Cost of sexual reproduction
A Socratic walk-through of the cost of sexual reproduction — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do most species pay to find a mate when a lineage that simply cloned itself would spread twice as fast?
Put an asexual mutant female into a sexual population and count. Both she and her sexual neighbours produce, say, four offspring. Hers are all daughters, all of whom reproduce. Her neighbour's four are, on average, two daughters and two sons, and the sons bear no young. Every generation, the clone's share of the population doubles relative to everyone else. That is not a marginal advantage; it is a rout, and it should be over in a few dozen generations.
It is not over. Most multicellular species reproduce sexually, and many that can clone themselves still stop periodically to have sex. So something is buying that twofold disadvantage back, every generation, in full. What could possibly be worth that much?
Reasoning it through
REASONING #Begin by pinning down what the twofold factor is actually a cost of. Notice that the arithmetic above never mentioned recombination, meiosis or mate-searching. It counted sons. The clone wins because half of her rival's reproductive investment goes into individuals who do not themselves bear offspring — so this is the cost of males, a consequence of anisogamy, not of sex as such. John Maynard Smith made the point sharply in 1978. A hermaphroditic or isogamous sexual species, where every individual bears young, does not pay a twofold cost at all, though it still pays for meiosis, mating and the risk of finding a partner. That already tells us the puzzle is not one puzzle.
It is also worth noticing what sets the size of the bill: the sex ratio itself. A population could in principle produce few males and pay much less — but a stable sex ratio is not chosen for the good of the population, and any departure from equal investment is undone by selection on individual parents. So the twofold cost is not a bad arrangement waiting to be optimised away; it is the standing charge left by an equilibrium that is itself stable. Whatever pays for sex must pay for it at that price.
Now, the answer everyone reaches for first, which is wrong as usually stated: sex produces variation, and variation lets the species adapt. Weismann argued something like this in the 1880s, and it fails for a specific reason — species do not reproduce, individuals do. A benefit that accrues to the lineage over ten thousand generations cannot rescue you from a competitor doubling every generation now. The variation must pay off fast, and to the individuals producing it.
So what can recombination do within a handful of generations? Two things, and they are different.
The first is combining. Suppose one beneficial mutation arises in one individual and another elsewhere. In a clonal population, those two lineages compete, and one of them is simply lost — the second mutation must wait to arise again on the winning background. With recombination, a single descendant can carry both. That is Fisher and Muller's argument from the 1930s. Its mirror image is Muller's ratchet: in a finite asexual population, the class of individuals carrying the fewest damaging mutations will eventually be lost by chance, and nothing can rebuild it, so the load ratchets upward permanently. Sex reconstitutes clean genomes from damaged parents routinely.
The second is being an unpredictable target. Parasites evolve to exploit whatever host genotype is common, and they do it fast because their generations are short. A clone is a large, fixed, well-advertised target; sexual reproduction generates rare gene combinations for which no parasite is currently prepared. This is the Red Queen hypothesis, and the reason it is taken seriously is that it makes checkable predictions about where sex should be maintained. In the New Zealand freshwater snail Potamopyrgus antipodarum, sexual and asexual forms live in the same lakes, and the sexual form is commoner in the habitats where trematode infection is heaviest. In an experiment by Morran and colleagues in 2011, Caenorhabditis elegans populations exposed to a coevolving bacterial pathogen maintained outcrossing, while populations facing a pathogen held evolutionarily still drifted toward self-fertilisation.
Which of these dominates is genuinely unsettled; the honest answer is probably "both, in proportions that differ by organism". A related clue sits in the phylogenies: asexual lineages arise constantly and are almost always evolutionary twigs — recently derived, not deep branches.
The analogy
THE ANALOGY #Think of independent workshops improving a machine. One works alone and must invent every improvement itself, in sequence, and can never undo a flaw already built into its only design. The others swap parts, so an improvement made in one shop can be married to an improvement made in another, and a batch of defective parts can be left out of the next build — at the price of handing half of every good design to a competitor.
workshops choose whom to trade with and what to keep, whereas recombination is blind — it breaks up good combinations exactly as readily as it assembles them, which is why the advantage only appears when something keeps changing which combination is good.
Clarifying the model
THE MODEL #Three refinements. First, the twofold cost is an upper bound, not a constant: where males provision offspring or defend territory, they are not reproductively idle and the cost shrinks accordingly.
Second, the benefit does not require sex every generation. Aphids, rotifers, many fungi and many plants reproduce asexually for long runs and sexually when conditions turn — crowding, stress, the end of a season. That pattern is itself evidence: an occasional shuffle captures most of the recombination benefit while paying the cost rarely, which is close to what a theory of periodic advantage predicts.
Third, and worth being explicit about, this is one of the places where evolutionary biology has a genuinely open problem rather than a settled story with a few loose ends. The mechanisms above are all supported; their relative weight is argued about, and the strongest evidence for any of them is regional and specific rather than universal.
A picture of it
THE PICTURE #How to readThis is a version-control diagram repurposed as a family tree, which is fair because the operations are the same ones. Follow the upper branch, the clone: it can only commit forward, so a useful change and a damaging one both stay locked into the same line for good. Follow the lower path: two lineages each acquire something useful, and the merge — which is what recombination is — puts them in a single descendant, while the following commit shows deleterious variants segregating out into offspring that selection removes. The clone's branch never merges, and that missing operation is the entire benefit being bought at twofold cost.
What became clearer
WHAT CLEARED #The twofold cost is a cost of producing males, not of shuffling genes, and it is set by a sex ratio that is itself stable against change — so it must be paid, not negotiated down. What pays it is that recombination performs two operations no clonal lineage can: joining useful variants that arose in different individuals, and shedding damaging ones that would otherwise ratchet upward forever. Whether the day-to-day pressure making that worthwhile is mutation load or coevolving parasites remains an open question, and the field's best evidence points different ways in different organisms.
Where to go next
ONWARD #- Why the same reasoning predicts female-biased sex allocation in hermaphrodites, where the cost of males largely disappears.
- How Muller's ratchet explains the degeneration of non-recombining chromosomes such as the mammalian Y.
Key terms
TERMS #| Term | What it means |
|---|---|
| Anisogamy | reproduction using two gamete sizes, one large and provisioned, one small and numerous, which is what creates distinct sexes. |
| Muller's ratchet | the irreversible accumulation of deleterious mutations in a finite population without recombination. |
| Fisher-Muller effect | the advantage of recombination in bringing separately arisen beneficial mutations into one genome. |
| Red Queen hypothesis | the proposal that sex is maintained by continual coevolution with parasites, which track common host genotypes. |
Every term the collection defines is gathered in the glossary.