As was reported in the review of the literature of Wichern et al. (2008), the 15N label is concentrated in the shoot tissue and in this study the amount of belowground excess 15N in roots, nodules and soil did not exceed 25 % of that applied. A few studies have followed the change in 15N enrichment of plant tissues and soil with time after labelling (McNeill et al. 1997, 1998; McNeill and Fillery, 2008; Gasser et al. 2015; Hupe et al. 2016; Rasmussen et al. 2019). As was recorded here (Fig. 3), all the studies showed that the aerial tissue decreased in 15N enrichment with time after labelling, owing to the dilution of the fixed amount of enriched N with increasing unlabelled N derived from soil and/or BNF. However, root DM and N accumulation with time have been found to be slow or even negative. Root DM decreased after 47 DAL and total root N decreased from a peak of 62 mg at 14 DAL to 30 mg plant− 1 at the final harvest (Figs. 1C and 2C). However, nodule DM accumulation was rapid from 25 to 47 DAL reaching 2.2 g at 47 DAL and decreased to 1.3 g plant− 1 at 70 DAP (Fig. 1D). Nodule total N showed a similar behaviour (Fig. 2D). As the 15N enrichment of the nodules changed very little with time, the N for nodule growth from 25 to 47 DAL (Fig. 4D) must have come from both labelled and non-labelled sources, the latter most likely being BNF. Many other studies have shown that nodules are lower in 15N enrichment than roots (e.g. Oghoghorie and Pate, 1972; Jensen, 1996; Russell and Fillery, 1996b) but there seems to be no other reports of the change in 15N enrichment of nodules with time.
The six papers cited where the 15N enrichment of different root cohorts was followed with time all reported results similar to those presented here (McNeill et al. 1997, 1998; McNeill and Fillery, 2008; Gasser et al. 2015; Hupe et al. 2016; Rasmussen et al. 2019). Root 15N enrichment, especially secondary or fine roots, varied little over time after two to three weeks after leaf-labelling. This is unexpected, but seems to be mainly due to very slow N accumulation of roots pre-established for 20 to 40 days and hence little dilution of the original fixed amount of excess 15N. Our results also show that while N accumulated by nodules increased from 22 mg plant− 1 at 25 DAL to 71 mg at 47 DAL and returned to 21 mg plant− 1 at 70 DAL, the 15N enrichment only varied from 0.150 to 0.159 to 0.111 atom % 15N excess over these three sequential harvests.
This relatively low variation in 15N enrichment of roots has been observed by all authors who made sequential harvests to study this. However, only in this study and those of Gardner et al. (2012), Gasser et al. (2015) and Rasmussen et al. (2019) was the soil sampled only one or two days after leaf-labelling. In all cases some of the applied 15N-enriched N was found in the soil after only 24 or 48 h after leaf-labelling. Gardner et al. (2012) reported that 1.3 to 1.4 % of the enriched N was recovered in the soil, when sub-clover (Trifolium subterraneum) or lucerne (Medicago sativa) leaves were labelled with 15N-enriched urea. Gasser et al. (2015) working with red clover (Trifolium pratense) found that 0.5 % of the excess 15N was recovered in the bentonite/sand mixture just 24 h after leaf-labelling. Rasmussen et al (2019), reported that the 15N-enriched N deposited in the soil in the first 24 h after leaf-labelling with 15N-enriched urea was equivalent to up to 5 % of unrecovered roots of white and red clover (T. repens and T. pratense, respectively). Gardner et al. (2012) suggested that the presence of 15N label in the first day after labelling was due to unrecovered roots which were impossible to separate from the soil, but even after 32 days there was no significant increase in the excess 15N recovered, which begs the question as to why there was no further increase in the quantity of unrecovered roots in the soil during this period. Rasmussen et al. (2019) also suggested that much of this short-term deposition of excess 15N in the soil could be due to unrecovered roots, but they recognised that unrecovered labelled roots could not explain the rapid appearance of 15N in neighbouring grass which corresponded to 6–8 % and 12 to 16 % of the excess 15N recovered outside the labelled plant (soil + grass). In contrast the rapid loss of a small proportion of excess 15N from red clover was attributed to leakage of soluble forms of 15N by Gasser et al. (2015).
