Life of Marsupials

By Hugh Tyndale-Biscoe

Over the past half a century research has revealed that marsupials – far from being ‘second class’ mammals – have adaptations for particular ways of life quite equal to their placental counterparts. Despite long separate evolution, there are extraordinary similarities in which marsupials have solved the challenges of living in such environments as deserts, alpine snowfields or tropical rainforests. Some can live on grass, some on pollen and others on leaves; some can glide, some can swim and others hop with extraordinary efficiency.

In Life of Marsupials, one of the world’s leading experts explores the biology and evolution of this unusual group – with their extraordinary diversity of forms around the world – in Australia, New Guinea and South America.

Joint winner of the 2005 Whitley Medal.

Included in Choice Magazine's 2006 Outstanding Academic Titles list.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Apr 22, 2005
• ISBN: 9780643099210

Book preview

LIFE OF MARSUPIALS

HUGH TYNDALE-BISCOE

© Hugh Tyndale-Biscoe 2005

All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO PUBLISHING for all permission requests.

National Library of Australia Cataloguing-in-Publication entry

Tyndale-Biscoe, Hugh.

Life of marsupials.

[New ed.].

Bibliography.

Includes index.

ISBN 0 643 06257 2 (Hardback).

ISBN 0 643 09199 8 (Paperback).

ISBN 0 643 09220 X (netLibrary eBook).

1. Marsupials. I. CSIRO Publishing. II. Title.

599.2

Available from

CSIRO PUBLISHING

PO Box 1139 (150 Oxford St)

Collingwood VIC 3066

Australia

Telephone: [phone]

Local call: 1300 788 000 (Australia only)

Fax: [phone]

Email: [e-mail]

Web site: www.publish.csiro.au

Front cover (clockwise from top left):

Gray four-eyed opposum (Hugh Tyndale-Biscoe); Stages in tammar wallaby development: the unattached vesicle (Ivan Fox) and a newborn tammar (LA Hinds); Chromosome painting (JAM Graves); Feathertail glider (Ederic Slater); Julia Creek dunnart with 60-day-old litter (PA Woolley and D Walsh).

Spine:

Yellow-footed rock wallaby (Esther Beaton).

Back cover:

Male honey possum on Banksia inflorescence (PA Woolley and D Walsh).

Set in Minion 10/12

Cover and text design by James Kelly

Typeset by J&M Typesetting

Printed in Australia by Ligare

Contents

Preface

1 What is a marsupial?

2 Reproduction and development

3 Opossums of the Americas: cousins from a distant time

4 Predatory marsupials of Australasia: bright-eyed killers of the night

5 Bandicoots: fast-living opportunists

6 Pygmy possums and sugar gliders: pollen eaters and sap suckers

7 Life in the trees: koala, greater glider and possums

8 Wombats: vegetarians of the underworld

9 Consummate kangaroos

10 Marsupials and people: past and present

References

Index

Preface

When the first edition of this book was written in 1970, the old debate about the inferior status of marsupials, compared to other mammals, was still active. The work reviewed then on a few species of marsupial in Australia and North America was beginning to dispel this idea but it still prevailed in other countries, particularly in the Northern Hemisphere. Thirty-five years later it is no longer an issue of importance. Now, much more is known about the past history and the present relationships of marsupials in Australia, New Guinea and South and Central America, so that the long evolution of this separate line of mammals is becoming much clearer. As well as this, there are now detailed studies on the physiology, reproduction, ecology and behaviour of representatives of all the main families of marsupials, so that comparisons and generalisations can be made with much more confidence.

In the first edition of this book it was also still possible to cover the whole literature on marsupials. That is no longer possible and a small book now must be selective in its coverage and its acknowledgement of sources. However, for most topics and for most groups of marsupials there are now excellent monographs or reviews that enable the interested reader to follow any topic further.

Several themes about marsupials have developed among the people who have studied them during the past 35 years and these resonate through all the work that is described here. The oldest of these themes is the remarkable convergence of adaptations seen in Australasian marsupials and mammals on other continents. When first seen by European explorers these similarities were thought to indicate close relationship but deeper understanding soon showed that these were two independent evolutionary lines responding to similar external imperatives. As well as these large convergences we can now recognise convergences between marsupials from Australasia and those from South America. Within the Australasian marsupials there are also convergences to similar food sources, such as the leaf eating koala, possums and ringtails, or the nectar-eating marsupials from four separate families. However, the most interesting outcome of the new work on marsupials has been a much greater appreciation of how marsupials have adapted to the special conditions of the Australian environment, its unpredictable climate, low fertility soils and unpalatable plants. It is an important and interesting aspect of the adaptive radiation of marsupials in Australia and raises the question how marsupials came to prevail in Australia but shared South America with other kinds of mammals: it also tells us how we must adapt to the land if we wish to live here in the long term, and what we must do to let these long time residents continue to live here also.

Because it is not possible for one person to command a knowledge of so many fields as this book covers, I have depended on the expert advice of colleagues in several fields: while taking full responsibility for what is written, I am deeply grateful for the generous help of Ken Aplin, Bill Foley, Jennifer Graves, Brian Green, Stephen Ho, Peter Janssens, John Kirsch, the late Richard Mark, Lauren Marotte, Jim Merchant, David Ride, Phil Waite, Mike Westerman and Patricia Woolley. For each reading several chapters as a non-expert and thereby helping me to express things more clearly than I otherwise would have, I sincerely thank Meredith McKinney and Nicola Tyndale-Biscoe.

I am also very pleased to acknowledge CSIRO: this great organisation has supported research on Australasian marsupials since 1950 and my own research for more than 40 years, so that much that is discussed in this book stems directly from that support. Then, when I began this book Brian Walker, Chief of CSIRO Wildlife and Ecology, offered me generous and stimulating hospitality to prepare it, and his successors in CSIRO Sustainable Ecosystems have graciously continued to do so, to its completion. At CSIRO Sustainable Ecosystems I have been wonderfully well supported by many people and I especially thank Alice Kenney for preparing the figures; Margaret Hindley, Trish Kelly, Megan Edwards and Inge Newman for tracking down difficult or unusual references with speed and efficiency; and Andrew Bishop, Brian Davis and Yechiam Marks for leading me courteously through the complexities of information technology. At CSIRO Publishing I thank Paul Reekie for great patience as deadlines passed and Nick Alexander and Briana Elwood for producing the finished work with diligence and despatch. I also thank Alexa Cloud for superb copy editing.

