Silicon nutrition mitigates the negative impacts of iron toxicity on rice photosynthesis and grain yield

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv

Silicon nutrition mitigates the negative impacts of iron toxicity on rice photosynthesis and grain yield Martielly S. dos Santosa,b, Lílian M.V.P. Sanglarda,c, Marcela L. Barbosaa, Filipe A. Namoratoa, Danilo C. de Meloa, William C.G. Francoa, Junior P. Pérez-Molinaa,d, Samuel C.V. Martinsa, Fábio M. DaMattaa,∗ a

Departamento Departamento de Biologia Vegetal, Universidade Federal de Viçosa, 36570-900, Viçosa, MG, Brazil Departamento de Ciências Biológicas, Universidade Estadual de Santa Cruz, Rodovia Jorge Amado Km, 16, Ilhéus, BA, Brazil c Australian Research Council Centre of Excellence in Plant Energy Biology, University of Western Australia, Crawley, WA, 6009, Australia d Escuela de Ciencias Biológicas- Universidad Nacional Costa Rica, 863000, Costa Rica b

A R T I C LE I N FO

A B S T R A C T

Keywords: Crop yield Growth Mineral composition Oryza sativa

Excess iron (Fe) is commonly observed in wetland rice (Oryza sativa L.) plants, impairing crop growth and productivity. Some information suggests that silicon (Si) can reduce Fe content in leaves and roots of rice (vegetative phase), but nothing is known if Si could mitigate the effects of Fe toxicity on rice production and photosynthesis. Here, we assessed the role of Si in alleviating the well-known effects of Fe toxicity on nutritional imbalances, biomass accumulation, photosynthesis and grain yield using two rice cultivars having differential abilities to tolerate excess Fe. Plants were hydroponically grown under two Fe levels (25 μM or 5 mM) and the nutrient solutions were amended with Si (0 or 2 mM). Under excess Fe were detected (i) nutritional deficiencies, especially of calcium and magnesium in leaves; (ii) negligible changes in grain nutritional composition, independently of Si application; (iii) decreases in net photosynthetic rates, stomatal conductance and electron transport rate, in parallel to decreased grain yield components (total grain biomass, 1000-grain mass, percentage of filled grains, number of grains per plant and harvest index), especially in the Fe-sensitive cultivar. These impairments were partially reversed by the application of Si. Results also suggest that Si alleviated the negative impacts of Fe on spikelet sterility. In summary, we conclude that the use of Si can be recommended as an effective management strategy to reduce the negative impacts of Fe toxicity on rice photosynthetic performance and crop yield.

1. Introduction

Fofana, 2009; Olaleye et al., 2001; Stein et al., 2009). In West African countries, for example, Fe toxicity can lead to rice yield losses of 50% (Audebert and Fofana, 2009). Although yield decreases can be partially avoided by using rice cultivars with superior tolerance to excess Fe, progress in developing high-yielding cultivars with adequate tolerance to Fe toxicity has been generally slow due mostly to large genotype × environment interaction and field heterogeneity, which makes rice selection ineffective (Sikirou et al., 2015). The critical level of Fe toxicity in plant tissues varies from 300 to 500 mg Fe kg−1 dry mass (Dobermann and Fairhurst, 2000). The excess of Fe can lead to direct toxicity in addition to causing nutritional imbalances (indirect effects), as found in rice cultivars during the vegetative and reproductive stages (Müller et al., 2015; Sahrawat, 2004). It is well established how these nutritional disorders (e.g., deficiency of P, K, Ca, Mg, Mn under excess Fe), usually evaluated in leaves and roots,

Iron (Fe) toxicity is recognized as one of the most widespread mineral disorders in lowland rice production (Audebert and Fofana, 2009) due to excess ferrous iron (Fe2+) formation in reduced soils. In the vegetative stage, the effects of Fe toxicity on rice plants have been associated with decreases in net CO2 assimilation rate (A) due to stomatal and non-stomatal limitations of photosynthesis (Pereira et al., 2013), including photochemical impairments of photosystem II (PSII) as noted through PSII photoinactivation and/or photooxidation (Müller et al., 2015; Pinto et al., 2016). In the reproductive stage, excess Fe leads to significant reductions in both the number of tillers and spikelet fertility, thus ultimately decreasing rice grain yields. These losses may be remarkable, depending on the cultivar, time of Fe toxicity, stress intensity and management strategy (e.g., mineral fertilization) (Audebert and

∗

Corresponding author. E-mail address: [email protected] (F.M. DaMatta).

https://doi.org/10.1016/j.ecoenv.2019.110008 Received 2 June 2019; Received in revised form 2 November 2019; Accepted 25 November 2019 0147-6513/ © 2019 Elsevier Inc. All rights reserved.

