Fitness of crop-wild sunflower hybrids affected by a range of competitive conditions

Background/Question/Methods

The study of gene flow and introgression between crops and compatible wild relatives deepens our understanding of how crop genetics influence the characteristics and evolution of wild populations.  Although we might expect selection against various crop traits in hybrid zones, in some species or environments particular crop alleles will be advantageous and crop allele introgression is expected.  We explored this process with crop and wild sunflower, which has a high chance of crop allele introgression since crop-wild hybrids can survive and reproduce.  Fitness of hybrids relative to the wilds, however, can vary across environmental conditions.  For example, F₁ crop-wild sunflower hybrids better tolerate competition than wild plants, potentially increasing the likelihood of crop allele introgression under greater competition.  Since biotic factors can affect the competitive environment in several ways, we investigated how variation in seed density, frequency of putatively more competitive crop-wild hybrids, and presence of interspecific competitors influenced the fitness of three sunflower crop-wild hybrid crosstypes (F₁, F₂, and BC) and their wild counterpart under experimental conditions at a field site in northeastern Kansas.  Seeds of each crosstype were planted into a replicated split-plot design in the fall and fitness components (survival and reproduction) were monitored the following year.

Results/Conclusions

Crosstypes differed in the probability that a planted seed emerged in the spring, but emerged seedlings of all crosstypes had similar and high probabilities of producing a mature seed head, though probabilities declined at higher densities.  The number of heads a plant produced varied by crosstype, but only under low density (wild and BC plants outperformed F₁ and F₂ plants).  Hybrid sunflowers with low quality heads (i.e., with herbivore presence) that experienced more competition produced the fewest seeds per head, while wild-produced, high quality heads on plants with little competition had the fewest.  Regarding overall fecundity, under low density, the estimated seed production of wild plants was highest, followed by that of BC and F₂ plants.  Under higher densities, the fecundity BC plants fell relative to wild and F₂ plants.  Throughout, F1 plants had the lowest fecundity.  Simulating seed production without seed predators increased seed production, but did not alter relative fitness.  Thus, under these realistically dense field conditions, the wild plants outperformed all crop-wild hybrids except the F₂s, even at high density.  In general the greater the crop contribution (BC>F₁) the lower the fitness with the exception that F₂s, with their larger seedlings, did better than expected.