Testudo graeca

Linnaeus, 1758
Spur-thighed tortoise

The rounded carapace (generally to 20 cm in graeca, but almost twice as large in ibera) is domed, highest behind the center, with abruptly descending sides, a broad cervical notch, and the posterior rim slightly to greatly flared and slightly serrated. The cervical scute is generally long and narrow. Vertebrals are broader than long; the 5th vertebral is expanded, but narrower than the 4th (the 5th vertebral is much broader than the 4th in T. hermanni). Vertebral and pleural areolae are rarely raised, and the shell appears smooth. Eleven marginals lie on each side, and a single undivided supracaudal scute is present; in some individuals the supracaudal is divided. Carapacial color varies from yellow or tan with black or dark-brown blotching to totally gray or black. The plastron is well-developed. Its forelobe is not or only slightly upturned and gently tapers toward the front; it is shorter and narrower than the kinetic hindlobe, which bears a posterior notch. The plastral formula is: abd > hum >< gul >< an >< pect > fem; the paired gular scutes are distinctly thickened with an inner epiplastral pocket, but do not or only slightly extend beyond the carapacial rim (in T. g. armeniaca the gulars apparently always extend beyond the carapacial rim). The bridge is broad, with a small axillary and one or two small or moderate-sized inguinals that do not touch the femoral. Plastron and bridge are yellow to greenish yellow, brown, or gray, with some dark-brown or black markings. The head is moderate in size with a nonprotruding snout and a weakly hooked, tricuspid upper jaw. Its prefrontal is rarely divided longitudinally, but the large frontal may be subdivided; other head scales are small. The head varies from yellow to brown, gray, or black, with or without dark spotting. Three to seven longitudinal rows of large, overlapping scales, covering well-developed osteoderms, lie on the anterior surface of each foreleg. The heel may lack spurlike scales, but each thigh has a large conical tubercle. The tail lacks a large terminal scale. Neck, limbs, and tail are yellowish brown to gray.
Mathey (1931) reported the karyotype contained 54-60 chromosomes, but the diploid chromosome number is 52, as in other Testudo (Bickham and Carr, 1983).
Males have longer, thicker tails with the vent beyond the carapacial rim.

Testudo graeca ranges from southern Spain across northern Africa from Morocco to Libya, and from the Balkan countries through Greece and Turkey eastward to Iran and through Syria, Lebanon and Jordan southward to Israel; records from Egypt are in error (Buskirk, 1996). This tortoise also occurs on Mallorca (Bruekers, 1995), Cyprus and several Aegean islands, and apparently has been introduced to the Canary Islands, Sardinia, mainland Italy and Sicily (Iverson, 1992).

