Anticoagulant Rodenticides in California Wildlife: What the State’s 2022 Testing Found

What the testing found

• The California Department of Fish and Wildlife tested 158 wild animals for anticoagulant rodenticide residues in 2022 across three sampling programmes, and 128 had detectable residues, or 81.0 percent.
• Birds showed the highest exposure and the highest harm. 30 of 34 tested birds, or 88.2 percent, had detectable residues, and 17 of those 30, or 56.7 percent, were confirmed cases of anticoagulant rodenticide toxicosis.
• Large game mammals showed high exposure and no confirmed harm. 12 of 15 tested, or 80.0 percent, had residues, and none of the 12 exposures was a confirmed toxicosis case.
• Small game and non-game mammals showed high exposure and low confirmed harm. 86 of 109 tested, or 78.9 percent, had residues, with 1 confirmed toxicosis case and 3 classified as undetermined because decomposition prevented diagnosis.
• Multiple compounds in the same animal is common. 46.7 percent of tested large game mammals carried two or more anticoagulant rodenticides, and one sub-adult female black bear from El Dorado County carried five.
• Prohibited compounds still dominate the residues. Brodifacoum, bromadiolone and difethialone were among the most commonly detected analytes, and all three have been prohibited for most California uses since 1 January 2021 under AB 1788.
• Bromethalin, a non-anticoagulant rodenticide, was also widely detected. 22 of 95 tested small game and non-game mammals, or 23.2 percent, had bromethalin exposure, and of those exposures 22.7 percent resulted in mortality or suspected mortality.
• These are not population estimates. Every animal in this dataset died and was submitted for examination, which is a severely biased sample. The report states that its figures represent a minimum number of cases.

What this dataset is, and what it is not

In short: This is a state laboratory’s annual necropsy record, published under a formal agreement between California’s wildlife and pesticide agencies. It counts animals that died and were tested. It does not estimate what share of living wildlife is exposed.

The figures in this report come from a single primary document: the California Department of Fish and Wildlife’s 2022 annual report on pesticide exposures and mortalities in non-target wildlife, prepared by the department’s Wildlife Health Laboratory and published on 31 July 2023. Its authors are Jaime Rudd, Krysta Rogers and Nicholas Shirky, with contributions from Deana Clifford and Brandon Munk.

The report exists because of a memorandum of understanding between the California Department of Pesticide Regulation, the County Agricultural Commissioners and the Department of Fish and Wildlife. Its stated purpose is to ensure that pesticides registered in California are used in a way that protects non-target wildlife “while recognizing the need for responsible pest control”. Toxicology testing was performed under contract by the California Animal Health and Food Safety Laboratory in Davis.

The document is unusually careful about its own limits, and quoting that carefully is the point of this analysis. It states that its data “represent a minimum number of reports for tested animals that died within the reported calendar year and are subject to change as new information becomes available”.

Wildlife arrives at the laboratory through the public, through rehabilitation centres, through department staff and law enforcement, and through other agencies. Of the 264 small game and non-game submissions in 2022, wildlife rehabilitators accounted for 35 percent and the department itself for 33 percent. Animals are submitted because someone found them dead or dying, which means every percentage below is a percentage of animals already selected for having died.

