Abstract A crop tree management study targeting green ash ( Fraxinus pennsylvanica Marsh.) was established in a 16-year-old, naturally regenerated, mixed-species, bottomland hardwood stand in southwest Tennessee. The goal was to maintain or improve ascendance of green ash into the overstory of a developing stand that otherwise might not occur. Three treatments were examined: a complete crown-touching release; the same release treatment along with a one-time, fertilizer application; and a control. Initial diameters, heights and crown size parameters were measured in 1996 and 18 years later. The two release treatments did not differ significantly in diameter and height growth response, but statistically outperformed the control treatment. Similarly, crown class scores did not differ between the two release treatments, but were significantly improved from the control. Both release treatments displayed a greater percentage of crop trees successfully competing for upper canopy status compared to the control. Using initial tree diameter as a reflection of crown size and position of crop trees, diameters of 12.5 to 20 cm were more likely to be a component of the overstory 18 years following crop tree treatments. Crop trees less than 12.5 cm in diameter were likely to remain in the subordinate crown classes. Trees greater than 20 cm in diameter were already in the overstory and maintained their upper canopy position. These results suggest that crop tree management can be an effective management tool for improving growth as well as improving overstory presence of green ash in pole-sized, mixed-species, bottomland hardwood stands.

The online version contains supplementary material available at 10.1007/s44391-026-00057-9.

Abstract Accurate prediction of forest growth and yield is essential for strategic planning in intensive plantation management. This study evaluates whether unsupervised anomaly detection can be used as a systematic data-quality layer in stand-level growth and yield modeling, after standard consistency checks have been applied. We used a multi-regional continuous forest inventory of hybrid Eucalyptus urophylla × Eucalyptus grandis plantations in Minas Gerais, Brazil (6,553 measurements from 1,749 permanent plots in three regions). Four unsupervised methods, Isolation Forest, One-Class SVM, Local Outlier Factor and a dense autoencoder, were applied within each region to identify multivariate anomalies in stand age, volume, basal area and dominant height. Using Isolation Forest with a contamination rate of 0.10, approximately 10% of records were flagged as anomalous, but this corresponded to the complete removal of 2–4% of plots, while most plots either retained all measurements or lost only a subset of them. We then calibrated nine nonlinear machine learning models to predict stand volume at the second measurement (V₂) under two scenarios: using the full dataset and using the anomaly-filtered dataset. Gradient Boosting and other tree ensembles achieved the best performance in both cases (R 2 ≈ 0.86–0.88, relative RMSE ≈ 13–15%), and differences in accuracy between full and filtered datasets were small (changes in R 2 < 0.03 and in relative RMSE < 1 percentage point). A sensitivity analysis across contamination levels from 0.05 to 0.20 confirmed that a 10% threshold offers a practical compromise, removing a consistent set of multivariate extremes while preserving plot coverage and predictive performance.

Abstract Amur honeysuckle ( Lonicera maackii [Rupr.] Herder) and other invasive shrubs represent a serious threat to the biodiversity and ecosystem services of forests in the eastern United States. Treating invasive shrubs can be costly and time consuming, highlighting the need for efficient techniques to control heavy invasions across large areas of forest. Mulching-heads have shown promise for efficiently controlling multiple species of invasive shrubs, including Amur honeysuckle, potentially aiding the restoration of suppressed herbaceous-layer plant communities. In this study, we examined how mulching intensity (i.e., treatment depth below the soil surface) affects Amur honeysuckle sprouting response, as well as the post-treatment response of the plant community. We found that deeper applications of a mulching-head inhibited the post-treatment regrowth of Amur honeysuckle sprouts. According to NMDS analysis, post-treatment species composition became more similar across plots and shifts in post-treatment axis values were associated with increased species diversity and floristic quality, a measure of species conservation value. Additionally, a canonical correspondence analysis revealed that deeper mulching-head intensities were correlated with increased herbaceous-layer plant species diversity and floristic quality. We conclude that mulching-head treatments are an effective technique for controlling invasive shrubs without degrading native plant communities, and deeper treatments reduce the rate of post-treatment sprouting.

Abstract Forest damage may increase in frequency, extent and severity as a consequence of climate change and global trade. To mitigate the negative consequences of forest damage caused by insects and fungi, pest management strategies need to be established. An important component of such strategies is the assessment of pest population densities to predict subsequent potential damage during outbreaks. In this study we evaluated different inventory strategies for assessing the population density of Scots pine ( Pinus sylvestris ) insect defoliators through cost-plus-loss analysis where the cost of the inventory is added to the expected losses, mainly from incorrect decisions due to inaccurate information. We assessed what inventory strategy minimized the expected cost-plus-loss for differentspatial distribution of pupae populations, forest stand ages and forest areas affected. The results showed that double sampling was a more efficient sampling strategy than a single-phase survey. With the double sampling strategy, a second phase sample was selected if the first phase results were inconclusive regarding whether pest mitigation efforts should be implemented. The average cost-plus-loss varied widely over all simulated defoliator pupae populations and stand ages. For the double sampling strategy, the average cost-plus-loss was between ~ 18.7 thousand SEK and 5.6 million SEK, while the alternative cost-plus-loss ranged between ~ 18.7 thousand SEK and 9.7 million SEK. The results underscored the importance of accurate information for decision-making in defoliator control, and that a double sampling strategy is better than a single sampling scheme for most scenarios evaluated. We found that the least inventory efforts should be used to evaluate defoliator populations in stands close to a scheduled harvest, and more efforts should be allocated to younger standsdue to the higher expected loss from incorrect decisions. In younger stands, salvage logging due to defoliation induced mortality resulted in higher costs.