This article is reproduced with the permission of New Scientist for exclusive use by Nova users.

Reading the entrails of evolution: Convergent evolution produces similarities that are more than skin deep. The digestive systems of marsupials and placental mammals show how alike such distant relatives can become when the need arises
15 April 1989
From New Scientist Print Edition.
Ian Hume

Marsupials, such as kangaroos and koalas, and placental mammals, such as horses and humans, arose from a common ancestor about 120 million years ago. They have gone their separate ways ever since. The two groups of mammals differ fundamentally in reproductive strategy. Marsupials give birth to young that still are embryonic, while newborn placental mammals are far more mature. But both are thoroughgoing mammals, and so the two generally occupy similar ecological niches.

Thus it is that the two kinds of mammal now provide some of nature's finest examples of convergent evolution - the process by which organisms of different types come to resemble each other, as they each strive to overcome similar ecological problems. Some of the convergences are obvious: the marsupial gliders resemble the flying squirrels; the thylacine 'tiger' was like a dog; and there is even a marsupial 'mole'. But it is becoming clear, now, that the convergences are more than skin deep. The more we explore marsupial nutrition and digestion, especially among the herbivores, the more remarkable and detailed the convergences appear.

To understand fully the nature and extent of these convergences we must look first at basic nutritional principles. All vertebrate animals need water, energy, protein, essential fatty acids, minerals and vitamins to stay alive. They do not all need exactly the same nutrients, but the differences are minor: for example, most vertebrates can synthesise their own vitamin C, but the primates, bats, guinea pigs and some perching birds and bony fish must obtain vitamin C ready-made in their diet. These animals lack the enzyme L-gulonolactone oxidase, which catalyses the final reaction in the biosynthetic pathway that creates the vitamin.

The significant differences between vertebrates lie in the relative amounts of the various nutrients, and in the ways in which they go about harvesting and extracting their requirements from the environment. These two factors, the amounts required and the digestive strategies employed, define the nutritional niche of an organism. The smaller the amount of any particular nutrient that the animal requires, and the more completely it can extract that nutrient, the broader its nutritional niche will be. For example, cats are more or less obliged to eat animal material because they have a very high requirement for protein and have only a short and simple digestive system. Their nutritional niche is very narrow.

Herbivores, on the other hand, often occupy very broad nutritional niches. Their nutrient requirements are often relatively low and their digestive systems are so complex that they can sustain themselves on fodder as unpromising as lichens, dry desert grass and the gummy, toxic and extremely fibrous leaves of Eucalyptus.

Herbivores rely heavily for their digestion upon microorganisms which live within their gut. Most of these are bacteria but sometimes protozoa and fungi are also present. The microbes' main task is to break down cellulose, which is the chief material in the walls of plant cells. Cellulose could be an excellent source of energy, precisely because it is the commonest carbohydrate in nature. Yet no vertebrate animal is known to produce a cellulase enzyme capable of digesting cellulose; only microbes and some lower invertebrates produce cellulases. Not all herbivores rely heavily upon cellulose for energy, but for many it is the principal source of energy. Without their cargoes of microbes they would be unable to make use of it at all. In addition, unless microbes break down the plant cell walls, the herbivore's own digestive enzymes would have difficulty gaining access to the cell's contents. The reason that herbivores generally have complex digestive tracts is to slow the passage of food, and so prolong contact with the microbes that help to digest it. The digestion of cellulose is an anaerobic process, of the kind commonly known as fermentation. All parts of the gut absorb the volatile fatty acids produced in the process and these provide the herbivore with energy.

Many of the animals that derive a significant proportion of their energy from the cell walls of plants are large: they include elephants, wildebeest, zebras and kangaroos. Large size confers two advantages on such 'fibrivores'. First, the amount of energy that a warm-blooded animal requires is related to its total surface area, because such animals use a large proportion of their energy in providing heat, and heat is lost through the surface. The volume of an animal increases in proportion to the cube of its linear dimensions, whereas its surface area increases only in proportion to the square of its linear dimensions. So compared with small ones, big animals have a smaller surface area relative to their volume, and although they need to eat more, they require less food per unit body weight. This applies to all nutrients - protein and vitamins as well as energy. Large herbivores, in general, should occupy broader nutritional niches.

Secondly, animals with large bodies can also have large guts, and so can contain large amounts of material that is fermenting slowly. Large herbivores generally retain their food in their guts for longer than smaller animals, and this maximises the opportunity for the breakdown of cellulose. Small herbivores generally depend for their energy more upon the contents of a cell than its walls, and so they must select plant material of higher quality. As a result, their nutritional niche is likely to be narrower.

