Measured Games · Paper 18 of 07 · Coevolutionary biology

Brood parasitism

After R. Dawkins and J. R. Krebs (1979); acceptance thresholds from N. B. Davies, M. de L. Brooke and A. Kacelnik (1996)

Abstract

A parasite that lays in another bird's nest and a host that can learn to spot it are locked in reciprocal selection: better mimicry favours more discriminating hosts, and more discriminating hosts favour better mimicry. The host's problem is one of signal detection rather than of incentives, because own eggs and parasite eggs produce overlapping distributions of appearance and any threshold rejects some of both. Where rejection errors are costly and parasitism is rare, accepting a good mimic is the better decision, so the arms race can rest rather than escalate without limit.

Keywordscoevolutionary arms race · signal detection · acceptance threshold · egg mimicry · field experiment

1Setup

Setup

A cuckoo lays in a reed warbler's nest. If the warbler spots the intruding egg and throws it out, the cuckoo has wasted the attempt. If it does not, the cuckoo chick hatches, evicts the warbler's own brood, and the warbler raises it at the cost of everything it had that season.

Selection therefore favours warblers that discriminate, and discriminating warblers favour cuckoos whose eggs look more like warbler eggs. That reciprocal pressure is the arms race. What makes it a decision problem rather than a story is the warbler's side: own eggs and cuckoo eggs do not sort cleanly. They produce overlapping distributions of appearance, so any rule for rejecting will sometimes throw out one of the warbler's own.

2The model

The model

# the host sees a difference d between an egg and the rest of the clutch own egg d ~ Normal(0, σ) cuckoo egg d ~ Normal(μ, σ) μ falls as mimicry improves # the host rejects anything above a threshold t false rejection α(t) = P(d > t | own) missed parasite β(t) = P(d ≤ t | cuckoo) # with parasitism rate p, and costs of each error E[cost](t) = (1-p)·α(t)·Cown + p·β(t)·Cbrood reject only where p · Cbrood · fcuckoo(d) > (1-p) · Cown · fown(d)

The threshold that minimises expected cost moves with the parasitism rate. Where cuckoos are common, hosts should reject aggressively and accept the occasional loss of their own egg. Where cuckoos are rare and mimicry is good, the arithmetic reverses and accepting is correct. Acceptance is therefore not always a failure to discriminate. It can be the right answer.

Figure 1Acceptance thresholds under overlapping distributions

This figure needs JavaScript. The panels below are live simulations rather than images, so with scripting turned off they stay empty and the controls do nothing. Everything the figure demonstrates is also stated in the surrounding text: the model is given in full in section 2 and the result in section 3.

Appearance of own and cuckoo eggs · the threshold and the two error types

Expected cost against threshold · hover to read any rule

Arms race · mimicry and host discrimination coevolving

Own eggs Cuckoo eggs Optimal threshold Mimicry

Charts are focusable. With a chart focused, use the left and right arrow keys to read values along the horizontal axis.

Parasitism rate

Mimicry, lower is better

Recognition noise σ

Cost of losing the brood

Optimal threshold—

Rejects own eggs—

Accepts cuckoo eggs—

Best rule—

Mimicry—

Generation0

Figure 1. The threshold is chosen by minimising expected cost over the two overlapping distributions, so it is computed from the signal-detection problem rather than assumed. Drop the parasitism rate towards zero and watch the optimal rule flip from rejecting to accepting even though the host's eyesight has not changed. Run arms race lets mimicry improve in response to host rejection and the host threshold respond in turn, which either settles or keeps escalating depending on where the costs sit.

3Results

Results

Dawkins and Krebs (1979) framed parasite and host as an arms race, and the framing carries an asymmetry worth naming: the cuckoo is running for its dinner and the host for its life, so selection does not press equally on both sides. Davies and Brooke turned the question into experiments, placing model eggs of varying similarity into nests and recording what hosts did.

