On February 9th, the US district court in Missoula, Montana heard oral arguments in the case challenging the 2014 USFWS decision not to list wolverines. The Missoulian published an article about the hearing, and it’s a pretty good outline, but since I was in the courtroom, I’m sharing my impressions and a more detailed synopsis as well. I’ll focus here on conveying what happened, and will share my own analysis later. I do have opinions about the debate around the science, and those will probably come through, but I will explain my reasoning and, I hope, convince readers as to why that reasoning is sound, in a subsequent post.
First, a bit of background: Back in August of 2014, the USFWS abruptly reversed course following a 2013 proposed listing rule (Proposed Wolverine Rule), and issued a decision not to list the wolverine as threatened under the ESA. A leaked memo (Region_6_Wolverine_Memo_5-30-14) revealed that the reversal seemed to be the result of the opinion of the assistant regional director of region six, in contradiction of a process of several years that found that wolverines were warranted for protection and should be listed. Environmental advocacy groups sued, charging that the decision was “arbitrary and capricious,” and contending that the decision should be thrown out. The defendants included the USFWS, and a number of parties referred to as “defendant-intervenors.” In this case, these parties included the states of Idaho, Wyoming, and Montana; Montana Fish, Wildlife, and Parks; state farm bureaus; several snowmobile associations; and – perplexingly- the American Petroleum Institute and the Montana Petroleum Institute.
To show that the decision not to list was in fact arbitrary and capricious, the plaintiffs had to prove that the reversal was not based on any sound scientific evidence, but rather on whims or machinations outside the bounds of the “best available science,” the standard tool used to determine whether a species warrants protection.
The notion of best available science is tricky. The ESA did not define it when the Act was written, and a definition added by amendment in 2005 – “…scientific data, regardless of source, that are available to the Secretary at the time of a decision or action for which such data are required by this Act and that the Secretary determines are the most accurate, reliable, and relevant for use in that decision or action.” – provoked ire from people who thought that it opened science to politicization by granting a government official the authority to decide what science counts and what doesn’t. This doesn’t even begin to touch on the larger debate about whether our process is effective for incorporating science into policy at all, or the even vaster debate, within the social sciences, about whether science that strives for neutrality is actually a practice of collective cultural self-delusion. Obviously, figuring out what defines good science is a complex issue.
Regardless, the question of best available science has plagued the latest round of the wolverine listing debate, just as the lack of scientific data plagued earlier rounds. The USFWS’ original proposed rule to list, like all such proposed rules, was subject to peer review. During this process, two out of the five researchers who were invited to submit opinions disagreed with the USFWS’ analysis of the science on the relationship between wolverine denning and late spring snowpack, and the models that projected decline in snowpack in wolverine range within the coming century. This dissent provoked an avalanche of further analysis and debate, which included, among other things, expert panels on climate modeling, rancorous attacks on various scientists and scientific institutions, the construction by the states of some fairly shaky arguments about wolverine population trends, and the sudden appearance on the scene of high-powered energy players with – we can speculate – a distinct interest in preventing the listing of a charismatic climate-sensitive carnivore. Months went by. The expert panel concluded that the climate modeling was sound. The decision looked like it would go forward. Then came the reversal; the leaked memo; a letter from the states commending the decision not to list but reiterating flawed arguments about what the science actually says; and the lawsuit.
The major scientific debates in this case – the points that were contended and argued over in the expert reviews, the letters, the memos, and the public forum – can be summarized as follows:
- What is the relationship between wolverines and snow, especially spring snowpack?
- What is the wolverine population trend in the US Rockies?
- Is genetic depression potentially a problem for wolverines?
- Are the climate models used to predict reduction in snowpack in the Rockies adequate, both scientifically, and as a justification to list?
The three papers most at issue in this discussion are Copeland et al 2010, which deals with the relationship between wolverines, late spring snow, and low summer temperatures; McKelvey et al 2011, which deals with projected retractions in wolverine habitat and dispersal corridors over the coming century; and Aubry et al 2007 (aubrywolverinedistribution), which deals with the historic and current range of wolverines. I’ve discussed these papers elsewhere on this blog and will dive back into them in subsequent posts, but for now, if you want to review them, there they are.
