Producing sufficient food to sustainably feed an estimated population of 9.7 billion people by 2050 is a significant challenge, particularly as yield trends are insufficient to meet projected global demands and the impacts of climate change on agriculture are increasing (Ray et al., 2013; Challinor et al., 2014; Cole et al., 2018). The increased frequency and severity of extreme weather events, such as drought, present a major threat to agriculture due to its sensitivity to weather parameters (Malhi et al., 2021; IPCC, 2023). For example, from 1964 to 2007, drought periods reduced global cereal yields by 5.1% and harvested area by 4.1% (Lesk et al., 2016). Wheat plays an essential role in global food security, providing the global population with approximately one fifth of dietary calories and proteins (Erenstein et al., 2022). However, its yields are susceptible to drought (Zampieri et al., 2017). Drought stress can reduce wheat yields at all stages of development, although terminal drought, where drought occurs during the flowering and grain-filling stages of development, typically causes the largest reduction in yields (Farooq et al., 2014). The intensity and duration of drought are also important contributors to the impact of drought stress on wheat yields (Farooq et al., 2014; Zhang et al., 2018). Consequently, there is an urgent need to improve the drought resilience of wheat.

Silicon (Si) fertilization has been shown to provide improved resilience against both drought and other stressors including salinity, high ultraviolet (UV) radiation, heavy metal toxicity, nutrient imbalances, pathogens and herbivory (reviewed by e.g. Coskun et al., 2016; Debona et al., 2017; Luyckx et al., 2017; Frew et al., 2018; Thorne et al., 2020). A range of mechanisms have been proposed for the alleviation of drought stress by Si; these include modifying gas exchange, reducing oxidative damage, improving photosynthetic rate, improving water uptake from the soil, increasing mineral nutrient uptake and regulating phytohormone synthesis. Whilst Si is the second most abundant element in soils, it is only accessible to plants as monosilicic acid, Si(OH)₄ (Raven, 1983; Gocke et al., 2013). This form of Si is soluble in the soil at pH< 9 and concentrations below 2 mM Si(OH)₄, with soils typically containing Si(OH)₄ concentrations of 0.1-0.6 mM (Epstein, 1994; Ma and Yamaji, 2006). Soils can therefore contain high levels of total Si, but low and potentially deficient levels of plant-available Si (Thorne et al., 2020).

Plants vary considerably in their abilities to accumulate Si (Hodson et al., 2005), with rice accumulating up to 10% Si by dry mass (Epstein, 1994). Seven of the ten crops with the highest global production are known Si accumulators, including rice, wheat, barley and maize (Guntzer et al., 2012). The ability to accumulate different levels of Si has been associated with Si transporters initially identified in rice, with several homologs being discovered in other species more recently (Mitani-Ueno and Ma, 2021). In rice, a combination of passive and active efflux transporters transport Si from the soil through the roots into the xylem (Ma et al., 2006, 2007; Huang et al., 2022). From here, it is translocated to the shoot via transpiration before being unloaded and deposited as hydrated silica (SiO₂·nH₂O) at target sites, again through a combination of passive and active transporters (Yamaji et al., 2008; Mitani-Ueno et al., 2023). This form of Si is immobile and cannot be redistributed once deposited (Epstein, 1994), though Si can be directed to specific sites within the plant (Thorne et al., 2023).

In addition to this strong link between Si uptake and transpiration, Si has been proposed to improve the resilience of plants to drought stress through alterations in gas exchange, with several studies showing that Si fertilization can impact both transpiration and stomatal conductance (g_s). For example, Si treatment has been observed to increase (Hattori et al., 2005; Ashfaq et al., 2024), have no impact on (Johnson et al., 2022) or decrease (Gao et al., 2005; 2006) g_s and transpiration in drought-stressed plants in a variety of species. A recent meta-analysis of 34 studies (excluding rice) explored the role of Si in plant water movement; here, the authors found that Si had no consistent pattern on these parameters in unstressed plants, but significantly increased g_s (although not transpiration) in drought-stressed plants (Cooke and Carey, 2023).

One mechanism through which Si has been proposed to impact transpiration is by altering root hydraulic conductance through the regulation of aquaporin activity (Liu et al., 2015). Another potential mechanism is the sub-cuticular deposition of Si, which has been suggested to reduce transpiration from the cuticle by acting as a physical barrier, although the contribution this could make to total leaf transpiration is small compared to water loss via the stomata (Agarie et al., 1998). Other studies have suggested that Si deposited in the cell walls of stomatal complexes may play a role in altering the gas exchange of Si-treated plants. For example, Ueno and Agarie (2005) proposed that silica deficiency in the cell walls of rice stomatal complexes could increase evaporative water loss from the epidermis, thus disrupting the generation of turgor pressure required for stomatal opening/closure. More recent research in tall fescue (Festuca arundinacea) has suggested that the deposition of Si in stomatal guard cells could promote increased stomatal sensitivity, mediated by K⁺ transporters (Vandegeer et al., 2021).

