Uncovering time-lagged causal mechanisms in hyperthermophilic composting using Peter-Clark momentary conditional...

Area of Science:

• Environmental Science
• Microbiology
• Biogeochemistry

Background:

• Composting is a complex process influenced by pile structure, microbial activity, and environmental factors.
• Time-lagged effects and feedback loops are critical but often overlooked in composting systems.
• Understanding causal relationships is key to optimizing composting efficiency and product quality.

Purpose of the Study:

• To apply the Peter-Clark Momentary Conditional Independence (PCMCI) causal discovery framework to hyperthermophilic composting.
• To reconstruct multi-lag causal networks for rice husk hyperthermophilic composting (RHC) and corn stalk hyperthermophilic composting (CSC).
• To elucidate the distinct regulatory architectures and causal drivers in RHC and CSC.

Main Methods:

• Utilized the PCMCI causal discovery framework on multi-source time-series data from RHC and CSC.
• Reconstructed multi-lag causal networks to identify time-lagged effects and feedback loops.
• Compared PCMCI with correlation analysis and structural equation modeling for causal inference.

Main Results:

• RHC exhibited a regulatory network where free air space (FAS) consistently drove moisture, pH, and nitrogen dynamics, supporting carbon degradation and humic substance formation.
• CSC showed a more constrained network, with microbial succession and nitrogen transformation being dominant; NH4+-N significantly delayed germination index, hindering maturity.
• PCMCI successfully resolved causal directionality, time delays, and feedback structures, outperforming traditional methods.

Conclusions:

• The distinct causal networks explain the superior performance of RHC over CSC.
• FAS is a critical upstream driver in RHC, while nitrogen dynamics pose a bottleneck in CSC.
• The PCMCI framework offers a transferable approach for mechanistic inference and process optimization in composting.