Information about activities. How to
                    join the TT. Discussion list, for members and others! Species-specific Care sheets Articles
                    from our Newsletter. Links to
                    important sites. Links
                    to members' sites. Search our
                    site! What has
                    been updated, and when.


Keeping & Breeding Tortoises in Captivity (1990)

Species Data and Captive Breeding Chapter

By A. C. Highfield

This is the original text from the 1990 edition of Keeping & Breeding Tortoises in Captivity. 

Note: There have been a number of taxonomic changes since this original text was published. See updated book and other articles on this website.

The information which follows is intended as a concise and practical guide to the basic requirements of some of the more commonly encountered terrestrial species from both temperate and tropical climates. It is not intended as a ‘complete’ guide to the maintenance and breeding of these species; it should however provide a good starting point and serve as a basis for further experimentation and progress in their captive husbandry. Most of the data presented is based upon actual records obtained from successful breeding groups and field studies of the species in the wild. Under the heading of each species or genus I have included a short taxonomic note. This is necessarily very brief, although in some instances the status of the taxon may be anything but straightforward. Where this is the case, I have included (I hope) sufficient information to enable identification for captive breeding purposes and enough description to avoid inadvertent hybridization between questionable forms.


General observations: Members of the genus Kinixys are distinguished by their uniquely hinged carapace which allows the rear of the shell to close giving additional protection to the tail and legs; a character which sets them apart from all other terrestrial tortoises.

Taxonomy: The status of Kinixys species is, at the present writing, in a state of considerable flux. Until only recently three species had been recognised by most authorities; Kinixys belliana, K.    homeana and K. erosa.   These last two being clearly distinguishable from the former. Kinixys belliana was assumed to have a wide distribution, from Senegal and northern Cameroon, via southern Africa to Madagascar; as is often the case where wide distributions are assumed, this in fact appears not to be the case. This apparently extensive distribution is illusory, and what passed for ‘identical’ “Kinixys belliana” in much of its assumed range actually comprises populations formed not only of several different subspecies, but also no less than 3 additional full species! In many respects, a situation which finds a parallel in North Africa with regard to Testudo graeca. The current checklist for Kinixys (based upon the pioneering work of Donald G. Broadley with this genus) therefore includes Kinixys homeana, K. erosa, K. belliana belliana, K. belliana nogueyi, K. belliana zombensis, K. spekii, K. lobatsiana and K. natalensis. It is quite likely that as studies progress additional forms will be identified. From the captive breeding point of view it is readily admitted that such a profusion of species creates some difficulties - not the least of which is obtaining a satisfactory identification of specimens. Fortunately, those species most likely to be encountered under captive conditions, K. homeana, K. erosa and K. belliana belliana are relatively easy to identify and in the case of unusual specimens study of specialist works on South African chelonia should provide the answer. Description: The maximum dimensions recorded for male K. belliana belliana are SCL (straight carapace length) 206mm and for female specimens 217mm. The carapace is a dull yellow-buff color with darker brown or reddish-brown central scute markings. The upper posterior section of the carapace is gently rounded. Kinixys homeana and K. erosa are very different from K. belliana in overall appearance, being an overall reddish-brown color with lighter rather than darker centers to the scutes, and in both cases these tortoises are much more sharply angular at the rear portion of the carapace than K. belliana. This characteristic is much more marked in K. homeana than in K. erosa and may be used to separate the two; in K. homeana there is typically a vertical descent from the 5th vertebral scute to the supracaudal. Kinixys erosa is the largest of the hinge-backs, with males often exceeding 315mm and females 250mm. K. homeana is a smaller species, with males rarely in excess of 208mm and females 220mm. Captive environment: It is not easy to generalize the requirements of Kinixys specimens, as they inhabit a surprisingly wide range of disparate biotypes from coastal plains and savannah to the edges of forests and swamps. Without knowing the origin of specimens it can therefore be extremely difficult to provide the sort of environment to which the animal is accustomed. Some experimentation is usually required, the environment being adjusted in accordance with the tortoises responses. As a starting point, a relatively high humidity should be provided with a daytime temperature in the order of 24-27 oC. Water should be available at all times, as Kinixys species like to soak and even swim on occasions. Those from forest habitats (e.g. K. erosa and K. homeana) prefer more shade than those from open or savannah habitats, but all tend to be secretive and appreciate a secure retreat area.  Most Kinixys tortoises tend to be more active during wet and rainy weather and in southern Africa K. belliana is known to aestivate during very dry periods. These tortoises frequently also show a tendency to feed and become active at dawn and dusk. In good weather Kinixys species can be maintained outside successfully in both Europe and the U.S.A. An outdoor pen is generally satisfactory provided it is well planted, kept moist, and has plenty of shade. A heated indoor retreat may however be required on all but the very warmest of summer nights. In cool weather these tortoises are best maintained under tropical greenhouse conditions. In South Africa, K. natalensis hibernates during the winter months. Diet: Kinixys species are typically omnivorous with a marked preference for mushrooms, slugs, snails, mixed fruit (especially banana), and earthworms. A good variety of green plant leaf material should also be provided. In a suitable outdoor enclosure the tortoises will find a high percentage of their food for themselves, spending a considerable time engaged in hunting down snails, worms and fresh shoots of weed. Breeding: Male Kinixys can be very aggressive and it may not be possible to maintain more than one per enclosure. K. erosa appear to be the most aggressive of all, and two males will sometimes inflict serious injuries upon each other if allowed to. Mating can likewise be a vigorous affair. Female K. belliana may lay more than one clutch during the summer, often at intervals of about 5-8 weeks.  A clutch typically comprises between 4-8 eggs, occasionally as many as 10 in the case of large individuals. The eggs are very elongate and usually measure approximately 38mm X 48mm and range from 23g to 32g in weight. There are inexplicable size variations in the literature pertaining to the eggs of Kinixys belliana and it seems probable that the various populations (and subspecies) in all likelihood produce eggs which are significantly different when considered statistically in a comparative sense. The hatchlings of K. belliana typically measure some 38-47mm SCL and weigh circa 17-25g at emergence - again there are significant differences between the different populations. The carapace hinge is not evident in hatchlings and juveniles. Artificial incubation is most successfully carried out at circa 30.5 oC at medium-high humidity (recommended 75-80%) in either a type I or II incubator. At 30-31 oC the incubation period is typically 90-110 days. In the wild the incubation period is variable according to the time of oviposition; it can take anything up to a year. At lower incubation temperatures the eggs may remain viable, but the incidence of dead-in-shell (DIS) increases and incubation time becomes much extended; in their excellent ‘South African Tortoise Book’ Richard Boycott and Ortwin Bourquin give an example of one egg which took 10 months to hatch when incubated at 25 oC and another egg which took 12 months. This latter publication is, incidentally, highly recommended to anyone keeping South African species in captivity.


General observations: This is one of the most common South American tortoises, with a wide distribution from Panama to northern Argentina. The descendants of introduced specimens now colonize a number of Caribbean islands although the population of Trinidad may be a natural one. The species is named on account of the bright red scales which adorn its legs in profusion. Taxonomy: At the higher level, there is much dispute about whether this species should be attributed to Geochelone or whether the sub-genus Chelonoidis should be elevated to generic rank to include them. Thus is it not unusual to see them listed under both headings. With a wide and fragmented distribution, G. carbonaria demonstrates considerable morphological variation. This is evident not only in the carapace shape but also in coloration. However, there are no currently recognised subspecies and at present the species is considered homogenous. Even so, in view of the variation already noted caution should be exercised in captive breeding and preferably this should be restricted to specimens from similar geographical origins.

