Disentangling mountain isolation from habitat influence in Scots pine: Cytonuclear evidence for drift and seed-limited connectivity

Small, isolated mountain populations are increasingly important for conservation and seed sourcing, yet their genetic distinctiveness may reflect either environmental selection or demographic processes such as drift and restricted connectivity. We tested these alternatives in Central European Scots pine (Pinus sylvestris L.) by comparing mountain populations (MTS) with lowland populations (LOW) and a low-elevation harsh-site reference population in Ojców National Park (OPN). We analysed 684 individuals using 13 mitochondrial markers and 37,095 nuclear SNPs. Mountain populations formed a mosaic of partially isolated genetic units broadly aligned with mountain massifs. Contemporary effective population size (N e ) among MTS ranged from approximately 2 to more than 500, and several small populations had extremely low estimates, indicating strong drift and demographic vulnerability. Nuclear heterozygosity nevertheless remained similar between MTS and LOW , consistent with extensive pollen-mediated gene flow, whereas mitochondrial variation showed sharper and more discontinuous structure in MTS, indicating restricted seed-mediated connectivity. OPN grouped with LOW in both marker systems, indicating that, in this reference system, harsh habitat and fragmentation at low elevation alone did not reproduce the mountain pattern. Conservative genome scans recovered a small, comparison-dependent set of differentiated loci, providing only modest indirect evidence for environment-associated divergence. Management should prioritise the smallest and most isolated populations. Reinforcement should aim to increase diversity using genetically and climatically compatible sources, preferably multiple nearby montane populations, while avoiding unmonitored mixing across strongly differentiated massifs.