Coming together but keeping apart: Genetic distancing in Geum
Walk along many country roads in May and June and you might chance across a hybrid swarm between Water avens (Geum rivale) and Wood avens (G. urbanum) occurring in the ditches and verges that line these roads. Such hybrid swarms are common in Britain and parts of Ireland wherever the two species come together. Though Wood avens tends to prefer freely draining soils and moderate shade, while Water avens normally occurs on wetter soils in shaded or open habitats, the two species frequently co-occur and hybridize in areas where their habitats overlap or are disturbed. In the east of Scotland, Water avens flowers mainly in May and reproduces by outcrossing, while Wood avens flowers mainly in June and predominantly selfs (Ruhsam et al. 2011). The flowering period of hybrids overlaps those of both species, peaking at the end of May and beginning of June.
Athough little was known about the genetic structure of these hybrid swarms until 10 years ago, we now have a much better idea of this, thanks to research conducted at Edinburgh University and the Royal Botanic Garden Edinburgh by Richard Ennos, Pete Hollingsworth and their students and colleagues during the last decade (Ruhsam et al. 2011, 2013; Jordan et al. 2018). Rather surprisingly, these hybrid swarms have been shown to contain adult plants that mainly comprise only four genetic types. These represent the two parents, their F1 hybrid and backcrossed offspring of the F1 to G. rivale. Despite the high fertility of the F1 hybrid, F2s and backcrossed offspring to G. urbanum are absent, even though they are easily produced from artificial crosses. In fact, a wide range of hybrid genotypes, including F2s, are produced in hybrid swarms but fail to survive to adulthood (Ruhsam et al. 2013).
The two species of Geum join a growing number of species pairs known to form hybrid zones or swarms that are composed of adult plants representing both parent species and highly fertile F1s, but few or no other hybrid classes, despite the fact that F2s and backcrossed offspring can be generated in such swarms (Abbott 2017). Why is this so? It is feasible that these later generation hybrid offspring are poorly adapted to the range of microhabitats present in hybrid swarms and are outperformed by parental types, i.e. there is strong ecological selection against them. Alternatively, they may exhibit low intrinsic fitness (poor vigour, survivorship and fertility) resulting from incompatible interactions between the parental genes they have inherited, although there is little supporting evidence for this in Geum (Ruhsam et al. 2011, 2013).
In the case of Geum hybrid swarms, the presence of both fertile F1s and fertile offspring of backrosses to G. rivale, leads to the expectation that genes might move from G. urbanum into G. rivale and that as a result many individuals of G. rivale should be introgressed, i.e. contain G. urbanum genes. However, a population genomic analysis has shown this isn’t true (Jordan et al. 2018). In fact, unexpectedly, some G. urbanum plants were found to contain a fragment of the G. rivale genome that possibly affects flower angle. Given the absence of backcrossing of F1s to G. urbanum, it is thought these introgressed plants are products of selfing by the F1.
Jordan and colleagues further showed that the two species of Geum diverged from each other approximately 2-3 million years ago and that hybridization between them has resulted in very little genetic exchange since that time. It is feasible, therefore, that habitat divergence acts as a very strong isolating barrier between these two species and that ecological selection against hybrids is a potent force in maintaining species identity and preventing their merger wherever they hybridize.
Abbott RJ (2017) Plant speciation across environmental gradients and the occurrence and nature of hybrid zones. Journal of Systematics & Evolution 55: 238-258.
Jordan CY, Lohse K, Turner F, Thomson M, Gharbi K, Ennos RA (2018) Maintaining their genetic distance: Little evidence for introgression between widely hybridizing species of Geum with contrasting mating systems. Molecular Ecology 27: 1214-1228.
Ruhsam M, Hollingsworth PM, Ennos RA (2011) Early evolution in a hybrid swarm between crossing and selfing lineages of Geum. Heredity107: 246-255. View this at https://www.nature.com/articles/hdy20119.pdf
Ruhsam M, Hollingsworth PM, Ennos RA (2013) Patterns of mating, generation of diversity, and fitness of offspring in a Geum hybrid swarm. Evolution 67: 2728-2740.
University of St Andrews
rja ‘at’ st-andrews.ac.uk