The influence of environmental factors on the detection and quantification of SARS-CoV-2 variants in dormitory wastewater at a primarily undergraduate institution

Abstract

Testing for the causative agent of coronavirus disease 2019 (COVID-19), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been crucial in tracking disease spread and informing public health decisions. Wastewater-based epidemiology has helped to alleviate some of the strain of testing through broader, population-level surveillance, and has been applied widely on college campuses. However, questions remain about the impact of various sampling methods, target types, environmental factors, and infrastructure variables on SARS-CoV-2 detection. Here, we present a data set of over 800 wastewater samples that sheds light on the influence of a variety of these factors on SARS-CoV-2 quantification using droplet digital PCR (ddPCR) from building-specific sewage infrastructure. We consistently quantified a significantly higher number of copies of virus per liter for the target nucleocapsid 2 (N2) compared to nucleocapsid 1 (N1), regardless of the sampling method (grab vs composite). We further show some dormitory-specific differences in SARS-CoV-2 abundance, including correlations to dormitory population size. Environmental variables like precipitation and temperature show little to no impact on virus load, with the exception of higher temperatures for grab sample data. We observed significantly higher gene copy numbers of the Omicron variant than the Delta variant within ductile iron pipes but no difference in nucleocapsid abundance (N1 or N2) across the three different sewage pipe types in our data set. Our results indicate that contextual variables should be considered when interpreting wastewater-based epidemiological data.

Importance: Testing for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), has been crucial in tracking the spread of the virus and informing public health decisions. SARS-CoV-2 viral RNA is shed by symptomatic and asymptomatic infected individuals, allowing its genetic material to be detected and quantified in wastewater. Here, we used wastewater-based epidemiology to measure SARS-CoV-2 viral RNA from several dormitories on the Appalachian State University campus and examined the impact of sampling methods, target types, environmental factors, and infrastructure variables on quantification. Changes in the quantification of SARS-CoV-2 were observed based on target type, as well as trends for the quantification of the Delta and Omicron variants by sampling method. These results highlight the value of applying the data-inquiry practices used in this study to better contextualize wastewater sampling results.

Keywords: SARS-CoV-2; droplet digital PCR; wastewater-based epidemiology.

Fig 1

Pairwise comparisons of the natural log of the target copies per liter by sampling method (grab, composite) and nucleocapsid gene target (N1 [black], N2 [yellow]). There was no significant difference between nucleocapsid gene targets within grab samples (Kruskal-Wallis, P = 0.46) or within composite samples (Kruskal-Wallis, P = 0.63). There was also no significant difference between grab and composite samples across both gene targets (Wilcoxon signed-rank test, P = 0.16).

Fig 2

Pairwise comparisons of the natural log of the target copies per liter by variant gene target (Delta, Omicron) and sampling method (grab [black], composite [yellow]). There was no significant difference between the sampling method for either variant of concern Delta (Kruskal-Wallis, P = 0.55) or Omicron (Kruskal-Wallis, P = 0.053). There was a significant difference between the detection of Delta and Omicron (Wilcoxon signed-rank test, P = 0.011).

Fig 3

Correlations between natural log target copies per liter for N1 and N2 gene targets for SARS-CoV-2 from dormitory wastewater samples and (A) total dormitory population, (B) vaccination rate per dormitory, and (C) total positive test results across university testing site(s). Correlations were tested using generalized linear models with Gaussian regressors.

Fig 4

Using a Wilcoxon rank-sum test, the mean log target copies of N1 per liter were compared pairwise across all 14 of the sampled dormitories. A P value < 0.05 is displayed in light blue, and a P value < 0.001 is displayed in dark blue. Each of the dormitories is coded with a letter (A–N), and the pipe type is color coded as ductile iron (black), polyvinyl chloride (PVC, yellow), vitrified clay (light blue), and unknown (green). The bathroom types are displayed below each color-coded letter, as either a single-unit bathroom (one-person outline), a shared-unit bathroom (four-person outlines), or a hall communal bathroom (building outline).

Fig 5

Correlations between natural log target copies per liter for N1 and N2 gene targets for SARS-CoV-2 from dormitory wastewater samples and (A) precipitation (inches) across all grab (black) and composite (yellow) samples using a generalized linear model with Gaussian regressor. (B) The correlation between natural log target copies per liter ≥9 threshold and precipitation (inches) across all grab (black) and composite (yellow) samples using a generalized linear model with Gaussian regressor. (C) Natural log target copies per liter of composite samples by air temperature (°C) using a generalized linear model with Gaussian regressor. (D) Natural log target copies per liter of grab samples by air temperature (°C) using a generalized linear model with Gaussian regressor.