WDNR errors in wolf science
Wisconsin wolf science debate
The following page explains why the state of Wisconsin's claims about the wolf population size, wolf mortality rates, and birth rates are not scientifically sound, replicable, or fundamentally credible. Specific criticisms of their most recent attempts are presented below in published, peer-reviewed rebuttals to Roberts et al. and Stauffer et al.
To summarize the big picture, tThe public should not trust those claims because the state and its allied authors ignore the principles of Open Science by not sharing data or methods, selective citation of the research they prefer no matter how weak, and by viewpoint discrimination against other research that challenges their views. Since 2013, we have been publishing our findings in stronger journals, using more rigorous independent review and stronger inference, sharing the data, and testifying to authorities in various ways. The Wisconsin DNR remains stubbornly wed to its poor science. Below we present our rebuttals to the two most recent efforts from Glenn Stauffer and Nathan Roberts. We summarize the latest errors in Wisconsin DNT wolf science then we provide all previous rebuttals of state wolf science that we have published in peer-reviewed articles.
Treves, A., Agan, SW, Langenberg, JA, Lopez-Bao, J.V., Louchouarn, NX. Parsons, DR, Rabenhorst, MF, Santiago-Ávila, FJ, 2024. Response to Roberts, Stenglein, Wydeven, and others. Journal of Mammalogy 2024 in press.
. Reply to Stauffer et al.: Uncertainty and precaution in hunting wolves twice in a year: in review at PLoS One available here.
In summary, our rebuttals to Roberts et al. and Stauffer et al, showed how
a. Stauffer et al.’s 2021 model of wolf occupancy in WI relies on data from previous years of winter tracking even if lethal management (such as a wolf-hunt) happened. Therefore, they are likely to be counting dead wolves as alive. That is why no one should trust the state estimate of the wolf population. If Stauffer and Roberts shared data as required by publication ethics, they might be able to support their estimates and their claims. But they have refused to share data again and again since 2012. And even if they did share their Datta we doubt it would show what they claim, so we suspect the state wolf population was lower than what they estimated in all previous years.
b. State estimates of birth rates are based on flawed science that hasn't passed peer review. The methods they use are inaccurate, even A. Wydeven admitted it in 2004, and they presume theirs are superior to mark-recapture methods published by Dick Thiel.
c. State estimates of wolf mortality are large, systematic under-estimates. Roberts and Stauffer both perpetuate a myth that radio-collared wolves who disappear just went off the air because of transmitter or battery failures despite the rate of disappearance being two- to three-fold higher than rates of disappearance for other (non-wolf) animal telemetry data.
d. Neither Stauffer nor Roberts share data so their claims are not credible by current scientific standards of the Open science movement. Neither they nor their co-authors transparently disclose potentially competing interests -- both financial and non-financial. These are breaches of publication ethics and research integrity according to the National Academies of science. We’re working with editors of multiple journals to flag their articles for concern or correction. The evidence for our claims is here. I will remove the disclosures they should have made once resolved. For now they are posted here .
The above evidence of undisclosed affiliations, and interests, both financial and non-financial. These are breaches of scientific integrity long described by the National Academies of Sciences, Engineering, and Medicine.
Note Stauffer et al. and Roberts et al., claim we are trying to silence them when we are obviously trying the opposite, to compel transparency about their potentially competing interests, see this article on competing interests.
See this editorial Treves, A. 2024. Authors declare no competing interests—really? Frontiers in Ecology and the Environment (guest editorial) to understand why disclosures of potentially competing interests are important and why current practices in scientific peer review need an overhaul.
They also routinely engage in selective citation which means they do not cite the work that contradicts their preferred findings and whenever they are compelled to cite contrary findings, they mention rebuttals to it. This creates the illusion that their science is ironclad while critics have been challenged. Such cherry-picking or selective citation also violates National Academies guidelines on research integrity. Ti also further undermines claims that Wisconsin DNR policy is informed by the best available science, let alone 'science-based'.
I explain why scientific integrity is so important to public confidence in our research community and good government at this page.
The DNR and NRB documents we cited from 2021.
The Supplementary Information from Treves & Louchouarn 2022.
Simulating study designs paper as a compressed file of scripts and data.