Description: A large and impressive tortoise, mature males typically measure some 300mm SCL and most of the females I have measured appear to be of similar size although much larger examples of both sexes have been reported. The principal dimorphic feature is the narrow ‘waist’ of males, which according to some observers are reminiscent of an overgrown peanut (plate xx). The tail of male specimens is also longer than that of females, and in males also the rear lobe of the plastron exhibits a much wider and flatter anal notch. Color does vary according to locality of origin, with Venezuelan specimens exhibiting a yellowish plastron with a browner carapace and Argentinean specimens a darker, almost black plastron and carapace. In all cases the central areola of the scutes are marked with yellow. The head is typically brightly marked with yellow scales, but some Venezuelan specimens possess instead a reddish-orange head. Captive environment: Red-foot tortoises inhabit grassland savannah and drier forest habitats throughout South America. In only a few localities does it impinge upon true tropical rainforest habitats.  In captivity, a moderately dry environment and temperature in the region of 21 C (night) to 27 C (day) will suit most specimens. Access to water should however be provided at all times. In most of Europe and North America Red-foot tortoises can be accommodated out of doors during the warm summer months, at least during the daytime, but in the winter and on cool nights indoor accommodation is also essential.     A well planted outdoor pen with access to a heated hut or greenhouse is probably ideal for this species which is relatively easy to manage in captivity. Diet: In the wild these tortoises prefer fallen fruits, flowers, green leaves and will occasionally take carrion when it is encountered. In captivity caution should be exercised when offering banana as they easily become addicted to it and meat products should similarly be very strictly rationed. Over reliance upon both items can result in serious dietary problems developing. Most Red-foot tortoises will readily take melon, orange, mango, hibiscus flowers, mulberry leaves,  vegetable greens and grapes.

Breeding:   The mating ritual of Red-foot tortoises is startlingly similar in many respects to that of T. horsfieldii; the male advances upon the female, circles her repeatedly, extends his neck near her face whereupon he proceeds to move his head rapidly from side to side with a strange jerky motion. This is accompanied by occasional biting, shoving and ramming actions. Eventually he mounts from behind and during copulation emits an extraordinary sequence of ‘clucking’ or ‘cackling’ sounds. Two male tortoises will often engage in combat, and most mating activity occurs either during or just after rain. Nesting and egg laying continues all year. The Red-foot tortoises in our own collection regularly lay eggs which measure on average 45mm long by 42mm wide and which weigh some 35-50g grammes. Typically a clutch consists of between 3-5 eggs and a female may lay 2-3 clutches each season.        At 30 C in a Type II incubator at moderately high humidity incubation takes on average 150-175 days. The hatchlings measure approximately 46mm long, are 42-44mm wide and weigh between 26-32g.


General: These tortoises are distinguished from G. carbonaria principally by the bright yellow-orange (rather than red) scales of the legs. They are also light golden-brown in color, and much larger. It is widely distributed in S. America and is found in Brazil, Ecuador, Columbia, Guyana, Venezuela, Peru, Guyana and Surinam. Like G. carbonaria a population also occurs on Trinidad and in much of its range it exists sympatrically with G. carbonaria. Taxonomy: G. denticulata was described earlier than G. carbonaria (in 1766 as opposed to 1824) and in many old accounts is to be found listed under the heading ‘Testudo tabulata’. As with G. carbonaria there is dispute over its generic attribution. It does not demonstrate the regional variation so evident in G. carbonaria. Description: Yellow-foot or Forest tortoises are considerably larger animals than Red-foot tortoises; the average mean adult SCL is 400mm but some very much larger specimens are occasionally encountered.      The record is believed to be circa 700mm.

Captive environment: G. denticulata inhabit true tropical and sub-tropical forest habitats; as such, they require higher humidity and more stable temperatures than G. carbonaria. Continuous access to drinking and bathing water is essential, and a frequent spray with ‘artificial rain’ is also helpful in maintaining good health and normal activity. Day and night temperatures in the order of 25-27 C are recommended.

Diet: As for G. carbonaria.

Breeding: Although superficially similar to G. carbonaria, Auffenberg has pointed out that the head movement differs in G. denticulata, consisting of a single sweep rather than in a series of jerky motions. Male G. denticulata also lack the ‘waist’ of G. carbonaria. Clutch densities have been reported as higher than G. carbonaria, as many as 15 on occasions, although most specimens appear to lay between 4-5. In the wild this tortoise does not appear to nest in the normal way, often leaving the eggs at least partially exposed. The eggs are similar in size to those of G. carbonaria and can be incubated under identical conditions. If anything, humidity should be a little higher. The hatchlings are very similar in both size and overall appearance to those of G. denticulata; so much so that it is not easy to tell them apart. At 30 oC the incubation period is usually between 140-160days.


General: Because of their North American and Mexican location, Gopher tortoises have attracted considerable interest and much scientific investigation; this has resulted in their being among the best known of all land tortoises. However, also because of their proximity to humankind they are under severe pressure from collecting and from habitat destruction. All species of Gopher tortoise are now protected by law and should not be disturbed. Large numbers remain in captivity, both in private and institutional hands and therefore they attract considerable interest from prospective captive breeders. It should be noted that the ‘obvious’ solution of releasing the large captive population back into the wild to replenish declining stocks is both genetically and biologically highly inadvisable. Under no circumstances should captive specimens be simply turned loose. Not only do they pose a potential genetic risk, but recent evidence suggests that they may contaminate already endangered  natural populations with fatal pathogens.

Taxonomy: Until recently it was generally thought that the taxonomy of Gopher tortoises was relatively straightforward; this has in fact proved to be very far from the truth. Four allopatric species are recognised         in most recent accounts; Gopherus agassizi (the Desert tortoise), Gopherus berlandieri (the Texas tortoise), Gopherus polyphemus (the Florida gopher tortoise) and Gopherus flavomarginatus (the Mexican or Bolson tortoise) which was itself only discovered in 1888 and not fully verified until 1959. Very recently however this apparently        simple picture has been completely overturned by studies which suggest that in fact G. agassizi and G. berlandieri should be considered       separately under the generic name Xerobates, with only G. polyphemus             and G. flavomarginatus continuing under Gopherus. Considering that Gopherus (Xerobates) are probably the most intensively studied tortoises in the history of herpetology, that such discoveries are still being made says much for our knowledge of the rest!. Even as this book goes to press, reports (as yet unconfirmed) of the discovery of a possible 5th species of tortoise in Baja, Mexico are being received.                     Not only that, but further (mitochondrial DNA)        results indicate that X. agassizi itself appears to comprise at least three genetic assemblages in discreet geographical zones as divergent from each other as X. agassizi is from X. berlandieri; clearly these tortoises have a far from simple ancestry, and a great deal of work still needs to be done if their phylogeny is finally to be unraveled. Description: Xerobates agassizi; SCL to 380mm, although most much smaller with females smaller than males. This species is principally distinguished by its narrow head and large hind feet.         Carapace coloration           is light chocolate brown. Gopherus polyphemus;  this tortoise is externally somewhat similar to X.   agassizi,   of almost identical color, but the head is much wider and typically the animal is smaller        (circa 250mm although some much larger specimens are encountered). Xerobates berlandieri; this is the smallest of the Gopher tortoises attaining a typical adult SCL of 160mm SCL although some larger specimens are seen (to about 215mm)     . The scutes have a much darker brown ground color than X. agassizi (often virtually black),   but feature prominent yellowish central areolae;      these can however be          less obvious in very ancient specimens.         The gular projection        is very well developed in males,    less so in females. Gopherus flavomarginatus; this tortoise is the largest terrestrial chelonian in north America, and attains the very considerable SCL of 400mm and        weighs in excess of 14kg. It has two very large black-pointed ‘spurs’ on the thighs, a brown and yellow carapace with yellowish-horn colored plastron and the rear marginals are markedly flared and serrate.

Captive environment: All Gopher tortoises are very highly adapted to their native habitat conditions and are by no means easy to maintain successfully - especially outside their natural bioclimatic range. In addition,           the various species do differ considerably in terms            of temperature          and humidity requirements. There are however common threads, one of which is their distinctive habit of residing in burrows - these can be up to 8 meters long in the case of G.     flavomarginatus, and up to          14m long in the case of G. polyphemus (although          X. berlandieri burrows rarely extend for more than 30-40cm). Not only do these provide a means of retreating from unfavourable weather conditions generally, but they provide a microclimate where humidity as well as temperature are relatively stabilised.

