Biodegradable Film Mulching Increases Soil Respiration: A Two-Year Field Comparison with Polyethylene Film Mulching in a Semi-Arid Region of Northern China

Abstract

Biodegradable film mulching is increasingly used to replace polyethylene in agriculture, but effects on soil respiration (SR) and components remain unclear, especially during degradation. This study investigated biodegradable mulching’s regulation of SR, root-derived respiration (RDR), and non-root-derived respiration (NRDR) under varying phases. A two-year field experiment was conducted in a rainfed maize system in northern China, comparing conventional tillage with biodegradable film mulching (BM), conventional tillage with polyethylene film mulching (PM), and conventional tillage without mulching (CT). Continuous measurements of soil CO₂ concentration (SCC), temperature, water content, and respiration components were used to assess dynamic responses. Results showed that BM enhanced SR and shifted peak timing, with the SR peaking at 106 days after sowing (DAS) under BM, 91.8 DAS under PM, and 91.2 DAS under CT, mainly through a more sustained RDR (BM peak at 103 DAS with a broader peak and greater cumulative RDR than PM and CT). As the biodegradable plastic film degraded, NRDR was higher during the degradation phase, consistent with a priming-like response. These phase-dependent effects suggest that BM first facilitates root growth then serves as a microbial substrate. Moreover, elevated SCC was positively associated with both RDR and NRDR, indicating that CO₂ may function as a regulatory signal rather than a passive byproduct of respiration. These findings reveal distinct temporal mechanisms by which BM influences soil carbon fluxes and offer mechanistic insights into the sustainable application of biodegradable film mulching. Future research should evaluate long-term effects on microbial community composition, soil carbon balance, and potential trade-offs with crop productivity and environmental risks.

1. Introduction

Polyethylene mulching enhances hydrothermal conditions and yields in arid and semi-arid regions [1,2], but its widespread use raises concerns over soil pollution, residues, and microplastics, threatening sustainability of plastic-covered agriculture [3,4]. Polyethylene film residues hinder water and nutrient flow and root growth, reducing productivity [5], while microplastics in plant tissues pose food chain and health risks [6]. As a promising alternative, biodegradable films have been introduced to replace polyethylene films due to their ability to degrade into environmentally benign compounds under field conditions [7,8]. Despite growing interest, the underlying ecological mechanisms and respiration consequences associated with biodegradable film application remain insufficiently understood.

Soil respiration (SR) is a key component of the terrestrial carbon cycle, reflecting the combined activities of plant roots and microbial communities [9]. It can be partitioned into root-derived respiration (RDR), associated with autotrophic root metabolism, and non-root-derived respiration (NRDR), associated primarily with microbial decomposition of soil organic matter [10,11]. The regulation of these two components is sensitive to changes in soil moisture, soil temperature, and soil CO₂ concentration (SCC), which are all modified by polyethylene film mulching practices [12]. Previous studies have shown that polyethylene film mulching enhances SR mainly through elevating surface temperature and water content, thus stimulating RDR [13,14]. Increased soil temperature can accelerate root metabolic rates and enzymatic activities, thereby enhancing autotrophic respiration [15,16]. Elevated moisture improves root vitality and promotes substrate transport and exchange at the root–soil interface [17]. However, whether biodegradable film exhibits similar effects during their intact phase remains unclear.

Another unresolved issue concerns the role of SCC itself. Under polyethylene or intact biodegradable film mulching, SCC can accumulate to high levels due to restricted gas exchange [18,19]. Elevated SCC levels in the rhizosphere may alter root respiration through feedback inhibition mechanisms, where excess CO₂ suppresses mitochondrial activity or alters pH and ion fluxes in root cells [20]. Moreover, high SCC may shift the composition and activity of microbial communities, potentially influencing NRDR and nutrient cycling [21]. Although the accumulation of soil CO₂ under plastic film mulching is a well-documented phenomenon, studies specifically examining its regulatory impact on RDR and NRDR remain limited [22]. Given the high SCC observed under both polyethylene and biodegradable film mulching, further investigation is needed to understand whether CO₂ acts merely as a byproduct or exerts feedback controls on soil biological processes [23]. Therefore, addressing these questions is crucial for evaluating the ecological sustainability of biodegradable film mulching technologies, particularly in the context of food security, climate mitigation, and microplastic pollution control.