And thus we come to Missoula on Tuesday morning, February 9th, at 9:30 am. The courthouse gallery was packed – the lady who signed me in and gave me my visitor tag told me that there were at least 50 people, and opined that there were “a lot of people interested in wolverines.” There were also a lot of attorneys, arrayed around their tables facing the judge’s seat – Tim Preso of Defenders of Wildlife and Matthew Bishop of Wild Earth Guardians for the plaintiffs, and the USFWS attorney and the assortment of lawyers for the defendant-intervenors.
Judge Dana Christensen began the proceedings by recounting, with articulate and detailed enthusiasm, his three encounters with wolverines in the wild, and segued into expressing his familiarity with the case, the briefs, and the wider issues at stake. He discussed the prior listing petitions and their outcomes. He mentioned having read Doug Chadwick’s book about the Glacier Park project, and said that he understood from the book that, “The folks who are committed to gathering the science are dedicated…I’ve concluded that this is a hard species to study, and it’s not surprising that we don’t know how many there are….or the exact data about their response to climate change.” Wolverines are a pretty obscure topic and it would have been easy to end up with a judge who didn’t even know what they were. Judge Christensen was admirably well-informed.
The plaintiffs opened by invoking the concept of best available science, and putting forth the argument that the papers used to justify the original listing rule – namely, Copeland et al. 2010 and McKelvey et al. 2011 – were subjected to criticism and then discounted on the basis of speculation rather than definite scientific evidence that they were flawed or incorrect. These two papers have been at the crux of all of the debates, and most of the criticism lobbed at them has had to do with lack of precision rather than lack of broad accuracy. Preso invoked a 2009 case, Tuscon Herpetological Society vs. Salazar, which determined that the government cannot dismiss threats to a species based on inconclusive science. As he stated again and again, in numerous ways, “The ESA does not demand perfect science – it demands the best available science.”
Preso also spent some time on the USFWS’ assumptions about population growth, and the flawed reasoning of relying on Aubry et al 2007 to justify the idea that the population would continue to grow. He argued that the USFWS and its attendant intervenors were conflating range expansion with population growth, and that there was no evidence that the population is growing. He highlighted the dangers of assuming that increased sightings of wolverines indicates that the wolverine population is increasing, mentioning that it was not surprising to see that wolverine sightings had been substantially higher during the years of a wolverine research study than in the years when the study was not operating.
Connectivity and genetics also received attention. Preso argued that connectivity problems should have been considered a primary rather than a secondary threat, referencing the lack of apparent connectivity between wolverines in the contiguous US and wolverines in Canada. The judge questioned this, and Preso responded with the published information that suggests that the trans-Canada highway and Canada’s trapping management regime do serve as effective barriers to free genetic exchange with wolverines in the US Rockies. From here, he ventured into a discussion about widely-accepted biological rules about the problems with inbreeding and genetic bottlenecking, and suggested that these rules most likely apply to wolverines as well. He was referring specifically to the so-called 50/500 rule, which states that a minimum effective (breeding individuals only) population of 50 individuals is needed to insure population survival over the short term, and 500 are needed over the long term. The current effective population of wolverines in the US Rockies is estimated, based on genetics, to be between 25 and 50.
Preso’s next theme was climate change, and the criticisms made of the Copeland et al 2010 paper. The judge asked several questions as Preso launched into his defense of the paper, the first about whether or not there was any published information to contradict either Copeland or McKelvey (Preso: “No.” Judge Christensen: “So it’s basically just criticism….”) Later, as Preso continued to explore the issues around McKelvey et al. 2011 and the question of the scale of the modelling, and stated that, “We know that wolverines are snow-obligate – ” the judge interrupted and said, “That’s a given in this case, so the question is, do we need to know exactly why?”
There was some discussion about another case that had to do with abrupt changes in decisions, which the judge brought up in order to ask whether Preso wanted to rely on this case for precedent; Preso said no and brought the discussion back to the Tuscon Herpetological case. The judge went on to affirm that this was a case about “change in policy needing to be based on good reasoning and logic.”
