Much of our current understanding of testosterone comes from studies of species that inhabit the northern temperate region of the globe, perhaps unsurprisingly, because northern temperate species are readily accessible to universities in the developed world. From these studies there is a reasonable amount of evidence that testosterone plays an important role in the production of song by male songbirds. But there are very few studies of birds in other regions and this is important because the role of testosterone is not necessarily conserved across species in these different environments. So we need to study tropical and subtropical species too.
And so I set out to ask:
does circulating testosterone regulate male song production in white-browed sparrow weavers in the kalahari?
Throughout the breeding season, I got up at 3am to monitor male song production (they sing their solo songs almost exclusively at dawn) and I caught the birds late at night to collect a blood sample (when individual males can be readily caught from their sleeping chambers). It was hard and tiring work that meant I had to sleep in short (3-4hr) shifts at night and in the middle of the day (when temperatures were over 30 degrees C!). I then spent weeks analysing the samples in a collaborator's laboratory in order to measure testosterone concentration, and more weeks analysing song files.
And when all the data was in, the answer came back: NO.
There was no evidence that circulating testosterone concentrations were directly related to dominant male song production. or that males who sang more had higher testosterone than males that sang less.
Digging into the literature it quickly became clear that correlative evidence for a direct relationship between song production and circulating testosterone was not all that common. On the other hand, experimental manipulation of circulating testosterone did affect song production in many cases. So did our result mean that there is a latitudinal difference in the role of circulating testosterone for male song production? Or was there more to it? If the relationship between circulating testosterone concentration and song production is not strong and consistent across studies, then perhaps there is something more complicated going on under the hood?
And indeed, recent work indicates that this is very much the case. There are at least 3 important reasons why a direct relationship between circulating testosterone concentration and song production might not be necessary or could be obscured:
1. variation in receptor expression - the androgen receptors that circulating testosterone bind to can vary in number or density, which could account for the effectiveness of circulating testosterone concentration without variation in circulating testosterone levels
2. local conversion - aromatase coverts circulating testosterone into estradiol, which may then bind to estradiol receptors at target sites, therefore the concentration of estradiol may be more directly related to behaviour than that of circulating testosterone
3. local synthesis - testosterone and other steroids can also be synthesised locally at target tissue sites, which means that peripheral measurements don't necessarily tell us an awful lot about what is going on in the brain or at other specific sites.
So what does this all mean? Well it means that there is still a lot more work to be done! We need to better incorporate an understanding of the complexity of neuroendocrine mechanisms into evolutionary hypotheses about the role that testosterone may play in life-history trade-offs and animal communication systems. Often, working hypotheses are based on the assumption that there is a simple concentration-dependent effect of testosterone, which need not be the case.
More generally, we need to take the approach that hormones are more than mechanistic links in the translation of genotype to phenotype. Due to their pleiotropic effects on gene expression, hormones can structure genetic variances and covariances that determine how a population responds to selection. But hormonal mechanisms are more than simply the circulating hormone acting alone, so we need to characterise the other components too.
Further studies of songbirds outside the northern temperate zone are now essential in order that we can begin to understand whether endocrine systems are conserved or divergent across species, and why this is the case.
You can read more about this work and the associated literature here in the open access article.
Male birdsong is known to perform two key functions: attracting a female and repelling rival males. In kalahari white-browed sparrow weavers (Plocepasser mahali), the dominant males monopolise access to breeding opportunities, both within- and extra-pair mating all goes to the dominant males: subordinate males don't get a look-in. Dominance tenures are also very long, therefore in order to breed, subordinate males must acquire a dominant position.
Dominant male white-browed sparrow weavers sing pretty much every day at dawn during the breeding season. Their territories are very densely-packed so a group's roosting tree might be just 100m from their neighbour. Waiting in the darkness at dawn, it is like sparrow weaver surround-sound as all the males in the area one-by-one begin their dawn song performance. You really get a feel for what's going on in the neighbourhood. So it occurred to me, perhaps subordinate males monitor the dominant males in the neighbourhood. This might give subordinates a good idea about their future prospects of promotion - if a dominant male's song performance is poor, then perhaps he is on his way out or is temporarily weakened and now is a good time to attempt a take-over.
So the first step was to ask: is dominant male dawn song performance a good indicator of current condition? I gave dominant males a small immune-challenge to see whether this would affect their dawn song performance. Would they stop altogether or would they sing less than the previous day? And across males, would the strength of their immune response be related to their dawn song performance?
The answer was yes... dominant males sang less the day after an immune challenge, but interestingly, they all still sang. The more a male sang on the day before the immune-challenge, the stronger was his immune-response to that challenge, suggesting that, at least in this species, dawn solo song might tell other birds something about a given dominant male's current immune status. The open access paper is available here if you want to know more.
The next step would be to find out whether subordinate males utilise this information during dispersal, in order to maximise their chances of winning the jackpot and becoming the dominant male in a group.
behaviour and evolution