The aim of this study was to evaluate how biodegradable film mulching influences SR and its components under field conditions, with particular attention to its intact and degradation phases. It was hypothesized that intact BM would enhance RDR via improved temperature and moisture, similar to polyethylene film mulching, whereas the degradation phase would stimulate NRDR through priming. Elevated SCC under plastic film mulching conditions was hypothesized to regulate both RDR and NRDR. To test these hypotheses, a two-year field experiment was conducted in a rainfed maize production system in northern China, comparing different plastic film mulching treatments. Continuous measurements of SCC, soil temperature, soil volumetric water content, SR, RDR, and NRDR were performed throughout the growing seasons. The objectives of this study were to determine whether biodegradable films in their intact phase regulate RDR via similar mechanisms to polyethylene film mulching by modifying soil temperature and water content; to explore whether the degradation phase of biodegradable films stimulates NRDR in a manner similar to a priming effect; to assess the regulatory role of SCC in modulating both RDR and NRDR; and to reveal the direct and indirect pathways by which mulching treatments influence SR components using structural equation modeling (SEM).

2. Materials and Methods

2.1. Site Description

The field experiment was conducted at the rainfed spring maize experimental station of the Department of Agro-Environment and Crop Water Productivity, Ministry of Agriculture, China, located in Shouyang County, Shanxi Province (37°51′ N, 113°05′ E; 1130 m a.s.l.). The region experiences a semi-arid continental temperate monsoon climate, with a mean annual precipitation of 470 mm, approximately 70% of which occurs between June and September. The 27-year mean annual temperature is 8.1 °C. The dominant cropping system is a single annual maize (Zea mays L.) crop grown from late April to early October. The soil at the site is classified as Calcaric-Fluvic Cambisols (IUSS Working Group WRB, 2006) with a sandy loam texture and cinnamon soil characteristics.

2.2. Field Experimental Design

Prior to the experiment, the site had been under a continuous spring maize (Zea mays L.) cropping system since 2010. The field trial was conducted during the 2016–2017 growing seasons using a randomized complete block design with three replicates per treatment. Three tillage–mulching treatments were applied: (i) conventional tillage with biodegradable film mulching (BM), (ii) conventional tillage with polyethylene film mulching (PM), and (iii) conventional tillage without mulching (CT). Each plot measured 104 m² (8 m × 13 m). In the CT treatment, all crop residues were removed, and plots were moldboard plowed to a depth of 15–20 cm following the autumn harvest. Seedbeds were prepared in mid-April using tine tillage. In the BM and PM treatments, plastic films were laid immediately after sowing and fertilization, with the edges firmly buried and compacted with soil. To prevent wind displacement, vertical soil belts were established every 2 m. Two weeks after sowing, small square holes (2 cm × 2 cm) were cut in the film to facilitate seedling emergence, and each hole was lightly covered with soil. For the PM treatment, a transparent, non-degradable polyethylene film (thickness: 0.008 mm; width: 0.8 m) was used (Guangzhou Kingfa Company, Guangzhou, China). For the BM treatment, a biodegradable film made of ecoflex^® polymer (thickness: 0.008 mm; width: 0.8 m) was used (BASF SE, Ludwigshafen, Germany). The application rate for both film types was 60 kg ha⁻¹. Plastic films in mulched treatments were removed immediately after crop harvest as part of standard field practice, with the objective of reducing residual film accumulation and minimizing potential impacts on soil nutrient dynamics.

Spring maize cultivar Qiangsheng No. 31 was sown at a density of 55,556 plants ha⁻¹, with a row spacing of 0.60 m and an intra-row spacing of 0.30 m. Seeds were planted to a depth of approximately 3–5 cm using a handheld hole-sowing and fertilization device with a 3 cm diameter. Fertilizer was simultaneously applied at sowing by band placement in the inter-row space using the same device. Sowing was conducted between 25 and 30 April, and harvest took place between 5 and 10 October in both years. All treatments received the same seeding and fertilization rates. Basal fertilization consisted of 150 kg N ha⁻¹ applied as urea [CO(NH₂)₂] and 75 kg P₂O₅ ha⁻¹ as diammonium phosphate [(NH₄)₂HPO₄]. Weed and pest control were also uniformly applied across treatments using 2,4-D butylate herbicide at 0.9 kg active ingredient (a.i.) ha⁻¹ and 40% dimethoate insecticide at 0.3 kg a.i. ha⁻¹, respectively. Across all treatments, fertilization rates and timings, tillage, irrigation, and pest control were kept identical, and no additional soil amendments were applied beyond the prescribed regime.

2.3. Measurement of Soil Temperature and Soil Moisture

Soil temperature was monitored at 5 cm and 15 cm depths in the middle of the crop rows using a temperature data logger (HIOKI LR5011, HIOKI E.E. Corporation, Nagano, Japan). The average of the two soil layer measurements represented soil temperatures in the 0–20 cm soil layer. Measurements were recorded at 30-min intervals throughout the maize growing season. In the BM and PM treatments, temperature sensors were installed beneath the plastic films.

Soil volumetric water content was measured every 7 days during the growing season at nine soil depths (in 0.2 m increments down to 1.8 m) in the middle of the crop rows using a FieldScout TDR-300 moisture meter (Spectrum Technologies, East Plainfield, IL, USA). For the BM and PM treatments, measurements were taken beneath the plastic film. All values were averaged over three replicates per measurement time.

