The famous fish beds of Lebanon: the Upper Cretaceous Lagerstätten of Haqel, Hjoula, Nammoura and Sahel Aalma

Haqel, Hjoula, Nammoura (all Cenomanian) and Sahel Aalma (Santonian) localities (Fig. 1) each represent Konservat-Lagerstätte in their own right, with distinct depositional environments, faunas and taphonomies and are described separately in detail below. It is, however, important to state that these may not be the only Lebanese Upper Cretaceous Lagerstätten. Localities such as Ain el Ghabour (عين الغبور) (near the town of Zeitoun (زيتون ); e.g. Forey et al. 2003), Maifouq (ميفوق ; e.g. Audo and Charbonnier 2013) and others yield exceptionally preserved fossils; however, these sites have not been the subject to major investigations, and less is known about their exact age, faunas, palaeoenvironment and collection history.

Occasionally, the geographical locality data of Lebanese fossils are very general, confusing different fossiliferous sites and/or erroneously combining sites. Detailed analysis of in situ facies, microfossils and sedimentary chemistry is lacking, making it difficult to trace the true provenance of Lebanese fossils from historical collections. Moreover, there are a plethora of translations and various spellings of site names throughout the literature. As Arabic is the official language of Lebanon, place names have Arabic (which we have included), English and French spellings on maps and road signs, which has led to some confusion. Here, we catalogued examples used in the literature (to help readers with internet search engine keywording), but we have attempted to use the direct anglicization of the respective Arabic town and village names from Lebanese road signs and maps where possible. We suggest standardization in future literature.

The first described, and probably most well-known Lebanese Lagerstätte, is Haqel (حاقل ). Today, fossils from Haqel are found in a small quarry (c. 700 m above sea level) located approximately 1.5 km east from the small town of Haqel (within the Keserwan–Jbeil Governorate), itself approximately 30 km NE of Beirut and around 12 km east of the sea (Fig. 2a). Within the scientific literature is a variety of spellings of Haqel; examples include ‘Hakel’ and ‘Hâqel’ (e.g. Dalla Vecchia et al. 2001; Audo and Charbonnier 2013).

The quarry at Haqel has a history of human resource extraction, with the quarried stone being used as building material from the fourth century (Capasso 2017). During extraction of this building stone, fossil fish were frequently recovered, often highly articulated, leading to a long documented human history, with examples of the fossils being reported from as long ago as the Roman occupation (see Box 1). Today, the site is still accessible, with tours and researchers visiting the quarry to prospect for fossils; however, it should be noted that the land the site is on is privately owned by several individuals.

The fauna at Haqel is dominated by small teleost fish, and is world renowned for its fossil ichthyofauna, with currently over 70 different genera of fish (c. 93 species) being described from this location (El Hossny et al. 2020; El Hossny and Cavin 2023). Moreover, Haqel is well known for its fish mass mortality slabs, which often contain large numbers of overlapping skeletons that are highly prized by collectors around the world (Forey et al. 2003; Fig. 2j). Although dominated by fossil bony fish, Haqel also boasts a high diversity of other fossil organisms, often preserved with soft tissues. These include sharks (e.g. Jambura et al. 2021), batomorphs (e.g. Cappetta 1980b; Forey et al. 2003), crustaceans (e.g. Garassino 1994), annelid polychaetes (e.g. Bracchi and Alessandrello 2005), cephalopods (e.g. Fuchs et al. 2009; Fuchs and Larson 2011a, b), echinoderms (e.g. Jaselli 2016) and disarticulated pterosaur fragments (Dalla Vecchia et al. 2001). Non-calcareous marine algae have been identified in thin sections from Haqel (e.g. Basson 1981; Krings and Mayr 2004). Terrestrial plants are also found at Haqel, although they are considered rare (Dalla Vecchia et al. 2001).

Although the age of the Haqel Lagerstätte has historically been debated (see El Hossny et al. 2020 for a review), the current consensus, based on ammonite and foraminifera biostratigraphy, dates the locality to the late Cenomanian (Hemleben 1977; Wippich and Lehmann 2004; Fuchs et al. 2009). The sediment of the Haqel quarry is characterized as yellow to grey laminated lithographic limestones that are hard, fine grained and well bedded (Hückel 1970, 1974; Hemleben 1977; Bracchi and Alessandrello 2005), with some distinct layers containing flint lenses or nodules (Fuchs et al. 2009; Capasso 2017; Fig. 2a). Assigned to the Sannine Formation (Walley 1997), these sediments are thought to have been deposited in a warm, shallow sea (Krings and Mayr 2004; El Hossny et al. 2020). Swinburne and Hemleben (1994) suggested that the carbonate was primarily pelagic in origin, deposited via suspension, and that the bedding laminae may have been caused by turbidity currents. However, an in-depth sedimentological analysis is required to confirm this.

