Aubier, Kokko and Joron (2019) showed with a three-locus, two-sex population-genetic model that when male and female mate preferences coevolve on the same ecological trait, genetic drift can trigger cyclic breakdowns of premating isolation ('preference cycling'), transiently raising hybridization rates and destabilizing speciation with gene flow. The result rests on a deliberately minimal model — sympatric ecotypes, free recombination, one choosiness locus per sex — and the authors explicitly flagged the conditions under which it should be re-examined: geographic context of divergence, alternative genetic architectures, multiple concurrent barriers to gene flow, empirical estimation of male courtship reallocation and of how accurately preferences are expressed, and the mapping from mating-system variation to speciation rates. No model or test of the two-sex coevolution addresses these conditions: subsequent theory has addressed geographic structure and genetic architecture only in one-sex choosiness models, and empirical work has quantified neither courtship reallocation nor preference cycling in nature. The problem — establishing whether preference cycling destabilizes reproductive isolation beyond the minimal sympatric model, and whether it operates in nature — remains open.

Background

In speciation with gene flow, two diverging populations must evolve barriers to exchange; the earliest is usually premating isolation, mating discrimination that prevents heterospecific pairings. A common genetic route is a 'magic trait': a trait under divergent ecological selection that also serves as the cue for mate choice, so that assortative mating (like with like) evolves as a by-product. Most theory has treated mate choice as a female trait, with males competing. However, males of many taxa are also choosy about whom they court, and a choosy male can do something a choosy female cannot: he can redirect the courtship effort he saves when he skips an unpreferred female toward preferred females. Aubier, Kokko and Joron (2019) modeled this. In their population-genetic model, an ecological locus under disruptive selection (selection favoring both extreme trait values over intermediates) is the mating cue; separate loci FF and MM control female and male choosiness (the strength of preference for mating with one's own ecotype, so-called homotypic preference). A choosy male reduces courtship of mismatched females to a small fraction ϵm\epsilon_{\mathrm{m}} of the nonchoosy level and can reallocate the saved effort toward preferred females to a degree measured by a parameter α\alpha (with α=1\alpha=1 meaning full reallocation and a strong mating advantage with preferred females); ϵf\epsilon_{\mathrm{f}} plays the analogous role for female acceptance of mismatched males. Because choosiness in either sex can generate assortative mating, the sexes' preferences substitute for one another: male choosiness relaxes the indirect selection maintaining female choosiness, so female choosiness can drift downward, which in turn disfavors male choosiness, producing episodes of random mating before selection rebuilds preferences. This 'preference cycling', initiated by genetic drift (random allele-frequency fluctuations in finite populations) and completed by selection, causes periodic breakdowns of premating isolation and transient bursts of hybridization. The result challenges the classical 'speciation continuum' picture in which barriers to gene flow accumulate monotonically: mutual mate choice can make premating isolation intrinsically unstable, and snapshots of isolating barriers between populations say little about their long-term stability. The authors were explicit that the finding is conditional on the minimal structure of the model — sympatry (one mixing population), two ecotypes, free recombination among the three loci, preference loci expressed in each sex — and identified the extensions and tests needed before the mechanism can be judged robust or empirically real.

Problem Statement

Determine whether the destabilization of reproductive isolation found by Aubier, Kokko & Joron 2019 — preference cycling from the coevolution of male and female mate choice — is robust and empirically real. The single overarching objective is to establish when, where, and with what long-term consequences mutual mate choice destabilizes premating isolation; the requested components jointly serve it: (1) theory — extend the two-sex choosiness coevolution model beyond its minimal sympatric, free-recombination setting to (a) explicit geographic structure of divergence (secondary contact after allopatry, partial range fragmentation, local spatial structure), (b) alternative genetic architectures, namely the 'two-allele' mechanism, in which preference alleles govern discrimination in heterospecific (between-ecotype) encounters specifically rather than setting the overall level of choosiness, and physical linkage among the choosiness loci (dominance variants were already shown in the source not to change the results qualitatively), and (c) systems with multiple concurrent barriers to gene flow (e.g. postzygotic incompatibilities, meaning the reduced survival or fertility of hybrid offspring), asking whether additional barriers can evolve between cycling-induced hybridization episodes fast enough to prevent populations from merging; (2) empirics — quantify, in naturally diverging or hybridizing populations, the male courtship-reallocation extent α\alpha, how accurately preferences are expressed (the residual mismatch rates ϵm\epsilon_{\mathrm{m}} and ϵf\epsilon_{\mathrm{f}}), and the spatial and temporal variation in preference strengths and hybridization rates needed to detect or refute preference cycling; and (3) macroevolution — test whether variation in mating systems across taxa, through its effect on the propensity for male mate choice and preference cycling, predicts patterns of speciation rates or hybridization. The theoretical components establish whether the cycling mechanism survives the conditions that prevail in nature; the empirical components parameterize and test it; the comparative component assesses its macroevolutionary footprint. Each component changes the interpretation of the others, so the robustness question is addressed only by treating all three components together. The theory extensions (a)–(c) are independent rather than conjunctive — resolving any one of them settles only whether cycling survives that particular departure from the minimal model — and the empirical (2) and macroevolutionary (3) components can likewise be resolved separately.

