On October 23, 2026, The Guardian reported a new lawsuit asking a federal judge to order action on surface-water pollution from atrazine. The report, as summarized, says the suit alleges that the EPA has for decades failed under the Clean Water Act to address this pollution, and it describes atrazine as the "hormone-disrupting pesticide atrazine." The legal questions will move through filings and briefs over a period this article cannot predict, and nothing here assumes how that will end. The research question is narrower and applies regardless of outcome: what does a surface-water monitoring record actually establish?
Any claim that a pesticide is widespread, persistent, or rising in a river rests on sampling records. Those records carry design choices that shape the answer before a single lab result is read. Frequency, timing, detection limits, method, and site selection each limit what a reader can conclude. A reader who treats a sample set as a basin-wide picture is usually asking more of it than it was built to deliver.
Frequency and timing decide what gets seen
Surface-water pesticide concentrations often change quickly. In watersheds where a herbicide is applied in a defined season, runoff tends to cluster after application and rainfall, and the peaks can last hours to days. A grab sample taken on a fixed monthly schedule may miss the pulse entirely, or catch the tail of it. The result is a record that looks calm while the stream experienced short, high exposures.
Composite sampling averages concentrations over a set period, which smooths the record but dilutes peaks, so a composite can understate a short event. Event-triggered sampling captures storm pulses, but it costs more, depends on forecasting or rapid response, and produces data concentrated in wet periods. Neither approach is wrong. Each answers a different question, and a dataset built on one approach cannot be read as if it were built on the other.
Detection limits and non-detects change the average
A result below the method's reporting limit is usually recorded as a non-detect, not as zero. How those values enter an average or a trend calculation changes the output. Substituting zero, substituting half the limit, or excluding non-detects each produce different means, and in a dataset where most results sit near the limit, the choice can move the apparent trend more than the underlying water does. A careful reader checks which convention the source used and whether the reporting limit changed during the record.
Detection limits also change over time as laboratories revise methods or switch equipment. A rising trend in detections can reflect a lower limit rather than more pesticide in the water. The record should show the limit for each year, and a trend that crosses a method change needs an explicit adjustment or a caveat.
Method and site consistency make a trend
A trend requires that the thing being measured stayed the same. That means the same analyte, the same extraction and analysis method, the same sites, and comparable sampling seasons across years. Adding a station downstream of a new land use, moving an intake, or changing the sampling day within a month can all produce changes that look like environmental shifts. Before a record is cited as a trend, the reader should confirm that the sites are fixed and that the method was not revised mid-series.
Absence of detection also needs context. A site that rarely records a pesticide may be upstream of little application, or it may be sampled outside the periods when the pesticide moves. The record shows what was looked for, where, and when. It does not show that the pesticide was absent from the water at other times.
What to check before citing a surface-water trend
Start with the sampling design: how many samples were collected per year, on what schedule, and whether any were event-triggered. Then check the reporting limit for each year and how non-detects were handled. Confirm that the sites and the analytical method held steady across the period. Finally, ask whether the sampling window overlaps the application and runoff season for the chemical in question.
For this lawsuit, the summary available here does not describe the underlying monitoring data, sites, or methods. Anyone relying on the case for a technical claim should read the complaint and any attached data directly before drawing conclusions about frequency, trend, or exposure. Until that review is done, the defensible statement is limited: the suit contends the federal response has been inadequate, and the monitoring record behind any such contention depends on design choices that the reader has to verify.