Machine learning methods for estimation the indicators of phosphogypsum influence in soil
The full understanding of the effect of mineral waste-based fertilizer in soil is still unrelieved, because of the extreme complex chemical composition and plethora of their action pathways. The purposes of this paper is to quantify the input of PG into the soil ecosystem process, considering the direct effects of PG as a whole on soil environment using of a plethora of chemical, toxicological, and biological tests.
Materials and methods
Greenhouse experiment includes different PG doses (0, 1%, 3%, 7.5%, 15%, 25%, and 40%) and two-time collection points after treatments—7 and 28 days. For each treatment and each time collection point, we measure (i) soil pH, bioavailable (H20 and NH4COOH-extractable) element content (S, P, K, Na, Mg, Ca, Fe, Zn, Sr, Ba, F); (ii) soil enzyme activities—dehydrogenase, urease, acid phosphatase, FDA; (iii) soil CO2 respiration activity with and without glucose addition; (iv) Eisenia fetida, Sinapis alba, and Avena sativa responses. Finally, we combine the ordinary chemical, toxicology, and biological measuring of soil properties with state-of-the-art mathematical analysis, namely (i) support vector machines (used for prediction), (ii) mutual information test (variable importance tasks), (iii) t-SNE and LLE algorithms (used for unsupervised classification).
Results and discussion
The results show similarity between the 0%, 1%, and 3% PG treatments in all collection times based on the toxicological and biological properties. Beyond 7.5% PG, some biological test was significantly inhibited in response to trace element stress. Among all tested parameters, soil urease activities, soil respiration activities after glucose addition, S. alba root lengths, and E. fetida survival rates show sensitivity to PG addition. Furthermore, the machine learning algorithms revealed that only several elements (mobile and water-soluble forms of Ca, Ba, Sr, S, and Na; water-soluble F) could be responsible to elevated soil toxicity for those indicators. SVR models were able to predict soil biological and ecotoxicity properties, and increasing numbers of randomly selected training examples from 50 to 90% of initial experimental data significantly improved model performance.
At this study, we demonstrate benefits of unsupervised machine learning methods for investigating toxicity of man-made substances in soil that can be further applied to risk assessments of various toxins, which are of significant interest to environmental protection.
KeywordsBioassay Biological properties Feature relevance Machine learning Pollution Regression Soil Trace element Waste
The research was supported by the Skoltech Next Generation Program and the Russian Found of Basic Research (Project No. 16-34-00063 mol_a). We are grateful to Anastasia Sharapkova (Rosetta Stone MSU) for proofreading the final version of the manuscript.
- Alvarenga P, Clemente R, Garbisu C, Becerril JM (2018) Indicators for monitoring mine site rehabilitation. In: Bio-Geotechnologies for Mine Site Rehabilitation. Elsevier, pp 49–66Google Scholar
- Ascari JP, Mendes IRN (2018) Desenvolvimento agronômico e produtivo da soja sob diferentes doses de gesso agrícola. Revista Agrogeoambiental 9Google Scholar
- Blum SC, Caires EF, Alleoni LRF (2013) Lime and phosphogypsum application and sulfate retention in subtropical soils under no-till system. J Soil Sci Plant Nutr 13(2):279–300Google Scholar
- Chaudhari MS (2016) Acute toxicity of diammonium phosphate to earthworm (Eudrilus eugeniae). J Entomol Zool Stud 4(6):501–503Google Scholar
- International Atomic Energy Agency Report (2013) Radiation protection and management of norm residues in the phosphate industry. 308 P. https://www-pub.iaea.org/MTCD/Publications/PDF/Pub1582_web.pdf. checked 19.11.2018
- ISO 14240-1:1 (1997) Soil quality - determination of soil microbial biomass - Part 1: substrate-induced respiration methodGoogle Scholar
- ISO 16072 (2002) Soil quality - laboratory methods for determination of microbial soil respirationGoogle Scholar
- ISO 19204 (2017) Soil quality - procedure for site-specific ecological risk assessment of soil contamination (soil quality TRIAD approach)Google Scholar
- Konarbaeva G (1997) Fluorine in the crusty solonetzes of Western Siberia and the impact of phosphogypsum on its content. Eurasian Soil Sci 30:977–981Google Scholar
- Kraskov A, Stögbauer H, Grassberger P (2004) Estimating mutual information. Phys Rev E 69. https://doi.org/10.1103/PhysRevE.69.066138
- Liu Z, Rong Q, Zhou W, Liang G (2017) Effects of inorganic and organic amendment on soil chemical properties, enzyme activities, microbial community and soil quality in yellow clayey soil. PLOS One 12. https://doi.org/10.1371/journal.pone.0172767
- Maaten LVD, Hinton G (2008) Visualizing data using t-SNE. J Mach Learn Res 9:2579–2605Google Scholar
- McBride M (1989) Reactions controlling heavy metal solubility in soils. Adv Soil Sci 26:1–56Google Scholar
- Morgado RG, Loureiro S, González-Alcaraz MN (2018) Changes in Soil Ecosystem Structure and Functions Due to Soil Contamination. Soil Pollution. Elsevier, pp 59–87Google Scholar
- OECD Guideline for testing chemicals 222 (2004). Earthworm reproduction test (Eisenia fetida/Eisenia andrei)Google Scholar
- Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blindel M, Prettenhofer P, Wiess R, Dubourg V et al (2011) Scikit-learn: machine learning in Python. J Mach Learn Res 12:2825–2830Google Scholar
- Pereira R, Cachada A, Sousa JP, Niemeyer J, Markwiese J, Andersen CP (2018) Ecotoxicological effects and risk assessment of pollutants. Soil Pollution. Elsevier, pp 91–216Google Scholar
- Russian National Report (2015) On the state and protection of the environment issued annually by the Ministry of Natural Resources and Environment (in Russian)Google Scholar
- Telesiński A, Siwczyk F, Zakrzewska H (2012) An attempt to determination of the 50% phytotoxicity threshold for different fluoride concentrations affecting the spring wheat (Triticum aestivum L.) and white mustard (Sinapis alba L.) seedlings. Fluoride 45(3/1):213–214Google Scholar
- Thalmann A (1968) Zur methodic derestimung der. Dehydrogenaseaktivität i. Boden mittels. Triphenyltetrazoliumchlorid (TTC). Landwirdschaft. Forschung 21:249–258Google Scholar
- Vapnik VN, Golowich S, Smola A (1997) Support vector method for function approximation, regression estimation, and signal processing. In: Mozer MC, Jordan MI, Petsche T (eds) Advances in neural information processing systems, vol 9. MIT Press, Cambridge, pp 281–287Google Scholar
- Vyshpolsky F, Mukhamedjanov K, Bekbaev U, Ibatullin S, Yuldashev T, Noble AD, Mirzabaev A, Aw-Hassan A, Qadir M (2010) Optimizing the rate and timing of phosphogypsum application to magnesium-affected soils for crop yield and water productivity enhancement. Agric Water Manag 97:1277–1286. https://doi.org/10.1016/j.agwat.2010.02.020 CrossRefGoogle Scholar
- Zhu J, Wang J, Ding Y, Liu B, Xiao W (2018) A systems-level approach for investigating organophosphorus pesticide toxicity. Ecotox Environ Saf 149:26–35Google Scholar