The palaeoenvironment of Haqel is poorly understood. Swinburne and Hemleben (1994) suggested that tectonic activity acting on the underlying carbonate platform formed a steep-sided, fault-bounded marine basin on the outer margin of the continental shelf, perhaps several hundred kilometres from the palaeocoastline. The fossiliferous plattenkalks were subsequently deposited in this basin (Swinburne and Hemleben 1994). The evidence for the basin's tectonic activity stems from an ‘anomalously thick sequence’ of olistostrome within the plattenkalk, which suggests an intermittent strike-slip opening and deepening of the basin (Swinburne and Hemleben 1994). The edges of the basin are poorly resolved, but the basin is thought to have been surrounded by submerged carbonate rudist or oyster reefs that may have even formed parts of the actual walls of the basin (Swinburne and Hemleben 1994) but evidence for this was not presented by the authors.

The fossiliferous areas of Haqel are reported to be extremely geographically restricted, which Hückel (1970) described as occurring in approximately 250 m wide bathymetric lows (termed ‘sinkholes’ by Hückel) that may have formed as a result of tectonic activity. Clear sedimentary evidence to support the ‘sinkhole’ hypothesis is lacking. Regardless, it has been suggested that the restricted bottom waters within the basin were anoxic, based on the high organic content and bituminous nature of the fossiliferous sediments, along with the lack of bioturbation and sessile bottom feeding organism fossils (Roger 1946; Patterson 1967; Hückel 1970; Hemleben 1977; Schram et al. 1999). However, Hemleben (1977) also described that some horizons within the fossiliferous zone do actually show evidence of bioturbation, and that fossil fish are not found in these horizons. This suggests that within the fossiliferous beds of Haqel, there was a periodic nature to the preservation of fossils. In the same study, Hemleben (1977) noted that fossil fish, especially the mass mortality slabs, were often restricted to distinct horizons, and he proposed that nutrient upwellings may have triggered blooms of dinoflagellates (termed a ‘red tide’). This potential kill mechanism would choke and kill pelagic organisms, which would then sink into the anoxic waters of the basin below, accumulating large numbers of carcasses in a bottom water devoid of scavengers. This hypothesis is frequently cited in the literature as to why soft tissue preservation is so common at Haqel, as the algal blooms would create environmental conditions suitable for soft tissue preservation via authigenic mineralization (e.g. Swinburne and Hemleben 1994; Schram et al. 1999; Wilson et al. 2016). Evidence for this hypothesis, sedimentary or otherwise, was not presented by the authors or in subsequent studies (Hemleben 1977). Detailed sedimentological, geochemical and taphonomic analyses of the fossiliferous sediments are required to accurately ascertain the drivers of preservation at Haqel.

Hjoula (حجولا) was first documented in the late 1800s (Lewis 1878), making it the third Lebanese Upper Cretaceous locality to be identified. It is sometimes paired with Haqel in the literature owing to their geographical proximity, faunal similarities and similar (if not the same) age. The fossils from Hjoula are found in a series of small limestone quarries located approximately 5 km south of Haqel, near its namesake town, Hjoula (حجولا, within the Keserwan–Jbeil Governorate). Similarly to Haqel, the name of this site in the scientific literature is inconsistent, including ‘Hajula’, ‘Hazhüla’, ‘Hjoûla’ and ‘Hadjoula’ (e.g. Lewis 1878; Dalla Vecchia et al. 2001; Capasso 2017; Miyashita et al. 2019). As Hjoula was described in the late 19th century during the period of fervent palaeontological interest from European and American fossil collectors, large historical collections of material from this site are now housed in many northern hemisphere natural history institutions (Capasso 2017; see Boxes 2 and 3). The site can still be visited and prospected today, although the land is privately owned.

Despite the fauna of the site having many similarities to that of Haqel and their close geographical locations, there are notable differences. Hjoula is dominated by crustaceans, particularly decapods, but articulated fish fossils are comparatively less common than at Haqel (Capasso 2017). This is also true of terrestrial plants (Capasso 2017). However, unlike Haqel, Hjoula's fossil assemblage includes insects (e.g. Maksoud and Azar 2021), nautiloids with possible soft tissues (e.g. Klug et al. 2021), exceptional coleoid cephalopods (e.g. Fuchs et al. 2009), echinoderms (e.g. Reich 2004), a batomorph embryo inside an egg case (Capasso and Yamaguchi 2023), a lone example of a myxinoid (hagfish; Miyashita et al. 2019), more complete pterosaur remains (e.g. Elgin and Frey 2011; Kellner et al. 2019) and undescribed turtle remains (Bazzi, pers. comm.).

Biostratigraphic data suggest that Hjoula and Haqel formed contemporaneously during the late Cenomanian (Hemleben 1977; Wippich and Lehmann 2004). Moreover, the geology of the two sites is similar, both being dominated by highly fossiliferous, laminated lithographic limestones (Hückel 1970; Fig. 2c). However, there are several key differences in lithology. For example, the limestones at Hjoula have been described as softer, more compacted, lighter in colour (yellow or grey) and completely devoid of flint nodules (Fuchs et al. 2009). The sediments at Hjoula have also been stated to be more bituminous than at Haqel, with higher quantities of pyrite, suggesting more prolonged and intense periods of bottom water anoxia (Swinburne and Hemleben 1994). Hückel (1970) did note that the fossiliferous zones at Hjoula are also highly geographically restricted and they suggested that, as at Haqel, small palaeogeographical ‘sinkholes’ may have increased the likelihood of soft tissue preservation.

