Silicon (Si), the second most abundant element in the earth’s crust, occurs in many minerals (Wedepohl, 1995; Ma and Yamaji, 2006; Schaller et al., 2021). Soil Si occurs as dissolved silicic acid, adsorbed Si, amorphous silica (ASi, including biogenic ASi such as phytoliths and diatom shells), and Si bound in primary and secondary silicate minerals (Schaller et al., 2021; Bezerra et al., 2025).

However, the amorphous silica (ASi) concentration in agricultural soils is often depleted, typically falling below 1%, and in many cases approaching 0% (Schaller et al., 2021; Schaller et al., 2025). The depletion of ASi in agricultural soils occurs because most crops accumulate Si, at crop-specific rates, with high Si accumulators including grasses and cereals (e.g., rice, wheat, maize, and sugarcane), medium accumulators including some legumes (e.g., soybeans), while many dicotyledonous crops, including most vegetables, tuber crops (e.g., potato), and oilseed crops (e.g., rapeseed, sunflower), are low Si accumulators (Guntzer et al., 2012; Rea et al., 2022), resulting in the removal of large amounts of Si from the field through crop harvest (Struyf et al., 2010; Schaller et al., 2021). At the same time, unlike essential nutrients, Si is rarely replenished through fertilization, leading to a progressive decline in soil Si contents, although recent field studies demonstrate that amorphous silica can be applied at field scale using naturally occurring materials such as diatomaceous earth (Madegwa et al., 2025; Schaller et al., 2025). Despite not being classified as an essential plant nutrient (Fertilizer and Service; (Mengel and Kirkby, 1987)), Si has been shown to affect soil properties, improve nutrient uptake, and thus enhance plant resilience against biotic and abiotic stresses (Pati et al., 2016; Galindo et al., 2021; Martos-García et al., 2025). For instance, it has been reported that ASi changes soil physical properties, such as bulk density or porosity, as well as water retention, where the extent of change depends on the initial soil texture (Barbosa et al., 2024). Moreover, studies found that ASi addition to soil, improved N uptake in young rice, olive, maize, as well as wheat plants, leading to increased growth and yield (Pati et al., 2016; Galindo et al., 2021; Martos-García et al., 2025). Similarly, Hoffmann et al. (2025) showed that ASi enhanced plant nitrogen uptake, potentially limiting nitrogen availability for microbial processes like nitrification and denitrification, which in turn led to a roughly 30% reduction in N₂O emissions. Additionally, ASi has been associated with higher soil CO₂ emissions (Hoffmann et al., 2025). This increase has been linked to enhanced biomass production and associated microbial and root respiration as a proposed mechanism, potentially resulting from ASi-induced increases in carbon inputs to the soil through root exudates and organic matter decomposition (Hoffmann et al., 2025). However, while previous studies have mainly been focused on the role of ASi in plant nutrition, stress tolerance, and the modification of soil physical properties, direct mechanistic links between ASi-induced changes in bulk density and WFPS and the resulting CO₂ and N₂O emissions under a controlled, standardized water input across contrasting soil textures remain poorly quantified.

ASi has a lower bulk density (0.056–0.230 g cm⁻³) (Liu et al., 2003) than most mineral soils (1.3–1.8 g cm⁻³) (Sagona et al., 2016). Its addition can thus reduce soil bulk density, thereby altering water-filled-pore space (WFPS). Changes in these properties can significantly impact GHG emissions by modifying microbial activity and gas diffusion in soils. For instance, a reduction in bulk density may improve soil aeration, potentially influencing N₂O and CO₂ flux dynamics.

Barbosa et al. (2024) found that ASi improved aggregate stability and modified pore size distribution, increasing meso- and micropore volume in coarse-textured soils, while no such effect was observed in finer-textured soils. While this change in pore structure enhances moisture retention, particularly in coarse-textured soils (Zarebanadkouki et al., 2022), it may also affect soil aeration, thereby lowering WFPS compared to soil without ASi under identical water input.