 In captivity, all Gopher tortoises tend to do badly if overcrowded and if mixed at random with other species. Ideally,  they should       be provided with an outdoor area of light but excavatable soil, lightly planted with scrub, grass and succulents. An overnight shelter        is essential,                 a small wooden hut is generally satisfactory if provided with a ramp for exit and entry. An even better option is to provide a natural or artificial burrow. Damp environments are categorically not suitable for Gopher tortoises - with the possible exception of         X. berlandieri          which I have found to positively thrive under more humid conditions and G. polyphemus which also prefers a relatively high level of ambient humidity . Cold and damp however are definitely to be avoided. In general however, most Gopher tortoises prefer it hot and dry, and will readily take advantage of both natural sunlight and artificial basking facilities. The only moisture in desert tortoise accommodation should comprise the drinking pool - this should  be 25-35mm deep and the water changed daily. Gopher tortoises hibernate during the winter (usually from November        or December through March). The best place for such hibernation is a natural or         artificial burrow. Temperatures for hibernation             as described previously and are similar to those for T. graeca and         T. hermanni. Diet: In their native habitat Gopher tortoises are virtually 100% herbivorous, with X. berlandieri alone showing some interest in the occasional snail and chance piece of carrion. The bulk of their diet consists of grasses and low growing herbs, flowers, succulents and cacti;      Opuntia pads and fruits are especially favoured but grasses generally comprise more than half of the total dietary intake.                In captivity it is vitally important that the basic chemical balance and profile of the natural diet is followed as closely as possible, even if exactly the same constituents may not be available (see Appendix I). Breeding: In such a diverse group of animals as the Gopher tortoises it is hardly surprising that considerable variation is found between the various forms in respect of their reproductive biology; X. agassizi typically lays clutches of 4-8 eggs (sometimes as many as 14)   each approximately 39mm x 44mm. Incubation in the wild takes approximately three and a half months. In captivity, they can be successfully incubated artificially at 30-31 C in a type I incubator at low       to medium humidity which typically results in emergence at 80-130 days. The hatchlings typically measure about 48mm SCL and weigh circa 28g. X. berlandieri have a lower average clutch density typically depositing 3 or 4 eggs - sometimes as few as 1, just occasionally       as many as  5. The eggs from a single clutch may be also laid in different nests,  singly. These eggs are quite different from those of        X. agassizi and when freshly laid are somewhat pliable.         They rapidly harden to normal consistency, are markedly elongate and usually measure circa 48mm        X 35mm. The hatchlings are smaller than those of         X. agassizi, measuring on average 40mm long and weigh about 21g. Somewhat surprisingly considering the shape of their egg, they are rounder and less elongate than X. agassizi hatchlings. They can be incubated at the same temperature as X. agassizi eggs but seem to require a much higher level of ambient humidity throughout the incubation process            if embryonic dehydration is to be avoided. A Type II incubator            is recommended          for this species. Gopherus polyphemus lays much more spherical, hard-shelled eggs measuring approximately 40mm in diameter in clutches ranging from 4 to 7. The incubation period in the wild varies from 80-110 days but little data is available on incubation times under artificial conditions. Gopherus flavomarginatus are unlikely to be encountered in captivity as it is estimated that the total world population is less than 10,000 individuals. However, their eggs, which closely approximate those of X. agassizi in size can be incubated under very similar conditions. It is interesting to note that the conservation captive breeding program for this tortoise in Mexico has made extensive and routine use of oxytocin injections to induce egg laying.


General: Box tortoises are so named because they have unusually kinetic plastron hinges which in most cases enable to tortoise to withdraw completely behind a protective “draw-bridge”. The location of these hinges are important, as they must not be confused with Hinge-back (Kinixys)         tortoises which achieve a similar result but by means of a hinge in the rear of the carapace. This is an effective defence mechanism and examples of the same evolutionary solution to a common problem are found not only in the New World genera Terrapene and Rhinoclemys, but also in the Asiatic genera Cuora and Cyclemys (Pyxidea).

Taxonomy:  American box turtles belong to the Emydid genus Terrapene. In the U.S two species occur each of which is further divided into sub-species         as follows; Terrapene carolina which is currently considered to have four geographical subspecies - Terrapene carolina carolina (the common or Eastern box turtle), T. c. triunguis (the Three-toed box turtle), T. c. major (the Gulf Coast box turtle) and T. c.  bauri (the Florida box turtle). The second U.S species is Terrapene ornata which has two subspecies - Terrapene ornata ornata (the Eastern ornate box turtle) and T. o. luteola (the Salt Basin or Texas ornate box turtle). Description: Box turtles have characteristically high domed carapaces which in some forms is somewhat ridged or keeled in the vertebral region. Terrapene carolina carolina occurs from Maine to the deep south; it is quite variable in coloration, but often features yellow or orange streaks and blotches on a brown colored ground.        T.          c. triunguis occurs in Georgia, eastern Texas and Missouri; again it is variable in color, some specimens are merely olive or reddish brown whilst others display a fine pattern of radial flecks on a brown-red ground. Its main diagnostic feature is that it has three toes on its hind legs rather than the usual four - although this is by no means an infallible indicator and it is not always easy for the inexperienced observer to tell T. c. carolina and T. c. triunguis apart. They are both of similar proportions, most adults attaining circa 120-130mm SCL and weighing around 470g. Indeed, any description of U.       S.    Box turtles can only be regarded in a general sense as a great deal of hybridisation appears to occur between certain races and evident intergrades are frequently encountered. The Gulf Coast box turtle, T. c. major is very much more distinctive in both appearance and size. This, the largest of the U.S. box turtles, occurs from south-west Georgia to eastern Texas. A female specimen in our own collection measures 160mm SCL and weighs 635g. T. c. major is uniformly dark in color, almost black, with radiating light marks and blotches on the carapace - although these can be absent on some individuals.        The feet of T. c. major are more evidently webbed than in other members of the T. carolina group. The last of the T. carolina assemblage is T.            c. bauri, the Florida box turtle. The carapace of this race is similar in appearance to that of T. ornata, but typically the plastron is unpigmented and featureless and it boasts two yellow head stripes in contrast to the characteristic orange-white irregular spotting of T. c. triunguis and T. c. carolina. The Ornate box turtle, Terrapene ornata is a smaller, very attractively marked species normally featuring distinctive bright green-yellow eyes. This Box turtle’s feet show little evidence of webbing, a testimony to its more typically terrestrial habits than T. carolina. The two subspecies are very similar, but are considered separate on the basis of a lack of plastral pattern, typically less distinct carapace radiations and yellowish head scales in T. o. luteola. This yellow coloration is more evident in females than males which are typically a more greenish yellow color.