Herbivores fall into two groups, depending on where in their gut microbial fermentation takes place. Some, the foregut fermenters, have an expanded and modified fore-stomach, that is either like a bag (sacciform), as in ruminants such as cattle and sheep, or else is partly tube-like (tubiform) as in langur monkeys. In the hindgut fermenters, fermentation takes place mainly in an enlarged caecum, which is a blind diversion of the gut between the small and large intestine; or in the large intestine itself, the colon. Those in which fermentation takes place mainly in the colon are typically large, and include dugongs and manatees, elephants and the odd-toed ungulates (horses, zebras, tapirs and rhinos). Fermentation in these animals takes place mainly in the front part of the colon (the proximal colon) and if they have a caecum at all, then this seems to function as an extension of the proximal colon, to form a single fermentation region.

Animals that do most of their fermenting in the caecum are typically small: lagomorphs, such as pikas, rabbits and hares; and rodents, such as beavers, porcupines and guinea pigs. In these animals, there is usually little microbial activity in the colon. The caecum fermenters have an interesting digestive strategy that enables them to tolerate diets that are higher in fibre than you might anticipate was possible for animals of their size. Colon fermenters tend to retain large particles in the gut for longer than they retain fine particles and fluid; this maximises the digestion of large particles, which are mainly cellulose. But most caecum fermenters retain fluid and fine particles selectively in the caecum, and excrete the large, poorly digestible, fibrous particles relatively rapidly.

Often, too, caecum fermenters practise coprophagy (that is, they eat their faeces), or caecotrophy - whereby they eat soft faeces, caecotrophs, derived from the contents of the caecum. Thus they salvage at least some of the protein and B-vitamins synthesised by the caecal microbes, and recycle them to the small intestine, where most amino acids and B-vitamins are absorbed. In this way small caecum fermen- ters increase their intake of valuable nutrients in order to meet their relatively high requirements. Colon fermenters do not eat faeces and so lose all the nutrients released by the microbes except the volatile fatty acids that are absorbed throughout the gut. But because colon fermenters are large, and need relatively little nutrient, they can generally get all that they need from the diet.

The examples so far have all been of placental mammals; but each strategy has its counterpart in marsupials. A few marsupials, such as the kangaroos and koala, have extra refinements of their own. To see how these similarities and differences have come about, we should briefly review marsupial history.

Marsupials are thought to have originated not in Australia, or even in South America (where there are still many species) but in North America, which now has only one remaining native marsupial, the opossum. They are thought to have 'rafted' their way to South America (on floating vegetation) and then moved on to Antarctica. At the time, Antarctica was not completely covered by ice, and it was connected both to South America and to Australia as part of the great southern continent, Gondwanaland. By the time Gondwanaland broke up about 45 million years ago the basic marsupial stock had established itself on the Australian land mass. As Antarctica moved south and became glaciated Australia and South America became refuges for marsupials.

The next 30 million years were important for mammals. Climates were changing rapidly worldwide, and mammals reached their greatest diversity. One event in particular played a vital role in the radiation of herbivorous mammals: the grasses appeared in the early Miocene epoch, about 25 million years ago. The rapid spread of the grasslands corresponds with the rise of the ruminants, and their apparent dominance over the odd-toed ungulates, in North America, Eurasia and Africa.

Throughout this time, however, Australia was isolated from the other continents. It too experienced enormous changes of climate, partly as a result of the worldwide changes, partly because it was constantly changing latitude, drifting north to reach its present position off the southeast tip of Asia around 15 million years ago. The rapid journey north produced a wide range of climatic and vegetational zones, and so of nutritional habitats, across the Australian continent. Just as other mammals radiated on other continents, so the marsupials radiated in Australia to exploit those habitats. The grasses also appeared in Australia in the Miocene, albeit later than on the other continents, and they too produced a radiation in the grazing animals: not in ruminants and perissodactyls, for there were none, but in kangaroos and wombats.

The ancestors of today's marsupial herbivores were tree-living insectivores or omnivores. The gut of the modern insectivorous and carnivorous mammals, whether placental or marsupial, consists of a simple, globular stomach, a short and simple small intestine, and a short and simple colon. Most placental carnivores have a small or rudimentary caecum, and so, too, do the remaining American marsupials; but the Australian marsupial carnivores, which make up the family Dasyuridae and include the misnamed marsupial cats, or quolls, lack a caecum.

Omnivores eat significant amounts of plants, as well as meat, and accordingly the caecum shows some limited development into an organ of fermentation. The first herbivores are thought to have arisen from omnivorous ancestors as the caecum, or some other part of the hindgut, developed into a fermentation organ that supplied a significant proportion of the animal's energy. So, fermentation in the hindgut is probably more primitive than in the foregut. The fact that some secondary fermentation takes place in the hindgut of all foregut fermenters supports this idea.

