Razib Khan One-stop-shopping for all of my content

March 31, 2010

When sickliness is manliness

Filed under: Behavior,Disease,Evolution,Evolutionary Genetics,Genetics,Immunity,Sex — Razib Khan @ 2:25 am

ResearchBlogging.orgBelow I note that sex matters when it comes to evolution, specifically in the case of how sexual reproduction forces the bits of the genome to be passed back and forth across sexes. In fact, the origin of sex is arguably the most important evolutionary question after the origin of species, and it remains one of the most active areas of research in evolutionary genetics. More specifically the existence of males, who do not bear offspring themselves but seem to be transient gene carriers is a major conundrum. But that’s not the main issue in this post. Let’s take the existence of males as a given. How do sex differences play out in evolutionary terms shaping other phenotypes? Consider Bateman’s principle:

Bateman’s principle is the theory that females almost always invest more energy into producing offspring than males, and therefore in most species females are a limiting resource over which the other sex will compete.

Female ova are energetically more expensive, and scarcer, than male sperm. Additionally, in mammals and other live-bearing species the female invests more time and energy after the point of fertilization but before the young exhibit any modicum of organismic independence (the seahorse being the exception). And, often the female is the “primary caregiver” in the case of species where the offspring require more care after birth. The logic of Bateman’s principle is so obvious when its premises are stated that it easily leads to a proliferation of numerous inferences, and many data are “explained” by its operation (in Mother Nature: Maternal Instincts and How They Shape the Human Species the biological anthroplogist Sarah Hrdy moots the complaint that the principle is applied rather too generously in the context of an important operationally monogamous primate, humans).

But the general behavioral point is rooted in realities of anatomy and life-history; in many dioecious species males and females exhibit a great deal of biological and behavioral dimorphism. But the direction and nature of dimorphism varies. Male gorillas and elephant seals are far larger than females of their kind, but among raptors females are larger. If evolution operated like Newtonian mechanics I assume we wouldn’t be theorizing about why species or sex existed at all, we’d all long ago have evolved toward perfectly adapted spherical cows floating in our own effluvium, a species which is a biosphere.

Going beyond what is skin deep, in humans it is often stated that males are less immunologically robust than females. Some argue that this is due to higher testosterone levels, which produce a weakened immune system. Amtoz Zahavi might argue that this is an illustration of the ‘handicap principle’. Only very robust males who are genetically superior can ‘afford’ the weakened immune system which high testosterone produces, in addition to the various secondary sexual characteristics beloved of film goers. Others would naturally suggest that male behavior is to blame. For example, perhaps males forage or wander about more, all the better to catch bugs, and they pay less attention to cleanliness.

But could there be a deeper evolutionary dynamic rooted in the differential behaviors implied from Bateman’s principle? A new paper in The Proceedings of the Royal Society explores this question with a mathematical model, The evolution of sex-specific immune defences:

Why do males and females often differ in their ability to cope with infection? Beyond physiological mechanisms, it has recently been proposed that life-history theory could explain immune differences from an adaptive point of view in relation to sex-specific reproductive strategies. However, a point often overlooked is that the benefits of immunity, and possibly the costs, depend not only on the host genotype but also on the presence and the phenotype of pathogens. To address this issue we developed an adaptive dynamic model that includes host–pathogen population dynamics and host sexual reproduction. Our model predicts that, although different reproductive strategies, following Bateman’s principle, are not enough to select for different levels of immunity, males and females respond differently to further changes in the characteristics of either sex. For example, if males are more exposed to infection than females (e.g. for behavioural reasons), it is possible to see them evolve lower immunocompetence than females. This and other counterintuitive results highlight the importance of ecological feedbacks in the evolution of immune defences. While this study focuses on sex-specific natural selection, it could easily be extended to include sexual selection and thus help to understand the interplay between the two processes.

The paper is Open Access, so you can read it for yourself. The formalism is heavy going, and the text makes it clear that they stuffed a lot of it into the supplements. You can basically “hum” through the formalism, but I thought I’d lay it out real quick, or at least major aspects.

This shows the birth rate of a given genotype contingent upon population density & proportions of males & females infected with a pathogen


These equations takes the first and nests them into an epidemiological framework which illustrates pathogen transmission (look at the first right hand term in the first two)


And these are the three models that they ran computations with


There are many symbols in those equations which aren’t obvious, and very difficult to keep track of. Here’s the table which shows what the symbols mean….


If you really want to understand the methods and derivations, as well how the details of how they computae evolutionarily stable strategies, you’ll have to go into the supplements. Let’s just assume that their findings are valid based on their premises.