Please cite this article as: Martielly S. dos Santos, et al., Ecotoxicology and Environmental Safety, https://doi.org/10.1016/j.ecoenv.2019.110008

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

2. Materials and methods

are reflected in rice crop yields (Audebert and Fofana, 2009; Olaleye et al., 2001; Stein et al., 2009) but much less is known on the effects of Fe toxicity on nutrient distribution to rice grains, which would have obvious impacts on human nutrition. Silicon (Si), per se, is not considered an essential nutrient for higher plants, although it is found at high concentrations in several species, especially those of the Poaceae family (Coskun et al., 2019). Rice, in particular, is a typical Si accumulating species and can be able to accumulate Si at concentrations as high as 10% of shoot dry mass (Ma and Takahashi, 2002). This ability has been associated with positive effects on rice grain yields given that Si can increase both resistance to lodging and the erectness of leaves, which allows better light transmittance through rice canopies with positive impacts on whole-plant photosynthesis (Tamai and Ma, 2008). Nevertheless, Si supply also seems to have a direct effect per se on rice crop yields. It was previously reported that Si application improves grain production by increasing straw biomass, the number of spikelets per panicle, and especially the percentage of filled spikelets, with no significant effect on the panicle number or the 1000-grain mass (Ma et al., 1989; Tamai and Ma, 2008). More recently we also reported an increased number of spikelets in Si-treated plants (Detmann et al., 2012); however, in contrast with the results of Ma and colleagues, a high crop yield was associated with higher grain number and higher 1000-grain mass, whereas both straw biomass and the percentage of filled spikelets remained unaltered (Detmann et al., 2012). Lavinsky et al. (2016) suggested that proper levels of Si in panicle structures would play an as yet unidentified role culminating with higher rice grain number and size. We proposed that higher crop yields in Si-treated rice plants led to an increased sink strength that, in turn, exerted a feed-forward effect on A that is fundamentally associated with increased leaf conductance to CO2 diffusion from the atmosphere to the carboxylation sites in chloroplasts (Detmann et al., 2012; Lavinsky et al., 2016). The positive effects of Si application on rice are also evidenced by increased tolerance to various biotic (Debona et al., 2017; Perez et al., 2014) and abiotic stresses such as toxic metals (Abbas et al., 2015; Coskun et al., 2019; Sanglard et al., 2014; Song et al., 2014). In fact, the negative impacts of toxic metals/metalloids (e.g. Cd, Zn, Al, Fe, As) have been attenuated by Si application given that Si can reduce the uptake, translocation and bioconcentration of these metals/metalloids in the shoot (Rizwan et al., 2016), with positive impacts on the photosynthetic performance, as observed by Sanglard et al. (2016) in rice plants challenged with As. In the case of Fe toxicity in particular, some information suggests that Si can mitigate the effects of toxicity on rice via decreased Fe concentrations in both leaf and root tissues, increased activity of the antioxidant system with concordant reductions in lipid peroxidation, which ultimately leads to a lower impact of Fe on the growth of Si-treated plants (Chalmardi et al., 2014; Dufey et al., 2014). These results, however, were obtained in rice plants at the vegetative growth stage and, therefore, little or nothing is known on the potential mitigating effects of Si in relation to excess Fe on the physiology and production of rice plants at the reproductive growth stage. Accordingly, little is known on the possible impacts of Si on rice grain nutritional composition under excess Fe. Recently, Nagula et al. (2016) reported that Si application (soil or foliar spray) led to decrease in the contents of Fe and Mn in rice grains, but major mineral elements were not analyzed in this study. Given the facts described above, we hypothesized that Si is able to mitigate the impairments of Fe toxicity on rice plants during their reproductive growth phase. To test this hypothesis we designed a hydroponic experiment using two rice cultivars with differential abilities to tolerate excess Fe, and the nutrient solutions were then amended with Si. Two major questions were addressed: (i) can Si application be able to alleviate the effects of Fe toxicity with positive effects on photosynthesis? (ii) will Si application improve rice grain yields and alter nutritional composition of rice tissues under Fe toxicity?

2.1. Plant material, growth conditions and experimental design The experiments were conducted in Viçosa (20°45′S, 42°54′W, 650 m altitude) in south-eastern Brazil. Two lowland rice cultivars (BRIRGA-409 and EPAGRI-109, characterized as sensitive and tolerant to excess Fe, respectively (Stein et al., 2009; Stein et al., 2014)) were used. Details of seed germination and early seedling growth (before transplanting) have been described by Dallagnol et al. (2011)). The plants were grown in non-aerated culture solutions, which were prepared based on Hoagland and Arnon (1950) with modifications. They included the following nutrients: 1.0 mM KNO3; 0.25 mM NH4H2PO4; 0.1 mM NH4Cl; 0.5 mM MgSO4.7H2O; 1.0 mM Ca(NO3)2; 0.30 μM CuSO4.5H2O; 0.33 μM ZnSO4.7H2O; 11.5 μM H3BO3; 3.5 μM MnCl2.4H2O; 0.1 μM (NH4)6Mo7O24.4H2O; 25 μM FeSO4.7H2O and 25 μM EDTA bisodic. The pH of the nutrient solutions was adjusted daily to 5.0. The plants were grown in plastic pots with 5 l volume in a greenhouse with controlled air temperature (30/25 ± 2 °C (day/ night)) and naturally fluctuating air relative humidity and photosynthetically active radiation (PAR) (maximum PAR was approx. 1500 μmol photons m−2 s−1 inside the greenhouse at midday). Si was supplied (2 mM; +Si plants) or not (0 mM; –Si plants) over the entire experiment as monosilicic acid, which was prepared by passing potassium silicate through cation-exchange resin (Amberlite IR120B, H+ form; Sigma-Aldrich, São Paulo, Brazil). Additionally, plants were divided into two groups: one received regular Fe supply (in the form of Fe-EDTA (FeSO4)) over the course of experiment (25 μM, –Fe plants), and in the other group Fe supply was progressively increased (0.5 mM per week) until reaching 5 mM (+Fe plants) at 75 days after transplanting (reproductive stage, R2 stage characterized by panicle elongation). The high Fe concentration corresponds to typical environmental conditions of excess Fe in relevant Brazilian rice producing areas (Stein et al., 2014). The plants were grown under these Fe levels until the end of the experiments. Photosynthetic measurements were performed at the R6 stage (grain filling, approximately at 117 days after transplanting), whereas growth traits and crop yield were assessed at the end of the cultural cycle, approximately at 132 days after transplanting. 2.2. Tissue nutritional analyses Flag leaves, roots and grains were collected and oven-dried at 60 °C until constant mass. Roots were previously washed in a dithio-citratebicarbonate solution (DCB) for removing the Fe that is adsorbed to the tissues. Root tissues were maintained under constant shaking in DCB solutions during 3 h. Grains were separated into husk and endosperm using a mechanical rice peeler. The oven-dried plant materials were ground from which Ca, Mg, P, K, S, Fe, Zn, Cu, P and K were extracted in an acidic medium containing concentrated HNO3 and 60% HClO4, v/ v (nitroperchloric digestion) at 95 °C. The contents of Ca, Mg, Fe, Mn and Zn were determined by atomic absorption spectrometry; K was analyzed by flame photometry; P was determined though the reduction of phosphomolybidate by vitamin C; and S was determined by turbidimetry. Further details for these analyses have been described elsewhere (Malavolta et al., 1989). Si contents were colorimetrically determined on 0.1 g of dried and alkali-digested tissues (Korndörfer et al., 2004). 2.3. Photosynthetic measurements Photosynthetic parameters were assessed in flag leaves after 42 days of the last excess Fe application. The net CO2 assimilation rate (A), stomatal conductance to water vapor (gs), and internal CO2 concentration (Ci) were measured simultaneously with chlorophyll (Chl) a fluorescence parameters on the attached flag leaves using two cross2

