Temple then tried to estimate the crushing force of a dodo’s gizzard by making a plot of body weight versus force generated by the gizzard in several modern birds. Extrapolating the curve up to a dodo’s size, he estimates that Calvaria pits were thick enough to resist crushing; in fact, the thickest pits could not be crushed until they had been reduced nearly 30 percent by abrasion. Dodos might well have regurgitated the pits or passed them along before subjecting them to such an extended treatment. Temple took turkeys—the closest modern analogue to dodos—and force-fed them Calvaria pits, one at a time. Seven of seventeen pits were crushed by the turkey’s gizzard, but the other ten were regurgitated or passed in feces after considerable abrasion. Temple planted these seeds and three of them germinated. He writes: “These may well have been the first Calvaria seeds to germinate in more than 300 years.” Calvaria can probably be saved from the brink of extinction by the propagation of artificially abraded seeds. For once, an astute observation, combined with imaginative thought and experiment, may lead to preservation rather than destruction.
I wrote this essay to begin the fifth year of my regular column in Natural History magazine. I said to myself at the beginning that I would depart from a long tradition of popular writing in natural history. I would not tell the fascinating tales of nature merely for their own sake. I would tie any particular story to a general principle of evolutionary theory: pandas and sea turtles to imperfection as the proof of evolution, magnetic bacteria to principles of scaling, mites that eat their mother from inside to Fisher’s theory of sex ratio. But this column has no message beyond the evident homily that things are connected to other things in our complex world—and that local disruptions have wider consequences. I have only recounted these two, related stories because they touched me—one bitterly, the other with sweetness.
Coenobita diogenes in the shell of Cittarium. Drawn from life by A. Verrill in 1900.
Postscript
Some stories in natural history are too beautiful and complex to win general acceptance. Temple’s report received immediate publicity in the popular press (New York Times and other major newspapers, followed two months later by my article). A year later (March 30, 1979), Dr. Owadally of the Mauritian Forestry Service raised some important doubts in a technical comment published in the professional journal Science (where Temple’s original article had appeared). I reproduce below, verbatim, both Owadally’s comment and Temple’s response:
I do not dispute that coevolution between plant and animal exists and that the germination of some seeds may be assisted by their passing through the gut of animals. However, that “mutualism” of the famous dodo and Calvaria major (tambalacoque) is an example (1) of coevolution is untenable for the following reasons.
1) Calvaria major grows in the upland rain forest of Mauritius with a rainfall of 2500 to 3800 mm per annum. The dodo according to Dutch sources roamed over the northern plains and the eastern hills in the Grand Port area—that is, in a drier forest—where the Dutch established their first settlement. Thus it is highly improbable that the dodo and the tambalacoque occurred in the same ecological niche. Indeed, extensive excavations in the uplands for reservoirs, drainage canals, and the like have failed to reveal any dodo remains.
2) Some writers have mentioned the small woody seeds found in Mare aux Songes and the possibility that their germination was assisted by the dodo or other birds. But we now know that these seeds are not tambalacoque but belong to another species of lowland tree recently identified as Sideroxylon longifolium.
3) The Forestry Service has for some years been studying and effecting the germination of tambalacoque seeds without avian intervention (2). The germination rate is low but not more so than that of many other indigenous species which have, of recent decades, showed a marked deterioration in reproduction. This deterioration is due to various factors too complex to be discussed in this comment. The main factors have been the depredations caused by monkeys and the invasion by exotic plants.
4) A survey of the climax rain forest of the uplands made in 1941 by Vaughan and Wiehe (3) showed that there was quite a significant population of young tambalacoque plants certainly less than 75 to 100 years old. The dodo became extinct around 1675!
5) The manner in which the tambalacoque seed germinates was described by Hill (4). who demonstrated how the embryo is able to emerge from the hard woody endocarp. This is effected by the swollen embryo breaking off the bottom half of the seed along a well-defined fracture zone.
It is necessary to dispel the tambalacoque-dodo “myth” and recognize the efforts of the Forestry Service of Mauritius to propagate this magnificent tree of the upland plateau.
