Detecting a Pathogen in Soil Is Not the Same as Predicting Disease, Study Finds

For potato growers, the sight of corky, pitted lesions on freshly harvested tubers is a familiar and costly disappointment. Common scab, a soil-borne disease caused by filamentous bacteria in the genus Streptomyces, does not usually kill the plant, but it can render tubers unmarketable. In the United States, potatoes with more than ten percent of their surface covered in scab lesions can no longer be used for processed products, translating into significant economic losses each season. Because the damage is only visible after harvest, producers have long sought a way to assess risk before planting, when management decisions could still make a difference. A new study published in Plant and Soil, however, delivers a sobering message about how difficult that goal may be to achieve, even with the most modern molecular tools.

Researchers at North Dakota State University set out to test a widely held assumption in plant pathology: that detecting the genetic activity of a pathogen in soil should be a better predictor of disease than simply detecting its DNA. DNA-based assays, which have been the standard approach for identifying Streptomyces scabies, the most prevalent pathogenic species in the United States, suffer from a fundamental limitation. They cannot distinguish between DNA from living, active cells and residual extracellular DNA left behind by dead organisms. Because extracellular DNA can persist in soil for months or even years under favorable conditions, DNA-based tests risk overestimating the presence of an active pathogen and inflating estimates of disease risk.

RNA, by contrast, is synthesized only by metabolically active cells and degrades rapidly after cell death, which is why many researchers have considered RNA-based detection a potentially more faithful indicator of pathogen activity. The research team, led by Rachel Clarkson and colleagues, designed a controlled growth chamber experiment to determine whether RNA-based assays could predict the severity of common scab before tubers form. They artificially infested a fine sandy loam soil collected from a potato-growing region near Glyndon, Minnesota, with a gradient of S. scabies densities, then planted radish, a well-established model host for the disease. Soil samples were collected both before planting and after harvest, and each sample was analyzed for both DNA and RNA using quantitative PCR and reverse transcriptase quantitative PCR.

The technical design of the study was careful and deliberate. The team targeted two genetic markers. The first was the txtAB operon, which encodes the non-ribosomal peptide synthetases responsible for producing thaxtomin A, the primary virulence determinant of pathogenic Streptomyces. Thaxtomin A inhibits cellulose biosynthesis in plant cells, allowing the bacterium to colonize developing tubers. The second marker was the 16S rRNA gene, a constitutively expressed housekeeping gene specific to S. scabies that serves as a proxy for the organism’s general metabolic activity. The logic was straightforward: if RNA-based detection of either marker correlated with disease severity, growers would have a reliable pre-season diagnostic tool.

The results confirmed that the researchers had successfully established a gradient of pathogen density across their experimental mesocosms. Transcribed 16S rRNA and genomic txtAB abundances both increased significantly and linearly with the volume of inoculum applied to the soil, with the genomic txtAB assay showing a particularly strong relationship. In other words, the molecular assays could clearly detect that some soils contained far more S. scabies than others, both before planting and, with some caveats, after harvest. The gradient was real, measurable, and consistent across the main detection targets.

Yet when the researchers compared these gene abundance measurements against the actual severity of scab lesions on the radish tubers, the predictive power collapsed. Disease severity, which ranged from zero to twenty-five percent of tuber surface area across the experiment, showed no significant relationship with any of the gene abundance measurements, regardless of whether the measurements were taken before planting or after harvest, and regardless of whether they were based on DNA or RNA. Even the transcribed 16S rRNA signal, which should have reflected the presence of metabolically active bacteria, failed to predict how badly the tubers would be affected. Soil pH and moisture content, analyzed as covariates, did not explain the discrepancy either.

One of the most instructive findings concerned the txtAB virulence gene itself. Transcribed txtAB was never detected in any sample, even at the end of the experiment when tubers were present. This makes biological sense: expression of the txtAB operon is facultative, triggered by cellobiose and cellotriose, cellulose subunits released only during active tuber growth. Before planting, no tubers existed, so no inducer molecules were available and no virulence gene transcription was expected. But even after five weeks of tuber development, the virulence transcripts remained below the detection threshold of quantitative PCR. In a post-hoc test, the team found that droplet digital PCR, a far more sensitive technique that partitions samples into thousands of microscopic droplets, could successfully amplify transcribed txtAB from post-harvest samples, suggesting the transcripts were present but simply too scarce for conventional qPCR to detect.

The copy number problem helps explain this failure. While the 16S rRNA gene exists in ten to twenty copies per S. scabies genome, txtAB is present in up to only three copies. In this experiment, genomic 16S rRNA abundance exceeded genomic txtAB abundance by more than a million-fold at the initial sampling. For genes present in such low quantities, standard qPCR may simply lack the sensitivity needed for reliable detection in complex soil matrices. The authors suggest that future work may need to explore alternative pathogenicity markers with higher copy numbers or longer cellular residence times, or to adopt digital PCR platforms as a standard tool for soil-based diagnostics of low-abundance virulence transcripts.

Beyond the technical limitations, the study reveals a deeper ecological truth: the presence of a pathogen, even in a metabolically active state, does not guarantee that it will express virulence. S. scabies, like many plant pathogens, only becomes dangerous when specific host-derived signals are present, and the timing of those signals relative to sampling matters enormously. If a soil sample is collected at the end of the growing season, the pathogen’s response to inducer molecules may already have attenuated as host signal production declined. The researchers also observed that the pathogen density gradient measured after harvest did not always match the gradient imposed at the start, underscoring how dynamic soil pathogen populations are over time and how difficult it is to capture a single snapshot that meaningfully predicts future disease.

The implications for agriculture are significant. Molecular assays are increasingly promoted as tools for assessing soil-borne pathogen risk, and this study suggests that gene abundance measurements alone, whether DNA-based or RNA-based, may be insufficient for predicting common scab severity. What is needed instead is a more complete understanding of soil-borne pathogen ecology: how soil physical, chemical, and biological properties interact with pathogen-host signaling to determine when a present and active pathogen actually becomes a disease problem. Until that ecological context is built into diagnostic frameworks, growers will continue to lack reliable pre-season tools for common scab management, and the gap between detecting a pathogen and predicting its impact will remain wide open.

Subject of Research: RNA- and DNA-based molecular detection of the soil-borne potato pathogen Streptomyces scabies and its ability to predict common scab disease severity

Article Title: Presence ≠ problem: pathogenicity decoupled from active pathogen abundance

Article References: Clarkson, R., Zitnick-Anderson, K., Vanderhyde, M., Longtin, S. M., & Butcher, K. (2026). Presence ≠ problem: pathogenicity decoupled from active pathogen abundance. Plant and Soil. https://doi.org/10.1007/s11104-026-09150-x

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09150-x

Keywords: Streptomyces scabies, common scab, potato, soil-borne pathogens, RNA-based detection, qPCR, rt-qPCR, txtAB, 16S rRNA, thaxtomin A, plant pathology, soil diagnostics