During September and October, 1991, a teamrepresenting the American Museum of Natural History, New York City,the British Museum (Natural History), London, and the Ministry of Oiland Mineral Resources of the Republic of Yemen, undertook avertebrate paleontological survey of several areas in Yemen. Fieldpersonnel included the authors of this report, plus Dr Mustafaal-Saruri of the Minerals Exploration Board, Aden Office, Dr MustafaAbdul Latif and Mr Mohammed Abdul Hamid, of the Ministry of Oil andMineral Resources, Minerals Sector, Sanaa, and other Ministry staffas assigned. The fieldwork was sponsored and facilitated by HisExcellency Ali Gabr Alawi, Deputy Minister of Oil and MineralResources of the Republic of Yemen and Chairman of the MineralsExploration Board, and Mr Othman Noman, Vice-Chairman of the MineralsExploration Board.
Our field research was initiated followingthe discovery by Mr Jamil Saif Raweh Al Dobai of a well-preservedfossil frog impression at the locality of Ar Rhyashia, southeast ofSanaa (see accompanying map). The Ar Rhyashia site represents afreshwater interbed in the Yemen Volcanic Series, the principalstratigraphic unit of north Yemen, and the anuran discovery was thefirst significant vertebrate fossil to be found in this stratigraphiccontext. The initial plan was simply to investigate Ar Rhyashia andthe surrounding region in detail, and to collect at fossiliferouslocalities; but by the time that fieldwork commenced the union of theformer YAR and PDRY had opened up new paleontological prospects inthe southern part of the combined Republic of Yemen. Accordingly,several stratigraphic contexts of potential interest were identifiedin southern Yemen, and prospection of these was added to theobjectives of an extended expedition.
The expedition's itinerary is shown in themap, and covered a wide area in both northern and southern Yemen. Our initial prospections were undertaken, as planned, in the regionof Ar Rhyashia. Three freshwater interbeds were identified in theYemen Volcanics at Ar Rhyashia: all were limey shales partlymetamorphosed by overlying basalt flows. A layer was identified thatcontained numerous anuran and other impressions, and some tencomplete anuran skeletons and one fish were collected in it. The ArRhyashia interbeds lie stratigraphically high in the Yemen VolcanicSeries, and we estimate them to be of earliest Miocene age, upwardsof 20 million years old. Samples of an underlying basalt werecollected for potential argon-argon dating, but proved contaminated. More extensive collecting and prospection at Ar Rhyashia were plannedthan were eventually achieved, since problems with unfriendly localelements rapidly curtailed our activities in the area.
The main object of our prospections was tolocate Tertiary and Quaternary mammal sites, which previousreconnaissances had failed to identify in northern Yemen. We thusshifted our attention southwards, since several early Tertiarysedimentary basins had been mapped along the southern Yemen coastbetween Aden and Mukallah. The first region to be prospected was theHamarah Basin (no. 4 on the map), approximately 300 km northeast ofAden. Here rocks of the Shihr Group are exposed: well-cemented redOligocene sandstones that unconformably overlie mildly gypsiferousand less well consolidated yellow sandstones of Eocene age (HadramawtGroup). Fossil worm casts were identified in the Oligocenesediments, but no bone was found.
Next prospected were rocks exposed in thesoutheastern and southwestern parts of the Mayfa'ah Depression. Inthe former area, yellow playa clastics of the middle Eocene HabshiyaFormation near Al Qurayn (no. 5 on the map) produced plant fossilsand invertebrates, plus some scraps of vertebrate bone. In thelatter, the red sandstones of the Miocene Irqah Formation, andunderlying yellowish sandstones probably belonging to the OligoceneFuwwah Formation, both proved unfossiliferous.