Preso also spent some time reiterating that California and Colorado, with single male wolverines in each state, should not count as inhabited range. He brought up the fact that some of the estimates of available habitat included habitat in these states, and that that habitat should be removed from analysis because it was not relevant to the population in the US Rockies or the Cascades.
At the end of Preso’s arguments, the judge asked him what he wanted the court to do, and Preso replied that he wanted the judge to apply the arbitrary-and-capricious standard and set aside the rule, remand the decision, and “wait and see what will happen.”
Matthew Bishop then spoke, and in some senses he seemed to contradict Preso’s arguments about not considering uninhabited range as part of the analysis for listing. He made an argument that the retractions in range during historical time should in fact be taken into account in determining the degree of threat faced by wolverines. This discussion revolved around a specific provision of the ESA, and Bishop was arguing that this should be applied; I haven’t had time to research it, so I’ll leave this here for now. Bishop also mentioned genetic issues, and inadequacy of existing regulatory mechanisms (one of the categories of threat under the ESA). Like Preso, he stated that the best available science in this case had been ignored.
Then came the defendants’ turn to speak. Trent Crable was arguing for the USFWS, and stated that since climate change was ruled the major and only primary threat, the science regarding wolverines and climate change should be the only issue under consideration. The USFWS had found that McKelvey et al was not enough, on its own, to show threat in “the foreseeable future.”
The judge at this point asked whether there was any published research to contradict Copeland or McKelvey, and Crable responded, “If by published, you mean peer-reviewed, then no.” The judge then asked, “What then was Noreen Walsh [the region 6 assistant director] relying on in making her decision?”
Crable replied, “Her understanding of what is needed to list, and the report provided by Dr. Torbit, in consultation with Andrea Ray of NOAA, saying that the modeling was insufficient for us to know what will happen to wolverine habitat.” This was in reference to an internal study on downscaling of climate models that the USFWS requested after the expert review panel found that the the snow modeling in Copeland and McKelvey was adequate. I have not seen this report, but it is referenced in Walsh’s memo, and was the linchpin in justifying the reversal. Crable explained that the report highlighted difficulties in modeling what will happen with precipitation on local scales; he said that modeling temperature changes was comparatively easy, but that precipitation projections were much more difficult, implying that our ability to understand future snowfall is compromised.
The judge produced emails that seemed to suggest that Walsh had made the decision to reverse prior to the submission of the Torbit and Ray report, and asked several questions to try to clarify the timeline. Crable said that Walsh had probably seen earlier drafts and/or talked with Torbit, since they are colleagues. The judge asked whether other scientists within USFWS or beyond were given time to respond. I don’t recall a direct answer to this question, but the implied answer was no, and as the exchange continued, the judge said, “So the same people who were tasked with the listing rule were then tasked with coming up with the exact opposite decision? That must have been distressing, if not demoralizing.” Crable stated that it was not the “exact opposite view, just a different conclusion.”
At this point, Crable steered the discussion back to McKelvey, saying that although McKelvey represents sophisticated science, the question was whether that single paper was enough to list. The USFWS contended that it wasn’t because it doesn’t tell us what will happen to habitat in the future, which is the real issue. He then stated that the burden of proof is on the USFWS to show why a species should be listed, not why it shouldn’t be listed, and said, “This is the law for listing under section 4.”
Despite his contention that other arguments about secondary threats should not be at issue, population trend was a theme with both Crable and the other lawyers at his table. Most of their arguments relied on Aubry 2007, a paper about historic range; they were using this paper to build an argument that wolverines have recovered in the US Rockies following an early-20th-century extirpation. Crable did venture into discussing genetic diversity and the 50/500 rule, stating that this rule was difficult to demonstrate in reality and that “small population sizes don’t necessarily mean that there is a threat,” and – again, drawing presumably on Aubry – “Population size is probably not substantially lower than prior to European colonization.”
I found the breadth of assumption in this statement so astonishing that I was scribbling notes and missed the exact flow of the next few moments of argument, but soon thereafter, the issue of trapping was raised. Montana maintained a trapping season – at first unlimited and then, after 2008, much more carefully managed – until it was shut down by court order several years ago, pending the listing decision. Crable said that “no trapping of wolverines is allowed,” in response to a line of discussion about the potential effects of trapping on the population. The judge said that he understood that the closure was the result of a court order, and Crable confirmed this, but said that the court order was lifted and that trapping had not resumed.