Results by group

1. 81.0 percent pooled exposure. Across the report’s three testing programmes, 128 of 158 animals tested for anticoagulant rodenticides had detectable liver residues in California in 2022. (Author calculation summing the report’s three group totals.)
2. 88.2 percent of tested birds. 30 of 34 birds tested had detectable anticoagulant rodenticide residues. (CDFW Wildlife Health Laboratory, 2022 annual report, Table 2.)
3. 56.7 percent of exposed birds were confirmed toxicosis cases. 17 of the 30 exposed birds were confirmed cases, the highest confirmed harm rate of any group in the report. (Same source.)
4. 1,211 bird remains were submitted to the laboratory for necropsy in California in 2022, of which 34 were tested for anticoagulant rodenticide exposure. (Same source, Table 1.)
5. 80.0 percent of tested large game mammals. 12 of 15 large game mammals tested had detectable residues, from 68 large game submissions. (Same source, Table 5.)
6. Zero confirmed toxicosis in large game. None of the 12 large game anticoagulant rodenticide exposures resulted in a confirmed toxicosis case. (Same source.)
7. 46.7 percent carried two or more compounds. 7 of 15 tested large game mammals had exposure to two or more anticoagulant rodenticides, regardless of generation. (Same source.)
8. Five compounds in one animal. One sub-adult female black bear from El Dorado County, California had detectable exposure to five different anticoagulant rodenticides. (Same source.)
9. 78.9 percent of tested small game and non-game mammals. 86 of 109 tested had detectable residues, with 1 confirmed toxicosis case and 3 undetermined. (Same source, Tables 10 and 11.)
10. 3.5 percent toxicosis suspicion rate among exposed non-game mammals. Anticoagulant rodenticide toxicosis was suspected in 3 of the 86 exposed animals, and could not be ruled in or out because of advanced decomposition. (Same source.)
11. Bobcats were the most submitted mammal. Bobcats accounted for the largest share of mammal samples submitted to the laboratory in California in 2022. (Same source, Table 11.)
12. 23.2 percent bromethalin exposure in non-game mammals. 22 of 95 tested animals, sampled across 34 California counties, had bromethalin exposure. (Same source, Table 12.)
13. 22.7 percent of bromethalin exposures involved death. Of the 22 exposed animals, 2 died and 3 were suspected mortalities, totalling 22.7 percent. (Same source, Table 13.)
14. 26.7 percent bromethalin exposure in large game. 4 of 15 tested large game mammals had bromethalin exposure, including 3 of 14 black bears. (Same source, Tables 7 and 8.)
15. Coumachlor was never detected. Coumachlor, a first generation anticoagulant, was not detected in any tested large game or non-game mammal liver sample in California in 2022. (Same source.)

Exposure by group

In short: All three groups show exposure near or above 79 percent. Confirmed harm differs by a factor of roughly 45 between birds and non-game mammals.

Source: California Department of Fish and Wildlife, Wildlife Health Laboratory, 2022 Summary of Pesticide Exposures and Mortalities in Non-Target Wildlife, published 31 July 2023. Pooled row calculated by the author. The small game and non-game group submitted 264 herptiles and mammals, of which 150 were tested for pesticide exposure and results were available for 109 at the time of publication.

What it means. The exposure rates are remarkably consistent across three very different groups of animals, sitting between 78.9 and 88.2 percent. The confirmed harm rates are not consistent at all. Birds, and specifically raptors, were confirmed as poisoned in more than half of exposed cases. Large game mammals, overwhelmingly black bears, showed no confirmed poisoning despite 80 percent exposure.

The likely explanation is body mass and dose. A red-tailed hawk that eats several poisoned rodents receives a far larger dose relative to its body weight than a black bear that does the same. The report also notes that diagnosis requires more than a residue: it relies on postmortem evidence of haemorrhage alongside the liver residue, and in badly decomposed animals that evidence is destroyed.

What limits it, and this is the central caveat. These are small samples of self-selected dead animals. 34 birds is not a survey of California birds. The pooled 81.0 percent figure combines three different sampling frames with different submission pathways and should be treated as a summary of this laboratory’s caseload, not as a population exposure estimate. Anyone citing a single number from this report should cite the group figure rather than the pooled one, and should say that the animals were submitted after death.

Which compounds were found

In short: The compounds most often detected include three that California prohibited for most uses from the start of 2021. The 2022 testing year is the second year of that prohibition.

The report divides anticoagulant rodenticides into two classes. First generation compounds include warfarin, coumachlor, diphacinone and chlorophacinone. Second generation compounds, which the report describes as “more toxic”, include brodifacoum, bromadiolone, difethialone and difenacoum.

Source: CDFW Wildlife Health Laboratory 2022 annual report, narrative accompanying Figures 2, 4 and 6.

What it means, stated carefully. Brodifacoum, bromadiolone and difethialone are second generation anticoagulants. California’s Assembly Bill 1788, the California Ecosystems Protection Act, was signed on 29 September 2020 and prohibited the use of brodifacoum, bromadiolone, difenacoum and difethialone in California from 1 January 2021, subject to exemptions. The animals in this report died during 2022, the second calendar year of that prohibition, and those compounds were still among the most frequently detected residues.

