Yin and Yang in the Bat World

"Yin", if you're wondering...

Although the term is used more casually in everyday English, in scientific terminology a "family" of animals is a specific rung on the taxonomic hierarchy. Cats (Felidae) are a family, but otters (Lutrinae) are just a subfamily within the larger weasel family. In turn, families are grouped into "orders", and the great diversity of mammals is indicated by the fact that the standard modern scheme recognises no less than 27 living orders of mammal.

In terms of the number of known species, the second-largest order of mammals, after the rodents, is that of the bats. There are over 1,400 named species of bat, currently grouped into at least 18 different families. Yet they're probably also the least studied of the major mammalian orders - although, to be fair, this may depend on your definition of 'major'. Compared with primates, or hoofed or large carnivorous mammals, we know relatively little about their evolution, and how the various sub-types of bat relate to one another.

Perhaps one of the first questions on this topic that we might want to ask is whether or not bats really exist as a group at all. The alternative possibility is that flight independently evolved twice among mammals, so that there was never some "first bat" from which all the others descend, and that some bats might therefore be more closely related to some terrestrial animal than they are to other bats. By analogy, for instance, we know that a fully aquatic lifestyle did evolve more than once among mammals and that, for example, dolphins are not related to manatees.

The reason that we might think this is true of bats is that there are at least two kinds of bat that are visibly, and behaviourally, very different from one another. Firstly, there are the enormous majority of bat species, which are fairly small animals, with strange-looking faces, often eating insects, and pinging things with their sonar. Then, there are the flying foxes, which are typically much larger, with dog-like faces, normal-sized ears, mostly eating fruit, and lacking true echolocation.

Sure, there's some variation here. Many regular bats also eat fruit (or nectar, or blood). Some relatives of flying foxes are, in fact, quite small, and a few can just about manage echolocation - although they do it in a different way than other bats do, and it isn't very good. But, nonetheless, taken together with other anatomical differences, there are clearly two kinds of bat, and these were traditionally named "microbats" and "megabats". But how related are they?

The fossil record isn't much help here, partly since bat skeletons, being light and fragile, don't preserve very well. Bats appeared almost immediately after the dinosaurs went extinct, and seem to have diversified very rapidly. But, especially around the late '80s, evidence began to appear that suggested that the two different kinds of bat really weren't all that closely related, with the megabats, for example, perhaps being close to tree-climbing primates, while microbats were sort of flying shrews.

It has, however, been clear since at least 2001 that this is wrong. In fact, it's worse than that, since the actual pattern doesn't look at all like what we'd expected. All bats do descend from a single common ancestor, and they are divided into two suborders. But it turns out that the megabats are just one branch within one of those suborders, apparently having lost the power of echolocation, perhaps because their larger size made it too energy-intensive. We don't even have sensible names for these two types of bat, since broadly similar-looking creatures exist in both branches; in the end we called them "Yin bats" and "Yang bats" because... well, that's two contrasting things, so why not?

The most popular of the eight recently proposed possibilities

This didn't, however, clear up how bats relate to other mammals. In fact, that's still not entirely clear. We do know that they fit in the same general corner of the mammalian family tree as the hoofed mammals, cetaceans, and large carnivores. But given that that includes horses, dolphins, and lions, it's a pretty big corner to be in and there is still debate as to which of those three might be the closest bat-relatives. The best bet is probably that bats split off before the others did, making them equally close to all of them, but it's not really settled.

Trying to work out how the different families of bat relate to one another, though, has proved even harder, and has been subject to considerable change over the last 20 years or so. Indeed, I said above that there are "at least" 18 families of bat, because while that's where we stood in 2000, there's reasonably strong evidence that we should now recognise 21, with the most recent having been defined in 2015.

Part of the problem is that, even when we're looking at species - a taxonomic level that should, at least in theory, represent something that actually exists in the real world - there are multiple different lines of evidence that can be used to figure out relationships and determine whether something is really distinct or not. Scientists, understandably, tend to use whichever method they're most experienced with, and different methods can produce conflicting results. If, however, multiple different lines of enquiry all give more or less the same answer, there's a fair chance that we're on to something.

These days, this idea of combining different sources of evidence to prune apart biological family trees is commonly called "integrative taxonomy"... although it's hardly a new idea. With bats, where much of the evidence is still debated, it can prove particularly useful.

For instance, a 2016 study of the moustached bats (Pteronotus spp.) of Central and South America started by looking at the genetic and molecular data on the six known species, and several of the subspecies. The authors compared this with detailed measurements of the animal's skulls, to confirm that species identified genetically really were physically distinct. Not only was this successful, but it also seemed to show that a number of supposed subspecies were more distinct than had been thought, potentially raising eight of them to full species status, and identifying two possible new species (as yet un-named) that hadn't been suspected before at all. What's perhaps particularly significant was that the pattern this revealed also matched one created purely by comparing the animal's echolocation calls, adding a third line of evidence.

Trying to figure out how higher level groups relate to one another can be trickier; there are no firm guidelines about what constitutes a "genus", a "tribe", or even a family. Even so, creating family trees can at least rule some possibilities out. An example here would be an examination of yellow-eared bats (Vampyressa spp.) in 2006 that used multiple lines of evidence to show that some members of the genus were more closely related to MacConnell's bat (Mesophylla macconelli) than they were to other supposed members of their own genus. The genus was accordingly split in two, with the two halves later turning out to be several steps apart.

With larger groups, such as families, problems can arise when we simply don't have a complete enough sample of all the various species to be able to place them accurately. This can even occur at the level of species. In the 1990s, there was debate as to the relationships within the long-eared bats (Plecotus spp.) since even analyses of the same gene kept coming up with different answers depending on who did the study. The mystery was solved in 2002, when a more complete data set showed that some of the samples belonged to a previously unknown species and that not accounting for this had given two different shapes to the family tree.

From a layman's point of view, much of this is complicated by the fact that most bat groups don't have readily known common names. We can understand cats, dogs, bears and so on - all families of the order Carnivora. But the families of bat have names like "slit-faced bats" and "disc-winged bats", and who would remember that short-tailed bats and long-tongued bats are both kinds of leaf-nosed bat, rather than, say, kinds of vesper bat?

It's probably fair to say that we are closing in on a reasonably consistent answer as to how the different kinds of bat are related. But there is still a way to go, but the sort of taxonomic stability that we see in many other mammals has yet to fully reach this, the second-largest of all mammalian orders.

[Photo by Shantanu Kuveskar, from Wikimedia Commons. Cladogram adapted from Tsagkogeorga et al. 2013.]