Among marsupial hindgut fermenters, the only true colon fermenter is the wombat, which is a large terrestrial grazer. Perry Barboza at the University of New England in Armidale, New South Wales, studied the digestive physiology and nutrition of wombats. In these marsupials, the colon provides 68 per cent of the total capacity of the digestive tract. Predictably, food remains in their guts for a long time. In common with other colon fermenters, the wombat retains coarse particles longer than fluid and fine particles; like the placental horse, then, it can make good use of fibrous grass. In both horses and wombats the colon has longitudinal bands of muscle (taeniae), with folds between them. The indentations (haustra) formed by these folds help to retain large particles of food.

Caecum fermentation is found in the possums and gliders, which are arboreal leaf-eaters. In the greater glider, Petauroides volans, fermentation takes place only in the caecum, which retains fine particles and fluid. Here, then, is yet another example of convergence between marsupials and placental mammals - this time with rodents and lagomorphs. The ringtail possums, species of Pseudocheirus, which are the smallest marsupial caecum fermenters, practise caecotrophy, in a way remarkably similar to the rabbit and other lagomorphs. Unlike the rabbit, however, the ringtail possum has no obvious anatomical specialisations of the proximal colon to explain how fluid is forced back into the caecum when hard faeces are formed.

Michael Chilcott, also at the University of New England, showed that caecotrophy makes an enormous difference to the nutrition of the animals that practise it. The energy recycled by ingestion of caecotrophs is equivalent to 58 per cent of the animal's intake of digestible energy, while the nitrogen recycled is 126 per cent of nitrogen intake. The extra nitrogen comes from protein that the microbes in the gut synthesise using nitrogen from sources such as urea.

In the size of its hindgut, and in its fixation upon the leaves of Eucalyptus as a source of food, the koala is unique. The English anatomist Sir Richard Owen remarked in 1868 that: 'In the koala the caecum and large intestines arrive at their maximum of development.' The koala exhibits the anatomical features of both caecum and colon fermenters. Steven Cork, at the University of New South Wales, showed that koalas ferment their food throughout the caecum and proximal colon, and selectively retain fluid and fine particles in both those regions. The koala does not eat its faeces or caecotrophs, except during the six weeks around the time of weaning. During this period the young takes a soft material known as 'pap' from its mother's cloaca. This may serve to inoculate the young koala's hindgut with the most appropriate microflora. It may also help the young to learn which species of Eucalyptus they should eat.

The koala's narrow preference for just a few species of Eucalyptus is legendary. Of the 500 or so eucalypts, it eats only about 20 regularly. In any one region, koalas will eat three or four species as staples, and perhaps another three or four now and again. Eucalyptus foliage is an unpromising food. It is low in nutrients such as nitrogen and minerals, and its fibre contains a lot of indigestible lignin. Eucalyptus leaves also contain 'anti-nutrients', agents that lower the availability of nutrients or are toxic. These include chemicals such as tannins, which reduce the availability of nitrogen, plus other phenolics and essential oils.

Although koalas are the most famous and the most committed eucalypt feeders, there are three others: the greater glider, the ringtail possum and the brushtail possum, Trichosurus vulpecula. Each of these species has physiological adaptations to counter the effects of anti-nutrients. They may be able to detoxify the essential oils in their livers. In addition, marsupials have a lower resting metabolic rate than most placental mammals, so they are more economical with energy. But the basal metabolic rate of the koala is much lower even than the 'marsupial mean', and of the four arboreal mammals that eat eucalypts, the koala is the only one in which this is so. As a result koalas do not need to eat as much as other mammals of comparable size. This is doubly advantageous, because detoxification requires energy and nitrogen - so the less eaten the better.

When the intake of food is low and the capacity of the digestive tract is large, the food passes through slowly, and microbes have plenty of opportunity to break down plant cell walls. Furthermore, because the animals that specialise in Eucalyptus leaves retain fluid and fine particles selectively, the digestive effort is concentrated on the components that potentially are more fermentable. In contrast, the brushtail possum, Trichosurus vulpecula, cannot select which components it retains. My colleague Bill Foley and I believe that this is the main reason why this animal is more dependent on other foods than are the other three specialist eucalypt feeders.

Eucalypt feeders also need relatively little protein (nitrogen). We found that the four Eucalyptus specialists need different amounts of nitrogen, probably depending on how much essential oil and phenolics their food contains. Generally, leaf-eating marsupials seem to need less nitrogen than most placental mammals.