What that does and does not establish. It does not establish illegal use. At least four explanations are consistent with the finding and the data here cannot separate them. AB 1788 carried exemptions, including for agricultural use and use under special permit, so lawful application continued. Second generation anticoagulants persist in liver tissue for months, so a 2022 death can reflect an exposure well before it. Product already in circulation before 2021 may have continued to be applied. And unlawful use is also possible. Attributing the residues to any one of these would go beyond the evidence.

What limits this section. The report identifies most commonly detected analytes in narrative text and in figures rather than in a compound-by-compound count table, so this analysis reports the ranking the source gives and does not assign numbers to individual compounds. The bird row above is deliberately vague on the not-detected column because the source sentence was truncated in our text extraction and we did not want to guess at the compound named.

The non-anticoagulant picture

In short: The debate concentrates on anticoagulants, but the laboratory also tested for bromethalin, organophosphates, carbamates, neonicotinoids, pyrethroids and fipronil. Bromethalin was found in roughly a quarter of tested non-game mammals.

103 wild non-game and small game mammals were tested for pesticides beyond anticoagulants in California in 2022, covering bromethalin, organophosphates and carbamates, neonicotinoids, pyrethroids, fipronil and fipronil sulfone.

Bromethalin is a neurotoxic rodenticide and is not an anticoagulant. Adipose tissue or brain from 95 animals across 34 California counties was tested for it. 22 animals, or 23.2 percent, had detectable exposure. Of those 22 exposures, 2 resulted in mortality and 3 in suspected mortality, which the report gives as 22.7 percent of exposures.

Two other results are worth recording because negative findings rarely get reported. A general toxicology panel on two raccoons from Sonoma and Tehama counties detected no toxic compounds. Twelve North American river otters were tested for neonicotinoids, pyrethroids, fipronil, fipronil sulfone and organophosphates, and among the five for which results were available at publication, none of the compounds was detected.

What it means. Bromethalin is frequently proposed as the alternative to anticoagulants in jurisdictions that restrict them. On this evidence it also reaches non-target wildlife, and its confirmed or suspected mortality share among exposed animals, at 22.7 percent, is higher than the confirmed toxicosis share among anticoagulant-exposed non-game mammals in the same report. That comparison deserves caution, because the two figures use different diagnostic standards, but it is a signal that substitution is not the same as solution.

What limits it. 95 animals across 34 counties is a thin sample per county. The river otter results were incomplete at publication, with results for 5 of 12. Bromethalin residues are measured in adipose or brain rather than liver, so the sampling is not directly comparable to the anticoagulant testing.

Three measures we calculated

1. The Exposure to Harm Ratio

Definition. Confirmed toxicosis cases divided by animals with detectable residues, within a sampling group, expressing how often detected exposure corresponded to diagnosed poisoning.

Formula. Exposure to Harm Ratio = confirmed toxicosis cases / animals with detectable residues.

Inputs. CDFW 2022 annual report, group totals from Tables 2, 5, 10 and 11.

Result. Birds 17/30 = 56.7 percent. Small game and non-game mammals 1/86 = 1.2 percent. Large game mammals 0/12 = 0 percent.

Why it matters. Public discussion of this topic almost always cites the exposure figure and almost never the ratio. The ratio is what distinguishes a residue finding from a mortality finding, and it varies by a factor of roughly 45 between birds and non-game mammals in the same report, in the same state, in the same year.

Limitations. The denominator and numerator come from different diagnostic processes: residue detection is a laboratory result, toxicosis is a pathologist’s judgement requiring intact tissue. Decomposition suppresses the numerator without affecting the denominator, so every ratio here is a lower bound. The report explicitly classifies three non-game cases as undetermined for this reason. With 12 and 30 animals in two of the three groups, a single case changes the ratio by several percentage points.

2. Post-Prohibition Residue Persistence

Definition. Whether compounds prohibited by a state restriction continue to appear among the most frequently detected residues in wildlife after the restriction takes effect, and for how long.

Method. Compare the compounds named as most commonly detected in a monitoring year against the compounds prohibited by statute, and record the interval between the prohibition’s effective date and the monitoring year.

