I recently got hooked on to the Behavioural Biology lecture series of Prof. Robert Sapolsky. It is absolutely addictive to watch, and highly recommended to anyone who is interested in evolutionary biology. In the third lecture of the series, he talks about a very interesting phenomenon called infanticide. The word is a combination of the words infant and homicide, and means intentional killing of infants by unrelated males. Simply put : males killing other males' new borns. But why would the males of any species kill other males' infants in a population, intentionally? The answer is pretty obvious when the concept of individual selection is applied. What happens is the following : You are a male in a group of males, and you guys just invaded a new population. You probably have a few more years left to reproduce, but the females in the population are either already pregnant or nurturing the infants of other males that were there before you. Following the concept of individual selection, you have only one purpose in mind : to pass the copy of your own genes onto the next generation as much as possible. But how can you do that if all the females are pregnant or dealing with their new borns, and therefore not avaliable?
Maybe a quick pause and some background information can be useful at this point. It's obvious that females can't get pregnant when they are already pregnant, but there is also a phase called lactation after birth, which is colloquially known as breastfeeding. During this phase, the female cannot start ovulating, and therefore cannot get pregnant again (called Lactational Amenorrhea). So dealing with a new born is not an obstacle only in terms of energy and time, but also an obstacle in terms of physiology.
Now it becomes easy to connect the dots : "the female that can potentially give birth to my kids, and pass on the copies of my genes, is dealing with some other guy's new born, and therefore I cannot get her pregnant", says the inner voice of the male. The solution becomes obvious : kill the new born, make the female available again for mating. Basically, it is a male "trick" to obtain sexual access to females who are otherwise preoccupied with maternal duties [1].
Sure, but what about the female? All that effort, energy, and nutrition put in the long months of pregnancy, and this new guy just comes and murders your new born? This costs a lot to the female biologically, and that kid was also her chance of passing the copies of her own genes! So as males came up with their own strategy, females also found a way to get around, which Sapolsky refers as the "one of the great all time ways to manipulate males".
The female solution is called Pseudo-Estrus. It means that the females act like they are ovulating, and ready to get pregnant with your baby, although they are already pregnant with some other guy's kid! They basically send physiological signals to males, such as swellings around their sexual body parts, and present themselves as they are ovulating. The male is happy, goes and mates with the female (who, don't forget, is already pregnant), and when the baby is out there - probably out there earlier than it is supposed to because she got pregnant with it way before - the male still thinks it's his baby, and does not get aggressive. Genius, right?! But at this point, another question rises : sure, males evolved the strategy of infanticide, females counteracted with Pseudo-Estrus, and then what? In fact, this also came up during the lecture, and one student asked : Why doesn't evolution favor the males that can figure out whether the baby is their theirs or not? You can watch the part where he explains the phenomenon of Pseudo-Estrus, and answers the question of that student below.
It is for you to judge, but with all due respect, Sapolsky gives an answer that was not very satisfactory to me. I continued watching the lectures (and again, they are amazing), but the question of why evolution stopped counteracting at that point bothered me a lot. You can see the conceptual consequences of such a question. Does coevolution have a limit? Is there an end to species (or sexes) developing counter strategies to pass on their genes, and converge to point where one of them falls for the other one's trick? If so, what determines this point of convergence?
Interestingly, some birds relieved me from my confusion :) I realized the connection while reading this amazing paper called "Arms races between and within species" by Dawkins and Krebs [2]. Apparently, this family of birds called Cuckoos has a very interesting reproductive strategy called brood parasitism. They mimic the eggs of the host they want to manipulate and leave their eggs to the host's nest so that the host raises the eggs like its own, while they fly on, enjoy life, and save the energy to reproduce more. You can see the photo (from Wiki article) of four clutches of reed warbler eggs, each containing one (larger) cuckoo egg below. They seem quite good at mimicry!
The paper goes on discussing the following paradox : why cuckoo hosts sometimes seem so clever in discriminating against poorly mimetic eggs, and at other times seem so stupid in feeding young clearly different from their own? Put in simpler words, the hosts (which cuckoos are manipulating) seem like they are quite good at discriminating the eggs, if they are mimicked even in a slightly poor way. But if they can't and the cuckoo bird makes it out of the egg, then the host acts like as if it's almost blind, and raises the bird anyway, although it looks completely different then it's own kids!
According to the authors, solution to this paradox lies in the combination of two concepts : evolutionary arms race and life-dinner principle. First concept, the arms race, means that there will always be counter-adaptation to manipulation, until... it becomes more costly for one side to go on adapting. This is where the life-dinner principle kicks in. One side has to be better, because counteraction might mean life to that species, whereas it might only mean dinner to the other. As in the case with foxes and rabbits : The rabbit runs faster than the fox, because the rabbit is running for its life, while the fox is only running for its dinner. For the case of cuckoos, I will directly quote the excerpt in the paper :
So far so good. But why this coevolution, this arms race stopped at this very particular point? It could have gone one step further by both sides, end at somewhere that the females are still better, right? So in some sense, this stable point of murder vs manipulation should work for both sexes.