Finally, I thank Marina, who read and commented on every chapter in draft and then read the proofs, and has sustained me throughout the whole saga as one year passed into another and the end remained a mirage too far away: thank you for everything.

Hugh Tyndale-Biscoe

January 2005

Chapter 1

What is a marsupial?

This is the story of a group of mammals that were isolated from the rest of the world for many millions of years. It is set on the great southern continent of Gondwana that stretched from the Caribbean to the islands of New Guinea and included the three present-day continents of South America, Antarctica and Australasia. The characters are the marsupial mammals and the plot is how they came to be there and how they adapted to the special conditions of their vast homeland.

Today marsupials only occur in Australasia and the Americas, although fossil marsupials have been discovered on every continent of the world. If they occurred on all the continents in the past, why are they not more widely distributed today? We first need to ask whether marsupials really share a common relationship closer than that to any other group of mammals. If they do, where did marsupials originate and how did they come to be where they are today?

The features that were described first by the Europeans were the pouch of the female and the extraordinarily small size of the young at birth. These are their two most distinctive features and reproduction is what sets marsupials apart from other mammals and permeates the life history of every species. But what confused the early European explorers was that many marsupials closely resembled mammals more familiar to them that follow similar life styles.

First European encounter with American marsupials

The first marsupial brought to Europe from America was a common opossum collected by Vincente Yañez Pinzón on his first voyage to the New World in 1500. He collected a female, which had young in its pouch, and later he presented it to Queen Isabella and King Ferdinand II in Grenada. By the end of the voyage to Spain the young were gone and the mother opossum dead but the Queen inserted her fingers into the ‘second belly’ of this strange creature from the New World. This extraordinary organ and the young it enclosed caused astonishment in scientific circles throughout Europe and led to speculation about how the tiny young reached the pouch – did they grow from the teats as buds, or were they blown into the pouch from the mother’s nostrils? Both ideas had a long currency but the equally astonishing fact that they crawl to the pouch unaided by their mother was not discovered for another 420 years. The strange appearance of this New World animal, with a fox-like head and a monkey’s hands, also puzzled European scientists, who described it as the monkey-fox or ‘simivulpa’.

First European contact with Australasian marsupials

In 1493 Pope Alexander IV divided the world between the two major European powers – Spain to the west and Portugal to the east – so it was Spanish explorers who discovered marsupials in the Americas, while Portuguese traders made the first observations on Australasian marsupials. The earliest description was by Antonio Galvao, Station Captain of the Portuguese settlement on Ternate in the Moluccas from 1536 to 1540. He brought back to Lisbon extensive notes from which he intended to write a treatise on the Moluccas. He wrote (Jacobs 1971):

Some animals resemble ferrets, only a little bigger. They are called Kusus. They have a long tail with which they hang from the trees in which they live continuously, winding it once or twice around a branch. On their belly they have a pocket like an intermediate balcony; as soon as they give birth to a young one they grow it inside there at a nipple until it does not need nursing any more. As soon as she has borne and nourished it, the mother becomes pregnant again.

This is a good description of the common cuscus, Phalanger orientalis, which still lives on Ternate. Galvao’s manuscript lay forgotten in the Jesuit Library at Seville for 400 years, until discovered and published by Father Hubert Jacobs in 1971. However, the manuscript may well have circulated in Europe because there are other references in the 17th century to animals similar to the opossum being found in the Moluccas (eg Piso 1648).

As the rivalry of Portugal and Holland for the rich takings of the Spice Islands increased, travellers also reported on the strange animals of New Holland and New Guinea. In 1606 Captain Don Diego de Prado y Tovar landed at San Millaus Bay, on the southern coast of New Guinea and wrote (quoted by Stevens 1930):

Here we killed an animal which is in the shape of a dog, smaller than a greyhound, with a bare scaly tail like that of a snake, and his testicles hang from a nerve like a thin cord; they say it was the castor, we ate it and it was like venison, its stomach full of ginger leaves and for that reason we ate it.

The species that most nearly fits this description (Calaby 1965) is the pademelon Thylogale brunii, which makes this the earliest European record of a member of the kangaroo family.

The Dutch Captain Pelsaert, wrecked on the inhospitable Abrolhos Islands off the west coast of Australia in 1625, described another animal with the same remarkable pouch and tiny offspring within it. This was the tammar wallaby, Macropus eugenii, the second member of the kangaroo family to be described by Europeans. Then on Dirk Hartog Island, the buccaneer, William Dampier in 1699, described the banded hare wallaby, Lagostrophus fasciatus:

A sort of raccoons, different from those of the West Indies chiefly as to their legs; for these have very short fore-legs; but go jumping upon them as the others do, and like them are very good meat.

Other navigators were also encountering similarly bizarre animals when they made landfalls on the coast of Western Australia and, from their reports, scientists in Europe recognised the similarities to the American animals. Pallas (1766) named the cuscus from the Moluccas Didelphis orientalis but Storr (1780) 14 years later noticed that the 2nd and 3rd toes or phalanges of their hind feet are partly fused and changed it to Phalanger orientalis, the name it has today. It has also given its name to the family of Australasian possums and cuscuses, the Phalangeridae.

Adaptive radiations on different continents

While all the early European observers were much struck by the pouch of female marsupials, they were also struck by the astonishing similarities of Australian marsupials to familiar mammals from Europe. In the scientific names they gave them they often used Greek prefixes that meant pouch or pocket and the scientific name of the familiar mammal that the marsupial resembled. So we find Pera-meles the pouched badger, Pera-dorcas the pouched antelope, Phascol-arctos, the pouched bear, Thylo-gale, the pouched hare and Thyla-cinus, the pouched dog.