The methods used for monitoring soil temperature and moisture were consistent with those described in a previous publication [24], with the exception that, in the present study, only the 0–20 cm soil layer data were used for analysis.

2.4. Determination of Soil CO₂ Concentration

The SCC in the 0–20 cm soil layer were determined using a custom-made soil gas sampling box in combination with gas chromatography. Each sampler consisted of a perforated plastic cube (2.5 × 2.5 × 2.5 cm³) connected to a three-way valve via transparent tubing; the cube was wrapped with 80-mesh nylon to prevent soil clogging. During sowing and mulching, two samplers per location were installed with the cube ends buried at depths of 5 cm and 15 cm, respectively; the valve stems were passed through the plastic mulch and left above the surface, and the penetration points were carefully sealed with soil. For each treatment, six samplers were installed—three at 5 cm and three at 15 cm—with a spacing of ≥5 m between sampler sets to avoid interference. Soil gas was collected between 09:00 and 11:00 at 7-day intervals throughout the maize growing season. Gas was drawn into evacuated flasks by inserting a needle at the three-way valve and momentarily opening the valve. SCC were analyzed within 24 h using a gas chromatograph (HP 4890D; Agilent Technologies, Santa Clara, CA, USA). The arithmetic mean of the 5 cm and 15 cm measurements was taken as the SCC for the 0–20 cm soil layer. SCC is reported in μmol·mol⁻¹ (ppm).

2.5. Determination of Soil Respiration and Its Components

SR and its components were determined using static chamber-gas chromatography in a stainless-steel chamber composed of a rectangular frame base (length × width × height = 0.5 m × 0.5 m × 0.5 m), a movable intermediate box of 0.5 m in length, width, and height, and a movable top of the same size. The side of the base was covered with round holes 2 cm in diameter to facilitate soil water transport and plant root growth. The outside of the top chamber was covered with white insulation material to minimize solar heating and internal air temperature fluctuations. To improve the accuracy of the static chamber method for measuring SR in the plastic film mulching conditions, the base was placed half on soil and half on plastic film. A modified trenching method was used to distinguish RDR from NRDR [9,25]. Before sowing, any plants including the root in the 1 × 1 m² area around the base was removed. The base was first driven into the soil and then dug back out, covered with nylon mesh bags with an aperture of less than 0.3 mm, and returned to the original position in the soil. Three bases were set in the three repeat experiments under each treatment condition. The base of the static chamber stayed in place throughout the year, and new plants in the base were periodically removed. During measurement, the static chamber was closed for 15 min, and three gas samples and air temperature readings were taken at 0, 5, and 15 min after closure. The frequency and method of gas sample determination were consistent with the method of CO₂ concentration determination. CO₂ flux was calculated using the following equation [26]:

$$\mathrm{F}=\mathsf{\rho }\mathrm{H}\times {\displaystyle \frac{\mathrm{d}\mathrm{c}}{\mathrm{d}\mathrm{t}}}\times {\displaystyle \frac{273}{273+\mathrm{T}}}$$

where F is the CO₂ flux (mg·m⁻²·h⁻¹), ρ is the gas density of CO₂ under a standardized state (g·L⁻¹), H is the height of top chamber, dc/dt is the rate of CO₂ gas concentration in the chamber, T is the air temperature in the chamber.

2.6. Statistical Analysis

Soil temperature was measured at 30-min intervals via automated sensors, while other variables were assessed manually. RDR was computed as SR minus NRDR. Negative values (3.61% of dataset), deemed invalid due to error or variation, were removed; missing data were imputed via linear interpolation (na.approx in zoo package, R). For consistency, all data were aligned to 7-day intervals during the maize growing season. Missing values for any variable on a given date led to exclusion of all data from that date, yielding a synchronized dataset. One-way ANOVA evaluated treatment effects (BM, PM, CT) on SCC, temperature, water content, SR, RDR, and NRDR, with least significant difference (LSD) tests at p < 0.05. The temporal dynamics of SCC, soil temperature, soil moisture, and the three respiration components (SR, RDR, NRDR) during the maize growing season were modeled using non-linear curve fitting. Lorentz or GaussAmp functions were fitted to seasonal trends based on visual suitability, and the curve fitting was performed using Origin software (Origin version 2024, OriginLab Corporation, Northampton, MA, USA). All parameters of the non-linear fitting include y₀ (baseline), xc (peak center), w (width), and A (amplitude).

