3.2. Effects of Soil Amendments on Agronomic Traits of Potted Rice
The growth and development of rice plants were significantly influenced by the application of soil amendments, as evidenced by changes in various agronomic traits throughout the growing season.
Figure S1 presents a visual representation of the potted rice growth dynamics at different time points. The images reveal that during the period from 28 April to 26 May 2023, rice growth was relatively slow. However, a rapid increase in plant height was observed between 26 May and 21 June 2023. This acceleration in growth coincides with the vegetative stage of rice, where the plant invests energy in leaf and tiller development.
The effects of different treatments on rice plant height are illustrated in
Figure 2a. The control treatment (treatment 1), which received no soil amendments, consistently showed the lowest plant height throughout the growing season. This observation underscores the importance of soil amendments in promoting rice growth. Among the amended treatments, treatment 6 (25 g conditioner + 75 g nZVI@BC per pot) resulted in the most pronounced increase in plant height. This finding suggests that the combination of soil conditioner and nZVI@BC at this specific ratio provides optimal conditions for rice growth. The positive effect of soil amendments on plant height can be attributed to several factors. The soil conditioner likely improved soil structure and nutrient availability, while nZVI@BC may have enhanced water retention and nutrient use efficiency [
24]. Moreover, the iron component of nZVI@BC could have addressed potential iron deficiencies, promoting chlorophyll synthesis and overall plant growth [
25,
26]. Interestingly, the effect of nZVI@BC on plant height showed a dose-dependent response. As the amount of nZVI@BC increased from 0 to 75 g per pot, plant height increased correspondingly. However, a slight decrease was observed at the highest nZVI@BC dose (100 g per pot), suggesting that there might be an optimal range for nZVI@BC application beyond which benefits may diminish or even become detrimental.
Figure 2b illustrates the effects of treatments on tiller number and effective panicle number. The addition of soil conditioner increased both total tiller number and effective panicle number compared to the control. However, the application of biochar, whether unmodified or as nZVI@BC, had a more complex effect. While it reduced the total number of tillers, it had a minimal impact on the number of effective panicles. This resulted in an increased panicle formation rate, particularly in the treatment with 10 g·kg
−1 nZVI@BC (treatment 5). This observation suggests that while biochar and nZVI@BC may not promote extensive tillering, they enhance the efficiency of tiller development into productive panicles. This could be due to improved nutrient availability and uptake, leading to better resource allocation within the plant [
27]. The higher panicle formation rate is a desirable trait as it can contribute to improved grain yield without the metabolic cost of maintaining unproductive tillers [
28].
The effects of treatments on panicle length and flag leaf length (
Figure 2c) closely mirrored the trends observed for plant height. The addition of soil conditioner significantly increased both panicle length and flag leaf length. The application of nZVI@BC further enhanced these parameters, with the most pronounced effect observed at 15 g·kg
−1 (treatment 6). The flag leaf, being the primary source of photosynthates for grain filling, plays a crucial role in determining grain yield [
29]. Therefore, the observed increase in flag leaf length suggests a potential for improved grain filling and yield.
Figure 2d presents the effects of treatments on grain number per panicle and 100-grain weight. Both parameters showed improvements with the addition of soil conditioner and biochar. The effect of nZVI@BC followed a similar trend to other agronomic traits, with an initial increase followed by a slight decrease at the highest application rate. Treatment 5 (10 g·kg
−1 nZVI@BC) appeared to be the optimal treatment for these yield components.
The observed improvements in yield components can be attributed to the synergistic effects of the soil conditioner and nZVI@BC. The soil conditioner likely improved soil structure and nutrient availability, while nZVI@BC may have enhanced water retention and nutrient use efficiency and potentially provided beneficial effects through the slow release of iron [
30]. The slight decrease in performance at the highest nZVI@BC application rate suggests that excessive levels may induce some stress, possibly due to changes in soil redox conditions or micronutrient imbalances [
31].
3.3. Effects of Soil Amendments on Soil Fertility of Potted Rice
The application of soil amendments had significant effects on various soil fertility parameters, as shown in
Table 2. Soil pH values ranged from 7.39 to 7.86 across all treatments, indicating slightly alkaline conditions. All treatments with amendments (2–7) showed slightly higher pH values compared to the control (treatment 1). This increase in pH could be attributed to the liming effect of the soil conditioner [
32], which contained limestone. The alkaline nature of biochar and potential alkaline surface groups on nZVI@BC may have further contributed to this pH increase.
