Péter Riczu

Spatial decision support for crop structure adjustment – a case study for selection of potential areas for sorghum (Sorghum bicolor (L.) Moench) production

One option for adaptation to climate change is to grow a wider variety of plant species. Sorghum (Sorghum bicolor (L.) Moench) is known to tolerate unfavourable environmental conditions, so it may be feasible to grow it on areas with extreme conditions to replace other species such as maize. Nowadays, spatial decision supporting systems primarily support the crop production process rather than crop structure adjustment. In this study, potential sorghum production sites in the Great Hungarian Plain were selected based on soil characteristics including genetic soil type, parent material, physical soil type, clay composition, water management, pH, organic matter content, topsoil thickness and fertility, as well as climatic data, particularly precipitation. For all the parameters the aim was to find the extreme values at which sorghum, which is less sensitive than maize, may still give an acceptable yield. By combining map layers of soil characteristics, it could be concluded that although the soil is suitable for sorghum on 40.46% of the Great Hungarian Plain, maize is generally a better choice economically. On the other hand, the soil conditions on 0.65% of the land are still suitable for sorghum but unfavourable for maize. As regards the precipitation demand of sorghum, May is the critical period; on 698,968 ha the precipitation required for germination was only recorded once in the period 1991-2010, so these areas cannot be considererd for sorghum. As a consequence, in an alternative crop rotation system sorghum could be competitive with maize, but both the soil and climate conditions and the demands of the crop need to be assessed. The lack of precipitation in critical phenophases significantly decreases the area where maize can survive. Sorghum, however, may produce an acceptable yield, as it is a drought-resistant species.

Precision Weed Detection using Terrestrial Laser Scanning Techniques

Weeds compete with cultivated plants for nutrients and water in agricultural fields and orchards. Site-specific precision plant protection requires optimal plant-specific pesticide combinations and quantities. The first step in precision weed control is to map and detect weeds. Active and passive remote-sensing tools are available for weed detection. This article presents a study of precision weed detection using three-dimensional terrestrial laser scanning in an apple (Malus domestica) orchard. The laser scanner was able to identify two monocotyledonous (Digitaria sanguinalis, Echinochloa crus-galli) and two dicotyledonous (Galinsoga parviflora, Portulaca oleracea) weed species. In addition to weed identification, weed coverage was also determined using three algorithms. The results indicate that laser scanning has the potential for fast, accurate weed detection and could support the development of a plant-specific, selective, precision weed-control system that reduces water and pesticide use in orchards.