Introduction

Mineral elements occupy an essential role in human nutrition (Huang et al. 2016). Deficiencies of micronutrients like zinc (Zn) and iron (Fe), recognized as hidden hunger, are the most prevalent health disorders worldwide, affecting nearly two billion people (Promchan et al. 2016). Rice is a major staple crop, supplying the bulk of calories and the majority of daily dietary nutrients for billions of people (Cakmak 2008; Waters et al. 2009). Therefore, enrichment of the staple crops like rice with essential nutrients, especially Zn and Fe, is currently a high-priority research area in crop science.

Biofortification of breeding crops with condensed micronutrients like Fe, Zn, and β-carotene through conventional breeding and modern biotechnology, is considered to be the most sustainable and cost-effective strategy for overcoming hidden hunger (Murgia et al. 2012). In addition, agronomic biofortification is viewed as complementary to breeding strategies for optimizing and ensuring the success of genetic biofortification in the long term (Cakmak 2008). Recent studies have shown that improving nitrogen (N) status of crops may play a pivotal role in uptake of mineral nutrients, and distribution and accumulation of mineral nutrients in edible parts of crops (Erenoglu et al. 2011). Soil and foliar application of N enhanced grain Zn concentration in wheat (Kutman et al. 2011). In rice, N application affected grain composition, and a moderate level of N application was favorable for promoting accumulation of micronutrients like copper (Cu), Fe, manganese (Mn), and Zn in brown rice (Hao et al. 2007). A previous study showed that N promotes Fe accumulation but depresses Zn accumulation in milled rice (Lin et al. 2014). However, the physiological mechanisms of how N affects remobilization of Fe and Zn in plants are largely unknown.

Flag and upper leaves are the main source organs of assimilates and mineral nutrients for grain filling (Sperotto et al. 2009; Zhang et al. 2017). In addition, non-leaf organs such as stems, branches, sheaths, glumes/bracts, and awns also have chloroplasts and photosynthetic activities, therefore these are also considered source organs. In wheat, leaves contributed 40%, glumes 23%, stem 23%, and roots 16% of the N remobilized to and accumulated in grains (Simpson et al. 1983). Pearson and Rengel (1994) suggested that wheat glumes are transient storage organs of Zn and Mn translocated from leaves and subsequently loaded into the developing grain. A study on rice revealed a substantial contribution rate of protein-N from bracts to grains, being as high as nearly 10% (Zhang et al. 2017). However, the remobilization of other nutrients like phosphorus (P), potassium (K), magnesium (Mg), Fe, and Zn from bracts and its contribution to grain mineral nutrient accumulation is largely unknown and requires a further investigation.

The current experimentation was conducted a pot experiment with three levels of N topdressing, and estimate the apparent contribution rate of N, P, K, Ca, Mg, Fe, Zn, Mn, and Cu from vegetative organs including bracts, leaf, and sheath to grains. The objective of the current study was to elucidate the role of rice bracts in accumulation of grain mineral nutrients. In addition, the study of N effect on remobilization of mineral nutrients from source to sink organs and their accumulation in grains was investigated, with the aim of developing an agronomic strategy for fortifying rice grains with Fe and Zn.

Materials and methods

Plant material and experiment design

The pot experiment was conducted in a field environment at the Danyang experimental station, Jiangsu Province, China (31°54′31″N, 119°28′21″E) in the summer of 2016, using a japonica rice cultivar viz. Huaidao 5. The pots were 30 cm in height and 34 cm in diameter, and were filled with 15 kg fine-grained soil. Seeds were sown in the nursery beds on May 30, and 4-week-old seedlings were transplanted to pots on June 28, with each pot having six seedlings.

