Introduction

Litter input is the dominant source of soil organic carbon (SOC) in most ecosystems. Even though litter decomposition driven by microorganisms leads to C mineralization, the process also leads to stabilization of SOC1,2. Microbial diversity is essential for driving soil processes3, with microbial diversity contributing to C cycling. Additionally, temperature is a critical factor affecting litter decomposition and subsequent SOC stabilization4, as it directly influences microbial activity and organic matter breakdown. Understanding the combined effects of elevated temperature and microbial diversity on the formation of stabilized SOC is vital for predicting the resilience of SOC stocks under climate change scenarios.

Litter decomposition plays a dual role in soil organic SOC dynamics. While it contributes to C loss through microbial respiration (releasing CO2 into atmosphere), microbes also assimilate nutrients and C compounds to build microbial biomass and other microbial-derived compounds (e.g., proteins, polysaccharides). Thess compounds can attach to mineral surfaces, facilitating the formation of stable organo-mineral compounds5. After death, microbial necromass can be stabilized by interacting with soil minerals6. This microbial-derived C accumulates in the mineral fraction, forming a significant part of stable SOC7,8. Moreover, organo-mineral associations and the proliferation of fungal hypha during decomposition benefit soil aggregation9. Microaggregates play a critical role in long-term SOC stabilization by physically protecting organic matter from microbial access10. In this study, we operationally considered SOC associated with microaggregates and minerals together as mineral-associated organic carbon (MAOC), aligned with other published studies11,12, since both fractions contribute to long-term SOC stabilization. Though particulate organic matter within macroaggregates decomposes quickly10, macroaggregates still play a crucial role in the occlusion of microaggregates and minerals13.

Microbes have complementary roles in soil C cycling14,15, making microbial diversity a key factor for SOC stabilization. Previous studies found positive relationships between microbial diversity and decomposition3,16, where rapid decomposition was accompanied by high microbial density and activity17, promoting formation of soil aggregates and MAOC18. However, there are also diverging results showing negative relationships19,20, likely due to antagonistic interaction among microbes competing for resources21. In benign environments with abundant resources, competition is expected to dominate over facilitation, as species can thrive independently and directly compete for the same resources22. Therefore, high diversity may lead to more intense competition than low diversity, resulting in less microbial biomass, and consequently decreased MAOC. Additionally, a few studies found microbial diversity could increase microbial carbon use efficiency (CUE)23, a proxy for microbial C allocation, describes the partitioning of C between growth and respiration. As CUE is a crucial factor influencing the C storage in the soil24, higher microbial diversity associated with higher microbial CUE could lead to more microbial biomass and necromass accumulating as MAOC25. However, studies on the effects of microbial diversity on the formation of MAOC are limited, and the results are overall not clear26 due to the interplay of biotic and abiotic factors. Moreover, some studies suggest that microbial communities across different diversity treatments have similar functions with respect to decomposition, but vary in terms of the resulting fate of SOC27,28. Therefore, the knowledge gap in the effects of microbial diversity on the formation of MAOC needs to be addressed.

Temperature is a major factor that influences microbial SOC processing. Elevated temperature stimulates microbial enzyme activity, increasing decomposition rate and CO2 release29. Climate models predict a 1–3.5 °C rise in global surface temperature by 210030, potentially leading to significant SOC losses31. However, decomposition stimulated by elevated temperature is constrain by substrate accessibility, as organic matter occluded within aggregates or bound to mineral surfaces is less accessible to microbes32. Limited data suggest that warming may reduce aggregate-associated C because of a decrease in binding agents for aggregation33, and warming might decrease MAOC by suppressing microbial growth and decreasing microbial CUE34,35. Moreover, elevated temperature leads to more active enzymes, accelerating decomposition of MAOC36. We therefore assume that elevated temperature could decrease MAOC.