In our study on soybean in the treatments leaf labelled with urea, by 2 DAL approximately 5 % of the 15N label had been transferred from the labelled leaf to the shoot and less than 1% was found in the roots. However, in this 48-h period the mean “leakage” to the soil was between 4.0 and 5.9 % of all labelled N, thus amounting to almost half of the N exported from the labelled leaf to the shoot, roots and soil (Table 1). There was a similar behaviour of the excess 15N deposited in the soil derived from labelled glutamine. At 2 DAP, only 2.8 % of the applied N was recovered in the roots while 11.6 % was released into the soil. As the amount of 15N excess in the roots of all treatments at 2 DAL was much lower than that found in the soil, the roots must be considered to be a conduit for enriched N rather than a source. This supports the hypothesis of Gasser et al. (2015) that this excess 15N was lost from the roots in solution.
Rasmussen et al. (2019) strongly criticised the conclusion of Gasser et al (2015) that the excess 15N found in the growth medium just 24 h after labelling was due to leakage of soluble forms of enriched N into the soil. However, in their own study they found enriched N in neighbouring grass roots just 24 h after leaf labelling. They attributed this to labelled N in “root exudates” which was clearly in solution for such rapid transfer. Gasser et al. (2015) included plants in their study which were not subjected to leaf labelling or any manipulation of the leaves. They showed that there was significantly more ammonium in the leaf-labelled plants, so that this increase in ammonium was definitely associated with the process of labelling the leaves. Rasmussen et al. (2019) suggested that the damage caused to the petioles during leaf-labelling may have been partially responsible of the short-term 15N leakage. In our study in two of the treatments labelled with enriched urea (ULL and USLL) the leaves were cut (leaf-flap technique – Khan et al. 2002) but the amount of excess 15N found in the soils after 24 h was not significantly different between these treatments and the treatment UEL where uncut leaf tips were immersed in the enriched urea solution.
It is logical to expect that the source of “rhizodeposits” should mostly be root exudates and products of root and nodule turnover and senescence. The expected chronological pattern would thus be a gradually increasing rate of loss of N from roots with time and especially considerable losses from senescent nodules. This was not the pattern observed in this study. The mean quantity of excess 15N deposited in the soil in the first 7 DAL from the urea labelled plants was 5.5 % of that fed to the leaves. Subsequently, over the next 63 days, this only increased further by 1.8 %, amounting to a total on average of 7.2 % (Table 1). Thus, the results suggested that most of the total excess 15N deposited in the soil at the final harvest came from initial leakage from the root, although some of this soluble labelled N could have been reabsorbed.
Labelling with 15N-enriched glutamine appeared to have no advantages over 15N-enriched urea. Initial uptake by the plant was faster but short-term deposition in the soil seemed more severe than was the case with the urea-labelling treatments. For the urea-labelling treatments, an average 4.6 % of the 15N label was deposited in the soil in the first two days; for the glutamine, the amount was estimated to be 11.6 % and, unlike the urea, some of the label appeared to be re-absorbed over the next eight days such that at 10 DAP the amount of excess 15N in the soil decreased to 7.2%.
Initially we utilized two strategies to calculate the NRRN both assuming that the sole sources of 15N excess in the soil were fine roots and nodules. In the first strategy, A, the fine roots were assumed to be the sole source of the N loss and in the second strategy, B, fine roots and nodules were the source. The calculated values for the NRRN, A and B, were respectively, 76 and 107 mg N for the ULL treatment, 61 and 82 mg N for the USLL, and 45 and 71 mg N for the UEL treatment. The estimates for the UEL treatment (intact leaf labelling) were significantly lower (P > 0.05) than those using the leaf-flap technique (treatment ULL).
