In line with the national goal of sustainable intensification of production, there is a need to increase grain yields and the utilization rate of both N-fertilizers and fungicides. With a large variability of the terrain, soil quality and water availability, growing conditions for cereals may vary significantly even within a single field. For both economic and environmental reasons, crop treatments (e.g. fertilization and disease control) should no longer be conducted uniformly but instead adapted to local plant requirements. The overall goal of the STRESSLESS project is to build a recommendation system for site-specific N fertilization and fungal disease control in cereals, using sensor based measurements of the solar radiation reflected from the crop.
About the project
This project addresses a week point of existing systems for site-specific N fertilization - the inability to distinguish between plant signaling of N demand and their response to other stressors (e.g. water under/oversupply and diseases). Moreover, we also face the challenge of detecting fungal diseases in cereals at pre-symptomatic stage. This project addresses also a big R&D challenge, which is to transform a mosaic of information from a range of technical devices into agronomic meaningful decisions.
During the first three seasons (2015-2017) we have carried out extensive work in field trials at Apelsvoll, Toten, where we have focused on the effects of reduced spring fertilization, water stress and plant protection on the plants' spectral properties and yield response. We have developed a drip irrigation system and a system with plastic tent to prevent from. We have carried out sensor measurements with many different optical sensors to capture the plant's spectral properties, and we took plant samples to determine biomass, nitrogen and water content. Registration of the disease symptoms has also been performed at various growth stages. Grain yield and quality were determined using standard analytical methods. The results showed that plant reflectance measurements and multivariate analysis can provide good estimates of both N-content and water content in cereal plants. The spectral signals required for modeling of both parameters were found in separate wavebands. We have also identified a novel spectral index based on selected wavelengths, which allows us to distinguish between different water status in the plants. This index has been called a "drought index", and it can be used to correct estimated values for N-status and thus improve the calculation of the optimum N-fertility level. We validated the models with data from the field trials performed at Apelsvoll in 2018. In addition, we used data collected on various farms with good results. We conducted measurement campaigns in several places in Østfold and Romerike, with wheat grown on different soil types and in various growth conditions. The dry weather conditions in Eastern Norway in 2018 induced drought stress in the crops, something that was positive for the project.
We carried out both greenhouse and field trials to investigate whether it was possible to identify early symptoms of septoria and powdery mildew infestation in wheat by means of spectral measurements. Despite the fact that our data set comprised of samples with a great variation in biomass and that the plants were affected by many stress factors, it was possible to identify weak spectral signals of powdery mildew. Based on this, we have developed a classification model, which we consider to be an important contribution to the ongoing international scientific debate on the feasibility of early fungal detection using optical tools.
Already in the process of project application, we were aware of the complexity of the goal to develop an overall agronomic decision support system that would address many types of in-field variations.
We delimited our approach to conducting annual, separate field trials to investigate yield responses to the three stress investigated factors (nitrogen, water and fungal infection). Results from the four seasons showed that the crops are most affected by the general water supply every year. The second most important thing for crop production is how the availability of both water and nitrogen develops during the same growing season. Our conclusion is that a robust decision support system requires more effort than plant growth data that can be obtained using various sensors. It must also be connected to both weather data and soil properties.
The development of the drought index that can be used to distinguish between different water status in the plants, as well as the powdery mildew detection model were communicated both nationally (YARA N-sensor meeting, Grain meeting) and internationally (NJF seminar at Agromek in Herning, DK). Both of the main findings in the project will now be published in international scientific journals.
Vi har stor tro på at resultatene som er oppnådd i prosjektet vil bli tatt i bruk både nasjonalt og internasjonalt av forskere og praktikere for kartlegging av kornåker. Våre metodiske tilnærminger kan også være til nytte i andre forskningsfelt relatert til fjernmåling. Indeksen relatert til vannstatus i planter som vi har identifisert er et svært viktig funn som har et klart potensial til å kunne forbedre dagens sensorbaserte gjødslingssystemer til korn. Kornbønder utstyrt med et optimalisert verktøy kan effektivisere produksjonen. Dette vil også være positivt for miljøet, med redusert avrenning og forurensing på grunn av mer effektiv bruk av gjødsel og plantevernmidler. Nytteverdien av prosjektet blir tydelig ved formidling på fagmøter for norske brukere av optiske sensorer, der interessen er meget stor. Økt forståelse av kompleksitet av plantenes egenskaper bidrar til kompetanseutvikling av hver enkelt bruker av denne teknologien.
The overall goal of the proposed project is to build a system for site-specific fertilization and fungal disease control in cereals, designed especially to handle combinations of crop-stressing factors. The project represents the second step in a three-step approach towards a market-ready product. The first step is a research project focused on concept development at high academic level (MULTISENS, to be finished in March 2015). This project (step two) aims at bringing research a step further towards the market, by utilizing the knowledge platform established in MULTISENS and the best and the most relevant sensors available today. Step three will be to complete the work by developing a prototype of a sensor-system, in close cooperation with an industry partner.
The interface between technology and agriculture is currently sub-optimal, thus representing a bottleneck of innovations in agricultural production systems. This project addresses a week point of existing systems for site-specific N fertilization - the inability to distinguish between plant N demand and other stressors (e.g. water and diseases). Moreover, we also face the challenge of detecting fungal diseases in cereals at pre-symptomatic stage. The most critical R&D challenge faced in this project, is to transform a mosaic of information from a range of technical devices into agronomical meaningful decisions.
The project will contribute to national competence building on the utilization of cutting-edge technology for agricultural purposes. The anticipated advances in N deficiency and disease sensing in cereals will take the research front a step forward. Moreover, the project is expected to contribute to the general goal of sustainable intensification, primarily by providing a system, which potentially may increase use efficiencies of N-fertilizers and fungicides, and which may detect diseases in their pre-symptomatic stages.
FFL-JA-Forskningsmidlene for jordbruk og matindustri