lpetrich

The <a href="http://www.secweb.org/asset.asp?AssetID=228" target="_blank">recent Kiosk article on bee evolution</a> is IMO rather shallow. There are several points that ought to have been made but that were not, or at least were not made in appropriate detail. Examples:

*** Hive organization
*** Queen-bee skills
*** Command hierarchy
*** Pollen baskets
*** Honeycomb geometry
*** Honey
*** Stings

Now for the details.

*** Hive organization:

Honeybees (Apis mellifera) live in hives of as many as 10,000 individuals; queens are dependent on their workers for nearly everything. It may thus seem that honeybee hives are an irreducibly complex system.

However, there are bee species with much simpler hive organization, such as bumbebees (Bombus), and there are even solitary species, like carpenter bees (Xylocopa) and sweat bees (Halictidae). It is not difficult to imagine a progression from solitary bees to honeybees:

Solitary bees: each female cares for her offspring individually, collecting pollen for her grubs to eat. This means that each one has to find or make a nest for herself.

Finding or making a nest can sometimes be difficult; thus, a bee's daughters may have to use their mother's nest. But the mother bee might not appreciate reproductive competition, and thus may suppress her daughters' reproduction. These daughters may then help their mother care for their mother's grubs, because they share much of their genome with their mother's other offspring.

The result is simple sociality, like that of bumblebees. A reproducing female (queen) founds a hive and cares for her grubs. When they grow up, they become workers, who help care for their mother's new grubs. Bumblebees live in temperate climates, whose winters are often too cold for the bees to be very active. The queens look for good places to hide and the workers die, ending the hives.

However, in warm climates, bees can build longer- and longer-lived hives, since there are no troublesome winters to force hives to be restarted each warm season. There is still the problem of founding new hives, since existing hives can be destroyed or get hopelessly fouled. Queens could still do this alone, but workers could provide valiable assistance. So departing queens could take some workers along with them. The result is that queens never have to found hives in isolation, meaning that their ability to do so can atrophy.

And the result is honeybees, whose queens are perpetually dependent on their workers.

*** Queen-bee skills

The article IMO overestimates queen honeybees. They are not known to be able to survive for any appreciable length of time away from their workers. However, as I'd mentioned, queens of other species can found hives and take care of the grubs until they become workers.

*** Command hierarchy

The article claimed that queen bees are the hives' commanders, as it were. However, there is no convincing evidence of that; hives appear to run like some virtuous-anarchist commune. Worker bees move from task to task during their lives, starting off ther adulthood by taking care of the grubs, then to helping build the hive, then to hive defense, and then to foraging. However, they will perform different tasks if they detect enough bees working at some task or if they detect not enough bees performing some other task.

*** Pollen baskets

These are simply two strips of extra-long hair along a bee's hind legs that surround an outward-facing bare strip; I fail to see why they are supposed to be such a big mystery.

As to why females have them and males do not, that is also no big mystery. They grow in response to "become female" hormones, and not in response to "become male" hormones; there are similar phenomena apparent in our species.

*** Honeycomb geometry

How do the bees manage to build such structures without doing some very sophisticated mathematics? There is a simple way of doing that, a way that does not require hive-scale draftsbeeship. When building a new honeycomb cell, build it as close as possible to some existing cells. The actual procedure may be more elaborate than that, like building next to the in-between point of two existing cells, but the general principle is likely to be correct.

*** Honey

This may seem very convenient for us, but bees make honey for their own use. They suck nectar from flowers and carry it in their front stomachs back to the hive, to feed to other bees, and to store it for lean times. That stored nectar is, of course, honey.

*** Stings

Stings are essentially modified ovipositors (egg-laying tools) that their solitary-wasp ancestors had had. One favorite place for such wasps to lay their eggs is on caterpillars and the like; the grubs then eat their prey as it contines to live.

However, some prey needs to be sedated in order to allow the wasp to lay eggs on them; the ovipositor gets modified to deliver a tranquilizer, thus becoming a sting.

Some such wasps eventually develop a taste for pollen, becoming bees. However, their stings remain useful, as weapons against predators.

Solitary and simple-social bees all have smooth stings, which can easily be reused. However, among honeybees, only queens have smooth stings -- the workers have barbed ones. These stings stick in the victim, and the escaping worker pulls out her guts as she tries to escape, dying some hours later.

