There's (bee) love in the air

A couple of weeks ago I came across a rather unconventional-seeming insect gathering on the sidewalk on my way home from the lab:

Bombus sp. mating pair (below), and harassing male (on top)

For many bees, fall is the mating season. The Bombus (bumble bee) life cycle is rather intriguing, so here's the condensed version: in the springtime, queens emerge from diapause (hibernation/overwintering) and begin foraging on the first flowers of the season. Each queen seeks out a suitable nesting site, starts building her nest and stocking it with pollen and nectar, and begins laying her eggs. Once the first brood of hatched larvae mature into workers, these workers take over foraging for the hive and the queen no longer leaves the nest. Instead she focuses her energy on laying more eggs. As the hives grow, workers may specialize further in their roles in the hive, some becoming foragers, others working primarily as nurses of the young brood, etc. For most of the season, the queen will lay eggs fertilized with the sperm she obtained from mating in the previous fall, which she stores in a dedicated sperm storage organ inside her body until she needs it. In bumble bees, fertilized eggs become females. During most of the season, the queen can suppress the sexual development of the females in her hive, so they remain incapable of laying eggs, and they serve as workers. But, toward the end of the season her control over workers starts to fail and the workers start developing their ovaries. During this period the queen will also lay some unfertilized eggs, and these unfertilized eggs become males (referred to as drones), and these drones leave the colony in search of mates. The queen either dies naturally or is killed by her workers, and the colony disbands.

Once the colony disbands, the new queens are looking for a mate and a place to hibernate; the males try to find queens to fertilize. When they do get together to mate, if the queen doesn't find the male distasteful (if she does, she will evade him and even sting him if he gets too pushy about trying to mate with her), then he will attach himself to her and transfer some sperm into her. This part of the process only takes a couple of minutes at most, but a male bumble bee will stay attached to the queen and then produce a mating plug, basically just a sticky sperm-free substance that, if effective, will block the path up to the organ she uses to store sperm, thereby preventing any subsequent males from providing sperm. The process of transferring the mating plug can take up to 80 minutes (average of just under 37 minutes in one species, [Duvoisin et al. 1999]), so for the majority of the time that the queen and male are copulating, really it's just the male hanging on transferring the mating plug to prevent other males from getting a chance to fertilize the queen's eggs.

So I filmed some bee porn. I watched these three on the sidewalk for about 25 minutes before I left because I was getting cold (and because the residents of the houses started giving me pointed, stern looks from their porches, since I was loitering with my camera out on the sidewalk across from an elementary school...). I may have come across them just at the beginning of copulation, so in theory I could have caught the moment when sperm transfer was occurring on camera, but I think it's much more likely that what I have filmed is the period during which the male transfers the mating plug.  Whether or not I did, I would have no way to tell whether the male is transferring sperm or a mating plug just from the video. It's pretty cool for a few minutes, but gets repetitive after  a while, so here's a short clip to give you an idea of what it looked like:

I was aware of the production of mating plugs in Bombus, so I was somewhat confused to see the bee threesome pictured above. If the mating plugs work in the usual way that they are expected to, as a mechanical blockage preventing subsequent males' sperm from entering the reproductive tract of the female, then there would presumably be no benefit in being that harassing male on top; even if he waits until the other male leaves and tries to mate with the female, unless he has some way of overcoming the mating plug, then mating with this female would just waste some of his limited sperm supply. So naturally, I had to wonder what might be going on here that would lead to this kind of behaviour. There were a number of possibilities to consider:

(1) it may be that this is an extremely rare behaviour because it is maladaptive, and I just happened to be in the right place at the right time to witness it;

(2) mating opportunities are so scarce for male Bombus spp. that even a really slim chance to father the next generation by being the second mate to a queen, would be better than trying to find another queen to mate with;

(3) the odds of siring some of the queen's offspring, even as a second (or later) mate to a queen aren't actually as bad as I am presuming based on the presence of the mating plug;

(4) the mating plug doesn't directly or entirely prevent other sperm from entering the running to sire the queen's eggs, but instead serves some other purpose;

(5) some other process or force that hasn't occurred to me.

I won't spend much time on (5) except to say that it's the thing that keeps scientists up at night ("What if the thing I found was actually the result of something else I totally haven't even imagined, that just happens to be associated with the thing I measured???"). This is the unknown unknown. We try to get around this with study designs that reduce the likelihood that this is occurring, but we can't be perfect.

