make use of the deep-sited fresh water (such as Helianthemum crassifolium subsp. sphaerocalyx and Lotus polyphyllos).

Furthermore, non-saline depressions occupy the major part of the land surface between the coastal ridges parallel to the coast such as the Abu Sir Depression, between the coastal and Abu Sir ridges; and the frontal plain, between the Gebel Mariut ridge and the Mariut tableland (Ayyad 1976). They are inhabited by Mediterranean endemics such as Anthemis microsperma, Muscari parviflorum, Verbascum letourneuxii, and Cynara cornigera.

The term ‘life form’ is distinct from the term ‘growth form’. The hereditary transmitted growth form is the morphology of the individual’s stem, whereas the environmentally influenced life form is the overall appearance of the plant community (Glime 2017). One spectacular phenomenon, to be noticed here, is the prevalence of perennials over annuals in the Mediterranean basin. This finding agrees to some extent with Zohary (1973) who assumed that predominance of sclerophyllous evergreen perennials is a characteristic trait that distinguishes this subsector from the adjacent ones. Therophytes flower during the rainy season, giving the coastal belt a temporary flush as a grassland desert. Whilst, woody shrubs and perennial herbs dominate during the longer dry periods giving the scrub vegetation of the area, which is scattered sparsely and grouped in patches (Zahran 2010). The prevalence of hemicryptophytes can be due to the widespread degradation of ecological conditions as a result of human activity and climate change (Habib & al. 2020). Phanerophytes came in the next order, a result that agrees with Kadi-Hanifi (2003), who elucidated that the aridity and openness of the environment reduce the number of phanerophytes.

In Mediterranean climates, the climatic factors most important in determining vegetation structure and composition appear to be the total amount of water available (in relation to potential losses) and the amount available in summer. In more continental Mediterranean areas, winter temperature also plays an important role. Reduced total water availability (in relation to evaporation) results in dominance by smaller plant forms. On the other hand, increased summer water stress results in smaller and harder leaves. The most striking feature of Mediterranean vegetation is its potential diversity of plant forms (Box 1982).

Flowering plants show a variety of sex forms (Barrett 2002). Only 6% of angiosperms are polygamous individuals, whereas hermaphrodites are the most represented (Tomaszewski & al. 2018). In the present study, most of the recorded species presented bisexual flowers (95.9%). Bisexual taxa are superior all over the world (Lewis 1941). Accordingly, they have variable possible chances of cross and self-pollination by autogamy (including cleistogamy) and geitonogamy giving many new individuals (Peedia, 2017). Bisexual flowers have parts such as bracts and pedicels that aid in insect attraction or organs that facilitate precipitation and uptake of pollen grains during animal pollination (Bawa & Beach 1981). Bram & Quinn (2000) reported that dioecy is rare angiosperms. The low number of dioecious taxa could be interpreted in the view that these taxa suffer a reproductive handicap, because their populations contain fewer seed-producing individuals (Queenborough & al. 2009).

Most Mediterranean endemics flower from May to March, while a low percentage flower from August to January. In Egypt, plants start to grow from November when temperatures and evaporation are low, and humidity is high enough to make soil water content suitable for plant growth till reaching flowering stage in March and April (Heneidy 2010).

Croteau (2010) described the role of movement factors like wind, water or animals in the passive dispersal of diaspores (i.e., seeds, spores or fruits) of flowering plants. These disseminules (i.e., diaspores) have some metamorphosis that help in transporting them from the parent plant to a new site. The predominance of ballochoric taxa (diaspores forcibly ejected from parent plant) may be due to the dry climate in Egypt. They are able to shatter when exposed to drought (El-Sheikh 1996). The commonness of pogonochoric taxa (diaspores with long hairs) reflects the suitability for wind dispersal in Egypt. The wide distribution of sarcochores (soft and fleshy diaspores) in the present study indicates that the principal mode of dissemination is the sarcochory by animals (i.e., zoochory).

Regarding the rarity forms, 60 taxa (92.3% of the total Mediterranean endemics) in the present study belonged to SNN cell, which means that these taxa are threatened. Chen & al. (2014) attributed the rarity of plants to internal causes such as genetic drift, pollination failure or unfeasibility of diaspores; or external causes such as human impacts or environmental factors like climate, soil or pathogens. Human impacts, roads, resorts or cultivation may be causes of habitat loss (Ahmed 2009). On the other hand, four Mediterranean endemics (Anthemis microsperma, Bellevalia sessiliflora, Filago mareotica and Thymbra capitata) belonged to the SNA cell. One of the recorded taxa (Verbascum letourneuxii) was the only Mediterranean endemic belonged to SWN; their population numbers are decreasing leading to habitat reduction, which in turn may cause the taxon to be extinct and move to one of the rare cells (SNA or SNN) (Rabinowitz 1981; Ahmed 2009). High percentage of rare taxa (SNN, SNA and SWN), whereas absence of common taxa (LNA, LNN and LWA) are indicators that support a management plan to conserve Mediterranean endemics in Egypt.