Captive environment: The various species and subspecies of U. S Box turtle occupy a variety of alternative habitats. These range from open woodlands in the case of T. carolina carolina to wet marshes in the case of T. c. major; given such a range of habitat preferences it is difficult to generalise even for a single species. However, despite their inclusion in the Emydid family they are all basically terrestrial in habit.  From the captive point of view, the main essential variable is the degree of preferred humidity and the time spent in or very near water. Some Box turtles, notably T. c. bauri and T. c triunguis are more aquatic than others; these turtles will swim in a pond and even dive to the bottom to forage among the weeds. They are most active in warm, wet weather - thunderstorms are especially welcomed.      Others, including T. ornata seem to prefer somewhat drier conditions, T. ornata in particular is essentially a prairie species with a preference for pastures and open woodlands. All Terrapene species can be safely maintained out of doors in the U.S.A and most of europe at least during the spring, summer and early autumn months. The advice sometimes given that a vivarium environment is ideal is not true. Provided outdoor temperatures are approximately similar to those experienced in their native habitat, then an outdoor environment is infinitely preferable. The Tortoise Trust has maintained all the U.S species of Box turtle in an outdoor terrarium area, very successfully, for several years. A vivarium environment is only employed with sick animals or those we wish to place under observation or to isolate for any reason.    A good outdoor terrarium should provide several square meters of land area, a reasonable sized pond for drinking and swimming, and should be thoroughly escape and predator proof - ours is constructed of plywood with a removable wire-mesh top cover. One part can be glazed,      thus creating a ‘mini-greenhouse’ for extra warmth. A good range of weeds and plants should be allowed to grow within the terrarium, and some hollowed out logs make ideal hides and retreats. Box turtles are generally shy creatures, and need to feel secure in order to feed well and to breed. Box turtles do hibernate, typically from November to March. We normally allow the Tortoise Trust colony to hibernate out of doors naturally, the turtles burying themselves deeply under some large half rotted logs in mud, loam and leaf litter. There are reliable accounts of some Box turtles choosing to hibernate over winter buried in the bottom mud of iced-over ponds; this is not a procedure which should be encouraged under normal captive conditions as natural ponds are biologically active to a much greater extent than artificial ones and anoxia could very easily occur if conditions are anything less than absolutely ideal. If they are to be overwintered, then a humid vivarium at a minimum of 21 C is necessary with additional basking facilities and preferably equipped with full spectrum fluorescent lighting.  Do not allow it to become too dry, or ear and eye problems are almost certain to develop. In the wild, under prolonged drought or heat-wave conditions Box turtles aestivate, disappearing underground for weeks on end. Diet: Terrapene species are omnivorous. In the wild they consume not only berries and other fallen fruit but also snails, insect larvae, earthworms, crickets, tadpoles, slugs and beetles in addition to toadstools and green plant material. Juveniles are noticeably more carnivorous in nature than fully grown adults. In captivity, if a natural outdoor type environment is adopted then a certain amount of the turtles dietary needs will also be met from natural resources; our own colony can often be observed hunting for small prey in the early morning and late evening, especially during or just after a heavy rainfall. Some supplementary feeding is however essential. This should comprise a wide range of berries and fruits such as cherries, apple, banana or melon plus additional vegetables and salad materials including cauliflower, green and red sweet peppers, lettuce, tomato, mushrooms and even surplus cooked potato. The animal protein requirement can be met from low fat dog food and whole dead mice, locusts and snails. Cat food is not recommended as a staple item as it is dangerously high in fat content. The food should be liberally dusted with a high ratio calcium supplement to counter the high levels of phosphorous present in the animal matter. Breeding: Sexing box turtles can be somewhat difficult; however, in T. carolina males tend to have a red eye whilst females have a yellowish brown eye; sometimes this also applies to T. ornata (especially T. o. luteola) but by no means always as a bright yellow-green eye is more usually seen in this species. Males also have longer and thicker tails than do females. There is some difference in plastral concavity, typically very evident in the T. carolina group but entirely absent in T. ornata. American Box turtles can all be captive bred under virtually identical conditions. T. c. carolina and T. c. triunguis for example typically lay 3-5 elongate eggs (occasionally as many as 8) which are somewhat leathery in texture and on average measure approximately 32mm by 20mm. These should be incubated in a Type II incubator at high humidity in a sphagnum moss substrate. If incubated at 26-28 C hatching usually occurs in 70-85 days. The hatchlings, on average, measure 28mm SCL. If the eggs are allowed to become too dry during the incubation process then embryonic dehydration will occur and the eggs will crumple and collapse. The mating process of Terrapene species is most curious; at one stage the males legs frequently become trapped in the posterior plastral hinge of the female - the process can also take several hours in contrast to the very rapid mating procedure of most terrestrial tortoises. This is preceded by a more typical biting, circling and shoving phase - during which the males frequently use their front legs to ‘spin’ the females around.


General: The Leopard tortoise is the second largest African mainland tortoise (after G. sulcata). The largest male ever recorded measured 656mm and weighed 43kg. The largest recorded female measured 498mm and attained a weight of 20kg. Most are substantially smaller, but it is not unusual to find specimens of both sexes in the 350-450mm length range and weighing above 15kg. G. pardalis occur from Sudan and Ethiopia and extend their range throughout southern Africa. It is named on account of its strikingly marked carapace which in practice constitutes an excellent camouflage. Taxonomy: The only currently disputed area of G. pardalis taxonomy concerns its division into two sub-species; G. pardalis pardalis and G. pardalis babcocki. Not all authorities accept that two clearly defined geographical races do in fact occur. However, the evidence for them is fairly compelling. What is not clear, and what might be the cause of at least some of the confusion, is whether these are the only races; certainly some keepers and field-workers I have spoken to are convinced that there may be more. Intergrades are in any case reported which further confuses the situation. There are consistent reports of infertility between pairs which are dissimilar in appearance although which theoretically belong to the same sub-species. The best fertility is obtained from pairs which are visually very similar in terms of overall body morphology, coloration and marking. Of the two recognised forms, G. p. pardalis occupies a limited range in Cape Province and in the southwestern sector of the Orange Free State, whereas G.      p. babcocki (sometimes known as the tropical leopard tortoise) enjoys a much wider distribution and appears to be subject to a higher degree of morphological variation than its Cape relative. Description: Whilst juveniles of the two currently recognised subspecies are relatively easy to differentiate, the same is not always true of adults. In juveniles, G. p. pardalis feature two or more blackish dots in the centre of the costal and vertebral scutes whereas G. p. babcocki typically feature only one. In the nominate form the plastral scutes are typically blotched with margins; in G.            p. babcocki the plastron is typically plainer and lacks the central spots. In adults G. p. pardalis are said to be typically flatter than G. p. babcocki which is highly domed as well as being considerably smaller in most cases - although this last stated character does not coincide with my own observations; I have seen some extremely large G. p. pardalis. Captive environment: In good weather G. pardalis should be allowed as much access to an outdoor grazing area as possible. Shade in the form of low growing shrubs and bushes should be included to allow retreat from the mid-day sun. During cold weather and over winter, a large heated shed, greenhouse or indoor penned area is essential which should attain daytime temperatures in excess of 20 C if activity is to be maintained. Spot or infra-red basking facilities will usually be quickly taken advantage of. Overnight temperatures should remain in excess of 10 C. In the wild, the southern race G. pardalis pardalis is known to hibernate or at least to experience a winter dormancy period, often seeking retreat from the cold in other animals’ discarded burrows. In captivity, most keepers prefer to keep the animals alert and feeding over winter by means of artificial light and heat. Diet: Leopard tortoises are very typical grazing herbivores; in the wild their diet consists very largely of grasses and succulents such as prickly pear. Several authors refer to their consuming bones and hyena faeces for their calcium content. In captivity G. pardalis should be maintained on a diet very high in fiber otherwise diarrhoea and intestinal parasite problems will quickly be encountered. For some years I have maintained a colony of seven G. pardalis babcocki on a diet consisting of natural graze (grass and assorted weeds) supplemented with cabbage and other coarse green leaf material in the winter when graze became in short supply. To supply fluid, a few tomatoes and cucumbers are included from time to time. This basic diet is heavily supplemented with ‘Vionate’ and ‘Nutrobal’ as these very large (and rapidly growing) tortoises have a prodigious demand for calcium. Breeding: Leopard tortoises, if provided with good accommodation and a well balanced diet, can be induced to breed quite readily in captivity. Males and females can be diagnosed by a number of characters; in males the tail is longer and the hind section of the plastron is depressed (only very slightly so in the case of G. p. babcocki); males are also somewhat more elongate and narrower than females. In G. p.     babcocki the males are smaller than females, whilst in G. p. pardalis males tend to be larger than females. Males kept together will often fight, levering at each other with the gular until one or other desists in defeat. Males court females in a similar manner, with much pushing and battering with the gular. Finally, the male mounts the female and accompanies the process with a great deal of deep throated ‘croaking’ and ‘grunting’. Females carrying eggs may go off their food for a while immediately prior to laying, and may dig one or more ‘trial nests’ in sunny areas of open ground. The nest itself usually measures about 25cm deep and accommodates the typical clutch of 8-10 eggs. Again, differences in clutch size are reported between the two recognised forms with G.           p. pardalis typically laying more eggs than G. p. babcocki. However, the absolute maximum recorded clutch size for G. p. pardalis is 18 eggs whilst figures as high as 30 have been reported for G.        p. babcocki which appears to contradict this. The rounded eggs of G.        p. pardalis typically measure some 43mm in diameter and weigh 50g or more whilst those of G. p. babcocki often are said to be smaller at circa 35mm with a typical weight of only 25g. However, this does not accord with my own observations based upon Kenyan and Tanzanian specimens where the average egg size is closer to 42mm with a weight of 45g. The hatchlings are also claimed to differ in size, those of G.       p. babcocki at about 38mm in length and weighing circa 17g whilst those of the southern race are said to be somewhat larger at approximately 48mm in length and up to 35g in weight. Once again however I know of 46mm hatchlings which weighed 33g produced by a female of Kenyan origin. Females may go on to produce several clutches per season - as many as six totalling 52 eggs in all has been recorded.   Incubation times in the wild are very variable, from 178 days in Zambia to 384 days in Natal. Eggs incubated artificially in a type I incubator at 28 C hatch in about 180 days; at 30 C hatchlings can be expected from 130 days onwards but most often emerge between 140 - 155 days after laying. Incubation humidity should be moderate. From the moment when the egg is first pierced by the hatchling it may be many hours or even days before it is finally ready to leave the egg; this is true of most tortoises including all Testudo, Furculachelys and Geochelone species. Once the immediate demand for oxygen has been met, the hatchling may remain in the egg whilst the remains of the egg-sac are absorbed. The only real danger (under normal captive circumstances) during this time is if the tortoise becomes dehydrated or if the mucous-like residue inside the egg literally glues the hatchlings mouth, nose or eyes up; mortality can occur if airways become blocked in this way. The problem is best prevented by maintaining adequate air humidity and by gently swabbing any obstructing matter away from the head and front legs using a damp cotton bud. Normally, human intervention will not be required but it is as well to be alert to the possibility. Juveniles grow quickly, and can reach sexual maturity within 5-6 years in the case of males,      somewhat older in the case of females which typically begin regular egg production as they attain a weight of 8kg or more.


