Spotlight on Zoo Science
July 14, 2003

Lactating on Eggs

National Zoo scientist Olav Oftedal offers a new scenario to explain how lactation—a defining trait of mammals—evolved.

gorilla nursing baby
Feeding young milk from mammary glands is a defining feature of mammals.

All female mammals secrete milk from mammary glands. Milk nourishes their babies until they are able obtain the nutrients they need from other food. In fact, this characteristic or process, called lactation, along with the presence of hair, defines a mammal. If the female of a species lactates, the animal is a mammal; if she doesn’t lactate, it’s not.

Despite lactation’s importance to our understanding of the biological class to which we belong, how lactation evolved has long puzzled scientists. Darwin’s theory of evolution by natural selection was even challenged by it. Because no other animals were known to do anything like this, it seemed difficult to imagine lactation evolving by “numerous, successive, slight modifications,” as Darwin proposed that evolution proceeded. One of Darwin’s critics asked. “Is it conceivable that the young of any animal was ever saved . . . by accidentally sucking a drop of scarcely nutritious fluid from an accidentally hypertrophied [skin] gland of its mother?”

While Darwin devised a plausible scenario in response, new information later proved it false. Subsequently, biologists have proposed at least ten hypotheses, but none of them are entirely satisfactory, according to a recent review by National Zoo scientist Olav Oftedal. In two papers published in 2002 in the Journal of Mammary Gland Biology and Neoplasia (references), Oftedal offers a new scenario to explain, in his words, “How could such an intricate process, involving radical innovation in both mother and suckling young, come into being?”

It is no easy matter to trace the evolution of a structure like the mammary gland and a process like lactation because it does not fossilize. Through painstaking examination of fossils, paleontologists have been able to “see” progressively more mammal-like forms evolve by following gradual changes in teeth and bones.

No such luck with mammary glands and lactation. Instead, to develop his scenario, Oftedal had to review mammalian evolutionary history as revealed by changes in bones and teeth, as well as amass evidence on the nature and evolution of skin glands, the difference between the eggs of birds and reptiles and those of mammals (yes, a few mammals lay eggs), ecological changes over evolutionary time, embryonic development of mammary glands, the functions of scales, skin, and hair, and more. In addition, he looked for clues in parental care exhibited by various vertebrates.

While lactation occurs only in mammals today, it may actually have appeared in ancestral forms well before mammals came on the scene. And its original function may not have been feeding young. Instead, Oftedal believes, it initially served to keep eggs moist during incubation.

The ancestors of animals that would eventually become mammals, called synapsids, first appeared about 310 million years ago, when they branched off from the earliest terrestrial vertebrates, called amniotes. (See diagram below.) Other amniotes were the ancestors of various other forms that would eventually become reptiles and birds (as well as dinosaurs and other now-extinct groups).

chart showing radiations from the Amniotes
Amnioties radiated into Synapida and Sauropsida. Sauropsida radiated into several taxa, some with living forms including turtles and birds and some now extinct. (Horizontal lines show when the taxa appeared and how long it lasted; unlabeled lines represent extinct taxa.) Synapsida radiated into Therapsida and several other now-extinct taxa. Therapsida radiated into Cynodontia and others. Cynodontia radiated into Mammaliformes and others; Mammaliformes into Mammalia and others, and, finally Mammalia radiated into Montremes, Marsupial, Eutheria (true mammals), and one now-extinct taxa.

What separated the amniotes from their ancestors was eggs with outer membranes and layers that permitted them to survive outside of aquatic—or at least very wet—environments. Surrounded by thin, parchment-like shells, however, these eggs were subject to drying out in the air like a wet sponge does.

The hard shell of birds' eggs keep  the contents from drying out.
The hard shells of birds' eggs keep the contents from drying out.

The ancestor of birds and some other groups solved this problem with the evolution of a hard, calcified egg shell—eventually the typical chicken egg. The synapsids appeared to have solved it by keeping the eggs in contact with skin kept moist by secretions from skin glands. Glandular skin was present in the ancestors of the amniotes and proved handy to meet this new need. And, it turns out that mammary glands are specialized skin glands, most similar to apocrine glands (other gland types are called sebaceous and eccrine), and both likely evolved from the same ancient gland type. Some apocrine glands make and secrete lipids and other complex organic molecules, and mammary glands also secrete lipids and the other complex organic components of milk.

How, though, did animals get from secreting a glandular substance to keep eggs moist to feeding their young the secretions of mammary glands?

The behavior of modern amphibians provides a clue to how this might have occurred. Some amphibians that nest in dry habitats carefully tend their eggs and may keep them moist through glandular secretions; glandular secretions may also provide protection from fungus and bacteria. Further, there are hints that the hatchlings of some caecilian species (legless amphibians) feed on their mother’s skin or skin gland secretions. As Oftedal says, “If verified, this remarkable discovery might provide a direct analogy for a transition from skin secretions as egg supplements to skin secretions as hatchling food.”

Platypuses are one of only three species of egg-laying mammals.
The platypus is one of only three species of egg-laying mammals.

Closer to home than amphibians, however, are the monotremes—one species of platypus and two species of echidnas that are the only egg-laying mammals. And, of course, by definition, female monotremes lactate. The ancestors of the living monotremes split off very early on from those of the rest of the mammals, indicating that lactation had evolved before that split, which occurred about 150 million years ago. The modern monotremes are certainly not identical to the first ones, having evolved their own unique specializations. But they may reveal a link between the waterers and feeders.

