Where did ants (and their social behavior) come from?

It’s been known for a while that ants evolved from a wasplike ancestor, but the relationships among living wasps and ants have been unclear.  An new online paper in Current Biology by Brian Johnson et al. (the senior author is my old friend Phil Ward at UC Davis), using a huge amount of sequence data, seems to have resolved the situation pretty definitively. The group they studied were the “aculeate” Hymenoptera, which includes all the “stinging” Hymenoptera.

Earlier workers suggested that ants evolved from ectoparasitic wasps that laid their egg on other insects, and whose larvae then hatched on the host or in its nest.  These wasps were obviously not “eusocial” (i.e., having reproductive queens with a nonreproductive female worker caste), leaving open the question of what factors prompted the evolution of that marvel of behavior, eusociality (mostly found in insects but also in naked mole rats and a few other species). What factors could prompt the evolution of a sterile worker caste? That was a question that puzzled Darwin in The Origin.

The new phylogeny, based on 308 to 5,214 genes, is strongly supported and is given below. It shows that the sister group of ants are the speciform wasps and bees (“Apoidea”), which is something nobody suspected. That group includes the familiar mud-dauber wasps, which you might have seen making their cylindrical nests on your house.

That novel finding gives one clue to the origin of eusociality in ants, for the Apoidea is a group characterized by two things: like ants, they build nests to contain their eggs, and they provision those nests by bringing prey back to feed the offspring.  The outgroups, scoliid wasps and their relatives, as well as the tiphioids and some pompiloids , retain their ectoparasitic habits, and none of them are eusocial. (Eusociality is indicated by asterisks in the diagram below.):

Picture 1

Figure 3 (from paper). Evolution of the Aculeate Hymenoptera Blue-green branches represent parasitoidism; orange branches represent nest construction and predation (with pollenivory and omnivory as derivative states thereof). Asterisks designate lineages containing eusocial species. Ants are entirely eusocial, but this is not true of all speciesof Vespidae and Apoidea. Biological information is from various sources, summarized in Gauld and Bolton [2] and Huber [13]. Names of superfamilies are modified from Pilgrim et al. [8]. Placement of Rhopalosomatidae is based on Pilgrim et al. [8] and Debevec et al. [11]. Images courtesy of Alexander Wild and Kurt Schaefer.

The results suggest that the common ancestor of ants and Apoidea exhibited two behavioral traits not shared with the ectoparasitoid Scolioidea, namely the building of a defensible nest and transport of prey to the nest; and that these were important prerequisites for the evolution of eusociality.

The offspring in such nests, of course, contain the genes of their mothers, and all of these species (including every group in the diagram above) are haplodiploid: the males are haploid, with a single set of chromosomes, while females have two sets.  The reproductive queen produces males from unfertilized eggs, and females from fertilized ones. It’s been hypothesized that this situation also favors the evolution of eusociality, as sterile workers share 3/4 of their genes with the sisters they help rear, which might give an evolutionary leg up to becoming sterile—using your efforts not to produce your own offspring, but to make more copies of your genes by helping your mother rear your sisters.

Now haplodiploidy is not sufficient to promote the evolution of eusociality, as most of the non-eusocial species above are also haplodiploid. And termites, which are not haplodiploid (both males and females have two sets of chromosomes), are also eusocial, with a reproductive queen and sterile workers. But previous work that I’ve discussed elsewhere suggests that haplodiploidy is nevertheless a factor important in promoting eusociality.

Rather, it may be the combination of haplodiploidy and the construction of a defensible and provisioned nest, which allows you to keep your offspring around to help you, that was pivotal in promoting the evolution of the eusocial lifestyle. This paper, at least, supports that two-factor hypothesis, and that’s why it’s important.

Edward O. Wilson and his colleagues have suggested that haploidiploidy and the “kin selection” it can cause was irrelevant in the evolution of eusociality, and the “nest” hypothesis was the important factor, along with selection among groups.  The Johnson et al. paper doesn’t speak to that hypothesis, except to support the idea that defensible, provisioned nests are important.

Rather, Wilson’s idea of group selection, and of the irrelevance of kin selection, seems unlikely for other reasons.  Those include previous work showing that eusociality evolved only in hymenopteran groups whose females mated only once (that causes increased relatedness in offspring, promoting kin selection), as well as the general unlikelihood of (and failure to demonstrate) group selection in any species. Finally, although the Nowak et al. paper in Nature claimed that, in their models, kin selection did not promote eusociality, they showed no such thing, for their models did not vary the level of relatedness among individuals.

You can read my posts earlier posts on the “group selection/eusociality” controversy here.

Carl Zimmer had a piece on this work in last Tuesday’s New York Times.

_____________

Johnson, B. R., M. L. Borowiec, J. C. Chiu, E. K. Lee, J. Atallah, and P. S. Ward. 2013. Phylogenomics resolves evolutionary relationships among ants, bees, and wasps. Current Biology, online, in press.  http://dx.doi.org/10.1016/j.cub.2013.08.050

18 Comments

  1. marksolock
    Posted October 19, 2013 at 11:14 am | Permalink

    Reblogged this on Mark Solock Blog.

  2. Posted October 19, 2013 at 11:28 am | Permalink

    Curious. Most of those wee beasties give me the heebie-jeebies, at least a bit. But not ants nor bees. I wonder if that’s in any way related to their relatedness.

    b&

  3. Mark Joseph
    Posted October 19, 2013 at 11:45 am | Permalink

    Why does the cladogram show two different twigs labeled the same way, “tiphioid-pompiloid wasps”? Is that an error, or (much more likely) some facet of taxonomy with which I’m not familiar? Thanks in advance.

