1. Introduction
Where it is environmentally and socio-economically possible, rice (
Oryza sativa L.) is grown under continuous soil flooding (CF) in the so-called lowland rice systems. Therefore, even if the average value of physiological water productivity of rice is comparable to that of the other major C3 cereal crops, higher total inputs of water are required, so that about 40% of the global irrigation freshwater is used in rice paddy fields [
1,
2,
3,
4]. Global warming and competition between lowland rice and other seasonal crops, together with industrial and civic requirements for water, may cause physical and/or economic water shortages for rice cultivation [
5]. It is expected that by 2050 several million hectares of currently lowland irrigated rice systems will experience water scarcity [
3]. A further critical issue in lowland rice systems is the establishment of ideal conditions for emission into the atmosphere of greenhouse gases, mostly methane [
6,
7]. It follows that the development of water-saving techniques is essential to improve the sustainability of irrigated rice systems [
3]. Among them, “Alternate Wetting and Drying” (AWD) is a management practice developed by the International Rice Research Institute (IRRI) [
8] that is becoming increasingly widespread.
In the CF condition, excluding a short period to allow weed control, constant pond water is maintained in the field until the pre-harvest drainage. In AWD, rice paddies are intermittently submerged and dried so that the upper soil layers switch from anaerobic to aerobic conditions several times during the crop growing season. In order to avoid yield reductions due to drought stress responses, in AWD the field is usually re-submerged before the values of the soil water potential (Ψ
w) in the rooting zone become lower than −20 kPa [
9]. Despite the original protocol issued by IRRI [
10], several variants of paddy field water management are reported as AWD; they differ in number, severity, and timing (referred to the crop phenological phases) of the dry periods, resulting in different effects on yield and water use efficiency [
9]. Several studies demonstrated that AWD could markedly (up to 90%) reduce the global warming potential of the gaseous emissions from rice fields and improve water use efficiency up to more than 60% [
9,
11]. Depending on severity and duration of soil dryness, number of the dry-wetting cycles along the season, cultivar adopted, and local weather trends, AWD does not affect [
12], lowers [
13] or even increases [
14] yield compared to CF. Recently, it has been put in evidence that soil properties and particularly duration and degree of soil drought induced by the dry periods are the most effective factors that influence yield under AWD [
9]. A study on a large panel of rice accessions usually grown under CF also suggested genotypic influences on growth and yield performance under AWD [
15].
Although it is well known that the soil water status during rice growth might affect qualitative traits of the grains [
16,
17], to our best knowledge only a recent paper [
18] reported on the possible effects of AWD on grain quality and nutritional value.
Due to the presence in the bran of several bioactive molecules able to protect against age-related pathologies [
19], the dietary consumption of brown grain rice is advised, expanding the market of this commodity. The accumulation of vitamins, health-promoting metabolites and mineral microelements in the grain is regulated by genetic and environmental factors as well as by their interactions [
19,
20]. Conventional and modern breeding tools, specific agronomic techniques and postharvest treatments offer opportunities to improve the nutritional value of rice grain. Nevertheless, the actual environmental sustainability of these solutions is not always ascertained [
21].
Aim of this work was to verify, by a side-by-side comparison, the hypothesis whether AWD, as influenced by the different environmental conditions in different seasons, could affect the concentrations of metabolites with health-related properties, minerals and toxic trace elements in the brown grain of three temperate japonica rice cultivars usually grown in Italy under CF. The three cultivars selected belong to different market segments: ‘Baldo’, largely used for the typical Italian dish risotto, ‘Gladio’, particularly suited for parboiling, and ‘Loto’, with several applications in the industrial production of rice-derived foods.
The study was conducted by relying entirely on the water supply (i.e., soil water status) available in the two years considered (2012 and 2013), as made possible uniquely by the weather conditions and the rotational turns in water distribution. In fact, the in advance-fixed irrigation turns stated in most of the European rice areas by the local Authorities that manage the distribution of water among fields can limit the possibility to reflood promptly when the soil Ψ
w drops below the value commonly considered as the threshold for a yield-safe AWD [
3,
9].