As the Australian continent was explored yet more remarkable similarities were discovered, the most extraordinary being the marsupial mole, Notoryctes typhlops, which looks and behaves like the golden mole of Namibia, Eremitalpa granti, especially in that both species ‘swim’ through dry sand, which collapses behind them leaving no burrow; and the exquisite numbat, Myrmecobius fasciatus, adapted to living on termites and ants. The tree-living marsupials of Australia and New Guinea show remarkable yet superficial likenesses to the various species of lemur of Madagascar (ie Daubentonia madagascarensis) and the tree sloths (ie Bradypus tridactylus) of South America. More remarkable yet are the many physiological similarities between marsupials and other mammals in such functions as fermentation of grass in the forestomach of kangaroos and ruminants. Apart from a pouch, what do all these species that are called marsupials have in common that tells us that they are uniquely related to one another and separate from all the other mammals that they variously resemble?

Distinctive features of marsupials

The person who first looked beneath the superficial similarities to find the fundamental criteria for determining relationships between mammals was de Blainville in 1816. He took as the defining character the anatomy of the female reproductive tract (Fig. 1.1). In marsupials there are two vaginae, two uteri and two oviducts, whereas in other mammals there is a single vagina, cervix and uterus and only the oviducts are paired. He named the marsupials the Di-delphia from the Greek words for two uteri and other mammals he called the Mono-delphia. He later realised that the platypus and echidna did not fit in either group, having a reproductive tract like that of birds and reptiles with a single opening for discharging products from the gut, the bladder and the gonads, and he termed them the Ornitho-delphia, or bird-uterus. His division of mammals has stood to this day, although his terms have not been retained. Instead, the Ornithodelphia are known as the Monotremata (Greek for one hole), the Didelphia are generally called the Marsupialia (Latin for pocket or pouch) and the Monodelphia are called the Placentalia, because of the great development of the placenta as an organ of exchange during pregnancy. These terms are also unsatisfactory because they are not exclusive. Thus, the marsupials share with the monotremes a single opening, the cloaca, for the discharge of all products. Second, not all female marsupials possess a well-developed pouch, whereas the female echidna, which is a monotreme, develops a pouch during lactation. And third, all marsupials have a placenta during intra-uterine development and in some species it is a complex structure with an intimate connection to the uterus.

To avoid this confusion of terms some biologists favour Huxley’s (1880) terms for the living mammals. He saw the three groups of mammals as evolutionary stages on the way to ‘true’ mammals, by which he meant the group of mammals to which we human beings belong. So he coined the terms Proto-theria, or first mammals (the monotremes), Meta-theria or halfway mammals (the marsupials) and Eu-theria or true mammals (the placentals). These terms themselves imply progress, and so are also unsatisfactory. It is probably best now to ignore their original meanings and accept the terms as neutral descriptors of the groups they have been assigned to: in this book I will refer to monotremes, marsupials and placentals.

Anatomy of the reproductive organs of marsupials

De Blainville was right to choose the anatomy of the reproductive organs as the central criterion of his classification because they are unequivocally distinct between the three groups of living mammals. To appreciate this we must consider the development of the kidney and the ducts that convey urine, and the genital ducts that convey gametes and embryos to and from the body.

If we compare a fetus from a tammar wallaby four days before birth and a human embryo at five weeks of gestation, for example, both have the same arrangement of kidney ducts, genital ducts and gonads (Fig. 1.1). Both the genital ducts and the kidney ducts (ureters) enter a common tube, the urogenital sinus, on its dorsal side and the future bladder is on the opposite, ventral side. In later development of both groups the ureters migrate to the ventral side to enter the bladder, while the genital ducts remain dorsal. This migration of the ureters only occurs in placentals and marsupials and probably arose as an adaptation for the more effective storage of urine in the bladder. In the monotremes, the ureters still discharge into the top of the sinus and the urine must pass across the sinus to enter the bladder, by a process still not understood.

Figure 1.1: The primary difference between marsupial and placental mammals. The relative position of the ureters and genital ducts and how it is derived from a common pattern of kidney and genital ducts at the indifferent stage in the embryo.

In the human, and all placental mammals, the ureters migrate outside and below the genital ducts, while in the tammar, as in all other marsupials, the ureters migrate inside and above the genital ducts. While the initial adaptation may have had more to do with excretion in the ancestral mammals, the route that the ureters take to reach the bladder unequivocally distinguishes living marsupials from living placentals. And it has had profound consequences for reproduction.

Its most obvious effect is seen in the female reproductive tract. Each oviduct transforms into the specialised regions of Fallopian tube, uterus, cervix and vagina (Fig. 1.1). In placental mammals the left and right oviducts join in the midline to form a single vagina and, in most species, also join to form a single cervix and a single uterus. The only portions that remain separate are the two Fallopian tubes, which each receive an egg from their respective ovary. In marsupials this joining of the oviducts cannot occur because the ureters pass between them, so there are two lateral vaginae, each arising from the common urogenital sinus posteriorly. Above the ureters the lateral vaginae loop back to the midline and become partially fused (Fig. 1.1). Then there are two cervices, two uteri and two separate Fallopian tubes. At copulation the semen is deposited in the lateral vaginae and the sperm pass through the two cervices and uteri to the Fallopian tubes, where fertilisation occurs. In the males of many species the glans of the penis is also divided and it is supposed that the semen is directed separately to each lateral vagina, but this has not been established. The same relationship of the ureters and the genital ducts in males means that the vas deferens in placental mammals loops over the ureter, whereas in marsupial males the arrangement is simpler (Fig. 1.1).

At birth the young marsupial passes through a new-formed canal in the tissues between the ureters, direct from the lateral vaginae to the urogenital sinus. In almost all marsupials this extraordinary arrangement, the pseudo-vaginal or birth canal, is re-formed at each birth: only in some kangaroos, Macropus, and the honey possum, Tarsipes rostratus, does it remain open after the first birth and in these species it is called the median vagina. So, female kangaroos have three vaginas, two for sperm and one for the young at birth. It has long been held that the small size of the divided uteri and the inadequacy of the birth canal are the reasons that marsupial young at birth are so very small. Whether or not this was the cause, the young of all marsupials at birth are much smaller than the most immature of placental young.

Figure 1.2: Relationship between maternal body weight and weight of the newborn marsupial and its stage of development in different species.