A. W. OWADALLY
Forestry Service, Curepipe, Mauritius
References and Notes
1. S. Temple, Science 197. 885 (1977).
2. Young Calvaria major plants that are 9 months old or more can be seen at the Forest Nursery in Curepipe.
3. R. E. Vaughan and P. O. Wiehe, J. Ecol. 19. 127 (1941).
4. A. W. Hill, Ann. Bot. 5, 587 (1941).
28 March 1978
The plant-animal mutualism that may have existed between the dodo and Calvaria major became impossible to prove experimentally after the dodo’s extinction. What I pointed out (1) was the possibility that such a relation may have occurred, thus providing an explanation for the extraordinarily poor germination rate in Calvaria. I acknowledge the potential for error in historical reconstructions.
I disagree, however, with the conclusion of Owadally (2) that the dodo and Calvaria were geographically separated. There have been virtually no bones of dodos or any other animals found in the uplands of Mauritius not because the animals were never there, but because the island’s topography does not cause alluvial deposits there. Catchment basins in certain lowland areas accumulated many bones of animals that were washed into these areas from the surrounding uplands. Accounts of early explorers, summarized by Hachisuka (3, p. 85), definitely refer to dodos occurring in the uplands, and Hachisuka makes a point of clarifying the misconception that dodos were strictly coastal birds. Early forestry records from Mauritius (4) indicate that Calvaria was found in the lowlands as well as on the upland plateau. Although native forests only occur in the uplands today, one of the surviving Calvaria trees is located at an elevation of only 150 m. Thus, the dodo and Calvaria may have been sympatric, making a mutualistic relation possible.
Taxonomic authorities on sapotaceous plants of the Indian Ocean region recognize seeds of Calvaria major, as well as the smaller seeds of Sideroxylon longifolium, from alluvial deposits of the Mare aux Songes marsh (5), but this has little relevance to the question of mutualism. Mutualistic species will not necessarily be fossilized together.
The Mauritius Forestry Service has only recently succeeded in propagating Calvaria seeds, and the unmentioned reason for their recent success strengthens the case for mutualism. Success was achieved when the seeds were mechanically abraded before planting (6). A dodo’s digestive tract merely abraded the endocarp naturally the same way the staff of the Mauritius Forestry Service does artificially before the seeds are planted.
The reference Owadally cites (7) is equivocal about the age of the surviving Calvaria trees because there is no easy way to accurately date them. Coincidently, Wiehe, the coauthor of the paper Owadally cites, was also my source of the estimated age of over 300 years for the surviving trees. I agree that there were more trees surviving in the 1930’s than today, which further suppports the notion that Calvaria major is a declining species and may have been so since 1681.
I erred in not citing Hill (8). However, Hill does not describe how and under what conditions he induced a seed to germinate. Without these details, his description is of little relevance to the question of mutualism.
STANLEY A. TEMPLE
Department of Wildlife Ecology,
University of Wisconsin-Madison.
Madison 53706
References and Notes
 
; 1. S.A. Temple, Science 197, 885 (1977).
2. A. W. Owadally, ibid. 203, 1363 (1979).
3. M. Hachisuka, The Dodo and Kindred Birds (Witherby, London, 1953).
4. N. R. Brouard, A History of the Woods and Forests of Mauritius (Government Printer, Mauritius, 1963).
5. F. Friedmann, personal communication.
6. A. M. Gardner, personal communication.
7. R. E. Vaughan and P. O. Wiehe, J. Ecol. 19, 127 (1941).
8. A. W. Hill, Ann. Bot. 5, 587 (1941).
I think that Temple has responded adequately (even triumphantly) to Owadally’s first three points. As a paleontologist, I can certainly affirm his arguments about the rarity of upland fossils. Our fossil record of upland faunas is exceedingly spotty; the specimens we do possess are generally found in lowland deposits, well worn and washed in from higher ground. Owadally was certainly remiss in not mentioning (point 3) that the Forestry Service abrades its Calvaria seeds before they germinate; for the necessity of abrasion lies at the heart of Temple’s hypothesis. But Temple was equally remiss in not citing the local Mauritian efforts, which, apparently, predate his own discovery.