As global temperatures rise, and microbial diversity decreases, the interaction between these factors may play a crucial role in determining the effectiveness of SOC stabilization. Microbial diversity is important for stabilizing ecosystem functioning37. Higher microbial diversity might buffer against temperature fluctuations, maintaining ecosystem functions38. To provide a more comprehensive understanding of SOC dynamics, particularly in the context of climate change and biodiversity loss, we performed an experiment with elevated temperature and microbial diversity as factors; we used initially microbe- and C-free soil to quantitatively assess C distribution in different fractions after litter decomposition of. Model soil systems, while simplified, can isolate specific components of complex soil processes by minimizing confounding factors inherent to natural soils39. In this study, we used a homogenized model soil (described in Methods) to control spatial heterogeneity and reduced variability caused by natural soil matrices, which often obscure relationships between microbial activity and C dynamics40. By employing model soil, this study uniquely isolates the effects of microbial diversity and temperature on SOC stabilization, providing mechanistic insights into how these factors influence SOC persistence under changing environmental conditions. We hypothesized: (1) Higher microbial diversity will result in increased litter decomposition and the formation of microbial-derived components that promote the formation of stabilized SOC. (2) Elevated temperature will increase litter decomposition but decrease stabilized SOC. (3) There is an interactive effect of microbial diversity and temperature on the formation of stabilized SOC, with elevated temperature causing a smaller reduction in stabilized SOC compared to ambient temperature.

Materials and methods

Experimental design

The experiment had a fully factorial design, with 6 unique combinations of the following treatments: 3 microbial diversity levels (low, medium, high), crossed with 2 temperature patterns (ambient and elevated temperature). Each treatment had 10 replicates for a total of 60 tubes. To ensure proper replication of the temperature treatment and avoid pseudoreplication, we used three independent incubators for each temperature treatment and included incubator as a random factor in the statistical analysis. We used 50 ml centrifugation tubes containing 30 g of soil, with hydrophobic vented caps to allow gas exchange. The water content was adjusted to 60% WHC. We replenished water loss in a sterile hood by injecting water with a syringe twice a week.

Model soil

The model soil was designed to replicate the texture of an Albic Luvisol (sandy loam) from an experimental grassland field at Freie Universitaet Berlin (Berlin, Germany), where live soil for making microbial inocula was collected (0–15 cm depth; total C: 18.7 g kg− 1, N: 1.2 g kg− 1)41. The composition included 7% kaolin clay, 5% quartz silt (< 250 μm, carbonate-free), 85.4% quartz sand (20–250 μm, carbonate-free), and 2.6% finely ground Acer platanoides litter (C: 426.1 g kg− 1, N: 6.7 g kg− 1, < 250 μm) by weight. We used finely ground leaves of Acer platanoides (C: 42.61%, N: 0.67%), the dominant litter type in study area, to represent organic matter. The model soil had no invertebrates, which play a key role in processing detritus and infusing the soil with physically smaller and chemically decomposed resources. Therefore, we mill-ground litter (Retsch MM 400; 30 Hz for 30 s) and sieved them through a 250 μm mesh. Subsequently, the litter was mixed thoroughly with the model soil. Following the method by Caesar-Tonthat (2002)42the mixture was thoroughly wetted with ultrapure water, dried at 60℃ for 48 h and the resulting soil cake was ground and sieved through 2 mm sieves, i.e. the model soil comprised primary particles, litter fragments not exceeding 0.25 mm, and model soil particles of various sizes not exceeding 2 mm. Subsequently, this mixture was autoclaved twice on two consecutive days.

Soil microbial inoculation (different diversity levels)

The soil inocula with different diversity levels were obtained from a previous experiment43: Briefly, we performed six serial dilution steps of the live soil (collected from the experimental grassland field described in the previous section) through 10− 6, using autoclaved soil as the diluent at each step. A portion of the diluted mixture from each step was used to inoculate the next dilution step, following the protocol of Franklin44. Soil inocula were incubated at room temperature for 2 months to achieve equivalent microbial biomass before use. Then, 100 undiluted (high diversity), 103 diluted (medium diversity) and 106 diluted (low diversity) soil were used as microbial inocula in our experiment. The microbial diversity levels were validated by using Illumina MiSeq high-throughput sequencing with fITS7 and ITS4 for fungi and 515f and 806r for bacteria43 (microbial community assembly were presented in Figure S1). Prior to inoculation, air-dried inocula were sieved (< 250 μm), and 1 g of each inoculum was homogenized with 30 g of sterile artificial soil to ensure uniform distribution.

