Sinervo et al 2000 - hypothesis that natural selection can favour two female strategies when density cycles. At low density, females producing many smaller progeny are favoured when the intrinsic rate of increase, r, governs population growth.
At peak density, females producing fewer, higher-quality progeny are favoured when the carrying capacity, K, is exceeded and the population crashes.
Decade long fitness studies and game theory demonstrated that 2 throat colour morphs (lizards maybe?) were refined by selection in which the strength of natural selection varied with density.
R strategies produced many eggs, and favoured at low densities.
K strategies produced large eggs and were favoured at high densities.
progeny size should also be under negative frequency-dependent selection in that large progeny will have a survival advantage when rare, but the advantage disappears when they become common.
Intrinsic causes of frequency and density dependent selection promotes an evolutionary game with 2-generation oscillation. every other generation, the cycle alternatively favours large progeny (crash year) or large clutches of small progeny (boom years). so progeny size under negative frequency-dependent selection. (large hatchlings are selected for when rare).
The alternative throat-colour morph in side-blotched lizards therefore may have arisen as a mutation in an endocrine gene, given the effect of hromones on the expression of orange vs yellow throat. So the locus had pleiotropic effecrs on physiology and behaviour.
UNLIKELY that all variation in clutch size and egg mass of morphs is due to pleiotropic effect of a single throat-colour locus. Many loci undoubtedly affect clutch size and egg mass.