Among marsupials, the only foregut fermenter (the equivalent of the ruminants) is the kangaroo. The kangaroo's expanded and differentiated forestomach looks more like the proximal colon of a horse than the rumen of a cow, however. The wall of the forestomach is organised into three longitudinal bands of muscle, the taeniae. Folds between the taeniae form the haustra. Contractions of the haustra propel food towards the rear, while at the same time the folds selectively retain particles rather than fluid. The overall effect is to squeeze fluid through the particles, so that the fluid passes through the digestive tract far more quickly.

The progress of particles and fluid through the gut varies and has important consequences for the nutrition of the animal. We can show how long each component takes to pass through the gut by labelling them with radioactive isotopes. In one experiment, we gave both kangaroos and sheep single doses of two radioactive markers. Chromium-51 complexed with EDTA labels the fluid part of the feed, while ruthenium-103 complexed with phenanthroline labels the particles. In sheep, the labelled fluid and particles appeared at their highest concentration in the faeces at roughly the same time. But in the kangaroos the highest concentrations of the marked fluid appeared in the faeces within less than 20 hours, while the marked particles took almost twice as long to reach their peak. In kangaroos, the flow of food through the stomach is basically tubular, and the marker does not mix with food from other meals. But in sheep, the marker mixes in with a large pool of food derived from several meals.

David Dellow at the University of New England examined the nutritional consequences of these different patterns of flow. He found that in the kangaroo, the main material fermented in the sacciform region of the stomach was soluble carbohydrate. When the food reached the tubiform region the main target of digestion was fibre. Perhaps because of this, most gut protozoa inhabit the sacciform region, and very few live in the tubiform forestomach. But kangaroos produce volatile fatty acids from cellulose as efficiently as does the ruminant sheep.

However, the colon-like stomachs of the kangaroo cannot retain food material as long as the sheep can do in its rumen, so the overall digestibility of fibre is almost always lower in kangaroos. This may seem a disadvantage, but they can compensate to some extent. Thus ruminants tend to eat less as the quality of feed declines, that is, as the fibre increases, but kangaroos are better able to maintain feed intake.

In the forestomach of wallabies the sacciform region is relatively larger and the tubiform region relatively shorter than in the larger kangaroos, so they are not able to maintain their feed intake in the same way. But because they can retain the food material longer in the sacciform forestomach, the overall retention time is about the same as in the much larger macropods, and they digest fibre to a similar extent.

A rich diet for small animals

As I noted earlier, smaller warm-blooded animals tend to have a higher metabolic rate, weight for weight, than larger ones, and so need more nutritious food. We would expect that small kangaroos would eat less fibrous plants than wallabies or the large kangaroos: and that they might even veer away from herbivory altogether. This indeed is what we find in the small rat-kangaroos of the family Potoroidae. The rufous rat-kangaroo, Aepyprymnus rufescens, has an enormous sacciform forestomach, and a very short tubiform forestomach. It eats both fungi and herbs, but the parts of the plants that it selects, such as seeds and roots, are much less fibrous than the leaves and stems that a kangaroo eats. The musky rat-kangaroo, Hypsiprymnodon moschatus, which belongs to its own subfamily, has a simple stomach, and basically eats insects or fungi.

Ian Wallis, at the University of New England, found that on a diet of chopped lucerne, designed to simulate the natural diet of a kangaroo, rat-kangaroos digested fibre as well as kangaroos do - but they could not eat enough of the food to maintain themselves. Yet, on a diet rich in starch, designed to simulate their natural diets, rat-kangaroos digested fibre very poorly. This led my colleague Carol Carlisle and me to examine the movements of food in rat-kangaroos by X-ray. We found that food often largely bypassed the sacciform forestomach. We suggest that the primary function of this region could be to store food, and that microbial fermentation is often of secondary importance. After all, small herbivores feeding in the open are very vulnerable to predators, and an ability to eat quickly and store what they eat could be very useful. Many rodents adopt a similar strategy, storing food in their cheek pouches.

Many of the nutritional strategies of marsupials seem to equip them especially well to life in harsh conditions. Because marsupials often have a lower basal metabolic rate than placental mammals, they need less energy and protein for maintenance. And although the digestive strategies of herbivorous marsupials resemble those of placental mammals, they also have unique features that help them to overcome problems that seem particularly trying, including, in the case of the koala, survival on a diet of eucalypt leaves.

Some biologists have suggested that marsupials are 'inferior' to eutherians. But this is no more than the prejudice of eutherian human beings. Marsupials should be seen as alternatives, which do almost all the things that eutherians can do, in ways that are sometimes strikingly different, and sometimes intriguingly similar - and in general, equip them to survive in conditions in which many a eutherian would surely perish.

Ian Hume is professor of biology at the University of Sydney, Australia.

From issue 1660 of New Scientist magazine, 15 April 1989, page

For the latest from New Scientiist visit www.newscientist.com