Temperature treatment

The ambient temperature was set at 25℃. We started the experiment with all treatments at 25 °C and incubated the tubes for 30 days, to allow full recovery of the microbial communities based on findings that inoculated soils achieve stable microbial biomass within 4 weeks45. After these 30 days the treatment with the elevated temperature was applied to half of the tubes. This was done by increasing the temperature by 2.5℃ from 25 to 27.5℃ and keeping the temperature constant at 27.5 °C for the remaining duration of the experiment, which was 45 days. Total incubation time was 75 days (30 + 45 days). The 75-day incubation period was selected based on established protocols for litter decomposition studies46, which demonstrates that this timeframe is sufficient to observe microbial-mediated carbon distribution. The 10 replicates of one factor combination were split into 3 groups and incubated in 3 individual incubators (temperature treatment: n = 3). The elevated temperature treatment (+ 2.5 °C) aligns with intermediate climate change projections for the 21st century, reflecting the predicted global surface temperature rise of 1–3.5 °C by 210030.

Microbial respiration

We measured microbial respiration as CO2 production rate (mg CO2-C kg− 1 dry soil h− 1) on day 75 of the incubation. Before the measurement, we flushed each of the tubes with CO2-free air for five minutes to standardize among experimental units47. After 5.5 h (preliminary measurements showed that after this time readouts were within the range of the calibration curve), we sampled 1 ml of air from the headspace of each tube, and injected this sample into an infrared gas analyser (LiCOR 6400xt). Microbial respiration and total C loss were used as a proxy for organic matter decomposition.

Harvest/destructive sampling

Soil samples were homogenized by spatula in a plastic bag, then dried at 40℃ for 48 h in a drying oven and subsequently stored at room temperature until the analysis of water stable aggregates and total C.

Water stable aggregates

We followed the protocol by Kemper & Rosenau48: the dried soil was sieved through a 2 mm sieve and we placed 4.0 g of soil into sieves for capillary rewetting in deionized water for 5 min. We used 0.25 mm sieves to test the stability of the soil fraction > 0.25 mm (macroaggregates) against water as disintegrating force. For the test, sieves carrying the wetted soil samples were placed in a wet-sieving machine (Eijkelkamp, Netherlands) and moved vertically (stroke = 1.3 cm, 34 times min− 1) for 3 min. The fraction (< 0.25 mm) left in the metal bin was transferred to a 0.053 mm sieve to repeat the wet-sieving again in order to obtain microaggregates. The fractions left on the sieves were dried at 60 ℃ for 24 h. There was no coarse matter (the sizes of sand and original organic matter fraction were smaller than 0.25 mm) in macroaggregates. The percentage of macroaggregates was calculated as: % water stable aggregates = macroaggregates (g) / 4.0 g x 100. We stored macroaggregates and microaggregates at room temperature until the analysis of SOC fractions.

Separation of SOC fractions

For this study, we divided SOC fractions into different groups to capture SOC protection by various mechanisms. The incubation lasted 75 days, which is the initial period of litter decomposition49. The labile C available during early stages of decomposition was partly mineralized, partly remained as unprotected SOC by microaggregates or minerals, and partly formed the stabilized SOC. We extracted the following SOC fractions50: free POC, free microaggregates-associated C (free microaggregates-C), free minerals-associated C (free minerals-C), macroaggregate-occluded POC, macroaggregate-occluded microaggregate-C, macroaggregate-occluded minerals-C (Fig. 1). We only focused on the SOC stabilized in aggregates and minerals, i.e. we did not measure dissolved organic C.