Swinburne and Hemleben (1994) suggested that Hjoula may have also been a steep-sided tectonically active marine basin palaeoenvironment surrounded by walls of shallower carbonate rudist or oyster reefs, but it is important to note that Hjoula is thought to be a completely separate marine basin from Haqel (Swinburne and Hemleben 1994) although the distance between these two basins during deposition has not, to our knowledge, been investigated. Some researchers have suggested that Hjoula was further from the palaeocoast than Haqel (Kellner et al. 2019) based on the absence of terrestrial plant material and insect fossils, although recent descriptions of insects and plants from Hjoula undermine this hypothesis (e.g. Capasso 2017; Maksoud and Azar 2021). As at Haqel, a detailed palaeoenvironmental investigation is required to fully understand the taphonomy and palaeogeography of the Hjoula palaeobasin.

Nammoura (نمورة) is the most recent Lebanese Upper Cretaceous plattenkalk to be described. Spellings in the literature include ‘Nammoûra’, ‘En Nammoura’, ‘Nmoura’ and ‘Al Nammoura’ (e.g. Rage and Escuillié 2000; Dalla Vecchia et al. 2002; Capasso 2017; Jambura et al. 2021). The Nammoura Lagerstätte is confined to a series of small quarries, which together take their name from the local village of Nammoura (نمورة) within the Keserwan–Jbeil Governorate, approximately 15 km SE of Haqel. Quarrying for building material began in 1915, with major expansion of the quarry occurring during the 1950s (Forey et al. 2003; Capasso 2017). In 1981, a road was constructed to the quarry, allowing motor vehicle access, which has greatly increased the output of rock, and fossil material, from the quarry (Forey et al. 2003). The plattenkalk at Nammoura is much more geographically extensive than at Haqel and Hjoula, extending for over 1 km, with exposure on both sides of the Abraham River (نهر إبراهيم ) (also known as the Adonis River). The river splits the quarry into two sites, one on either side of the valley, with only a restricted section on one side of the river bank seeming to be fossiliferous (Forey et al. 2003; Capasso 2017). Currently, the land is private and off limits without express permission from the quarry owners, but collecting fossils from the quarries is possible, with some pre-arranged tours being allowed to prospect the site.

Nammoura, although rich in fossil diversity, has been described as less fossiliferous than Haqel and Hjoula (Capasso 2017). The most abundant fossils found at Nammoura are small teleost fish, although mass mortality fossil slabs are not found at Nammoura (Forey et al. 2003). It has also been consistently reported in the literature that the fidelity of preservation of fish is much higher than that of Haqel and Hjoula; the fossils are commonly fully articulated and often have preserved soft tissues such as skin, muscles and eye spots (Bannikov and Bacchia 2000; Dalla Vecchia et al. 2002; Forey et al. 2003; Capasso 2017). The site is also well known for its terrestrial plant fossils (Dilcherm and Basson 1990; Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002; Fig. 2f). Other fossil groups found at Nammoura include insects (e.g. Nel et al. 2004), crustaceans (although these are rare according to Dalla Vecchia et al. 2002), echinoids, cephalopods (Jattiot et al. 2015), bivalves (Dalla Vecchia et al. 2002), testudines (Tong et al. 2006), marine squamates (e.g. Dal Sasso and Renesto 1999), a single fossil avialan (Dalla Vecchia and Chiappe 2003; Cau and Arduini 2008), disarticulated singular avialan feathers (e.g. Dalla Vecchia et al. 2002; Cau and Arduini 2008), a partial arm of a non-avialan theropod dinosaur (Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002) and coprolites (Jattiot et al. 2015).

Biostratigraphic data from microfossils indicate that units deposited on top of the Lagerstätte are from the middle to late Cenomanian, indicating that the site at Nammoura is slightly older than those at Haqel and Hjoula (Dalla Vecchia et al. 2002). Analysis of the major outcrops along the Abraham river by Dalla Vecchia et al. (2002) divided the Lagerstätte into eight lithological units, and found that fossiliferous beds were restricted to a single wedge-shaped unit of gently sloping, fine-grained, bluish grey to green lithographic limestones with some small chert lenses and limited evidence of bioturbation (termed ‘Unit 3a + b’). These fossiliferous limestone beds at Nammoura have been reported as being significantly different from the limestones at Haqel and Hjoula, with both Dalla Vecchia et al. (2002) and Forey et al. (2003) describing the limestone at Nammoura as more of a ‘classic’ lithographic limestone with plane-parallel lamination and tabular bedding more closely resembling the plattenkalks of the Solnhofen Lagerstätte, Germany (Fig. 2b). It has also been reported that the greenish colour of the limestone is different enough that fish fossils from this site can be easily distinguished from those recovered at Haqel or Hjoula (Dalla Vecchia et al. 2002).