Inputs. CDFW 2022 annual report analyte rankings; California AB 1788, effective 1 January 2021.

Result. Three of the four compounds prohibited by AB 1788, namely brodifacoum, bromadiolone and difethialone, were still among the most commonly detected analytes two years after the prohibition took effect.

Interpretation. A residue-based monitoring programme cannot be expected to show an immediate response to a use restriction. Liver persistence, statutory exemptions, existing stock and the lag between exposure and death all delay any signal. Operators and policymakers reading annual wildlife reports for evidence that a restriction worked should expect a multi-year lag and should not read the first two years as failure.

Limitations. This is a single monitoring year compared against a single statute. Establishing an actual trend requires the equivalent reports for 2021, 2023, 2024 and 2025, which we did not analyse here. The report does not state exposure dates, only death dates.

3. Testing Coverage Rate

Definition. The share of animals submitted for necropsy that were actually tested for pesticide residues, which bounds how much any exposure rate can tell you.

Formula. Testing Coverage Rate = animals tested for anticoagulant residues / animals submitted for necropsy in the same group.

Inputs. CDFW 2022 annual report, Tables 1, 4 and 9 for submissions.

Result. Birds 34/1,211 = 2.8 percent. Large game mammals 15/68 = 22.1 percent. Small game and non-game mammals 109/264 = 41.3 percent.

Why it matters. The 88.2 percent bird exposure rate rests on 2.8 percent of submitted birds. Animals are selected for toxicology testing when something about the case suggests poisoning, which means the tested subset is enriched for exposure by design. This single ratio is the strongest argument against treating any figure in this report as a population estimate, and it is not reported anywhere else that we could find.

Limitations. The report does not publish its criteria for selecting which submitted animals go to toxicology, so the degree of enrichment cannot be quantified. Some submissions arrive too decomposed to test, which lowers coverage for reasons unrelated to case selection.

Related: how pest control’s rodenticide use changed after the 2021 restriction is measured in California’s rodenticide ban, measured.

Related research: five years of wildlife exposure against structural use; US poison center rodenticide exposures, 2014 to 2024.

Download the data

The tables in this report are published as CSV files under a Creative Commons Attribution 4.0 licence. You may republish, remix and build on them, including commercially, with attribution to The Pest Control Brief and a link to this page.

• California wildlife rodenticide testing by group, 2022 (CSV, 4 rows)

How to cite this report

The Pest Control Brief, “Anticoagulant Rodenticides in California Wildlife: What the State's 2022 Testing Found”, 2 October 2026. https://thepestcontrolbrief.com/rodenticide-wildlife-exposure-data/

References

1. California Department of Fish and Wildlife, Wildlife Health Laboratory. “2022 Summary of Pesticide Exposures and Mortalities in Non-Target Wildlife”, 2022 annual report, published 31 July 2023. Authors Jaime Rudd, Krysta Rogers and Nicholas Shirky, with contributions from Deana Clifford and Brandon Munk. nrm.dfg.ca.gov document 241290 (PDF). Accessed 25 September 2026. Every exposure, test, submission and toxicosis count in this report is read from this document.
2. California Assembly Bill 1788 (2019-2020), the California Ecosystems Protection Act. leginfo.legislature.ca.gov, AB 1788. Accessed 25 September 2026. Signed 29 September 2020, effective 1 January 2021.
3. California Department of Pesticide Regulation. “New Second Generation Anticoagulant Rodenticide Prohibitions and Allowed Uses: Questions and Answers”. cdpr.ca.gov. Accessed 25 September 2026.
4. California Assembly Bill 2552 (2023-2024), the Poison-Free Wildlife Act. Signed 25 September 2024, effective 1 January 2025. Added chlorophacinone and warfarin to the state prohibition, with public health and agricultural exemptions.
5. California Animal Health and Food Safety Laboratory, Davis. Performed the toxicology testing reported in the CDFW document under contract.

How the figures were read

Source. One primary document: California Department of Fish and Wildlife, Wildlife Health Laboratory, “2022 Summary of Pesticide Exposures and Mortalities in Non-Target Wildlife”, 2022 annual report, 24 pages, published 31 July 2023, retrieved from the department’s document server on 25 September 2026. Regulatory dates come from California Assembly Bill 1788 (2019 to 2020 session) and Assembly Bill 2552 (2023 to 2024 session).

