Sperm competition and the Kamikaze Sperm Hypothesis

This is the very controversial and highly sensational scientific theory that the book Sperm Wars was based upon. It was originally put forward in the early 1990′s by Robin Baker and Mark Bellis, and later adopted by various self-styled Seducers, as a justification for their even more bizarre notions about alpha biology amongst humans. I shall endeavour to recount the theory here, with comments and analysis:

Apparently, reasonably young, healthy, heterosexual couples have sex approximately every three days. Perhaps this is coincidental, but this mating frequency maintains an almost continuous supply of sperm inside the female reproductive tract. And so any extra-pair copulations may then lead to Sperm Competition (unless the female had not been having sexual intercourse with her husband, and she only has one additional partner). However, it is very difficult to say exactly how often humans engage in adulterous relationships, and even more difficult to say how frequently a woman might end up mixing sperm from two or more men. For this to happen, a woman would have to have sex with two different men, within approximately a five day period.

Baker and Bellis report that while intra-pair copulations are divided more or less evenly throughout a woman’s reproductive cycle; by contrast, extra-pair copulations are more frequent when women are most fertile (even if fertilization is thwarted by the use of birth control). Baker and Bellis say “at some time in their lives the majority of males in western society place their sperm in competition with sperm from another male and the majority of females contain live sperm from two or more different males.” further, they estimate that in Great Britain1 4 to 12% of children are conceived by “sperm that has prevailed in competition with sperm from another male.” This would be consistent with standard estimates2 of ‘paternal discrepancy‘ (calculated to be about 10%) amongst human beings generally. And, on the surface, this might suggest an opportunity for sperm competition. Additionally, when almost 4,000 sexually experienced women (having had sex at least 500 times) were surveyed1, 1 in 200 claimed to have had sexual intercourse with two different men within half an hour of each other on at least one occasion; Within 24 hours, the figure jumped to almost 30%. Again, suggesting more opportunities for sperm competition.

Many insects and birds have a ‘last male advantage‘. This means that the last male to copulate with a particular female is more likely to fertilise most of her eggs. In the case of mammals, things are rather less straight forward; and so, it is currently completely unknown (particularly amongst humans) whether the first male, the last male or any other male in-between that mates with a female has any advantage at all.

It is quite possible that amongst human beings the probability of fatherhood may simply be a case of depositing the most sperm into the vagina; and then frequent regular sex with a woman could be seen as ‘topping up‘ her reproductive tract, replacing sperm that has become old or disabled since the last copulation. This would keep a pretty much constant supply of sperm inside the female, and would be adaptive under any circumstances, because it would maximize the chances of fertilization (this would be especially true, if the woman might have sex with other men too). Furthermore, from a sperm competition perspective, masturbation could then be a way of making sure that the sperm available to be ejaculated in sexual intercourse have a long ‘shelf life’: If the body removes older sperm, then whatever is left should be fresh, and the younger sperm would be better able to compete, and especially more able to penetrate cervical mucus.

Baker and Bellis go on to claim that women can (unconsciously) control the outcome of a sperm competition, and we accept that since women evaluate men by many criteria (emotional, intellectual, physical, financial etc.) while searching for the appropriate qualities in a parent, colleague, friend, lover, protector etc. that there is no reason why a woman should not find herself socially paired with one man, but disposed to have extra-pair sex with another man.

Women are quite likely to be concerned about the overall quality of the male who fertilizes her eggs; and the internal reproductive tract3 of a women produces many barriers, including anti-sperm antibodies that can interfere with fertilization by immobilizing, or even destroying sperm and by impairing their ability to penetrate the egg, while other antibodies act against the egg’s membrane to prevent early egg cleavage and development. The key point here is that these antibodies do not necessarily reduce fertility; instead they diminish the fertility of certain male-female pairings. Therefore, a woman may enhance her reproductive success by seeking a different sexual partner, while retaining her social marital partner (all quite possibly unconsciously).

Another behavioural strategy that may be likely is that of ‘flowback‘. Up to one-third of the seminal fluid ejaculated within the vagina leaks out within a few minutes, and semen can also be expelled with substantial force when urinating, as opposed to dribbling out after sex when a woman stands up (or even if she remains laying down). About 12% of the time, this flowback results in the expulsion of virtually all the sperm ejaculated inside the reproductive tract; so, women are certainly capable of exercising some control over sperm.