Geographic Variation
Two main groups are often recognized within the Testudo graeca complex—a group from North Africa and southern Spain, and a group from Eurasia and the Middle East.
The North African group usually is considered monotypic, consisting only of Testudo graeca graeca. This tortoise is characterized by a cream to straw-colored (sometimes reddish, depending on the substrate) ground color, a plastral pattern that is usually rich in contrast, with well-delimited and symmetrical dark blotches, a black and yellow head pattern (sometimes lacking black pigment), and a scutellate 1st vertebral. Highfield and Martin (1989b, c) resurrected Testudo whitei Bennett, 1836, a large (to 29.2 cm) tortoise from Algeria, and described the small (rarely exceeding 13.5 cm) Libyan T. flavominimaralis (which Andy C. Highfield, pers. comm., now considers invalid). Bour (in David, 1994) and Perälä (1999) examined the whitei type and concluded it does not correspond to the large tortoises depicted in Highfield and Martin (1989b). However, Perälä (1999) postulates that "future research will most certainly verify that this form deserves taxonomic recognition". Highfield (1990) created a new genus for another small (to 13.4 cm) tortoise from Tunisia, Furculachelys nabeulensis, based on a supposed unique suprapygal configuration, in which the first element bifurcates posteriorly to embrace the second. Perälä (1999) verified the morphological distinction of this population, but also demonstrated that the suprapygal structure is neither constant nor unique for tortoises from Tunisia and eastern Algeria, and that nabeulensis is a junior synonym of Testudo mauritanica Duméril and Bibron, 1835.
The Eurasian and Middle Eastern group differs from the nominate subspecies in having a horn-colored to brownish ground color, a usually inconspicuous plastral pattern (with asymmetrical blotches if the pattern is more prominent), a uniformly colored head (but with a black 'mustache' on the rhamphotheca in terrestris), and a 1st vertebral scute with rather straight lateral sides. The general appearance of legs and head is rather stout when compared to the more slender T. g. graeca. The Middle Eastern spur-thighed tortoise Testudo g. terrestris Forsskål, 1775 occurs in Israel, Lebanon, Syria and Jordan. It is small (to 20 cm), elevated dorsally, has some expansion but little flaring of the posterior marginals, and yellow pigmentation on the side and top of the head and on the limbs. A population from Antakya, Turkey, was described as T. antakyensis Perälä, 1996. Tortoises from this area have a conspicuous, centrally divided horny tubercle on the tail tip (not claw-like and longish as in hermanni) and the 5th vertebral typically broader than any other vertebral. Fritz et al. (1996) synonimized antakyensis with terrestris, but a recent morphometric study by Perälä (1999) shows that terrestris as currently recognized consists of several distinct taxa. T. g. ibera Pallas, 1814, the Asia Minor spur-thighed tortoise, occurs from the central Balkans to the Black Sea, Turkey, and the southern Caucasus to Iran. It is larger (to 35 cm), is slightly flattened dorsally, has some expansion and slight flaring of the posterior marginals, and a unicolored tan or gray to blackish head. T. g. anamurensis Weissinger, 1987b, the Anamurum spur-thighed tortoise, is found on the southwest coast of Turkey from the Bay Mountains in Antalya eastward toward Mersin. With a mean size of 22.7 cm (Jarmo Perälä, pers. comm.) this is a large race, distinguished by a long, trapezoid-shaped, very flattened carapace (height 45.9%, mid-body width 67.5% of carapace length; Jarmo Perälä, pers. comm.) with scattered small blotches and strongly flared posterior marginals, a dark head with yellow marks, and a cream-colored chin and throat. The status of T. g. anamurensis seems equivocal, as Bruekers found both ibera and anamurensis occupying the same habitat near Side (Bruekers, 1997) and Alanya (Jaco Bruekers, pers. comm.). The Iranian spur-thighed tortoise T. g. zarudnyi Nikolskii, 1896 occurs in eastern and southern Iran, and possibly in adjacent Afghanistan and Pakistan. It is large (to 27.5 cm), elongated, elevated dorsally, with much expansion, flaring, serration, and slight upturning of the posterior marginals, and a plain gray-brown head. Nikolskii's spur-thighed tortoise T. g. nikolskii Chkhikvadze and Tunijev, 1986 is restricted to the northwestern Caucasus region. Its domed, convex-sided carapace reaches 21.0 cm in length and has a low, somewhat, pointed keel on each vertebral, four rows of large, rounded scales on each foreleg (pointed in T. g. ibera), and long, pointed, clear claws on each forefoot (short and black in T. g. ibera). T. g. armeniaca Chkhikvadze and Bakradze, 1991, the Araxes spur-thighed tortoise, is found along the Araxes (Araks) Valley west of Megri at the base of Mount Ararat near the Caspian Sea in Armenia. The rounded yellow-brown to dark brown or black carapace (to 23 cm) has a pattern of dark spots if light colored or dark and yellow lines if darker, and is flattened (height less than 50% of carapace length), with no marginal flaring laterally and only a slight amount posteriorly. The head is dark brown or black. T. g. armeniaca shares several features generally associated with T. horsfieldii, and according to Perälä (1999) "this taxon can certainly not be viewed a Testudo graeca, and it ought possibly be classified into a new genus".
Furthermore, Chkhikvadze (1989a) listed T. g. pallasi from Dagestan (a nomen nudem) and Bodenheimer (1935) proposed the name Testudo floweri for the tortoises inhabiting southwestern Israel, describing them as "the tiny variety found in the Negeb...". Wermuth (1958) placed T. floweri in the synonymy of T. terrestris, but Bour (1988b) resurrected the name as a full subspecies of T. graeca without explanation, and David (1994), apparently on Bour's advice, assigned it to subspecific level under T. terrestris. Contrary to the opinion of Crumly (1988), the single adjective "tiny" can not be accepted as a description; therefore, the name T. floweri is a nomen nudum, and can not be used as a valid name for any tortoise population.

North African Testudo g. graeca occurs from sea level to approximately 1,900 m in the Atlas Mountains. Generally it occupies semidry grassland and brush areas, but also can be found among coastal dunes, marshland borders, rocky, brushy hillsides, and pine woods. Eurasian and Middle-Eastern tortoises live from sea level to about 2,700 m on plateaus and mountains. They most often occur on dry open steppes, barren hillsides, and wastelands where vegetation varies from sea dune grasses to scrub thorn or dry woodlands.