Female platypus and echidnas lay eggs, which they incubate for about 12 days. When the young hatch, they feed on milk produced by the mother’s mammary glands. In a significant departure from all other mammals, however, the mammary glands do not open into a nipple from which young suck milk. Instead, the glands open in a hairy patch and the young suck the milk from the hair and skin.

Oftedal suggests, however, that the glands may also be secreting water and other substances onto the eggs during the incubation stage. This is not known with certainty, but there are suggestive observations.

During incubation, monotreme eggs are covered with a moist, sticky substance of unknown origin. Moreover, before they are laid, the eggs have three eggshell layers; afterwards, a fourth layer appears that is quite different from the earlier layers, and it seems to be applied to the egg surface in a fluid form. Thus, it is entirely possible that both the sticky substance and the fourth eggshell layer are secretions of the mammary glands.

This dual mammary gland function of coating the eggs and feeding young may also help explain why the monotremes lack nipples: secretions from a largish, flat patch might more easily coat an egg than secretions from the single point of a smallish nipple.

Hair and Milk

Oftedal’s scenario also suggests a novel original function of hair. Many apocrine and sebaceous skin glands are associated with hair, and these glands and hair follicles develop in tandem, the entire complex being called the “apo-pilo-sebaceous unit” (pilo refers to hair). The glandular secretions, including some related to reducing water loss, thus end up coating the hairs.

Echidna's mammary glands open in hairy patch.
Echidnas' mammary glands open in a hairy patch.

As noted above, monotreme mammary glands are associated with hairs and marsupial mammary glands start out hairy but later the hair is shed from the nipple. The mammary glands of the rest of the mammals are not associated with hairs but, in early development, hair formation appears to be actively inhibited around the mammary area. In both of these later groups, hair on the nipples would interfere with suckling, so there would be selection for the hair to be lost before feeding young became an issue or for the hair never to grow at all.

Oftedal proposes that hairs may first have been associated with glands because the hairs wicked and spread secretions over eggs, while also reducing air circulation around the eggs. The egg-laying monotremes thus still have hair associated with the mammary patches to do this job, but once other mammals gave birth to live young, neither the hairs nor the mammary patch were required for egg tending. Suckling from non-hairy nipples became more efficient.

This all leads to the conclusion that hair may first have evolved in the ancestors of mammals to aid in egg-tending, and that hair first appeared at the glandular patches on animals’ undersides in contact with eggs. Only later did hair become associated with other skin glands as the ability to grow hair was borrowed to insulate the rest of the body surface.

If Oftedal is correct, it means that mammary glands and hair—those defining traits of mammals—first evolved well before mammals did, and served entirely different functions than they do today. It also reveals a key insight into how evolution proceeds. Biologists note that evolution does not engineer new features out of whole cloth. “Instead, wrote one scientist, “it cobbles together new features by tinkering with existing ones in a way that would have made Rube Goldberg proud.”

Wet and Warm

Another surprising insight that emerges from Oftedal’s scenario is the role that moisture-producing mammary glands may have played in the evolution of endothermy. Mammals (and birds) are endotherms, which means they generate heat internally, through a high metabolic rate, and usually maintain a fairly high and constant body temperature regardless of ambient temperature (within reasonable limits); this is often called being “warm-blooded.”

(Reptiles, in contrast, are ectotherms. They obtain heat from outside sources, like the sun, and regulate their temperature through behaviors such as basking or seeking shade. Their body temperature is roughly the same as the ambient temperature.)

Scientists believe the first mammals resembled today's shrews.
Scientists believe the first mammals resembled today's shrews, which are active predators of insects.

Endothermy is considered to be a pre-condition for the kind of sustained activity that is advantageous for hunting prey and escaping predators. Endothermy and related features such as enhanced respiratory function, rapid growth, improved locomotion, and feeding specializations gradually evolved in the synapsid line that eventually led to mammals, and first achieved mammal-like parameters well before true mammals appeared. The upshot of all of this is that at some point synapsids began incubating eggs at above-ambient temperatures with their internally generated heat.

Endothermic incubation offers several advantages, including rapid growth and development and shorter incubation times. But it would also pose additional problems for parchment-shelled eggs already subject to drying out because higher temperatures increase the rate at which water is lost from eggs, just as a wet sponge dries more quickly in the dryer than it does on the kitchen sink. Moreover, endothermic incubation means that eggs have to be kept off the ground because the warmth imparted to the eggs by the parent’s body would be offset by the heat lost to the soil.

One solution to this problem is incubating parchment-shelled eggs in a warm, moist pouch, as echidnas do. (The ancestors of birds solved it with hard-shelled eggs.) But carrying fragile eggs in a pouch would have been a real drag on an active predatory lifestyle, and would limit both egg size and clutch size to what a mother could comfortably carry. Instead, Oftedal believes, lactation enabled mothers to park their eggs in a nest while they foraged and then replenish the eggs’ moisture when they returned. In fact, this would have been an advantage even for ectotherms and likely preceded, indeed made possible, the evolution of endothermy in the synapsid line.

—Susan Lumpkin


Oftedal, O. T. 2002. The mammary gland and its origin during synapsid evolution. Journal of Mammary Gland Biology and Neoplasia, 7(3):225-252.

Oftedal, O. T. 2002. The origin of lactation as a water source for parchment-shelled eggs. Journal of Mammary Gland Biology and Neoplasia, 7(3):253-266.

Note to Media: If you would like more information about this project, or any of the Zoo's conservation and science programs, please contact the Zoo's Office of Public Affairs.

Page Controls