We aimed to tackle a common problem in post-fire tree mortality where the number of trees that survived surpasses the number of dead trees. Here, we investigated the factors that affect Korean red pine (Pinus densiflora Siebold & Zucc.) tree mortality following fires and assessed the statistical effects of class-balancing methods when fitting logistic regression models for predicting tree mortality using empirical bootstrapping (B = 100,000). We found that Slope, Aspect, Height, and Crown Ratio potentially impacted tree mortality, whereas the bark scorch index (BSI) and diameter at breast height (DBH) significantly affected tree mortality when fitting a logistic regression with the original dataset. The same variables included in the fitted logistic regression model were observed using the class-balancing regimes. Unlike the imbalanced scenario, lower variabilities of the estimated parameters in the logistic models were found in balanced data. In addition, class-balancing scenarios increased the prediction capabilities, showing reduced root mean squared error (RMSE) and improved model accuracy. However, we observed various levels of effectiveness of the class-balancing scenarios on our post-fire tree mortality data. We still suggest a thorough investigation of the minority class, but class-balancing scenarios, especially oversampling strategies, are appropriate for developing parsimonious models to predict tree mortality following fires.

Understanding the mechanisms of vegetation recovery following severe ecological disturbances is essential for effective forest management. The relative importance of environmental predictors across different disturbance types and temporal phases remains poorly quantified. We utilized a comprehensive nationwide dataset of 194 restoration plots in South Korea to evaluate the factors influencing total vegetation cover. We employed random forest regression models to analyze topographic, soil, and stand structural variables across four disturbance types and two recovery periods. Model performance revealed that short-term recovery and mechanically unstable roadside slopes are highly stochastic and strictly limited by physical terrain features, including elevation and slope. Conversely, the long-term recovery phase and severely degraded environments, such as mined sites, are strongly associated with biological structural traits and soil nutrient dynamics. Partial dependence plots demonstrated severe non-linear threshold mechanisms where total cover exhibited a sharp positive response to canopy height in quarries and an abrupt negative response to high soil carbon-to-nitrogen ratios during long-term recovery. These results indicate that structural canopy facilitation and microbial nutrient immobilization co-develop with advanced community expansion rather than acting as independent drivers. Ecological restoration strategies must therefore transition from immediate physical site stabilization to proactive soil nutrient management and canopy facilitation as succession advances.

We utilized the USDA Forest Service Forest Inventory and Analysis (FIA) database, encompassing 95,602 trees within 3,307 plots throughout most types of forest ecosystems in the contiguous U.S, to investigate the mixed patterns of structural diversity – productivity relationship (SDPR) by analyzing the growth of existing trees under the main and interactive effects of SD with basal area, stand age and annual precipitation with other factors. We also compiled the studies of SDPR published during 1974 - 2022 and compared the mean annual precipitation between studies showing the negative and positive SDPR relationship. Our findings revealed: 1) SDPR was negative at low SD but became positive at high SD; 2) SDPR became more positive with greater basal area, and the effect of basal area on productivity strengthened with SD; 3) SDPR was more positive with stand development; and 4) SDPR was more positive with higher precipitation at low SD but became negative at high SD. The negative SDPR in wetter conditions aligns with published studies showing higher precipitation in negative SDPR cases. Our findings identified the drivers of the mixed patterns of SDPR to enhance the understanding of forest dynamics from a SD perspective.