Baker and Bellis somehow managed to convince a number of women to capture flowback after both extra-pair copulations and intra-pair copulations; the results they found showed a lower level of sperm retention after intercourse with their main partner. According to Baker and Bellis “women achieve higher sperm retention during extra-pair copulations by reducing their frequency of non-copulatory orgasms via masturbation” So, contractions during female orgasm may actually push out semen and thus by masturbating less (and therefore having fewer orgasms) women end up retaining more sperm from extra-pair copulations.

Insects and birds have sperm storage organs, whereas mammals lack any similar organs. However, it has been claimed that sperm are stored in ‘cervical crypts‘ which are tiny cavities lining a woman’s cervix. It is further claimed that from here they could be released over a period of time, after sex. This would be important for human sperm competition, because the ability to store sperm would encourage sperm competition among successive males. The key question (which scientists are still debating) is how long after ejaculation can human sperm remain viable? and thus capable of competing to fertilize a woman’s egg. Estimates vary from two to ten days; so if we were to assume a median value of five to six days to be true, that would mean that if a woman had sex with someone within five or six days of having sex with another man, the sperm of the two men could be in competition.

Ultimately, a ‘Sperm competition‘ begs the question: How does one win? For a woman, this would be by making the best choice (i.e. having the opportunity to choose more than one man), as well as being able to select ‘good’ sperm, possibly by setting up a competitive situation. For a man, winning would be simply by having ones sperm succeed in fertilizing a woman’s egg.

A male strategy may therefore be as simple as making lots of sperm, we could assume that fertilizing the egg was a type of lottery (i.e. own more tickets and you’re more likely to win). Biologists have therefore asked “What is the best strategy for males, in terms of dividing their sperm between intra-pair copulations and extra-pair copulations?”. And Geoffrey Parker concluded from a detailed mathematical model that males with a partner should generally ejaculate more sperm during extra-pair copulations than intra-pair copulations (assuming that the male was able to maintain an adequate sperm level inside his main sexual partner). The only exception being4 when a male has determined that his partner has been unfaithful, in which case he should increase his sperm numbers; although it is still to be proven whether or not this actually occurs. However, fertilization may be less a lottery than a race; in which case it would be important to make sperm that move quickly.

Baker and Bellis in their ‘Kamikaze Sperm Hypothesis‘ have proposed that it is a war, in which men’s sperm literally do battle with each other, and that only a small percentile of human sperm are actually intended to function as ‘egg-getters‘, which fertilize eggs. Whilst the balance are Kamikazes on a suicide mission, whose goal is simply to stop the sperm of other males. In addition to these ‘Blockers‘ (sperm with coiled and kinky tails) there are supposedly others that go on ‘search-and-destroy’ missions. These, it is claimed carry out chemical warfare, via specialized structures known as acrosomes, which are located on the tip of the sperm.

A normal human male produces a wide variety of different types of sperm (amorphous, coil-tailed, crook necked, double headed (bicephalous), pin-sized, short-tailed, weirdly shaped). As an example, the smallest sperm in a single human ejaculate can have 14% of the volume of the largest, and there is more variation within a single human ejaculate than in the mean sizes of sperm from all the different primates put together.

This wide variety of different sperm types has always been thought to be because sperm are so difficult to produce (about 30% of human sperm are acknowledged to be defective in some way). And may be a reason why males produce so many of them; but if natural selection had acted on males to produce ‘egg-getters,’ why should so many sperm be lame, defective, slow or deformed?

Baker and Bellis argue that semen should be seen as another human organ, comparable to the immune system. As such, it would be made up of many different types of highly specialized cells, all of which work together to get an important job done. Amongst rats, sperm from copulation will form a copulatory plug which will get in the way of sperm from the next male; and the sperm which form that plug are those which have smaller heads, and which are more likely to be decapitated. Human sperm does not form copulatory plugs, but there could be an evolutionary advantage for men whose sperm was especially nasty to anyone else’s. Baker and Bellis would suggest that older sperm serve as blockers, guarders, seek-and-destroyers, kamikazes etc.) Further, Baker and Bellis claim that when sperm from two different men are mixed, they could see sperm stuck together in the process of killing each other. They also saw an increase in the proportion of acrosome-reacted sperm (including that they had used their acrosome in the killing process), and they also saw many more dead sperm in mixtures than in single-sperm samples that had been separated and then re-mixed.

