Assessing microclimate conditions of surface soil layers to improve weed emergence modelling

Thermal or hydrothermal models for weed emergence prediction are useful tools for Integrated Weed Management (IWM) and require an accumulation of Growing Degree Days (GDD) to be calculated through the comparison of base temperature for germination with daily average soil temperature. Consequently, the accuracy of measurements or estimations of soil temperature, which has a strong depth-dependant variability, affects the predictive quality. Emergence models for arable conditions adopt soil temperature measured at a depth of 3-5 cm, but this may not be adequate for no-till conditions. Daily and hourly soil temperatures were measured at depths of 2, 5, 10, 20 and 50 cm during the period of weed emergence and the respective means (Td and Th) were calculated. Accumulations of GDD were computed for Abutilon theophrasti, Chenopodium album and Sorghum halepense with values of Td and Th measured at depths of 2, 5 and 10 cm. The emergence curves calculated for each species with the AlertInf model were compared to identify estimation variability due to the adoption of values of soil temperature measured at different depths. Despite the great depth-dependent differences observed for Td and Th, differences among the emergence curves estimated for each species were not significant from the point of view of weed control. The adoption of Th measured at a depth of 5 cm could be the best compromise to guarantee model accuracy without complicating measurements. However, further studies are required to adjust and calibrate models developed for arable fields to no-till conditions.