Family Values


I'm sure by now you've heard of the ginormous spider web that was spun in Texas. The thing was huge -- 200 yards long -- and it was spun by multiple different species. That interspecific collaboration got Bill Poser thinking, so he blogged about it at Language Log:

The web covers hundreds of square meters. Not only was it built by hundreds of spiders, who normally build isolated webs and eat each other if they get too close, but entomologist Allen Dean reports that they belong to twelve different families! We're talking massive inter-species communication here folks, and not particularly closely related species either. It is comparable to communication and collaboration between human beings (family Hominidae) and Lar Gibbons (family Hylobatidae).

Twelve different families! To show you how diverse these spiders are, Poser reveals what species we'd have to collaborate with to accomplish such inter-species team work. Only he underestimates the scale of the difference. You see, Poser falls victim to taxonomic bias, assuming that a family in one taxon (spiders) is equivalent to a family in another taxon (primates).

Before we get any further, allow me to remind you of the major taxonomic categories used for classifying animals:








To use humans as an example, we're in the kingdom Animalia, aka the animals. Our phylum is Chordata, which is made up predominantly of the vertebrates (ie, things with backbones). Our class is Mammalia, our order is Primates, our family is Hominidae, and our species name (made up of the genus and species titles) is Homo sapiens. Many mnemonics have been developed to aid students in remembering the taxonomic groupings, playing off the first letter in each word. My favorite: King Plays Cards On Fat Girls' Stomachs.

Anyway, mammalian families aren't all that diverse. The family Hominidae contains only four extant species (humans, chimps, gorillas, and orangutans). If they were invertebrates, we'd probably classify them as a genus (or sub genus). But, because they're closely related to us, they get their own family (that's taxonomic bias). Other mammalian families are more diverse; Muridae (mice, rats, and gerbils) contains hundreds of species, and it's the largest of all mammalian families. As a point of comparison, the genus Drosophila has over a thousand species. Yes, a single insect genus is more specious speciose that the largest mammalian order.

But counting species isn't the only way to determine diversity within a taxon. We can also estimate the age of the taxon using fossils and DNA. Let's look again at the Muridae. The species that make up this family last shared a common ancestor approximately 25 million years ago (mya). To go back to our invertebrate example, the species in the genus Drosophila last shared a common ancestor approximately 60 mya. Once again, the invertebrate genus is more diverse than the mammalian family.


Hopefully I've shown you that comparing taxonomic categories across different taxa is fairly meaningless. What gets named as a family in one taxon contains less diversity than a genus in another taxon. But what about the spiders? What would be the equivalent collaborator for humans to accomplish the same amount of inter-family collaboration that we see in the spider web in Texas? A group of entomologists at Texas A&M sampled spiders from the web and classified them based on their families and species. The three most abundant families were:

  1. Tetragnathidae (long-jawed orb weavers)
  2. Salticidae (jumping spiders)
  3. Araneidae (orb-weaver spiders)

All three of these families belong to the suborder Araneomorphae, which was thought to have radiated in the late Paleozoic or early Mesozoic Era (reference), or about 250 mya. Tetragnathidae have been observed as early as the Cretaceous Period, which came at the end of the Mesozoic Era, about 150 mya. We'll use this date as a lower-bound estimate of the divergence times for the spiders found in the web in Texas. What follows is therefore a search for a taxon with the minimum amount of divergence from humans to achieve the equivalent amount of divergence between the spiders in the huge web.


Mammals diverged from birds and reptiles about 300 mya. This date can be used to calibrate a molecular clock, which can then be used to estimate the divergence dates between different vertebrate lineages provided we can sequence DNA from extant representatives of those lineages. Kumar and Hedges used this approach to estimate the divergence times between humans and various other vertebrates. At about 150 mya, the human lineage (and all other eutherian mammals) diverged from the marsupials. This was followed by a rapid radiation of the eutherians, giving rise to the mammalian orders we see today. Humans and gibbons (Poser's chosen equivalent collaborator) diverged about 15 mya.

Contrary to what Poser claims, humans collaborating with gibbons is not equivalent to the various spider families working together to build I gigantic web. To accomplish such an inter-taxon collaboration, humans would have to work with kangaroos, koalas, or possums. I'm not sure how they do things in Australia, but I doubt they've got Tasmanian devils working construction.

More like this

"a single insect genus is more specious that the largest mammalian order"
That's kind of a judgment mean "speciose."

My favorite mnemonic for the Linnaean taxonomic hierarchy:
Kissing Passionately Can Open For Great Sex.

By Sven DiMilo (not verified) on 14 Sep 2007 #permalink

Thanks for the improved estimate. However, I don't think that it is entirely fair to say that I got it wrong. The fault is really that of biologists who continue to use a patently inadequate system of taxonomy, namely one that reifies categories like "family", as if they had some meaning, when they don't. Without the detailed information about which species are involved and the estimates of divergence times, to which I didn't have access. the traditional Linnean taxonomy provides the only available distance measure.

So, why don't biologists bite the bullet and forget about these meaningless node labels?

So, why don't biologists bite the bullet and forget about these meaningless node labels?

We don't have the energy to overcoming the inertia of the current taxonomic system. Plus, changing taxonomy requires dealing with taxonomists, which can be an arduous task.

I'm a fish taxonomist. I think the species I recognize are real entities, so let's not discuss species concepts here. I think a genus or family, etc, should be monophyletic. It should have some morphological uniformity, and should be diagnosable in such a ways as to separate it from any closely related genus or family. If a group is proposed as part of a general phylogenetic scheme, it should be at the same level as its sister group. If the sister group is a genus then the group, even if only a single species, should be a separate genus.

None of this says that genera, families, etc. have, or should have, equivilent numbers of species. So, as said, comparing numbers of families in different groups can be quite misleading. It has also been pointed out that well studied groups tend to have more species, and sometimes genera, than poorly studied groups. Drosophila being a fine example.

We have described a new genus for a species originally described in a genus where it really did not belong. Our genus was later synonmized with a genus in a different sufamily, but now it is back, I think about where it belongs. It is the single species sister group of another genus according to DNA. Another new genus for a single species. Not related the way we thought, according to DNA. Its sister group is another genus.

By Jim Thomerson (not verified) on 17 Sep 2007 #permalink

Great post! Having just completed my biology degree the memory of having to learn hundreds of species along with their appropiate classification is still fresh. I remember that after we had learnt a lot of plant taxonomy in first year they reclassified a lot of species and we had to learn a new system in second year - it was so frustrating. And our tutors kept telling us this was just because they wanted all equal taxons represent equal categories, which I think is more or less impossible anyway. Any system concosted by humans is bound to be anthropocentric, and in the end, what's wrong with that? In the end I'm kind of sceptic that a spider is going to walk up to a taxonomist any time soon to complain that spider families aren't equal to Mammalian ones.

By Uschi Symmons (not verified) on 22 Sep 2007 #permalink