Spatial genetic structure, i.e. the non-random spatial distribution of genotypes, arises in populations because of different processes including spatially limited dispersal and selection. Knowledge on the spatial genetic structure of plant populations is important to assess biological parameters such as gene dispersal distances and the potential for local adaptations, as well as for applications in conservation management and breeding. In their work, Niskanen and colleagues demonstrate a multifaceted approach to characterise the spatial genetic structure in two replicate sites of a continuously distributed Scots pine population in South-Eastern Finland. They mapped and assessed the ages of 469 naturally regenerated adults and genotyped them using a SNP array which resulted in 157 325 filtered polymorphic SNPs. Their dataset is remarkably powerful because of the large numbers of both individuals and SNPs genotyped. This made it possible to characterise precisely the decay of genetic relatedness between individuals with spatial distance despite the extensive dispersal capacity of Scots pine through pollen, and ensuing expectations of an almost panmictic population.

The authors’ data analysis was particularly thorough. They demonstrated that two metrics of pairwise relatedness, the genomic relationship matrix (GRM, Yang et al. 2011) and the kinship coefficient (Loiselle et al. 1995) were strongly correlated and produced very similar inference of family relationships: >99% of pairs of individuals were unrelated, and the remainder exhibited 2nd (e.g., half-siblings) to 4th degree relatedness. Pairwise relatedness decayed with spatial distance which resulted in extremely weak but statistically significant spatial genetic structure in both sites, quantified as Sp=0.0005 and Sp=0.0008. These estimates are at least an order of magnitude lower than estimates in the literature obtained in more fragmented populations of the same species or in other conifers. Estimates of the neighbourhood size, the effective number of potentially mating individuals belonging to a within-population neighbourhood (Wright 1946), were relatively large with Nb=1680-3210 despite relatively short gene dispersal distances, σg = 36.5–71.3m, which illustrates the high effective density of the population. 

The authors showed the implications of their findings for selection. The capacity for local adaptation depends on dispersal distances and the strength of the selection coefficient. In the study population, the authors inferred that local adaptation can only occur if environmental heterogeneity occurs over a distance larger than approximately one kilometre (or larger, if considering long-distance dispersal). Interestingly, in Scots pine, no local adaptation has been described on similar geographic scales, in contrast to some other European or Mediterranean conifers (Scotti et al. 2023).

The authors’ results are relevant for the management of conservation and breeding. They showed that related individuals occurred within sites only and that they shared a higher number of rare alleles than unrelated ones. Since rare alleles are enriched in new and recessive deleterious variants, selecting related individuals could have negative consequences in breeding programmes. The authors also showed, in their response to reviewers, that their powerful dataset was not suitable to obtain a robust estimate of effective population size, Ne, based on the linkage disequilibrium method (Do et al. 2014). This illustrated that the estimation of Ne used for genetic indicators supported in international conservation policy (Hoban et al. 2020, CBD 2022) remains challenging in large and continuous populations (see also Santo-del-Blanco et al. 2023, Gargiulo et al. 2024).

References

CBD (2022) Kunming-Montreal Global Biodiversity Framework. https://www.cbd.int/doc/decisions/cop-15/cop-15-dec-04-en.pdf

Do C, Waples RS, Peel D, Macbeth GM, Tillett BJ, Ovenden JR (2014). NeEstimator v2: re-implementation of software for the estimation of contemporary effective population size (Ne ) from genetic data. Molecular Ecology Resources 14: 209–214. https://doi.org/10.1111/1755-0998.12157

Gargiulo R, Decroocq V, González-Martínez SC, Paz-Vinas I, Aury JM, Kupin IL, Plomion C, Schmitt S, Scotti I, Heuertz M (2024) Estimation of contemporary effective population size in plant populations: limitations of genomic datasets. Evolutionary Applications, in press, https://doi.org/10.1101/2023.07.18.549323

Hoban S, Bruford M, D’Urban Jackson J, Lopes-Fernandes M, Heuertz M, Hohenlohe PA, Paz-Vinas I, et al. (2020) Genetic diversity targets and indicators in the CBD post-2020 Global Biodiversity Framework must be improved. Biological Conservation 248: 108654. https://doi.org/10.1016/j.biocon.2020.108654