Associations between environmental factors (temperature, moisture) and respiration components were explored via Pearson correlations and Mantel tests per treatment, with matrices and networks visualized using the R packages linkET (version 0.0.7.4) and ggcor (version 0.9.8.1). SEM examined direct/indirect effects of mulching on environmental factors and respiration. Piecewise SEM was implemented using the R packages piecewiseSEM (version 2.3.0.1) and lme4 (version 1.1-37) and refined by AIC and d-separation tests; standardized coefficients were calculated, and effects partitioned into direct, indirect, and total. Treatment-specific pathways were validated with Bayesian SEMs using the R package brms (version 2.22.0): four MCMC chains (10,000 iterations, 4000 warm-up), convergence checked (R-hat, ESS), posteriors summarized, and models compared via LOOIC. All analyses used R 4.3.2 unless noted.

3. Results

3.1. Soil Volumetric Water Content

Soil volumetric water content was significantly affected by different treatments during the maize growing season (Figure 1a). The PM treatment exhibited the highest average soil water content, followed by CT, while BM showed the lowest values. Statistical analysis confirmed significant differences among all treatments (p < 0.05, LSD_0.05 = 0.97).

The temporal variation in soil volumetric water content under different treatments displayed modest fluctuations over the growing season (Figure 1b). Gaussian amplitude fitting revealed distinct response patterns: the PM and CT treatments showed higher baseline water contents, while BM had a lower baseline. The peak in soil volumetric water content occurred earlier under BM at 87 days after sowing (DAS) and later under PM at 120 DAS, while CT showed no clear peak. The GaussAmp model exhibited only moderate fitting accuracy (R² = 0.09–0.16), reflecting the inherently variable nature of field soil moisture dynamics.

3.2. Soil Temperature Under Different Mulching Treatments

The distribution of soil temperature across the maize growing season was moderately influenced by different treatments (Figure 2a). The PM treatment exhibited the highest average soil temperature, followed by BM, whereas CT had significantly lower temperatures compared to PM (p < 0.05, LSD_0.05 = 0.75), but was not significantly different from BM.

The temporal dynamics of soil temperature under different treatments are illustrated in Figure 2b. Lorentz function fitting revealed a modest early-season temperature peak around 18–19 DAS under both plastic film mulching treatments (BM and PM), with baseline temperatures (y₀) ranging from 19.3 °C to 19.6 °C. The CT treatment presented unstable and biologically implausible fitting parameters, indicating that temperature in CT soils fluctuated more irregularly and was not well captured by the model. The goodness-of-fit was relatively low across all treatments (R² = 0.05–0.09), reflecting the inherent variability in field temperature data.

3.3. Soil CO₂ Concentration

The SCC exhibited significant variation among the three treatments throughout the maize growing season (Figure 3a). On average, the BM treatment resulted in the highest SCC, significantly exceeding those observed under PM and CT (LSD_0.05 = 905.86, p < 0.05), while the difference between PM and CT was not statistically significant.

The temporal variation in SCC followed a unimodal pattern for all treatments, with peaks occurring during the mid-growth stages (Figure 3b). Peak times were 92.9 DAS for BM, 88.8 DAS for PM, and 95.8 DAS for CT. Among them, BM showed both the highest peak value and the largest area under the curve (A), indicating stronger and more sustained CO₂ accumulation. The Lorentz models yielded moderate to good fits, with R² values ranging from 0.41 (CT) to 0.54 (BM), suggesting reasonable model adequacy.

3.4. Soil Respiration

SR differed significantly among the three treatments throughout the maize growing season (Figure 4a). BM exhibited the highest average SR rate, followed by PM, with CT showing the lowest levels. One-way ANOVA confirmed significant differences among all treatments (p < 0.05, LSD_0.05 = 121.86), indicating that mulching—particularly with biodegradable films—substantially promoted SR, likely by enhancing microbial and root activity.

The temporal dynamics of SR are shown in Figure 4b. Lorentz function fitting effectively captured the seasonal respiration patterns, with peak timing at 106 DAS under BM, 91.8 DAS under PM, and 91.2 DAS under CT. Notably, BM exhibited the largest peak width and the greatest area under the curve, suggesting more prolonged and intensive respiration activity. The goodness-of-fit was acceptable, with R² values ranging from 0.20 under CT to 0.37 under BM.

3.5. Root-Derived Respiration

RDR differed significantly among the three treatments over the course of the maize growing season (Figure 5a). The BM treatment exhibited the highest RDR levels, significantly exceeding those under CT (p < 0.05, LSD_0.05 = 108.38), likely due to improved soil conditions supporting root development and respiration. Although the difference between BM and PM was not statistically significant, both plastic mulching treatments (BM and PM) promoted greater root-respired CO₂ fluxes compared to the CT treatment.

