on the diet of the dodo are based mainly on speculation. Some sailors' accounts talk of watching dodos wade into water-pools to catch fish. They have been described as "strong and greedy" hunters. What really fascinated the visitors to Mauritius, however, was the fact that dodos seemed to eat stones and iron frequently and with no trouble. It is now surmised that the rocks eased digestion. The dodo did eat seeds and fruits.
Stanley Temple, a University of Wisconsin ornithologist, hypothesized that the extinction of the dodo was responsible for the near extinction of the tambalacoque or calvaria tree (Sideroxylon grandiflorum, formerly Calvaria major). He proposed that dodo and tambalacoque represented an obligation animal-plant mutualism in which the tambalacoque seeds had to pass through the dodo digestive system before they could germinate. Temple assumed the tambalacoque's fruit, with its thick and hard endocarp, would mechanically prevent germination unless worn down by the dodo digestive system. He drafted a flock of turkeys to serve in a germination experiment. Of the seventeen tambalacoque pits he force-fed to the turkeys, his dodo-substitute, only three germinated. Temple also claimed that no tambalacoque trees were less than 300 years old, but had no data to support that other than second-hand estimates. Tambalacoque trees have no annual rings so their age is not easy to determine. Temple published in a very prestigious journal that gave his appealing hypothesis added credibility and widespread attention. The hypothesis has been widely adopted as fact by many biology books and webpages as an example of an obligate animal-plant mutualism. If his hypothesis was true, it would be a good example of the consequences of the extinction of a species in an ecosystem.
After his publication several scientists have rebutted Temple's hypothesis. There exist tambalacoque trees in the wild on Mauritius that are less than 300 years old. There were also not 13 trees left in 1973 as he claimed, but several hundred. The decline in the tambalacoque population has been caused by other factors such as large-scare deforestation, destruction of seeds by fungal diseases, and introduced plants and animals. Temple also overlooked reports on tambalacoque seed germination by Hill (1941) and King (1946), who found the seeds germinated without abrading. Tambalacoque is analogous to Peach. Both have a hard endocarp, surrounding the seed but the endocarp naturally splits along a fracture line during germination. Temple made a fundamental error by not having a control treatment of uneaten pits in his germination experiment with turkeys. It is now also known that tambalacoque seeds do germinate naturally without the dodo and without artificially abrading the endocarp. While dodos may have eaten tambalacoque fruits, there is no solid evidence they did. Nor is there solid evidence that the dodo was absolutely required for seed germination. This was a great story when it was first reported in the early 1970's, but it was proven to be false.