Variable application of pre-emergent herbicides in winter rapeseed

Herbicides account for almost 50% of the consumption of plant protection products in the Czech Republic, and it is still common agricultural practice to apply a single dose of herbicide to the entire plot. Nevertheless, selective (zonal) applications of herbicides are already appearing in operational practice today, in which the herbicide is applied only to the weed plant or its local location. This application saves between 10 and 90% of the herbicide, depending on the level of weed infestation. However, in the vast majority of cases, these are only post-emergence applications, which are conditioned by the previous detection of weed plants in the crop or in the intercrop period. Very good results are achieved with selective applications of herbicides to weed plants such as field sorrel, common datura, outbreaks of creeping fescue or common sedge. Pre-emergence herbicides are applied before the emergence of crops and weeds, and it is therefore not possible to physically detect weed plants as in post-emergence applications. The possibilities for selective applications are therefore very limited and are carried out very rarely based on historical identification or predisposition of weed infestation.

Fig. 1 The use of variable applications appears to be very effective in winter rapeseed

A general condition for good efficacy of pre-emergence applied soil herbicides is sufficient soil moisture, on the contrary, their effectiveness may fail in dry conditions. Variability of soil properties can cause different sorption of herbicides. Herbicides are more phytotoxic to crops on light and sandy soils, which are also characterized by a low sorption capacity. Mikulka (2022) recommends applying lower doses of herbicides on such soils. On the other hand, heavy soils with a high sorption capacity bind very strongly the active substances of herbicides. However, the professional literature does not provide a mechanism for changing the dose of herbicide, whether to change the dose of active substance using the amount of water - this method is available today or the concentration of the active substance in the spray liquid. This method is technically and financially expensive. However, many authors agree that each herbicide will have a certain type of behavior in different soil conditions and it is necessary to verify the sorption of the herbicides used. At the same time, the spatial distribution of weed plants within the plot is not uniform. The description of the behavior of herbicides in the environment is very complex, because the processes of their binding to soil particles or organic matter are influenced by a large number of factors.

From the point of view of determining the level of sorption capacity within a plot, a detailed determination of the variability of soil properties is essential. Soil sampling is considered to be one of the most accurate methods of determining soil properties. An important evaluation parameter is obtaining representative samples. In precision agriculture, sampling ranges from 1 to 5 ha per sample, depending on the level of variability. Another variation is the determination of a sampling network, which ideally reflects the relief or production zones of the plot. The relatively high costs of soil sampling and subsequent laboratory analysis lead to the development of indirect measurement methods using so-called on-the-go sensors, which could detect critical soil properties at each location of the plot as these sensors move around the plot (Adamchuk 2011). Although the literature describes a wide range of design concepts and measurement methods, in operational practice, mainly eclectic and electromagnetic sensors based on measuring electrical resistance, capacitance or conductivity of the soil have proven themselves. Mapping electrical conductivity of the soil is a simple and relatively inexpensive tool for measuring soil properties. The measurement is based on the ability of a substance to conduct an electrical current and is expressed in units of millisiemens (mS/m). Brant et al. (2020) describe correlations between soil conductivity and the occurrence of common sedge plants.

Fig. 2 Measuring set equipped with EM 38–MK2 probe for measuring soil conductivity

Soil conductivity maps show a high correlation with yield maps, which document crop yield within a plot using combine harvesters. According to a number of authors, soil properties, expressed as soil conductivity, explain up to 85% of yield variability. Different soil properties are mainly related to infiltration and water retention in the soil, which has a direct impact on plant stress during vegetation and thus yield. Light soils are generally characterized by lower water capacity, which is the main limiting parameter for achieving the desired yields of cultivated crops in the Czech Republic. At the same time, as already mentioned, light soils with low water capacity have lower sorption capacity and it is desirable, according to the cited authors, to reduce the amount of herbicide in these parts of the plot.

Harvesting crops using combine harvesters is associated with harvest losses, which are optimally at the level of 1% of the harvested production. The effort to reduce harvest losses below 1% is associated with a decrease in working speed and thus with a decrease in the area performance of the combine harvester. It is therefore necessary to find the optimal ratio between area performance (ha/hour) and percentage harvest losses (%). An important variable in this calculation is the price of commodities (CZK/t), which fundamentally changes the calculation of harvest losses. Harvest losses (so-called "stubble") can cause significant problems in the subsequent crop, especially in winter rapeseed, which is sown immediately after the harvested previous crop. With an average yield of winter and spring wheat of 6.07 t/ha, harvest losses at the level of 1% remain at approximately 60.7 kg/ha of harvest losses. It should be noted that in operational practice, due to poorly adjusted combine harvesters, operators, crop conditions, etc., harvest losses are often significantly higher and can reach 5%, which is approximately 303.5 kg/ha of harvest losses. In this case, the loss amount is similar to the seeding rate when sowing the crop.

Pre-emergence herbicide applications in winter rapeseed are carried out in the Czech Republic on an area exceeding 70% of the sown areas. These applications are currently carried out with a uniform dose of spray, when a combination of herbicides is used against dicotyledonous and monocotyledonous weeds is used. If pre-emergence application is not possible, it is replaced by early post-emergence. During this period, there is often high competition between weeds and winter rapeseed. The results of the verification of the relationship between weed pressure and soil properties to date show a high correlation between harvest losses and soil properties, which are expressed by the production potential of the plot (Fig. 3). The use of variable applications and their effect on limiting weed pressure with harvest losses is currently being operationally verified. The ongoing results show the suitability of changing the herbicide dose depending on the yield level (zone) of the plot. In practice, a prescription map for pre-emergence or early post-emergence application would be created based on the production potential of the land or a yield map from a combine harvester, which would reflect the current soil conditions of the land through yield (t/ha).

Fig. 3 High correlation between the production potential of the plot (left) and the post-harvest losses of the previous crop (right).

Verification of the possibility of using (zonal) selective post-emergence applications did not bring satisfactory results compatible with operational use due to the staged emergence of harvest losses. A compromise and a significant reduction in consumption together with a higher efficiency of pre-emergence herbicides is the combination of band and variable application of pre-emergence herbicides, when a commonly available application technique can be used for application, which has a suitable nozzle and seeding unit spacing, or the application can be carried out directly during the sowing of the given crop using a liquid tank and an application frame behind the seeding units. Herbicide-untreated areas do not pose a risk during the emergence of winter rapeseed and during the subsequent, today common, corrective autumn herbicide intervention, an area application can be carried out, suppressing even pre-emergence-untreated areas of the land.

The economic benefits of variable applications are questionable, especially for herbicides; rather than savings, we should talk about better use, sorption of herbicides with a positive response to the cultivated crops. On the other hand, with band applications, at a crop spacing of 50 cm, up to 50% herbicide savings occur. Another calculated benefit may be the combination of application with crop sowing and reduction of herbicide application costs with guaranteed timeliness and accuracy of herbicide intervention. The technical solution itself for variable, band or selective (zonal) applications is subsequently transferable and usable in other crops, such as sugar beet or corn.

Fig. 4 A seeding unit equipped with a front liquid tank and an application frame behind the seeding units allows for band and variable application of pre-emergent herbicide.

The pressure of low purchase prices of commodities, rising input prices and pressure from state institutions to reduce plant protection products and fertilizers is forcing farms to seek savings also in the area of herbicide protection. Currently, the Center for Precision Agriculture at the ČZL and cooperating farms are also testing herbicide strategies in winter rapeseed with the aim of reducing herbicide consumption.

Dedication: This work was created within the framework of the NAZV project QL24020309.