This sounds pretty convincing, but other studies from different mammals and birds reported nothing resembling warfare. And, subsequently Harry Moore, working in collaboration with colleagues at a local fertility clinic, reran Baker’s and Bellis’s experiment, but at a much higher level of sophistication. They differentially labelled sperm from the different males to see whether the agglutination Baker and Bellis had reported was between sperm from different males, as it would have to be. But, whatever agglutination there was turned out to be random with respect to the sperm sample; also, mixing sperm from different men caused no induction of the acrosome reaction, and no increase in sperm mortality.5

Some evidence of male sperm competition comes from analysing the detailed makeup of the sperm ejaculate. Human ejaculation takes place in a series of three to nine spurts, and through mammoth efforts, scientists have examined split ejaculates, obtained by capturing a few squirts of semen from various stages of ejaculation. In the experiment, early and late spurts were different. The final squirts actually containing a spermicidal substance which may ambush those sperm of the next male to ejaculate into the same female. At the same time, chemicals6 present in the first half of the ejaculate contribute some protection to sperm in the second half, and possibly against any chemicals deposited by the final spurts of a preceding male.

It is well known that men produce larger ejaculations when their sex lives are interrupted, and then resumed. This is most likely because seminal fluid has had a chance to accumulate over time; and with any emptying, the volume of ejaculation becomes less, so it’s not surprising that Baker and Bellis discovered that when a man spends time away from his female partner, he produces more sperm per ejaculate, once sexual relations have been restored. More interesting is their discovery (from analysing condom contents) that sperm concentration is higher when males engage in sexual intercourse than when they masturbate. Additionally, during sex, they claim that the amount of sperm transferred is adjusted according to the risk of sperm competition, especially how long it has been since the last copulation with the same woman, and even how much time the two have spent together recently. To quote Baker and Bellis1 “Males may not look very sophisticated in the moments leading up to and during ejaculation but… some very sophisticated adjustments are taking place.”

Evidence from primates suggests that human males are less adapted to competing with the sperm of other men, than male chimpanzees, which have to deal with polyandrous, promiscuous females (human beings are less prone to extra-pair copulations than chimps) As evidence, human sperm concentration diminishes more rapidly, with repeated ejaculations. In one intriguing experiment, men engaged in a ‘10-day depletion experience,‘ (averaging 2.4 ejaculations per day). Afterwards, their sperm output7 remained below their earlier pre-depletion levels for more than five months! By comparison, male chimps8 can ejaculate every hour for five hours, after which their sperm count is only halved, and they recover very rapidly. From this we can conclude that whatever the importance of sperm competition in humans, it is likely not as pronounced as it could be.

Biologists, especially Alexander Harcourt, have criticized specific parts of Baker and Bellis’s work more directly, pointing out that mammals are unlikely to produce designated, non-fertilizing sperm for a number of reasons:

  • Given a high natural loss, males may be unable to afford production of sperm that are certain not to be potential fertilizers, and whose supposed aggressive / defensive properties may never actually be called for.
  • Secretions from the accessory gland of males appear sufficient by themselves to coagulate semen and generate copulatory plugs (in other mammals); a male who used these secretions to form a copulatory plug, and continued to produce fertilizing sperm, instead of diluting his ejaculate with Kamikazes would then be at an evolutionary advantage.
  • Males of polyandrous species are under much more intense sperm competition than in monandrous species, in which females mate with only one male. However, a review of research findings shows that polyandrous species do not produce a greater number, or even a higher percentage of deformed (non-fertilizing) sperm. Neither do they produce more slow-swimmers, which you would expect to see if such Kamikazes were designed to stay behind, and fight with a competitor’s sperm.

Harcourt concludes9 that sperm competition (at least in mammals) occurs via what ecologists call ‘scramble competition,’ in which contestants struggle individually towards a goal, irrespective of the competition, as opposed to ‘contest competition,’ in which individual contestants would seek to physically better their opponents.