Distinct temporal patterns of RDR were observed among treatments (Figure 5b). The respiration peak occurred later under BM (103 DAS) than under PM (82.5 DAS) but earlier than under CT (123 DAS). BM also exhibited the broadest peak width, indicating a more sustained root-respiration phase. By contrast, PM peaked earlier, whereas CT showed a delayed but narrow peak, suggesting a brief root-respiration response. Cumulative RDR, as indicated by the area under the curve, was substantially greater under BM compared to PM and especially CT, underscoring the effectiveness of biodegradable film mulching in sustaining root activity. Model fits were acceptable for BM and PM (R² = 0.17), while the poorer fit under CT (R² = 0.05) likely reflects higher variability in root respiration in no-mulched soils.

3.6. Non-Root-Derived Respiration

NRDR varied significantly across treatments during the maize growing season (Figure 6a). The BM treatment exhibited the highest average NRDR rate, significantly exceeding those under PM and CT (p < 0.05, LSD_0.05 = 59.65), while no significant difference was detected between PM and CT.

The temporal dynamics of NRDR were well described by Lorentz function fits (Figure 6b), revealing delayed and sustained microbial activity under BM. The peak respiration occurred the latest under BM (120 DAS), compared to PM (107 DAS) and CT (102 DAS), suggesting prolonged microbial stimulation under biodegradable film mulching. However, the cumulative NRDR, as indicated by the area under the curve, was highest under PM, followed by BM and CT, indicating that while BM prolonged microbial activity, PM achieved the highest cumulative microbial respiration. The model fit was strongest under BM (R² = 0.40), followed by PM (R² = 0.34) and CT (R² = 0.19), reflecting clearer and more structured temporal responses under mulched conditions.

3.7. Relationships Between Soil Respiration Components and Environmental Factors

The correlations among three respiration components (SR, RDR, NRDR, SCC) and environmental factors (soil volumetric water content, soil temperature) differed under the three treatments (Figure 7a–c). Pearson and Mantel tests jointly revealed treatment-specific interaction patterns.

Under BM treatment (Figure 7a), SR exhibited a strong positive correlation with RDR (r > 0.9, p < 0.01) and a moderate correlation with SCC (r ≈ 0.6, p < 0.01), indicating that both root-derived and CO₂-mediated processes contributed substantially to SR. Significant Mantel correlations were observed between soil temperature and both SR and RDR (Mantel’s r > 0.2, p < 0.05), suggesting that temperature was a key environmental driver of root respiratory activity under biodegradable film mulching.

In the PM treatment (Figure 7b), SR remained positively correlated with RDR (r > 0.9, p < 0.01), while its correlation with NRDR was weak and statistically insignificant. Notably, soil volumetric water content showed a significant positive Mantel correlation with NRDR (Mantel’s r ≈ 0.2–0.4, p < 0.05), indicating that microbial respiration under polyethylene film mulching was sensitive to soil moisture. RDR and NRDR exhibited a negative Pearson correlation, albeit non-significant. SCC showed moderate positive correlations with both RDR and SR (r ≈ 0.4–0.6, p < 0.05), suggesting a partial regulatory role of SCC in total SR.

In contrast, the CT treatment (Figure 7c) exhibited a relatively reduced network connectivity among environmental and respiration variables. Although SR remained strongly correlated with RDR (r > 0.9, p < 0.01), other Pearson correlations, such as those between SCC and SR or RDR, were only moderate (r ≈ 0.5, p < 0.01). Notably, Mantel tests revealed minimal or no significant spatial correlations across most variable pairs, suggesting that the absence of plastic film mulching led to more variable and less spatially coordinated respiration–environment relationships.

3.8. Structural Equation Modeling of Treatment Effects on Soil Respiration Components

The SEM (Figure 8a) assessing the effects of tillage–mulching treatments on SR and its components (RDR and NRDR) demonstrated a good fit to the data (Fisher’s C = 10.161, p = 0.75, df = 14, n = 360). The model explained 84% of the variance in soil temperature (conditional R² = 0.84, marginal R² = 0.01), 63% in soil volumetric water content (conditional R² = 0.63, marginal R² = 0.20), 88% in SCC (conditional R² = 0.88, marginal R² = 0.09), 54% in NRDR (conditional R² = 0.54, marginal R² = 0.06), 40% in RDR (conditional R² = 0.40, marginal R² = 0.04), and 82% in SR (conditional R² = 0.82, marginal R² = 0.81).

Significant direct paths (p < 0.05) included positive effects from soil volumetric water content to SCC (standardized path coefficient = 0.289), soil volumetric water content to RDR (0.128), SCC to SR (0.139), RDR to SR (0.283), NRDR to SR (0.409), and RDR to NRDR (0.047). Treatments exerted significant direct effects on multiple variables: BM negatively affected soil volumetric water content (−0.277,) but positively influenced SCC (6.402), RDR (1.963), NRDR (3.314), and SR (4.032); PM positively affected soil volumetric water content (0.210), SCC (6.233), RDR (1.830), NRDR (3.095), and SR (4.071); CT positively influenced SCC (6.096), RDR (1.729), NRDR (3.079), and SR (3.972). No significant direct effects were observed from treatments to soil temperature (all p > 0.05) or from soil temperature to soil volumetric water content and RDR. From CT to soil volumetric water content was also non-significant.

Effect partitioning revealed varying direct, indirect, and total effects of treatments on NRDR, RDR, and SR (Figure 8b). For NRDR, BM exhibited the highest total effect (3.404), driven primarily by a direct effect (3.314) with a minor positive indirect effect (0.090); PM showed a total effect of 3.182 (direct: 3.095, indirect: 0.087); and CT had a total effect of 3.160 (direct: 3.079, indirect: 0.081). For RDR, BM had the highest total effect (1.929), with a strong direct effect (1.963) offset by a small negative indirect effect (−0.035); PM displayed a total effect of 1.863 (direct: 1.830, indirect: 0.033); and CT showed a total effect of 1.730 (direct: 1.729, indirect: 0.002). For SR, PM yielded the highest total effect (5.474), comprising a direct effect (4.071) and indirect effect (1.404); BM followed with a total effect of 5.458 (direct: 4.032, indirect: 1.425); and CT had the lowest total effect (5.312; direct: 3.972, indirect: 1.340).

3.9. Pathway Differentiation Across Treatments: SEM-Based Insights

To elucidate treatment-specific mechanisms regulating SR and its components, separate SEMs were fitted for each treatment, revealing distinct causal relationships among soil temperature, soil volumetric water content, SCC, RDR, NRDR, and SR (Figure 9).

For the BM treatment (Figure 9a), the model explained 60% of the overall variance (R² = 0.60, LOOIC = 486.6, n = 120). Variance explained for individual variables was 0.699 for soil temperature, 0.566 for soil volumetric water content, 0.976 for SCC, 0.891 for NRDR, 0.765 for RDR, and 0.927 for SR. Significant positive paths included NRDR to SR (posterior median = 0.46) and RDR to SR (0.406). Non-significant paths encompassed soil temperature to soil volumetric water content (−0.02, ns), soil temperature to RDR (0.064, ns), soil volumetric water content to RDR (0.221, ns), soil volumetric water content to SCC (0.021, ns), RDR to NRDR (0.058, ns), and SCC to SR (0.066, ns).

For the PM treatment (Figure 9b), the model explained 62% of the overall variance (R² = 0.62, LOOIC = 479.9, n = 120). Variance explained for individual variables was 0.806 for soil temperature, 0.609 for soil volumetric water content, 0.974 for SCC, 0.948 for NRDR, 0.767 for RDR, and 0.839 for SR. Significant positive paths included NRDR to SR (0.206), RDR to SR (0.432), and SCC to SR (0.241). Non-significant paths included soil temperature to soil volumetric water content (−0.156, ns), soil temperature to RDR (−0.209, ns), soil volumetric water content to RDR (0.17, ns), soil volumetric water content to SCC (0.006, ns), and RDR to NRDR (−0.089, ns).

For the CT treatment (Figure 9c), the model explained 65% of the overall variance (R² = 0.65, LOOIC = 203.3, n = 120). Variance explained for individual variables was 0.924 for soil temperature, 0.559 for soil volumetric water content, 0.992 for SCC, 0.937 for NRDR, 0.846 for RDR, and 0.951 for SR. Significant positive paths included soil temperature to RDR (0.293), NRDR to SR (0.495), RDR to SR (0.44), and SCC to SR (0.08). Non-significant paths included soil temperature to soil volumetric water content (−0.072, ns), soil volumetric water content to RDR (0.219, ns), soil volumetric water content to SCC (−0.016, ns), and RDR to NRDR (0.069, ns).

4. Discussion

4.1. Effects of Different Mulch Treatments on Soil Hydrothermal Conditions and CO₂ Concentrations

The observed variations in soil volumetric water content, soil temperature, and SCC across treatments (Figure 1, Figure 2 and Figure 3) highlight the differential impacts of mulching on soil microenvironments, supporting the hypothesis that intact films improve hydrothermal conditions for plant growth. PM maintained the highest soil moisture due to reduced evaporation and enhanced retention, as shown in meta-analyses from northwestern China indicating up to 20–30% increases in water storage in semi-arid zones [1,2,3]. Conversely, BM showed the lowest moisture, likely from partial degradation increasing permeability and water loss mid-to-late season, aligning with findings that biodegradable films degrade after ~65–70 days in semi-arid fields [24,27]. This may reduce long-term water holding by 1.5–3.0% compared to PM [28]. CT exhibited intermediate levels, reflecting natural fluctuations in rainfed systems that worsen drought stress [29,30].

Soil temperature followed a similar pattern, with PM inducing the highest averages, followed by BM, due to the greenhouse-like effect of intact films trapping solar radiation and reducing heat loss [31]. The early-season peaks under BM and PM (18–19 DAS) suggest rapid warming benefits for seedling establishment, corroborating hypotheses that mulching accelerates soil heating in semi-arid climates by 2–3 °C, thereby advancing phenological stages and improving crop emergence rates [32]. In the later stages of the growing season, especially after crop canopy closure, the fragmentation of biodegradable film under BM may have led to increased soil moisture fluctuations, which in turn influenced soil temperature dynamics [33]. However, overall temperature trends under BM remained largely consistent with those observed under PM, suggesting that temperature regulation was not substantially compromised by BM degradation.

Notably, BM produced the highest SCC (peak 92.9 DAS), surpassing PM and CT, challenging assumptions of passive accumulation from restricted exchange. This rise likely arises from microbial decomposition during degradation, where ecoflex polymers act as labile carbon, stimulating heterotrophic respiration [34,35]. Phase-dependent dynamics support evidence of amplified emissions via priming in later growth [18], with oasis studies reporting 18.2% CO₂ increases from reduced diffusivity and microbial activity [36]. Optimized biodegradable films might cut emissions by 10–20% via better organic matter integration, stressing film composition [23]. These hydrothermal and CO₂ shifts set the stage for differential respiration responses, underscoring the need for phase-specific management—such as hybrid mulching systems—to mitigate potential greenhouse gas emissions while maximizing water and temperature benefits in semi-arid agroecosystems. Across treatments, SCC co-varied with both RDR and NRDR, suggesting that SCC may act as a potential regulatory cue rather than a sole driver. Mechanistically, the CO₂↔HCO₃⁻ interconversion catalyzed by carbonic anhydrases in soils can modify microsite pH and dissolved inorganic carbon availability, thereby influencing extracellular enzyme kinetics and microbial metabolism [37,38,39]. In addition, elevated SCC can indirectly reshape rhizosphere communities through plant C allocation, potentially altering substrate supply to microbes [21,40].

4.2. Mechanistic Responses of Soil Respiration and Its Components to Different Mulching Treatments

The mechanistic responses of total SR, RDR, and NRDR to mulching treatments (Figure 4, Figure 5 and Figure 6) reveal phase-dependent effects, fully validating the hypothesis that BM promotes RDR early on while stimulating NRDR later, with implications for carbon cycling in dryland farming. BM showed the highest average SR (Figure 4a) with a prolonged peak (106 DAS), reflecting sustained efflux from autotrophic and heterotrophic sources, aligning with early hydrothermal benefits similarly to PM, though subsequent degradation may introduce complex and uncertain impacts [41].

For RDR, BM and PM both elevated rates compared to CT (Figure 5a), with BM showing the broadest and latest peak (103 DAS), suggesting a more extended root activity facilitated by initial moisture and temperature gains. This supports the hypothesis that intact biodegradable films boost autotrophic respiration via enhanced root vitality, substrate availability, and enzymatic kinetics, akin to polyethylene mulches, where studies report 20–30% increases in root respiration due to warmer, moister rhizospheres [42,43]. The delayed peak under CT implies slower root development without mulch protection, potentially linked to reduced nutrient uptake and oxidative stress from temperature fluctuations [44].

NRDR under BM was markedly higher (Figure 6a), with the latest peak (120 DAS), suggesting a priming-like effect during degradation, where film fragments serve as microbial substrates, accelerating organic matter decomposition and enzyme production. In principle, biodegradable mulching film fragments can influence microbial respiration through two non-exclusive pathways: (i) pre-incorporation microclimate effects (barrier-driven changes in temperature, moisture, and gas diffusion) and (ii) post-incorporation carbon inputs from biodegradable fragments and additives that are metabolized by soil microbes [34,35]. The latter pathway aligns with the classical priming effect—stimulation of native SOM mineralization by inputs of fresh, labile C—extensively documented across ecosystems [45,46,47,48]. This phase-specific stimulation contrasts with PM, where cumulative NRDR was highest but peaked earlier, likely due to stable microbial habitats without degradation inputs, fostering consistent heterotrophic activity [49]. CT’s lower NRDR underscores mulch’s role in microbial activation, but BM’s priming may enhance soil carbon sequestration if decomposed residues incorporate into stable humus fractions, as evidenced by a recent study showing net carbon gains in dryland systems under continuous biodegradable mulching [23]. These findings extend prior work on biodegradable mulches, highlighting potential negative allelopathic effects on plants from residues but positive microbial priming, which could increase soil microbial biomass by 15–25% over multiple years [50]. Overall, BM modulated the timing of respiration components—higher RDR during the intact phase and higher NRDR during the degradation phase—thereby re-timing seasonal CO₂ efflux.

