Predicting Escherichia coli levels in manure using machine learning in weeping wall and mechanical liquid solid...

Area of Science:

• Environmental Science
• Agricultural Engineering
• Microbiology

Background:

• Dairy manure management is crucial for public and environmental health.
• Alternative Dairy Effluent Management Strategies (ADEMS) are being developed to improve efficiency.
• Solid separation systems, like mechanical solid-liquid separators (SLS) and gravitational weeping walls (WW), are key components of ADEMS.

Purpose of the Study:

• To compare the chemical, physical, and biological parameters of dairy effluent treated by SLS and WW systems.
• To assess the effectiveness of these systems in reducing microbial pollution, using Escherichia coli as an indicator.
• To develop a predictive model for effluent quality and microbial risk.

Main Methods:

• Pilot study with 96 samples analyzing various parameters (sodium, potassium, salts, volatile solids, pH, E. coli).
• Comparison of solid and liquid fractions from SLS and WW systems.
• Development and validation of a machine learning model (E-C-MAN) for predicting E. coli levels.

Main Results:

• Significant differences in chemical, physical, and biological parameters (including E. coli) were observed between SLS and WW systems.
• The solid fraction of dairy effluent showed the lowest E. coli levels, indicating potential for microbial pollution control.
• The E-C-MAN model demonstrated a reliable framework for predicting E. coli in treated manure.

Conclusions:

• SLS and WW systems differentially impact dairy effluent characteristics.
• Separating the solid fraction of manure is effective in reducing E. coli levels.
• The E-C-MAN machine learning model offers a promising tool for predicting microbial pollution risk in ADEMS.