- They assume no sexual selection
- They assume unlimited male gametes, so total reproductive skew where one male fertilizes all females is possible
- Fecundity is inversely correlated with population density
- Total population growth is ultimately dependent on females, they are the “rate limiting” sex
- Total population growth is proportional to density
- There is no acquired immunity
- There is no evolution of the pathogen in this model

Basically the model is exploring a quantitative trait which exhibits characteristics in relation to resistance of acquiring the pathogen and tolerance of it once the pathogen is acquired. In terms of the “three models,” the first is one where there is resistance to the pathogen, individuals recover from infection and decrease pathogen fitness. The second is one of tolerance, individuals are infected, but may still reproduce while infected. Note that the ability to resist or tolerate infection has a trade off, reduced lifespan (consider some forms of malaria resistance). The third model shows the trade off of tolerance and resistance.

The “pay off” of the paper is that they show that the male evolutionarily stable strategy (ESS), that is, a morph which can not be “invaded” by a mutation, may be one of reduced immune resistance in certain circumstances of high rates of infection. There is an exploration of varying rates of virulence, but there was no counterintuitive finding so I won’t cover that. In any case, here’s the figure:


The text is small, so to clarify:

1) The two panels on the top left are for model 1, and show variation in male and female recovery from infection left to right (resistance)

2) The two panels on the bottom left are for model 2, and show variation in male and female fecundity when infected left to right (tolerance)

3) The four panels on the right are for model 3, and show variation in recovery in the top two panels and fecundity in the bottom two, with male parameters varied on the left and female on the right

The vertical axis on all of the panels are male infection rate, the horizontal the female infection rate. Circled crosses (⊕) indicate regions (delimited by solid lines) where females evolve higher immunocompetence than males. The lighter shading indicates a higher value of the trait at ESS (recovery or fecundity). Note that the two top left panels show a peculiar pattern for males, the sort of counterintuitive finding which the model promises: when infection rates among males are very high their resistance levels drop. Why? The model is constructed so that resistance has a cost, and if they keep getting infected the cost is constant and there’s no benefit as they keep getting sick. In short it is better to breed actively for a short time and die than attempt to fight a losing battle against infection (I can think of possible explanations of behavior and biological resistance in high disease human societies right now). It is at medium levels of infection rates that males develop strong immune systems so that they recover. The bottom right portion of panel which shows variation in male resistance illustrates a trend where high female infection results in reduced immune state in males. Why? The argument is simple; female population drops due to disease result in a massive overall population drop and the epidemiological model is such that lower densities hinder pathogen transmission. So the cost for resistance becomes higher than the upside toward short-term promiscuous breeding in hopes of not catching the disease. Another point that is notable from the panels is that males seem to be more sensitive to variation in infection rates. This makes sense insofar as males exhibit a higher potential variance in reproductive outcomes because of the difference in behavior baked into the model (males have higher intrasexual competition).

One can say much more, as is said in the paper. Since you can read it yourself, I commend you to do so if you are curious. Rather, I would like a step back and ask: what does this “prove?” It does not prove anything, rather, this is a model with many assumptions which still manages to be quite gnarly on a first run through. It is though suggestive in joint consideration with empirical trends which have long been observed. Those empirical trends emerge out of particular dynamics and background parameters, and models can help us formalize and project abstractly around real concrete biological problems. The authors admit their model is simple, but they also assert that they’ve added layers of complexity which is necessary to understand the dynamics in the real world with any level of clarity. In the future they promise to add sexual selection, which I suspect will make a much bigger splash than this.

I’ll let them finish. From their conclusion:

We assessed the selective pressures on a subset of sex-specific traits (recovery rate, reproductive success during infection and lifespan) caused by arbitrary differences between males and females in infection rate or virulence (i.e. disease-induced death rate). In so doing, we covered a range of scenarios whereby sex-specific reproductive traits such as hormones and behaviour could plausibly affect the exposure to infection…r the severity of disease…First, we showed that changes in the traits of either sex affect the selective pressures on both sexes, either in the same or in opposite directions, depending on the ecological feedbacks. For example, an increase in male susceptibility (or exposure) to infection favours the spread of the pathogen in the whole population and therefore tends to select for higher resistance or tolerance in both sexes if the cost of immunity is constitutive. However, above a certain level of exposure, the benefit of rapid recovery in males decreases owing to constant reinfection (we assume no acquired immunity). This selects for lower resistance in males, ultimately leading to the counterintuitive situation where males with higher susceptibility or exposure to infection than females evolve lower immunocompetence…A similar pattern arises if the cost of immunity is facultative, in the form of a trade-off between rate of recovery and relative fecundity during infection (model (iii)): if males happen to be more susceptible (or exposed) to infection than females, they are predicted to evolve a longer infectious period balanced by higher sexual activity during infection than females.

Restif, O., & Amos, W. (2010). The evolution of sex-specific immune defences Proceedings of the Royal Society B: Biological Sciences DOI: 10.1098/rspb.2010.0188

Powered by WordPress