General: One of the worlds most distinctive tortoises, this strikingly marked animal has for many years been sought after by collectors and illegal export continues to represent a significant threat along with habitat loss and utilisation for food. It has been recorded infrequently in Pakistan, with the main area of distribution in India and Sri Lanka where population density is said to be still good. Taxonomy: This tortoise appears to be closely related to the almost unknown Geochelone platynota of Burma, from which it differs principally by featuring radiating lines on the plastron and by having more rays on the costals. According to some authorities G. elegans has a more conical form to the vertebrals and costals than G. platynota, but in fact this character is very variable even within G. elegans and may be related to geographical origin. Description: The maximum recorded size of G. elegans is circa 350mm, but most specimens are considerably smaller, usually in the region of 250mm SCL. The carapace features the radiating ‘star’ pattern from whence it takes its name, a characteristic duplicated on the plastron. Each of the costal and vertebral scutes has a large, yellow central dot surrounded by a series of radiating yellow stripes. Captive environment: G. elegans is found in the wild inhabiting dry, scrub forest areas, the borders of sandy deserts and even man-made wastelands. It also inhabits grassy hillsides and the borders of cultivated areas. It appears in this respect a robust and adaptable species, yet in captivity it is generally regarded as extremely sensitive and delicate. Certainly, it does not mix well with other species and is best maintained in isolated groups. In other respects, the tortoise appears to do well if kept under very similar conditions to G. pardalis. In most of western Europe and n. America it can be allowed out of doors in good weather, and an enclosure which is well planted with grass is ideal. Overnight, a heated indoor retreat will generally be required. Otherwise treat as for G.   pardalis. Diet: As for G. pardalis. Breeding: G. elegans are not a particularly easy tortoise to induce to breed (or even mate) in captivity although it can be done. The most consistently successful captive breeding results have been achieved within the species natural bioclimatic range in India and Sri Lanka. Although this species habitat is typically dry, stony and thorny for most of the year, it is subject to seasonal rains or monsoons;   it appears to be the on-set of this rainy season which initiates interest in mating.     During this period (which occurs in India in June) the animals become especially active and feed extensively upon the new shoots of vegetation. They can often be seen marching in small groups in some areas, a lone female pursued by several males. In captivity interest in mating can sometimes be stimulated by either a natural downpour or by extensive spraying with a hose. Males rarely exhibit mutual animosity, and aggression is not often observed in this species. Females are typically larger in size than males (circa 290mm compared to 230mm) and achieve sexual maturity at about 10-12 years of age. Males can demonstrate sexual maturity in 3-5 years under captive conditions. The eggs appear to be of very variable dimensions,     although as in most accounts no geographical origin is disclosed for the specimens it could be that size varies with location. Most eggs measure about 42mm X 31mm, although records indicate a range from 38mm to 50mm in length and from 27mm to 39mm in width. Egg weight is similarly variable from 22g to as much as 38g. Females lay several clutches per year, typically three, but sometimes more.     A normal clutch consists of 3 to 6 eggs. Incubation periods in the wild demonstrate considerable variability depending upon how late or early in the season laying occurs; data from captive specimens within the bioclimatic range indicates that it can take as little as 47 days or as long as 147 days. Artificially incubated eggs in a Type I container at 28 C hatch in about 100 days; at 30 C incubation takes circa 75 days. Incubation humidity should be medium to low. The average length of hatchling G. elegans is 35mm and the average weight 15-16g (recorded minimum = 12g, recorded maximum = 22g). The hatchlings lack the distinct ray markings of the adults initially, this first becomes really evident at about 12 months of age.


General: Testudo hermanni enjoys a relatively wide distribution in the form of two recognised currently subspecies which includes eastern Spain, southern France, Italy and the Balearic islands, the Balkan peninsula, Yugoslavia, Albania, Bulgaria, Romania, Greece and Turkey. It is also found on Corfu, Sicily and Sardinia. Taxonomy: There are, as indicated above, two currently recognised sub-species of Testudo hermanni. However, these are not as frequently cited in most field guides and other works of reference ‘Testudo hermanni hermanni’ in respect of the eastern (Balkan) race and ‘T.      h. robertmertensi’ in respect of the western (French) race. In fact, by virtue of taxonomic priority the western race is actually the nominate form and should be cited as Testudo hermanni hermanni GMELIN 1789 with a designated Type Locality of Collobrieres, France. At the same time, the eastern race should be cited as Testudo hermanni boettgeri MOJSISOVICS 1889 with a designated Type Locality or terra typica of Orsova, Romania. The western population, Testudo hermanni hermanni is relatively homogenous and is certainly very distinctive. The recent discovery of      a southern Italian ‘miniature’ form which is extraordinarily small (adults are typically less than 115mm SCL) but which in other respects is visually identical to normal Italian or French T. h. hermanni may complicate this state of affairs in time. Even so it is clear that generally speaking the western populations of T. h. hermanni are very convergent in all external characters. The same cannot be said of the eastern sub-species T. h. boettgeri which is by no means contiguous and which displays an alarming range of shapes,      sizes, colors and patterns within what is supposed to be a single (subspecific) form . It is almost certain that further research will eventually lead to the identification of certain populations which will require separate taxonomic consideration. It may also be necessary to view the entire T. hermanni conglomerate as a complex of many very divergent individual populations than as two simple and clearly defined geographical sub-species. Description: Testudo hermanni was first differentiated from T. graeca on account of the horny tip it bears upon its tail and its lack of thigh tubercles. Unlike T. graeca, T. ibera and all other Mediterranean or Asiatic terrestrial species with the exception of T. horsfieldi it has a fixed and rigid xiphiplastron. It is typically a flattish animal with a broad, low carapace which bears black markings upon an olive-yellow base. The western population of T. h.    hermanni are typically smallish tortoises circa 120-130mm SCL whilst certain eastern populations of T. h. boettgeri can easily reach double this length.     In both cases there is marked dimorphism, with western males attaining a typical absolute maximum of 165mm and females 190mm.    I have measured several (female) tortoises from Bulgaria however which exceed 260mm SCL and which weigh in excess of 3,400g. Most of the tortoises of Yugoslavia, although considerably larger than T.            h. hermanni at an average of 180-200mm SCL do not attain such dramatic dimensions as that. Curiously, these ‘giant’ Hermann’s tortoises from eastern europe have all possessed 4 claws on all feet - a character usually associated with T. horsfieldi. Hybridisation is here ruled out, as they do not occur sympatrically in the region. It is frequently alleged that it is possible to distinguish between the western and eastern populations of T. hermanni by determining if the supracaudal shield is divided or undivided. It is also sometimes claimed that T. hermanni can be distinguished from T.      graeca using the same criteria. In fact, neither claim is true and this character is of little value in specific determination. The plastral markings of Testudo hermanni hermanni are characteristically formed of two almost solid dark bands running longitudinally down the plastron.         Every specimen of this sub-species examined by the author (several hundred in both France and Italy) have possessed this feature.      The plastral markings of the eastern form T. hermanni boettgeri appear to be somewhat more variable. Some specimens examined have possessed plastrons with dense markings which almost approach that of T.          h. hermanni so this character should not employed in isolation to diagnose speciation. Most bear a diffuse series of blotches    however. The groundcolor of T. h. hermanni is typically a bright golden yellow.   This contrasts sharply with most specimens of the eastern T. h. boettgeri where the groundcolor could best be described as a greenish-yellow. Similarly, the carapace markings of the western population seem to be unusually clear and well defined compared to most eastern specimens. Captive environment: The natural habitat of T. hermanni is evergreen Mediterranean oak forest; this habitat has however been substantially degraded and reduced as a result of human activity. Present-day populations are therefore found in (secondary) maquis type environments - coarse, arid and scrubby hillsides. In only a very few places does T. hermanni continue to inhabit surviving primary forest.        In captivity T.   hermanni is a comparatively resilient and adaptable tortoise. It seems equally at home in arid or even moderately damp environments but plenty of sun and warmth is essential. An ideal captive situation would provide a large outdoor enclosure, planted with low growing herbs and shrubs on well drained soil. There should be a slope or gentle hill to encourage basking and to provide a nesting site.         For overnight accommodation, a wooden hut can be provided although most tortoises will make a ‘scrape’ under a suitable bush. In the wild, T. hermanni hibernate from November to April. Diet: The natural diet of T. hermanni consists of herbaceous and succulent plants native to the mediterranean zone. In captivity as wide a range of weeds and green vegetation as possible should be provided. Lettuce alone is far from an adequate substitute. A natural grazing area is much better than artificial feeding. Although most T. hermanni enjoy fruits, an excess should not be given as their sugar content can increase the probability of digestive problems      and diarrhoea. T. hermanni require a diet which is very high in fibre and will often take dried leaves in preference to fresh. Just occasionally in wet weather T. hermanni will take advantage of a passing slug or snail but this does not comprise a significant dietary component. They should not be provided with any meat-based food items in captivity; if they are, then serious consequences may be encountered in the long-term. During heavy rain T. hermanni raise themselves on their back legs and place their noses to the ground in order to drink from shallow puddles; typically they also void urine at the same time.        As with many arid-adapted reptiles they normally choose not to lose fluid unnecessarily until it can be easily replenished. It is important to note that as with all mediterranean tortoises, seasonal variations in the quantity, quality and constituency of the food intake is an important factor; in spring abundant moist food is available in the form of fresh shoots and flowers but in summer the land becomes dry and unable to support much green vegetation. During this period the tortoises consume food with a higher dry-weight ratio than they do earlier in the year. In extreme conditions (as experienced in north African mediterranean zones for example) the tortoises may actually aestivate during this barren period. The early autumn rains result in a renewal of green vegetation allowing for a final feeding period before hibernation. Breeding: The breeding behaviour of T. hermanni is very different from that of all other mediterranean species. Unlike T. graeca or T. ibera male hermanni do not engage in the same degree of violent ‘ramming’        activity during courtship, instead resorting much more to head and leg biting which can become quite vicious in nature to the extent of occasionally drawing blood. The horny tip of the tail is also used to stimulate the females cloacal region. Males are also frequently observed apparently ‘resting’ on females backs almost mechanically stroking the carapace of the female with the front legs; this behaviour, which is conducted in a seemingly trance-like state can persist for hours. It is not known what purpose it serves, if any. Mating is a vigorous affair, accompanied by high pitched ‘squeaks’ from the male during copulation. Egg laying in the wild takes place from April to June and hatching usually occurs immediately following the first rains of September. There is a very marked difference both in clutch density and egg size between the western T. h. hermanni and eastern T. h. boettgeri; the average clutch size of T. h.     hermanni in France is 3, whilst in eastern T. h. boettgeri it is typically 5-8 and can be as high as 12. The eggs of T. h. hermanni are fairly small at 30mm long by 24mm wide on average, whilst those of T.       h. boettgeri are by comparison enormous at 40mm long by 29mm wide. Hatchling T. h. hermanni typically weigh 9-10g or so, whilst eastern T. h.                boettgeri hatchlings weigh in at an average of 12-14g. Statistically and in taxonomic terms, these are very significant differences. Both sub-species frequently lay more than one clutch per year.        In captivity artificial incubation at 30.5-31 C normally results in hatching at about the 8th or 9th week.