Owadally’s fourth point, however, represents the potential disproof of Temple’s claim. If “quite a significant population” of Calvaria trees were less than 100 years old in 1941, then dodos cannot have assisted their germination. Temple denies that so young an age has been demonstrated, and I certainly have no additional insight that can resolve this crucial question.
This exchange highlights a disturbing issue in the transmission of news about science to the public. Many sources cited Temple’s original story. I did not find a single mention of the subsequent doubts. Most “good” stories turn out to be false, or at least overextended, but debunking doesn’t match the fascination of a clever hypothesis. Most of the “classic” stories of natural history are wrong, but nothing is so resistant to expurgation as textbook dogma.
The debate between Owadally and Temple is too close to call at the moment. I’m rooting for Temple, but if Owadally’s fourth point is correct, then the dodo hypothesis will become, in Thomas Henry Huxley’s inimitable words, “a beautiful theory, killed by a nasty, ugly little fact.”
28 | Sticking Up for Marsupials
I AM ANNOYED that the rapacious ways of my own species have irrevocably prevented me from seeing the dodo in action, for a pigeon as large as a turkey must have been something else, and stuffed, moldy specimens just don’t carry conviction. We who revel in nature’s diversity and feel instructed by every animal tend to brand Homo sapiens as the greatest catastrophe since the Cretaceous extinction. Yet I would argue that the rise of the Isthmus of Panama a mere two to three million years ago must rank as the most devastating biological tragedy of recent times.
South America had been an island continent throughout the Tertiary period (for seventy million years before the onset of continental glaciation). Like Australia, it housed a unique suite of mammals. But Australia was a backwater compared with the range and variety of South American forms. Many survived the onslaught of North American species after the isthmus rose. Some spread and prospered: the opossum moved as far as Canada; the armadillo is still making its way north.
Despite the success of a few, extirpation of the most dramatically different South American forms must be ranked as the dominant effect of contact between mammals of the two continents. Two entire orders perished (we group all modern mammals into about twenty-five orders). Think how our zoos would have been enriched with a liberal sprinkling of notoungulates, a large and diverse group of plant-eating mammals, ranging from rhino-sized Toxodon, first exhumed by Charles Darwin on shore leave from the Beagle, to rabbit and rodent analogues among the typotheres and hegetotheres. Consider the litopterns with their two subgroups—the large, long-necked camel-like macrauchenids and the most remarkable group of all, the horse-like proterotheres. (Proterotheres even repeated some of the evolutionary trends followed by true horses: three-toed Diadiaphorus preceded Thoatherium, a single-toed species that outdid Man ’O War by reducing its vestigial side toes to a degree never matched by modern horses.) They are all gone forever, victims in large part of faunal disruptions set in motion by the rising isthmus. (Several notoungulates and litopterns survived well into the glacial epoch. They may even have received their coup de grâce from early human hunters. Still, I do not doubt that many would still be with us if South America had remained an island.)
The native predators of these South American herbivores also disappeared completely. The modern carnivores of South America, the jaguars and their allies, are all North American interlopers. The indigenous carnivores, believe it or not, were all marsupials (although some flesh-eating niches were occupied by the phororhacids, a remarkable group of giant birds, now also extinct). The marsupial carnivores, although not as diverse as placental carnivores in northern continents, formed an impresive array, from fairly small animals to bear-sized species. One lineage evolved in uncanny parallel with the saber-toothed cats of North America. The marsupial Thylacosmilus developed long, stabbing upper canines and a protecting flange of bone on the lower jaw—just like Smilodon of the La Brea tar pits.
Although it is not commonly bruited about, marsupials are not doing badly in South America today. North America may only boast the so-called Virginia opossum (actually a South American migrant), but opossums in South America are a rich and varied group of some sixty-five species. In addition, the caenolestids, pouchless “opossum rats,” form a separate group with no close affinity to true opossums. But the third great group of South American marsupials, the carnivorous borhyaenids, were completely wiped out and replaced by northern cats.
