تأثیر تغییر اقلیم بر سلامت خاک: تحلیل واکنش میکروبیوم به گرمایش با بهره گیری از مدل های یادگیری ماشین

نوع مقاله : مقاله پژوهشی

نویسندگان

1 گروه علوم خاک دانشکده کشاورزی دانشگاه فردوسی مشهد مشهد ایران

2 عضو هیات علمی بخش پایش و بهسازی منابع خاک و آب موسسه تحقیقات خاک و آب کشور کرج ایران

3 گروه کامپیوتر، جهاد دانشگاهی خراسان رضوی، مشهد ایران

10.22059/ijswr.2025.397431.669962

چکیده

تغییر اقلیم اثرات گسترده‌ای بر عملکرد خاک دارد؛ به‌ویژه از طریق حساسیت تنفس میکروبی به دما که با ضریب حساسیت دمایی (Q10)  سنجیده می‌شود. این پژوهش با هدف شناسایی عوامل مؤثر بر این ضریب و ارزیابی کارایی الگوریتم‌های یادگیری ماشین انجام شد.در این مطالعه از داده‌های مربوط به ۳۳۲ نمونه خاک از ۲۹ کشور جهان استفاده شد و ضریب حساسیت دمایی در سه سطح کم، متوسط و زیاد طبقه‌بندی گردید. شش الگوریتم شامل جنگل تصادفی، ماشین بردار پشتیبان، نزدیک‌ترین همسایگی، رگرسیون لجستیک، تقویت گرادیانی و شبکه عصبی چندلایه به‌کار گرفته شدند. عملکرد مدل‌ها با شاخص‌های دقت، سطح زیر منحنی مشخصه عملکرد گیرنده (ROC) و سطح زیر منحنی دقت–بازخوانی (PRC) مقایسه شد. نتایج نشان داد جنگل تصادفی بالاترین کارایی را داشته است؛ به‌گونه‌ای که با دقت /0 ± 1170/0، مقدارROC  برابر با 8150/0 ± 080/0 و مقدار PRC برابر با 7530/0 ± 0960/0 بر سایر مدل‌ها برتری داشت. برای تفسیر نتایج از روشSHAP  به‌عنوان رویکرد هوش مصنوعی توضیح‌پذیر استفاده شد. این تحلیل نشان داد تنفس ناشی از گلوکز، غنای قارچی و شوری خاک مهم‌ترین عوامل مؤثر بر ضریب حساسیت دمایی هستندد. این یافته‌ها با ارائه تصویری روشن از نقش ترکیبی ویژگی‌های خاکی و میکروبی، توان الگوریتم‌های یادگیری ماشین و روش‌های توضیح‌پذیر را در تحلیل پاسخ خاک به گرمایش جهانی نشان می‌دهند.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

The Impact of Climate Change on Soil Health: Analyzing Microbiome Response to Warming Using Machine Learning Models

نویسندگان [English]

  • Reyhaneh KeshikNevisRazavi 1
  • Elham Farahani 2
  • Narges Abedinzadeh 1
  • Mohammad Abdolahi 3
1 Department of Soil Science, Faculty of Agriculture, Ferdowsi University of Mashhad Mashhad Iran
2 Assistant professor, Soil Reclamation and Sustainable Land Management Department,, Soil and Water Research Institute of Iran, Karaj, Iran
3 Computer Department, Jihad Daneshgahi, of Khorasan Razavi, Mashhad Iran
چکیده [English]

Climate change has extensive effects on soil functioning, particularly through the temperature sensitivity of microbial respiration, which is measured by the temperature sensitivity coefficient (Q10). This study aimed to identify the factors influencing this coefficient and to evaluate the performance of machine learning algorithms. Data from 332 soil samples collected across 29 countries were used, and the temperature sensitivity coefficient was classified into three levels: low, medium, and high. Six algorithms—including Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, Gradient Boosting, and Multilayer Perceptron—were applied. Model performance was compared using accuracy, the area under the Receiver Operating Characteristic curve (ROC), and the area under the Precision–Recall curve (PRC). The results indicated that Random Forest had the highest performance, achieving an accuracy of 0.1170 ± 0, ROC of 0.8150 ± 0.080, and PRC of 0.7530 ± 0.096, outperforming the other models. The SHAP method was employed for interpreting the results as an explainable AI approach. This analysis revealed that glucose-induced respiration, fungal richness, and soil salinity were the most influential factors affecting the temperature sensitivity coefficient. These findings provide a clear picture of the combined role of soil and microbial features and demonstrate the potential of machine learning algorithms and explainable methods in analyzing soil responses to global warming.

کلیدواژه‌ها [English]

  • Climate change
  • Explainable Artificial Intelligence (XAI)
  • Machine learning
  • Soil microbial community

Introduction:

Climate change is having profound effects on soil health and its functioning, particularly by influencing microbial respiration and enhancing carbon release through rising temperatures. The Q10 index, which indicates the temperature sensitivity of respiration, is a crucial metric for understanding soil carbon dynamics in the context of global warming. This study leverages machine learning models and explainable AI (using SHAP) to predict Q10 values based on a combination of biological, chemical, and physical soil properties. By employing a three-level classification system for Q10, the study offers a more nuanced perspective on microbial response and delivers interpretable insights into soil-climate interactions.