Given the impact of Si treatment on gas exchange varies between different experiments, the specific mechanistic steps that occur between Si fertilizer application and an altered gas exchange response remain poorly understood (Coskun et al., 2016). Furthermore, it is unclear how alterations in gas exchange can lead to improvements in drought resilience. Despite these complexities, however, many studies do report significant biomass and/or yield increases with Si treatment under drought stress in wheat (e.g. Bukhari et al., 2021; Ayed et al., 2022; Johnson et al., 2022), though some do not (Thorne et al., 2021), and responses are often genotype-specific (Thorne et al., 2021; Ayed et al., 2022; Christian et al., 2023).

This study investigated mechanisms underpinning this variation by quantifying the impact of Si treatment on wheat water use and drought resilience. To do this, we first assessed the impact of Si treatment on two wheat genotypes under well-watered and drought conditions. We then explored potential factors driving the observed genotype-specific responses to Si treatment, focusing on differences in stomatal density. Finally, we investigated how the type and duration of drought (watering regime) affected the ability of Si to provide drought alleviation.

In this study, we aimed to assess the impact of Si treatment on wheat water use and drought resilience under different watering regimes. By comparing several different genotypes and contrasting watering regimes, we found that the impact of Si treatment on wheat water use and drought resilience is strongly affected by genotype and watering regime. We also explored the impact of manipulating stomatal density on the wheat response to Si treatment to determine whether differences in stomatal density could account for genotype-specific responses. Our results suggest that stomatal density alone cannot be responsible for genotype-specific yield responses to Si treatment, but that stomatal density could play a key role in the impact of Si treatment on g_s and stomatal function.

To explore the role of Si in wheat water use, we first grew two high Si accumulating genotypes (H3 and H5) under 0 mM or 1.8 mM Si treatments. A recent meta-analysis of 34 studies (excluding rice) found that supplemental Si had no significant impact or consistent pattern on transpiration or g_s in unstressed plants (Cooke and Carey, 2023). We therefore hypothesized that under well-watered conditions, plants would exhibit minimal g_s differences. This appeared to be correct for H3 during IRGA-induced light shifts ( Figure 1B ). However, in H5 we observed considerable increases in g_s under well-watered conditions with Si, particularly in closing g_min ( Figure 1D ). This suggests that Si may be reducing the ability of stomata to fully close under low light conditions in a genotype-specific manner. This could be due to the deposition of Si in the cell walls of stomatal guard cells, an observation that has been made in several species (Sakai and Thom, 1979; Kim et al., 2002; Motomura et al., 2004; Ueno and Agarie, 2005; Law and Exley, 2011; Pierantoni et al., 2017). Our findings contrast with those of a study in tall fescue, where guard cell Si deposition was linked to a reduction in g_s in non-stressed plants and an increase in stomatal sensitivity (Vandegeer et al., 2021), suggesting that responses may be species-specific.

To test the role of Si treatment in these wheat genotypes under drought stress, we withheld water at the booting stage of development. In droughted plants, Si has been shown to increase (Hattori et al., 2005; Cooke and Carey, 2023; Ashfaq et al., 2024), decrease (Gao et al., 2005; 2006) or have no impact (Johnson et al., 2022) on transpiration and/or g_s, or even simultaneously increase transpiration whilst decreasing g_s (Chen et al., 2011). Here, we found that Si-treated plants used significantly less water than control plants during stress ( Figure 1E ), corroborating reports that Si can act to reduce g_s and transpiration under drought stress. However, this reduction in water use did not necessarily lead to improved drought resilience. Indeed, the impact of Si on yields varied considerably between H3 and H5, with significant Si-associated yield increases under drought stress for H5 and non-significant reductions in yields for H3 ( Figure 2 ). This supports several recent studies where the impact of Si on wheat yields under drought stress has been shown to be genotype-specific (Thorne et al., 2021; Ayed et al., 2022; Christian et al., 2023).

To investigate whether the differing response to Si in H3 and H5 could be linked to their contrasting stomatal densities ( Figure 1A ), we compared the responses of a reduced stomatal density transgenic Fielder line (TaEPF1OE) and a Fielder control to Si treatment. We reasoned that, if the different responses of H3 and H5 to Si treatment were reflected in the experiments involving TaEPF1OE and Fielder, higher stomatal densities could be a key factor in driving beneficial responses to Si treatment. In the IRGA-induced light shifts, TaEPF1OE and Fielder ( Figures 3B–D ) behaved similarly to H3 and H5 ( Figures 1B–D ) respectively, suggesting that stomatal density plays an important role in the impact of Si treatment on g_s and stomatal function. However, aside from the differences in gas exchange, minimal differences were observed between the reduced stomatal density and background lines. This suggests that stomatal density alone cannot be responsible for genotype-specific responses to Si and, therefore, that additional factors must be involved. These findings highlight the need to identify mechanisms that can directly translate to improvements in yields.