As WFPS regulates the balance between aerobic and anaerobic conditions (Linn and Doran, 1984; Butterbach-Bahl et al., 2013), changes in WFPS due to ASi can substantially impact N₂O and CO₂ emissions. In well-aerated soils, CO₂ emissions typically peak around 60% WFPS, where microbial respiration is optimized, and decline at lower (<60%) or higher (>60%) WFPS due to oxygen or water limitation (Linn and Doran, 1984). For N₂O, different microbial processes may dominate across varying WFPS. Nitrification also has its optimum around 60% WFPS but produces a relatively small amount of N₂O. With higher WFPS (80%–100%), a decrease in oxygen availability favors denitrification (the dominant pathway for N₂O production), resulting in higher N₂O emissions (Linn and Doran, 1984; Davidson and Verchot, 2000). However, above 80% WFPS, increased oxygen depletion might also result in complete denitrification, converting N₂O to N₂ (Davidson and Verchot, 2000; Wu et al., 2017). At the same time, isotopic and microscale studies show that denitrification can already occur at about 60% WFPS, and both nitrification and denitrification may co-occur within aggregates due to microsite oxygen limitation (Müller and Clough, 2014; Wrage-Mönnig et al., 2018; Schlüter et al., 2025). In contrast to N₂O and CO₂, methane (CH₄) plays a minor role in well-aerated agricultural soils, which typically act as a net sink for atmospheric CH₄ due to microbial methane oxidation (Hansen et al., 2024). Changes in soil physical properties and aeration induced by ASi addition could therefore potentially enhance CH₄ uptake in the upper soil layers, although CH₄ fluxes are expected to remain small compared to N₂O and CO₂. In addition to oxygen dynamics, N₂O formation depends strongly on carbon and nitrate availability and specific microbes (Davidson et al., 2000); in coarse-textured soils with low soil organic carbon (SOC), limited availability of organic carbon and nitrate can constrain rates and alter typical WFPS emission patterns, as well as N₂O: N₂ ratios.

Given the strong impact of WFPS, aeration, and nitrogen availability on microbial processes and GHG emissions, understanding the actual effect of ASi addition is critical, especially in the case of texture-diverse agricultural soils where responses might differ. However, it is still not certain whether ASi shifts bulk density and pore water status in the same direction in coarse-versus fine-textured soils, and how such shifts correspond to CO₂ and N₂O emissions. For example, Maillet et al. (2025) demonstrated that soils with lower meso- and macroporosity exhibited higher WFPS and greater N₂O emission, underlying the critical role of pore connectivity and size distribution in regulating gas exchange and microbial processes.

Hence, this study aims to investigate how ASi addition affects soil bulk density, WFPS, and the emission of key GHGs, particularly N₂O and CO₂, which are most relevant in mineral and agricultural soils. We hypothesize that (i) under equivalent water input, ASi addition to soil changes the soil bulk density and thereby alters WFPS compared to soils without ASi addition (Experiment I) and ii) these ASi-induced shifts in bulk density and WFPS will be associated with changes in CO₂ and N₂O emissions during incubation (Experiment II). However, the magnitude and direction of these changes may depend on the soil’s texture. To test these hypotheses, two experiments (Experiment I and II) were performed using two different soil types varying in their silt and clay content: Luvisol (modest silt and clay content), which refers to intermediate clay fractions of ∼15–20% relative to other studied soils, and Arenosol (low silt and clay content). In experiment I, soils were treated without (controls) and with different amounts of ASi (w/w relative to soil dry weight) to determine its effect on soil bulk density and WFPS. Here, higher ASi rates were used to resolve influences soil bulk density and the derived WFPS across a broader addition range. In experiment II, soils amended with 1% ASi and unamended (controls) were incubated to determine ASi’s impact on soil N₂O and CO₂ emissions. We used 1% ASi (relative to soil dry weight) in the incubation experiment (Experiment II) because this level is within the range of ASi contents typically found in non-agricultural soils (∼1.2% on average) and is higher than what is often observed in cultivated soils, where ASi is frequently depleted to <1% in majority zero (Schaller et al., 2025). Bulk density and WFPS in Experiment II were inferred from the standardized packing and water addition procedures, based on the responses quantified in Experiment I. By addressing the impact of ASi addition on bulk density, WFPS, and finally N₂O and CO₂ emissions, we seek to provide new insights into the potential of ASi-based soil amendments as a measure to modify soil physical conditions and thereby contribute to the mitigation of agricultural N₂O and CO₂ emissions, depending on soil texture and management.