Nammoura has a different depositional environment from the aforementioned sites, and was probably located closer to the inner continental shelf (see Swinburne and Hemleben 1994; Fig. 1). The fossiliferous limestone beds at Nammoura are geographically more extensive than at either Haqel or Hjoula. The individual beds have been reported as distinctly lens-shaped, and do not appear to be tectonically constrained (Dalla Vecchia et al. 2002; Forey et al. 2003). It is likely that these limestones formed within a pre-existing palaeo-basin on top of the underlying, gently sloping, clinoform, shallow-marine platform (Dalla Vecchia et al. 2002; Forey et al. 2003). Dalla Vecchia et al. (2002) postulated that the fossiliferous basin was probably restricted, with associated stagnant water and oxygen depletion; there is also little evidence of coarse sediment input and a rarity of planktonic and nektonic organisms. Furthermore, they reported that there is evidence, based on crystalline overgrowths of plant material, to suggest hypersalinity, although this hypothesis requires further investigation (Dalla Vecchia et al. 2002). Unlike the two previously described Lagerstätten, there are no distinct fossil ‘hotspots’ at Nammoura (Forey et al. 2003) and a lack of ‘mass mortality’ fossils, so the stagnant, hypersaline water could have been a ‘passive’ kill mechanism. Interestingly, Dalla Vecchia et al. (2002) found little evidence of rapid inputs of sediment, suggesting that rapid burial could not be a kill mechanism at this site. Taphonomically, swift burial is often described as a prerequisite for soft tissue preservation (see Clements and Gabbott 2021); however, it is unlikely that carcasses were buried swiftly at Nammoura. Dalla Vecchia et al. (2002) suggested that the salinity and stagnant water may have inhibited scavenging and that cyanobacterial mats could act to trap macro-organic material and, in concert with the proposed anoxic hypersalinity, created conditions that retarded decay of soft tissues; however, Dalla Vecchia et al. (2002) pointed out that no sedimentary evidence of cyanobacterial mats has yet been identified.

The abundance of terrestrial organisms and plants has been used to suggest that Nammoura was proximal to the palaeocoast (Krassilov and Bacchia 2000; Rage and Escuillié 2000; Cau and Arduini 2008); however, palaeogeographical investigations indicate that the coast was hundreds of kilometres to the SE (Hückel 1970; Philip et al. 1993; Dalla Vecchia et al. 2002). Dalla Vecchia et al. (2002) suggested that the shallow relief of the limestone platform would have been susceptible to eustatic oscillations, and that emergent carbonate islands may have acted (1) to restrict seawater flow into the Nammoura area, allowing hypersaline conditions to form, and (2) as a potential source of terrestrial fossils.

After Haqel, the second Lebanese Upper Cretaceous Lagerstätte to be described was Sahel Aalma (ساحل علما). This site yields spectacular fossils, many of which were collected during the 19th and early 20th century. Unfortunately, analysis of the depositional environment is limited, and future investigation will be difficult owing to lack of access today. Located approximately 20 km NE of Beirut, and 21 km SW of Haqel, the village of Sahel Aalma (علما ساحل , within the Keserwan–Jbeil Governorate) overlooks the sea (c. 1 km east of the shoreline) and is known for its traditional Lebanese houses and lemon tree groves. As for other Lebanese sites, there are spelling variations in the literature including ‘Sâhel Aalma’ (e.g. Dalla Vecchia et al. 2001) and, more commonly, ‘Sahel Alma’ (e.g. Capasso 2017). Sahel Aalma has a long human history of fossil discovery dating back to at least the Crusades (see Box 1). Sahel Aalma has historically been a well-collected site, with significant historical collections at several major European museums (see Boxes 3 and 4). The Sahel Aalma Lagerstätte preserves a diverse range of fossil marine organisms including high numbers of fish (including deep-water forms; e.g. Davis 1887; Gayet et al. 2003), crustaceans (e.g. Audo and Charbonnier 2013), encrusting epibionts (e.g. Petit and Charbonnier 2012), cartilaginous fishes (e.g. Werdelin 1986), coleoid cephalopods (including the oldest known octopus; e.g. Woodward 1883; Fuchs 2006) and arthropods (Dalla Vecchia et al. 2001; Capasso 2017). Most fossils appear as flattened compressions, although soft tissue fossils (phosphatized muscles, gills, etc.) have been frequently reported.

The age of the Sahel Aalma site is well constrained. A detailed investigation of the foraminifera assemblage was undertaken by Ejel and Dubertret (1966), who refined the period of deposition to be during the late Santonian, making it approximately 10 myr younger than Haqel and Hjoula. The fossiliferous outcrop at Sahel Aalma is relatively geographically small (Ejel and Dubertret 1966) and, unfortunately, it is no longer possible to access the outcrop. After the mediaeval monastery adjacent to the outcrop fell into ruin, the land was purchased by a private company, which restored the monastery as a hotel (Capasso 2017). Owing to the lack of access, modern palaeoenvironmental studies have not been undertaken, and so the depositional environment is even more poorly resolved than for Haqel and Hjoula. The host rock is described as a chalky laminated limestone (Ejel and Dubertret 1966; Audo and Charbonnier 2013) with a soft, fine, powdery texture (Capasso 2017), but there is little more information in the literature. Audo and Charbonnier (2013) suggested that, based on the faunal assemblage including deep-water arthropods and chondrichthyans, Sahel Aalma was probably deposited in water depths exceeding 150 m, which would be greater than the depths at Haqel and Hjoula.

