Forecasting models: from data to the field
Today farmers have to deal with several daily issues, so using advanced technologies to record and use data is essential. In this sense, understanding the potential of forecasting models and how they work is a prerequisite for those who want to manage their fields professionally.
The current context
Climate change in recent years has caused a reorganization of protection strategies, especially for those crops where management of many diseases and insects is required. Due to alternating frosts and droughts, and the occurrence of abundant and poorly distributed rainfall, it becomes essential to anticipate the occurrence of diseases and insect infestations in order to be able to react quickly and limit possible damage. In recent years, the interest of the Decision Support System (DSS) has also increased due to regulations (Directive 2009/128/C) concerning the reduction of chemical inputs in the context of integrated pest management. To this must be added all the other commitments that Europe and beyond impose on us regarding environmental sustainability such as: Green Deal, Pesticide Package and Farm to Fork strategies. All this is happening in a context where the Utilised Agricultural Area (UAA) and the number of molecules for crop protection management are decreasing, air pollution, soil erosion and the number and severity of diseases and insects are increasing, the latter due to the non-rational use of plant protection products (PPP) and globalization, not to mention the increase in world population that will lead all actors in the agri-food system to increase the quantity and quality of food in the near future.
What is a model?
A plant disease model is a simplification of a real system consisting of the pathogen, the host plant and the environment. Models are used in epidemiology to describe, understand, predict and compare epidemics and their components. A practical help for growers comes from disease prediction, which is an estimation process in past, current or unknown future situations.
For the development of forecasting models, general climatic data such as minimum, average and maximum temperature and relative humidity, daily and monthly precipitation and number of rainy days, wind speed and duration, dew point, hours of sunshine are needed as input. Depending on the complexity of the data, historical climate data (e.g. on a 30-year basis) are also used. In addition, data on crops and pests (pathogens and/or phytophages) may also be requested. For example, the following parameters can be used: for the crop – leaf moisture, plant density, phenological stages, maximum leaf area index, vapour pressure deficit, net radiation above and below the canopy, susceptible phenological stages; for fungal pathogens – the rate of spore release, maturation and germination, their number on the leaves, the amount of primary and secondary inoculum and the rate of colonisation. For insects (e.g. L. botrana on vines) – flight patterns of male individuals (detected with pheromone traps) and distribution of male catches between generations.
How and where is it used in agriculture?
As the quality and computational power offered by computers improves day by day, models are increasingly incorporated into DSS, which are tools that assist users in tactical and operational decision making in crop protection, in order to better plan applications with PPPs in defence strategies. Models can also be part of local disease warning systems. The use of forecasts for crop protection against fungal diseases and pests is not so new (e.g. rule of 3-10 for downy mildew of the grapevine).
The areas of application in agriculture are different, from crop growth and development to crop productivity, water balance, and defence against biotic and abiotic adversities.
Advantages and disadvantages
The main advantages of DSS are the possibility to analyse different scenarios, to better understand physical and biological processes, to carry out manipulations on the real system in order to verify hypotheses on its functioning, to evaluate the effect of possible external interventions in order to modify the behaviour of the system, to favour the exchange of data and to allow a better use of data and resources. In contrast, the critical aspects consists of a use often limited by the development of software packages, and the necessity of verifying and validating procedures in different years.
The forecasting models of xFarm Lab
The forecasting models of xFarm Lab are conceived, designed, calibrated and validated to achieve a single objective: to give farmers an additional technical tool that does not simply complement the others already available (e.g. agronomic, biological, physical, chemical) but coordinates them all to increase the efficiency and effectiveness of protection strategies and cultivation techniques, in order to increase the quality and quantity of production.
In the context of integrated crop protection, the use of DSS is a valuable tool to minimise the number of applications of PPPs. This concept is valid both for farmers with small plots of land and for agri-food chains that have the possibility, on a large scale, to amplify the positive effect resulting from a rational and efficient management of resources.
The primary objective of xFarm Lab is to support the entire agri-food system through digitalisation and technologies applied to agriculture in order to safeguard the sustainability of production processes in both agronomic and economic terms.
Dubey, P.K., Singh, G.S., Abhilash, P.C. (2019). Adaptive agricultural practices: Building resilience in a changing climate. Springer.
Pertot, I.; Caffi, T.; Rossi, V.; Mugnai, L.; Hoffmann, C.; Grando, M.S.; Gary, C.; Lafond, D.; Duso, C.; Thiery, D.; Mazzoni, V.; Anfora, G. (2016). A critical review of plant protection tools for reducing pesticide use on grapevine and new perspectives for the implementation of IPM in viticulture. Crop Protection, (), S0261219416303489–.
Rossi, V., Giosuè, S., Caffi, T. (2010). Modelling plant diseases for decision making in crop protection. In Precision crop protection-the challenge and use of heterogeneity (pp. 241-258). Springer, Dordrecht.