Indeed, the evolutionary history of the mammalian way of providing nutrients for young is difficult to ascertain on the basis of the usual techniques: Fossils and comparative anatomy. The soft parts involved don’t fossilize well, and there are not enough “intermediates” living today to develop a plausible story of the evolutionary transitions linking egg-laying to live birth and lactation.
A new study recently published in PLoS Biology brings us a long way towards understanding this set of evolutionary events.
PLoS provides an “Author Summary” for published papers, which are sometimes, sadly, only slightly rewritten technical abstracts. But the Author Summary for this paper is nice and clear:
Egg yolk contains the nutrients required for the development of the nonmammalian vertebrate embryo. These nutrients derive by and large from a single set of proteins, vitellogenins, which are produced in the liver and provide or transport amino acids, lipids, phosphorous, and calcium to the egg. Mammals have evolved new nutritional resources for their developing and early offspring, such as lactation and placentation. However, the evolutionary timing and molecular events associated with this major phenotypic transition are not well understood. In this study, we have investigated the evolutionary fate of the three ancestral vitellogenin-encoding genes in mammals. Using detailed evolutionary analyses of genomes from the three major mammalian lineages (eutherian “placental” mammals, marsupials, and monotremes), we found that these genes progressively lost their functions and became pseudogenes relatively recently during mammalian evolution (the most recent inactivation event occurred roughly 30-70 million years ago). Monotremes, which lactate yet lay small parchment-shelled eggs, even retained a functional vitellogenin gene, consistent with their intermediate reproductive state. Our analyses also provide evidence that the major milk resource genes, caseins, which have similar functional properties as vitellogenins, appeared in the common mammalian ancestor ∼200-310 million years ago. Based on our data, we suggest that the emergence of the alternative resources for the mammalian young–lactation and then placentation–only gradually reduced the need for egg yolk resources (and hence functional vitellogenin genes) in mammals.
200–310 million years ago is a LONG time ago. But that is in accord with the slit between living mammals and their nearest non-mammalian relatives. These early “mammal like reptiles” known as the Therapsids, arose between about 250 and 300 million years ago. This does not mean that Therapsids had mammary glands, but it might mean that the earliest glandular transfer of nutrients to young could have been found among these early critters.
Th timing of the loss of egginess would be, according to this study, much later, perhaps prior to 70 million years ago. By 70 million years ago, mammals had already diversified, so this finding conforms to expectations from other genetic studies and the fossil record. (see this)
Regarding the timing, the authors suggest:
… while placentation, … appears to have allowed for the complete loss of egg yolk resources … the emergence of nutritive lactation may have reduced the selective pressure on yolk-dependent nourishment ever since early in the last common mammalian ancestor. Thus, the initial driving force for the reduction in nutrient content of the mammalian egg was probably lactation, the key feature of mammals.
Brawand, D., Wahli, W., Kaessmann, H. (2008). Loss of Egg Yolk Genes in Mammals and the Origin of Lactation and Placentation. PLoS Biology, 6(3), e63. DOI: 10.1371/journal.pbio.0060063