4. Discussion
Among the agronomic techniques able to improve the sustainability of the rice systems, AWD complies the increasing requirement of water saving practices. Nevertheless, knowledge of the possible effects of the soil water status induced by this irrigation regime on the qualitative traits of the rice grain is still scanty.
The present work intended to deepen the knowledge on this topic by verifying the hypothesis whether AWD, as influenced by the different environmental conditions experienced in different seasons, could affect the concentrations of metabolites with health-related properties, minerals and toxic trace elements in the brown grain of three temperate japonica rice cultivars.
The very different weather conditions in 2012 and 2013 and the rotational turns in water distribution imposed by the local Irrigation Water Managers determined two very different AWD conditions (
Figure 1) than can be classified as severe AWD and mild AWD [
9] in 2012 and 2013, respectively.
In rice, panicle initiation and flowering are the developmental stages most sensitive to environmental conditions, and particularly to low soil Ψ
w and air temperature. Although the phenological patterns of the three cultivars are different (‘Loto’ and ‘Gladio’ early cultivars, ‘Baldo’ a late one), in 2012 both panicle initiation and flowering of all cultivars occurred in flooded soil condition. In 2013 all cultivars reached the panicle initiation stage during a period of mild soil drying and flowered in flooded soil condition. In this context, it appears reasonable to exclude that the differences concerning the yield responses to AWD of the three cultivars (‘Baldo’ and ‘Gladio’ sensitive, ‘Loto’ insensitive) are primarily due to differences in the soil water content during critical developmental phases. The yield losses in ‘Baldo’ and ‘Gladio’ were not related to the severity of the AWD drying periods, since in ‘Gladio’ the negative effect was even more pronounced in the mild AWD condition of 2013. The yield components were affected differently by AWD in the three cultivars: The number of seeds per panicle, particularly in ‘Gladio’, and the weight of the grains, particularly in ‘Baldo’, contributed to a greater extent to the observed yield losses in AWD. Seeds harvested from ‘Gladio’ grown in AWD were smaller because they were basically shorter (
Table 1).
The negative effects of AWD on yield and its components, when present, seemed related rather to the cooler temperatures during the growing periods than to the severity of soil drying (
Figure 1 and
Supplemental Figure S1). In 2013, in fact, independent of the value of soil Ψ
w, the concomitant low temperatures and absence, in AWD, of ponded water with its thermal buffering effect exposed more severely the plants to the effects of the daily temperature excursions, that negatively affect yield [
32]. In the same year, the generalized delay in the flowering time of the three cultivars in AWD in comparison to CF shifted the grain-filling period towards cooler days. Since the overall length of the filling time (
Table 1) of each cultivar was unchanged, the temperature factor could explain the greater effect on grain weight recorded in 2013 in the two sensitive cultivars (
Table 1).
Possible negative effects of AWD on yield have been ascribed to decreased leaf photosynthetic rate and general metabolic activity, influenced by the leaf N nutritional status [
12,
33]. The Dualex® CHL and NBI indexes were similar in AWD and in CF (
Table 2); therefore, we might reasonably exclude that in our conditions AWD determines early leaf senescence and/or N deprivation, capable to affect the leaf source function for the developing grains.
Apparent amylose content and N-protein concentration are indicators of rice grain quality, affected by genetic and environmental determinants, like temperature and N availability during the grain-filling period [
34]. These parameters were not modified substantially by AWD in any cultivar (
Table 2), allowing us to exclude effects on the sink strength of the developing grains [
12].
Flavonoids, tocols, γ-oryzanol, and phytic acid are major antioxidants in rice brown grain [
30]. The concentrations of total flavonoids in AWD tended to increase (
Table 3). The essentially constant values of CHL and NBI indexes suggest, in agreement with Cerovic et al. [
23], that the flavonoid concentrations in the flag leaf did not change in any genotype and water management regime, allowing to hypothesize that the microclimate condition established by AWD might stimulate the biosynthesis of flavonoids in the grains.
