Nitrogen limitation in high-yield soybean: Seed yield, N accumulation, and N-use efficiency

Abstract

Recent reviews about soybean nitrogen (N) requirement have speculated about a possible N limitation in high-yield environments (>4.5 Mg ha−1). However, these studies did not provide definitive experimental data to test that hypothesis and results from the literature are conflicting. To fill this knowledge gap, we evaluated N limitation across 13 high-yield soybean environments in Argentina and USA. Each experiment included a ‘zero-N’ treatment, which forced the crop to rely on biological N2 fixation and indigenous soil N, and a ‘full-N’ treatment, which provided an ample fertilizer N supply during the entire crop cycle based on site-specific yield potential. Accumulated N in aboveground dry matter (ADM) measured in a N-omission maize plot grown adjacent to the soybean experiments was used to determine indigenous soil N supply. Soybean seed yield, protein and oil concentration, ADM, harvest index (HI), accumulated N in ADM, N harvest index (NHI), and seed N were measured in both treatments at physiological maturity. A simple conceptual framework relating ADM with accumulated N was used to assess treatment differences in ADM, accumulated N, and N-use efficiency (NUE). To account for treatment differences in seed biomass composition, mass-based ADM was expressed in glucose equivalents (ADMe). Seed yield ranged from 4.6 to 6.7 Mg ha−1 (full-N) and 4 to 5.8 Mg ha−1 (zero-N), with the full-N treatment averaging 12% (0.6 Mg ha−1) and 3% (9 g kg−1) higher seed yield and protein concentration, respectively. The full-N treatment exhibited 18% (70 kg N ha−1) and 14% (1.6 Mg ha−1) greater accumulated N and ADM, respectively, compared with the zero N treatment, without changes in HI and NHI, but slightly lower NUE (29 versus 30 kg ADM kg−1 N in full and zero-N, respectively). However, NUE differences between treatments became indistinguishable when ADM was expressed as ADMe (45 kg glucose kg−1 N). The (full-N minus zero-N) seed yield difference and the indigenous soil N supply were negatively associated. This research documented (i) the existence of N limitation in high-yield soybean, (ii) that seed yield and protein concentration increases were physiologically associated with changes in accumulated N (but not in NUE and/or NHI), and (iii) that the degree of N limitation is modulated by indigenous soil N supply. Findings from this study can serve as a basis to identify environments with the largest N limitation and thus guide N management in soybean.