So why are worker honeybees kamikaze stingers and other bees not?

There are two reasons.

The first is that worker honeybees are more "expendable" than workers of other bee species, since their hives are bigger; loss of a few bees is an acceptable trade-off for superior stinging capability. A small hive cannot afford to lose as many workers, and a queen who dies when she stings can no longer reproduce. Thus, queen honeybees continue to have smooth stings.

The second is that honeybee hives are simply bigger, and thus a more attractive target for big predators -- a bear may want to eat honey from a honeybee hive, but may not have much interest in a bumblebee hive or a solitary-bee nest, because those are too small. Thus, big hives may need a forcefulness of stinging that is overkill for small hives and solitary-bee nests.

Secular Elation

You should have posted in the feedback forum as well. There, you would likely recieve a response from the writer of that article.

Dr.GH

I think that the Cheng article should be rewritten, and authorship shared between lpetrich and Cheng. And based on what I read, I would not think that it is unfair to give lpetrich lead authorship.

I also seem to recall a recent thread on bees with an extended discusion of the fossil record which rebutted a creationist tract.

Ah. Found it at ARN. some good points, and a number of good links and other references.

<a href="http://www.arn.org/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=13;t=000270" target="_blank">http://www.arn.org/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=13;t=000270</a>

The revised kiosk article should address this as well.

[ September 21, 2002: Message edited by: Dr.GH ]</p>

Duvenoy

Excellent!

I have but one quibble, and it's nit-picking.

They're 'larvae' rather than 'grubs'.

Thanks for a great analysis!

doov

MrDarwin

I agree with you to a large extent, but it's quite possible that solitary bees are descended from social bees, not vice versa. The problem is that the Hymenoptera as a whole--ants, wasps, hornets, bees--are largely social (with the exception of some of the "primitive" groups like sawflies and parasitic wasps). It's an interesting question whether social behaviors have evolved independently in these different groups, or the common ancestor of the ants + wasps + bees clade was social, with some groups (all bees, I believe) reverting to solitary behavior. If I have time I'll look into it.

Non-praying Mantis

Ummm... Nope.

As an entomologist, I disagree with your statement that Hymenoptera as a whole are social. Most species, families, and genera of Hymenoptera are not social, such as sawflies and parasitic wasps. There are way more families of non-social, parasitic wasps than anything else in the Hymenoptera.

As to how many times eusociality (what entomologists call the social structure of bees, wasps, and ants) evolved in Hymenoptera, that is an open question, but IIRC, most entomologists suggest that eusociality evolved several times in the Hymenoptera, even though this would not be as "parsimonious" in clades of the Hymenoptera.

Eusociality has also evolved another time independently in insects, the termites.

NPM

MrDarwin

NPM, you caught me before I had a chance to modify my posting! You are quite correct, the vast majority of Hymenoptera are not social, although the most familiar and well-known are (bees, wasps, ants).

I've been doing some digging and I'm writing something up right now about the evolution of social behavior in Hymenoptera, but in a nutshell it looks like Ipetrich was correct, that eusociality arose independently in Apoidea (bees) and Vespoidea (wasps, hornets, ants). But I do find it interesting that these two groups are each other's closest relatives, suggesting that the common ancestor of these groups had some suite of characters predisposing its descendants to the evolution of eusociality.

(BTW I have an undergraduate degree in entomology myself, although I leaned more toward Coleoptera than Hymenoptera. Had I gone on in entomology, I would have liked to have studied the origin and early evolution of the insects. I am now a botanist.)

MrDarwin

Okay, I don't want to spend all day on this so I'm posting what I have so far, with a few interesting links I found, about <a href="http://tolweb.org/tree?group=Hymenoptera&contgroup=Endopterygota" target="_blank">Hymenoptera</a>.

In Hymenoptera, eusociality is found only within one small subgroup of Hymenoptera, the Aculeata. Within this group, most phylogenies seem to agree that Apoidea (bees) are most closely related to Vespoidea (paper wasps, hornets, ants). From the <a href="http://tolweb.org/tree" target="_blank">"Tree of Life" website:</a>



See also <a href="http://www.life.uiuc.edu/whitfield/annual_review98.pdf" target="_blank">here</a> (pdf file) for some info on Hymenopteran phylogeny, primarily with reference to parasitic Hymenoptera.