This study investigated how ASi addition can modify soil bulk density and the resulting WFPS proxy, and how these physical changes are associated with N₂O and CO₂ emissions in soils with different silt contents. WFPS was calculated rather than measured continuously, and bulk density and WFPS in the incubation vessels were inferred from the standardized packing and water addition procedures based on responses quantified in Experiment I. This approach provides a controlled mechanistic framework for interpreting gas-flux responses, but it does not capture potential variability in soil structure or moisture dynamics that may occur in the field. Consequently, while relative treatment differences are meaningful under controlled conditions, absolute values of bulk density, WFPS, and their direct influence on N₂O and CO₂ emissions remain uncertain. Therefore, gas flux responses observed in Experiment II as a result of ASi addition are interpreted as a consistent mechanistic pattern seen in the bulk density and WFPS index responses from Experiment I under the same fixed water addition, rather than serving as direct evidence of the conditions within the vessels. For this reason, oxygen availability and its influence on gas flux are discussed as an inferred mechanism.

Findings from Experiment I support our hypothesis that ASi addition alters the soil WFPS through changes in soil bulk density, but the response depends on the initial soil texture and ASi dose (Table 1; Figure 3). Under standardized packing, the response and changes in soil bulk density at 1% ASi was small and texture specific, with Arenosols increasing and Luvisols decreasing in bulk density, thus the WFPS, whereas at higher ASi addition (5%–10%), the bulk density and WFPS response converged across soils, indicating that increasing ASi dose can shift the packing and the resulting WFPS proxy in a way that dominates initial textural differences (Table 1; Figure 3). Thus, the direction and magnitude of ASi-induced changes in WFPS via changes in bulk density are not uniform but depend on the soil’s initial texture and underlying pore structure.

Against this background, we propose two contrasting effects through which ASi may alter bulk density and thus WFPS: (i) when added to finer-texture Luvisol soils, ASi may decrease bulk density due to its low density and its fine particle size that can shift the pore size distribution. These may have reduced the packing efficiency and increased the total pore volume of the soil under standardized compaction (Figure 4f), which under fixed water input would lower the fraction of pore space filled with water and thus lower the WFPS (ii) when added to coarse-textured Arenosol soils, ASi may have reduce macroporosity by clogging larger pores, thereby enhancing the bulk density and WFPS (Barbosa et al., 2024). Depending on the amount of added ASi and which of these effects dominates at a particular soil texture, adding ASi may either increase or decrease bulk density, thus affecting WFPS.

Additionally, texture analysis showed that ASi behaves like silt- and clay-sized particles, increasing the fine fraction while proportionally reducing sand content (Table 1). This shift towards finer textures may have reduced macroporosity in sandy Arenosol (Figure 4e), thereby increasing WFPS. A likely consequence of the increase in fine fraction in the coarse-texture soils is that water may be retained more readily at the applied water input because smaller pores hold water more strongly within the large pores typical of sandy soils. In contrast, in the already finer-textured Luvisols, ASi slightly lowers WFPS, potentially due to changes in pore structure, such as improved pore volume (Figure 4f) that lowers WFPS under fixed water input. This difference suggests that in finer-textured soils, the ASi-induced changes are less about adding new fractions and more about changes in packing and pore space, such that the total volume can increase without increasing the fraction of pore space filled with water under same water input. Thus, ASi may influence not only the amount of pore space through bulk density, but also the balance between pores that mainly transmit water and pores that retain water, which may help explain why WFPS can increase in Arenosols but decrease in Luvisol under identical water input. These textural effects highlight that the direction of ASi’s influence on bulk density and WFPS is texture-dependent and determined by how ASi modifies the internal pore network and the relative distribution of micro-, meso-, and macropores.

The resulting aeration shifts match the observed decrease (Arenosols) or increase (Luvisols) in CO₂ and N₂O emissions observed during experiment II. We found that 1% ASi addition to Arenosols caused an increase in bulk density, while Luvisols showed a decrease. Neither the increase nor the decrease in bulk density at 1% ASi treatment was significant (Table 1). However, at increased ASi levels (10%), both Arenosols and Luvisols showed a significant reduction in both bulk density and WFPS, suggesting that higher ASi levels may have increased the proportion of finer pores or enhanced internal porosity within soil aggregates, leading to a decrease in bulk density and WFPS.

In our second hypothesis, we proposed that ASi-induced WFPS changes via bulk density alteration would affect N₂O and CO₂ emission dynamics. In Experiment II, contrasting responses to N₂O and CO₂ emissions in Arenosols and Luvisols following ASi treatment were found, partially confirming the second hypothesis. The variations can be attributed to differences in soil texture that influence how ASi alters pore structure and bulk density, subsequently modulating the WFPS and oxygen availability.

