Double crop soybeans management: a review, field studies, and modeling

Abstract

Double-crop (DC) soybean (Glycine max (L.) Merr.) systems, is an alternative to sustainably intensify production in agricultural land. However, DC system is subject to different environmental conditions relative to the one faced by full-season soybean. To better understand the effect of management practices on DC responses, and to learn how to improve desirable characteristics and minimize non-desirable outcomes, three approaches were chosen for the study of DC soybean. Chapter 1 was a systematic literature review. The objectives were to (i) quantify attainable yield for DC soybean benchmarking against full-season (FS) soybean; (ii) determine and build probabilistic response models on the effect of previous wheat productivity on DC soybean yields; and (iii) detect and rank factors influencing DC soybean yields via a decision inference tree analysis. Analysis showed that the yield gap between FS and DC soybeans increased from -31 to 1160 kg ha⁻¹ as FS yield improved from 1500 to 3000 kg ha⁻¹. Even though the proportion of variation accounted for wheat yields in the DC soybean/wheat yield ratio was low (R2 = 0.15), the probability of soybean yield being equal to wheat yield was 0, 20, 30, and 55% for wheat yields of ≥6, ≥4 and < 6, ≥2 and < 4, and < 2 Mg ha⁻¹. Inference tree analysis indicated that the major factors impacting success of the DC system was wheat yield, soybean planting date and maturity group. The second chapter aimed to evaluate the effect of the management practice treatments on seed quality. Seven management practice treatments were tested in each planting date: 1) common practice (no inputs), 2) no seed treatment; 3) non-stay green (without fungicide/insecticide); 4) high plant density (45 m⁻²); 5) wide rows (75 cm); 6) N effect (without late-season fertilizer N); and 7) kitchen sink, considering all the inputs evaluated in previous treatments. Protein, oil, fatty acids (stearic, palmitic, oleic, linoleic and linolenic), fiber and ash were analyzed. Oil content presented a negative relationship with protein content. Monounsaturated (oleic) and saturated (stearic and palmitic) fatty as well as total fatty acid yields were increased as inputs and seed yield increased. There were no differences for seed composition and quality for planting times for the majority of the analysis. Lastly, as seed filling duration increases, fatty acids concentrations decrease, increasing final seed content. Chapter 3 aimed to evaluate responses to management practices in field and run simulations for long-term responses of the variables in other environments. Seed yield in DC soybean can vary among years and is dependent on the management practices applied. Greater inputs can have positive influence in yield. However, greatest differences in yield, were observed for planting earlier, right after wheat harvest, beginning of June relative to the late planting date in June. Initial soil moisture had significant effect on yield, being negatively affected as initial soil moisture was 40% in comparison with 90%. Weather greatly affected seed yield for DC soybean. The greatest differences were observed for dry and warm weather, when late planting greatly impaired yield.