Small size at birth

Females of the largest living marsupials, the eastern grey kangaroo, Macropus giganteus, and the red kangaroo, Macropus rufus, weigh 28 kg and deliver one young that weighs 830 mg, half the size of a newborn mouse. The newborn kangaroo is 0.003% of its mother’s weight, compared to a newborn mouse or human baby that are about 5%: almost a two thousand-fold difference. Most marsupials at birth are between 200 mg and 400 mg (Fig. 1.2) but those of small dasyurid marsupials are under 10 mg and the very smallest newborn marsupial is that of the honey possum at 4 mg. It is little wonder that the first European explorers to see these tiny creatures were unable to believe that they were born in the normal manner or could travel unaided to the pouch and there find and attach to a teat. It seemed impossible that the young marsupials could be active and possess sensory organs or have sufficient neuromuscular coordination to move independently. But they do and, even more remarkably, they control the onset of their own birth like the much more advanced placental young do (see Chapter 2).

One reason often cited for the small size of marsupials at birth is that pregnancy is very short, compared to placental counterparts. Although it is true that in some species pregnancy is very short (less than two weeks), in other species it is longer than in a comparable sized placental species. The real point is not the duration of pregnancy but the mass of young produced at the end of pregnancy. All placental mammals bring forth very much larger young, at a more advanced stage of development, because the major growth phase of the young occurs during pregnancy, via a well-developed placenta. By contrast, in marsupials almost all growth and development occurs after birth during the long and complex period of lactation, when large changes in the volume and composition of the milk occur that support the changing needs of the young. Thus, the placental female makes her major investment in reproduction during pregnancy, the female marsupial makes hers during lactation. This difference has important consequences in the ecology of the species.

Control of sexual differentiation in marsupials

Superficially the external genitalia of marsupials and placentals are similar but the control of development to the adult form is different in the two groups of mammals. In placental mammals sexual differentiation of the fetus takes place during gestation when the external appearance of both sexes is the same, the so-called indifferent stage. Later under the influence of testosterone secreted by the developing testes of male fetuses the genital tubercle develops into a penis and the scrotum forms behind it; in the absence of testes the same structures develop in female fetuses into the clitoris and the outer lips of the vulva, respectively; nipples and mammary glands are formed in both sexes and retained through life. The whole cascade of change from the indifferent stage to the adult form is controlled by the expression of one gene on the tiny Y chromosome, called the ‘sex determining region’ of the Y chromosome, or SRY gene. Possession of this gene directs the gonads to differentiate into testes, which then secrete testosterone and transform the other genital structures to the male pattern: in the absence of the SRY gene the fetus becomes female. The normal complement of sex chromosomes in placentals and in marsupials is two X chromosomes in females and one X and one Y chromosome in males, but an individual with one X and no Y chromosome (XO) is female and an individual with two X and one Y chromosome (XXY) is male.

By contrast in marsupials the scrotum forms as two bulges in front of the genital tubercle, there is nothing equivalent to the outer lips of the vulva in female marsupials, and the pouch and mammary glands only differentiate in females. Male marsupials have a Y chromosome and the SRY gene, which directs the differentiation of the gonads to become testes and secrete testosterone leading to the differentiation of the internal genitalia and the genital tubercle to the male form: absence of the SRY gene results in the female form of the internal genitalia and genital tubercle. However, the developing testes do not control the differentiation of mammary glands, pouch or scrotum. Indeed, scrotal bulges develop in genetic males and mammary glands and pouch in genetic females many days before the gonads can be distinguished as ovary or testis, and these organs are not affected in their later development by sex hormones.

The present thinking is that these external organs in marsupials (mammary glands, pouch or scrotum) are controlled directly by the sex chromosome constitution of the tissues themselves, particularly the X chromosomes (Cooper 1993). Thus, possession of one X chromosome, as in a normal male marsupial, leads to differentiation of scrotal bulges; and possession of two X chromosomes, as in a normal female marsupial, leads to differentiation of mammary glands and a pouch (Renfree et al 1996a). In genetically abnormal tammars, XO individuals have female organs internally, as in placental species, but externally they do not have mammary glands or a pouch but do have a well-developed, empty scrotum. Conversely, XXY tammars have male organs internally and a well-developed penis, in accordance with their possession of a Y chromosome, but instead of a scrotum, they have a small pouch and mammary glands, in accordance with their possession of two X chromosomes (Sharman et al 1990).

Thus, the control of sexual differentiation in marsupials has followed a different path from that followed by placentals, although the end result is deceptively similar.

Physiological differences between marsupials and other mammals

Marsupials, like other mammals and birds, maintain their body temperature at a fairly constant level. However, the normal body temperature of marsupials is about 35.5°C, which is 2.5°C lower than that of most placentals, which in turn are lower than birds (Table 1.1). It is not clear why this should be so but it does appear to be something that is genetically determined and affects the lives of marsupials as profoundly as their mode of reproduction. To appreciate this we need to understand the underlying physiological process of which body temperature is an outward manifestation. Since the rate of chemical reactions doubles for every rise of 10°C, the rate in marsupials must be about 25% lower than in placentals, which in turn must be about 25% lower than passerine (song) birds. This is clearly seen in the cost of maintaining a constant body temperature, which rises with increasing basal body temperature (BBT) (Table 1.1).

Table 1.1: A comparison of the standard metabolic rate of terrestrial vertebrates

Standard metabolic rate (SMR; kJ/kg⁰.⁷⁵ per day); basal body temperature (BBT; °C) (after Dawson and Hulbert 1970).

To maintain their body temperature mammals and birds expend the least amount of energy when the surrounding temperature is nearly the same as their body temperature. When the surrounding, or ambient, temperature is lower more energy is required to generate heat and when it is higher than the body temperature more energy is expended in cooling devices, such as panting and sweating. The ambient temperature where minimum energy is used by the non-feeding mammal at rest is called its thermo-neutral zone. This minimum value is termed the standard metabolic rate (SMR) and it represents the energy required to maintain essential functions of the living body at a constant body temperature. It is usually determined as the oxygen (O2) consumed or carbon dioxide (CO2) produced under these conditions in a given period of time (Table 1.2).

