Feeding the growing global population requires an annual increase in food production. This requirement suggests an increase in the use of pesticides, which represents an unsustainable chemical load for the environment. To reduce pesticide input and preserve the environment while maintaining the necessary level of food production, the efficiency of relevant processes must be drastically improved. robotic systems for effective weed and pest control aimed at diminishing the use of agricultural chemical inputs, increasing crop quality and improving the health and safety of production operators. To achieve this overall objective, a fleet of heterogeneous ground and aerial robots was developed and equipped with innovative sensors, enhanced endeffectors and improved decision control algorithms to cover a large variety of agricultural situations. This article describes the scientific and technical objectives, challenges and outcomes achieved in three common crops.

The dispersal of Avena spp. (A. fatua and A. sterilis) by natural dissemination and by agricultural operations was studied in four experiments conducted in Spain and Britain. Natural dispersal was very limited, with a maximum dispersal distance of 1.5 m. Dispersal was higher in the geographic direction that was downwind than in any of the other three geographic directions. Although plant movement was very small under notillage, an annual patch displacement of 2-3 m in the tillage direction was observed under conventional soil tillage. Ploughing downhill resulted in much larger dispersal distances than ploughing uphill. In the crops studied, combine harvesters dispersed few Avena spp. seeds, because of the fact that the plants had shed most of their seeds (>90%) before harvest. The percentage of seeds available to be dispersed by the combine was dependent on the harvest time. Although combine harvesting may not contribute much to short-distance dispersal, it may play an important role in long-distance dispersal. In our studies, isolated plants were located up to 30 m from the original sources. This small proportion may have a significant effect on the distribution of the weed within a field, acting as foci for new patches.

El artículo seleccionado no se encuentra disponible por ahora a texto completo por no haber sido facilitado todavía por el investigador a cargo del archivo del mismo.

Weed monitoring is the first step in any site-specific weed management programme. A relatively large variety of platforms, cameras, sensors and image analysis procedures are available to detect and map weed presence/abundance at various times and spatial scales. Remote sensing from satellites or aircraft can provide accurate weed maps when the images are obtained at late weed phenological stages. Cameras located on unmanned aerial vehicles (UAVs) have been shown to be adequate for early-season weed detection in a variety of wide-row crops, providing images with relatively high spatial resolutions. Alternatively, weed detection/ mapping systems from ground-based platforms can achieve even higher resolutions using a variety of nonimaging and imaging technologies. These ground systems are suited, in some cases, for real-time site-specific weed management. Despite this rich arsenal of technologies, their commercial adoption is, apparently, low. In this study, we describe the state of the art of remotely sensed and ground-based weed monitoring in arable crops and the current level of adoption of these technologies, exploring major constraints for adoption and trying to identify research gaps and bottlenecks.

Summary Predictive empirical models of the timing of emergence were developed for ten major weed species in maize crops. Monitoring of seedling emergence was performed over two years in two maize fields located in Central Spain and Tagus Valley in Portugal. Thermal time was used as the independent variable for predicting cumulative emergence. Different non‐linear growth curves were fitted to the data sets of cumulative percent emergence for the different species, sites and years using genetic algorithms. Based on their emergence patterns, weed species were arranged into three groups. Species with early‐season emergence (Abutilon theophrasti, Xanthium strumarium, Datura stramonium, Datura ferox, Sorghum halepense, Digitaria sanguinalis and Echinochloa crus‐galli) reached 70% emergence with less than 700 growing day degrees (GDD). Species with whole‐season emergence (Cyperus rotundus and Solanum nigrum) started early their emergence processes but the emergence continued throughout the maize life‐cycle; they required up to 1300 GDD to reach 70% emergence. The only species with late‐season emergence was Sonchus oleraceus; it required more than 1300 GDD to reach 70% emergence. The results obtained in our experiments have shown a good synchrony between the predictions obtained in different years in the same site. However, no single model was able to predict the timing of emergence in two sites with different environmental conditions, challenging the hypothesis that a single general model, based on temperature only, can be used to predict weed emergence in different geographical locations.