Anyway, witnessing this bee threesome gave me lots of questions and very few answers, so I have been reading up for a few weeks in preparation for this post. I was able to discount (1) as a possibility fairly quickly, as I found a bumble bee mating study which directly references the frequent presence of harasser males (in a laboratory setting) during copulation [Duvoisin et al. 1999], though they note that none of them successfully displaced the first male during mating in their observations. At any rate, this is a behaviour that has been documented previously, so it's very unlikely that I witnessed a one-off event.

Some previous work suggests that (2) is a contributing factor to the observed phenomenon of bee threesomes: male bumble bees very rarely get a chance to reproduce. This alone wouldn't explain the behaviour, though, since the harassing male has to hang around harassing this mating pair, which, if he's got no chance of siring any of the queen's eggs, is a complete waste of time that might be better spent looking for another queen to mate with. When I moved this mating threesome off the sidewalk--I was worried they would get trampled--I did notice that the harassing male seemed to be trying to displace the first male, but he didn't succeed. He may have waited until the pair were done mating and then tried his luck with the queen afterward, but if so there has to have been a good reason to do so. I did get some footage of his attempt to disrupt the first male:

(3) and (4) are related issues. If the mating plug is imperfect as a mechanical blockage, then a subsequent male partner still has some chance of siring her offspring; similarly, if the mating plug serves a different purpose, then subsequent males' sperm might still have a chance as well. One indirect but very telling question to examine (3) would be to ask whether all the offspring produced by one queen are half- or full-siblings. If they're all full siblings, then she must only have a single successful mate, who gets to sire all of her offspring. If they can be half siblings, then that means that multiple males can sire her offspring (i.e., that subsequent males definitely have a chance). Multiple patrilines (groups of offspring with the same father) have been documented in some bees, notably honeybees, but very little is currently known about how closely related individuals in bumble bee hives are. Certainly it is theoretically possible to find multiple patrilines within a bumblebee colony, and Lierch and Schmid-Hempel (1998) created colonies with multiple patrilines and showed that Bombus terrestris colonies with multiple patrilines are more resistant to parasites; a related study by Shykoff and Schmid-Hempel (1991) showed that pathogen transmission between bumble bee workers is lower in hives with more patrilines; consequently, multiple mating could be beneficial for queens, whose likelihood of producing a strong enough hive to reach end of season and produce new queens and males is higher if their resistance to parasitism is greater, e.g., through the presence of multiple patrilines. So, from the female perspective it may be better to have more mates and thus a more genetically diverse group of workers in the hive. However, the relationship between colony success and number of patrilines is not linear: Baer and Schmid-Hempel (2001) found that colonies with either just one, or more than four, patrilines did better than those with 2 or 3, and they speculate that this may be due to problems with worker social structure (i.e., conflicts between workers of different patrilines) with 2-3 patrilines present in the colony. Some older research has shown that more related honeybee workers are more likely to undertake the same kinds of colony work (i.e., that the tasks that individual workers end up doing within the colony structure are partly determined by genetics and/or relatedness) (Robinson and Page 1989), which raises the possibility that more patrilines can result in improved division of labour within the colony, but I don't think any such pattern has been shown in bumble bees.

The Duvoisin et al. (1999) paper mentioned above for discusses possible alternative functions of the mating plug, though they reject many of them and do not make any firm commitment to any particular function. It may be a form of nuptial gift (nourishment to improve the female's reproductive success), as is found in some organisms (e.g., a male bringing a female food), but this is unlikely as they found no useful nutrients in the mating plugs in their study. Another possibility is that the mating plug might actually serve more to prevent backflow of sperm, and thereby its loss, than to prevent subsequent males from transferring sperm. These researchers did note that the duration of copulation, which greatly exceeds the time needed to transfer sperm into the female's genital tract, may be related to the time needed for the sperm to get to the sperm storing organ inside the queen; if so, then this might be a secondary function of the mating plug (it wouldn't be the primary function, as you don't need a mating plug to stay attached to the queen for a longer period so if this were the main important thing, we would expect no mating plug but long mating duration). Fortunately, there's a very interesting analysis that suggests a possible function of the mating plug that is not mechanical interference with subsequent males: chemical behaviour modification! Baer et al. (2001) found a chemical in the mating plugs of one bumble bee species that reduced queens' willingness to accept other mates afterward. In other words, the mating plug may make the queen more likely to refuse other males and so allow a male to monopolize opportunities to fertilize her eggs through behavioural modification rather than through mechanical interference.

There are so many more amazing things to talk about in bumble bee mating, but I'm going to leave off here in the interest of not overwhelming everybody (including myself). I would just like to end by saying that there are so many questions left to answer in biology. It's a fascinating field, and every day I am confronted with the vastness of scientific knowledge, and the even greater vastness of things we have yet to understand. It's a wonderful time to be a scientist!