There are large numbers of fossil species that have been described from the four Upper Cretaceous Lebanese Lagerstätten (e.g. Figs 2–4). An in-depth description of all these species is beyond our scope; however, here we present brief overviews of several key fossil groups. There are several groups we do not cover here, including gastropods, bivalves, foraminifera and sea cucumbers (e.g. Hückel 1970; Krings and Mayr 2004; Reich 2004; Wippich and Lehmann 2004).

The majority of fossil plant descriptions from the Lebanese Lagerstätten are based on samples from Nammoura (e.g. Fig. 2f). Twelve species of ferns, gymnosperms (conifers and ginkgophytes) and angiosperms have been described from this locality, often represented by single leaves, small tufts and branches (Dilcherm and Basson 1990; Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002) although it is angiosperms, such as Sapindopsis, that dominate the floral association (Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002). Nammoura is the type locality for several species, including Nammouria cretacea and Nammourophyllum altingioides (Krassilov and Bacchia 2000). Interestingly, the flora found at Nammoura is substantially different from the Early Cretaceous and Turonian floras of other Middle Eastern fossil sites, and is, in fact, more closely related to coeval floras found in contemporaneous North American and central European sites (Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002). Many of the plants from Nammoura are xeromorphic (adapted to live in drought conditions), indicating that the climatic conditions of Cenomanian Lebanon may have been similar to the present-day Mediterranean climate (Krassilov and Bacchia 2000; Dalla Vecchia et al. 2002).

Arthropods are common in the Lebanese Lagerstätten, especially decapod crustaceans (Garassino 1994; Pasini and Garassino 2023; Fig. 3e). Some of the most studied are the ‘crest-bearing shrimps’ from Sahel Aalma, which have been the focus of several descriptive publications (e.g. Brocchi 1875; Audo and Charbonnier 2013). These charismatic arthropods are preserved as flattened composite impressions but commonly are associated with exquisite soft tissues such as muscles, gills, eyes, intestines with associated ingested matter and coprolites, and even preserved eggs (Audo and Charbonnier 2013). Multiple species of lobsters (Astacidea) have also been collected from the Haqel and Hjoula localities (Garassino 1994; Charbonnier et al. 2017a), including slipper lobsters (Scyllaridae). Studies of exceptionally well-preserved slipper lobster fossils, complete with their delicate petaloid antennae, from Hjoula have led to a greater understanding of the ontogeny and evolution of this captivating group (Haug et al. 2016). Another group that is uncommon in the fossil record but found in the Lebanese Lagerstätten are isopods, of which two different species have been reported from the Sahel Aalma locality: Cirolana garassinoi and C. cottreaui (Feldmann 2009; Feldmann and Charbonnier 2011). The Lebanese Lagerstätten also host a diverse selection of species of the Thylacocephala. Despite being one of the most enigmatic clades of arthropods, eight genera have been described from the Haqel, Hjoula and Sahel Aalma localities, and this represents most of the Cretaceous members of this mysterious clade (Dames 1886; Roger 1946; Lange et al. 2001; Charbonnier et al. 2017b). Exceptionally well-preserved specimens of horseshoe crabs (Xiphosurids, Tachypleus syriacus) have also been described from Haqel and Hjoula (Lamsdell and McKenzie 2015; Bicknell et al. 2019).

Terrestrial arthropods have also been found at the Lebanese Lagerstätten, predominantly insects from the Hjoula and Nammoura localities. These include a species of cockroach and many dragonflies (Nel et al. 2004; Vršanský and Makhoul 2013; Azar et al. 2019; Azar and Nel 2023). The large number of fossil insects recovered from these localities is unusual for Upper Cretaceous fossil sites, indicating that these outcrops are among some of the best in the world for collecting insect fossils of this age (Maksoud and Azar 2021).

Although polychaetes are rare in the fossil record, 17 species (in six families and seven genera) have been described from the Lebanese Lagerstätten by Bracchi and Alessandrello (2005). As would be expected, most of the Lebanese fossil examples are represented by only the recalcitrant parts of the organism (chaetae, jaws, etc.) occasionally associated with body outlines or impressions. However, several polychaete samples with remarkably preserved phosphatized musculature have been described (Bracchi and Alessandrello 2005). In fact, a polychaete specimen from Hjoula preserved with 3D phosphatized musculature has allowed the reconstruction of the body wall, gut and parapodia, allowing identification of the polychaete to the family level (Amphinomidae) (Parry et al. 2015).

Ectocochleate cephalopod fossils (ammonoids and nautiloids) have been found at all of the Lebanese Lagerstätten. Although the vast majority of these are isolated shells, it is not uncommon to find examples with in situ aptychi (jaws). Moreover, there are reports of ectocochleate cephalopod preserved with soft tissues from Lebanon. Wippich and Lehmann (2004) reported a heteromorphic ammonite from Haqel with in situ aptychi and potential soft tissue interpreted as a stomach with contents. Klug et al. (2021) reported two taphonomically wonderful fossil ‘mushes’ from Hjoula that are diagnosed as the fossilized internal anatomy of nautiloids (albeit sans shell). These fossils may represent the remains of a partially eaten (a pabulite) or scavenged nautiloid cephalopod but, in either case, further analyses may explain these bizarre taphomorphs.