Rocks of the Wadi Fuwwah (no. 6 on the map)and of the nearby Wadi Ambakhah, near Mukallah, included poorexposures of the Habshiya Formation. These were unfossiliferous, aswere the Oligocene and possibly earlier sandstones and siltstones ofthe Wadi Brum (no. 7 on the map), 10 km to the southwest of Mukallah,and the lagoonal marly shales of the Wadi Buwaysh (no. 8 on the map),20 km to its northeast. Similarly unproductive were the greennearshore shales of the Paleocene Umm Ar Radamah Formation (no. 9 onthe map), about 100 km northeast of Mukallah, and deposits of theShihr Group in the Wadi Huwayrah (no. 10 on the map, page 13), 50 kmfrom Mukallah. The same disappointing situation prevailed in theShihr Group deposits exposed near 'Irs (no. 11 on the map) in theWadi Haj'r, and in the Jurassic Kohlan Sandstones of the sameregion.
More fossiliferous were the rocks exposed inthe Wadi Kaninah, where the middle Eocene Habshiya Formation is wellexposed. These are nearshore marine sands and sandy silts with ahematitic surface litter and numerous fossil wood fragments and otherplant fossils. Mollusks and gastropods were found here as well aswood. To the east of Kaninah village (no. 12 on the map), furtherexposures of the poorly consolidated yellow nearshore sands of theHabshiya Formation, stratigraphically above the principal fossil woodzone, yielded Turritella and oyster fossils, plus some more fossilwood. These rocks also produced our first discoveries of significantvertebrate bone in this stratigraphic context, including parts ofteleosts, rays, chelonians, cetaceans and sirenians.
Everywhere they were exposed in the Mayfa'ahtrough, sediments of the early to middle Eocene Jiza and RusFormations proved unfossiliferous. However, in this area theKaninah-Jizwal sequence (no. 13 on the map) of the later HabshiyaFormation did produce some plant and invertebrate fossils. In theWadi Haj'r, limestones of the Shuqrah Formation (early late Jurassic)are exposed above the Kohlan sandstones. Near Jabal Zulma Ba-Thalab(no. 14 on the map), 40 km to the north of the main Aden-Mukallahroad, a partial crocodylian skeleton was found. Unfortunately, thissignificant find was embedded in a block that had fallen from high ona cliff face, so extensive prospection of the deposits from which ithad come was not practicable.
1. Zulma Ba-Thalab, southernYemen.
At this locality in the Wadi Haj'r, some40 km to the north of the Aden-Mukallah road, a crocodylian fossilwas found in limestones of the early late Jurassic Shuqrah Formation. This 165 million year-old partial skeleton consists of parts of theposterior postcranium of a teleosaurid thalattosuchian crocodile andincludes the diagnostic osteoderms. It is the oldest vertebratefossil currently known from the entire Arabian peninsula, and ischiefly of importance on this account. This marine crocodile,estimated to have been about three meters long when alive, belongs toa family which is common in early late Jurassic marine depositsworldwide. The block of limestone in which it was found embedded hasbeen recovered and transported to the premises of the MineralsExploration Board in Aden, where it awaits study and installation ina proposed museum.
2. Ar Rhyashia, northern Yemen.
At Ar Rhyashia, approximately 23 km south ofRada', limey and partly metamorphosed shaley freshwater interbeds inthe Yemen Vocanic Series have yielded a fauna consisting principallyof well-preserved anuran skeleton impressions. This is the firstsignificant vertebrate fauna to have been discovered anywhere in theYemen Volcanics, the principal stratigraphic unit of northern Yemen. Circumstances unfortunately prevented detailed mapping of the ArRhyashia localities, and basalt samples collected for argon-argondating were, as noted, contaminated. However, the Ar Rhyashiyainterbeds lie stratigraphically quite high in the Volcanic Series,and are estimated to be of earliest Miocene age. Some ten anuranspecimens in an excellent state of preservation have been collected,and have been submitted to Ms Amy Henrici of the Carnegie Museum ofNatural History for detailed study. This study has not as yet beencompleted, but preliminary analysis suggests that all specimensbelong to the same frog species, an undescribed member of Pipidaethat is apparently closely related to species of the living genusXenopus. One fish skeleton was also collected, and is currentlyunder study by Dr John Maisey of the American Museum of NaturalHistory. It appears to be a neopterygian amiid or a teleost ofundetermined affinity; more certain identification must await furtherstudy.
