Measuring recent effective gene flow among large populations in Pinus sylvestris: Local pollen shedding does not preclude substantial long-distance pollen immigration

Abstract

The estimation of recent gene flow rates among vast and often weakly genetically differentiated tree populations remains a great challenge. Yet, empirical information would help understanding the interaction between gene flow and local adaptation in present-day non-equilibrium forests. We investigate here recent gene flow rates between two large native Scots pine (Pinus sylvestris L.) populations in central Iberian Peninsula (Spain), which grow on contrasting edaphic conditions six kilometers apart from each other and show substantial quantitative trait divergence in common garden experiments. Using a sample of 1,200 adult and offspring chloroplast-microsatellite haplotypes and a Bayesian inference model, we estimated substantial male gametic gene flow rates (8 and 21%) between the two natural populations, and even greater estimated immigration rates (42 and 64%) from nearby plantations into the two natural populations. Our results suggest that local pollen shedding within large tree populations does not preclude long-distance pollen immigration from large external sources, supporting the role of gene flow as a homogenizing evolutionary force contributing to low molecular genetic differentiation among populations of widely distributed wind-pollinated species. Our results also indicate the high potential for reproductive connectivity in large fragmented populations of wind-pollinated trees, and draw attention to a potential scenario of adaptive genetic divergence in quantitative traits under high gene flow.

Microgeographic variation in early fitness traits of Pinus sylvestris from contrasting soils

“…We consider more likely that substrate-driven divergent selection exists, but that gene flow is counteracting it, thus precluding early-stage microgeographic adaptation to substrate in the study populations. Again, we cannot establish whether divergent selection could be weak enough to be overcome by gene exchange, but this hypothesis would be consistent with the high (8−21%) estimates of pollen gene flow at the recruitment stage between the two nearby study populations (Jiménez-Ramírez et al, 2021), and with the fact that in a different scenario with negligible gene flow (i.e., very distant populations) a significant genotype × calciumsupply interaction in seedling growth was found consistent with local adaptation to calcareous soils (Kavvadias and Miller, 1999). Moreover, edaphic adaptation likely involves many different loci and functions (Guggisberg et al, 2018), which would tend to enhance the constraining effects of gene flow on adaptive divergence (Yeaman and Guillaume, 2009).…”

“…On the other hand, it could be hypothesized that variation in the level of pollen immigration from external populations might also help explain the contrasting results of the two studies. While we know that pollen immigration from nearby plantations into the two native provenances has been rather high during recent years (Jiménez‐Ramírez et al, 2021), it might have been lower when seeds were collected by Agúndez et al (1994), at a time when plantations were younger and might had not yet reached full reproductive maturity. We don't believe, however, that pollen immigration from the plantations represents a serious confounding factor, as planted stands are hardly distinguishable genetically from native ones (Jiménez‐Ramírez et al, 2021), most likely because they were established with local seeds, so that the historical pattern of among‐provenance gene exchange has probably not been significantly altered.…”

“…We studied two nearby Mediterranean mountain populations of P. sylvestris in central Spain, which grow less than five km away from each other on contrasting soil types: (1) the Campisábalos population (CAL) on soils developing over Cretaceous calcareous and dolomitic parent rocks (calcareous soils); and (2) the Galve de Sorbe population (SIL) on soils over Ordovician quartzite and slate parent rocks (siliceous soils) (Figure 1). Despite contemporary gene flow estimates at the recruitment stage between the two populations being high (Jiménez-Ramírez et al, 2021), measures of seedling fitness traits in nursery experiments have revealed differences between CAL and SIL of similar magnitude than those observed for pairs of Iberian provenances separated hundreds of kilometers (Agúndez et al, 1994). A subsequent multi-site provenance trial showed sustained genetic divergence among the two provenances in tree survival and growth rates at seven years of age, as well as contrasting genotype × environment interactions (Alía et al, 2001).…”

“…In this study, we looked for evidence of local adaptation to calcareous versus siliceous soil types in Pinus sylvestris L. (Pinaceae), using as study system two nearby Mediterranean populations connected via pollen flow (Jiménez-Ramírez et al, 2021). We conducted a greenhouse experiment where we sowed seeds from both populations on both types of natural local substrate and tested for variation in early seedling performance among open-pollinated maternal families originating from either of the two edaphic provenances.…”

Does the seed fall far from the tree? Weak fine-scale genetic structure in a continuous Scots pine population

“…Our study sites have regenerated naturally after cutting to an unknown, but likely lower than 2000 trees/ha density of seed trees. However, as pollen and seed dispersal from surrounding forests (Jiménez-Ramírez et al . 2021) likely increases the effective density, we used 2000 trees/ha as our starting point, and estimated σ g assuming ratios of effective to census density of 0.25 ( D e = 500 trees/ha) and 0.5 ( D e = 1 000 trees/ha).…”