Objective(s):

The objective of this study was to predict microbial Q10 temperature sensitivity using various machine learning algorithms, incorporating biological, chemical, and physical soil data. Additionally, the research aims to pinpoint the most influential factors impacting Q10 through interpretable AI methods such as SHAP, thereby improving our understanding of soil carbon behavior in response to climate change.

 

Methods:

This research utilized a dataset containing soil biological, chemical, and physical property data to classify microbial Q10 values into three sensitivity levels. Multiple machine learning models, including Random Forest, XGBoost, SVM, and MLP, were applied, with class imbalance addressed using resampling techniques like SMOTE, SMOTEENN, and ADASYN. Model performance was assessed using precision, recall, and F1-score, leveraging nested cross-validation. SHAP analysis was conducted to interpret feature importance and model decisions.

Results:

Ensemble models, particularly Extra Trees, Random Forest, and XGBoost, outperformed other models in accuracy and AUC. Extra Trees achieved the best results, with an accuracy of 0.653 ± 0.117, AUROC of 0.815 ± 0.080, and AUPRC of 0.753 ± 0.096. Decision Tree and KNN performed the weakest. Feature importance analysis, both embedded and SHAP, highlighted microbial traits such as Bacteria_Positive and Mean_Glucose as key predictors. SHAP plots also revealed threshold effects for chemical features like SOC and Alkane. Statistical tests confirmed the superiority of ensemble models over simpler approaches (p < 0.05).

 

Conclusions

This research employs machine learning and explainable AI to identify critical factors affecting Q10 in soils. Key microbial traits, including Bacteria_Positive and Mean_Glucose, and chemical features like SOC and Alkane, were found to be influential. SHAP analysis provided transparency in model predictions, uncovering complex relationships between soil properties and microbial respiration. These insights can inform soil management practices and climate adaptation strategies. Future research could focus on incorporating physical and hydraulic soil properties to refine Q10 predictions for specific ecological conditions.

Author Contributions

KeshikNevisRazavi, S.R. contributed to the conceptualization, methodology, software development, investigation, data curation, writing of the original draft, and manuscript review and editing. Farahani, E. was involved in validation, writing of the original draft, review and editing of the manuscript, supervision of the research process, and overall project administration. Abedinzadeh, N. participated in software development, data curation, and writing of the original draft. Abdolahi, M. contributed to the methodology, software implementation, validation, formal analysis, investigation, data curation, writing of the original draft, review and editing, visualization of results, and research supervision. All authors have read and approved the final version of the manuscript.

Data Availability Statement

The dataset used in this study is publicly available on:

https://figshare.com/articles/dataset/The_soil_microbiome_governs_the_response_of_microbial_respiration_to_warming_across_the_globe/20776243?file=42356520

Acknowledgements

The authors would like to thank the original authors of the publicly available dataset used in this study. We also appreciate the support provided by our affiliated institutions and colleagues who contributed insightful discussions during the research process.

 

Ethical Considerations

This study is based on secondary data analysis of a publicly available dataset. No human or animal subjects were involved, and therefore no ethical approval was required.

Conflict of Interest

The authors declare that there is no conflict of interest regarding the publication of this paper.