It indeed does. Think of the following scenarios : if all males were so clever to avoid the manipulative females, and kill the infants of other males, they could drive their species to extinction. On the other hand, if all the males fall for this Pseudo-Estrus strategy and nobody understands which kid belongs to whom, then no male kills any infants! So between these two extremes, you can imagine that there will be a certain amount of confusion among males that is good for the continuation of the species. A healthy (!) ratio of murder over manipulation...
Since I am obsessed with these evolutionary sweet spots, had the code this one and see what comes out :) There goes the link to the Github project. The code is just a simple MATLAB script, and basically simulates an agent-based system where there is a certain "confusion rate" among males. This confusion rate is equal to the probability of any infant not being murdered by a male within the lactation phase of the mother. If the infant is murdered while the mother is still in the lactation phase, she becomes available for mating earlier than she would have if she continued breastfeeding. If the infant is not murdered, the mother completes the whole lactation phase, and becomes available for mating afterwards. You can find more details about the modelled scenario in the comments of the script. I set the parameters such that it would be consistent with a human population living in the Paleolithic era (life expectancy of 33 years, pregnancy of 40 weeks, etc.) for fun :)
One interesting thing to look at is the behaviour of this system given different confusion rates among males (which is equal to the probability of not killing other males' infants). You can see the population size plots across 4 different time scales below : 5, 10, 20, and 50 years, where the confusion rate is denoted by \(r\).
If the confusion rate is around \(60\%\), the population (approximately) stays at stable population size. Below that confusion rate, it basically drives itself to extinction. As the confusion rate goes higher, you can observe the exponential growth of the population. This is of course an artifact of not implementing a carrying capacity in the code, and the \(r\) value that would change the end result depends on all the other parameters in the system, but you get the point. In the end, looks like confusing males to a certain extend works for the greater good of the species... :)
[1] Palombit Ryne A. Infanticide as Sexual Conflict: Coevolution of Male Strategies and Female Counterstrategies.
[2] Dawkins R., Krebs J. R. Arms races between and within species.
(more references for general reading is provided in the github repository.)
Maybe a quick pause and some background information can be useful at this point. It's obvious that females can't get pregnant when they are already pregnant, but there is also a phase called lactation after birth, which is colloquially known as breastfeeding. During this phase, the female cannot start ovulating, and therefore cannot get pregnant again (called Lactational Amenorrhea). So dealing with a new born is not an obstacle only in terms of energy and time, but also an obstacle in terms of physiology.
Now it becomes easy to connect the dots : "the female that can potentially give birth to my kids, and pass on the copies of my genes, is dealing with some other guy's new born, and therefore I cannot get her pregnant", says the inner voice of the male. The solution becomes obvious : kill the new born, make the female available again for mating. Basically, it is a male "trick" to obtain sexual access to females who are otherwise preoccupied with maternal duties [1].
Sure, but what about the female? All that effort, energy, and nutrition put in the long months of pregnancy, and this new guy just comes and murders your new born? This costs a lot to the female biologically, and that kid was also her chance of passing the copies of her own genes! So as males came up with their own strategy, females also found a way to get around, which Sapolsky refers as the "one of the great all time ways to manipulate males".
The female solution is called Pseudo-Estrus. It means that the females act like they are ovulating, and ready to get pregnant with your baby, although they are already pregnant with some other guy's kid! They basically send physiological signals to males, such as swellings around their sexual body parts, and present themselves as they are ovulating. The male is happy, goes and mates with the female (who, don't forget, is already pregnant), and when the baby is out there - probably out there earlier than it is supposed to because she got pregnant with it way before - the male still thinks it's his baby, and does not get aggressive. Genius, right?! But at this point, another question rises : sure, males evolved the strategy of infanticide, females counteracted with Pseudo-Estrus, and then what? In fact, this also came up during the lecture, and one student asked : Why doesn't evolution favor the males that can figure out whether the baby is their theirs or not? You can watch the part where he explains the phenomenon of Pseudo-Estrus, and answers the question of that student below.
It is for you to judge, but with all due respect, Sapolsky gives an answer that was not very satisfactory to me. I continued watching the lectures (and again, they are amazing), but the question of why evolution stopped counteracting at that point bothered me a lot. You can see the conceptual consequences of such a question. Does coevolution have a limit? Is there an end to species (or sexes) developing counter strategies to pass on their genes, and converge to point where one of them falls for the other one's trick? If so, what determines this point of convergence?
Interestingly, some birds relieved me from my confusion :) I realized the connection while reading this amazing paper called "Arms races between and within species" by Dawkins and Krebs [2]. Apparently, this family of birds called Cuckoos has a very interesting reproductive strategy called brood parasitism. They mimic the eggs of the host they want to manipulate and leave their eggs to the host's nest so that the host raises the eggs like its own, while they fly on, enjoy life, and save the energy to reproduce more. You can see the photo (from Wiki article) of four clutches of reed warbler eggs, each containing one (larger) cuckoo egg below. They seem quite good at mimicry!
The paper goes on discussing the following paradox : why cuckoo hosts sometimes seem so clever in discriminating against poorly mimetic eggs, and at other times seem so stupid in feeding young clearly different from their own? Put in simpler words, the hosts (which cuckoos are manipulating) seem like they are quite good at discriminating the eggs, if they are mimicked even in a slightly poor way. But if they can't and the cuckoo bird makes it out of the egg, then the host acts like as if it's almost blind, and raises the bird anyway, although it looks completely different then it's own kids!
According to the authors, solution to this paradox lies in the combination of two concepts : evolutionary arms race and life-dinner principle. First concept, the arms race, means that there will always be counter-adaptation to manipulation, until... it becomes more costly for one side to go on adapting. This is where the life-dinner principle kicks in. One side has to be better, because counteraction might mean life to that species, whereas it might only mean dinner to the other. As in the case with foxes and rabbits : The rabbit runs faster than the fox, because the rabbit is running for its life, while the fox is only running for its dinner. For the case of cuckoos, I will directly quote the excerpt in the paper :
There are a few good points in this excerpt to apply to our initial question of why males fall for Pseudo-Estrus. First one is the asymmetrical potentials for evolution. As Sapolsky also vaguely mentions at the end of his answer, it is too much to expect such a cognitive ability to evolve in a primate system. So as hosts being good, but cuckoos being better, males are good, but females had to be better. Why had to? Because of the life-dinner principle. Maybe not exactly a life vs dinner situation at this point, but more generally, unequal selection pressures acting on males and females. A male can go and mate god knows how many females, and pass on his genes quite frequently, but in comparison it is extremely costly for a female to go through the whole period of pregnancy.It may be believable that an individual foster-parent could be hypnotized, addicted, or otherwise manipulated into feeding a cuckoo nestling which obviously bears no resemblance to one of her own. But surely, it will be said, over evolutionary time natural selection would put a stop to that: selection would favour hosts who have nervous systems resistant to the manipulation. Yes, but exactly the same thing could in theory be said of mimetic deception: mimicry cannot possibly evolve, because selection will always favour individuals who are not fooled by it! This, of course, is where the arms race idea comes in. Any nervous system can be subverted if treated in the right way. Any evolutionary adaptation of the host nervous system to resist manipulation by cuckoo nestlings simply lays itself open to counter-adaptation by the cuckoos. Selection acting on cuckoos will work to find whatever chinks there may be in the hosts' newly evolved psychological armour. Host birds may be very good at resisting psychological manipulation, but cuckoos might become even better at manipulating.
It does indeed seem easy to imagine that cuckoos might have a built-in advantage in their arms race with their hosts. The 'life-dinner principle' could apply: a cuckoo in the nest has to manipulate its host successfully or it will surely die; its individual foster-parent will benefit somewhat if it resists manipulation, but it still has a good chance of future reproductive success in other years even if it fails to resist this particular cuckoo.
So far so good. But why this coevolution, this arms race stopped at this very particular point? It could have gone one step further by both sides, end at somewhere that the females are still better, right? So in some sense, this stable point of murder vs manipulation should work for both sexes.
It indeed does. Think of the following scenarios : if all males were so clever to avoid the manipulative females, and kill the infants of other males, they could drive their species to extinction. On the other hand, if all the males fall for this Pseudo-Estrus strategy and nobody understands which kid belongs to whom, then no male kills any infants! So between these two extremes, you can imagine that there will be a certain amount of confusion among males that is good for the continuation of the species. A healthy (!) ratio of murder over manipulation...
Since I am obsessed with these evolutionary sweet spots, had the code this one and see what comes out :) There goes the link to the Github project. The code is just a simple MATLAB script, and basically simulates an agent-based system where there is a certain "confusion rate" among males. This confusion rate is equal to the probability of any infant not being murdered by a male within the lactation phase of the mother. If the infant is murdered while the mother is still in the lactation phase, she becomes available for mating earlier than she would have if she continued breastfeeding. If the infant is not murdered, the mother completes the whole lactation phase, and becomes available for mating afterwards. You can find more details about the modelled scenario in the comments of the script. I set the parameters such that it would be consistent with a human population living in the Paleolithic era (life expectancy of 33 years, pregnancy of 40 weeks, etc.) for fun :)
One interesting thing to look at is the behaviour of this system given different confusion rates among males (which is equal to the probability of not killing other males' infants). You can see the population size plots across 4 different time scales below : 5, 10, 20, and 50 years, where the confusion rate is denoted by \(r\).
If the confusion rate is around \(60\%\), the population (approximately) stays at stable population size. Below that confusion rate, it basically drives itself to extinction. As the confusion rate goes higher, you can observe the exponential growth of the population. This is of course an artifact of not implementing a carrying capacity in the code, and the \(r\) value that would change the end result depends on all the other parameters in the system, but you get the point. In the end, looks like confusing males to a certain extend works for the greater good of the species... :)
References
[1] Palombit Ryne A. Infanticide as Sexual Conflict: Coevolution of Male Strategies and Female Counterstrategies.[2] Dawkins R., Krebs J. R. Arms races between and within species.
(more references for general reading is provided in the github repository.)