One of the fossil ‘jewels’ of the Upper Cretaceous Lebanese Lagerstätten are the coleoid cephalopods (e.g. Fig. 3d, g and h). The first to be described by Woodward (1896) was Palaeoctopus newboldi, found at Sahel Aalma. Palaeoctopus was the oldest and only known pre-Cenozoic member of the octobrachia until Fuchs et al. (2009) described three new species of octobrachian cephalopods, themselves from the slightly older Haqel and Hjoula localities. Many of the eight-armed coleoid cephalopods from these sites, such as Keuppia and Styleoctopus, are phosphatically preserved with exceptional fidelity (see Fuchs et al. 2015; Box 2). The extensive work by Dirk Fuchs and colleagues to describe and interpret these fossils has been instrumental in furthering our understanding of the anatomy and evolution of these early crown group octopodes.

Although other coleoids, such as the vampyromorph Dorateuthis syrica, have been described from Sahel Aalma (Woodward 1883; Lukeneder and Harzhauser 2004), it should be noted that fossil decabrachians (squids) have not been found in the Lebanese Lagerstätten. These animals are known to have existed, and presumably lived in the region (Clements et al. 2016). The absence of this group is most probably a taphonomic bias related to their mode of life requiring increased levels of buoyancy control chemicals (such as ammonia), which would prevent their carcasses generating the conditions required for soft tissue phosphatization during decay (see Clements et al. 2016).

A recent paper described a stunningly preserved Myxinoidea (hagfish) from Hjoula (Miyashita et al. 2019). Tethymyxine tapirostrum represents one of the few known examples of a fossil hagfish, and is so well preserved that numerous soft tissues have been observed using synchrotron X-ray fluorescence, including cartilage, aspects of the branchial and digestive systems, slime glands and possibly a heart. The data obtained from this specimen have had a major impact on our understanding of the phylogenetic affinities of Cyclostomes (Miyashita et al. 2019).

Cappetta (1980a) provided the most comprehensive account of sharks from the Upper Cretaceous Lebanese Lagerstätten. His synopsis identified 20 species diagnosed from isolated teeth as well as partial and complete skeletons, which is significant considering the otherwise tooth-dominated fossil record of sharks (Cappetta 2012). This makes the Lebanese Lagerstätten deposits some of the best in the world for studying sharks from the final stages of the Mesozoic. Notably, these have yielded individual shark species from different ontogenetic stages, body size classes and ecological niches, and on occasion the preservation of organ structures and possible biomolecules. From the Sahel Aalma locality, a total of 14 species have been described so far, including two mitsukurinid lamniforms (Davis 1887; Arambourg 1952; Fig. 2i), two hemiscylliid orectolobiforms (Pictet and Humbert 1866; Signeux 1949), three scyliorhinids (Fig. 2e) and one triakid carcharhiniforms (Davis 1887; Cappetta 1980a), two somniosids and two squalid squaliforms (Pictet 1850; Davis 1887; Signeux 1950; Feichtinger et al. 2022), one pristiophorid pristiophoriform (Woodward 1932; Fig. 3f) and at least one hexanchid hexanchiform (Davis 1887; Cappetta 1980a). Conversely, four species have been described from Haqel; these include Pseudocorax kindlimanni (Jambura et al. 2021), Cantioscyllium decipiens (Pfeil 2013), Mesiteia emiline (Cappetta 1980a) and Pararhincodon lehmani (Cappetta 1980a). The orectolobiform Cantioscyllium decipiens has also been described from Hjoula (Pfeil 2013). Pfeil (2021) recently illustrated and briefly commented on additional material recovered from all aforementioned localities, including Nammoura, which have produced two species of lamniforms including Haimirichia amonensis and Squalicorax sp. Based on the listed taxa provided by Pfeil (2021), the Hjoula locality has the highest taxonomic diversity of sharks, and Nammoura has the least. It is worth noting that despite some taxonomic similarity between the sites, squatiniforms are found only from Haqel, heterodontiforms only from Hjoula, and pristiophoriforms and hexanchiforms only from Sahel Aalma, an indication of bathymetric and habitat differences.

Batomorpha is present across all of the Lebanese Lagerstätten, often preserved with articulated skeletons, including rajids (skates) from Haqel and Hjoula (e.g. Figs 2h and 3b) and Raja davisi from Sahel Aalma (Davis 1887; Fowler 1958). Several species of Rhinobatidae (guitarfish) have been recovered from Haqel and Sahel Aalma (Davis 1887; Woodward 1889; Cappetta 1980b; Brito and Dutheil 2004; Kachacha et al. 2017). Beautifully articulated fossils of the extinct clade Cyclobatidae are represented by at least three species of Cyclobatis that are frequently found at Hjoula, Haqel and Nammoura (Egerton 1844; Davis 1887; Cappetta 1980b). Other iconic Lebanese batomorph fossils include those of the Sclerorhynchoidei or ‘sawskates’ (Greenfield 2021). Two species of these rajiform rays have been described from Hjoula and are, unlike most fossil examples of this group, known to be fully articulated with preserved musculature and distinct body outlines (Hay 1903; Cappetta 1980b).