Fig. 1
figure 1

Diagram depicting the workflow of the separation of different SOC fractions used in the experiment. Bulk soil was subjected to wet sieving to separate macroaggregate (250–2000 μm), free microaggregate (53–250 μm), and free minerals (< 53 μm). Density separation method was used to separate free microaggregates and free POC. Macroaggregate fraction was then further separated into occluded POC and the combination of occluded microaggregate and occluded minerals by using density separation method.

Full size image

We wet-sieved the soil with a 250 μm sieve and collected the material that passed through the sieve for the determination of free POC and free microaggregates. The material that stayed on the sieve was collected for the determination of macroaggregate-occluded C.

We subsequently followed and modified a protocol by Plaza51to separate SOC fractions. To separate free microaggregates and free POM, the microaggregates of 2 replicates were combined due to the limit amount of microaggregates in each replicate, transferred to a 15 ml tube, and we performed a density separation with the following steps: (i) we added 12 mL of sodium polytungstate (SPT) solution (Smetu, Germany) at a density of 1.85 g mL− 1 to the sample (ii) rotated the tube at 1 revolution/s for 30s, (iii) centrifuged it at 2500 g for 30 min, (iv) vacuum-filtered the supernatant (47 mm, Glass microfiber filters - GF/A Grade 1.6 μm pore size, Whatman, UK) and finally (v) rinsed the material retained on the filter (free POC) with 300 mL of distilled water. Steps i) to v) of this density separation were repeated in order to more completely capture the free POC. The remaining soil in the tube was designated as the heavy fraction, referring to free microaggregates. To clean the SPT left in the heavy fraction, the fraction was transferred into a 50 ml tube, suspended in 30 mL distilled water and centrifuged at 3000 g for 5 min; the supernatant was siphoned off and these steps were repeated three times.

To gain macroaggregate-occluded POC, macroaggregate-occluded microaggregates and macroaggregate-occluded minerals, we passed the macroaggregates through a 250 μm mesh using gentle pressure so as to minimize macroaggregate destruction (2 replicates were combined due to the limited amount of macroaggregates in each replicate); since our soil consisted of components smaller than 250 μm, macroaggregates could be easily passed through the sieve without applying much pressure. We followed the same procedure as above to separate the light fraction (macroaggregate-occluded POC) and heavy fraction (macroaggregate-occluded microaggregates and minerals). All fractions were dried at 60℃ for 48 h and weighed. The C contents were measured by CN Analyzer (Vario EL Cube; Elementar).

Since we did not collect free minerals, the free mineral-C was calculated as subtracting the C within all the fractions we collected above from the total C. As we refer MAOC as C within microaggregates and C sorption to minerals, we defined macroaggregate-occluded microaggregate and macroaggregate-occluded minerals-C as macroaggregate-occluded MAOC, and free microaggregates-C and free minerals-C as free MAOC. The total POC was the combination of free POC and macroaggregate-occluded POC, and the total MAOC was the combination of free MAOC and macroaggregate-occluded MAOC (Fig. 1).

Statistical analysis

We used a Shapiro Wilk’s test52 and Bartlett test53to check the normality of residuals and the homogeneity of variances, respectively. We analyzed the effects of microbial diversity and elevated temperature using a linear mixed-effects model with incubators as a random factor, followed by two-way ANOVAs, with a p-value cutoff of 0.05, and considered p-values that fell between 0.05 and 0.10 as marginally significant. To test for significant differences (p < 0.05) between treatments, we performed Tukey’s HSD post-hoc multiple comparisons using the “emmeans” package. We used t.test to detect significant differences (p < 0.05) between ambient and elevated temperature at each diversity level. We used Pearson correlation to obtain associations between variables. We used piecewise structural equation models to evaluate the indirect effects of microbial diversity and elevated temperature on SOC stabilization through microbial activity (microbial respiration), and the percentage of macroaggregates. The goodness-of-fit statistics were analyzed using Fisher’s C statistic. Following Grace (2006)54, we assessed the conceptual model (full model, Figure S2) versus reduced models by using the AIC to select the final model among alternative models. The final model had the lowest AIC value. The overall model fit was evaluated, with a good fit indicated by a statistically nonsignificant Fisher’s C (p > 0.05). We implemented structural equation models using the “piecewiseSEM” package55. All statistical analyses were conducted in R 4.2.056.