Table 1.2: Resting body temperature (TBody), oxygen consumption and standard metabolic rate (SMR) of seven marsupials, arranged according to body mass

(Hume 1999).

AGowland (1973).

The SMR values can be converted to energy used (Joules) if the composition of the food being eaten is known. For carbohydrate one molecule of sugar and 6 molecules of O2 are converted into 6 molecules of water and 6 molecules of CO2. This is a respiratory quotient (RQ) of 1 and the water formed is called metabolic water. For fat the RQ is 0.7 and more metabolic water is formed, while for protein the RQ is 0.8 and less metabolic water is produced. If it is assumed that the animal’s food is a mixture of carbohydrate, fat and protein, with an average RQ of 0.8, then 1 mL of O2 consumed is equivalent to 21 Joules of energy and 1 mL of CO2 produced is equivalent to 26 Joules. The values for O2 consumed by a range of marsupials of increasing body size are given in Table 1.2. Although the body temperature of all the species is much the same, the smallest species consumed 12 times more O2 per gram of body tissue than the largest species. Why is the cost of living for the small species greater than for the large species?

Body size in relation to metabolic processes

For all animals there is an important relationship between body mass and standard metabolic rate, which is much more pronounced for birds and mammals than for other animals. In its simplest terms the mass of the body increases by the cube power, whereas the surface area increases by the square power, so the small species has a relatively larger surface area than the larger species. Since all metabolic functions occur at surfaces the smaller species has a relatively higher metabolic rate than the larger one. This affects all sorts of physiological functions. For instance, the heart rate of the smallest mammals are about 1000 beats per minute, compared to about 70 beats per minute for humans and fewer than 10 beats per minute for large whales. Again, the strength of muscle and bone depends on the cross-sectional area, so the strengths of these tissues increase by the square power also, so that a comparison between a small and a large mammal shows the small one to be proportionately much stronger than the larger one. For instance, a female antechinus, weighing 30 g, can carry a litter of young that weighs more than she does, whereas the female kangaroo ejects her single young from the pouch when it weighs about one-tenth of her own weight.

Conversely, because the small mammal has a proportionately greater surface area than the larger species, it loses heat and water across its skin and lungs more readily. This greater energy and water loss must be made up from the food and water ingested, so that less is available for synthesis into stored material. Hence, small animals cannot survive starvation for as long as large ones. For example, a small marsupial, such as a dunnart, Sminthopsis, consumes the equivalent of its body weight each day, whereas a person consumes about 1%, so a dunnart cannot survive more than a few days without food, whereas a person can survive for several weeks. For the same reason, one large mammal takes much longer to exhaust its food supply than an equivalent mass of many small mammals. Hence the argument: 10 rabbits eat as much as one sheep.

There is, thus, in body mass a nice balance of advantages and disadvantages. Under favourable conditions the small mammal converts food more rapidly and the population proliferates faster than the large species. In adverse times, however, the population of a small species will decline rapidly, as its members succumb to the lack of food or the adverse environment, whereas the members of a large species can withstand adversity for much longer. In the later chapters of this book the importance of body size in the economy of different marsupials will recur often.

Comparing mammals of different size

In order to compare the performances of animals of different diets, life styles and ancestry it is necessary to agree on mathematical functions that reduce the variability due to body mass. The formula that has generally been adopted is:

y = bxk

where y is the size-dependent variable (eg O2 consumption, heart rate or food consumption), x is body mass in kilograms, b is the intercept constant and k is the slope (Fig. 1.3).

For metabolic rate the exponent 0.75 is still the best approximation for k, despite considerable variability among species. Thus, to compare the SMR of a range of mammals of differing size, the formula used is kJ/kg⁰.⁷⁵ per day. Using this formula the average SMR of 56 marsupials from small 7 g dasyurids to large 29 kg kangaroos is 204 kJ/kg⁰.⁷⁵per day (Table 1.1). The SMR values for individual species vary from 140 for some desert dasyurids (eg Dasycercus cristicauda, 160 for the koala, Phascolarctos cinereus, to 310 for the tiny planigale, Planigale ingrami. This compares with the average SMR for a group of 272 placental mammals, 289 kJ/kg⁰.⁷⁵ per day, ranging from the house mouse to the elephant. Thus, the average value for marsupials is about 70% of the average value for placental mammals.

Figure 1.3: The relationship between body mass (on a log scale) and daily energy consumption for marsupials and placentals. Closed lines show standard metabolic rate and dashed lines show field metabolic rate. Note that the slope of the field metabolic rate for marsupials is not parallel to that for placentals, the smallest species having a much higher metabolic scope.

In Table 1.1 these values are compared to other terrestrial vertebrates, measured at their thermo-neutral zone and also at 38°C, the body temperature of placental mammals. What is evident from this is that the reptiles, which do not control their body temperature, have a much lower metabolic rate than the mammals and birds under both conditions. More interestingly, each group of mammals and birds has a characteristic level and each is positively correlated with body temperature. The monotremes have the lowest body temperature and the lowest metabolism, while the passerine birds have the highest. The high values of the latter are probably associated with the special requirements of flight.

During the past 30 years there has been much discussion about the significance of the apparent difference between marsupials and placentals, some people arguing that life style and diet may be more significant factors in determining SMR than ancestry (see especially Lee and Cockburn 1985, McNab 1986, 1988). Thus, some placentals, such as sloths, Bradypus, have a SMR below the marsupial average and some marsupials, such as planigale and the honey possum have a SMR of, respectively, 106 and 158% of the placental average. Nevertheless, the mean value for marsupials is 30% below the mean for placental mammals, which suggests that there is a basic underlying difference in SMR, although food habits and activity are sometimes strong enough to mask phylogeny.

It is not clear why the set point should vary between different kinds of animals, nor what controls it. Hulbert and Else (1999) have observed that the plasma membranes of the mitochondria and cells of vertebrates with high metabolic activity have a relatively high proportion of polyunsaturated lipids, whereas those with lower metabolic activity have membranes that are relatively mono-unsaturated. They have proposed that these differences may be the basis of the pacemaker for metabolism. This suggests that there may be a fundamental difference between the cell membranes of marsupials and placentals, which affects all metabolic processes, including respiration, excretion, water balance, nerve conduction and growth rate. The lower setting of the rate in marsupials is reflected in several other physiological functions: their heart rates, adjusted for body mass, are lower than for placentals, as are nitrogen requirements in some species. The lower SMR of marsupials also means that they have lower food requirements and water turnover rates, which may confer special advantages in adverse conditions or arid environments.