For example, the incorporation of ASi in Arenosols resulted in an increase in WFPS, from 67.5% to 70.4% in Arenosol I and from 82.2% to 94.5% in Arenosol II (Figures 4c,d). This rise in WFPS may be attributed to a reduction in mesopore and macropore volume due to ASi addition. These changes likely limit oxygen availability and diffusion within the soil, thereby inhibiting microbial respiration, essential oxygen-dependent pathways for CO₂ production, supporting the decrease in CO₂ emissions in Arenosol (Figures 4b,d). These findings are in line with established relationships, indicating that microbial activity and GHG emissions decrease at high WFPS in response to oxygen limitation (Linn and Doran, 1984; Butterbach-Bahl et al., 2013).

We showed that N₂O emissions were highest at elevated WFPS (∼80%) compared to those at lower WFPS (∼60), irrespective of treatment (Figure 4c). An Earlier study by Linn and Doran (1984) identified this moisture range as one where oxygen limitation favors denitrification, leading to high N₂O production. However, studies by Müller and Clough (2014), Wrage-Mönnig et al. (2018), and Schlüter et al. (2025) indicate that denitrification can already occur at around 60% WFPS even in sand or weakly aggregated soils. This suggests that both nitrification and denitrification likely occurred concurrently in our soils, with denitrification becoming increasingly dominant as WFPS increases. Müller and Clough (2014), Wrage-Mönnig et al. (2018), and Schlüter et al. (2025) further showed through isotopic and microscale studies that changes in pore connectivity and microsite oxygen distribution, rather than bulk water content alone, regulate the balance between nitrification and denitrification. These findings provide a mechanistic understanding for our observations, suggesting that ASi-induced structural changes can influence microsite oxygen dynamics in a similar manner (Müller and Clough, 2014; Wrage-Mönnig et al., 2018). Consistent with microscale studies, showing that pore connectivity and oxygen heterogeneity, rather than bulk water content alone, regulate concurrent nitrification and denitrification (Müller and Clough, 2014; Wrage-Mönnig et al., 2018), the mechanism summarized in our conceptual diagram (Figure 2c), together with the observed flux patterns shown in Figures 4c,d, suggest that ASi-related structural changes may modify gas fluxes by altering microsite oxygen availability. Oxygen limitation within the aggregates, even at intermediate moisture levels (60% WFPS), can lead to overlapping nitrification and denitrification activity. Thus, the effect of ASi on pore structure could modify the extent of these micro-oxic zones, shifting the relative contribution of different N₂O-producing pathways under varying oxygen conditions. However, as the incubation was conducted with disturbed, repacked soil, pore connectivity and aggregate structure may differ from intact field conditions; nevertheless, under fixed water input and standardized compaction, spatially variable oxygen conditions are likely, given that air and water-filled pores coexist.

The high N₂O fluxes at elevated WFPS therefore, reflect enhanced denitrification, which generally produces more N₂O than nitrification processes at lower moisture levels (Davidson et al., 2000). However, with ASi treatment, further increases in WFPS beyond ∼80% (up to ∼95%) were associated with a slight reduction in N₂O emissions compared to the control (Figures 4a,c), even though emissions at these high WFPS remained greater than at lower WFPS. This pattern suggests that while denitrification was a key contributor to the high N₂O flux at elevated WFPS, the more extreme oxygen-limited condition induced by ASi at >80% WFPS may have promoted more complete denitrification, thus reducing net N₂O production toward dinitrogen (N₂) production (Linn and Doran, 1984; Davidson and Verchot, 2000). This shift in WFPS might have played a significant role in substantially reducing overall N₂O emissions found in Arenosols over time. This suggests that beyond a certain threshold, an increase in WFPS could limit oxygen-dependent microbial processes, thereby reducing both CO₂ and N₂O emissions.

Maillet et al. (2025) reported that soils with lower meso-and macroporosity, associated with WFPS, showed higher N₂O emissions. In our study, this pattern was not uniform: in Luvisol, at lower WFPS under ASi treatment coincided with higher N₂O emission compared to the untreated control (Figures 4a,c), whereas in Arenosol, under higher WFPS, due to ASi likely clogging meso- and macropores, coincided with lower N₂O emission. These results indicate that microsite oxygen dynamics, rather than total WFPS alone, govern N₂O production, and that the same WFPS can yield different emission outcomes depending on soil texture and pore connectivity (Wrage-Mönnig et al., 2018; Schlüter et al., 2025).