    • andreschuiteman
      Posted October 19, 2013 at 12:46 pm | Permalink

      It’s not an error. It means that the tiphioid-pompiloid wasps are a paraphyletic group, at least according to this cladogram.

      • Mark Joseph
        Posted October 19, 2013 at 12:58 pm | Permalink

        Well, now that you say that, it’s obvious. Thanks!

        • Diane G.
          Posted October 19, 2013 at 12:58 pm | Permalink

          But I’m glad you asked.

          • Mark Joseph
            Posted October 19, 2013 at 2:16 pm | Permalink

            :-)

            “I have learned just enough to perceive my ignorance, and, ashamed of my defects in everything, I wish to seize the opportunity of remedying them.” (Michael Faraday)

  4. Diane G.
    Posted October 19, 2013 at 12:44 pm | Permalink

    sub

    • teacupoftheapocalypse
      Posted October 19, 2013 at 3:28 pm | Permalink

      ditto

  5. Bob Carlson
    Posted October 19, 2013 at 12:48 pm | Permalink

    It shows that the sister group of ants are the speciform wasps and bees (“Apoidea”), which is something nobody suspected.

    I forwarded this post to a few entomologists, including a couple of hymenopterists. One of the latter remarked “I thought most hymers believe ants came from wasps!” That was my impression as well.

    The group they studied were the “aculeate” Hymenoptera, which includes all the “stinging” Hymenoptera.

    Although one unlikely to be stung by Hymenoptera other than the aculeates, many of the larger species of the largest family of Hymenoptera are able to sting humans. Henry Townes, the elder brother of Charles, worked mostly on the family Ichneumonidae. He said that the stings of certain tropical Cryptinae were more potent than those of most aculeates and could remain sore for several days. And they can even cause serious reactions. In the early 70’s, a secret service agent brought a specimen to my office in the National Museum of Natural History to get it identified. It was a female of the genus Pimpla. It had gotten trapped inside his shirtsleeve in his hotel room in eastern Europe and had stung him causing a severe reaction. Females of many smaller Ichneumonidae don’t have the strength to penetrate human skin with their ovipositors, but they try when trapped. In fact, even though they lack the equipment, male ichneumonids will try to sting when trapped.

  6. Posted October 19, 2013 at 1:05 pm | Permalink

    I too was pretty amazed at this finding. There is a famous fossil insect in Cretaceous amber that is considered to be a transitional form between ants and wasps. Specifically, this insect seems to be an ant b/c its waist is modified into a petiole, but it also had wasp-like mandibles. You can see it at See Fig. 3. The article there explains that E.O. Wilson had described this fossil in 1967, and that he had linked ants specifically with the tiphiid wasps. I do not know if he had a specific reason to do that other than the view that tiphiids can resemble ants.

  7. Torbjörn Larsson, OM
    Posted October 19, 2013 at 1:21 pm | Permalink

    Rather, it may be the combination of haplodiploidy and the construction of a defensible and provisioned nest, which allows you to keep your offspring around to help you, that was pivotal in promoting the evolution of the eusocial lifestyle.

    Speaking of defensibility, when checking the terms I stumbled on this unreferenced claim:

    “It is commonly believed that the possession of a venomous sting was one of the important features promoting the evolution of social behavior, as it confers a level of anti-predator defense rarely approached by other invertebrates.[citation needed]” [ http://en.wikipedia.org/wiki/Aculeata ]

    But see also the mentioned termites and Bob Carlson’s comment.

    Is that “commonly” believed, i.e. is that an important factor for defensibility alongside nesting ability itself? Or is it the indefensible [sic] leap it looks like?

    • Posted October 19, 2013 at 2:27 pm | Permalink

      The stinger of hymenopterans is derived from the ovipositor, which is a tool used by most females to drill into something (wood, soil, or host insect) to lay eggs. Even in the ‘primitive’ species that do not sting the ovipositor is a definite though modest weapon for self defense. One can see how evolution for building burrows or other nests in non-social wasps would mean that the ovipositor was less needed as a drill for egg laying. Then, they could evolve this structure to be a dedicated weapon by exaptation. But the process was probably completed before the appearance of eusocial wasps and their descendant bees and ants. Having stingers probably helped make large eusocial colonies possible.

  8. Torbjörn Larsson, OM
    Posted October 19, 2013 at 1:24 pm | Permalink

    Also, great article! I heart ants (most of the time) and bees (most of the time).

  9. gerdien
    Posted October 20, 2013 at 5:01 am | Permalink

    from the Carl Zimmer article:
    “Dr. Boomsma agreed with this idea. “Colony life with altruistic helpers could only evolve after nests and parental care had evolved first,” he said.”

    This seems to be what Nowak et al maintained. Boomsma is a very prominent researcher of sociality in ants, including kin selection theory.

    • Posted October 20, 2013 at 5:10 am | Permalink

      That’s not all that Nowak at al. maintained, though they may be right on this issue. But that issue wasn’t the controversial one, for they also asserted these propositions, which were widely criticized:

      1. Relatedness (i.e., kin selection) was not important. It could in fact be just as important as nest building‚or more.

      2. Kin selection was not a type of natural selection.

      3. Selection among groups rather than among individual genotypes was important in this scenario.

      4. Kin selection is not an important evolutionary “force”.

  10. Bill Gilliland
    Posted October 20, 2013 at 6:43 am | Permalink

    “all of these species (including every group in the diagram above) are haplodiploid: the males are haploid, with half a set of chromosomes, while females have two sets.”

    Minor nitpick: Haplodiploid males don’t have half a set of chromosomes, they have a single set of chromosomes.


Follow

Get every new post delivered to your Inbox.

Join 29,406 other followers

%d bloggers like this: