Remove a species from an ecosystem and other species tend to suffer. Take the giant Madagascar tortoise, for example. The two species of giant tortoises on Madagascar went extinct centuries ago, but their loss is still being felt today. According to new research, the extinction of these tortoises robbed one of the island's iconic baobab tree species of its most important seed dispersers, a situation from which the trees still have not recovered.

As I wrote a few weeks ago, all six of Madagascar's native baobab tree species relied on large animal species to eat their fruit and disperse their seeds. Unfortunately, most if not all of those large animals have gone extinct, leaving the trees with fewer opportunities for seeds to be carried in the animals' guts to new germination sites, a problem compounded by modern logging, agriculture and development, which have reduced and strictly limited baobab habitat. Although some baobab species remain relatively common today all six Madagascar species face low reproduction rates and have few opportunities for seed dispersal. In the case of the species known as the fony baobab (Adansonia rubrostipa), which grows in the dry forests of western Madagascar, a 2007 study (pdf) found that the growth of new seedlings required seeds reaching areas that weren't already covered by older baobab trees or heavy shrubbery, a phenomenon that the researchers characterized as "relatively rare."

It has long been theorized that giant tortoises (Aldabrachelys grandidieri and A. abrupta) served that dispersal role for fony baobabs until humans ate the animals into extinction roughly 2,000 years ago. A team of researchers from the University of Antananarivo in Madagascar and CIRAD, an agricultural research center in France, wanted to test that theory. Obviously there were no native Madagascar giant tortoises to participate in the experiments, but an analogous species was readily available. The Aldabra giant tortoises (A. gigantea) living on the islands of the nearby Republic of Seychelles, located just north of Madagascar in the Indian Ocean, are close relatives to the extinct Madagascar species and fill a similar ecological role.

The researchers collected baobab fruits that had recently fallen from A. rubrostipa trees and fed them to five Aldabra giant tortoises residing at Tsimbazaza Botanical and Zoological Park, also in Madagascar's capital city, Antananarivo. The tortoises appeared to enjoy the fruits, an indication that the Madagascar species probably would have had a taste for them as well. The next step was to wait for the tortoises to defecate the seeds, a process which took between 15 and 23 days. Tortoises in the wild can travel a fair amount of distance over two to three weeks, so this defecation time frame supported the theory that the Madagascar tortoises could have served as effective seed dispersers. Previous research into Galápagos giant tortoises has shown that they disperse seeds in their feces more than four kilometers from parent plants. This is especially important for baobab trees, since seeds that fall and stay too close to their parents often don't have enough room or sunlight to germinate and thrive.

Finally, the researchers tested the seeds that had passed through the tortoises' guts to see if they would germinate and compared the germination rate with seeds that had simply been washed of their surrounding fruit pulp. The tortoises’ stomach acid had somewhat scarred the defecated seeds but they still germinated, although at a slightly slower rate than seeds that had not been swallowed. The researchers theorize that the defecated fony baobab seeds may enter a dormancy phase which parallels that of the continental African baobab species, seeds from which can remain dormant for up to three years and sprout when ecological conditions are most advantageous for growth. The combination of food preference, digestion time and dormancy phase all supported the idea that the extinct Madagascar tortoises would have served as primary and effective dispersers for fony baobab seeds.

The research, published August 5 in the African Journal of Ecology, does more than just prove that baobab trees relied on extinct Madagascar tortoises. It also lays some of the groundwork for the reintroduction of giant tortoises back to the island. One of the paper's authors, CIRAD conservation biologist Miguel Pedrono, is also part of a team that plans to import 300 juvenile Aldabra giant tortoises to western Madagascar. The animals will live in a pen for five years where their eating habits will be closely monitored. After that it is hoped that they will breed and disperse, filling the ecological niche once filled by their extinct cousins. Pedrono and his co-authors described their plan in the March issue of Biological Conservation, where they called it the first project to restore an island's ecology by importing a relative of an extinct megafauna species and wrote this translocation could be "a pragmatic and cost-effective tool to contribute to halting the ongoing extinction processes in parts of western and southern Madagascar, and would further understanding of the role of these species in prehuman Madagascan ecosystems."

Could the effects of a millennia-old human-caused extinction truly be reversed? Only time and tortoises will tell.

Previously in Extinction Countdown:

Photos: Aldabra giant tortoise by Michelle Bender. Adansonia rubrostipa by Andrew Hankey. Both used under Creative Commons license