Under CF, the total tocols concentrations in the brown grain were close to the lowest limit of the range (15–60 mg kg
−1) reported for non-pigmented rice varieties [
30]. The antioxidant properties of tocols contribute to stress tolerance in plants, explaining their accumulation under environmental constraints [
35,
36]. In ‘Baldo’ and ‘Loto’, AWD markedly increased the total tocols concentrations and seemed to increase all T3s, with particular regard to γT3 (the most abundant tocol in brown grain). In ‘Gladio’, lack of effect of AWD on total tocols concerned all of their components (
Table 3 and
Table 4). In our material, the increases in the concentrations of Ts and T3s in AWD were particularly apparent in ‘Loto’ and essentially absent in ‘Gladio’ (
Table 4). This effect may be related to a higher general sensitiveness of this genotype to the stressful environment of AWD; it may be hypothesized that the growth condition generated by this water regime somehow affects different specific regulatory points of the tocols metabolic pathway, known to be finely tuned [
37,
38].
Rice γ-oryzanol is a mixture of 25 so far identified ferulate esters of triterpene alcohols with antioxidant properties [
26]. Out of the three cultivars studied, only ‘Loto’ showed a marked AWD-induced enhancement of total γ-oryzanol (
Table 3) mainly due to a significant increase in 24-MeCAF (
Table 5), consistent to what described for rice brown grain under different stressful environmental conditions [
35]. A wide genetic variability in the environment-related accumulation of γ-oryzanol is described in several japonica rice cultivars [
39], possibly explaining the peculiar behavior observed in ‘Loto’.
The antioxidant activity of rice brown grain usually falls in the range 0.2–20 mmol TEAC kg
-1 [
30]. Only in ‘Loto’ this activity was increased in AWD (
Table 3), consistent with the effects of this water regime on the antioxidants levels. It is interesting to stress that ‘Loto’, insensitive to AWD in terms of yield (
Table 1), showed a higher capability to respond, through the accumulation of antioxidant compounds, to the probable stress conditions induced by this water regime.
Phytic acid is acknowledged to have antinutritional but also antioxidant properties [
31,
40]. In our material, the levels of phytic acid were not affected by AWD in any cultivar and in any year (
Table 3), according to the general lack of effect of the water regime on the concentrations of total P (
Table 6).
Whilst excellent source of carbohydrates, rice grains are poor in inorganic micronutrients whose dietary supply dramatically drops after processing to produce white grains. Consumption of brown rice could enhance the intake of essential inorganic micronutrients, but precautions should be taken since toxic trace elements can also accumulate in the bran. On this basis, several efforts to increase the presence of beneficial microelements (e.g., Fe and Zn) and limit the accumulation of hazardous ones (e.g., As and Cd) in rice grains have been made by applying specific fertilization practices, exploiting the existing natural variation within germplasm, or using transgenic approaches [
41]. Since also the soil redox conditions affect the solubility and availability for plants of several mineral elements, AWD, by altering, compared to CF, the soil redox status, may represent an interesting strategy to control the concentrations of healthy/hazardous mineral nutrients in the grain. In our material, AWD did not substantially modify the accumulation of K
+, Ca
2+, and Mg
2+ even if soil aerobiosis is expected to diminish their release into the soil solution, on the contrary favored by anaerobiosis in prolonged flooding [
42]. In spite of the aerobic/anaerobic fluctuations in the soil, since the amounts of these cations are always relatively high in relation to the plant requirements, changes in their uptake and accumulation are not expected.
The ionomic results indicate that, in our conditions, AWD, compared to CF, caused changes in the grain concentrations of those elements whose redox status and/or solubility are a function of the soil redox status. In major detail, alternation from anaerobic to aerobic soil conditions would determine: i) shifts of Cu
2+, Ni
2+, and Cd
2+ from their insoluble sulfide salts to the soluble sulfate ones [
43,
44,
45]; ii) shifts of the As(III) water-soluble, mobile species to As(V), strongly adsorbed on soil (hydr)oxides and poorly mobile [
46] and iii) changes in the balance between soluble Mn
2+ salts and insoluble Mn oxides [
47]. The dynamics of Fe and Zn species as a function of the soil redox conditions and the interaction of these elements with the root system are quite complex and can alter their bioavailability [
47,
48], explaining the null or weak effect of the aerobic periods on their accumulation in the grain.
