Cumulative frequency-dependent selective episodes allow for rapid morph cycles and rock-paper-scissors dynamics in species with overlapping generations

Abstract

Rock-paper-scissors (RPS) dynamics, which maintain genetic polymorphisms over time through negative frequency-dependent (FD) selection, can evolve in short-lived species with no generational overlap, where they produce rapid morph frequency cycles. However, most species have overlapping generations and thus, rapid RPS dynamics are thought to require stronger FD selection, the existence of which yet needs to be proved. Here, we experimentally demonstrate that two cumulative selective episodes, FD sexual selection reinforced by FD selection on offspring survival, generate sufficiently strong selection to generate rapid morph frequency cycles in the European common lizard Zootoca vivipara, a multi-annual species with major generational overlap. These findings show that the conditions required for the evolution of RPS games are fulfilled by almost all species exhibiting genetic polymorphisms and suggest that RPS games may be responsible for the maintenance of genetic diversity in a wide range of species.

Keywords: alternative strategies; evolutionary game theory; frequency-dependent selection; polymorphisms; sexual selection.

Figure 1.

Colour morphs, morph frequency cycles and experimental design. (a) In the European common lizard, males exhibit colour morphs that behave like a single putative locus with three alleles (orange, o, yellow, y and white, w), producing six distinct phenotypes (homozygotes (oo, yy and ww) and heterozygotes (yo, wy and wo)). (b) Schematic showing 3-year RPS cycle of the common lizard. Morph frequencies of the mature cohort (black dots, t_x) exhibit 3–4 year cycles. The maturing cohort's colour frequency is depicted with grey dots (t″_x). The morph frequency in the subsequent year (t_{x + 1}) results from the newly mature cohort (t″_x) plus the surviving adults from the previous year (i.e. the surviving t_x adults). (c) Mature and maturing male cohorts were introduced into six experimental populations. Adult colour frequency (08) did not differ among RPS phases (see text). In three populations, we simulated an advancement of the morph frequency cycle towards w (white dots) by releasing maturing lizards (08″) with a w frequency bias and, in the remaining populations, advancement towards o by releasing maturing lizards with o bias (orange dots). Dashed arrows indicate the frequency difference among mature and maturing cohorts, and black arrows indicate the annual change in colour morph frequency of the mature cohorts from the time of their release (t_x; 08) until the following summer (t_{x + 1}; 09, when the invading cohort completed maturation). Grey shaded areas depict the standard errors of the experimental colour morph frequencies.

Figure 2.

Siring success of male morphs with respect to RPS phase. Morph siring success (number of females with which a male morph sired offspring) as a function of male colour morph for (a) populations cycling towards high o frequency and (b) populations cycling towards high w frequency. o and y scores represent the number of putative colour alleles (note that in autumn 2008 no yy males were found in the field). Dots represent observed values and dot size is proportional to the number of observed values. Surfaces represent predictions generated using the statistical model.

Figure 3.

Progeny survival of colour morphs with respect to siring success. Survival of offspring sired by the male morph with the highest siring success (wy males in populations cycling towards o and oo males in populations cycling towards w; figure 2) and of offspring sired by the remaining morphs. Stars and associated error bars indicate mean ± s.e. per group. The survival probabilities for each of the two morph groups and for each of the six populations are shown. Lines connect survival probabilities measured in the same population. Populations cycling towards high o frequency are depicted with orange dots, and those cycling towards high w frequency with white dots.