The soil type was clay soil, containing 25.7 g kg−1 organic matter, 1.1 g kg−1 total N, and 23.8 mg kg−1 available P. The total concentrations of soil P, K, Ca, Mg, Fe, Zn, Mn, and Cu were 519.8, 26298.1, 1140.1, 3077.5, 14621.5, 50.2, 286.8, and 14.1 mg kg−1, respectively. The EDTA extractable K, Ca, and Mg concentrations were 89.7, 89.6, and 601.6 mg kg−1, respectively, and the DTPA extractable Fe, Zn, Mn, and Cu were 40.1, 1.7, 266.9, and 2.5 mg kg−1, respectively. The EDTA extractable K and Zn were in the medium level for the clay soil, while others were among the high levels, as classified by Bao (2000).

Before transplanting, basal fertilizers were applied as 1.0 g N, 1.2 g P2O5, and 0.9 g K2O per pot. N topdressing was split equivalently at two stages, the panicle initiation stage (4th leaf age in reverse order) and the middle stage (2nd leaf age in reverse order). Three levels of N topdressing were applied as follows: (1) LN, low N level (0.8 g N/pot); (2) MN, medium N level (1.6 g N/pot); and (3) HN, high N level (2.4 g N/pot). Each N treatment had 50 pots, and water management was implemented according to approaches for high yielding (Chen et al. 2016).

Sampling and chemical analysis

Samples of the top three leaves and their sheaths, and grains and their bracts were collected at the beginning of grain filling, the 7th day after anthesis (7 DAA), and at maturity (35 DAA). The bracts included lower glume, upper glume, lemma, and palea, as described by Zhang et al. (2017). Enzymes were deactivated by heating at 105 °C for 1 h in an oven. Then the samples were dried to a constant weight at 75 °C for 48 h. After that, dry matter was weighed and samples were milled into powder and stored in plastic bags at room temperature for mineral analysis.

Nitrogen concentration was determined using the semimicro-Kjeldahl method. Inductively coupled plasma optical emission spectrometry (ICP-OES 710; Agilent Technologies, USA) was used to determine the contents of P, K, Ca, Mg, Fe, Zn, Mn, and Cu. Samples (~ 0.50 g) were wet digested with 10 mL HNO3–HClO4 mixed acid (3:1). Argon was used as the make-up gas, and parameters of ICP-OES were as follows: radio frequency power, 1.1 KW; plasma gas flow, 15.0 L min−1; auxiliary gas flow, 1.50 L min−1; nebulizer pressure, 200 N; delay time, 15 s; flush time, 10 s; and read time, 5 s. Standard solutions of the eight mineral elements (1000 ppm, Merck, Germany) were used to obtain the calibration curve, with correlation coefficients more than 0.999 in this study. Each sample was measured in triplicate.

Statistical analysis

Samples were analyzed in triplicate and mean values were used for comparison. The analysis of variance was performed using Least Significant Difference (LSD) test in SPSS 17.0 (Statistical Product and Service Solutions, IBM).

Results

Dry matter accumulation of leaf, sheath, bracts, and grain

Dry matter weight decreased slightly in leaves and their sheaths with the progress of grain filling (Table 1). Similarly, dry matter in bracts was slightly lower at 35 DAA than at 7 DAA (Table 1), indicating the translocation of assimilates in these organs. Nitrogen significantly promoted dry matter accumulation in leaves, with HN exhibiting the largest influence.

Table 1 Dry matter accumulation of leaf, sheath, and bracts and grain yield under three N treatments
Full size table

As expected, N topdressing significantly increased grain yield per plant, leading to more panicles per plant and more grains per panicle (Table 1). Although MN resulted in heavier grains, there was no significant difference between the high and low N treatments. Grain yields of MN (15.27 g/plant) and HN (16.46 g/plant) were significantly higher than that of LN (12.30 g/plant). Although the value of HN was higher than that of MN, there was no statistically significant difference between them.

