Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Individual study: Trial release of Madagascar ploughshare tortoise Geochelone yniphora at Beaboaly, northwestern Madagascar

Published source details

Pedrono M. & Sarovy A. (2000) Trial release of the world's rarest tortoise Geochelone yniphora in Madagascar. Biological Conservation, 95, 333-342


The Madagascar ploughshare tortoise Geochelone yniphora is the world's rarest tortoise, with less than 1,000 wild individuals remaining in remote isolated patches of bamboo scrub habitat in an area of about 100 km² around Baly Bay, northwestern Madagascar. The population decline has been caused by frequent man-made fires, and collection for food and the pet trade. In the 1980's, a captive breeding program, consisting of confiscated captive animals, was initiated at the Ampijorao Forestry Station, northwest Madagascar. The captive population increased annually and by 1997 the first captive-raised juveniles were ready for experimental reintroduction in the Baly Bay National Park. In this study, the results of a trial release aimed at maximising the success of future reintroductions are presented.

Tortoise selection: Released individuals needed to be large enough to withstand predation (e.g. by the non-native African bush pig Potamochoerus larvatus and Indian palm civet Viverricula indica, and the native Madagascar buzzard Buteo brachypterus) and environmental factors (climatic and food shortage). However, they had to be sufficiently young so as not to impair their ability to settle at a site. For these reasons, juvenile first generation offspring (the oldest captive-bred tortoises) were used. Individuals of similar body dimensions and weight were chosen. Thus, tortoises were 8-9 years old, had carapace lengths of 198-208 mm, and body weights of 1.91-2.04 kg.

Tortoises were screened for diseases to ensure they were healthy and could not spread disease to wild tortoises. They were then implanted with AVID passive integrated transponders for individual identification. Lastly, five similarly sized, wild tortoises were marked at Ambatomainty as controls.

Release site: Tortoises historically occurred at Beaboaly (west of Baly Bay), but due to human harvesting and habitat destruction they had not been reported for over 20 years. The local Sakalava people were aware of the tortoise, its rarity and supported its protection. The site is several km² with suitable bamboo scrub and dry deciduous forest habitat, dominated by the tree Terminalia boivinii, the orchid tree Bauhinia pervillei, and bamboo Perrierbambos madagascariensis.

Release: Soft-release reintroduction was used. In January 1998, the five tortoises were acclimatised on-site for 4 weeks in a 4 x 4 m pre-release pen, which was situated at the centre of the release area. For the first two weeks they were provided with water and vegetables, with only water in the third week, and with nothing in the fourth week. Tortoises were released in February 1998, which is the wet season and therefore has favourable conditions with food, water, and vegetation cover available. They were placed in a group in the centre of the release site early in the morning and once released, human contact was minimised and no intervention was made.

Monitoring: Each individual at Beaboaly (released) and at Ambatomainty (wild) had an AVM transmitter (AVM Instrument Company Ltd.) glued to the final vertebral scutes with epoxy resin. From the date of release, all tortoises were located daily until January 1999, using a LA12Q receiver with H-shaped antenna. Their position was mapped, the carapace surface temperature was measured using a thermocouple probe connected to an electronic thermometer HI 935005 (James Scientific Instruments Ltd.), and it was noted whether the tortoise was in the open or under vegetation. Every 15 days, tortoises were measured and weighed with a Pesola scale and Haglof Mantax tree callipers (Forestry Suppliers Inc.). Additionally, foraging was observed for six hours per individual per week.

Post-release dispersal & home range: All tortoises settled near the release site and displayed site fidelity. The tortoises did foray from the release site (maximum distance = 746 m), but they always returned to the release area and after 325 days, tortoises were 339 ± 144 m from the release point. Indeed, the released tortoises returned to the same location (26% of daily locations) significantly less than wild tortoises (44% of daily locations). During the dry season, released tortoises entered aestivation (a period of inactivity) one month later than wild tortoises, but both released and wild tortoises became active in September. Finally, released and wild tortoises did not differ significantly in their distance moved daily or in their total area of use.

Foraging success & behavioural patterns: Tortoises selected and fed on native plants within minutes of release. Average body mass changes were similar for wild and released tortoises. Furthermore, there was no significant difference in the distribution of behaviours, i.e. time spent resting, feeding and walking, between wild and released tortoises. At night both groups burrowed into the leaf litter. Finally, carapace surface temperature was not significantly different between wild and released tortoises, indicating that released tortoises are able to regulate their body temperature effectively in wild surroundings.

Conclusions: The trial release was a success, with all released tortoises surviving for one year until the end of monitoring. Furthermore, they behaved and foraged in a similar manner to wild tortoises, which suggests that they had adapted to their release site rapidly.

Note: If using or referring to this published study please read and quote the original paper. This is available from Please do not quote as a case as this is for previously unpublished work only.