Conversely, the advantage for placentals of a higher body temperature and SMR is faster nerve conduction and smooth muscle contraction, faster growth rates and faster reproduction, with the trade off in higher food requirements. In those environments where soils are rich, the climate is equable and food predictable the higher metabolic and reproductive rate of placentals is advantageous. However, in less benign environments, where the soils are infertile and climate unpredictable, the lower metabolism of marsupials may confer an advantage. The spectacular radiations of marsupials in South America and Australia alongside placentals may in part have been due to environmental constraints in these continents that favoured species that conserved limited resources.

Field metabolic rates

A major development in the last 20 years has been the measurement of field metabolic rate (FMR) in 28 species of marsupial, using isotope dilution techniques (see Box 1.1). This is a much more informative measure of a species’ actual metabolic needs than SMR and the difference between SMR and a species’ maximum metabolic rate provides a measure of its metabolic scope. While most marsupials have a low SMR, compared to most placentals, the FMRs of species under 100 g body mass are much the same, which means that their metabolic scopes are greater (Fig.1.3). For instance, the fat-tailed dunnart, Sminthopsis crassicaudata, has a FMR up to seven times higher than its SMR (Nagy et al 1988). It is unusual for placental mammals to have FMR more than three times SMR, which means that the dunnart’s metabolic scope is considerably greater than equivalent placental species. This may be a definite advantage in extreme environments, since the low SMR conserves resources of the animal at rest, without limiting its metabolism when active.

Box 1.1: Measuring field metabolism

To compare the metabolic strategies that different species have evolved we need to understand the various components of metabolism. These are metabolic rate, water turnover, and the nutrients (eg nitrogen) required for an animal to maintain itself in its natural environment and reproduce. Each of these three components can be measured in captive animals in feeding trials, and by the use of respirometers. The use of water labelled with radioactive or other identifiable isotopes of hydrogen and oxygen give more realistic measures from free-living animals. Doubly labelled water (³H2O, H2 ¹⁸O), composed of tritium (³H2) or deuterium (²H2), and a non-radioactive isotope of oxygen (¹⁸O), can provide measures of respiration (CO2 production), water turnover, fat deposition and food consumption. If the sodium content of the food is known, food intake can also be estimated in the same way, using an isotope of sodium (²²Na).

The technique is to capture the animal, inject it with a known quantity of the particular isotope and then let the isotope equilibrate with the animal’s body tissues, a process that usually takes 2 to 6 hours. Then a blood sample is taken to measure the concentration of the isotope distributed in the body. This initial ratio of labelled to unlabelled isotope at equilibration gives a measure of total body content, or pool size, of that element:

Pool size = concentration of isotope injected × sample volume concentration of isotope in sample

The animal is then released and, after some days, when it is recaptured, a second sample of blood is taken and the concentration of the isotope measured again. The difference between the initial and subsequent concentrations is a measure of the dilution that has taken place in the elapsed time by respiration, and by the ingestion and excretion of water and food by the animal. These measures can be converted into field metabolic rate, water turnover rate and food consumption, respectively. For instance, oxygen turnover is measured by the dilution of the ¹⁸O isotope compared to the common isotope ¹⁶O, using a mass spectrometer.

Because each of these rates is substantially affected by the body mass, different species can be compared only if the values are expressed by an allometric exponent. For metabolic rate and food intake the exponent is usually taken to be the three-quarter power of body mass (kg⁰.⁷⁵), for water turnover it is kg⁰.⁸. However, both of these are approximations based on placental mammals and standard conditions. Green (1997) has shown that there are wide differences between species of marsupial and within a species at different times of the year, and that other exponents may more accurately reflect reality. However, we will follow convention here when comparing species in each Chapter.

In order to estimate the field metabolic rate of an animal we need to measure the volume of O2 consumed or CO2 produced, and to know the available energy content of the food. The mean energy content of the food varies according to the relative composition of carbohydrate, fat and protein. For example, it is 3–4 kJ/g for insects and 6 kJ/g for mammalian flesh. However, the net energy that can be used by the animal is less than this because about 10% is lost in faeces and 8% in urinary excretion. The balance is the net metabolically available energy of the food and, with this information and the SMR (calculated as kJ/kg⁰.⁷⁵ per day), the daily food requirements of the animal can be estimated. By comparing these several measures in different species of marsupial in later Chapters we can understand how each species is using the available resources to meet its basic requirements at different periods of its life cycle or in different environments.

These several attributes of marsupial metabolism will be discussed when considering the adaptations of particular species in later chapters.

Relationships within marsupials

If we accept that marsupials are a distinct group of mammals, what are the relationships within the group, and how have they evolved? For 100 years marsupials have been classified on the basis of the number and kinds of teeth and on the number of digits on the feet. These characters are undoubtedly affected by the mode of life of the possessors but have the advantage of being available in fossil remains and can be accurately dated and used to calibrate rates of change in other criteria, based on living animals. Most information comes from the structure of the molar teeth, whereas the structures of the simpler canines and incisors have only been used to determine total dentition. There may also be subtle differences in structure of the enamel and other features of these teeth that have not been disclosed so far. In classifying living marsupials the anatomy of the soft tissues can also be used to distinguish between related species or groups, as can the form and number of chromosomes.

In the past 30 years new techniques in biochemistry and molecular biology have greatly extended our understanding of the relationships of modern day marsupials: amino acid analyses and immunological techniques opened up new ways to compare relationships of living species by comparing proteins, and analyses of base sequences in nuclear and mitochondrial DNA has provided an even more powerful means to determine relationships of living marsupials, and to provide increasingly precise measures of the time since related groups diverged in the past.

We will begin by considering anatomical characters and then see how relationships based on the newer techniques corroborate or refute the classification of marsupials based on anatomy.