General: The Afghan tortoise has to date been little studied and much remains unclear concerning its biology, taxonomy and ecology.       In Pakistan T. horsfieldi occurs in Baluchistan and in very low densities in the North West Frontier Province. It also occurs in the U.S.S.R where it is subject to heavy exploitation. Taxonomy: At the species level there is little dispute; at genus level its taxonomic status is very confused. The morphological evidence is conflicting. The cranial characters suggest a close affinity to Testudo, but the carapace osteology is sufficiently different for it to be allocated to the genus Agrionemys by several   authorities. Externally, it expresses a marked phenetic relationship to T. hermanni with which it shares the morphological features of fixed xiphiplastron and horny tip to the tail - although this latter is reduced compared to T. hermanni. Description: The carapace is typically flattish, roundish and yellow-green or olive in coloration. There are some ill defined dark brown markings on the larger scutes. The plastron is typically blotched with black, or may be black all over on some examples.      There is typically a group of enlarged scales or tubercles to each side of the tail; these appear to be larger on males than females. The tail has a hard, horny tip. The feet all have four claws.  The skin is yellowish. A breeding pair of male and female T. horsfieldi now maintained by the Tortoise Trust measure 140mm590g and 187mm1,390g respectively. Captive environment: T. horsfieldi inhabit very arid, barren and rocky areas frequently in excess of 1500m altitude. They burrow deeply into sandy or clay soil in valleys and the walls of dried-up river beds or ravines.     They often initiate their burrows under clumps of grass and extend the burrow for up to 2 meters excavating an enlarged chamber at the end.     They are most active immediately after hibernation in March and April. In the extreme heat of summer the tortoises spend most of the day in retreat deep within their burrows, often emerging only for an hour or two in the morning or late afternoon to forage. In this, they closely parallel American Gopher tortoises whose lifestyle they in many respects emulate. In captivity, T. horsfieldi cannot tolerate damp but otherwise can be successfully maintained under similar conditions to T. hermanni. Burrowing facilities should however be provided. This species hibernates during the winter. Diet: As for T. hermanni. Some reports state that T. horsfieldi rarely take grass, but I have found that they very much enjoy young green shoots and consume it avidly. Breeding: In the wild most mating occurs early on in the season. Males are frequently very aggressive. They chase females, biting at their head and legs with sufficient ferocity to regularly draw blood. Two males will often fight viciously if confined together in the breeding season, but at other times seem to co-exist peacefully. Males court females by means of a strange ‘head nodding’ ritual, staring directly into the females face whilst simultaneously jerking their head up and down in a rapid motion. They emit a series of high pitched squeaks during copulation. The eggs are usually laid within 8 weeks of successful mating, and typically measure some 47mm long by 34mm wide but considerable variability is seen between individual females. Eggs usually weigh between 23-25g. Clutch density is usually between 3-5, but again this is very variable. Two to three clutches are often deposited during a season. In the wild, incubation usually takes 80-110 days, but if incubated at moderate humidity in a Type I incubator at a more or less constant temperature of 30.5 C 60-75 days is more usual. The new hatchlings typically measure between 32-34mm long and weigh from 9-12g. In their first year they tend to grow more quickly than hatchlings of either T. hermanni or T. ibera.