Soil organic matter (SOM) content varied from 44.1 to 59.3 g·kg
−1. Interestingly, the control treatment showed the highest SOM content. This unexpected result might be due to the decomposition of original soil organic matter being accelerated in the amended treatments [
33], possibly due to increased microbial activity stimulated by the amendments.
Total nitrogen (TN) content ranged from 2.63 to 3.40 g·kg
−1, with the control treatment again showing the highest value. This could be explained by the potential immobilization of nitrogen by biochar and nZVI@BC, which have high C/N ratios. The temporary sequestration of nitrogen by these amendments might have reduced the measurable TN in the amended treatments [
34].
Available phosphorus (AP) showed a marked increase in all amended treatments compared to the control, ranging from 137 to 281 mg·kg
−1. This substantial increase (1–2 times that of the control) can be attributed to the phosphorus content in the soil conditioner and the potential of biochar and nZVI@BC to enhance phosphorus availability through their high surface area and adsorption properties [
35].
Available potassium (AK) also increased in the amended treatments, ranging from 40.6 to 68.3 mg·kg
−1. This increase is likely due to the potassium content in the soil conditioner (potassium feldspar and potassium hydroxide) and the potential of biochar and nZVI@BC to retain and slowly release potassium [
36].
Available manganese (AM) showed slight variations across treatments, ranging from 58.6 to 66.8 mg·kg−1. The presence of nZVI@BC did not seem to significantly affect manganese availability, suggesting that it did not interfere with manganese dynamics in the soil.
Available zinc (AZ) ranged from 0.99 to 1.78 mg·kg−1, with no clear trend across treatments. This suggests that the amendments did not have a substantial impact on zinc availability in the soil.
Available copper (AC) showed an increase in most amended treatments, ranging from 13.2 to 19 mg·kg
−1, with a notable spike in treatment 5 (34.6 mg·kg
−1). This increase could be due to the presence of copper in the soil conditioner or the potential of biochar and nZVI@BC to influence copper availability through adsorption and desorption processes [
37].
The overall results indicate that the soil amendments, particularly the combination of soil conditioner and nZVI@BC, had significant effects on soil fertility parameters. The increases in available phosphorus and potassium are particularly noteworthy, as these are essential macronutrients for rice growth. The slight alterations in soil pH towards alkalinity may be beneficial for nutrient availability in what were likely originally acidic paddy soils. However, the decrease in total nitrogen in amended treatments warrants further investigation. It may be necessary to supplement nitrogen fertilization when using these amendments, especially in the early stages of rice growth, to compensate for potential nitrogen immobilization. The stability of micronutrient availability (Mn, Zn, Cu) across most treatments suggests that the amendments do not negatively impact micronutrient dynamics. The spike in available copper in treatment 5 could potentially be beneficial for rice growth, as copper is an essential micronutrient, but care should be taken to avoid copper toxicity at higher application rates.
3.4. Effects of Soil Amendments on Heavy Metal Content in Potted Rice Soil
The impact of soil amendments on heavy metal content in the potted rice soil is presented in
Table 3 and
Table 4. The results show that the concentrations of most heavy metals (Hg, As, Pb, Cr) were well below their respective limiting values across all treatments, indicating that these metals did not pose significant pollution risks in the experimental soil.
Mercury (Hg) concentrations were consistently low, around 0.2 mg·kg−1, which is well below the limiting value of 1.3 mg·kg−1. The application of soil amendments, including nZVI@BC, did not significantly alter Hg concentrations, suggesting that these amendments do not mobilize or immobilize Hg to a notable extent under the conditions of this study.
Arsenic (As) concentrations ranged from 4 to 5 mg·kg−1, far below the limiting value of 40 mg·kg−1. The slight variations observed across treatments were not substantial, indicating that the soil amendments did not significantly influence As dynamics in the soil.
Lead (Pb) concentrations were stable around 20–25 mg·kg−1, well below the limiting value of 70 mg·kg−1. The consistency of Pb concentrations across treatments suggests that the amendments did not significantly affect Pb mobility or availability in the soil.
Chromium (Cr) concentrations ranged from 40 to 50 mg·kg−1, which is considerably lower than the limiting value of 150 mg·kg−1. The variations observed across treatments were minor, indicating that the soil amendments did not substantially influence Cr dynamics in the soil.