Rocks of the Paleocene through EoceneHadramawt Group are widely if not extensively exposed in a series ofdepressions that lie perpendicular to the shoreline in the Shawba andHadramawt Governorates of southern Yemen. Of these rocks, only thoseof the middle Eocene Habshiya Formation proved at all fossiliferous,and then only at two localities among the many that were prospected. Near Al Qurayn, in the southeastern part of the Mayfa'ah Depression,Habshiya exposures yielded some plant and invertebrate fossils, butonly scraps of vertebrate bone. More important was a locality eastof Kaninah village, which lies in the Wadi Kaninah, approximately 25km north of Jawl Ba-Hawa, the principal town of the Wadi Haj'r. Herea coarse yellow unconsolidated sandstone of nearshore facies producedplant, invertebrate and vertebrate fossils.
Plant fossils comprise the first fossilflora known from southern Yemen, and include internal casts of fruitsof a genus cf. Anona (custard apples or magnolids), fruitsthat may be in the waterlily family, and a fruit possibly in theCucurbitacea (cucumbers), as well as numerous fragments offossil wood. Together these fossils indicate the proximity of atropical rainforest, suggesting an onshore environment similar tothat inferred from isolated mammal teeth for the late Eocene ofsouthwestern Oman, and equivalent to the renowned Eocene-OligoceneFayum fauna and flora of Egypt.
Invertebrates include oysters andTurritella, and a substantial list of vertebrates includes theteeth of a ray (possibly Myliobatis), cranial and dentalelements of large teleost fish, chelonian bones, dugong ribs andscapula, and a cetacean rib. This fauna includes the first fossilmammals known from anywhere in Yemen. All elements of the flora andfauna agree with the notion that this facies of the HabshiyaFormation was deposited in a nearshore marine environment thatinterfaced with a tropical rainforest environment onshore via amarshy, lagoonal shoreline.
The reconnaisance reported here has producedthe oldest vertebrate fossil yet known from the Arabian peninsula,the only vertebrate fauna yet known from deposits associated with theTertiary Yemen Volcanic Series, the first mammal fossils known fromYemen, and the first fossil flora to be reported from southern Yemen. It has also pointed towards the best prospects for futurepaleontological prospection in Yemen by mammalian paleontologists. Our surveys indicate that onshore and nearshore deposits of earlierTertiary age become thicker and better-exposed as one moves northeastalong the southern Yemen shoreline, toward Oman. Given thatnearshore marine deposits in the area of Oman that borders on Yemenhas recently started to produce mammal teeth that are comparable withthose found in the remarkable late Eocene to Oligocene deposits ofEgypt's Fayum, we would predict that the best prospects that Yemenoffers for significant finds in Paleogene mammalian paleontology liefurther to the east than we were able to explore on this occasion, inthe area between Ash Shihr and the frontier with Oman. This area isremote, little visited, logistically difficult, and geologically notwell known; but extrapolation of the observations we made along thecoast to the southwest of Ash Shihr suggests that exploration of thisregion would be highly worthwhile.
The authors wish to thank the NationalGeographic Society for its generous financial support of the researchreported here. We would also like to record our gratitude to DeputyMinister Ali Gabr Alawi and Deputy Chairman Othman Noman for theirenthusiastic sponsorship and facilitation of this work. Our othercolleagues at the Ministry of Oil and Mineral Resources, especiallyDr Mustafa As-Saruri, but also Mr Mohamed Mukrid, Mr Mohammed AbdulHamid, Dr Mustafa Abdul Latif, and Dr Iqbal Al Jailani, are alsodeserving of our warmest thanks, as is Dr Hamed El-Nakhal, of SanaaUniversity, whose advice and support have been invaluable throughoutour researches in Yemen. And yet again we wish to register ourappreciation to the American Institute for Yemeni Studies and to DrJeffrey Meissner for providing logistical support and invaluableadvice.