For a few elements, our results are different from those recently reported by Xu and coworkers [
18], who compared CF with a moderate soil dry irrigation. In particular, for Mn and Zn their results are essentially consistent with ours, whereas these authors reported that Cu and Fe concentrations decreased in the water-saving regime and in our conditions the Cu concentrations increased and Fe did not show any change (
Table 6). These discrepancies may be possibly due to differences in the soil physico-chemical properties and in the rice genotypes considered.
Different effects of AWD on the concentrations of Cd and As are reported in the literature. The behavior observed in our conditions (increased Cd accumulation and decreased As accumulation in AWD in both years) is in agreement with the results reported by Norton et al. [
49], but is different, for what concerns Cd, from the results by Xu et al. [
18], who did not observe any effect of AWD, and, for what concerns As, from the results of Carrijo et al. [
50], who observed effects on As accumulation only in severe AWD. Most probably these differences depend on the different timing and severity of the imposed AWD condition.
For both Cd and As we observed also a strong seasonal effect. The higher Cd concentrations in AWD in 2012 compared to 2013 may be due to the different duration and severity of the third dry period, which occurred during the grain-filling phase. Indeed, in rice the largest fraction of Cd in the grain derives directly from the roots through a xylem–phloem bypass in the upper node [
51]. Under -oxic conditions, the lower availability of Mn in the soil induces overexpression of the NRamp5 Mn transporter, which is also the main route for the entry into the root of Cd [
51], whose solubility is enhanced in aerobic soils. This -oxic condition was of longer duration and higher severity in 2012 than in 2013, explaining the significant season effect reported for Cd. For As, to our best knowledge such a detailed picture is not reported. We might only speculate that the general lower accumulation of As in AWD in 2013 was due to multiple, possibly interacting factors, like the metabolism of As-reducing microorganisms, O
2 availability, soil pH and redox conditions, involved in the equilibrium of As(V)/As(III) [
52] and all possibly affected by the different weather conditions.
Consistent with their low concentrations in the soil used in our experiments (
Supplemental Table S1), the levels of Cd (particularly under AWD) or As (particularly under CF) in the brown grain of each cultivar were much lower than the maximum EU limits (0.2 mg kg
−1 for both) for commercialization of white rice. Therefore, according to our results, consumption of brown rice would not constitute a hazard for the consumer’s safety for what concerns Cd and As. Different considerations should be made for rice grown on soils with higher levels of the available fractions of these elements: AWD would be counterproductive for Cd but advisable for As.
Finally, considering the allergenic risks related to Ni and the increase in its levels observed in all cultivars under AWD, attention should be devoted to the consumption of brown rice from plants grown under AWD.
5. Conclusions
Due to the consumer’s increasing preference towards healthy eating habits, the brown rice market is currently expanding justifying the adoption of practices able to improve its nutritional quality while also preserving the environment.
AWD induced in the rice varieties considered an improvement of the nutritional quality of the brown grain (higher concentrations of health-promoting molecules like components of vitamin E, γ-oryzanol and flavonoids), which could indeed represent a plus in the rice market, with marked (‘Baldo’ and ‘Gladio’) or null (‘Loto’) effects on yield.
Concerning inorganic microelements for human nutrition, AWD did not, or only slightly, alter the grain density of the most important ones, i.e., Fe and Zn.
AWD exerted opposite effects on potentially toxic/allergenic elements, favoring the accumulation of Cd or diminishing that of As in the grain. This factor must be considered where soils and/or surface water contain these toxic elements. The marked increases in Ni levels observed in AWD should also be considered carefully when addressing to consumers who require Ni-free food.
AWD seemed to slow down the rate of plant growth and development through exposure to cooler temperatures for the absence of ponded water in the field. In temperate rice growing areas, the choice of rice cultivars for AWD should therefore rely on the ones suited better to low temperatures than to drought. In our conditions, ‘Loto’ showed the best maintenance of yield and grain nutritional quality under AWD.
The present study offers a promising perspective for combining the environmental-friendly water saving technique AWD with improved rice grain nutritional quality.