Chimaeriformes (chimaera or ghost sharks) are known from the Sahel Aalma locality. This material is rare and has been attributed to one species, Harriotta lehmani, although there are fossil remains of a larger, indeterminate species (Werdelin 1986).

Historically, the Upper Cretaceous Lebanese Lagerstätten are famous for the quality and quantity of actinopterygian fossils found here. The ichthyofauna of all four sites described in this review is dominated by crown-group teleosts; however, the latimeriid sarcopterygian Macropomoides orientalis has been recovered from both Haqel and Hjoula (Woodward 1942), and exceedingly rare amiaforms (stem-bowfins) have been found at Haqel (Pictet 1850; Davis 1887; El Hossny et al. 2020). Probable stem-teleost taxa are also known from the Upper Cretaceous of Lebanon, including three subclades of the extinct Pycnodontiformes; the Coccodontidae, Gebrayelichthyidae, Gladiopycnodontidae and Pycnodontidae, all of which are endemic to the Cenomanian of Lebanon (Poyato-Ariza and Wenz 2005; Taverne and Capasso 2014a, b, 2015, 2020; Cawley and Kriwet 2019). Other probable stem-teleost examples include the Aspidorhynchiformes (shield-snout forms), which are long, barracuda-like fish notable for their elongated rostrums formed from fused premaxillae, and have been found in Nammoura (Agassiz 1833–43; Forey et al. 2003). There are also icthyodectiforms, such as Eubiodectus libanicus, known from Haqel, Hjoula and Nammoura (Pictet and Humbert 1866; Forey et al. 2003). Many clades of probable crown-teleosts are represented in the Lebanese Lagerstätten, and have been extensively covered in other reviews (see Forey et al. 2003). Similar to the pcynodontiformes, the Protobramidae (a subclade of the extinct Tselfatiiformes) are endemic to the Cenomanian seas of present-day Lebanon (Taverne and Gayet 2005). Five members of the Pantodontidae, which has one living representative, have also been described from Lebanon (Taverne and Capasso 2012; Taverne 2021). Members of Elopomorpha are found as fossils in Lebanon, including Megalopidae (tarpons), Albulidae (bonefishes) and Anguilliformes (eels), all of which have been recovered from the four localities focused on in this review (e.g. Woodward 1901; Hay 1903; Goody 1969; Forey 1973; Belouze et al. 2003; Forey et al. 2003). Among the most fascinating fossil elopomorphs from Lebanon is the oldest known species of Notacanthidae (deep-sea spiny eel), Pronotacanthus sahelalmae, from Sahel Aalma (Davis 1887; Sepkoski 2002).

Members of the Clupeidae (herrings and sprats) have been reported from Haqel and Hjoula, and Chanidae (milkfish) are known from Haqel and Sahel Aalma (Heckel 1849; Pictet and Humbert 1866; Davis 1887; Woodward 1901; Gayet 1993). A wide variety of species belonging to Aulopiformes (lizard fish) are known from all four localities, including the armoured Spinascutichthys pankowskiae from Nammoura (Davis 1887; Forey et al. 2003; Murray et al. 2022) and the beautiful flying fish-like Exocoetoides minor and Cheirothrix spp. from both Haqel and Hjoula (e.g. Fig. 3c). Cheirothrix lewisii (Davis 1887) has wonderfully preserved enlarged pectoral fins and ventral lobe of the tail fin. It is hypothesized that Cheirothrix was able to make self-propelled leaps out of the water and then glide short distances, similar to the extant Exocoetidae. Members of Myctophiformes (relatives of the modern deep-sea lanternfish and blackchins) are also known from Haqel, Hjoula and Sahel Aalma (Woodward 1901; Hay 1903; Forey et al. 2003). Finally, there are numerous species of Acanthomorpha known from all four localities, many of which are close relatives of lampriformes (oarfish, ribbonfish, etc.), Polymixiidae (beardfishes), beryciforms and trachichthyiforms (Davis 1887; Forey et al. 2003; Delbarre et al. 2016).

The Lebanese Lagerstätten have yielded several specimens of marine squamates. Among the most scientifically important of these are the four-legged and long-tailed Pontosaurus (Dal Sasso and Renesto 1999; Caldwell and Dal Sasso 2004; Caldwell 2006). At least one specimen of this genus, found in Nammoura, consists of a complete skeleton preserved with an extensive number of scales that differ in size and shape (Caldwell and Dal Sasso 2004). Additionally, the four-legged and long-tailed marine lizard Aphanizocnemus libanensis is probably from the Lebanese Lagerstätten, but it is unknown which locality it is from (Dal Sasso and Pinna 1997; Dal Sasso and Renesto 1999). Another wonderful Lebanese reptile fossil is the transitional form Eupodophis descouensi, a species of marine snake that had vestigial hind limbs, and has been described from Nammoura (Rage and Escuillié 2000). Additional specimens of this species have been recovered from Haqel (Rieppel and Head 2004).

The only testudine to be described from the Lebanese Lagerstätten is the beautifully preserved holotype of the protostegid chelonioid Rhinochelys nammourensis found at Nammoura (Tong et al. 2006). Its taxonomic placement is uncertain, as recent analysis by Scavezzoni and Fischer (2018) has recovered it as more closely related to Notochelone, Protostega and other protostegids than to European species of Rhinochelys (including the type species). Other Lebanese turtle fossils (known to the authors) exist in private collections and are currently undescribed (e.g. Fig. 3i).