Results

We found that microbial diversity, temperature and the interaction of those factors had significant effects for response variables (Table 1), and we discuss effects by SOC fraction in the following sections.

Table 1 Two-way ANOVA detecting the effects of microbial diversity and temperature on SOC fractions.
Full size table

Total C and microbial respiration

Microbial diversity had a significant effect (Table 1) on total C. At ambient temperature, low diversity had 6.17% and 4.79% lower total C compared to medium and high diversity, respectively. At elevated temperature, these reductions were 6.26% and 3.93%, respectively (Fig. 2a). Both microbial diversity and temperature had significant effects on microbial respiration. Low-diversity treatments had the highest respiration rates, exceeding medium and high diversity by 35.70% and 20.13% at ambient temperature, and by 50.26% and 31.93% under elevated temperature, respectively. Elevated temperature increased respiration by 24.54%, 12.48%, and 13.41% for low, medium, and high diversity, compared to ambient temperature (Fig. 2b).

Fig. 2
figure 2

Effects of microbial diversity and temperature on total carbon (a) and microbial respiration (b). Boxplots represent 25th and 75th percentile, median. Black diamonds represent the mean of each treatment, filled circles represent outliers, blue represents ambient temperature, orange represents elevated temperature. Uppercase letters indicate significant differences among microbial diversity levels (p < 0.05).

Full size image

Total POC and MAOC

Microbial diversity had significant effects on total POC (Table 1). At ambient temperature, low diversity had 13.06% and 8.05% lower POC than medium and high diversity, respectively. At elevated temperature, these reductions were 17.32% and 12.75%, respectively (Fig. 3a). Both microbial diversity and temperature had non-significant effects on total MAOC (Fig. 3b).

Fig. 3
figure 3

Effects of microbial diversity and temperature on total POC (a) and total MAOC (b). Boxplots represent 25th and 75th percentile, median. Black diamonds represent the mean of each treatment, filled circles represent outliers, blue represents ambient temperature, orange represents elevated temperature. Uppercase letters indicate significant differences among microbial diversity levels (p < 0.05).

Full size image

Macroaggregate-occluded POC and MAOC and the percentage of macroaggregates

Microbial diversity had significant effects on microaggregate-occluded POC (Table 1). At ambient temperature, low diversity had 20.3% and 1.5% lower microaggregate-occluded POC compared to medium and high diversity, respectively. At elevated temperature, these reductions were 8.7% and 6.3%, respectively (Fig. 4a). Microbial diversity and temperature had a marginally significant interactive effect on macroaggregate-occluded MAOC (p = 0.054). Ambient temperature had higher macroaggregate-occluded MAOC than elevated temperature with the effect size depending on microbial diversity: the largest difference (38.36%) occurred at medium diversity, followed by 8.78% at low diversity and 9.40% at high diversity (Fig. 4b). Microbial diversity and temperature had a significant interactive effect on the percentage of macroaggregates, showing the same pattern as that of macroaggregate-occluded MAOC, with the largest difference between temperature treatments (19.58%) occurred at medium diversity, followed by 1.7% at low diversity and 3,14% at high diversity. This similarity was supported by a strong positive correlation between the two variables with a strong correlation between the two variables (r = 0.779, p < 0.001; Fig. 4c; Figure S3).