Cadmium (Cd) concentrations, however, were at or slightly above the limiting value of 0.3 mg·kg
−1 in most treatments. This suggests that the experimental soil had a pre-existing Cd contamination issue. Interestingly, treatments with nZVI@BC showed slightly lower Cd concentrations compared to the control, with values ranging from 0.26 to 0.30 mg·kg
−1. This reduction, although small, indicates that nZVI@BC may have some capacity to immobilize or adsorb Cd in the soil [
38]. During the Cd test, the soil was sieved. Some of the larger particles of nZVI@BC were sieved out, which reduced the total Cd content of the soil used for digestion. We further tested diethylenetriamine pentaacetic acid (DTPA)-extractable Cd in soil. Heavy metals in soil extracted by DTPA are considered the most available fraction for plant uptake [
39]. The Cd concentration in the filtrates of the aforementioned sequential extraction, the DTPA extraction test, were measured by an atomic absorption spectrophotometer [
40]. After treatment, the concentration of DTPA-Cd decreased slightly when adding BC to the soil. The reduction in DTPA-Cd was more pronounced when adding nZVI@BC to the soil, from 0.06 mg kg
−1 in CK to 0.03 mg kg
−1 in treatment 4 and to 0.01 mg kg
−1 in treatments 6 and 7. Similar to our results, Lu et al. [
41] found that the application of bamboo biochar was able to reduce the contents of DTPA-Cd by 0.16 mg kg
−1 in contaminated soil.
The heavy metal pollution indices (
Table 4) provide a more comprehensive view of the overall heavy metal status in the soil. The single-factor pollution indices (
Pi) for most metals were well below 1, except for Cd, which ranged from 0.867 to 1.067. This further confirms that Cd was the primary heavy metal of concern in this soil. The Nemerow comprehensive pollution index (
Pt) ranged from 0.661 to 0.808 across treatments. According to common interpretation standards, these values indicate a low to moderate level of heavy metal pollution, primarily driven by the Cd content. The slight variations in
Pt across treatments suggest that while the soil amendments did not dramatically alter the overall heavy metal status, they may have had subtle effects on metal mobility and availability [
42]. The lower Cd concentrations and pollution indices in treatments with nZVI@BC (particularly treatments 5 and 6) suggest that this amendment may have some potential for Cd immobilization. This effect could be attributed to the high surface area and strong adsorption capacity of nZVI@BC, which can bind Cd and reduce its bioavailability in the soil.
The soil amendments, particularly nZVI@BC, did not negatively impact heavy metal dynamics in the potted rice soil. Most heavy metals remained at low, non-polluting levels across all treatments. The slight reduction in Cd levels in nZVI@BC treatments is a promising finding, suggesting that this amendment may have potential applications in mitigating Cd contamination in rice paddy soils. However, given the pre-existing Cd contamination in the experimental soil, further long-term studies are needed to fully evaluate the efficacy of nZVI@BC in Cd immobilization and its potential impacts on Cd uptake by rice plants.
3.5. Effects of Soil Amendments on Soil Microbial Communities in Potted Rice
The application of soil amendments, particularly nZVI@BC, had significant impacts on the soil microbial communities in the potted rice experiment. These effects were observed across various taxonomic levels and were reflected in both the diversity and composition of bacterial and fungal communities.
Table 5 presents the number of bacterial taxa identified at different classification levels for each treatment. The control treatment (A1) exhibited the highest number of bacterial phyla and classes among all treatments. This suggests that the unamended soil harbored a diverse bacterial community at higher taxonomic levels. However, at the order, family, and genus levels, treatments A3 and A4 showed the highest diversity. This indicates that while the addition of soil amendments may have reduced the diversity at higher taxonomic levels, it promoted diversity at lower taxonomic levels, possibly by creating new niches or altering resource availability.
The variations in bacterial diversity across treatments suggest that different combinations and concentrations of soil amendments can selectively influence bacterial communities. The increase in lower-taxonomic-level diversity in treatments A3 and A4 could be attributed to the combined effects of the soil conditioner and either biochar (A3) or a low concentration of nZVI@BC (A4). These amendments may have provided additional carbon sources, altered soil pH, or changed the soil’s physical structure, thereby creating conditions favorable for a wider range of bacterial species [
43].
Table 6 shows the number of fungal taxa identified at different classification levels for each treatment. For fungi, the control treatment (A1) had the highest number of phyla and orders, while treatment A5 had the most classes, and treatment A3 had the highest number of families and genera. This pattern differs from that observed for bacteria, highlighting the distinct responses of bacterial and fungal communities to soil amendments. The high fungal diversity at the family and genus levels in treatment A3 (soil conditioner + biochar) suggests that this combination may have created particularly favorable conditions for fungal growth and diversification. This could be due to the increased carbon availability from biochar, which many fungi can efficiently utilize [
44], combined with the improved soil structure and nutrient availability from the soil conditioner.