The traditional view—though I dedicate this essay to opposing it—attributes the extirpation of carnivorous marsupials to the general inferiority of pouched versus placental mammals. (All living mammals except marsupials and the egg-laying platypus and echidna are placentals.) The argument seems hard to beat. Marsupials flourished only on the isolated island continents of Australia and South America where large placental carnivores never gained a foothold. The early Tertiary marsupials of North America soon disappeared as placentals diversified; South American marsupials took a beating when the Central American corridor opened for placental immigration.
These arguments of biogeography and geological history gain apparent support from the conventional idea that marsupials are anatomically and physiologically inferior to placentals. The very terms of our taxonomy reinforce this prejudice. All mammals are divided into three parts: the egg-laying monotremes are called Prototheria, or premammals; placentals win the prize as Eutheria, or true mammals; the poor marsupials lie in limbo as Metatheria, or middle mammals—not all quite there.
The argument for structural inferiority rests largely upon differing modes of reproduction in marsupials versus placentals, bolstered by the usual smug assumption that different from us is worse. Placentals, as we know and experience, develop as embryos in intimate connection with a mother’s body and blood supply. With some exceptions, they are born as reasonably complete and capable creatures. Marsupial fetuses never developed the essential trick that permits extensive development within a mother’s body. Our bodies have an uncanny ability to recognize and reject foreign tissues, an essential protection against disease, but a currently intractable barrier to medical procedures ranging from skin grafts to heart transplants. Despite all the homilies about mother love, and the presence of 50 percent maternal genes in offspring, an embryo is still foreign tissue. The maternal immune system must be masked to prevent rejection. Placental fetuses have “learned” to do this; marsupials have not.
Marsupial gestation is very short—twelve to thirteen days in the common oppossum, followed by sixty to seventy days of further development in the external pouch. Moreover, internal development does not proceed in intimate connection with the mother, but shielded from her. Two-thirds of gestation occurs within the “shell membrane,” a maternal organ that preven
ts the incursion of lymphocytes, the “soldiers” of the immune system. A few days of placental contact follow, usually via the yolk sac. During this time, the mother mobilizes her immune system, and the embryo is born (or, more accurately, expelled) soon after.
The marsupial neonate is a tiny creature, equivalent in development to a rather early placental embryo. Its head and forelimbs are precociously developed, but the hind limbs are often little more than undifferentiated buds. It must then undertake a hazardous journey, slowly pulling itself along through the relatively great distance to mother’s nipples and pouch (we can now understand the necessity of well-developed forelimbs). Our embryonic life within a placental womb sounds altogether easier and unconditionally better.
What challenge can then be offered to these biogeographical and structural accounts of marsupial inferiority? My colleague John A. W. Kirsch has recently marshaled the arguments. Citing work of P. Parker, Kirsch contends that marsupial reproduction follows a different adaptive mode, not an inferior path. True, marsupials never evolved a mechanism to turn off the maternal immune system and permit a completed development within the womb. But early birth may be an equally adaptive strategy. Maternal rejection need not represent a failure of design or lost evolutionary opportunity; it may reflect an ancient and perfectly adequate approach to the rigors of survival. Parker’s argument goes right back to Darwin’s central contention that individuals struggle to maximize their own reproductive success, that is, to increase the representation of their own genes in future generations. Several highly divergent, but equally successful, strategies can be followed in (unconscious) pursuit of this goal. Placentals invest a great deal of time and energy in offspring before their birth. This commitment does increase the chance of an offspring’s success, but the placental mother also takes a risk: if she should lose her litter, she has irrevocably expended a large portion of her life’s reproductive effort for no evolutionary gain. The marsupial mother pays a much higher toll in neonatal death, but her reproductive cost is small. Gestation has been very short and she may breed again in the same season. Moreover, the tiny neonate has not placed a great drain upon her energetic resources, and has subjected her to little danger in a quick and easy birth.
The Panda’s Thumb Page 26