Fig. 4
figure 4

Effects of microbial diversity and temperature on macroaggregate-occluded POC (a), macroaggregate-occluded MAOC (b), and percentage of macroaggregates (c). Boxplots represent 25th and 75th percentile, median. Black diamonds represent the mean of each treatment, filled circles represent outliers, blue represents ambient temperature, orange represents elevated temperature. Different lowercase letters above represent significant differences at p < 0.05. Uppercase letters indicate significant differences among microbial diversity levels (p < 0.05), asterisk indicates significant difference between ambient and elevated temperature (*, p < 0.05). For variables with significant interaction effects, multiple comparisons were conducted across all six treatment combinations (3 microbial diversity levels × 2 temperature regimes), lowercase letters indicate significant differences among six treatments (p < 0.05).

Full size image

Indirect effects of microbial diversity and elevated temperature on SOC stabilization

The structural equation model (Fig. 5a) showed that microbial diversity had an indirect positive effect (r = 0.15) on macroaggregate-occluded POC via decreasing microbial respiration. Temperature had an indirect negative effect (r= -0.12) on macroaggregate-occluded POC via increasing microbial respiration. Microbial diversity had an indirect negative effect (r= -0.39) on macroaggregate-occluded MAOC via decreasing macroaggregates. Temperature had a direct negative effect (r= -0.31, p < 0.01) on macroaggregate-occluded MAOC.

Fig. 5
figure 5

Structural equation model illustrating the pathways of microbial diversity and temperature in regulating SOC distributions in macroaggregates (a) and soil (b). The arrow direction denotes the hypothesized direction of causation, orange and blue reflect negative and positive, respectively. The bar represents the effect size of each factor on POC or MAOC (n = 30). The number next to the arrow is the corresponding standardized coefficient r with significance levels indicated (*p < 0.05, **p < 0.01 and ***p < 0.001, 0.05 < p ≤ 0.10 was considered as marginally significant.). These findings underscore the nuanced interplay between microbial diversity, microbial activity, percentage of macroaggregates and the formation of MAOC in soil. The piecewise SEM fit our data well: (a) Fisher’s C = 14.672, p-value = 0.549; (b) Fisher’s C = 13.806, p-value = 0.464.

Full size image

Microbial diversity had an indirect effect on total POC (r = 0.39) via decreasing soil respiration and macroaggregates (Fig. 5b). Temperature had an indirect negative effect (r= -0.10) on total POC via increasing microbial respiration. Microbial diversity had a direct effect (r = 0.45, p < 0.01) on total MAOC and an indirect negative effect (r= -0.24) on total MAOC via decreasing macroaggregates.

Discussion

Contrary to our hypothesis that higher microbial diversity would result in increased litter decomposition and formation of microbial-derived components that promoting stabilized SOC, we observed that low microbial diversity was associated with enhanced decomposition and macroaggregate-occluded MAOC. Our findings support the hypothesis that elevated temperature increases litter decomposition, but decreases the formation of stabilized SOC, indicated by increased microbial respiration and decreased macroaggregate-occlude MAOC. Additionally, we found interactive effects between microbial diversity and temperature on the formation of macroaggregate-associated MAOC, with the greatest difference between ambient and elevated temperature occurring in soils with medium microbial diversity.

Effect of microbial diversity on the formation of MAOC

We found that low microbial diversity was associated with enhanced decomposition, this is often attributed to the dominance of fast-growing microbes and less interspecific competition for substrates, which tend to have higher metabolic activity57,58. In our study, low-diversity communities also had a higher abundance of Firmicutes and Bacteroidetes, which are known to be fast decomposers59,60 (Figure S1). In contrast, interspecies competition may have suppressed decomposition in soils with medium and high diversity61. This was evidenced by lower remaining total C and POC after incubation compared to the initial C content derived exclusively from the added litter. Because POC is highly susceptible to decomposition, its decomposition primarily contributes to C loss62. The degradation and resynthesis of the POC resulted in the production of small-size microbially processed organic matter particles and microbial-derived C63, which would be incorporated into microaggregates and associated with minerals63. This process also promoted the formation of macroaggregates64. Therefore, low microbial diversity increased macroaggregate-occluded MAOC due to faster decomposition, which increased macroaggregates and internal MAOC.