The differences in taxonomic diversity between bacteria and fungi across treatments underscore the complex and often contrasting responses of these two major groups of soil microorganisms to soil amendments. These results emphasize the importance of considering both bacterial and fungal communities when assessing the impacts of soil amendments on soil microbial ecology.
Figure 3 illustrates the number of bacterial and fungal OTUs for each treatment. For bacteria (
Figure 3a), treatment A4 showed the highest number of OTUs, followed by a decreasing trend as the amount of nZVI@BC increased in subsequent treatments. This suggests that there may be an optimal concentration of nZVI@BC for promoting bacterial diversity, beyond which the diversity begins to decline. The initial increase in bacterial OTUs with the addition of nZVI@BC could be due to the creation of new microhabitats on the biochar surface and the potential slow release of nutrients or electron donors from the nZVI component [
45]. However, at higher concentrations, nZVI@BC might start to exert inhibitory effects on some bacterial groups [
46], possibly due to changes in redox conditions or the release of inhibitory compounds.
For fungi (
Figure 3b), treatments A2 and A3 showed the highest number of OTUs, both exceeding the control. This indicates that the addition of soil conditioner alone (A2) or in combination with biochar (A3) created favorable conditions for fungal diversity. The subsequent decline in fungal OTUs with increasing nZVI@BC concentration suggests that fungi may be more sensitive to the effects of nZVI than bacteria. This could be due to the potential antifungal properties of nanoparticulate zero-valent iron or changes in soil chemistry that are less favorable for fungal growth.
These results align with the earlier observations on taxonomic diversity and further emphasize the differential responses of bacteria and fungi to soil amendments. The contrasting trends in bacterial and fungal OTUs with increasing nZVI@BC concentration highlight the importance of finding a balance that can maintain or enhance the diversity of both groups, as they play complementary roles in soil ecosystem functioning.
Figure 4 presents the relative abundance of bacterial communities at the class level across different treatments. The top 10 most abundant bacterial classes across all treatments were Gammaproteobacteria, Alphaproteobacteria, Vicinamibacteria, Actinobacteria, Thermoleophilia, Blastocatellia, Acidimicrobilia, Methylomirabilia, Gemmatimonadetes, and Bacteroidia. These 10 classes accounted for approximately 75% of the total bacterial sequences. Treatment A2, which received only the soil conditioner, showed the highest overall abundance of these bacterial classes. This suggests that the soil conditioner had a generally positive effect on bacterial growth and proliferation, possibly due to improved soil structure and nutrient availability. However, the addition of nZVI@BC appeared to have a slight suppressive effect on overall bacterial abundance, with the effect being relatively consistent across different nZVI@BC concentrations.
The dominance of Proteobacteria (both Gamma- and Alpha-) across all treatments is consistent with their known prevalence in agricultural soils. These classes include many plant-growth-promoting bacteria and play crucial roles in nutrient cycling. The high abundance of Actinobacteria, known for their ability to decompose recalcitrant organic matter, suggests active organic matter turnover in the soil. The presence of Methylomirabilia, capable of methane oxidation, indicates the potential for methane mitigation in these rice paddy soils. The consistent presence of Gemmatimonadetes, often associated with arid soils, in these paddy soils is interesting and warrants further investigation.
Figure 5 illustrates the relative abundance of fungal communities at the class level across different treatments. The top 10 most abundant fungal classes were
Agaricomycetes, Sordariomycetes, Eurotiomycetes, Dothideomycetes, Leotiomycetes, Microbotryomycetes, Tremellomycetes, Orbiliomycetes, and Saccharomycetes. These classes accounted for approximately 90% of the total fungal sequences, indicating a more concentrated dominance compared to the bacterial communities.
Agaricomycetes was the most dominant fungal class, accounting for about 50% of the fungal sequences across treatments. This class includes many saprotrophic fungi capable of decomposing complex organic matter, particularly lignin. The high abundance of Agaricomycetes suggests a significant potential for organic matter decomposition in these soils, which could contribute to nutrient cycling and soil fertility. Interestingly, the relative abundance of Agaricomycetes showed a non-linear response to nZVI@BC addition. It initially increased with nZVI@BC concentration, peaking in treatment A5, before declining at higher concentrations. This trend aligns with the observed patterns in rice agronomic traits, where treatment A5 often showed the best performance in terms of plant growth and yield components. The high abundance of Sordariomycetes, which includes many plant pathogens as well as endophytes, suggests a complex fungus–plant interaction in these soils. The presence of Eurotiomycetes and Dothideomycetes, which include many stress-tolerant fungi, indicates the potential resilience of the fungal community to environmental stressors.