General: This tortoise occurs throughout Turkey, north-eastern Greece, Bulgaria, Romania, western Iran, Syria, Jordan, Iraq and the Republic Georgia in the U.S.S.R. from where the type specimen was taken in 1814. There is a considerable degree of morphological variation throughout this very considerable range, principally in respect of color but also in size; specimens from asiatic Turkey and Syria tend to be very much lighter and often feature a bright yellow head and limbs.      In north-western Turkey T. ibera tends to be much more melanistic, and on occasions almost entirely black. An introduced colony of T.      ibera also exists on Sardinia where it is sympatric with not only similarly introduced T. hermanni but also T. marginata. Taxonomy: In 1946 the German herpetologist Robert Mertens designated Testudo ibera PALLAS 1814 as a sub-species of Testudo graeca LINNAEUS 1758 but this suggestion appears seriously flawed in the light of recent investigations. There are in fact major structural,     biotypic and behavioural differences between T. graeca and T. ibera which Mertens (and most subsequent authors) have completely overlooked. Thus, this tortoise should be considered separately from T.          graeca which it only very superficially resembles. Within the very large and widely distributed population of T. ibera however there are a considerable number of very distinct geographical forms or races some of which occur in relative isolation, some of which appear to represent extremes of clines; only a few of these have so far received separate systematic recognition. To date only Testudo (graeca) nikolskii from northwest Transcaucasia and T. (graeca) anamurensis from southwestern Turkey have been proposed. It should be noted that although assigned by their respective authors to ‘Testudo graeca’ if T. ibera is considered a true and separate biological species these would instead be amended to T. ibera nikolskii and T. ibera anamurensis. Of the lighter colored and frequently much smaller Syrian and Jordanian forms very little field data currently exists and these are sometimes in error referred to as examples of ‘Testudo graeca terrestris’ with which it also allegedly intergrades (for more details of which see under ‘T. graeca’). Description: T. ibera is readily distinguished from true T. graeca; it is considerably larger with males typically reaching 180mm SCL and females 201mm SCL, although even larger examples are very frequently seen. It is also much broader and flatter than T. graeca which is characteristically high domed in lateral profile (the length-height-width ratios of T. graeca and T. ibera are totally dissimilar).    The groundcolor of T. ibera is quite variable but ranges from a greenish horn color to light orange-brown. The carapace markings are brownish-black and the vertebral and costal scutes typically feature a dark central areola with anterior and lateral borders. Very aged specimens sometimes lose the outer layer of keratin revealing large irregular bright orange areas underneath. Another carapace character which also distinguishes T. ibera from T. graeca is found at the 1st vertebral scute which is very angular compared to the rounded form of T. graeca. The head (and underlying cranium) of T. ibera is quite different from that of T. graeca, the nose in particular is broader and blunter and the eyes are characteristically much larger by comparison. The limbs are by comparison to T. graeca much thicker and shorter (there are acute osteological differences involving certain limb bones) . The supracaudal shield is only occasionally divided and the thighs feature either a single or double ‘spur’.            Some geographic populations have significantly flared or serrate posterior marginals, occasionally upturned or reverted, a character which is more prevalent in males than females. Captive environment: Generally as for T. hermanni . Testudo ibera are a relatively tenacious and robust species capable of withstanding considerable extremes of climate. Along with T. hermanni they tend to do well            under captive conditions and breed very readily. Diet: As for T. hermanni. Breeding: For best breeding success, pairs should be closely matched on the basis of general coloration and markings. Although dissimilar pairs can and do produce viable offspring, this is noticeably less consistent than is the case if pairs are more closely matched. Clutch density ranges from 4-12 but more typically numbers 6-8. Large females tend to produce higher clutch densities than smaller specimens.  The eggs of T. ibera throughout its range appear remarkably consistent in both size and weight; on average they measure 36mm long by 30mm wide and weigh about 18-20g. They can be incubated in a Type I incubator in a medium humidity environment and at a more or less constant temperature of 31 C can be expected to hatch in 60-80 days. The hatchlings typically weigh in the region of 14-16g and measure 32-34mm SCL. Contrary to popular belief the hatchlings can usually be hibernated without difficulty - indeed, it is preferable that they should hibernate if at all possible. However,    under artificial conditions the very greatest care must be taken to ensure temperatures remain at between 4-6 C and that excessive weight loss does not occur. Due to the reduced body mass of hatchlings, their core temperature fluctuates much more rapidly than that of adults in response to sudden environmental temperature changes - a temperature deviation which might cause no problem for a large adult can easily kill a tiny hatchling. Nor should hibernation periods be excessively long - between 65-80% of the normal typical hibernation period in the wild is usually the advisable maximum. We believe however that even a short hibernation is beneficial - provided it is correctly managed. There is simply no rom for error. If overwintering is preferred for the first year, then a heated and illuminated vivarium will be required.


General: Testudo graeca was for years in europe considered the archetypal ‘pet tortoise’. So much so that it became generally known as the ‘common tortoise’. However, in reality this creature is little known and is very far from common. It is found only in certain limited areas of north Africa and a small population also occurs in southern Spain. I have examined osteological material of north African descent from the Balearic islands.

Taxonomy:        ‘Confusing’ would be the most appropriate description. Despite being described for the first time in 1758 even today few are aware of its true nature. For many years, Testudo graeca was considered to be the only land tortoise in n.      Africa with the exception of Testudo kleinmanni which is found in Egypt and Libya. Other mediterranean tortoises were considered to be closely related to it and were thus regarded as sub-species. These included Testudo (graeca) ibera, Testudo (graeca) zarudnyi and Testudo (graeca) terrestris. This latter is itself the subject of yet more taxonomic confusion not to say mayhem in that it sometimes also passes for Testudo (graeca)  floweri - yet another alleged form which to all intents and purposes is completely unknown and which entirely lacks any adequate type description. Both T. ibera and the rarely encountered Iranian T. zarudnyi are in this authors opinion full and separate biological species bearing only the most distant relationship to the African T. graeca. Meanwhile T. (graeca) terrestris simply does not exist in the commonly understood sense. The name is simply a convenient a ‘catch-all’ for a variety of currently poorly defined and little      known tortoises which inhabit a herpetologically almost unstudied region. It should be noted that the allegedly ‘tiny’ terrestris form has never been scientifically described in      any meaningful way and that some authors versions of this mysterious creature measure in excess of 250mm SCL!. If this situation was not complex enough, it now transpires that the Testudo graeca described by Linnaeus is far from the only tortoise species which occurs in north Africa. These recently discovered (or in one case, re-discovered) tortoises are really very different in almost every respect from the Linnaean holotype. It is beyond the scope of this particular book to even begin to describe their complex morphology, osteology and phylogeny; however it is worth noting that at their most extreme these ‘new’ tortoises include distinct geographical forms which range from ‘giants’ which attain an SCL of 345mm and weigh almost 5kg (thus equaling T. marginata in length and totally eclipsing it in body mass) to true ‘pygmy’ or miniature species which even as fully grown adults measure only 130mm and weigh less than 400g. Certain aspects of their osteology have caused some to be placed in an entirely new genus Furculachelys. Others remain in Testudo. The current checklist of north African tortoises therefore includes not only Testudo graeca (Western Algeria and Morocco) and Testudo kleinmanni (Libya and Egypt) but also Furculachelys whitei (central coastal Algeria), F. nabeulensis (Tunisia) and T. flavominimaralis (Libya). This latter little tortoise incidentally, with its bright yellow elongate carapace, bright black eyes, bright yellow ‘masked’ head and yellow scaled legs is one of the several entirely different tortoises which have frequently been mistaken for Testudo graeca terrestris on the erroneous basis that as they are obviously not a ‘normal’ T. graeca they can’t be anything else!.

Description: True Testudo graeca are a relatively small tortoise but by no means as small as T. flavominimaralis or F. nabeulensis from which they differ not only osteologically but also in terms of their external morphology and marking. Males typically attain no more than 145 mm SCL (the average is 130mm) and weigh circa 535g. Females are considerably larger and demonstrate a high degree of dimorphism; 180mm SCL and a weight in the region of 1,300g would not be untypical. The carapace is not at all flattish but is highly domed. The groundcolor of the scutes is bright yellow and features an irregular series of small black-brown flecks or spots. The areola of each large scute has a central black dot which is bordered laterally and anteriorly. The scales of the head and legs are yellow. The frontal vertebral scute is roundish in form, somewhat pinched in or depressed on the lower half. The rear marginals are not at all flared or serrate, even on males. The supracaudal is not projected and is not ventrally introflexed. There are small tubercles or ‘spurs’ on the upper thighs, next to the tail.

Captive environment: T. graeca (and all north African species generally) are extremely environmentally sensitive and cannot tolerate any mishandling. They usually tend to do badly in captivity and very few survive in the long-term. If mixed with other more robust, competitive and aggressive species such as T. hermanni or T. ibera they rapidly succumb to alien diseases and stress. They should be maintained separately. They require conditions of higher ambient humidity than T. hermanni or T. ibera. In the wild when conditions become too hot and dry they aestivate underground. If deprived of this facility they soon develop respiratory diseases. Some authors have mistaken T. graeca aestivating in August (when temperatures regularly soar into the high 90’s F) as the beginning of their hibernation period!. In fact, some coastal populations do not naturally hibernate at all - it never gets cold enough. The coldest months are January and February where it only just falls below 60 F. In north Africa I have found tortoises grazing and otherwise behaving perfectly normally on Christmas day. Inland populations at the higher altitudes are more inclined to hibernate - but only for a very short while. Extended hibernation periods can be fatal and they should not be subjected to them.