In contrast, the addition of ASi to Luvisol soil resulted in a decrease in WFPS, from 65.7% to 62.7% in Luvisol I and from 64.6% to 58.8% in Luvisol II (Figures 4a,c). This adjustment in WFPS moved the conditions closer to the optimal WFPS range (40%–60%) for microbial nitrification and aerobic respiration (Linn and Doran, 1984). While Linn and Doran (1984) showed that nitrification dominates under such conditions, Schlüter et al. (2025) demonstrate that small oxygen-deficient microsite zones can still develop even within well-aerated soils, allowing localized denitrification to occur concurrently with nitrification. Thus, in ASi-treated Luvisol, the lower WFPS likely improved aeration and oxygen diffusion, promoting aerobic microbial respiration, while maintaining conditions that permit limited denitrification and associated N₂O formation from localized anoxic zones. Compared to the untreated controls, these changes could explain the slightly higher CO₂ emissions and marginal difference in N₂O fluxes observed in Figures 4b,c. For Luvisols, no significant increase in N₂O emissions was observed; the small sample size (n = 4) may have reduced statistical power; increasing the sample size could demonstrate differences, through increased sensitivity, however. While ASi may contribute to lower N₂O emissions by altering bulk density and WFPS, improved plant nitrogen uptake (Hoffmann et al., 2025) could further reduce available mineral N for microbial nitrification and denitrification, being a potential way to make agriculture more sustainable (Schaller et al., 2025). Mineral N (N_min) concentrations were more or less similar across treatments (Table 1), suggesting that differences in initial substrate availability are unlikely to be the primary driver of the contrasting N₂O and CO₂ responses following ASi addition. This is consistent with the experimental setup, as soils were sampled at the end of the growing season and no additional mineral N was applied prior to incubation. Under these controlled conditions, texture- and dose-dependent changes in soil physical properties (bulk density, inferred pore connectivity, and calculated WFPS) therefore provide a more likely explanation for the observed gas-flux patterns than variation in mineral N availability.

It is still not clear, however, what the observed effect of ASi on WFPS and associated N₂O and CO₂ emissions would mean under field conditions. Several aspects of our incubation design limit direct extrapolation: soils were disturbed and repacked; incubation was conducted at constant temperature with fixed headspace flow; WFPS was calculated rather than measured continuously, and the use of standardized water addition and a single WHC reference to compare treatments. Moreover, Experiment II did not include plants, and therefore, root uptake and rhizosphere effects that can reshape the moisture and oxygen microsite were not present. Under field conditions, denitrification dynamics and the balance between N₂O production and reduction may differ due to intact aggregates, variable organic matter availability, and transient wetting-drying and precipitation events. In addition, other ASi-induced processes, such as enhanced crop growth and improved N uptake, may dominate or counteract the mechanisms observed under controlled incubation conditions, thereby changing the N₂O and CO₂ responses across texture and moisture regimes.

In conclusion, our study particularly highlights ASi’s effect on soil bulk density and WFPS proxy, and these effects are associated with texture-dependent changes in N₂O and CO₂ emissions by shifting towards or from optimal conditions for microbial respiration, nitrification, and denitrification under controlled incubation. In Experiment I, ASi effects on bulk density and WFPS were texture- and dose-dependent. In Experiment II, the contrasting N₂O and CO₂ responses between soil types coincided with the direction of ASi-induced shifts in the calculated WFPS, supporting a texture-specific role of ASi in modulating aeration controls on gas fluxes under controlled incubation conditions. These results suggest that ASi may be promising for modifying aeration-related controls on gas fluxes in coarse sandy soils, while in finer-textured soils it could have opposite effects. Field effects will heavily depend on actual soil texture, water regime, and management, and direct field observations are required to assess whether these patterns hold under realistic conditions. Under our experimental conditions, the highest N₂O emissions coincided with the highest WFPS (>80%), where oxygen limitation favored denitrification as the dominant pathway. However, considering that partial oxygen limitation and denitrification can already occur around 60% WFPS, especially in coarse-textured soils, ASi-induced changes in packing-related physical condition may influence N₂O production not only under saturated but also under moderately moist conditions. These considerations highlight the need for field-based studies across contrasting textures and moisture regimes to test whether the texture-dependent mechanism due to ASi treatment under controlled incubation conditions translates to managed soils under realistic soil structure, fluctuating water availability, and plant-soil interaction. Future work should therefore include ASi amendment under (i) field trials across texture gradients with crops present and (ii) coupled assessment of plant responses, mineral N dynamics, and CO₂ and N₂O fluxes.