Natural History
In the northern part of their range, North African tortoises hibernate until March or early April. A reversed activity cycle occurs in the south, where they aestivate throughout the summer. Mating occurs in April-May and again in the fall, but this is highly variable according to locality and altitude. Nesting takes place in May or June; more than one clutch may be deposited each season. Clutches contain 3-4(7) eggs; large Algerian tortoises ("whitei ") lay clutches of 12-14 eggs (Andy C. Highfield, pers. comm.). The 30-35 mm hatchlings emerge in the fall.
Eurasian and Middle Eastern tortoises generally nest in May or June, but nesting has been observed from April to July. During courtship, the male butts and taps the female; while mating, he emits high-pitched sounds (Baran and Atatür, 1997). The eggs of Turkish T. g. ibera are large (36 x 30 mm) compared to those of North African T. g. graeca, and oval rather than spherical (Highfield, 1992). Breeding of captive Israel T. g. terrestris was studied by Lapid and Robinzon (1996). Reproductive activity and oviposition depended on environment temperature, with nesting in April-June. As many as 3 clutches of 1-7 (average 4) eggs were produced a year. Male plasma testosterone levels were high the entire study period, but levels about doubled in July (2 ng/ml to 4 ng/ml), and continued to rise to a mean peak of 12.8 ng/ml in November. In females, plasma estradiol values (4.1-70.2 pg/ml) were much lower than those for testosterone in males, and no clear seasonal pattern was evident, except a peak in August. The 30-35 mm young hatch in late summer or fall (Nikolskii, 1915), and by October-November both adults and hatchlings are entering hibernation (Highfield, 1992).
Over its range, T. graeca consumes many varieties of herbaceous plants and grasses.

Taxonomic Comment
David (1994) considered T. graeca monotypic, elevated T. terrestris to specific status, and assigned the former subspecies of T. graeca (ibera, zarudnyi, nikolskii, anamurensis) to it. He credits these changes to Bour (1988b: 14), but Bour listed these and terrestris as races of T. graeca, not T. terrestris. However, David's 1994 arrangement may be accepted as a reflection of Bour's re-evaluation of the genus as of 1994. The splitting of Testudo graeca into the species T. graeca and T. terrestris, and the subsequent division of the subspecies of T. graeca between them has not been widely accepted, partly because the name terrestris is controversial. Fritz et al. (1996) consider color differences between T. g. graeca and T. g. terrestris to represent merely adaptations to local conditions, such as substrate color, according to Gloger's rule, and not as differentiations of systematic importance. However, the significant morphological variation found by Perälä (1999) indicates that a taxonomic revision is inevitable.
James R. Buskirk (pers. comm.) states that there are several problems with terrestris as an acceptable species epithet for any variety of Testudo: "Wermuth (1958) placed Forsskål's Testudo terrestris in the "Libanon-Gebirge" [= mountains of Lebanon, which he incorrectly placed inside Israel], but Gasperetti et al. (1993) retraced the itinerary of Forsskål and concluded it did not include any portion of Lebanon, Palestine or Syria. Anderson (1896), likewise familiar with Forsskål's itinerary, interprets him having mentioned—not described—a species of terrestrial chelonian from southwestern Arabia, from today's Yemen. In other words, Forsskål was admittedly passing on incomplete information, and his choice of language (according to Latin scholars L. Sussman and P. McBreen of the University of Florida) in that paragraph (Forsskål, 1775, p. 12) is so poor that comments about Testudo marina and T. terrestris are difficult to separate. There is no type locality other than Wermuth's (1958) misplaced "Libanon-Gebirge", nor is there a type specimen of T. terrestris. Furthermore, it is not clear whether Forsskål saw "land turtles" in Cairo [= "Kahirae", misunderstood as Kabirae by Wermuth (1958) and as Kabira(?) by Gasperetti et al. (1993)] or in Yemen. In as far as Forsskål's troubling passage refers to a non-marine chelonian, i.e. Testudo terrestris, he may have been the very first European to behold Yemenite Pelomedusa subrufa". Highfield (1998) too commented on this topic and is currently preparing a proposal for suppression of the name terrestris (Andy C. Highfield, pers. comm.).

IUCN Red List Status (1996)
Vulnerable (A1cd). Testudo graeca nikolskii is considered Critically endangered (A1abcde+2bcde).