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

Fig. 1. The effects of silicon, Si (0 or 2 mM: Si or + Si, respectively), and iron, Fe (25 μM or 5 mM: Fe and +Fe, respectively), on the leaf, root and grain (husk and endosperm) concentrations of Si (A–B) and Fe (C–D) of two rice cultivars [tolerant cv. ‘EPAGRI-109’ and sensitive cv. ‘BR-IRGA-409’ to excess Fe] grown in nutrient solutions. n = 5 ± SE.

2.4. Growth traits and crop yield

calibrated portable open-flow gas exchange systems (model LI-6400XT) equipped with a fluorimeter (LI-6400-40, LICOR Inc., Lincoln, NE, USA). Measurements were conducted from 10:00 to 14:00 h (solar time), which is when A was at its maximum, under artificial PAR, i.e., 1000 μmol photons m−2 s−1 at the leaf level and an external CO2 concentration of 400 μmol mol−1 air. During the measurements, the block temperature was set at 25 °C and leaf-to-air vapour pressure deficit was maintained at approx. 1.0 kPa, while the amount of blue light was set to 10% of PAR in order to maximize the stomatal aperture. Previously dark-adapted (30 min) leaf tissues were illuminated with weak modulated measuring beams (0.03 μmol m−2 s−1) to obtain the initial fluorescence (F0). Saturating white light pulses of 8000 μmol photons m−2 s−1 were applied for 0.8 s to ensure maximum fluorescence emissions (Fm), from which the variable-to-maximum Chl a fluorescence ratio, Fv/Fm = [(Fm − F0)/Fm)], was calculated. In lightadapted leaves, the steady-state fluorescence yield (Fs) was measured following a saturating white light pulse (8000 μmol m−2 s−1; 0.8 s) that was applied to achieve the light-adapted maximum fluorescence (Fm’) following the procedures described by Genty et al. (1989). The actinic light was then turned off, and a far-red illumination was applied (2 μmol m−2 s−1) to measure the light-adapted initial fluorescence (F0’). Using the values of these parameters, the capture efficiency of excitation energy by open PSII reaction centers (Fv'/Fm’) was estimated as Fv'/Fm’ = [(Fm’ – F0')/Fm']; the coefficient for photochemical quenching (qP) was calculated as qP = [(Fm’ − Fs)/(Fm’ – F0')]; and the actual quantum yield of PSII electron transport (ΦPSII) was computed as ΦPSII = (Fm’ – Fs)/Fm’, from which the electron transport rate (ETR) was calculated as ETR = ΦPSII*PAR*f*α; f is a factor that accounts for the partitioning of energy between PSII and PSI and is assumed to be 0.5, which indicates that the excitation energy is distributed equally between the two photosystems, and α is the leaf absorptance by the photosynthetic tissues and is assumed to be 0.84 (Maxwell and Johnson, 2000).

At the end of the experiment, plants were harvested and separated into stems, leaves, roots, panicles and grains. Total leaf areas were measured with an area meter (Area Measurement System, Delta-T Devices, Cambridge, UK). Plant tissues were then oven-dried at 60 °C until constant mass, after which the dry masses of the vegetative and reproductive parts were determined. The specific leaf areas were calculated by the ratio between total leaf area and leaf biomass. The number of leaves, panicles and tillers, the total grain yield, the percentage of filled spikelets, the 1000-grain mass, and the harvest index were also determined. 2.5. Statistical analysis The data obtained were analyzed using a completely randomized design following a 2 × 2 × 2 factorial scheme (two cultivars x two Si levels x two Fe levels) with six plants in individual pots (48 in total) per treatment combination serving as conditional replicates. The data were subjected to an analysis of variance (three-way ANOVA with all main factors evaluated as fixed factors; see Tables S1 and S2) performed using the MIXED procedure of SAS (version 9.1) and an α = 0.05. Regression analysis was used to examine the relationships between some variables. In addition, a multivariate analysis was performed to examine the correlations among photosynthetic and production traits. Data were subjected to a principal component analysis (PCA) in order to reduce dimensionality of data containing a large set of variables in addition to maximizing the variance of linear combinations of these variables. Results of this analysis were examined with a biplot graph using the first two principal components (PC1 and PC2) derived from the PCA. The treatments that produced similar plant responses were clustered using a multivariate technique of grouping analysis according to the method of Tocher, which is based on average Euclidean distances (Rencher, 2002). PCA analysis was performed using the software 3

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

package “factor extra” R (Kassambara and Mundt, 2016).

and harvest index; Table S2, Table 1). These reductions were more pronounced in the Fe-sensitive cultivar; with exception of biomass of vegetative organs, these reductions were partially reversed by the application of Si, as particularly noted by the higher grain biomass of +Si + Fe plants than in their –Si + Fe counterparts (decrease of 44% for the Fe-sensitive cultivar, and 26% for the Fe-tolerant cultivar). In addition, Si application improved the phenotypic appearance of panicles and grains in +Fe plants, particularly in the Fe-sensitive cultivar (Fig. S3). Regardless of Fe treatments, Si had a direct positive effect on A and gs (in parallel with an ETR invariability in the –Fe plants) (Fig. 2, Table S2); the values of these parameters were, on an average, 20% higher in +Si–Fe plants than in their –Si–Fe counterparts. Regardless of the applied treatments, A correlated significantly with gs (r2 = 0.87, P ≤ 0.01) and ETR (r2 = 0.92, P ≤ 0.001) (Fig. 3A–B). In addition, there was also a close relationship between Fe contents with A (r2 = 0.86, P ≤ 0.01) and with gs (r2 = 0.79, P ≤ 0.01) (Fig. 3C–D). In parallel, the higher grain yield that was observed in the +Si–Fe plants was associated with a higher number of grains per plant and higher percentage of filled grains with an increase in the 1000-grain mass. These results were reflected in a higher harvest index in the +Si plants than in the –Si individuals independently of cultivar (Table S2, Table 1). To gain an integrated view of plant performance in terms of photosynthesis and grain production in response to the treatments, a PCA was performed using the variables that responded significantly to the factors Si and/or Fe. The first two PCs explained 66.5% and 79.6% (respectively for the Fe-tolerant and Fe-sensitive cultivar) of the observed variance (Fig. 4). This analysis evidenced the negative effects of Fe on photosynthetic performance and grain yield, as well as the mitigating effects of Si on attenuating these impairments, especially in the Fe-sensitive cultivar (negative correlations between Fe content and all of the analyzed variables in contrast to Si, i.e., positive correlations). In addition, within each cultivar, the plants were grouped (and separated) according to the four treatment-combinations we applied. The plant groups + Si–Fe and –Si + Fe were located at the ends of PC1, while the –Si–Fe and +Si + Fe plant groups were located closer to the origin of the biplot (Fig. 3). Once again, the best separation between groups in the Fe-sensitive cultivar than in its tolerant counterpart highlights the stronger effects of Fe, and the mitigation ability of Si on the negative effects of excess Fe, on the photosynthetic performance and the crop yield parameters.

3. Results 3.1. Si reduced the Fe content in leaves and roots with no effects on the grains As expected, Si application promoted increase in Si contents in roots, leaves and husks (Table S1, Fig. 1A–B). Regardless of cultivars, the highest Fe contents were observed in leaves and roots under excess Fe, whereas only discrete increases in Fe content were observed in the husk (Fig. 1C–D). Notably, Si application was able to mitigate Fe toxicity as evidenced by the reduced Fe levels in the leaves and roots in +Si + Fe plants from both cultivars (Table S1, Fig. 1C–D); however, Si did not affect the content of Fe in the husk. Regardless of treatments, no significant changes in Si and/or Fe contents were observed in the endosperm (Table S1). 3.2. The excess of Fe affected mostly the mineral composition of leaves, a fact that was partially reversed by Si The leaf contents of some nutrients such as Ca, Mg, S, Cu and Mn were reduced under excess Fe, but less markedly in +Si plants (Table S1, Fig. S1, Fig. S2). These reductions were stronger (> 50%) for Ca and Mg contents in both cultivars, and for S content in the Fe-sensitive cultivar In contrast, little, if at all, alterations in nutrient pools were observed in rice roots, husks and endosperms. In the case of endosperm, for example, there were only slight changes in P contents (reductions in –Si + Fe plants relative to –Si–Fe plants) (Table S1, Fig. S1). 3.3. The negative impacts of Fe toxicity on photosynthesis and crop yield were alleviated by Si Under adequate Fe levels (–Fe), cultivar differences were noted in some morphological and production traits; the Fe-tolerant cultivar displayed greater number of leaves and leaf area, root biomass, number of panicles, 1000-grain mass, harvest index and number of grains per plant than its sensitive counterpart (Table S2, Table 1). Fe toxicity led to significant decrease in gas exchange and chlorophyll a fluorescence parameter values (A, gs, ETR, qP and Fv’/Fm’; Fig. 2, Table S2), total biomass and agronomic traits related to crop yields (grain biomass, 1000-grain mass, percentage of filled grains, number of grains per plant

Table 1 The effects of silicon, Si (0 or 2 mM: Si or + Si, respectively), and iron, Fe (25 μM or 5 mM: Fe and +Fe, respectively), on the yield components of two rice cultivars [tolerant cv. ‘EPAGRI-109’ and sensitive cv. ‘BR-IRGA-409’ to excess Fe] grown in nutrient solutions. n = 5 ± SE. Fe-tolerant cv.

Fe-sensitive cv.

-Fe

Height (cm) Leaf number Tiller number Panicle number Leaf area (m2) Specific leaf area (m2 kg−1) Leaf biomass (g) Culm biomass (g) Root biomass (g) Root: shoot ratio (g) Total biomass (g) Grain biomass (g) 1000-grain weight (g) Filled grain percentage (%) Harvest index (g. g−1) Grain number plant −1 Panicle number plant −1

+Fe

-Fe

+Fe

-Si

+Si

-Si

+Si

-Si

+Si

-Si

+Si

95 ± 0.3 154 ± 43 37 ± 1.3 31 ± 0.8 0.27 ± 0.01 9.6 ± 0.5 29.2 ± 1.1 76.1 ± 4.5 16.4 ± 1.2 0.16 ± 0.01 192 ± 7 70.8 ± 1.8 24.3 ± 0.3 67.9 ± 0.3 0.37 ± 0.01 3618 ± 96 117 ± 3

94 ± 1.1 163 ± 9 40 ± 0.9 31 ± 1.8 0.27 ± 0.01 9.6 ± 0.3 28.4 ± 0.8 78.6 ± 1.8 17.3 ± 1.0 0.16 ± 0.01 205 ± 2 81.1 ± 1.6 26.1 ± 0.2 72.4 ± 2.5 0.39 ± 0.01 4103 ± 60 134 ± 8

90 ± 0.6 101 ± 2 37 ± 0.9 32 ± 1.0 0.19 ± 0.01 7.5 ± 0.3 24.9 ± 0.8 56.3 ± 1.8 11.7 ± 0.2 0.15 ± 0.00 141 ± 9 48.0 ± 1.9 23.5 ± 1.1 57.3 ± 1.2 0.34 ± 0.01 3269 ± 101 101 ± 1

92 ± 0.8 104 ± 2 38 ± 2.5 32.4 ± 0.8 0.19 ± 0.01 7.8 ± 0.5 25.4 ± 1.3 58.2 ± 1.9 12.1 ± 0.7 0.15 ± 0.01 156 ± 4 60.6 ± 2.2 25.7 ± 0.1 60.7 ± 1.0 0.36 ± 0.01 3550 ± 93 111 ± 5

97 ± 1.9 115 ± 5 34 ± 1.2 26.2 ± 0.2 0.22 ± 0.01 8.8 ± 0.3 24.9 ± 1.0 68.9 ± 2.3 14.6 ± 0.8 0.15 ± 0.01 151 ± 2 41.4 ± 1.6 23.4 ± 0.7 63.0 ± 0.4 0.28 ± 0.01 2616 ± 49 104 ± 2

98 ± 0.7 118 ± 7 36 ± 0.9 26.4 ± 0.4 0.21 ± 0.02 8.5 ± 0.6 24.9 ± 0. 67.5 ± 1.5 16.6 ± 0.4 0.18 ± 0.00 166 ± 2 56.7 ± 2.2 26.3 ± 0.7 78.5 ± 1.7 0.34 ± 0.01 3139 ± 70 121 ± 3

93 ± 0.5 74 ± 3 24 ± 1.0 22.0 ± 0.4 0.12 ± 0.01 5.6 ± 0.2 21.7 ± 1.5 48.9 ± 3.5 7.4 ± 0.5 0.11 ± 0.01 102 ± 5 24.4 ± 0.2 18.4 ± 1.0 48.2 ± 0.5 0.24 ± 0.01 2385 ± 85 104 ± 2

94 ± 0.5 73 ± 2 25 ± 0.8 23.4 ± 0.5 0.12 ± 0.01 5.7 ± 0.3 21.3 ± 0.5 50.4 ± 2.1 8.9 ± 0.4 0.13 ± 0.01 116 ± 3 35.2 ± 1.1 21.7 ± 0.4 58.1 ± 1.5 0.30 ± 0.01 2525 ± 37 108 ± 3

4

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

Fig. 2. The effects of silicon, Si (0 or 2 mM: Si or + Si, respectively), and iron, Fe (25 μM or 5 mM: Fe and +Fe, respectively), on the net CO2 assimilation rate, A (A–B), stomatal conductance, gs (C–D), electron transport rate, ETR (E–F), photochemical quenching coefficient, qP (G–H), and capture efficiency of excitation energy by open photosystem II reaction centres, Fv’/Fm’ (I–J) of two rice cultivars [tolerant cv. ‘EPAGRI-109’ and sensitive cv. ‘BR-IRGA-409’ to excess Fe] grown in nutrient solutions. n = 5 ± SE. 5

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

Fig. 3. The correlation between net CO2 assimilation rate (A) with stomatal conductance (gs) (A) or electron transport rate (ETR) (B), and the correlation of A with leaf Fe contents (C) and gs with leaf Fe contents (D) of two rice cultivars [tolerant cv. ‘EPAGRI-109’ (circles) and sensitive cv. ‘BR-IRGA-409’ (triangles) to excess Fe] grown in nutrient solutions. Open and solid symbols refer to –Si plants and +Si plants, respectively.

