This study proposes a strategy for selecting materials for Abies nephrolepis, a key subalpine species, to restore damaged forests in Mt. Gariwangsan while considering genetic diversity. The genetic diversity distribution of A. nephrolepis in Korea was evaluated using nuclear simple-sequence repeat (nSSR) markers and the genetic differences between populations were analyzed. A suitable genetic restoration population was selected and a strategy for selecting restoration materials was presented. The genetic diversity indices of 326 individuals across 10 populations, and the genetic relationships between populations, were analyzed using AMOVA, principal coordinate analysis, UPGMA, and STRUCTURE analysis. The genetic diversity of A. nephrolepis populations averaged 0.791, indicating relatively high genetic diversity. The results showed that the genetic differentiation pattern of A. nephrolepis was divided into two groups: Mt. Jungwangsan and other populations. Therefore, genetically diverse restoration materials can be secured from the Mt. Gariwangsan population to restore damaged areas. We found that individuals distributed within approximately 10 m were autocorrelated due to their high genetic similarity, whereas they were randomly distributed in the range of 15 m or more. The optimal sampling strategy involved collecting seeds from a minimum of 25 individuals within 10 m. This study suggests that genetic diversity in restoration material selection is necessary to enhance adaptability to environmental changes.

Taxus cuspidata is a threatened subalpine conifer in South Korea, necessitating evidence-based restoration strategies to counter the impacts of climate change. In this study, we assessed 13 natural populations using 15 polymorphic nuclear simple sequence repeat (nSSR) markers developed in Taxus species and spatial autocorrelation analysis to provide a scientific foundation for conservation. The results showed an intermediate level of genetic diversity, with the Mt. Gariwangsan population exhibiting higher diversity. This highlights its priority as a source for restoration materials. Bayesian clustering supported four distinct management units. Spatial autocorrelation analysis revealed significant positive genetic structure within approximately 50 m, indicating a localized genetic patch size. Based on these results, we suggest maintaining a minimum 50 m sampling distance during seed collection to avoid collecting closely related individuals and to reduce the risk of genetic homogeneity in restoration materials. Such restoration strategies informed by spatial genetic structure and broader genetic data are critical for enhancing the long-term resilience of T. cuspidata in the face of accelerating environmental shifts.