مراجع

Ahrens, B., et al. (2020). Temperature responses of soil respiration are contingent on the microbial community composition. Global Change Biology, 26(7), 3732–3746.
Ahsan, M. M., Mahmud, M. P., Saha, P. K., Gupta, K. D., & Siddique, Z. (2021). Effect of data scaling methods on machine learning algorithms and model performance. Technologies, 9(3), 52.
Allen, D. E., Singh, B. P., & Dalal, R. C. (2011). Soil health indicators under climate change: A review of current knowledge. In B. P. Singh, A. L. Cowie, & K. Y. Chan (Eds.), Soil health and climate change (pp. 25–45). Springer.
Aydın, Y., Işıkdağ, U., Bekdaş, G., Nigdeli, S. M., & Geem, Z. W. (2023). Use of machine learning techniques in soil classification. Sustainability, 15(3), 2374.
Biecek, P., & Burzykowski, T. (2021). Explanatory model analysis: Explore, explain, and examine predictive models. CRC Press.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32.
Cascio, D., Taormina, V., Cipolla, M., Bruno, S., Fauci, F., & Raso, G. (2016). A multi-process system for HEp-2 cells classification based on SVM. Pattern Recognition Letters, 82, 56–63.
Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). ACM.
Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, 20(3), 273–297.
Crowther, T. W., Van den Hoogen, J., Wan, J., Mayes, M. A., Keiser, A., Mo, L., Averill, C., & Maynard, D. S. (2019). The global soil community and its influence on biogeochemistry. Science, 365(6455), eaav0550.
Ferrer, L. (2022). Analysis and comparison of classification metrics. arXiv preprint arXiv:2209.05355.
Geurts, P., Ernst, D., & Wehenkel, L. (2006). Extremely randomized trees. Machine Learning, 63(1), 3–42.
Guo, G., Wang, H., Bell, D., Bi, Y., & Greer, K. (2003). KNN model-based approach in classification. In R. Meersman, Z. Tari, & D. C. Schmidt (Eds.), On the move to meaningful internet systems 2003 (pp. 986–996). Springer.
Haaf, D., Six, J., & Doetterl, S. (2021). Global patterns of geo-ecological controls on the response of soil respiration to warming. Nature Climate Change, 11(7), 623–627.
Ibrahimi, E., Lopes, M. B., Dhamo, X., Simeon, A., Shigdel, R., Hron, K., et al. (2023). Overview of data preprocessing for machine learning applications in human microbiome research. Frontiers in Microbiology, 14, 1250909.
Lal, R. (2011). Soil health and climate change: An overview. In B. P. Singh, A. L. Cowie, & K. Y. Chan (Eds.), Soil health and climate change (pp. 3–24). Springer.
LaValley, M. P. (2008). Logistic regression. Circulation, 117(18), 2395–2399.
Mahecha, M. D., Reichstein, M., Carvalhais, N., Lasslop, G., Lange, H., Seneviratne, S. I., et al. (2010). Global convergence in the temperature sensitivity of respiration at ecosystem level. Science, 329(5993), 838–840.
Meyer, N., Welp, G., & Amelung, W. (2018). The temperature sensitivity (Q10) of soil respiration: Controlling factors and spatial prediction at regional scale based on environmental soil classes. Global Biogeochemical Cycles, 32(3), 306–323.
Nottingham, A. T., Whitaker, J., Turner, B. L., Salinas, N., Zimmermann, M., Malhi, Y., & Meir, P. (2015). Climate warming and soil carbon in tropical forests: Insights from an elevation gradient in the Peruvian Andes. Bioscience, 65(9), 906–921.
Novielli, P., Romano, D., Magarelli, M., Bitonto, P. D., Diacono, D., Chiatante, A., Lopalco, G., Sabella, D., Venerito, V., Filannino, P., et al. (2024). Explainable artificial intelligence for microbiome data analysis in colorectal cancer biomarker identification. Frontiers in Microbiology, 15, 1348974.
Ozenne, B., Subtil, F., & Maucort-Boulch, D. (2015). The precision–recall curve overcame the optimism of the receiver operating characteristic curve in rare diseases. Journal of Clinical Epidemiology, 68(8), 855–859.
Patil, A., & Lamnganbi, M. (2018). Impact of climate change on soil health: A review. International Journal of Chemical Studies, 6(1), 2399–2404.
Rolnick, D., Donti, P. L., Kaack, L. H., Kochanski, K., Lacoste, A., Sankaran, K., et al. (2022). Tackling climate change with machine learning. ACM Computing Surveys, 55(2), 1–96.
Saez-Sandino, T., Garcia-Palacios, P., Maestre, F. T., Plaza, C., Guirado, E., Singh, B. K., et al. (2023). The soil microbiome governs the response of microbial respiration to warming across the globe. Nature Climate Change, 13(11), 1382–1387.
Steyerberg, E. W. (2019). Coding of categorical and continuous predictors. In Clinical prediction models (pp. 175–190). Springer.
Swain, P. H., & Hauska, H. (1977). The decision tree classifier: Design and potential. IEEE Transactions on Geoscience Electronics, 15(3), 142–147.
Tong, D., Li, Z., Xiao, H., Nie, X., Liu, C., & Zhou, M. (2021). How do soil microbes exert impact on soil respiration and its temperature sensitivity? Environmental Microbiology, 23(6), 3048–3058.
Wen, P., Xu, Q., Yang, Z., He, Y., & Huang, Q. (2024). Algorithm-dependent generalization of AUPRC optimization: Theory and algorithm. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(9), 5062–5079.
Wilhelm, R. C., van Es, H. M., & Buckley, D. H. (2022). Predicting measures of soil health using the microbiome and supervised machine learning. Soil Biology and Biochemistry, 164, 108472.
Winkler, J. P., Cherry, R. S., & Schlesinger, W. H. (1996). The Q10 relationship of microbial respiration in a temperate forest soil. Soil Biology and Biochemistry, 28(8), 1067–1072.
Xu, M., et al. (2021). Interactive effects of soil properties on temperature sensitivity of microbial respiration. Soil Biology and Biochemistry, 156, 108220.
Yang, S., Wu, H., Wang, Z., Semenov, M. V., Ye, J., Yin, L., Wang, X., Kravchenko, I., Semenov, V., Kuzyakov, Y., et al. (2022). Linkages between the temperature sensitivity of soil respiration and microbial life strategy are dependent on sampling season. Soil Biology and Biochemistry, 172, 108758.
تحقیقات آب و خاک ایران
دوره 56، شماره 10
دی 1404
صفحه 2809-2823

فایل ها

هم رسانی

ارجاع به این مقاله

آمار