Figure 6 presents heat maps of the species composition for (a) bacteria and (b) fungi at the genus level based on the hierarchical clustering of community relative abundance. For bacteria, the control treatment (A1) showed high expression of genera such as RB41, Latescibacterota, and SC-I-84. In contrast, treatments with soil amendments (A2–A7) were characterized by high abundance of genera, including 11–24, TRA3-20, Blastococcus, MB-A2-108, Massilia, Brevundimonas, Entotheonellaceae, Pseudonocardia, and Pedomicrobium. This shift in bacterial community composition suggests that the soil amendments created conditions that favored different bacterial groups. The genera that became abundant in the amended treatments are known to include many species with beneficial traits for plant growth and soil health. For instance, Massilia includes species known for their plant-growth-promoting abilities, while Pseudonocardia contains members capable of degrading complex organic compounds.
For fungi, the control treatment (A1) was dominated by genera such as Paraconiothyrium, Montagnula, Mycoarthris, and Plectosphaerella. Treatment A2, which received only the soil conditioner, showed a distinct fungal community dominated by Papulaspora, Wiesneriomyces, Purpureocillium, and Thelonectria. Treatments A3–A7, which received various combinations of soil conditioner, biochar, and nZVI@BC, each showed unique fungal community compositions with little overlap. This suggests that fungal communities are highly sensitive to the specific combination and concentration of soil amendments, particularly to the presence and concentration of nZVI@BC.
The distinct fungal communities observed across treatments indicate that the soil amendments, especially nZVI@BC, have a strong influence on fungal community composition. This could be due to changes in soil physicochemical properties, carbon availability, or direct interactions between fungi and the nanoparticles.
Table 7 and
Table 8 present various alpha diversity indices for bacterial and fungal communities, respectively, across different treatments. For bacteria, treatments A3, A4, and A5 showed the highest Chao1 indices, with A4 having the highest value (4934.77). This suggests that these treatments, which involve moderate levels of soil amendments, support the highest bacterial species richness. The observed species richness shows a similar trend, confirming that these treatments indeed support a more diverse bacterial community.
The Simpson index for bacteria was consistently high (close to 1) across all treatments, indicating low dominance and high evenness in the bacterial communities. The Shannon index, which considers both richness and evenness, was highest for treatment A4 (11.3656), further supporting the notion that this treatment provides optimal conditions for bacterial diversity.
For fungi, treatments A2 and A3 showed the highest Chao1 indices (286.018 and 294.219, respectively), indicating that the addition of soil conditioner alone or in combination with biochar supports the highest fungal species richness. However, the Simpson index was lowest (indicating highest diversity) for treatment A5 (0.862208), suggesting that moderate levels of nZVI@BC promote a more even distribution of fungal species.
The fungal Shannon index was highest for treatment A4 (4.93188), mirroring the trend observed in bacteria. This suggests that this particular combination of soil amendments creates conditions that support high diversity in both bacterial and fungal communities.
The Goods_coverage index was consistently high (close to 1) for both bacteria and fungi across all treatments, indicating that the sequencing depth was sufficient to capture the majority of microbial diversity in these samples.
Figure 7 presents the beta diversity analyses for bacterial and fungal communities using PCoA and NMDS. The PCoA based on Bray–Curtis distances (
Figure 7a,b) showed that the first two coordinates explained 18.5% and 17.7% of the variation in bacterial communities and 26.1% and 19.1% in fungal communities, respectively. For bacteria, the treatments showed relatively small differences in community composition, with no clear clustering pattern. This suggests that while the soil amendments did influence bacterial community composition, the effects were subtle and did not lead to dramatic community shifts.
In contrast, fungal communities showed more pronounced differences between treatments, with the control treatment (A1) being notably distinct from the amended treatments. This indicates that fungal communities were more sensitive to the soil amendments than bacterial communities, undergoing more substantial compositional changes in response to the treatments.
The NMDS analysis (
Figure 7c,d) provided complementary insights into community composition differences. The stress values for bacterial and fungal NMDS were 0.091 and 0.207, respectively, indicating a good representation of the multidimensional data in two-dimensional space, particularly for bacteria.
The NMDS plot for bacteria shows a gradual shift in community composition across treatments, with treatments A4 and A5 occupying an intermediate position between the control and the treatments with higher nZVI@BC concentrations. This suggests a dose-dependent effect of nZVI@BC on bacterial community composition.
For fungi, the NMDS plot reveals more distinct clustering of treatments, with the control and treatments A2 and A3 forming one cluster, and the nZVI@BC treatments (A4–A7) forming another. This further supports the observation that fungal communities are more strongly influenced by the presence of nZVI@BC than bacterial communities.