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.

An effective method of artificial incubation for mediterranean tortoise eggs

A C Highfield


The incubators used in our Mediterranean tortoise captive breeding program are based upon a fairly simple design. This comprises a faced chipboard box measuring approximately 18" X 24" X 12". Heat is provided by a 60W industrial quality ceramic heating element located in the base (although a similarly powered heat-pad is just as effective), and temperature control is provided by a proportional electronic module which avoids the major fluctuations experienced with simple on-off type thermostats (Highfield, 1987). The design is capable of maintaining temperatures to +/- 0.25 degrees centigrade if required (this facility can prove useful in Environmental Sex Determination experiments). The probe from the temperature module is placed at the same height as the egg trays, and a second probe leads to a digital thermometer located on the front panel. This can include an audible high or low temperature warning if required. Small electronic digital temperature modules are now fairly widely available from electronic stores such as Radio Shack at very reduced prices compared with a few years ago. An even more useful development is that of combined humidity-temperature modules. These are absolutely ideal for incorporation into incubators and allow a degree of control previously unattainable in anything outside laboratory settings.
The incubators are each fitted with two lids, one made of heat resistant transparent plastic and the second of cork-lined plywood. To inspect the eggs only the outer lid need be lifted, thus preserving temperature stability during extended periods of observation (but see notes re; ventilation). The eggs are not buried but rest in plastic ice-cream or margarine containers on a substrate of sterile peat, light soil and gravel. Contrary to much published advice, sand is not a good material in which to bury eggs during incubation. I have encountered many cases involving the suffocation of emerging hatchlings where this has been used; a lighter mixture of sandy soil or peat is much safer.
A third transparent plastic lid rests gently on each (to prevent unexpected hatchlings climbing out and injuring themselves). This is perforated to permit air to reach the eggs.
Humidity is provided by placing a tray of water containing sponges in the incubator alongside the egg trays. The level of humidity does not appear to be especially critical with the hard-shelled eggs of
Testudo species, but we usually maintain it at around 70-80%. Each incubator holds approximately 30 eggs - we prefer to use multiples of these smaller incubators rather than use fewer larger incubators and (literally) place all our eggs in one basket!. So far, an incubator has never failed but it is obviously far safer to distribute any risk in this way. It also facilitates incubating batches at different temperatures. As a general starting point, 31 degrees C. is suggested for most Testudo species.
The time required for incubation is related to temperature, higher temperatures producing progressively shorter times. At 31 degrees C most
Testudo species hatch in 8-10 weeks. Very high temperatures, although producing shorter incubation times can result in severely deformed hatchlings. For maximum success and safety, it is best to keep incubation temperatures within the range 29-32 oC. Tortoise eggs, like crocodile eggs, are subject to environmental sex determination (ESD). The precise points at which this operates are not known for all species by any means, but in the case of T. hermanni for example eggs incubated at or below 29.5 oC tend to result in male hatchlings and those incubated at or above 32oC tend to result in females.

Oxygenation & incubation conditions

One potential problem area often overlooked by those involved in captive breeding chelonians is the requirement of developing eggs for oxygen; admittedly this demand is low in the early phase of development but it does increase (Bellairs, 1969). Where anoxia of eggs is allowed to occur an increased number of 'dead-in-shell' incidents will be noted. Embryonic anoxia can also result in hatchlings leaving the egg early (a rise in blood Co2 levels almost certainly acts as the 'trigger' for emergence in tortoises as it does in birds). Where eggs are incubated in conditions where oxygenation and gas-exchange potentials are limited hatchlings will be seen to leave the eggs in a weak, feeble condition often bearing unusually large egg-sacs. Incubation in sealed incubators or in inadequately ventilated containers can drastically increase both pre-hatching and post-hatching mortalities. It may also contribute to neo-natal deformities. The practice of incubating eggs in sealed or almost-sealed containers is therefore to be strongly discouraged. Conversely, the airflow of most bird egg incubators is far too high for the successful incubation of chelonian eggs and can easily lead to excessive drying and even to embryonic dehydration. Infertile eggs dehydrate most rapidly due to the lack of formation of protective internal membranes, but even fertile eggs can suffer fatal dehydration if subjected to extended periods of high airflow. We have found that if the incubator is opened once per day for a 30 second ventilation period this is quite adequate to flush any Co2 build-up and prevents any problems of this nature manifesting.
Whilst it is easier to maintain high levels of humidity in a sealed environment tests have shown that Co2 levels accumulate rapidly in such situations. Embryonic anoxia is quite possibly the hidden factor responsible for many "dead-in-shell" and premature hatchlings encountered by captive breeders where other explanations (e.g incubator failure or parental genetic incompatibility) can be discounted.
Where a sealed incubation environment has been used and high levels of mortality encountered, a change to a more natural and aerated incubation method will often produce an immediate and dramatic improvement in survivorship. It is a myth that because eggs are buried underground they require no oxygen or do not need to ventilate waste gasses - in fact, soil oxygen levels of nesting sites are usually good and permeability comparatively high.