Other parts of the theories for specialist sperm were severely criticised10, when in a review of Baker and Bellis’s book, Roger Short pointed out that the sperm which had been designated ‘egg-getters’ were about as likely to be egg-getters as they were to contain little men. Twenty years previously, Short and three colleagues had published a paper in the scientific journal Nature, pointing out that these large-headed sperm were in fact production errors carrying twice the normal chromosome complement, and as a consequence they were incapable of producing a normal embryo. Somehow, Baker and Bellis had overlooked this publication.11

The fact that males produce12 such vast quantities of sperm may therefore be because fertilization is a simple ‘raffle,’ rather than a direct competitive struggle. A raffle would still involve sperm competition, but a competition in which the contestants compete by buying as many tickets as possible, rather than by tearing up each other’s entries. Another answer may be that males make lots of sperm simply because, considering their very high mortality rate (even without sperm competition) it benefits men to make as many sperm as possible, if fertilization is to occur at all. The warmth and moisture of the vaginal environment is ideal for bacteria to enter the female body, so the low pH is as harsh on human sperm as it is on other intruders. Phagocytes roam through every woman’s reproductive tract, and many sperm will end up being absorbed into the uterine wall. Sperm have a long way to swim, as well as the fact that about half will swim up the wrong fallopian tube, whilst a fertile egg awaits at the other.

The reproductive adaptations proposed by Baker and Bellis are rather more extreme than seems likely to have occurred by evolution. It is therefore not surprising that, so far, many of Baker’s and Bellis’s results have failed to stand up to scrutiny, and many of the rest of their conjectures are still far from proven.

In summary, the importance of sperm competition may be overblown and the incredible view of sperm competition that Baker and Bellis have perpetuated, whilst catching the public’s imagination, is little more than a sexual fantasy.


  1. Baker, R. R., and M. A. Bellis, Human Sperm competition (London: Chapman & Hall, 1995)
  2. Macintyre, S., and A. Sooman, ‘Nonpaternity and prenatal genetic screening’, Lancet (1992), 338: 839
  3. Kanada, M., T. Daitoh, K. Mori, N. Maeda, K. Hirano, M. Irahara, T. Aono, and T. Mori., ‘Etiological implication of autoantobodies to zona pellucida in human female infertility’, American Journal of Reproductive Immunology (1992), 28: 104-109; Ahmed K., and R. K. Naz., ‘Effects of human antisperm antibodies on development of preimplantation embryos’, Archives of Andrology (1992), 29: 9-20
  4. Parker, G. A., ‘Sperm competition: Sneaks and extra-pair copulations’, Proceedings of the Royal Society of London (1990), B, 242: 127-133
  5. Moore, H. D. M., M. Martin and T. R. Birkhead, ‘No evidence for killer sperm or other selective interactions between human spermatozoa in ejaculates of different males in vitro’, Proceedings of the Royal Society of London (1999), B, 266: 2343-2350
  6. Lindholmer, C., ‘Survival of human sperm in different fractions of split ejaculates’, Fertility and Sterility (1973), 24: 521-526
  7. Freund, M., ‘Effects of frequency of emission on semen output and an estimate of daily sperm production in man’, Journal of Reproduction and Fertility (1963), 6: 269-286
  8. Marson, J., D. Gervais, S. Meuris, R. W. Cooper, and P. Jouannet, ‘Influence of ejaculation frequency of semen characteristics in chimpanzees’, Journal of Reproduction and Fertility (1989), 85: 43-50
  9. Harcourt, A. H., ‘Sperm competition and the evolution of nonfertilizing sperm in mammals’, Evolution (1991), 45: 314-328
  10. Birkhead, T. R., H. D. Moore, and J. M. Bedford, ‘Sex, science and sensationalism’, Trends in Ecology and Evolution (1997), 12: 121-2
  11. Short, R. V., ‘Review of R. R. Baker and M. A. Bellis, Human Sperm competition: Copulation, Masturbation and Infidelity’, European Sociobiological Society (1997), 47: 20-23; Seuanez, H. N., A. D. Carothers, D. E. Martin, and R. V. Short, ‘Morphological abnormalities in spermatozoa of man and great apes’, Nature (1977), 270: 345-7
  12. Parker, G. A., ‘Why are there so many tiny sperm? Sperm competition and the maintenance of two sexes’, Journal of Theoretical Biology (1982), 96: 281-294