4.3. Environmental Drivers and Pathway Regulation of Soil Respiration Components Under Different Mulching Treatments

Correlation and SEM analyses (Figure 7, Figure 8 and Figure 9) elucidate the environmental drivers and regulatory pathways, confirming CO₂’s feedback role and treatment-specific mechanisms, while revealing nuanced interactions that deepen our understanding of mulch-mediated carbon fluxes. Under BM, strong SR-RDR correlations (r > 0.9) and temperature-SR/RDR Mantel links highlight temperature as a primary driver of autotrophic processes, supporting hydrothermal hypotheses by accelerating root metabolic rates and enzyme activities like cytochrome oxidase [51]. PM’s moisture-NRDR associations emphasize heterotrophic sensitivity to water, potentially via improved substrate diffusion and microbial hydration, whereas CT’s weaker networks indicate decoupled responses without mulch, leading to stochastic respiration patterns influenced by precipitation events [52].

Pooled SEM (Figure 8) explained substantial variance (e.g., 82% in SR), with positive paths from moisture to CO₂/RDR and CO₂/RDR/NRDR to SR, validating CO₂’s regulatory feedback—potentially via pH shifts, ion fluxes, or microbial signaling that modulates respiration enzymes—beyond byproduct status [9]. BM’s negative indirect effect on RDR (−0.035) via moisture loss during degradation offsets direct positives, illustrating a trade-off where late-stage permeability increases CO₂ efflux but reduces water-mediated root support [53,54]. PM’s slightly higher total SR effect (5.474) compared to BM (5.458) reflects a balanced contribution of direct and indirect pathways, likely amplified by consistent CO₂ accumulation that minimizes feedback inhibition more effectively than in BM, where degradation-induced CO₂ spikes may enhance regulatory suppression [18,34].

Treatment-specific SEMs (Figure 9) revealed distinct regulations: BM’s strong NRDR/RDR-to-SR paths (0.46/0.406) suggest microbial-root synergy without CO₂ mediation, driven by degradation-primed carbon inputs; PM’s CO₂-to-SR path (0.241) indicates CO₂-driven feedbacks, possibly through rhizosphere acidification enhancing nutrient solubilization; CT’s temperature-to-RDR (0.293) emphasizes abiotic controls, with weaker CO₂ roles due to better gas exchange. These align with SEM applications in mulching studies, where physicochemical properties explain 60–80% of respiration variability, but a Bayesian approach adds robustness by accounting for uncertainty in semi-arid data [19,55]. Limitations of this study include the lack of destructive root measurements at key phenological stages, the absence of repeated soil organic carbon (SOC) stock measurements on an equivalent soil mass basis with bulk density for depth-explicit estimates (0–20 and 20–40 cm), unquantified residue and root-C inputs, and the absence of direct microbial indicators such as the fungal-to-bacterial ratio, extracellular enzyme activities, or sequencing-based composition. The experiment covered only two seasons at a single site, and therefore inferences about net C storage or sequestration cannot be drawn from CO₂ fluxes alone. To advance understanding, future research should integrate metagenomic approaches to elucidate microbial community functions and their roles in soil respiration, apply isotopic tracing to refine carbon-flux pathways and distinguish autotrophic from heterotrophic sources [56], and pair flux partitioning with root metrics and repeated SOC monitoring using the equivalent soil mass approach across multiple years. In addition, long-term studies are needed to evaluate whole-system carbon balances under climate variability, and the development and testing of hybrid mulching strategies tailored to drylands in northern China could help optimize carbon retention while minimizing CO₂ emissions.

5. Conclusions

This two-year field experiment in a rainfed maize system in northern China’s semi-arid region demonstrates that BM exerts phase-dependent effects on SR and its components, offering a sustainable alternative to PM while addressing plastic pollution concerns. Key findings reveal that BM significantly enhances SR by promoting RDR during the intact phase through improved soil hydrothermal conditions, akin to PM, and is associated with higher NRDR during the degradation phase, consistent with a priming-like response. Under mulching, elevated SCC is interpreted as a potential regulatory cue—not merely a byproduct—with positive associations to both RDR and NRDR, consistent with SEM results.

Overall, the phase-specific responses indicate that BM re-times seasonal carbon efflux—with higher RDR during the intact phase and higher NRDR during the degradation phase. Any implication for long-term soil-carbon retention remains uncertain and depends on the balance between carbon inputs and respiratory losses, which were not quantified in this study. Compared with PM and CT, BM offers dual-phase benefits while mitigating microplastic risks and maintaining yield—supporting hydrothermal conditions in semi-arid systems. Nevertheless, trade-offs—including mid-season moisture loss and possible residue-related constraints—underscore the need for optimized film formulations and management before broader deployment.