Davies, Brooke and Kacelnik (1996) supplied the decision model. A host cannot see whether an egg is a cuckoo's; it sees a degree of oddity, and must set a rule. Because recognition is error-prone, the optimal rule depends on how often parasitism actually happens and how much each kind of mistake costs. Their conclusion was that reed warblers should accept mimetic eggs when parasitism rates are low, and that observed acceptance need not indicate an inability to discriminate.

That produces a genuine resting point. The starting state is no rejection and no mimicry; hosts evolve rejection, cuckoos evolve a mimetic egg, and it can then pay hosts to accept it provided rejection is costly enough and parasitism rare enough. The arms race can stop, and where it does the equilibrium looks from the outside like host incompetence.

4Evidence

Evidence

Davies and Brooke's fen experiments in the mid 1980s are the foundation, and the subsequent literature has tested the same ideas across many host species. The robust result is that hosts are more likely to accept mimetic eggs and to reject poorly matching ones, which is the minimum the model requires.

The comparative pattern is stronger evidence than any single population. Host species with a long history of parasitism, or under heavier parasitism pressure, show better egg recognition and higher rejection rates. Populations with no history of parasitism show little or no recognition ability at all, which is what a coevolutionary account predicts and a fixed-trait account does not.

The system also produces defences beyond rejection. Hosts increase variation in egg appearance between clutches, which acts as a signature that makes forgery harder, and reduce variation within a clutch, which makes an odd egg stand out. Both are strategies operating on the signal itself rather than on the decision rule, and both are documented in the field.

5Limitations

Limitations

The main interpretive difficulty is well known in the field. Where a host accepts parasitic eggs there are two explanations: evolutionary lag, meaning coevolution has not had time to produce recognition, or evolutionary equilibrium, meaning acceptance is currently the better decision. The lag hypothesis has been criticised precisely because it makes no strong predictions and is very hard to falsify, which means acceptance on its own cannot settle the question.

The model above also treats mimicry as a single number and appearance as one dimension. Real egg matching involves colour, spotting, pattern and ultraviolet reflectance, and hosts weigh them differently, so a one-dimensional threshold is a considerable simplification of what a warbler is actually doing.

Finally, defences at the chick stage are largely absent even in hosts that discriminate well at the egg stage, which the simple arms-race picture does not explain. Work on fairy-wrens and bronze-cuckoos has found nest desertion in response to a lone parasite chick, so the picture is still developing.

References

Brooke, M. de L., & Davies, N. B. (1988). Egg mimicry by cuckoos Cuculus canorus in relation to discrimination by hosts. Nature, 335(6191), 630–632. https://doi.org/10.1038/335630a0
Davies, N. B., & Brooke, M. de L. (1989). An experimental study of co-evolution between the cuckoo, Cuculus canorus, and its hosts. I. Host egg discrimination. Journal of Animal Ecology, 58(1), 207–224. https://doi.org/10.2307/4995
Davies, N. B., Brooke, M. de L., & Kacelnik, A. (1996). Recognition errors and probability of parasitism determine whether reed warblers should accept or reject mimetic cuckoo eggs. Proceedings of the Royal Society B: Biological Sciences, 263(1372), 925–931. https://doi.org/10.1098/rspb.1996.0137
Dawkins, R., & Krebs, J. R. (1979). Arms races between and within species. Proceedings of the Royal Society B: Biological Sciences, 205(1161), 489–511. https://doi.org/10.1098/rspb.1979.0081
Kilner, R. M., & Langmore, N. E. (2011). Cuckoos versus hosts in insects and birds: Adaptations, counter-adaptations and outcomes. Biological Reviews, 86(4), 836–852. https://doi.org/10.1111/j.1469-185X.2010.00173.x
Langmore, N. E., Hunt, S., & Kilner, R. M. (2003). Escalation of a coevolutionary arms race through host rejection of brood parasitic young. Nature, 422(6928), 157–160. https://doi.org/10.1038/nature01460

Verification Every reference on this page has been checked against Crossref, so the works cited are real and correctly identified. The claims this page makes about what those sources report have not yet been independently checked. Of the nine papers audited so far, eight required a correction, so treat the specifics here as unconfirmed.