Concentrations of mineral nutrients in leaf, sheath, bracts, and grain

Concentrations of N, P, K, Mg, Zn, and Cu in leaf, sheath, and bracts decreased as grain filling progressed, and concentrations were higher at 7 DAA than at maturity of 35 DAA (Tables 2, 3). In contrast, concentrations of Ca and Fe were lower at 35 DAA than at 7 DAA. The concentration of Mn increased in leaf but decreased in sheath and bracts from 7 DAA to 35 DAA.

Table 2 Concentrations of macronutrients in leaf, sheath, bracts, and grain under three N treatments
Full size table
Table 3 Concentrations of micronutrients in leaf, sheath, bracts, and grain under three N treatments
Full size table

In general, N topdressing had a positive influence on concentrations of mineral nutrients in leaf, sheath, and bracts both at 7 DAA and at 35 DAA. Concentrations of the nine mineral nutrients in grains decreased as grain filling progressed, being higher at early stage than at late stage (Tables 2, 3). This is attributed to the dilution effect of grain dry matter accumulation, which had a higher rate than those of the nine mineral nutrients. Of interest was that the concentrations of Ca, Mg, Fe, and Mn dropped drastically from 7 DAA to 35 DAA, indicating that these mineral nutrients were translocated to grains mainly at early stage of grain filling. In addition, high N topdressing significantly lowered Fe concentration in mature grains (35 DAA) as compared with LN; however, concentrations of Fe and Zn were not significantly altered by MN (Table 3).

Noticeably, the grain P concentrations of HN treatments were significantly higher than that of LN at 7 DAA (Table 2). Although the P concentrations of grains between HN and LN were not significantly different, there was a relatively lower effect of HN compared to LN at 35 DAA. In addition, the grain P concentrations of HN treatments were statistically lower than that of MN (Table 2), which was consistent with our previous study (Bi et al. 2013).

Net accumulation of mineral nutrients in leaf, sheath, and bracts

The net accumulation of each mineral nutrient was calculated by subtraction of its content (mg/plant or µg/plant) at 35 DAA by its content at 7 DAA. As shown in Fig. 1, the value of net accumulation of N, P, K, Mg, Zn, and Cu in leaf, sheath, and bracts were all below zero, indicating that these mineral nutrients were remobilized from these vegetative organs during grain filling. In contrast, net accumulation of Ca and Fe in leaf, sheath, and bracts, and of Mn in leaf, were all positive numbers, implying that these vegetative organs were also sinks of Ca, Fe, and Mn, similar to the developing grains.

Fig. 1
figure 1

Net accumulation of macronutrients and micronutrients in leaf, sheath, bracts and grain during grain filling (between 7 and 35 days after anthesis) and its variation with N treatments. Data within a column followed by different lowercase letters are significantly different at P < 0.05. Low rate of N topdressing (LN, 0.8 g N/pot); medium rate of N (MN, 1.6 g N/pot); high rate of N (HN, 2.4 g N/pot)

Full size image

Nitrogen topdressing positively affected the net accumulations of N, Ca, Fe, and Mn in leaves and sheaths, while that of P, K, Zn, and Cu was depressed (Fig. 1). In addition, N fertilization had a positive effect on the net accumulation of P in rice bracts but a negative effect on net accumulation in leaves and sheaths.

Contribution rate of leaf, sheath, and bracts to grain mineral nutrient accumulation

As shown above, leaf, sheath, and bracts are source organs, supplying N, P, K, Mg, Zn, and Cu for grain filling. Therefore, we calculated the apparent contribution rate of leaf, sheath, and bracts to grains in terms of the six mineral nutrients. The contribution rate of each mineral nutrient was calculated using the following formula: (apparent translocation of mineral nutrient of a vegetative organ from 7 DAA to 35 DAA)/(mineral nutrient accumulation at 35 DAA − accumulation at 7 DAA in grain). Note that apparent translocation of each mineral nutrient from a vegetative organ was the opposite of net accumulation (Tables 2, 3).

