Tweaking the mammal family tree with ancient DNA and fossil proteins

By Peter D. Heintzman

Mammals are those hairy, milk-producing, warm-blooded animals that can be found in nearly all the major ecosystems on Earth. They comprise around 5,400 living species and range in size from tiny monkeys and bats, little more than an inch long, to enormous whales, some of which are the largest animals to have ever lived.

Mammals can be split into three groups, which differ by the way their young develop. Young of the egg-laying mammals, which include the platypus and echidna, hatch from eggs that have been laid by their mother (figure 1). On the other hand, the marsupial mammals, such as kangaroos, koalas, and wombats, give birth to live but underdeveloped young, who continue their development in their mother’s pouch. The final group, to which humans belong, is the placental mammals, which develop their offspring internally and give birth to fully developed young. In this post, I will discuss what we know about the family tree of living and some recently extinct placental mammals.

Figure 1. Example species from the three groups of mammals: the egg-laying platypus, the marsupial kangaroo, and the placental cheetah. Image credits: platypus – Klaus (CC BY-SA 2.0), kangaroo – PanBK (GFDL/CC BY-SA 3.0), cheetah – schani (CC BY-SA 2.0); all via Wikimedia Commons or Flickr.
Figure 1. Example species from the three groups of mammals: the egg-laying platypus, the marsupial kangaroo, and the placental cheetah. Image credits: platypus – Klaus (CC BY-SA 2.0), kangaroo – PanBK (GFDL/CC BY-SA 3.0), cheetah – schani (CC BY-SA 2.0); all via Wikimedia Commons or Flickr.

How do we reconstruct family trees?

To reconstruct a family tree, we need to know how different animals are related to one another. We assess the relatedness of different species by comparing shared characteristics – more closely related species, which share a more recent evolutionary common ancestor, should be more similar to one another than a more distantly related species. For example, in contrast to elephants, mice and rats are small, and have fast metabolisms and continuously growing front teeth. We can then be confident that a mouse and a rat share a more recent common ancestor than either does with an elephant.

There are two commonly-used types of evidence for understanding how different animals are related to one another: anatomical features, such as the constantly growing teeth of rats and mice, and DNA and protein sequence information – species with similar sequences are likely to be more closely related to one another than to a species with a more distinct sequence. In practice, hundreds of anatomical features and thousands to millions of sequence characters are used to reconstruct family trees – more information tends to provide more accurate predictions of relationships.

In the case of the placental mammal family tree, neither anatomical features nor sequence data are foolproof sources of information. They give conflicting answers, both with some areas of uncertainty (figure 2). This is because some anatomical features and sequence characters may be shared between two very different groups through convergent evolution, in which a feature or character evolves independently, rather than being inherited through common ancestry. Fins, for example, are anatomical features that have evolved independently as an adaptation to swimming in seals, whales and dolphins, and sea cows, which belong to very different, unrelated mammal groups. Although not perfect, sequence characters are a preferable source of information as there are typically far more of them available than anatomical features for reconstructing family trees.

Figure 2. The family tree of placental mammals based on anatomical features (left), and DNA and protein sequence information (right). Major groups are color-coded based on DNA and protein sequence information. Dashed lines indicate uncertain branches. Trees are based on those in Springer et al. (2004), dos Reis et al. (2012), and Welker et al. (2015).
Figure 2. The family tree of placental mammals based on anatomical features (left), and DNA and protein sequence information (right). Major groups are color-coded based on DNA and protein sequence information. Dashed lines indicate uncertain branches. Trees are based on those in Springer et al. (2004), dos Reis et al. (2012), and Welker et al. (2015).

Fossils generally do not contain DNA or proteins, as these have been broken down and lost in the millions of years since the fossil was a living animal. We therefore tend to rely on anatomical features to insert fossils into family trees, which can cause difficulty if few anatomical features are known. For some fossils, however, we can retrieve ancient DNA (if younger than a million years) and fossil protein sequence information (if less than five million years old). The likelihood of retrieving this information is increased if the fossils are preserved in cold, dry, and dark conditions, which reduce the rate at which DNA and protein molecules are broken down.