Hatchlings pierce the eggshell using an egg-tooth like appendage ( actually an egg-caruncle) gradually enlarging this opening by biting small pieces from the eggshell and pushing with the front legs.
Immediately hatching begins the eggs should be kept under continuous observation. Hatching can take some time; between 2-5 hours is average for
T. graeca, T. hermanni and T. marginata. Once access to air has been gained, the young tortoise will often stay in the egg for a day or more gradually gaining in strength and allowing time for the egg-sac to be properly absorbed. If a hatchling is in obvious trouble and is clearly weakening then careful assistance can be given. Provided that hatching is in full progress giving such aid will not do any harm. Do not ever initiate hatching artificially - even if it appears that the 'normal' incubation period is long past and hatching is overdue. Tortoises have been hatching without assistance from humans for millions of years and really don't need our unwelcome (and usually untimely) interference. If you find yourself having to break open eggs to get live hatchlings then look again at your incubation technique. You are going badly wrong somewhere along the line.

Comparative egg morphology

There are considerable morphological differences in the construction of eggs between various species - even within the mediterranean region. These differences are consistent and can be of use in taxonomic diagnosis (Highfield). For example, the eggs of the east european Testudo hermanni boettgeri MOJSISOVICS 1889 (Bour, 1987) typically measure some 38mm X 30mm and are markedly elongate. By comparison the eggs of Testudo ibera PALLAS 1814 typically measure 36mm X 29mm and those of the true Testudo graeca LINNAEUS 1758 from Oran, Algeria are a diminutive 29mm X 28mm and are roundish rather than obviously elongate. The more easterly Algerian Testudo whitei BENNETT 1836 (Highfield & Martin, 1989) produces eggs which are very much larger and a different shape than those of T. graeca at 36mm X 28mm. There are similar differences to be seen in the hatchlings, with true Testudo graeca L. 1758 typically emerging at 27mm/8g, Testudo ibera PALLAS 1814 at 33mm/13g, T. hermanni boettgeri MOJSISOVICS 1889 at 33mm/12g and Testudo whitei at 34mm/14g. Early phase growth is also variable between species, but very consistent within species (Highfield, in press)..
By this point readers familiar with the general literature on Mediterranean tortoises may have noted that the nomenclature used above is not consistent with that of currently accepted works of reference which state merely that in the Mediterranean region there are but three species,
T. graeca, T. hermanni and T. marginata. This is in fact not correct, and in actuality the term 'T. graeca' as usually employed is a catch-all definition which includes a number of very diverse tortoises from discreet geographical areas some of which are undoubtedly full and separate species in their own right. Few of these have been studied in any detail until very recently, which may explain why they have been incorrectly classified for so long. In particular, Robert Mertens hypothesis of the alleged sub-specific divisions of Testudo graeca, where T. ibera and T. zarudnyi were synonymised and relegated is seriously in error; these tortoises are tremendously divergent both morphologically and genetically from the n. African species and are zoogeographically not at all closely related as Mertens suggested (Mertens, 1946).
Clearly, this has important implications for anyone involved in captive breeding these tortoises. The best practical advice has to be to confine breeding efforts only to accurately matched pairs of the same species (i.e, do not attempt cross-breeds between Turkish
T. ibera and Moroccan T. graeca for example). My own evidence is that typically unmatched pairs are unlikely to produce fertile eggs, but when they - exceptionally - do so, the young appear to be sterile. All is by no means completely cut and dried in this respect, but the need to exercise due caution is I think self evident. Anyone involved in captive breeding endangered animals (and all of the tortoises mentioned above fall into this category) has a duty and obligation to preserve intact the genetic integrity of the species; cross-breeding, whether accidental or deliberate runs counter to that obligation.

Although primarily concerning the incubation of eggs from Testudo species, most (if not all) of the above observations apply equally to other species. Obviously soft-shelled eggs frequently require higher levels of ambient humidity, but the same design of incubator and incubation environment has proved successful with many Geochelone species. At present, we also have eggs from Malacochersus tornieri, Chelonoidis carbonaria and T. horsfieldi incubating under similar conditions.

Literature cited

  • Bellairs, A. (1969) The Life of Reptiles. Weidenfeld & Nicholson. London.
  • Bour, R. (1987) L'identite' des Tortues terrestres europeennes: Specimens-types et localite-types. Revue fr. Aquariol 13 (4):111-122
  • Highfield, A. C. (In press) Studies in Comparative Morphology of the eggs of Tortoises, genus Testudo and their relevance as Characters in Diagnostic Taxonomy.
  • Highfield, A. C. (In press) Comparative early phase growth in Palearctic land tortoises genus: Testudo.
  • Highfield, A. C. (1987) Electronic temperature Measurement and Control for Incubators and Vivaria. The Herptile. Journ. International. Herpet. Soc. Vol.12 (4):130-133
  • Highfield, A. C & Martin, J. (1989) Testudo whitei BENNETT 1836; New light on an old carapace - Gilbert White's Selborne tortoise re-discovered. Journ. Chelonian Herpetology 1(1):13-22
  • Highfield, A.C. (1990) Preliminary report on the Taxonomic, Biotypic and Conservation status of the Land Tortoises of Tunisia. Tortoise Survival Project, London.
  • Iverson, J. B. (1982) Biomass in Turtle populations; a neglected subject. Oecologia (Berl) (55):69-76.
  • Lambert, M. R. K. (1984) Threats to Mediterranean (West Palaearctic) tortoises and their effects on wild populations: an overview. Amphibia-Reptilia 5:5-15.
  • Mertens, R. (1946) Uber einige mediterrane Schildkrotenrassen. Senckenbergiana Biol. 27:111-118.

(c) A. C. Highield 1987