However, medium diversity under ambient temperature represented an exception to this trend. Medium diversity at ambient temperature exhibited lower decomposition, as evidenced by higher total C, but higher macroaggregates compare to low diversity. This aligns with previous studies reporting humped relationships between microbial diversity and their functions due to higher diversity leading to more intense competition65,66. We also hypothesize that the unique microbial community composition at medium diversity facilitated macroaggregate formation and SOC stabilization. According to Yang43, the medium-diversity inoculum had a higher relative abundance of Basidiomycota and Ascomycota compared to other diversity levels (Figure S1). These fungal taxa are well-documented as effective soil aggregators, producing dense hyphal networks that enhance soil particle cohesion67. It is also reported that some microbial communities are more likely to produce glue-like agents (e.g. microbial extracellular polysaccharides) at certain temperatures, whilst decomposition is kept low68. Thus, it is possible that the medium-diversity microbial community under ambient temperature facilitated macroaggregate formation, but in parallel did not actively decompose organic matter and lost less C compared to the low-diversity microbial community.

Although a microbial community with a higher diversity exhibited slower decomposition, and may be linked to reduced production of microbial-derived components, it had a direct positive effect on total MAOC. It has been proposed that necromass production is not necessarily related to necromass stabilization69, and there are only few studies that directly address the process of necromass stabilization in soils. It is suggested that high diversity can increase microbial CUE39, producing necromass that is more resistant to decomposition70. The inverse effects that microbial diversity exert on the decomposition rate and the fate of decomposition, may lead to a weak relationship between microbial diversity and MAOC formation, which has been previously observed28. Further research is needed to directly address the processes of necromass stabilization and destabilization in soils and their influence on SOC stabilization.

Effects of elevated temperature on the formation of MAOC

Elevated temperature decreased the formation of macroaggregate-occluded MAOC, despite slightly increased decomposition, as reflected by microbial respiration. This discrepancy is likely due to the direct regulation of MAOC formation by temperature. One possible explanation is elevated temperature decreased microbial CUE. At higher temperatures, microbes tend to respire a larger proportion of assimilated C for energy maintenance, reducing CUE71. As a result, less C is allocated to microbial biomass and microbial-derived products, which are critical precursors for MAOC formation72. Additionally, elevated temperatures may accelerate the turnover of microbial biomass, leading to faster loss of microbial-derived C and reduced accumulation in MAOC73,74. Another contributing factor is the simplified organic matter composition of the model soil used in this study. Different organic compounds vary in their sensitivity to temperature75, and the labile organic matter in our study may have fueled rapid microbial growth and respiration at elevated temperatures, further reducing CUE and limiting the stabilization of C as MAOC.

Interactive effects of microbial diversity and elevated temperature on the formation of MAOC

We also found an interactive effect between microbial diversity and temperature, with the greatest difference between ambient and elevated temperature appearing in soils with medium microbial diversity. At elevated temperature, microorganisms that are efficient aggregators seem to be sensitive to warming and were suppressed76. This sensitivity likely explains the pronounced temperature effects in soils with medium diversity, where efficient aggregators are dominate and play a critical role in soil aggregation but are vulnerable to temperature increases. A limitation of our study is that we did not measure microbial community composition after incubation, preventing direct detection of changes under elevated temperature. Nevertheless, our results suggest that the relationship between microbial diversity and MAOC formation is temperature-dependent.

In conclusion, our findings demonstrate the complex relationship between microbial diversity, temperature, and SOC stabilization. Low microbial diversity enhanced decomposition and macroaggregate formation, while medium diversity under ambient temperature facilitated more effective SOC stabilization. However, these benefits were reduced under elevated temperature, highlighting the context-dependent nature of microbial diversity-function relationships and underscoring the need for tailored soil management strategies to enhance SOC stabilization under different environmental conditions. Further research is needed to investigate how microbial diversity and soil aggregation influence SOC stabilization under other global change factors, providing critical insights for improving predictions of SOC dynamics and guiding sustainable land management practices in a changing environment.