 What can we learn from ancient DNA and fossil protein sequence information?

Using ancient DNA sequence information, we have learned that one of the last camels to live in North America, the extinct Camelops, was more closely related to living Asian and African camels than to South American llamas and alpacas (figure 3). In contrast, anatomical features were suggestive of Camelops being related to alpacas and llamas. We have also discovered that the extinct giant deer Megaloceros was more closely related to living fallow deer, rather than red deer as suggested by anatomical features.

Perhaps the most impressive insight was the placement of two ‘South American Native Ungulates’, Macrauchenia and Toxodon. These weird, extinct mammals have a very confusing mix of anatomical features not seen in anything alive today. This had resulted in conflicting ideas about their placement in the mammal tree, such as being near elephants, or near horses, rhinos, and tapirs, or near pigs, camels, cows, and deer. Using fossil protein sequence data, it was clearly shown that Toxodon and Macrauchenia were actually most closely related to the group comprising living horses, tapirs, and rhinos.

Fig3_blog_post_v3
Figure 3. Recently extinct mammal species (top row) whose placement in the mammal family tree using anatomical features was either overturned or clarified using ancient DNA or protein sequence information. Their closest living relatives are shown in the bottom row. Image credits: Western camel – Sergiodlarosa (CC BY 3.0), Giant deer – Charles Robert Knight (Public domain), Macrauchenia – Olllga/Kobrina Olga (CC BY 3.0), dromedary camel – Jjron (GFDL/CC BY-SA 3.0), Bactrian camel – Bernard Gagnon (CC BY-SA 3.0), fallow deer – NotFromUtrecht (CC BY-SA 3.0), horse – Chinneeb (CC BY-SA 3.0), tapir – Jim Winstead (CC BY 2.0), rhino – Alanb (CC BY 2.0); all via Wikimedia Commons or Flickr.

Through these small tweaks and improvements that ancient DNA and fossil protein sequence information is providing to the mammal family tree, we are gaining a more complete understanding of the diversity and evolution of this important and successful group.

Higher level classification of living mammals:

Mammalia: all mammals

Monotremata: egg-laying mammals

Marsupialia: marsupials

Placentalia: placental mammals

Afrotheria: red group in figure 2

Xenarthra: yellow group in figure 2

Euarchontoglires: green group in figure 2

Laurasiatheria: blue group in figure 2

Suggestions for further reading:

dos Reis et al. (2012) Phylogenomic datasets provide both precision and accuracy in estimating the timescale of placental mammal phylogeny. Proceedings of the Royal Society, Series B 279, 3491–3500.

Heintzman (2013) Patterns in palaeontology: an introduction to ancient DNA. Palaeontology [Online] 3, 1-10.

Heintzman et al. (2015) Genomic data from extinct North American Camelops revise camel evolutionary history. Molecular Biology and Evolution 32 (9), 2433–2440.

Lister et al. (2005) The phylogenetic position of the ‘giant deer’ Megaloceros giganteus. Nature 438, 850–853.

Llamas et al. (2015) Late Pleistocene Australian Marsupial DNA Clarifies the Affinities of Extinct Megafaunal Kangaroos and Wallabies. Molecular Biology and Evolution 32 (3), 574–584.

Orlando et al. (2013) Recalibrating Equus evolution using the genome sequence of an early Middle Pleistocene horse. Nature 499, 74–78.

Rybczynski et al. (2013) Mid-Pliocene warm-period deposits in the High Arctic yield insight into camel evolution. Nature Communications 4, 1550.

Springer et al. (2004) Molecules consolidate the placental mammal tree. Trends in Ecology and Evolution 19 (8), 430–438.

Welker et al. (2015) Ancient proteins resolve the evolutionary history of Darwin’s South American ungulates. Nature 522, 81–84.

Advertisements