This was the conclusion of Crable’s argument, and he took a seat, while the lawyer from Wyoming, whose name was Peterson, stood up to speak for the states of Wyoming, Idaho, and Montana. He talked about the environmentalists using whatever means necessary in “attempting to achieve their goal, which is listing the wolverine.” He argued an equivalence in uncertainty over population growth and the uncertainty in Copeland and McKelvey, saying that “environmentalists” were applying a double standard by invoking lack of evidence of population growth to justify listing, while ignoring uncertainty in Copeland and McKelvey that suggested that wolverines might not be threatened. He said that it was the responsibility of the USFWS to decide what represented an acceptable level of uncertainty for a particular decision. “At heart,” he said, “This is about a difference of opinion” – again, neatly implying a scientific equivalence between the two sides of the discussion.
Peterson brought up the fact that wolverines had been extirpated and then rebounded over the past century as evidence of population growth, in spite of a continuing trapping season, and then said that “there was no trapping when the decision was made,” reiterating Crable’s implication that trapping was not something to worry about. Here, Judge Christensen interjected, bringing up emails sent by the USFWS about potential reactions by various parties to the reversal of the decision to list. Among these, the section about the reaction of the states proclaimed that they were unlikely to object, since “many of the arguments for the withdrawal of the listing decision originated with the states.” The email also clearly said that Montana intended to reopen the trapping season. The judge questioned the intent of the states. Peterson quickly deferred to the lawyer for Montana, who stood up and stated that “We’d like to keep a limited and carefully controlled trapping season on the table.”
The states took a seat, and the lawyer for the “non-government defendant intervenors,” a coalition representing farm bureaus and snowmobile associations, took his turn at the podium. This was by far the wackiest segment of the hearing, because the lawyer, Blevins, argued that wolverines in the US Rockies should not be considered a listable entity since they are a subspecies with conspecifics in Eurasia, and in order to list them, we would need to do a review of the global population. I’m going to let my bias out to play here and get it over with so I can focus on more relevant discussions in subsequent posts – this argument was ridiculous, but in light of who the lawyer was representing, the absurdity may have an explanation. As I understand it, he was attempting to get into the record a nitpicking dispute over semantic ambiguity in the ESA, which is certainly an object of almost totemic hatred among farm bureaus in the western US. I’m pretty sure I saw people of significant stature rolling their eyes when he stood up. During the rebuttal, Crable disavowed association with Blevins’ line of argument. Enough said.
Next up were the “energy intervenors,” the American Petroleum Institute, and the Montana Petroleum Institute. The lawyer for the Montana Petroleum Institute, D’Angelo, stood up to present the energy industry’s arguments, but before he could begin, the judge interjected to ask, “Please tell me why the American Petroleum Institute and the Montana Petroleum Institute have a dog in this fight?”
D’Angelo responded by citing concerns about restrictions on operating in wolverine habitat if the species is listed, and then quickly went on to state that the USWFS started with a conclusion that wolverines should be listed, and then backfit evidence to that conclusion.
The judge here interjected again to ask, “Where’s the evidence in the record? That’s a serious contention.”
D’Angelo brought up emails from Shawn Sartorius, who wrote the listing decision. In the courtroom, Sartorius was quoted as writing that “wolverines will have a proposed rule,” and D’Angelo argued that this was evidence that the USFWS was biased. Again, I’m going to get this out of the way by contextualizing where the lawyer failed to do so. I don’t know what was going on in Sartorius’ mind when he wrote that, but these emails were sent at a specific moment in the listing debate. In 2010, wolverines were deemed warranted-but-precluded, which meant that a scientific analysis found that there was justified evidence of a threat, but that the USWFS did not currently have the resources to list the species, particularly in light of more immediate and habitat-based threats facing other species. In essence, and perhaps with some defensible logic, the decision said that the resources of the agency should be applied to species who are threatened by something that the ESA is capable of dealing with (in situ habitat-based threats) rather than something that the ESA is not authorized to regulate (carbon emissions). A warranted-but-precluded ruling places a species in limbo, which can last for many years. Shortly after the 2010 wolverine decision, environmental groups sued over undecided ESA cases, including warranted-but-precluded decisions, and won an order that all of these cases had to be decided within a limited time frame. The wolverine was one of the first to go up for consideration. The McKelvey et al 2011 paper that is the object of so much dispute had entered the literature in the interim, but beyond that there was no new peer-reviewed science to consider. Since wolverines had already been found warranted for listing, and McKelvey added substantiation to the idea of threat, Sartorius’ emails at this point are less evidence of a nefarious plot to rig scientific data to fit a pre-ordained, agenda-based conclusion, and more a logical outgrowth of the fact that a very recent review of the science had reached a particular scientific conclusion that suggested that a listing rule would follow.