How figures were taken. All exposure counts, test counts, submission counts and toxicosis counts are read directly from the report’s narrative text and numbered tables. Where the narrative and a table express the same quantity, both were checked against each other before use.

Denominator discipline. The report expresses some percentages against animals tested and others against animals exposed. Two bromethalin figures illustrate the trap: 23.2 percent is 22 exposed animals out of 95 tested, while 22.7 percent is 5 mortalities out of 22 exposed. Both appear in the source and they are not interchangeable. Every percentage in this report names its denominator.

What we derived. Three quantities are ours and are labelled as such: the pooled 81.0 percent exposure rate across all three groups, the Exposure to Harm Ratio, and the Testing Coverage Rate. Each is a simple arithmetic combination of the report’s own published totals.

A discrepancy we did not resolve. In our text extraction of the large game bromethalin tables, the county detail and species detail rows appear to show one confirmed toxicosis case while the total rows for the same tables show zero. This may be a column alignment artifact of extracting text from a PDF rather than an inconsistency in the source. We therefore report large game bromethalin exposure, 4 of 15, and make no claim about large game bromethalin toxicosis. Readers needing that figure should consult the original document.

Limitations. Every animal in this dataset died and was submitted for examination, so the sample is selected on the outcome. Animals are chosen for toxicology testing when poisoning is suspected, which enriches the tested subset. Sample sizes are small, at 34 birds, 15 large game mammals and 109 other mammals. The findings are California only and do not transfer to other states, whose rodenticide rules differ. The report is a single year and this analysis does not establish a trend. Residue persistence means a detection does not date the exposure. The report itself states its figures are a minimum. Exposure is not poisoning, and this analysis does not claim that 81.0 percent of anything was harmed.

What we deliberately did not do. We did not estimate wildlife population exposure. We did not compare exposure rates against rodenticide sales or application volumes, because the report contains no such data. We did not take figures from advocacy organisations or from press coverage of this report, including widely circulated exposure percentages attributed to other years and other studies, because we could not verify them against a primary document in this analysis.

Last reviewed: 25 September 2026.

Primary source

Tier 1, primary government data. California Department of Fish and Wildlife, Wildlife Health Laboratory, 2022 Summary of Pesticide Exposures and Mortalities in Non-Target Wildlife, published 31 July 2023. Toxicology performed by the California Animal Health and Food Safety Laboratory, Davis. California Assembly Bill 1788 and Assembly Bill 2552, for prohibition scope and effective dates. California Department of Pesticide Regulation guidance on second generation anticoagulant rodenticide prohibitions and allowed uses.

Tier 2, credible secondary. None of the figures in this report derives from a secondary source.

Tier 3, journalism. Not used.

Excluded. Advocacy campaign pages were excluded as sources for any figure. Frequently quoted exposure percentages from other California studies, including figures for mountain lions and bobcats circulating at between 85 and 100 percent, were excluded because we could not verify them against the underlying primary studies within this analysis. Their exclusion is not a judgement on their accuracy.

Guidance for reporters

Most quotable statistics

1. 128 of 158 wild animals tested by California’s state wildlife laboratory in 2022 had detectable anticoagulant rodenticide residues, or 81.0 percent.
2. Among exposed birds, 56.7 percent were confirmed poisoning cases. Among 12 exposed large game mammals, none were.
3. One sub-adult female black bear in El Dorado County carried five different anticoagulant rodenticides.
4. Three compounds prohibited in California since January 2021 were still among the most commonly detected residues in 2022.
5. Only 34 of 1,211 birds submitted for necropsy were tested for rodenticide residues, which is 2.8 percent.

Data limitations to quote alongside them

Every animal was submitted after death, so this is not a population sample. Animals are selected for testing when poisoning is suspected. Sample sizes are 34 birds, 15 large game mammals and 109 other mammals. The report is California only, covers one year, and states its own figures are a minimum. Detectable residue is not the same as poisoning.

Fields in the published CSV files

Year; taxonomic group; county; species; number submitted; number tested; number exposed; exposure rate; confirmed toxicosis; undetermined toxicosis; compounds detected; compound generation; tissue tested.