Loiselle BA, Sork VL, Nason J & Graham C (1995) Spatial genetic structure of a tropical understorey shrub, Psychotria officinalis (Rubiaceae). American Journal of Botany 82: 1420–1425. https://doi.org/10.1002/j.1537-2197.1995.tb12679.x

Santos-del-Blanco L, Olsson S, Budde KB, Grivet D, González-Martínez SC, Alía R, Robledo-Arnuncio JJ (2022). On the feasibility of estimating contemporary effective population size (Ne) for genetic conservation and monitoring of forest trees. Biological Conservation 273: 109704. https://doi.org/10.1016/j.biocon.2022.109704

Scotti I, Lalagüe H, Oddou-Muratorio S, Scotti-Saintagne C, Ruiz Daniels R, Grivet D, et al. (2023) Common microgeographical selection patterns revealed in four European conifers. Molecular Ecology 32: 393-411. https://doi.org/10.1111/mec.16750

Wright S (1946) Isolation by distance under diverse systems of mating. Genetics 31: 39–59. https://doi.org/10.1093/genetics/31.1.39

Yang J, Lee SH, Goddard ME & Visscher PM (2011) GCTA: a tool for genome-wide complex trait analysis. The American Journal of Human Genetics 88: 76–82. https://www.cell.com/ajhg/pdf/S0002-9297(10)00598-7.pdf

Population geneticists frequently use the genetic and genotypic information of a population sample of individuals to make inferences on the reproductive system of a species. The detection of clones, i.e. individuals with the same genotype, can give information on whether there is clonal (vegetative) reproduction in the species. If clonality is detected, population geneticists typically use genotypic richness R, the number of distinct genotypes relative to the sample size, to estimate the rate of clonality c, which can be defined as the proportion of reproductive events that are clonal. Estimating the rate of clonality based on genotypic richness is however problematic because, to date, there is no analytical, nor simulation-based, characterization of this relationship. Furthermore, the effect of sampling on this relationship has never been critically examined.
The paper by Stoeckel, Porro and Arnaud-Haond [1] contributes significantly to the characterization of the relationship between rate of clonality and genetic and genotypic parameters in a population. The authors use an extensive individual-based simulation approach to assess the effects of rate of clonality (fully sexual, fully clonal and a range of intermediate levels of clonality, i.e., partial clonality) on genetic and genotypic parameters, considering variable population size, sample size, and numbers of generations elapsed since population initiation. Based on their simulations, they derive empirical formulae that link for the first time the rate of clonality to the genotypic richness and to the size distribution of clones (genotypic parameters), as well as to the population inbreeding coefficient and to a metric of linkage disequilibrium (genetic parameters). They then use the simulated data to assess the accuracy of their predictions. In a second phase, the authors use a Bayesian supervised learning algorithm to estimate rates of clonality from the simulated data.
The authors show that the relationship between rate of clonality and genotypic richness is not linear: genotypic richness decreases slowly with increasing clonality, a large drop in genotypic richness is only seen for rates of clonality ≥ 0.90. Genetic parameters are only sensitive to high rates of clonality. The practical implications of these results are that genotypic and genetic parameters can complement each other for the estimation of rates of clonality, with genotypic parameters most useful throughout most of the range of clonality values and with genetic parameters complementing them meaningfully at higher values. The most meaningful practical result of the paper is the demonstration of sampling bias on the estimation of genotypic richness. Commonly used population sample sizes in population genetics studies (n ≤ 50) lead to great overestimation of genotypic richness, which consequently leads to a severe underestimation of the rate of clonality in most systems, irrespectively of whether they have reached stationary equilibrium. Only in small populations, these effects are attenuated.
Biologists interested in the estimation of the rate of clonality will find this paper highly useful to design their sampling, and to choose their statistics for inference in a meaningful way. This paper also calls for a careful reappraisal of previously published works that infer rates of clonality from genetic data, and highlights the prime importance of complementary information on species life history data for a correct understanding of partial clonality.

References

[1] Stoeckel, S., Porro, B., and Arnaud-Haond, S. (2019). The discernible and hidden effects of clonality on the genotypic and genetic states of populations: improving our estimation of clonal rates. ArXiv:1902.09365 [q-Bio] v4 peer-reviewed and recommended by Peer Community in Evolutionary Biology. Retrieved from http://arxiv.org/abs/1902.09365v4