Si uptake occurs both passively, driven by transpiration, and actively (reviewed by e.g. Ma and Yamaji, 2015; Kumar et al., 2017b; Mandlik et al., 2020; Mitani-Ueno and Ma, 2021); however, it remains unclear whether Si accumulation is limited by the rate of transpiration or the rate of active uptake. If Si is primarily taken up by passive processes, it could be expected that by reducing stomatal density, transpiration is reduced and thus Si concentration may also be reduced. However, our results show that significantly reducing stomatal density had no impact on leaf Si concentration ( Supplementary Figure 2D ). These data add to the existing literature that highlights the importance of active transport in Si uptake and deposition (Liang et al., 2006; Gocke et al., 2013; Kumar et al., 2017a; McLarnon et al., 2017).

The yield data from this second experiment show that Si treatment can significantly increase yields under both stressful and non-stressful conditions in some genotypes ( Figure 4 ), contrasting with the idea, presented by Coskun et al. (2019), that Si does not promote plant growth per se but does protect against stress. Whilst the majority of literature demonstrates the benefit of Si fertilizers under stressful conditions (e.g. see meta-analyses by Cooke and Leishman, 2016; Johnson et al., 2024), some studies have observed benefits under non-stressful conditions (e.g. Camargo et al., 2021; Malik et al., 2025), although it is important to recognize that a truly stress-free environment rarely exists (Frew et al., 2018). Regardless, these findings suggest that even plants growing in relatively ‘non-stressed’ environments can still benefit from Si fertilization.

To test the role of watering regime on the wheat response to Si treatment, we subjected two genotypes (H5 and Fielder) that responded positively to Si under the water withholding drought stress to a prolonged reduced rSWC drought stress. It was hypothesized that the observed increases in g_s with Si in H5 ( Figure 1B ) and Fielder ( Figure 3B ) during IRGA light shifts could help improve wheat drought resilience under a more prolonged drought stress by allowing plants to maintain g_s and thus A. However, under this contrasting watering regime, Si had no significant impact on most of the parameters studied, including stomatal density, steady-state g_s, iWUE and total water use during drought ( Figure 5 ). In contrast to the increased yields under drought stress observed with Si in the first two experiments, here, we observed non-significant reductions in aboveground biomass with Si.

These differing results could be due to the altered Si treatment regime in the contrasting drought treatments, with plants in the rSWC experiment receiving less Si, and thus accumulating less Si overall, than plants in the water withholding experiments ( Supplementary Figures 1D , 2D , 3E ). Several studies have found that the benefits of Si fertilizer vary with the way in which Si is applied, e.g. the type of Si fertilizer (Thakral et al., 2024; Christian et al., 2025), the application rate and/or concentration (Johnson et al., 2022; Shamshiripour et al., 2022; Cheraghi et al., 2024; Foresti et al., 2024; Malik et al., 2025), or the lifecycle stage/timing of application(s) (Ma et al., 1989; Lavinsky et al., 2016). In our study, the contrasting watering regimes were implemented at different wheat lifecycle stages, either at booting (water withholding) or throughout vegetative growth (rSWC). In rice, Si is re-directed to the husks during flowering by the action of specific transporters (Yamaji and Ma, 2009; Yamaji et al., 2015), with significant grain yield penalties being observed if the hyper-accumulation of Si in the husk is prevented (Tamai and Ma, 2008). If this is the case in wheat, it might explain why Fielder and H5 responded more positively to Si fertilizers under drought stress at the booting stage of development than during vegetative growth.

The contrasting responses of H5 and Fielder to Si treatment under differing water regimes could also be linked to the potential costs of Si uptake. A recent study proposed the existence of commonly ignored direct costs associated with Si accumulation, specifically highlighting examples where increased leaf Si has been negatively correlated with growth rate (de Tombeur et al., 2023). In our experiments, perhaps the cost of accumulating Si in leaf tissue under prolonged reduced rSWC was too great a proportion of diminished resources, leading to slight reductions in biomass with Si. The observed biomass responses in our study may also relate to inherent differences in a plant’s response to contrasting drought stresses. For example, the priority may be to rapidly close stomata under a short-term drought stress and then minimize overall transpiration in the long-term by reducing shoot growth, with complicated feedback processes mediating these distinct strategies (Tardieu et al., 2018). It may be that Si is only beneficial for a subset of these processes that vary between drought types. Overall, the contrasting yield responses to Si under different watering regimes warrants further attention so that farmers can make informed choices when applying Si fertilizers as a mitigation for drought stress.

We found that the impact of Si treatment on wheat water use and drought resilience is strongly affected by both genotype and watering regime, helping to explain why different studies have come to different conclusions on the potential benefits of Si. Our results further show that whilst stomatal density may play a role in the impact of Si treatment on g_s and stomatal function, stomatal density alone cannot be responsible for genotype-specific yield responses to Si. It is important for future work to characterize the mechanisms responsible for these genotype-specific or stress-specific responses. If Si-fertilization is to be used as an effective and robust agronomic practice, farmers need to know: (a) if their selected cultivar is one that can benefit from Si addition under drought stress; and (b) when and how to apply Si fertilizers to the field to achieve the best protection against different types of drought stress.