4. Discussion

chlorophyll pools in rice leaves (Peng et al., 2019). It should be noted that lower ETR values were associated with lower values of qP, suggesting a lower use of the absorbed radiant energy in photochemical processes, as well as a lower energy dissipation in the form of heat (as noted by the decreases in Fv’/Fm’); in any case, given that the Fv/Fm ratio remained at approx. 0.80 (data not shown), photosynthetic photoinhibition under excess Fe is unlikely to have occurred (Logan et al., 2007) in this study. To the best of our knowledge, this is the first report examining the ability of Si to mitigate the effects of Fe toxicity at the reproductive growth phase of rice plants. This mitigating effect was at least in part related to the lower Fe accumulation in +Si plants, particularly in the sensitive cultivar. Si application has been reported to reduce Fe precipitation on rice root surface under excess Fe (Fu et al., 2012); in addition, Si can decrease Fe uptake in lowland rice via an increased oxidation power of roots, thus ultimately decreasing Fe2+ availability (Fageria, 2014). Given that Si did not affect the total leaf area, it is suggested that the higher grain biomass (26–44%) and higher harvest index of the +Si + Fe plants compared to their –Si + Fe counterparts should have been largely associated with a greater photosynthetic capacity per unit leaf area (coupled with increases in gs and ETR), i.e. greater source capacity. In addition, Si application is believed to increase the sink strength (i) via decreased transpiration of grain husks (Ma and Yamaji, 2015) which can be translated into better water status (and high hydrostatic pressures) and thus improved growth, and (ii) by stimulating the formation of more grains per plant (6–9%), which might be associated with a role of Si in panicle fertility (Lavinsky et al., 2016). A high coordination between the source capacity and sink demand under high Si should ultimately lead to an improved crop yield regardless of excess Fe. It should be noted that Si application, per se, has been associated with increases in yield and harvest index in rice cultivated under unstressed conditions; and to support this increased production, a feed-forward effect on photosynthesis that was coupled with increases in leaf conductance has been observed (Detmann et al., 2012; Lavinsky et al., 2016). These results, obtained with other rice cultivars, are consistent with the data of this current study. Taken together, our

The Fe-tolerant cultivar was better able to accumulate biomass and to produce higher grain yields than its sensitive counterpart. Differences in tolerance to Fe toxicity appear to be at least in part associated with the inefficiency of the sensitive cultivar to avoid excess Fe levels in its tissues when compared with the Fe-tolerant cultivar (Silveira et al., 2007). Larger grain yield of the Fe-tolerant cultivar was associated not only with a larger leaf area, but also with a higher photosynthetic capacity per unit of leaf area (and hence greater photosynthetic capacity at the whole-plant level), and with an improved assimilate partitioning for the grains (higher harvest index). Regardless of cultivar, Fe levels in both the flag leaves and roots under excess Fe were above the critical level of toxicity (Dobermann and Fairhurst, 2000). Despite increase in husk Fe contents as particularly found in the Fe-sensitive cultivar, no changes in Fe content, as well as for any other nutrients (except P in the sensitive cultivar), were observed in the endosperm. It is therefore tempting to suggest that there was a notable nutritional homeostasis in the rice grains independently of the external Fe supply, as also noted elsewhere (e.g., Frei et al., 2016). Therefore, there seems to be little avenues to increase Fe levels in the grain and expected that the cultivation of this species in areas with excess Fe will be impacted in terms of growth and production (see below), but without creating any appreciable benefit in terms of food safety via increases in Fe content in the grain. The effects of Fe toxicity on nutritional imbalances in rice are relatively well established (Sahrawat, 2004; Stein et al., 2009, 2014), as observed here with decreases in Ca, Mg, S, Cu, and Mn contents, especially in the leaves. This response is a type of antagonistic interaction promoted with the inhibition of cation uptake by excess Fe (Sahrawat, 2004). Thus, effects of Fe toxicity on production, including depressed growth, increased spikelet sterility, reduced grain filling and biomass, are consistent with the results presented here. In addition to these changes, reductions of A occurred with concordant decreases in gs and ETR. Decreases in ETR in response to excess Fe may be associated with Mg deficiency (Fig. S1) which in turn is likely to decrease 6

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

Fig. 4. The first two axes of a principal component (PC) analysis of all four treatment-combinations (0 or 2 mM: Si or + Si, respectively; and iron, Fe (25 μM or 5 mM: Fe and +Fe, respectively) of two rice cultivars [tolerant cv. ‘EPAGRI-109’ and sensitive cv. ‘BR-IRGA-409’ to excess Fe] (A–B). The correlation coefficients for all traits are represented by Eigenvectors (linear correlation between a variable in a main component) (solid arrows). The segregation of rice plants following relative scores of the groups based on four treatment-combinations in relation to the two principal components (PC1 and PC2) (C-D). The numbers indicate abbreviations as follows: (1) leaf Si; (2) leaf Fe; (3) net CO2 assimilation rate; (4) stomatal conductance; (5) electron transport rate; (6) actual quantum yield of photosystem II; (7) photochemical quenching coefficient; (8) 1000-grain mass; (9) total grain biomass; (10) percentage of filled grains; (11) grain number per plant; (12) grain number per panicle; (13) harvest index (A-B). The ellipsis represents the interval confidence (α = 0.05) for each treatment-combination (C–D).