References

  1. Aubier TG, Kokko H, Joron M (2019) Coevolution of male and female mate choice can destabilize reproductive isolation. Nature Communications 10:5122. DOI 10.1038/s41467-019-12860-9. https://doi.org/10.1038/s41467-019-12860-9
  2. Aubier TG, Joron M (2020, v2 2021) Costs of choosiness can promote reproductive isolation in parapatry. bioRxiv preprint. DOI 10.1101/2020.11.12.379461. https://doi.org/10.1101/2020.11.12.379461
  3. Aubier TG, Bürger R, Servedio MR (2023) The effectiveness of pseudomagic traits in promoting premating isolation. Proceedings of the Royal Society B 290:2022.2108. DOI 10.1098/rspb.2022.2108. https://doi.org/10.1098/rspb.2022.2108
  4. Kyogoku D, Yamaguchi R (2023) Males and females contribute differently to the evolution of habitat segregation driven by hybridization. Journal of Evolutionary Biology 36:515-528. DOI 10.1111/jeb.14156. https://doi.org/10.1111/jeb.14156
  5. Servedio MR, Hermisson J (2020) The evolution of partial reproductive isolation as an adaptive optimum. Evolution 74:4-14. DOI 10.1111/evo.13880. https://doi.org/10.1111/evo.13880
  6. Rushworth CA, Wardlaw AM, Ross-Ibarra J, Brandvain Y (2022) Conflict over fertilization underlies the transient evolution of reinforcement. PLoS Biology 20:e3001814. DOI 10.1371/journal.pbio.3001814. https://doi.org/10.1371/journal.pbio.3001814
  7. Qvarnström A, Veen T, Husby A, Ålund M, Weissing FJ (2023) Assortative mating in an ecological context: effects of mate choice errors and relative species abundance on the frequency and asymmetry of hybridization. The American Naturalist 201:125-137. DOI 10.1086/722156. https://doi.org/10.1086/722156
  8. Maisonneuve L, Beneteau T, Joron M, Smadi C, Llaurens V (2021) When do opposites attract? A model uncovering the evolution of disassortative mating. The American Naturalist 198:625-641. DOI 10.1086/716509. https://doi.org/10.1086/716509
  9. Cotto O, Day T (2021) The evolution of age-specific choosiness when mating. Journal of Evolutionary Biology 34:477-485. DOI 10.1111/jeb.13750. https://doi.org/10.1111/jeb.13750
  10. Jablonszky M, Barta K, Krenhardt K, Nagy G, Zsebők S, Garamszegi LZs (2025) Experimental evidence of slight alteration of male Collared Flycatcher songs based on the female audience. Ibis 167:1028-1042. DOI 10.1111/ibi.13410. https://doi.org/10.1111/ibi.13410
  11. Yan L, Huh NJ, Ibañez IV D, Rosenthal MF, Hedin M, Elias DO (2025) Latent preference for red ornamentation drives interspecific mating in nascent jumping spider species (Habronattus americanus group). Proceedings of the Royal Society B 292:2024.2754. DOI 10.1098/rspb.2024.2754. https://doi.org/10.1098/rspb.2024.2754
  12. Mussmann SM, Saltzgiver MJ, Diver TA, Gilbert EI, Farrington MA, Schleicher BJ (2026) Semipermeable species boundaries permit gene flow but prevent a hybrid swarm among sympatric catostomids of a desert ecosystem. Heredity. DOI 10.1038/s41437-026-00867-0. https://doi.org/10.1038/s41437-026-00867-0
  13. Pärssinen V, Hulthén K, Langerhans RB, Nilsson PA (2026) Divergent ecological adaptation in allopatry leads to behavioral isolation through female resistance. BMC Ecology and Evolution 26:23. DOI 10.1186/s12862-026-02504-0. https://doi.org/10.1186/s12862-026-02504-0