Chemical and microbial characterization of sugarcane mill mud for soil applications

Abstract

Sugarcane mill mud/filter cake is an activated sludge-like byproduct from the clarifier of a raw sugar production factory, where cane juice is heated to ≈90°C for 1-2 hr, after the removal of bagasse. Mill mud is enriched with organic carbon, nitrogen, and nutrient minerals; no prior report utilized 16S rRNA gene sequencing to characterize the microbial composition. Mill mud could be applied to agricultural fields as biofertilizer to replace or supplement chemical fertilizers, and as bio-stimulant to replenish microorganisms and organic carbon depleted by erosion and post-harvest field burning. However, mill mud has historically caused waste management challenges in the United States. This study reports on the chemical and microbial (16S rRNA) characteristics for mill muds of diverse origin and ages. Chemical signature (high phosphorus) distinguished mill mud from bagasse (high carbon to nitrogen (C/N) ratio) and soil (high pH) samples of diverse geographical/environmental origins. Bacterial alpha diversity of all sample types (mill mud, bagasse, and soil) was inversely correlated with C/N. Firmicutes dominated the microbial composition of fresh byproducts (mill mud and bagasse) as-produced within the operating factory. Upon aging and environmental exposure, the microbial community of the byproducts diversified to resemble that of soils, and became dominated by varying proportions of other phyla such as Acidobacteria, Chloroflexi, and Planctomyces. In summary, chemical properties allowed grouping of sample types (mill mud, bagasse, and soil-like), and microbial diversity analyses visualized aging caused by outdoor exposures including soil amendment and composting. Results suggest that a transient turnover of microbiome by amendments shifts towards more resilient population governed by the chemistry of bulk soil.

Fig 1. Chemical analysis of bagasse, soil, and mud collected from sugarcane production facilities.

(a) Principal components analysis and (b) heatmap with complete linkage clustering. Results were normalized to within row variance with the colored scale bar representing standard deviations from within-row means [23].

Fig 2. Average relative abundance of 16S rRNA genes from the 5 most abundant phyla.

Fig 3. Alpha diversity analyses of samples based on observed 16S rRNA gene (97% similarity).

Fig 4. Correlations between C:N ratios and alpha diversity metrics.

For ease of comparison, Faith’s index values were divided by 100.

Fig 5. Beta diversity analyses of samples based on the relative abundances of the top 10 phyla.

(a) Principal component and (b) heatmap with complete linkage clustering.