Diet: Similar to T. ibera, T. marginata and T. hermanni but flowers appear to be a more important dietary constituent. These tortoises are all coprophagous to a greater or lesser extent, feasting whenever the opportunity arises on mammalian dung - in north Africa sheep, goat and camel droppings are all taken with relish. It is quite likely that this does in fact play an important biochemical and digestive role but to date this aspect has been little studied.

Breeding: Unlike T. hermanni, T. marginata or T. ibera, north African T. graeca are rarely bred at all in captivity. In addition, very little data is available on their natural reproductive biology or behaviour.  In the last 12 months however the Tortoise Trust has succeeded in captive breeding north African tortoises from several different localities including true T. graeca. The eggs are much smaller and rounder than those of T. ibera typically measuring 30mm long by 27mm wide. The hatchlings weigh a diminutive 7-8g and measure circa 28mm SCL. A typical clutch consists of 4-5 such eggs. Incubation is best accomplished at between 30.5 C and 31.5 C in a Type I unit at medium humidity. At this temperature range incubation takes approximately 68-80 days. The hatchlings are light brownish-yellow in color without distinct markings which develop as they grow. By 1 year of age they are approximately half the size of T. ibera hatched at the same time.


General: Testudo marginata is one of the largest circum-mediterranean tortoises (comparable only to Algerian F. whitei which surpasses it in body mass) and is distinguished not only by its extremely flared, serrate and projecting rear marginals (from which it takes its name) but also by its unusually marked plastron which is unique among the tortoises of the region (see however my note concerning T. kleinmanni). The main population concentration of T. marginata occurs in the Peloponnese and along the associated Greek coast to Mount Olympus. Smaller populations occur on several Aegean islands, and an introduced population occurs on Sardinia. A small population also survives to this day in Tuscany, Italy where it is assumed to have been introduced by the Etruscans. Taxonomy: This tortoise was described by Schoepff in 1792 and has since acquired a number of synonyms including ‘T. nemoralis’ and the rather descriptive ‘T. campanulata’. The Marginated tortoise is distinctive both visually and zoogeographically and its status has rarely been the subject of any dispute. Early records of alleged ‘T. marginata’ from north Africa (of which, surprisingly, there are plenty) are the result of confusion either with the similarly sized F. whitei of Algeria or equally surprisingly with the diminutive Testudo kleinmanni; despite all the other differences this little tortoise does feature an approximately similar set of markings on its plastron - not similar enough to cause any confusion you would have thought, but nonetheless this mistake has occurred several times.

Description: In addition to the obviously flared posterior marginals noted above, T. marginata feature several other key distinguishing features. The soft skin of the upper limbs and tail is typically very creamy pale; much more so than on either T. ibera or T. hermanni and very noticeably so in young animals. The plastron of T. marginata characteristically features a large brown-black triangular marking on each of the larger scutes. The head is relatively small and the eyes are also smaller and more almond-shaped than those of T. ibera.         It has been claimed that T. marginata have 5 claws on all feet and that the rear plastral lobe is not mobile; neither statement is in fact true. Adult females typically measure circa 240-280mm SCL and weigh on average between 2-3kg, whilst males are normally somewhat longer than females of equivalent age measuring 250-295mm SCL but weigh in the same range. Length for length females typically weigh more than a male of identical SCL but are considerably broader in girth. The assumed homogenous genetic continuity of T. marginata is thrown into some doubt however by recent morphometric results which indicate that in at least some populations adult males do not exceed 230mm SCL and females do not exceed 215mm. Other dimorphic indicators typical of males in this species include tail length, a considerably narrower mid-line and more prominent marginal flares.

Captive  environment: As for T. ibera and T. hermanni Diet: As for T. ibera and T. hermanni.

Breeding: T. marginata breed readily in captivity; possibly the fact that there appears to be relatively little genetic diversity from one part of their range to another (with the exception of the ‘miniature’ populations noted above) plays a part in this. Certainly, almost any male will mate successfully with almost any female and viable eggs will usually result. Clutch density is typically 8-10 (occasionally more) and the eggs measure on average 30.5mm long by 28mm in width. They normally weigh between 16-18g. If incubated at circa 31 C hatching usually occurs in 60-70 days. Incubation humidity does not appear to be especially critical, good results having been noted in low (- 65%), medium (80%) and high (90% +) environments. One of the most successful breeders of T. marginata I know has used a simple Type II incubator of the most basic design for several years and invariably succeeds in hatching large numbers of this species every year. The hatchlings usually measure some 30mm SCL and weigh between 10-14g. Juveniles are roundish in overall shape, lacking both the elongate body form and flared marginals of the adults.


General: Malacochersus tornieri, the Pancake or Soft-shelled tortoise, is in many ways the most unusual tortoise in the world. It is certainly one which demonstrates a remarkable degree of adaption. Not only is it almost completely flat, but throughout life it remains soft and surprisingly flexible - a capability it employs to useful effect in defense by wedging itself in rocky fissures, expanding itself and thereby ‘jamming’ itself in very effectively. In order to effect this remarkable ability, the underlying bones are fenestrated. It is in addition rated as possibly the worlds fastest tortoise, and surprised specimens will usually run for cover rather than seek withdrawal into their shells for protection . Its distribution is limited to Kenya and Tanzania where it occurs in thorn covered rocky outcrops to an altitude in excess of 1600m. Taxonomy: This tortoises taxonomy and nomenclature is not currently in dispute. There are no recognised sub-species. Description: Both males and females are flat and entirely lack any evidence of the usual carapace ‘dome’. Females are only marginally deeper in the body than males. The normally encountered maximum size of males is circa 167mm (height 36mm) whist that of females is 177mm (height 45mm). A male and female in our collection measure 160mm long by 34mm high and 170mm long by 43mm high respectively. The male weighs 360g and the female 550g. The ground color of the carapace is typically a golden brown or horn color with a distinct pattern of radiating black-brown marks. Females tend to have more of a rayed pattern than males which are characteristically more ‘mottled’   in appearance.

Captive environment: Pancake tortoises prefer a dry, rocky environment in captivity as they do in the wild. If a large, artificial ‘mountain’ can be provided with plenty of retreats and cracks then so much the better. These tortoises will spend hours scrambling over even the most precipitous terrain. Peak activity occurs in the morning and early evening, much of the day being spent in retreat under a convenient rocky ledge. They are inclined to be gregarious in habit, with several tortoises ‘stacking’ themselves on top of the other in a particularly favoured retreat. Two males placed together will however fight viciously during the breeding season, biting and snapping at each others heads and attempting to turn each other over. Their agility is such however that they can right themselves with little difficulty. Daytime temperatures should be in the order of 25 - 29 C although as is frequently the case with tortoises which inhabit high, dry places Malacochersus tornieri can withstand considerably lower temperatures overnight. In captivity a minimum of 13 C is recommended. Damp is not much appreciated and should be avoided. This tortoise rarely drinks water, seemingly obtaining almost all of its fluid requirement from its food.

 Diet: As for G. pardalis with a preference for grass and succulents. In captivity cabbage, lettuce, tomatoes and cucumber are taken readily. They seem little interested in most fruits, although melon is often an exception to this general rule. The diet should be supplemented as required with multi-vitamins and calcium - especially for egg laying females and juveniles.

Breeding: M. tornieri mate in a state of high excitement, the male vigorously snapping and biting at the females head and legs as he quickly and repeatedly circles her. Eggs are laid singly, and often at intervals of 6-8 weeks. A female in our own collection regularly produces such eggs which on average measure 47mm long by 31mm wide and weigh 35g. At 30 C incubated in a Type I container the incubation period is typically 140 days, but this does appear to be quite variable sometimes taking much longer. Incubation humidity should be medium to low. In the wild, eggs are laid in July or August, but the natural mean incubation period remains unknown. In captivity, females may choose to nest either during the daytime or early evening digging a fairly normal nest approximately 100mm in depth, or alternatively simply secreting the egg in a convenient nook or cranny among the rocks. Hatchlings upon emergence are strangely not particularly odd looking, and are even somewhat domed. Their profile is not dissimilar to that of T. hermanni for example. The hatchlings have a bright yellow groundcolor and plastron with deep brown-black spots on the vertebral and costal scutes. Typically, newly emerged hatchlings measure about 40mm in length and weigh about 16-18g.

(C) 1990-2004 by A. C. Highfield. All rights reserved.