Relationships based on anatomy

Teeth

Teeth develop in the jaws of young mammals as a coalescence of cells derived from the base of a deep groove of surface epithelium and the underlying tissues in the jaw. The initial tooth bud develops as a central core and an overlying cap of epithelium in which dentine and enamel is later laid down. The tooth bud rises to the surface of the jaw as it develops, eventually piercing the groove and erupting; other tooth buds follow successively behind it in the jaw. In the front teeth the structure remains simple with a single cusp and a single root, but the back teeth develop more cusps and roots. These secondary cusps develop, like the primary one, as subsidiary buds and their position relative to the primary cusp is characteristic of the particular tooth and for the particular species.

This pattern of development has been understood for a long time but in the last 10 years the way in which the pattern is controlled genetically has become clearer through the expression of developmental genes in the tooth bud (for review see Thesleff and Sharpe 1997). One group of genes determine that the tooth bud will develop into a complex molar instead of a simple front tooth, while other genes determine the position of the secondary buds relative to the primary bud. What is interesting is that the expression of the genes precedes by one or two days the first morphological changes in the cells of the jaw, indicating that the information about pattern is set up before the buds begin to form and not as a result of interaction between adjacent tooth buds in the jaw. Thus, gene expression predicts future cusp pattern and affects the very early stages of tooth development. In evolutionary terms small changes in gene expression could have profound effects on tooth morphology.

As an example of what may become possible, Jernvall et al (2000) compared the expression of specific genes in the formation of molar teeth in mouse, Mus musculus, and vole, Microtis rossiameridionalis, which evolved from a common ancestor in the early Miocene epoch 20 million years ago. They showed that the big differences in the molar pattern between mice and voles is the result of small changes in the expression of the genes that determine the relative position of the secondary cusp, whether it is parallel to the primary cusp (in mice) or diagonal to the primary cusp (in voles).

What is the relevance of this to marsupials? Marsupial teeth develop in the same manner as the teeth of placental mammals, reflecting their common ancestry more than 120 million years ago. It is, therefore, reasonable to assume that the same or similar genes control the pattern of molar cusps in marsupials. More than this, the discovery that the cusp pattern of molar teeth is controlled by the expression of particular genes brings closer the day when the evidence of palaeontology, based largely on the morphology of molar teeth, and the evidence of relationships based on molecular genetics, can be integrated more precisely.

Teeth in mammals have two primary functions: the front teeth are used to bring food into the mouth, while the back teeth are used to process it for digestion, either by cutting or grinding the food into fine pieces. Teeth may also be used in aggressive or defensive displays, for grooming, or for grasping the young. The front teeth are single rooted and comprise a variable number of chisel-shaped incisors and a single set of pointed canines. Behind the canines are up to four larger teeth, each with two roots, called premolars and behind the premolars are up to four sets of molars, each with three or four roots. The structure of teeth reflects the uses to which they are put in different species and they can therefore tell us a considerable amount about the life of the possessor. This is especially useful when examining fossil specimens, which often consist only of a few teeth, because these are the most durable part of the body. The size and shape of the front teeth can tell us whether the mammal catches moving prey and what sort of prey, or whether its diet is largely of plant material. Carnivorous species have many sharp incisors, prominent canines and many sharp points on the premolars and molars. Their molars intersect in such a way as to provide many shear surfaces, like scissor blades, and less emphasis on flat, opposed surfaces where food can be crushed or ground small. In herbivorous species, by contrast, the canines are small or absent, the incisors form two opposing rows of chisel-like teeth with which herbage can be cut, the molars and premolars have fewer sharp points and shear surfaces but have a much larger area for grinding.

Sanson (1985) has suggested that the different types of dentition reflect the forces required to penetrate the bodies of prey species. For instance, the impact strength of bone is about 2 kg/cm² while that of insect cuticle is 23 kg/cm², so puncturing the cuticle with sharp points is an easier option than crushing it. The impact strength of plant cell walls, however, is about 76 kg/cm², which explains why the teeth of herbivores have such highly developed crushing surfaces. It also accounts for the different anatomy of the jaws of carnivores and herbivores: carnivores have the jaw muscles grouped at the back of the jaw, which allows for a wide gape but delivers less force than the forward disposition of the jaw muscles of herbivores, such as kangaroos (see Chapter 9).

Since the earliest mammals and all the ancestral marsupials and placentals were small insectivores, the dentition of small carnivores resembles more closely the ancestral dentition of mammals from which the several types of herbivorous dentition have evolved. In the study of teeth of mammals the molars provide the greatest amount of information about diet, about relationships between species, and about the evolution of the main lines of descent from the Cretaceous period ancestors. Because of their importance in the subsequent discussion on the evolution of marsupials, some knowledge of the anatomy of molars is necessary.

Figure 1.4: Probable evolution of molar teeth in marsupials and placentals from pantothere ancestors 250 million years ago. The lower left molar is shown in occlusal view, stippled, and the upper molar is superimposed in outline. In the upper molar the primary cusps are the (pr) protocone, (pa) paracone and (me) metacone, to which were later added posteriorly the (mcl) metaconule and laterally the (st) stylar shelf. In the lower molars the primary cusps are the (prd) protoconid, (pad) paraconid and (med) metaconid, to which were later added posteriorly the (end) entoconid and (hyd) hypoconid. Data from Romer (1966) and Archer (1976).

The structure of molar teeth can be traced from the earliest mammals, which had triangular, three-cusped teeth in the upper and lower jaws arranged so that, seen in surface view, the apex of each upper tooth (protocone) pointed inwards while the apex of each lower molar (protoconid) pointed outwards (Fig. 1.4). When the mouth closed the upper and lower molars fitted closely between each other providing a zigzag shear surface. This can still be seen in the skulls of small carnivorous marsupials (see Fig. 4.2) and placentals. At an early stage of mammalian evolution, before the separation of marsupials and placentals, the simple arrangement of three cusps was extended. In the lower jaw two additional cusps developed on the posterior face of the molars, the entoconid and hypoconid, thereby providing a basin into which the apex, or protocone, of the upper molar fitted. This provides a grinding surface in addition to the shearing component of the molar teeth. In the upper molars additional cusps developed, both between the primary three cusps and also on the outside of them. The outer series of five small cusps, called the stylar shelf, do not meet complementary parts of the lower molars so that their function in mastication is unclear. Although the earliest placentals had two stylar cusps, later placentals do not, whereas all early marsupials had the full complement of five cusps, most of which are retained in the present day American opossums (Fig. 1.4) and dasyurids (see Fig. 4.2). The presence of the stylar shelf is a diagnostic tool in differentiating between fossil placentals and fossil marsupials.

