Use of machine learning to establish limits in the classification of hyperaccumulator plants growing on serpentine, gypsum and dolomite soils
Metadatos
Mostrar el registro completo del ítemEditorial
Universidad Complutense de Madrid
Materia
Hyperaccumulators Serpentinophytes Nickel Phytoremediation Artificial intelligence Inteligencia artificial
Fecha
2021-03-08Referencia bibliográfica
Mota-Merlo, M. & Martos, V. 2021. Use of machine learning to establish limits in the classification of hyperaccumulator plants growing on serpentine, gypsum and dolomite soils. Mediterr. Bot. 42, e67609. [https://dx.doi.org/10.5209/mbot.67609]
Resumen
So-called hyperaccumulator plants can store heavy metals in quantities a hundred or a thousand times higher
than typical plants, making hyperaccumulators very useful in phytoremediation and phytomining. Among these, there are
many serpentinophytes, i.e., plants that grow exclusively on ultramafic rocks, which produce soils with a great proportion
of heavy metals. Even though there are multiple classifications, the lack of consensus regarding which parameters should
be used to determine if a plant is a hyperaccumulator and the arbitrariness of established thresholds elicits the need to
propose more objective criteria. Therefore, this work aims to refine the existing classification. To this end, plant mineral
composition data from different vegetal species were analyzed using machine learning techniques. Three complementary
approaches were established. Firstly, plants were classified into three types of soils: dolomite, gypsum, and serpentine.
Secondly, data about normal and hyperaccumulator plant Ni composition were analyzed with machine learning to find
differentiated subgroups. Lastly, association studies were carried out using data about the mineral composition and soil type.
Results in the classification task reached a success rate of over 75%. The clustering of plants by Ni concentration in parts per
million (ppm) resulted in four groups with cut-off points in 2.25, 100 (accumulators) and 3000 ppm (hyperaccumulators).
Associations with a confidence level above 90% were found between high Ni levels and serpentine soils, as well as between
high Ni and Zn levels and the same type of soil. Overall, this work demonstrates the potential of machine learning to analyze
plant mineral composition data. Finally, after consulting the IUCN’s red list as well as those of countries with high richness
in hyperaccumulator species, it is evident that a greater effort should be made to establish the conservation status for this
type of flora.