As shown in Table 4, averaged across N treatments, the contribution rates of leaf to grains of N, P, K, Mg, Zn, and Cu were 23.51, 4.70, 21.93, 16.58, 9.79, and 7.19%, respectively. For sheath to grains the contribution rates were 14.69, 11.86, 32.48, 18.27, 26.67, and 4.90%, respectively. Considerable contribution rate of bracts for N, Mg, and Zn was detected, with mean values of 5.96, 12.56, and 12.34%, respectively. Also, the calculation showed minor contribution of bracts with respect to P, K, and Cu, with mean values of 0.99, 3.90, and 3.05%, respectively.

Table 4 Contribution rate of leaf, sheath, and bracts of six mineral elements to grains under three N treatments (%)
Full size table

Nitrogen topdressing lowered the contribution rate of all the mineral nutrients for leaf and sheath. For instance, the contribution rate of leaf N to grains was 27.55% for LN but 20.44% for HN. Also, the contribution rate of K in sheath was 42.26%, 29.77%, and 25.41% for LN, MN, and HN, respectively. In addition, N application tended to decrease the contribution rate of bracts except for P.

Discussion

Remobilization of mineral nutrients from bracts to grain

Nitrogen absorption by cereals occurs mainly before anthesis, with a value of about 85% for rice (Li et al. 2014). Therefore, the major part of grain N is transferred from the vegetative organs. Typically, only 10–30% of N in mature cereal is retained in the senesced leaves, with the major part being remobilized to sink organs like grains (Gregersen 2011). In rice, the remobilization (redistribution) rate of leaf N and its contribution rate to grains were 76.88 and 21.74%, respectively (Zhao et al. 2015). Similar to leaf, bracts of rice spikelets are also important source organs, contributing heavily to grain N accumulation. A previous study showed a considerable contribution rate of protein-N in bracts to grain filling, with values as high as 10.19 and 9.77% for superior and inferior grains, respectively (Zhang et al. 2017). In the present study, averaged contribution rates of total N in leaf, sheath, and bracts to grains were 23.51, 14.69, and 5.96%, respectively. The values are relatively lower than that of Zhang et al. (2017), probably because the current study assayed the total N content of the samples from 7 DAA to maturity, while the former measured the samples in protein-N content from 0 DAA to maturity, a highly removable fraction of leaf total N. Nevertheless, these results jointly suggest an important role of rice bracts in providing N for grain filling.

Magnesium is a crucial component of chlorophyll required for plant photosynthesis. In soybean and wheat, seed/grain Mg appears to be mobilized from leaves, partially because of its high phloem mobility (Maillard et al. 2015). Zinc is also essential for rice plants, and more than half of grain Zn is remobilized from that accumulated and stored in leaf, sheath, and stem before anthesis (Wu et al. 2010; Yilmaz et al. 2016). In this study, a considerable contribution rate of Mg and Zn nutrients of the bracts was detected, with values as high as 12.56 and 12.34%, respectively, in comparison with that of the leaf (16.58 and 9.79%), and sheath (18.27 and 26.67%). This result suggested that the bracts of rice are an important transient organ for the loading of Mg and Zn into developing grains.

In rice, about 25% of P is absorbed and accumulated during grain filling, and the remobilization from vegetative organs accounts for about 20% of the grain P (Li et al. 2014; Julia et al. 2016). Potassium uptake by cereals occurs mainly before anthesis, at about 90% for rice, and vegetative organs including leaves, stem, and sheaths contribute nearly 60% of grain K at maturity (Li et al. 2014). The extent of Cu remobilization within plants was species dependent, and showed the lowest remobilization in rice compared with other crops (Barunawati et al. 2013; Maillard et al. 2015). In the current study, a minor contribution of bracts to grain P, K, and Cu was detected, with contribution rates of 0.99, 3.90, and 3.05%, respectively. Similarly, other vegetative organs like leaf and sheath contributed a small part of P and Cu. These findings indicate that the major part of grain P and Cu should originate from other sources like stem or from root uptake after anthesis, as shown in Fig. 2.

