Occasional direct-seeding of grain legumes in Organic Agriculture in Germany and Brazil

Abstract

The aim of this work was to test strategies, which improve nutrient supply and weed control in occasional direct-seeded (DS) grain legumes in Organic Agriculture. The effect of intra-row fertilisation of rock phosphate (RP) and several sulphur fertilisers on crop growth, yield and nutrient uptake was studied in field trials with faba bean (<em>Vicia faba</em>) in Germany (two sites in NRW, 2011 and 2012) and with soybean (<em>Glycine max</em>) in Brazil (three sites in Paraná, 2012 and 2013). The second part of this work investigated weed control with natural herbicides (NH). Therefore, eighteen screening trials in the field and five field trials with soybean and common bean (<em>Phaseolus vulgaris</em>) were conducted in Brazil. Additionally, the effect of different amounts of oats straw residue on faba bean and weed shoot growth were examined as a second trial factor within the 2012 fertiliser trials in Germany. <br /> Soils at all <u>fertiliser trials</u> in Germany exhibited sufficient P-contents (LUFA C) and the intra-row<br /> fertilisation of RP (50 kg/ha^-1 P) did not have any effect on faba bean growth and P-uptake, neither alone nor in combination with elemental sulphur, which by way of in situ acid formation can enhance RP solubility. On the contrary, the sulphur fertilisers potassium sulphate, gypsum and elemental sulphur did result in a markedly increased S-uptake in both trial years, while faba bean growth and yield were only affected positively and in part significantly in the second trial year. The clear effect was assumed to be due to low soil sulphate contents during initial crop development, caused by low sulphate adsorption and hence high leaching rates during winter and also due to low soil organic matter (SOM) mineralisation rates under DS management. However in Brazil, soybean did not react to S-fertilisation with increased crop growth, yield or S-uptake in any trial. The sufficient S-supply was explained with the high contents of adsorbed sulphate in soil and high SOM mineralisation rates under tropical climate. P-fertiliser application was also ineffective at most sites due to sufficient to high soil P-contents. Only on a field with low P-contents at site <em>Ponta Grossa</em> P-fertilisation had a positive effect on P-uptake, crop growth and yield. In both years this positive effect was increased by simultaneous application of elemental sulphur, which presumably increased solubility of RP. <br /> In screening trials with <u>natural herbicides</u> pine oil and acidic acid were found to be the most potent active ingredients with a cell membrane disrupting (CMD) effect. While the sole spray application of high amounts of NaCl resulted in relatively weak plant damage, it was found that the combination of CMD and NaCl resulted in a strongly enhanced weed control. Additionally, a meristem damaging systemic effect of NaCl was identified, which was particularly strong for dicot weed species. In consequence, CMD amounts and hence application costs could be reduced drastically. Furthermore, it was determined repeatedly that emulsifiers influence formulation efficacy strongly at the rates commonly applied in NH and therefore these inert ingredients have to be considered as active ones in NH formulations. In the DS soybean field trial in Londrina it was determined that at a fixed total amount of AI (90 L/ha^-1 pine oil or limonene and 90 kg/ha^-1 NaCl) two concentrated applications resulted in a higher weed control than the three diluted ones. Weed control in this trial was not satisfactory, though, and crop growth as well as yield were markedly reduced compared to the clean control treatment. In the DS soybean trial<br /> in <em>Ponta Grossa</em> 50 L ha^-1 pine oil or limonene and 50 kg/ha^-1 NaCl were applied once, twice and three times. Even after three applications weed drymass remained relatively high compared to the weedy control. Nevertheless, soybean shoot drymass was not affected strongly by weed infestation and with three applications grain yield was close to the clean control. Nevertheless, results in this trial were influenced by crop damage due to contact with NH spray, because no working protective screen was available during application. In two field trials with conventionally tilled common bean, weed control was satisfactory in all spray treatments. Crop growth and yield were comparable to those of the weed free control.