Supplementary Data published for J Mammalogy paper on WI wolf mortality: Supporting Information SD1-SD5. Please note that on 11 June 2017, one file was updated to correct a typo discovered by FSA.
For Wisconsin Department of Natural Resources, Endangered Resources Bureau Population Reports before 2016 accompanying the article on gray wolf mortality in Wisconsin. Note: not all years were available for download before the WDNR archived or removed the reports. Please email Adrian Treves if you have unedited reports from the missing years, which I can add to this public repository.
Population Reports 2016-2021. Accessed 1 April 2021.
Between 2004 and today, Dr. Adrian Treves and scientists in the Carnivore Coexistence Lab have been independently reviewing and questioning wolf science presented by the Wisconsin DNR. We have found it wanting over and over again in transparency and scientific rigor.
Wisconsin DNR continues to claim that killing wolves protects livestock despite reams of evidence from inside the state and outside the state that the claim is invalid. Here again the state refuses to share all data transparently or conduct the appropriate robust analyses to support their claim.
Treves, A., L. M. Elbroch, J. T. Bruskotter, 2024.
Evaluating fact claims accompanying policies to liberalize the killing of wolves, peer-reviewed chapter TBA, In press. Alpha Wildlife Publications, Canada.
We also presented alternative explanations for why killing predators can lead to increases in domestic animal losses.
Elbroch L & Treves A. (equal co-authors) 2023.
Why might removing carnivores maintain or increase risks for domestic animals? Biological Conservation 283:110106.
The method for estimating Wisconsin wolf abundance as advocated by Stauffer et al. 2021 -- called the scaled occupancy model -- is not robust. We showed how it was imprecise, inaccurate, insensitive to changing conditions, and irreproducible because of subjective inclusion and exclusion criteria. The state estimate of wolf abundance should not be trusted because the method counts dead wolves as fi they were alive and wad validated for years without legal wolf-killing then applied to years with current and prior legal wolf-killing. We explain fully below.
Treves, A., Santiago-Ávila, F.J. 2023Estimating wolf abundance with unverified methods. Academia Biology 1 doi 10.20935/AcadBiol6099 Compressed source documents
from the state (WDNR 2022 population reports, greensheet and Stauffer et al. 2021).
Evidence accumulates that liberalizing wolf-killing does not improve coexistence. Evidence from Michigan wolves (Louchouarn), red wolves (Again et al.; Santiago-Ávila et al.), Mexican wolves (Louchouarn et al. 2021 and 2025), and Wisconsin wolves (Chapron & Treves 2016; Santiago-Ávila & Treves 2022; Treves et al. 2013, 2017, Treves & Bruskotter 2014; Browne-Nuñez et al. 2015; Hogberg et al. 2015). In the face of all these independent lead authors, independent datasets, independent lines of evidence (human attitudes, population dynamics, individual wolf survival), the Wisconsin DNR has stubbornly refused to support its own claims or acknowledge our findings.
Santiago-Ávila FJ, Agan S, W.,, Hinton JW, Treves A. 2022.
Evaluating how management policies affect red wolf mortality and disappearance. Royal Society Open Science 9:210400. https://doi.org/10.1098/rsos.210400.
Treves, A., Louchouarn, N.X. 2022.
Uncertainty and precaution in hunting wolves twice in a year. PLoS One, 2022. 17(3): e0259604.
A third-party raised a concern about a typographical error in one of our sources (Thiel et al. 2009), so for due diligence we re-analyzed our data and published a comment here
However our conclusions did not change.
The Supplementary Information from Treves & Louchouarn 2022.
And see the WDNR documents since the disastrous 2021 February wolf-hunt. The DNR and NRB documents we cited from 2021.
Don’t judge the roar by its echo: Tests of assumptions, tools and policies for human-carnivore coexistence in North America. PhD dissertation, August 2023, Carnivore Coexistence Lab, Nelson Institute for Environmental Studies, University of Wisconsin-Madison.
We reviewed evidence that most participatory stakeholder processes in conservation are unjust, inconsistent with the US public trust doctrine, and deliver narrow interest group victories by design. This applies to Wisconsin DNR stakeholder and advisory groups, even the Natural resource Board itself.
Santiago-Ávila, F.J. & Treves, A. 2022.Poaching of protected wolves fluctuated seasonally and with non-wolf hunting. Scientific Reports 12:e1738. https://doi.org/10.1038/s41598-022-05679-w.
