6 | CONCLUSIONS AND PERSPECTIVE
Within-population genetic variation, both natural and restored, is crucial for the response to short-term environmental stresses and long-term evolutionary change. Although the levels ofH e are often correlated with fitness (Oostermeijer et al., 1994; Reed & Frankham, 2003; Szulkin et al., 2010), H e of NGV is poorly correlated with heritability (h 2 orH 2) of quantitative traits (AGV). As discussed above, the relationship of H e toh 2 or H 2 is often very weak, while the relationship between F ST andQ ST is comparatively stronger; thus,F ST could be considered a proxy ofQ ST. However, whenever logistically possible, common garden and/or transplant studies are strongly recommended to quantify patterns of adaptive genetic variation and differentiation (de Villemereuil et al., 2016; Sork, 2018; Capblancq et al., 2020). The most comprehensive studies conducted so far are generally those carried out with many commercially important tree species (e.g., eucalypts, oaks, poplars, pines, and spruces), as plants (including propagules) with well-adapted genotypes are used to replant clear-cut areas (Depardieu et al., 2020). Exemplifying this, 14 out of 18 entries for woody perennials (seven genera) used in our analyses belong to the eucalypt-oak-pine-spruce-poplar group (De Kort et al., 2013). More studies on Q STF STcomparisons are needed, particularly on rare woody species and common herbaceous species, to avoid bias in the inferences, as well as to balance entries among the different life forms. Although the differences between herbaceous and woody plants regarding F STwere large in the data reviewed here, similar average values forQ ST were relatively common in the data set used (De Kort et al., 2013). With a larger dataset, one could also expect some generalizations to emerge concerning theQ STF ST relationships regarding life history characteristics and morphological/anatomical traits. Such generalizations could aid conservation managers and practitioners in using neutral F ST estimates to predict approximate Q ST values and aid the conservation and restoration of plant species. Multiple approaches, including molecular markers (NGV), quantitative traits and/or quantitative trait loci coding for traits and contemporary genome-wide association approaches in the context of a common garden experiment, and environmental variation (e.g., designation of climatic zonation) are needed to gain comprehensive insights into conservation of herbs and trees (de Villemereuil et al., 2016; Rodríguez-Quilón et al., 2016; Sork, 2018).