Fig. 2
figure 2

Schematic presentation of allocation of mineral nutrients to grains (a) and the contribution of vegetative organs of leaf, sheath, and bracts to grains (b). Dashed blue line indicates root uptake, solid red line indicates remobilization of minerals from leaf and sheath, and dashed black line indicates remobilization of minerals from bracts. Data are averages of N treatments. Other sources include remobilization of minerals from stem and root uptake. LG lower glume, LM lemma, PL palea, RC rachilla, UG upper glume

Full size image

Calcium is less mobile in the phloem, and its continuous accumulation in leaf even after anthesis has been well established (Himelblau and Amasino 2001; Maillard et al. 2015). Iron is inefficient for remobilization between vegetative organs and grains, and a study on rice showed that only 4% of total shoot Fe was translocated to grain (Marr et al. 1995). Due to poor phloem mobilization, Mn appears to be an immobile nutrient in leaves of wheat (Pearson and Rengel 1994) and barley (Maillard et al. 2015). In the current study, the net accumulation of Ca, Fe, and Mn increased during grain filling, not only in the grains but also for the vegetative organs, in particular the leaves, indicating that leaves are also a sink of these mineral nutrients. Interestingly, concentrations of Ca, Fe, and Mn in grains decreased dramatically during grain filling, with the late stage (35 DAA) showing a two to threefold decrease in comparison with the early stage (7 DAA). Since the leaf, sheath, and bracts are also sink organs, other sources including remobilization from stem and uptake by root should be crucial for the accumulation of Ca, Fe, and Mn in rice grains.

In summary, the patterns of the remobilization of N, P, K, Mg, Zn, and Cu during rice grain filling were summarized according to the results of the current study (Fig. 2). Of note is that the contributions of leaf, sheath, and bracts only account for part of the amount of these mineral nutrients in grains, and other sources including remobilization from stem and uptake by root are also essential for grain mineral nutrient accumulation.

Effect of N fertilization on grain mineral nutrients

Nitrogen fertilization, especially topdressing at the panicle initiation stage, is widely adopted as a high-yielding technology by rice growers (Ogawa et al. 2016). While rice plant tends to maintain the homeostasis of mineral nutrition in grains within a specific range, application of N fertilizers can alter the balance. For rice grain, the medium level of N application promoted the accumulation of Cu, Fe, Mn, and Zn in brown rice (Hao et al. 2007). Lin et al. (2014) found that addition of N fertilizers tended to lower Zn accumulation in brown rice, while the trends for Cu, Fe, and Mn followed no specific pattern. Jaksomsak et al. (2017) reported that increasing N rate increased grain Zn concentration and yield in the high-yield/low grain Zn cultivars, but depressed grain Zn concentration and increased grain yield in the low-yield/high grain Zn cultivars. In this study, grains produced at the MN topdressing level had higher Fe and similar Zn concentration in comparison with those at the HN level. The Zn concentration of the japonica cultivar Huaidao 5 used in the current study was similar to that designated as low grain Zn cultivars by Jaksomsak et al. (2017), but exhibited a different pattern, suggesting genotypic differences in the response of grain Zn to N fertilization. Importantly, there was no statistically significant difference in grain yield between MN and HN, indicating the possibility of appropriate N management for maintaining Fe and Zn level under high yielding conditions.

Conclusions

This study quantified the contribution of rice bracts to grain in terms of minerals, and examined the effect of nitrogen topdressing. The role of bracts was clarified, providing a substantial portion of N, Mg, and Zn for grains, while being a sink organ of Ca and Fe, as their net increase during grain filling. In addition, N topdressing depressed the remobilization of N, Mg, Zn, K, and Cu from bracts to grains but increased that of P. Collectively, these findings suggests the necessity of studying on bracts when elucidating the physiological mechanism underlying grain filling, a crucial process for the formation of both rice yield and quality.