Ridges may form between these several cusps and the pattern of the ridges and the relative sizes of the cusps are used to determine phylogenetic relationships between extinct and living mammals. In addition the fine surface structure of the teeth and surface scratch marks can provide further information about relationships and the type of food that was processed by the living animal.

All the living American marsupials and the Australian carnivorous species have long snouts bearing a battery of simple, sharp-pointed teeth. In each jaw there are four molars, three premolars and one prominent canine, as well as four or five incisors in the upper jaw and three in the lower jaw (Fig. 1.5). On this criterion they are grouped together as the Polyprotodontia (meaning many front teeth). The bandicoots are included in this group on this criterion but they are more omnivorous than the dasyurids and their molar teeth are squared up with an extra cusp at the back, called the metaconule (Fig. 1.4 and 5.2).

Figure 1.5: Marsupial relationships based on teeth and feet. Upper panel to show representatives of didactyl and syndactyl feet, the latter differentiated by the small, paired digits 2 and 3 on the pes; also note the large digit 4 in kangaroos; lower panel shows representative skulls of polyprotodont and diprotodont species, differentiated by the presence of 4 or 5 incisor teeth (i) in each jaw in front of the canines (c) in all polyprotodonts and only one in each lower jaw of diprotodonts; all have four molars (m) and up to three premolars (pm). After Jones (1924) and Tyndale-Biscoe (1973).

The herbivorous species of Australasia have fewer premolars, the canines are small or absent and there are only one to three incisors in the upper jaw and a single large pair of incisors in the lower jaw. With fewer teeth, there is often a gap between the front teeth and the cheek teeth, called the diastema, which enables the herbage to be presented by the tongue to the battery of grinding molars.

The single pair of incisors in each lower jaw is the defining character for this diverse group, which is called the Diprotodontia (meaning two front teeth). Again, there is a superficial link to one group of South American marsupials, the Caenolestidae, in which the first incisors are large and procumbent, but in them the other incisors are present. Furthermore, in the embryologi-cal development of the teeth of the Diprotodontia, the large incisor is the second, not the first, incisor so this is another case of convergence rather than indicating a close relationship.

Within the Diprotodontia further distinctions can be made on the basis of teeth. As in bandicoots, they have the metaconule in the upper molars, but have lost the stylar shelf, and their lower molars are also squared up by the loss of the paraconid at the front (Fig. 1.4). Also the grinding surface of the molars is more rounded than the sharp pointed cusps of the Polyprotodontia and are variously shaped into crescents or ridges (see Fig. 7.7). This development of a grinding battery of molars reaches its greatest development in two groups of grass eaters: in the wombats the incisors in the upper jaw are reduced to one pair and all the teeth grow continuously through life in a similar manner to the teeth of rodents and rabbits; in the large kangaroos the molar teeth are high crowned, like those of sheep, and the whole battery moves forward in the jaw, the molars being used successively as the more anterior ones are worn down and shed (see Fig. 9.3). Elephants do this too, but on a grander scale.

Foot structure

When foot structure is used as the criterion for grouping marsupials, all the American marsupials and the dasyurid marsupials of Australia share a common feature of hind feet with five separate, subequal digits (Fig. 1.5). This is termed didactyly, or separated digits. The remaining Australian marsupials have digit 1 of the hind foot reduced to a nubbin and digits 2 and 3 partly fused and together equal in size to digit 5, and this is termed syndactyly, or fused digits. In some species, especially the large kangaroos, digit 4 is much larger and longer than the other digits (Fig. 1.5) and takes the main thrust during jumping. The semi-fused digits 2 and 3 are used in grooming by some species (see Fig. 9.2e), but whether they evolved primarily for this function or represent a progressive reduction for speedier locomotion, as in the evolution of horses and ruminants, is not clear. A long-standing paradox in understanding marsupial relationships is that bandicoots (Peramelomorphia) have a dentition like the American and Australian carnivorous marsupials but have a foot structure that is apparently identical with the Diprotodontia. This paradox is slowly being resolved in favour of bandicoots and the Diprotodontia having acquired fused toes independently. Although this is a very remarkable convergence, it is under a simple genetic control and so could have arisen more than once.

Ankle bones

Szalay (1982) proposed that the anatomy of the ankle bones could differentiate between American and Australasian marsupials: in all the Australian species that he examined there was a single articular facet where the ends of the long bones attach to the bones of the ankle joint, whereas in the two main families of American marsupials he found two facets. The only American species that had undivided facets like the Australian species was a species that lives in Chile, the monito del monte, Dromiciops gliroides; this alerted biologists to look for closer links between American and Australian marsupials, which has gained strong support from the molecular studies (see Marsupial relationships based on protein analysis). While subsequent work has confirmed a closer relationship of Dromiciops with Australasian marsupials than with American species, the criterion itself has not proved to be as consistent as Szalay thought. Hershkovitz (1992) examined a much larger series of species than Szalay and found that both types of ankle joint occur among the American marsupials and among the Australasian species, so it is not an exclusive character.

Brain anatomy

Links between the two halves of the forebrain provide another way to distinguish relationships between seemingly similar marsupial groups. Two nerve tracts, or commissures, link the cerebral hemispheres of the forebrain: the large anterior commissure, which also links the two olfactory lobes of the forebrain, and the smaller hippocampal commissure (Johnson 1977). In placen-tals a third commissure, the corpus callosum, links the cerebral cortex of each side, but it is absent from monotremes and marsupials. Within the marsupials a clear distinction can be made between the Diprotodontia and the Polyprotodontia: the former group has an additional tract of fibres, called the fasciculus aberrans, that extends the links of the anterior commissure between the two sides of the cerebral