After the disastrous wolf-hunt of Feb2021 in Wisconsin, we published a peer-reviewed estimate of the wolf population size. This is still the only peer-reviewed estimate in existence although the WDNR claims Stauffer et al. 2021 which is a method offers the state estimate and is peer-reviewed. We showed how unscientific that method is (see above for our 2023 paper in Academia Biology).
Treves, A., F.J. Santiago-Ávila, and K. Putrevu 2021.
Quantifying the effects of delisting wolves after the first state began lethal management. PeerJ, 9:e11666, doc 10.7717/peerj.11666.
We went back through the history of Wisconsin wolf population size estimates. We found an error all the way back in 1995 that set in motion certain inaccuracies and invalid assumptions. These were exacerbated between 2000 and 2004 when the WDNR refused to acknowledge that volunteer trackers counted different numbers of wolves in census blocks than did government biologists counting in the same census blocks. They refused to study which count was more accurate (either could be) or account scientifically for the variability and uncertainty this produced for the statewide estimate. We also showed how Wisconsin wolf policy was founded on an invalid assumption of density-dependent growth, a wild of state carrying capacity, and value judgments about maximum sustainable yield. The below peer-reviewed, publication demonstrates how Wisconsin wolf policy is riddled with concealed value judgments that hunting wolves is appropriate, hidden by a misleading veil of scientific modeling.
Treves A, Paquet PC, Artelle KA, Cornman AM, Krofel M, Darimont CT. 2021.
Transparency about values and assertions of fact in natural resource management. Frontiers in Conservation Science: Human-Wildlife Dynamics, 2:e631998, doc 10.3389/fcosc.2021.631998.
Lopez-Bao, J.V., Chapron, G., Treves, A. 2017.
The Achilles heel of participatory conservation. Biological Conservation 212: 139-143.
Soon after correcting the record on Wisconsin wolves, we published a more general article showing the fairly simple algebra that explains why all US wolf populations have been miscalculating mortality rates, under-estimating poaching rates, and misidentifying the worst cause of mortality (poaching). The below article has not been rebutted. It shows that wolves in the Northern Rockies, Mexican wolf recovery area, red wolves in North Carolina, and Wisconsin wolves have lower risk of legal mortality and higher risk of poaching mortality than governments have acknowledged. We also showed qualitatively that the major cause of Alaskan wolf mortality in Adams et al. 2008 was unregulated, unreported killing.
Treves, A., Artelle, K.A., Darimont, C.T., Parsons, D.R. 2017.
Mismeasured mortality: correcting estimates of wolf poaching in the United States. Journal of Mammalogy 98(3): open access at DOI: https://doi.org/10.1093/jmammal/gyx052.
We aired state wolf mortality data so anyone could re-analyze it or critique our methods in the article below that took 6 years to finish and relied on multiple independent data verifiers to make sure we ha don't made errors. We showed that poaching was under-estimated by the state, that radio-collared wolves were dying or disappearing at unusually high rates and earlier than nonhuman ('natural') causes. We explained why disappearances of radio-collars could not ALL be explained by mechanical issues or migration out of state. The state attempted to suppress and retract this study but the journal held firm. The debate over these data continued into 2024 (see above).
Treves, A., J.A. Langenberg, J.v. López-Bao, M.F. Rabenhorst 2017. Gray wolf mortality patterns in Wisconsin from 1979 to 2012 Journal of Mammalogy 98(1): DOI:10.1093/jmammal/gyw145
While exploring the idea of public trusteeship for wolves, we examined how agencies should and should not act. We discussed agency capture by narrow interests with Wisconsin DNR as a case study.
Treves, A., Chapron, G., Lopez-Bao, J.V., Shoemaker, S., Goeckner, A., Bruskotter, J.T. 2015.
Predators and the Public Trust. Biological Reviews doi: 10.1111/brv.12227.
In 2016, Dr. G. Chapron and Dr. A. Treves published 'Blood doe snot buy goodwill' igniting a scientific debate that still rages in 2025. Some of the history of that debate is recounted below. In 2025, I would say the preponderance of evidence supports the notion that liberalizing wolf-killing increases poaching, reduces tolerance, and slows wolf population growth beyond that expected from the legal killing. All the evidence points to an increase in illegal killing.