The judge didn’t get into any of this, he let the matter sit, and D’Angelo went on to assert that, “It’s not reasonable to dispute that the population is increasing….and projected to increase.” He referenced an Inman paper from 2013 and Aubry 2007 to back his claim, but the judge again began to question him, saying that, “Within the [proposed listing] rule, scientists disagree [about population trajectory], I can’t see where everyone agrees that population will increase forever – maybe you weren’t saying that?”
D’Angelo replied, “I wasn’t. I apologize.”
The judge said, “It sounded like you were.”
D’Angelo managed to recontextualize his argument and put some time scale boundaries on his contentions about population growth. He then ventured into a criticism of Copeland and the snow modeling, stating, among other arguments, that the obligate relationship between wolverines and snowpack was “on a denning scale,” and that “Copeland describes where wolverines are, not what they need.” He said that the authors on Copeland et al “drew a line around the wolverine population” and then backfit the snow data. He concluded by saying that the paper was good for predicting where wolverines are found, but not as a premise for the McKelvey paper, which relied on Copeland to model habitat loss.
With this, the defendants concluded, and the rebuttals began.
During the rebuttal, Preso talked about modeling and uncertainty, saying that McKelvey was not a stand-alone, that the expert panel convened by the USFWS agreed with him, and then said that the USFWS had in the past relied on the same kind of modeling to conclude that pikas were not warranted for listing – that decision stands. Preso stated, regarding the contention that Copeland shows where wolverines live but not what they require, that it was unreasonable to assume that wolverines were living in places that didn’t provide what they needed. He again talked about uncertainty in population trend, citing a number of studies including Inman 2013 to highlight the lack of any kind of evidence about current demographics. He reiterated the lack of connectivity with Canada, and said that the 50/500 rule was “basic biology” and that it should apply to wolverines. The judge briefly questioned him about the uncertainty in modelling precipitation versus the relative ease of modeling temperature. Bishop then spoke again, briefly, about the fact that trapping was “still on the table” by the admission of the lawyer for Montana. The judge sought some clarification about the concerns over trapping, saying that he understood that incidental take was a possible source of mortality on top of the prospect of a managed season, and Bishop confirmed this.
Crable then gave his rebuttal, drawing again on Aubry to contend that the population is increasing and will continue to do so. He said that decline in snowpack doesn’t mean that wolverines are threatened, nor does inbreeding or genetic depression, and that there was not enough evidence of any of it to warrant listing, He again came back to population growth, saying that “estimates have gone up in published literature,” that there was no evidence that the population was declining, and that – again – there were no grounds to list.
He then ventured into the trapping discussion, saying that Inman had argued at some point – I missed the paper reference – that trapping was not a problem and might even be good for the effective (breeding) population, “because if you take out a male and he’s replaced by multiple other males, it will increase the effective population.”
The judge intervened here and said, “I don’t suppose that you’re suggesting that we resume trapping in wolverine habitat to increase the population?”
Crable looked somewhat sheepish and said no, that wasn’t what he was suggesting.
He concluded by saying that this was not “a coin flip” between equally robust sets of evidence. He emphasized again that a single study (McKelvey) was not enough to justify the listing, and said, “just because they [the USFWS] said that there was another way to do this doesn’t mean it was a coin flip.”
With that, the hearing concluded. The judge stated that he had “a lot of work to do” on this decision. We will probably have to wait at least six months for a ruling.
Thoughts and comments welcome. Stay tuned for some analysis over the next week or so.
The Wolverine Blog 