Support and Evaluation of Graduate, Brazil (CAPES, Finance Code 001).

results suggest that the positive effects of Si on photosynthesis and crop yield were manifested not only in the –Fe plants but also in the +Fe plants, especially in the sensitive cultivar, as highlighted by our PCA. In summary, our results support the hypothesis that Si is capable of mitigating the negative impacts of Fe toxicity on rice plants at the reproductive growth stage by reducing the accumulation of Fe in both leaves and roots, in addition to indirectly alleviating other nutritional deficiencies such as those of Ca and Mg. It should be noted that Fe toxicity (and Si application) had little or no effect on the grain nutritional composition. The + Si plants displayed lower Fe contents and higher crop yields which in turn were supported by increased A coupled with enhanced gs and ETR. In addition, it is likely that Si reduced the negative impact of Fe on spikelet sterility due to the possible role of Si in panicle fertility (Lavinsky et al., 2016). In view of these facts, it is recommended the use of silicate fertilizers in rice-producing areas under excess Fe as an effective management strategy to reduce the negative impacts of Fe toxicity on rice photosynthetic performance and grain yield.

Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.ecoenv.2019.110008. References Abbas, T., Balal, R.M., Shahid, M.A., Pervez, M.A., Ayyub, C.M., Aqueel, M.A., Javaid, M.M., 2015. Silicon-induced alleviation of NaCl toxicity in okra (Abelmoschus esculentus) is associated with enhanced photosynthesis, osmoprotectants and antioxidant metabolism. Acta Physiol. Plant. 37, 2–15. Audebert, A., Fofana, M., 2009. Rice yield gap due to iron toxicity in West Africa. J. Agron. Crop Sci. 195, 66–76. Chalmardi, Z.C., Abdolzadeh, A., Sadeghipour, H.K., 2014. Silicon nutrition potentiates the antioxidant metabolism of rice plants under iron toxicity. Acta Physiol. Plant. 36, 493–502. Coskun, D., Deshmukh, R., Sonah, H., Menzies, J.G., Reynolds, O., Ma, J.F., Kronzucker, H.J., Bélanger, R.R., 2019. The controversies of silicon's role in plant biology. New Phytol. 221, 67–85. Dallagnol, L.J., Rodrigues, F.A., DaMatta, F.M., Mielli, M.V.B., Pereira, S.C., 2011. Deficiency in silicon uptake affects cytological, physiological, and biochemical events in the rice Bipolaris oryzae interaction. Phytopathology 101, 92–104. Debona, D., Rodrigues, F.A., Datnoff, L.E., 2017. Silicon's role in abiotic and biotic plant stresses. Annu. Rev. Phytopathol. 55, 1–23. Detmann, K.C., Araújo, W.L., Martins, S.C.V., Sanglard, L.M.V.P., Reis, J.V., Detmann, E., Rodrigues, F.A., Nunes-Nesi, A., Fernie, A.R., DaMatta, F.M., 2012. Silicon nutrition increases grain yield, which, in turn, exerts a feed-forward stimulation of photosynthetic rates via enhanced mesophyll conductance and alters primary metabolism in rice. New Phytol. 196, 752–762.

Acknowledgements This research was supported by the Foundation for Research Assistance of Minas Gerais State, Brazil (FAPEMIG, Grant CRA-APQ02132-15) and by the National Council for Scientific and Technological Development, Brazil (CNPq, Grant 308652/2014-2) granted to FMD. This study was financed in part by the Brazilian Federal Agency for 7

Ecotoxicology and Environmental Safety xxx (xxxx) xxxx

M.S. dos Santos, et al.

Olaleye, A.O., Tabi, F.O., Ogunkunle, A.O., Singh, B.N., Sahrawat, K.L., 2001. Effect of toxic iron concentrations on the growth of lowlands rice. J. Plant Nutr. 24, 441–457. Peng, Y.Y., Liao, L.L., Liu, S., Nie, M.M., Li, J., Zhang, L.D., Ma, J.F., Chen, Z.C., 2019. Magnesium deficiency triggers SGR-mediated chlorophyll degradation for magnesium remobilization. Plant Physiol. 181, 262–275. Pereira, E.G., Oliva, M.A., Rosado-Souza, L., Mendes, G.C., Colares, D.S., Stopato, C.H., Almeida, A.M., 2013. Iron excess affects rice photosynthesis through stomatal and non-stomatal limitations. Plant Sci. 201, 81–89. Perez, C.E.A., Rodrigues, F.Á., Moreira, W.R., DaMatta, F.M., 2014. Leaf gas exchange and chlorophyll a fluorescence in wheat plants supplied with silicon and infected with Pyricularia oryzae. Phytopathology 104, 143–149. Pinto, S.S., Souza, A.E., Oliva, M.A., Pereira, E.G., 2016. Oxidative damage and photosynthetic impairment in tropical rice cultivars upon exposure to excess iron. Sci. Agric. 73, 217–226. Rencher, A.C., 2002. Methods in Multivariate Analysis. Wiley, New York. Rizwan, M., Meunier, J.D., Davidian, J.C., Pokrovsky, O.S., Bovet, N., Keller, C., 2016. Silicon alleviates Cd stress of wheat seedlings (Triticum turgidum L. cv. Claudio) grown in hydroponics. Environ. Sci. Pollut. Res. 33, 1414–1427. Sahrawat, K.L., 2004. Iron toxicity in wetland rice and the role of other nutrients. J. Plant Nutr. 27, 1471–1504. Sanglard, L.M.V.P., Martins, S.C.V., Detmann, K.C., Silva, P.E.M., Lavinsky, A.O., Silva, M.M., Detmann, E., Araújo, W.L., DaMatta, F.M., 2014. Silicon nutrition alleviates the negative impacts of arsenic on the photosynthetic apparatus of rice leaves, an analysis of the key limitations of photosynthesis. Physiol. Plant. 152, 355–366. Sanglard, L.M.V.P., Detmann, K.C., Martins, S.C.V., Teixeira, R.A., Pereira, L.F., Sanglard, M.L., Fernie, A.R., Araújo, W.L., DaMatta, F.M., 2016. The role of silicone in metabolic acclimation of rice plants challenged with arsenic. Environ. Exp. Bot. 123, 22–36. Sikirou, M., Saito, K., Achigan-Dako, E.G., Dramé, K.N., Ahanchédé, A., Venuprasad, R., 2015. Genetic improvement of iron toxicity tolerance in rice – progress, challenges and prospects in West Africa. Plant Prod. Sci. 18, 423–434. Silveira, V.C., Oliveira, A.P., Sperotto, R.A., Espindola, L.S., Amaral, L., Dias, J.F., Cunha, J.B., Fett, J.P., 2007. Influence of iron on mineral status of two rice (Oryza sativa L.) cultivars. Braz. J. Plant Physiol. 19, 127–139. Song, A., Li, P., Fan, F., Li, Z., Liang, Y., 2014. The effect of silicon on photosynthesis and expression of its relevant genes in rice (Oryza sativa L.) under high-zinc stress. PLoS One 9, e113782. Stein, R.J., Duarte, G.L., Spohr, M.G., Lopes, S.I.G., Fett, J.P., 2009. Distinct physiological responses of two rice cultivars subjected to iron toxicity under field conditions. Ann. Appl. Biol. 154, 269–277. Stein, R.J., Lopes, S.I.G., Fett, J.P., 2014. Iron toxicity in field-cultivated rice, contrasting tolerance mechanisms in distinct cultivars. Theor. Exp. Plant Physiol. 16, 135–146. Tamai, K., Ma, J.F., 2008. Reexamination of silicon effects on rice growth and production under field conditions using a low silicon mutant. Plant Soil 307, 21–27.