Chapron, G. and A. Treves 2016a and b, 2017a and b.
In 2016, Chapron & Treves published the first challenge to the notion that legal killing reduces illegal killing of wolves, a notion which had been put forward by The US government in a court case about protections for gray wolves. Chapron & Treves (2016a,b) presented population dynamic evidence that the opposite occurred, liberalizing wolf-killing increased wolf-poaching. This was challenged without new evidence by Pepin et al., Stien, and Olson et al. in 2017, so the debate simmered until 2020 when Dr. Francisco Santiago-Ávila et al. presented estimates of gray wolf survival and incidence of death and disappearance among radio-collared Wisconsin wolves. The data presented no evidence that liberalizing wolf-killing did anything good for the wolves as individuals or as a population. Those researchers formulated the hypothesis of facilitated illegal killing. That hypothesis was replicated for Mexican gray wolves in Arizona and New Mexico by Louchouarn et al. 2021 and for North Carolina red wolves by Santiago-Ávila et al. 2022. Also in 2022 and 2023, two new analyses of Wisconsin and Michigan gray wolves reaffirmed the inter-annual increases in deaths and disappearances during periods characterized by different policies, different mammal-hunting seasons, and different political administrations (Santiago-Ávila & Treves 2022; Louchouarn 2023). Breck et al. (2023) published a contrary view of Mexican gray wolf removals but their analysis has also been questioned on multiple methodological grounds. Although the scientific debate is unresolved, the discussion has advanced measurably in the rigor of scientific analysis and the quality of evidence. In my opinion the balance of evidence is that policies and politics that reduce the value of wolves or encourage lawlessness contribute to wolf-poaching. And liberalizing and expanding wolf-killing has been the major cause of devaluing wolves in many regions.
Since 2018, the debate in Nordic countries has been equivocal with Liberg et al. (2020) arguing that regulated wolf-hunting lowered poaching, a conclusion challenged by Treves et al. (2020) on the basis of the same data. Also in Finland, Surtainen & Kojola (2017, 2018) reported that legal killing removed gray wolves before poachers could do so and that regulated killing was not a solution to illegal killing in Finland. Although the scientific debate is unresolved, the discussion has advanced measurably in the rigor of scientific analysis and the quality of evidence. In my opinion the balance of evidence is that policies and politics that reduce the value of wolves or encourage lawlessness contribute to wolf-poaching. And liberalizing and expanding wolf-killing has been the major cause of devaluing wolves in many regions.
See the Publications page for references to Treves, Chapron, Santiago-Ávila, and Louchouarn. For other references, please contact
Evidence was accumulating by 2014 that tolerance for wolves actually declined when the government made it legal and easier kill wolves.
Treves, A. Bruskotter, J.T. 2014.
Tolerance for Predatory Wildlife. Science 344: 476-477.
The below article was the first evidence that the assumption that lethal management improved tolerance for wolves was invalid. We went on to study that again and again with Dr. C. Browne-Nuñez, Jamie Hogberg, Dr. F. J. Santiago-Ávila, Dr. S. Agan, and Dr. N. Louchouarn all corroborating that the state and federal claims were unsupported.
Treves, A., Naughton-Treves, L., Shelley, V. 2013.Longitudinal Analysis of Attitudes Toward Wolves. Conservation Biology 27: 315–323
Wydeven and I show that pack size estimates are confounded by migration and births, so none of those estimates can be considered independent of each other. That makes estimates of birth rate and even population growth rates suspect. It took us years to understand how that was confusing Stenglein, Wydeven, and van Deelen's estimates of population dynamics.
For example, we showed that Stenglein & van Deelen made errors and failed to share data: Treves A, Darimont CT, Santiago-Ávila FJ. 2023. Comment on correcting Stenglein & van Deelen 2016 & comment on 2022 correction to Stenglein & van Deelen 2016. PLoS One Comments. 2022. 10.13140/RG.2.2.20984.52484.
Wydeven, A.P., Treves, A., Brost, B., Wiedenhoeft, J. 2004.
Characteristics of wolf packs in Wisconsin: Identification of traits influencing depredation. Pp. 28-50 in People and Predators: From Conflict to Coexistence. N. Fascione, A. Delach, M. Smith, eds. Island Press, Washington, DC.