Distribution and Ecological Traits of Cotoneaster integerrimus in South Korea

Although the rare plant <i>Cotoneaster integerrimus</i> is distributed across Eurasia, ecological information on its isolated populations at the easternmost range limit in Korea has been entirely lacking. This study was conducted to (1) characterize the environmental characteristics of the habitat of Korean <i>C. integerrimus</i> populations and (2) predict potential habitats via a simple species distribution model (SDM) based on ridge logistic regression and presence-background data, providing a foundation for effective conservation strategies. To this end, we analyzed habitat type, topography, and light conditions through field surveys and combined these data with an SDM fitted to six known occurrences on limestone ridges. Results revealed a clear ecological divergence; the Korean population is biased toward partial shade and north-facing slopes within the forest understory, in contrast to European populations inhabiting open, rocky sites. This distribution pattern is interpreted as a local adaptive strategy that reduces exposure to hot and humid summer conditions. Furthermore, a unique morphological trait not reported in European populations was identified: dense persistent hairs that remain until seed maturity. The SDM analysis showed moderate discrimination (training AUC = 0.784) and indicated that high elevation and ridge topography (Topographical Position Index, TPI) acted as key habitat factors, whereas annual mean temperature was the strongest limiting factor. Mapping the upper decile (top 10%) of predicted suitability within the limestone belt highlighted a small, spatially restricted set of high-elevation ridges as candidate microrefugia and survey priorities. This study suggests that the Korean <i>C. integerrimus</i> population may have undergone local adaptation due to isolation. Furthermore, this population is considered both a Geographical Peripheral Population (GPP) and a glacial relict, and is assessed to be vulnerable to climate change. Given that the SDM is based on only six occurrences and shows variable performance among spatial folds, all spatial predictions and variable effects should be regarded as exploratory and spatially conservative rather than as definitive habitat projections. These findings, therefore, support the urgent need to establish in situ and ex situ conservation strategies that preserve this geographically peripheral population as an irreplaceable component of the species' genetic diversity.