Dobermann, A., Fairhurst, T., 2000. Rice: Nutrient Disorders and Nutrient Management. International Rice Research Institute, Manila. Dufey, S.G., Ingabire, A., Lutts, S., Bertin, P., 2014. Silicon application in cultivated rices (Oryza sativa L. and Oryza glaberrima Steud) alleviates iron toxicity symptoms through the reduction in iron concentration in the leaf tissue. J. Agron. Crop Sci. 200, 132–142. Fageria, N.K., 2014. Mineral Nutrition of Rice. CRC Press, Boca Raton. Frei, M., Tetteh, R.N., Razafindrazaka, A.L., Fuh, M.A., Wu, L., Becker, M., 2016. Responses of rice to chronic and acute iron toxicity, genotypic differences and biofortification aspects. Plant Soil 408, 149–161. Fu, Y.-Q., Shen, H., Wu, D.-M., Cai, K.-Z., 2012. Silicon-mediated amelioration of Fe2+ toxicity in rice (Oryza sativa L.) roots. Pedosphere 22, 795–802. Genty, B., Briantais, J.M., Baker, N.R., 1989. The relationship between the quantum yield of photosynthetic electron-transport and quenching of chlorophyll fluorescence. Biochim. Biophys. Acta 990, 87–92. Hoagland, D.R., Arnon, D.I., 1950. The Water Culture Method for Growing Plants without Soils. California Agricultural Experimental Station, Berkeley. Kassambara, A., Mundt, F., 2016. Factoextra, extract and visualize the results of multivariate data analyses. R package version 1.0.3. Available online: https://cran.rproject.org/web/packages/factoextra/index.html. Korndörfer, G.H., Pereira, H.S., Nolla, A., 2004. Análise de Silício: Solo. Planta e Fertilizante. Universidade Federal de Uberlândia, Uberlândia. Lavinsky, A.O., Detmann, K.C., Reis, J.V., Ávila, R.T., Sanglard, M.L., Pereira, L.F., Sanglard, L.M.V.P., Rodrigues, F.A., Araújo, W.L., DaMatta, F.M., 2016. Silicon improves rice grain yield and photosynthesis specifically when supplied during the reproductive growth stage. J. Plant Physiol. 206, 125–132. Logan, B.A., Adams, W.W., Demmig-Adams, B., 2007. Avoiding common pitfalls of chlorophyll fluorescence analysis under field conditions. Funct. Plant Biol. 34, 853–859. Ma, J.F., Takahashi, E., 2002. Soil, Fertilizer, and Plant Silicon Research in Japan. Elsevier Science, Amsterdam. Ma, J.F., Yamaji, N., 2015. A cooperative system of silicon transport in plants. Trends Plant Sci. 20, 435–442. Ma, J.F., Nishimura, K., Takahashi, E., 1989. Effect of silicon on the growth of rice plant the different growth stages. Soil Sci. Plant Nutr. 35, 347–356. Malavolta, E., Vitti, G.C., Oliveira, S.A., 1989. Evaluation of the Nutritional State of Plants: Principles and Applications. Potafos, Piracicaba. Maxwell, K., Johnson, G.N., 2000. Chlorophyll fluorescence – a practical guide. J. Exp. Bot. 51, 659–668. Müller, C., Kuki, K.N., Pinheiro, D.T., Souza, L.R.S., Silva, A.I.S., Loureiro, M.E., Oliva, M.A., Almeida, A.M., 2015. Differential physiological responses in rice upon exposure to excess distinct iron forms. Plant Soil 391, 123–138. Nagula, S., Joseph, B., Gladis, R., 2016. Silicon nutrition to rice (Oryza sativa L.) alleviates Fe, Mn and Al toxicity in laterite derived rice soils. J. Indian Soc. Soil Sci. 64, 297–301.

8