A comparison of statistical and dynamic modeling of wheat (Triticum aestivum L.) fungal diseases under the climate change

Crop simulation models are complementary tools to illustrate and explore the dynamics of climatic variants and to develop innovational crop management system under miscellaneous climatic conditions in a crop based ecosystem. Production of wheat as staple food is impinged due to increasing population, biotic and abiotic stresses like climatic events including increased temperature, droughts, floods and variability in rainfall pattern due to global warming. Climate is likely an acute driver of evolutionary changes in pathogen and host populations by interfering with host plant-pathogen interaction, population dynamics and gene expression. In this perspective, current study was designed to analyze wheat crop as host to explore the dynamics of foliar blight Alternaria triticina and Drechslera sorokiniana under varying climatic conditions on wheat growth, development and productivity for Islamabad region of Pakistan by comparing and evaluating two modeling methods, i.e. Statistical (R model) and Dynamic (APSIM model). The impacts of fungal diseases on wheat yield under historical long term climatic data (1961-2012) and for forecasting of wheat crop under A2 scenarios (2100) were studied using farming system APSIM model. Efficiency of both models was evaluated using model validation skill scores including RMSE, d-stat and R-2. For statistical modeling, the values for pathogen occurrence for Alternaria leaf blight were with 50% with d-stat 0.43 and RMSE 11.36 and for Drechslera leaf blight was 63% with d-stat 0.75 and RMSE 8.81. For dynamic modeling, the values for pathogen occurrence for Alternaria leaf blight were with 97% with d-stat 0.79 and RMSE 1.11 and for Drechslera leaf blight was 93% with d-stat 0.80 and RMSE 1.40. The finding of the results led to the conclusion that dynamic model was more suitable and efficient to simulate wheat crop development and growth. As disease module is absent in dynamic model APSIM therefore future management options could be predicted by linking the future climate (2100) with host-pathogen interactions. Similarly, analysis of climatic scenarios that change over time might be performed in dynamic model. Consequently, it can be used further for decision making to mitigate the ill effects of climate extremes and design future management strategies based upon model outcomes. The results justified the significance of models as beneficial and necessary tools in management strategies and decision making processes and highlighted that both models might be used to predict disease occurrence and impact studies but dynamic model APSIM could be used more significantly with good validation skill scores than statistical modeling. This study is novel as there is no already published data/work regarding the modeling and simulation among the fungal diseases of wheat crop particularly from Pakistan.