Upcycling human excrement: the gut microbiome to soil microbiome axis

Abstract

Human excrement composting (HEC) is a sustainable strategy for human excrement (HE) management that recycles nutrients and mitigates health risks while reducing reliance on freshwater, fossil fuels, and fertilizers. A mixture of HE and bulking material was collected from 15 composting toilets and composted as 15 biological replicates in modified 19-liter buckets under mesophilic conditions with weekly sampling for one year. We hypothesized that (i) the microbiome of 1 year old compost would resemble that of a soil and/or food and landscape waste compost microbiome more closely than the original HE; and (ii) the human fecal indicators, Escherichia coli and Clostridium perfringens, would be undetectable after 52 weeks using qPCR and culturing. This investigation identified unique successional trajectories within buckets (i.e. biological replicates) and significant shifts in microbial communities around 25 weeks across buckets, with reductions in fecal-associated taxa and increases in environmental taxa indicating effective composting. We present a comprehensive microbial time series analysis of HEC and show that the initial gut-like microbiome of HEC systems transitions to a microbiome similar to soil and traditional compost but that pathogen risk assessment is important if thermophilic temperatures are not achieved. This study also produced the highest resolution composting microbiome data to date, establishing a baseline for HEC optimization and thermophilic composting studies while serving as a resource for bioprospecting for enzymes and organisms relevant to upcycling waste.

Keywords: 16S; amplicon; biosolids; compost; compost toilet; human excrement; latrines; manure; microbiome; soil; sustainability.

Figure 1

Distinct trajectories of HEC microbiomes along the gut-to-soil microbiome axis across four representative composting buckets, as visualized by two-dimensional PCoA plots. Two-dimensional principal coordinate analysis (PCoA) was applied to the unweighted UniFrac distance matrix computed between all samples including soil and food and landscape waste compost (FLWC) reference samples. The trajectory through the ordination space for individual buckets is plotted with the statistical mean of ordination results across all 15 buckets (denoted by stars connected by a line) at each composting time point—defined as the number of weeks since HEC was retrieved from the participant and actively managed (Fig. S1). Averages are not points within the ordination, but rather overlaid on the ordination plot. The average trajectory exhibits a progressive movement away from HE along Axis 1 over the course of one year, with final TPs clustering closer to reference samples. Images of the composting toilets corresponding to each bucket are included to illustrate the variability between systems. (A) Bucket 1’s transition rapidly diverges from HE and its rate of transition appears to decrease around week 11 with minimal further variation. (B) Bucket 4’s initial transition occurs rapidly, beginning and ending less similar to fecal material than average. (C) Bucket 5 begins more similar to HE and has the greatest extent of change relative to all buckets ending with a microbiome profile similar to reference samples and B4. (D) Bucket 7’s transition appears stunted, beginning and ending more similar to HE than average. These microbiome composition data do not offer sufficient resolution to provide clear identification of all pathogens or provide insights into microbial activity; future studies will address these aspects. Definitions: Extent—Euclidean distance between two data points in the ordination. Rate—extent divided by the difference in time between two data points in the ordination. A higher rate is illustrated by consecutive data points spaced farther apart, indicating rapid changes. Conversely, a lower rate is characterized by data points that remain relatively close. Trajectory—successional pattern of the microbiomes, illustrated by consecutive data points connected by a line on the ordination plot for each bucket. Some trajectories span the entire plot, reflecting substantial changes, while others display trajectories that cluster tightly, indicating reduced variability (i.e., high stability) in microbiome composition. Plots for all 15 replicates are presented in Fig. S2. Created in BioRender.

Figure 2

qPCR and culturing data for select buckets. (A) Mean copy number of E. Coli across all 16 composting buckets at each composting TP. Elevated variability during the first 15 weeks reflects heterogeneity in abundance across systems. Copy numbers declined over time, likely due to increasingly unfavorable environmental conditions for E. Coli. (B) Mean copy number of C. Perfringens across all 16 composting buckets at each composting TP. Elevated variability during the final 15 weeks reflects heterogeneity in abundance across buckets. Copy numbers increased over time, likely due to increasingly favorable environmental conditions for C. Perfringens. (C) Mean most probable number (MPN/g) of E. Coli across all 16 composting buckets at each composting TP. No culturable E. Coli was detected after TP-25. (A–C) Error bars represent ±1 standard deviation. Data points above the 99.9th percentile were excluded to reduce the influence of outliers, except in (C), where insufficient data precluded this filtering. (D) Copy numbers (solid line) of the uidA gene found in E. Coli normalized by 16S rRNA gene copy number data from the BactQuant assay (indicative of total bacterial load). By TP-52, each bucket remained below the LOQ. (E) Copy numbers (solid line) of the C. Perfringens group normalized by BactQuant. By TP-52, each bucket remained above the LOQ. A moving average period of four TPs was applied to (D) and (E). A pseudo-count, defined as the floor of the log10 of the minimum, minus one, was added to all data points but varied by plot because the minimum nonzero quantification varies by plot. The y-axis scale is not consistent across buckets to highlight individual variability in copy numbers. Significant fluctuations in data are attributed to sampling variability, given that TaqMan PCR analysis was conducted in triplicate. The LOQ is 304 (dashed line) normalized by BactQuant and fluctuates over time as a result. The limit of detection (LOD) of 30.4 (not displayed in the figure), was also normalized by BactQuant. Normalized data for all buckets can be found in Fig. S4 and raw results of the BactQuant assay are presented in Fig. S5. Non-normalized data for all buckets can be found in Fig. S6. (F) Biweekly samples (odd TPs) were cultured in EC broth with 4-methylumbelliferyl-β-D-glucuronide (MUG) for the detection of E. Coli. Data is presented linearly and the y-axis is not consistent across buckets to highlight the variability in the composting systems. Culturing data for all buckets can be found in Fig. S7. Created in BioRender.