Few other authors have compared the effect of different labelling techniques on the estimates of NRRN. Khan et al (2002) investigated different concentrations of urea and different methods of labelling (petiole or leaf-flap) fababean (Vicia faba), chickpea (Cicer arietinum), mungbean (Vigna radiata) and pigeonpea (Cajanus cajan) but did not calculate values of NRRN. Other authors have also studied the efficiency and distribution of 15N label in legume and plants (Merbach et al, 2000; Hertenberger and Wanek, 2004; Yasmin et al. 2006) but none of them made comparative estimates of NRRN in plants grown in soil. An exception was the study of Chalk et al. (2002) who found that the injection of labelled N into the hollow stem of Sesbania rostrata resulted in an estimate of NRRN of only 13 % of whole plant N compared to 42 and 56 % for leaf and root immersion, respectively. In this present study, the plants labelled with glutamine gave significantly higher estimates of NRRN than those labelled with urea.
There are three previous evaluations of NRRN for soybean. Rochester et al. (1998) performed a study in the field in New South Wales, Australia. The plants were labelled via the petiole three times at three-week intervals. At pod-filling stage the plant shoots accumulated 431 kg N ha− 1 and belowground N (in recovered roots and NRRN) was estimated to be 275 kg N ha− 1. At maturity, total shoot N increased to 525 kg N ha− 1 and RBGN was estimated to have decreased to 162 kg N ha− 1. This indicates that the RBGN at maturity was approximately 24 % of total plant N. However, their final conclusion was that approximately 40 % of total plant N (the belowground N present at pod-filling stage) remained in the soil and this estimate of 40 % for RBGN thus appears to be unjustified from the results from this study (Rochester et al. 1998).
A further study on soybean by Laberge et al (2009) was performed at two sites in the field in Nigeria, Ibadan in the West and Zaria in the North. Plants were leaf-labelled with urea twice within a three-week period. Soybean yields were not high, at 1032 kg and 1571 kg ha− 1, for Ibadan and Zaria, respectively. The NRRN was estimated to be 11.6 and 18.0 % of the total accumulated plant N, respectively.
Zang et al (2018) estimated the NRRN of soybean in China. The plants were grown in soil in pots in the greenhouse. They used the stem labelling (cotton wick) technique and added 15N-enriched urea solution every two weeks after planting. One harvest was performed at grain maturity and they estimated that NRRN was 23.5 % of whole plant N.
The evidence in our study suggests that the excess 15N derived from the labelled urea that was found in the soil in the first 2 to 7 days was due to leakage, or some form of 15N deposition provoked by the leaf-labelling process, and not to root senescence or other rhizodeposition which would occur in undisturbed plants. In this case the amount of excess 15N deposited in the soil from 7 DAL until the final harvest at 70 DAL falls from 166.1 µg to 40.5 µg excess 15N plant− 1. Using these data, the estimates of NRRN-A and NRRN-B become 14.8 and 22.2 mg plant− 1, respectively, 1.9 and 2.7 % of total plant N. As loss of N from nodules was very significant, we assume that the estimate NRRN-B of 22.2 mg plant− 1 is more likely to be close to actual N loss to the soil. Even so, the proportion of N lost as NRRN (2.7 %) is almost an order of magnitude lower than the estimates of NRRN for soybean by Rochester et al (1998), Laberge et al (2009) and Zang et al (2018) that ranged from 12 to 24 % of total plant N. This estimate is also far lower than those for most other legumes using 15N labelling of leaves or stems (Wichern et al. 2008; Fustec et al. 2010).
The total N in roots and nodules recovered from the soil by sieving amounted to a mean of 51.9 mg plant− 1. Adding this to the estimate of NRRN-B of 22.2, the total residual belowground N (RBGN) was 74.1 mg plant− 1, or 9.1 % of plant total N.
Soybean in Brazil in 2020 had a mean yield of 3,250 kg grain ha− 1. Assuming a typical plant density of 330,000 plants per ha and a concentration of N in the grain of 6.5 %, each plant would accumulate approximately 750 mg N in the grain (250 kg N ha− 1) and assuming a nitrogen harvest index (NHI) of 80 %, a total N accumulation of 935 mg (313 kg N ha− 1) is estimated. In this study the total N accumulated by the plants at the final harvest reached 800 mg which indicates that the plants were almost as large as in the average field crop. For the average Brazilian soybean crop the total residual belowground N after soybean harvest (RBGN) would amount to approximately 28 kg N ha− 1 of which 20 kg N ha− 1 is derived from recovered roots and 8.5 kg N ha− 1 from non-recoverable root N (NRRN).