TY - JOUR
AU - Mathur, Preet
AB - Macrophomina phaseolina, a soil saprophytic plant pathogen of global distribution and wide host range, was studied in relation to current and future (2050 and 2070) climate change scenarios, soil variables, and habitat heterogeneity indices (HHI). On 285 geographically thinned, presence-only data, we used R program-based Ensemble Species Distribution Modelling (ESDM) and eight individual algorithms to do ensemble modelling. When compared to other algorithms and ensemble outcomes, our study demonstrated that Random Forest (RF) was the best predictive individual algorithm. As a consequence, we utilized RF to assess this species’ habitat appropriateness, niche width, niche overlap, and area occupied within pre-defined habitat classes. In the present and 2050 Bio-Climatic (BC) periods, isothermality was recognized as the most significant element, whereas annual mean temperature was indicated as the most important regulating factor during BC-2070. According to HHI, the population of this species drops monotonically as the coefficient of variation increases. With the exception of 15 to 30 cm, depth soil predictor demonstrated that sand percentage had the least influence on the pathogen’s habitat at all examined depths. Silt played a vital function at varied depths. The findings of ESDM with the combined current data set demonstrated that climatic factors outperformed HHI and soil variables in terms of dispersion.
TI - Global distribution modelling of macrophomina phaseolina (tassi) goid: a comparative assessment using ensemble machine learning tools
JF - Australasian Plant Pathology
DO - 10.1007/s13313-023-00927-7
DA - 2023-07-01
UR - https://www.deepdyve.com/lp/springer-journals/global-distribution-modelling-of-macrophomina-phaseolina-tassi-goid-a-Q3V4VJaQDH
SP - 353
EP - 371
VL - 52
IS - 4
DP - DeepDyve
ER -