However, most references also seem to agree that eusociality arose independently in the Apoidea and Vespoidea. And within these groups, there is a very <a href="http://faculty.washington.edu/sodonnel/pdfGlandsdef.pdf" target="_blank">wide range of social behavior</a> and <a href="http://faculty.washington.edu/sodonnel/pdfCastedetermin.pdf" target="_blank">caste determination</a>.

The evidence for more than one origin of social behavior in hymenopterans, but all within closely related groups, suggests that something about the ancestor of those groups had some characteristic that favored eusociality. But overall it looks like ipetrich was correct, and that the range of social behaviors we see in bees today probably do reflect the gradual evolution of social behavior within bees, if not within Hymenoptera as a whole.

I wanted to provide more about the origins of eusociality itself, but didn't find enough online to make the time spent worthwhile. But looking at outgroups--Chrysidoidea and Ichneumonoidea--strongly suggests that the common ancestor of the eusocial insects was parasitic. (In fact, although the most well-known and widely recognized Hymenoptera are social insects, a huge number of species in the order are parasitic.)

Sphecid wasps (e.g., mud-daubers, cicada killers) appear to be basal within the Apoidea (there is some disagreement as to whether they belong within Apoidea, or should be their own group, Sphecoidea, as in the tree above). Sphecid wasps are solitary and parasitic; many are ground nesting, but others build nests. It does suggest independent origins of eusociality in Apioidea and Vespoidea from parasitic ancestors. The phylogeny suggests that nest-building (mud-daubers, etc.) may have already been present in the common ancestor, although it was probably solitary and free-living. Sphecid wasps are certainly more like the social wasps in that they hunt for food and bring it back to a nest (albeit alive and paralyzed), rather than laying an egg on their prey and leaving the larvae to fend for themselves (as do most other parasitic wasps). Furthermore, these wasps can use their ovipositor as a stinger for self-defense, as the social hymenopterans do; the poison was apparently originally used to paralyze prey, but proved to be very useful for self-defense, first of the individual wasp, then of the nest. Perhaps eusociality got its start when several of these females began to nest together, providing group protection to more or less separate nests.

(edited, as usual, for @#$%&* links)

(edited again to add that I really shouldn't have called the Sphecidae "parasitic"; rather, they are predatory, and the larve devours the paralyzed prey alive from the outside, rather than developing inside the (not paralyzed) living prey as with many other parasitoid hymenoptera.)

[ September 23, 2002: Message edited by: MrDarwin ]</p>

MrDarwin

In a nutshell, what I was getting at was that whenever we find a species or group of related species that have unusual and complex combinations of characters and can't figure out how it could have evolved, it is often instructive to look to that group's closest relatives. In many cases we find slight variations on, or precursors to, these seemingly "irreducibly complex" systems.

One such example is the bombardier beetle; both the chemicals and morphological structures are found in their closest relatives, but in a slightly different form, and used in slightly different ways (e.g., the chemicals are combined to produce nauseating or caustic, but not hot, compounds).

Non-praying Mantis

Just one quick comment...

I would suggest that the above characteristic which makes the Hymenoptera more inclined to evolve sociality would be their method of sex determination in offspring.

In humans (and in many other animals), sex determination is made by the random assortment of certain chromosomes in the fertilized egg. In mammals this chromosome is the "Y" sex chromosome, which is much smaller than its homologous partner, the "X" sex chromosome. Males receive one copy of the X and one of the Y while females receive two copies of the X. This makes a physiological "choosing" of the sex of offspring rather difficult.

OTOH, Hymenoptera all have a sex determination based on ploidy status. Diploids, such as most animals (including humans), have two sets of chromosomes, each represented as a homologous pair. These chromosomes all code for similar traits, but the actual genes in the chromosomes themselves can be different. Haploids, such as sex cells (eggs and sperm) in animals, only have one set. When fertilization of the egg occurs, the sperm's set combines with the egg's set, thus making a brand new combination in the offspring.

Hymenoptera determine sex based on this ploidy status, with males being haploid and females being diploid. This means that a hymenopteran mother can "choose" the sex of her offspring by using a fertilized egg for a female, or an unfertilized egg for a male. This might make sociality more feasible because not only do hymenopteran males share almost 100% of their genes with their mother, making kin selection much more likely, and because the female could choose to produce female workers or male reproductives at any time.

(...and a long explaination!)

NPM