Answers are grounded in this paper and cite their sources.
Investigates heavy metal levels in Pinus nigra across various sites, indicating a reliable biomonitoring method.
Heavy metals are persistent environmental contaminants that accumulate in soils and vegetation, posing significant risks to ecological systems and human health. Pinus nigra needles are widely recognized as effective biomonitors for reflecting spatial and temporal variations in atmospheric heavy metal deposition. However, the complex, nonlinear interactions among multiple pollutants, environmental factors, and site-specific conditions limit the effectiveness of conventional statistical approaches in accurately modeling and predicting contamination patterns. This study investigated the spatial and seasonal distribution of heavy metals in soils and Pinus nigra needles across different environmental settings in Adıyaman, Türkiye, including urban traffic zones, an organized industrial area, a cement factory vicinity, and a clean reference site. Metal concentrations were determined using inductively coupled plasma mass spectrometry (ICP–MS) following standardized acid digestion procedures. To address the limitations of traditional methods and capture complex nonlinear relationships, advanced machine learning (ML) algorithms—multilayer perceptron, Random Forest, XGBoost, LightGBM, CatBoost, and Gradient Boosting—were employed to model elevation based on heavy metal concentrations. The dataset was divided into training (80%) and testing (20%) subsets, and model performance was evaluated using R2, RMSE, MAE, MAPE, and EVS. Among the models, XGBoost exhibited superior predictive performance. Excluding Cd, Cr, and Cu, it achieved R2 = 0.9996 (RMSE = 0.068) in training and R2 = 0.9526 (RMSE = 17.77) in testing. Including these metals further improved performance to R2 = 0.9999 (RMSE = 0.054) for training and R2 = 0.9890 (RMSE = 5.55) for testing. The results confirm that Pinus nigra needles are reliable bioindicators of heavy metal accumulation. More importantly, the integration of biomonitoring data with ML techniques provides a powerful framework for capturing complex environmental interactions and improving predictive accuracy, thereby supporting more effective environmental monitoring, risk assessment, and sustainable management strategies.
No takes yet. Share an insight, caveat, or question.
Dere et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: