New methods
As genomic techniques are increasingly accessible, several methods can contribute to the estimation of fitness and characterization of filial generations in wild populations. For example, field-based pedigrees are commonly used in studies of birds and mammals (e.g. Marr et al. 2002, Martinig et al. 2020), and offer a promising perspective for other taxa (e.g. plants, Ellstrand 214). Alternatively, parental line genotyping has been used to identify admixed offspring (e.g. Bull & Sunnucks 2014), and even further discriminate among filial generations (e.g. Fitzpatrick et al. 2016), allowing for fitness comparisons in systems where individual-based datasets are impractical. More recently, new statistical tools have been developed that enable accurate genetic assignment of individuals into demes within genetically structured metapopulation systems (e.g. Kuismin et al. 2020). In combination with other field-based data, classification of individuals into immigrant, residents, and several filial generations is possible (e.g. Saatoglu et al. 2021). Therefore, these methods have the potential to enormously contribute to the feasibility of heterosis studies in natural populations. Furthermore, these methods can be combined with traditional experimental approaches such as controlled crosses and reciprocal transplants, significantly enriching the understanding of ancestry of individuals in cross-sectional studies.
Genomic tools may allow further advancements of the field by enabling field studies to assign continuous hybrid categorizations without explicit knowledge of filial generation categories (e.g. Aase et al. 2022). For instance, longevity and reproductive success were shown to positively correlate with Buerkle’s hybrid index (Buerkle 2005) in the bighorn sheep (Ovis canadensis ; Miller et al. 2012). In turn, although “pure” non-local individuals of the perennial plantArenaria grandiflora had higher fitness than “pure” locals, admixtured individuals with higher proportions of locally-sourced genetic composition presented the highest fitness, suggesting a complex, non-linear, relationship between fitness and Buerkle’s hybrid index (Zavodna et al. 2015). Moreover, genomic methods have been widely used to estimate relationships between fitness and heterozygosity in inbreeding depression studies. Notably, heterozygosity and Buerkle’s hybrid indexes parallel the concepts of hybridization and source indexes introduced by the line-cross theory (Box 1). Therefore, theoretical developments elucidating the relationship between these indexes are crucial, and will open new avenues of research by allowing the application of line-cross theory, as well as its extensions incorporating estimates of local adaptation (Schneemann et al. 2020), to the study of wild populations.
Finally, because selection removes variation from the population, fitness estimates may be positively biased, resulting in apparent positive heterosis, due to the loss of less fit offspring before estimates are recorded. Consequently, fitness observed in further generations will result from a biased sample of the previous generations’ genotypes (e.g. see Thompson et al. 2022a). The application of line-cross theory in studies of wild populations, therefore, requires that future theoretical developments incorporate changes in allele frequencies across generations as to account for the effects of selection when estimating the expected fitness of hybrid offspring.