Press summary

California’s state wildlife laboratory tested 158 wild animals for anticoagulant rodenticide residues in 2022 and found detectable residues in 128 of them, or 81.0 percent, according to an analysis of the department’s annual report by The Pest Control Brief. Exposure rates were similar across birds at 88.2 percent, large game mammals at 80.0 percent and other mammals at 78.9 percent, but confirmed poisoning differed sharply: 56.7 percent of exposed birds were confirmed toxicosis cases, against none of 12 exposed large game mammals. One black bear in El Dorado County carried five different anticoagulant compounds. Three compounds prohibited for most California uses since January 2021 were still among the most commonly detected residues in the second year of that prohibition, which the analysis attributes to a combination of statutory exemptions, tissue persistence and exposure lag rather than to any single cause. Only 2.8 percent of submitted birds were tested, so the figures are not population estimates.

Five headlines editors could use

1. Four in five wild animals tested by California’s wildlife lab carried rat poison
2. One black bear, five different rodenticides
3. Exposure is not poisoning: what California’s wildlife testing actually shows
4. Banned rat poisons still topped the detection list two years after the ban
5. The denominator problem in California’s rodenticide data

What readers ask about this data

What share of California wildlife is exposed to rat poison?

This dataset cannot answer that. It reports that 128 of 158 animals tested after death had detectable residues, which is 81.0 percent of a sample selected for having died and then selected again for suspected poisoning. A population exposure estimate would require testing living animals at random.

Does exposure mean the animal was poisoned?

No, and the report is explicit about the distinction. Exposure means detectable residue in the liver at any concentration. Toxicosis requires postmortem evidence of harm alongside the residue. Among exposed birds 56.7 percent were confirmed toxicosis cases, among exposed non-game mammals 1.2 percent were, and among exposed large game mammals none were.

Why are birds so much more likely to be poisoned than bears?

The most likely reason is dose relative to body mass. A raptor consuming poisoned rodents ingests a far larger dose per kilogram of body weight than a black bear doing the same.

Are second generation anticoagulants legal in California?

Assembly Bill 1788 prohibited use of brodifacoum, bromadiolone, difenacoum and difethialone in California from 1 January 2021, subject to exemptions including agricultural use and use under special permit. Assembly Bill 2552, signed 25 September 2024, added chlorophacinone and warfarin from 1 January 2025, with exemptions for public health and agriculture. Operators should confirm current requirements with the Department of Pesticide Regulation, because the exemptions and their conditions are the part that changes.

If the compounds are prohibited, why are they still being detected?

At least four explanations fit the data and it cannot distinguish between them: lawful use under the statutory exemptions, persistence of these compounds in liver tissue for months after exposure, application of product already in circulation before the prohibition, and unlawful use. Attributing the detections to any single cause would go beyond the evidence.

Is bromethalin a safer alternative?

It is not an anticoagulant, and it also reached non-target wildlife in this dataset. 22 of 95 tested non-game mammals had bromethalin exposure, and of those exposures 22.7 percent involved confirmed or suspected mortality. That comparison should be read with care because the diagnostic standards differ between the two compound classes.

Which animals were most often submitted?

Bobcats accounted for the largest share of mammal samples submitted to the laboratory in 2022. Across the small game and non-game group, wildlife rehabilitators submitted 35 percent of the 264 cases and the department itself 33 percent.

Does this apply outside California?

No. California has the most restrictive anticoagulant rodenticide rules in the United States and one of the few state wildlife laboratories publishing an annual pesticide exposure report. Other states’ rules differ, and most publish nothing comparable.

What does this mean for a pest control operator?

Practically, that the evidence base driving rodenticide restriction is residue monitoring of dead wildlife, that it shows high exposure and variable harm, and that the regulatory direction in California has been to extend restrictions rather than relax them. Our coverage of restricted use pesticides and what chemicals exterminators use sets out the product context, and rodent removal costs covers the treatment side.

Where can I read the original report?

It is published by the California Department of Fish and Wildlife Wildlife Health Laboratory as its 2022 annual report on pesticide exposures and mortalities in non-target wildlife, dated 31 July 2023. We recommend reading the tables directly rather than relying on any secondary summary, including this one.