The probability of detecting host-specific microbial source tracking markers in surface waters was strongly associated with method and season.

Fecal contamination of surface waters presents significant human health and environmental risks. While many methods for understanding fecal contamination cannot differentiate between human and animal sources, microbial source tracking (MST) marker methods are used to identify fecal sources. To understand how to best employ MST marker data for managing fecal contamination risks, the present study compiled previously collected MST marker data ( = 12,878 samples) from across North America; each sample in the data set had been tested for fecal contamination from one or more of five sources (avian, canine, human, ruminant, swine). Using these data, this study aimed to characterize associations between non-methodological and methodological factors and detection of host-specific MST markers and determine how methodological differences may complicate the interpretation of these associations between studies. Variance partitioning analysis was used to quantify the variance in host-specific MST marker detection attributable to non-methodological and methodological factors. Conditional forest and regression analysis were utilized to assess the association between detection and select non-methodological and methodological factors. Between 10% (canine) and 38% (human) of total variance was uniquely attributable to non-methodological factors for any of the fecal sources considered, while between 50% (human) and 84% (canine) of variance could not be attributed to either methodological or non-methodological factors. This highlights the need for standardization of methods for MST marker detection across studies but also suggests that other factors, beyond those non-methodological variables considered here, influenced variation in the likelihood of MST marker detection.IMPORTANCEThis study underscores complications associated with comparing findings from studies that used different methodologies to detect the same fecal targets and highlights the difficulties associated with using non-comparable data to generalize findings and develop science-based risk management plans. These findings highlight the need for standardization of sampling and laboratory methods across microbial source tracking marker studies. Our findings build on previous research to suggest that one-size-fits-all approaches to managing fecal hazards in surface waterways may not be appropriate; instead, strategies tailored to specific water sources and conditions at the time of water use may be more effective.