Despite their fragmentary nature, pterosaurs from the Lebanese Lagerstätten are among the best known from the Levant region during this time period (Arambourg 1954; Frey and Martill 1996; Tchernov et al. 1996; Pentland and Poropat 2023). One example is a partial forelimb recovered from Haqel, which has been assigned to Pteranodontoidea indet. (Dalla Vecchia et al. 2001). A more complete pterosaur specimen recovered from Hjoula consisting of various postcranial bones was later described as Microtuban altivolans. This represents one of the few non-azhdarchid azhdarchoids from the Late Cretaceous (Elgin and Frey 2011). The best preserved pterosaur from the Lebanese Lagerstätten is Mimodactylus libanensis. The holotype, found at Hjoula, is an almost complete skeleton and is the namesake of the clade Mimodactylidae (Kellner et al. 2019).

The only avian species described from the Lebanese Lagerstätten is the avisaurid enantiornithine Enantiophoenix electrophyla. The holotype was recovered from Nammoura, and consists of disarticulated postcranial elements as well as carbonized feather remains (Cau and Arduini 2008). The original description of the holotype is the first record of a Mesozoic avialan from northern Gondwana (Dalla Vecchia and Chiappe 2003). Isolated fossil feathers have also been found at Nammoura (Dalla Vecchia et al. 2002). Lastly, there is a report of a theropod dinosaur manus from Nammoura; however, its whereabouts are currently unknown and it is considered lost to science (Dalla Vecchia et al. 2002).

The vast diversity of organisms and the exceptional fidelity of preservation of the fossils found in the Upper Cretaceous Lagerstätten of Lebanon have been a subject of fascination for over 1600 years. However, there is still much that is unknown about the organisms and the palaeoenvironments of the Lebanese Lagerstätten. For some taxa, their most recent descriptions are from over a hundred years ago (e.g. Lewis 1878; Woodward 1901; Hay 1903), and several descriptions are not in English, creating a potential barrier to dissemination of results. With the application of modern analytical techniques (electron microscopy, UV investigation, chemical analyses, etc.) detailed anatomical re-evaluation of taxa can be undertaken, greatly expanding our knowledge of the anatomy of Cretaceous species, and helping to elucidate their phylogenetic relationships and modes of preservation. Furthermore, the diversity of organisms found in the Lebanese Lagerstätten (and the large number of taxa that are endemic to these Lagerstätten) allows for large dataset investigations. For instance, network analyses have recently been utilized to uncover food web structure in well-preserved fossil communities (e.g. Swain et al. 2021), and considering how many taxa are known from the Lebanese localities, especially Haqel and/or Hjoula, such investigations could help clarify trophic interactions in the Cretaceous.

Moreover, despite the amount of fossiliferous material available for study in museums around the world, few studies have investigated the depositional palaeoenvironments of the Upper Cretaceous Lebanese Lagerstätten. Site-specific investigations are required, including detailed geological mapping, sedimentological analysis, chemostratigraphy and taphonomic work. This work would constrain the depositional environments and could examine untested hypotheses commonly cited in the literature, such as the almost 50-year-old red tides kill mechanism hypothesis (Hemleben 1977). These investigations should also be used as an opportunity to develop local Lebanese palaeontology, collaborate with Lebanese palaeontologists and celebrate the magnificent Lebanese Lagerstätten.

The Arabic abstract was largely written by S. Stephanos, and edited by S. Stephanos, H.G., T.E.H. and M.B. with contribution from F. Saleh. We would like to thank A. Clements, E. Dunne, E. Bernard, Z. Hughes, S. Maksoud, S. Wheatley and J. Knüppe for their assistance with development of the paper. Thanks go to P. Donoghue for inviting this review, and to F. Saleh for a constructive review of the text, figures and Arabic abstract. NHMUK photographs copyright belongs to The Trustees of the Natural History Museum. This is Paleobiology Database official publication number 490

HG: conceptualization (lead), data curation (supporting), investigation (lead), writing – original draft (lead), writing – review & editing (lead); MB: conceptualization (supporting), data curation (supporting), formal analysis (equal), investigation (supporting), methodology (supporting), writing – original draft (supporting); TEH: conceptualization (supporting), investigation (supporting), writing – original draft (supporting), writing – review & editing (supporting); NA: conceptualization (supporting), data curation (lead), formal analysis (lead), investigation (supporting), methodology (lead), writing – original draft (supporting); PAS: conceptualization (supporting), writing – original draft (supporting); TC: conceptualization (lead), data curation (lead), formal analysis (lead), investigation (lead), methodology (supporting), writing – original draft (lead), writing – review & editing (lead).

H.G. is funded by a University of Bristol PhD Scholarship. T.C. is funded by a Leverhulme Early Career Fellowship (ECF-2019-097). M.B. is funded by a Wallenberg Foundation Postdoctoral Scholarship KAW 2022.0330.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper

All data generated or analysed during this study are included in this published article and its supplementary information file.