Robert T. Lackey

Department of Fisheries, Wildlife, and Conservation Sciences

Oregon State University

Corvallis, Oregon  97333

Citation:  Lackey, Robert T.  2022.  What Would Happen to Columbia River Basin Wild Salmon Runs If Hatchery Stocking and Fishing Ended?  Published in an Oregon State University Blog, September 7.



            Wild (non-hatchery origin) salmon and steelhead in the Columbia River Basin are not doing well.  Basin-wide, wild salmon and steelhead are roughly 2-4% of their pre-1850s levels.  Even the total (wild plus hatchery) current salmon and steelhead run is now less than 10-20% of the pre-1850s levels.  About 140 million ocean-going juvenile salmon and steelhead are released annually from fish hatcheries in the Columbia River Basin.  Currently, the annual harvest (i.e., from commercial, recreational, and tribal fishing) is strongly dominated by fish of hatchery origin.

            Even without major human intervention, salmon runs (whenever I refer to “salmon,” I include steelhead) are notoriously variable annually and decadally.  Still, the overall downward trajectory has been apparent over the long term (150 years).  Interestingly, the current low overall salmon abundance has not shown an obvious continuing downward trend for several decades.  Nor has the trend been plainly upward.

           Deciding how important preserving wild (vs. hatchery origin) salmon runs and/or sustaining fishing (vs. the myriad competing public policy priorities such as flood control, energy production, agriculture, etc.) are decisions that society continues to make either explicitly or implicitly.  Of course, scientists do not make such tradeoffs, but they should play an important, even essential, role as sources of impartial scientific information and assessments.

            One important role is providing expert judgment about complex scientific questions that are beyond the capability of the traditional scientific method.  Thus, scientists are understandably uncomfortable when asked to answer “scientific” questions about policy-relevant questions that the scientific method cannot answer.

            To help provide relevant “expert” scientific professional judgment to policy and management deliberations, I conducted a “thought experiment” by asking Columbia River Basin salmon and steelhead scientists how (1) stopping fishing and (2) closing hatcheries would most likely affect the remaining wild runs.

            In the persistent political controversy over salmon policy tradeoffs, some advocates argue that one or both changes (i.e., stop fishing and close hatcheries) would increase the abundance of wild salmon in the Basin — perhaps even increase the wild plus hatchery salmon abundance.  Probably the most well-known example of “close the hatcheries to save wild salmon” advocacy is the Patagonia-produced film Artifishal:  The Fight to Save Wild Salmon, which millions of people have viewed.  Many other organizations push similar policy perspectives, although not with a billionaire’s financial backing or marketing platform.

            Other advocacy organizations push the assertion that continued fishing and hatchery stocking are not currently among the major causes of the very low levels of wild salmon in the Columbia River Basin.  An example of such advocacy is the video Benefishal, a direct response to the Patagonia-funded film.  Such advocates typically argue that the main reason that any Basin salmon fishing still exists is because of hatcheries.

            Either view — and the many nuances between the two — might seem counterintuitive until the array of past and current policy drivers are considered.

Ecological Context

            As a start, consider how much the physical and biological features of the Columbia River Basin have changed since the pre-1850s.  Unlike native salmon and steelhead, non-native anadromous fish species thrive in the highly altered aquatic environment.  A simple statistic captures the sobering reality.  In 2021, the Independent Scientific Advisory Board of the Northwest Power and Conservation Council (ISAB) determined that “ . . . in 2019, about 30 times more shad passed over Bonneville Dam than salmonids (around 7.5 million vs. 0.25 million).”  American shad, a non-native anadromous species, is one of many non-native fish species that thrive in the present-day Columbia River Basin.  American shad, walleye, crappie, smallmouth bass, largemouth bass, bluegill, yellow perch, channel catfish, and northern pike are all relatively new species to the Basin.  Their success is not surprising to biologists because these species are well adapted to current (and likely future) ecological conditions within the Basin.  In stark contrast to native salmon and steelhead, these new arrivals thrive by expanding into much of the highly altered Basin aquatic environment.

            It is not that salmon are particularly delicate species, but like all species, they have specific environmental requirements.  They can also adjust to some unfavorable conditions.  For example, salmon do “stray” in an attempt to find better environments.  In one well-known example, the famous Bridge of the Gods blockage of the Columbia River around 600 years ago blocked salmon for several years.  When salmon eventually could move past the barrier, they re-colonized the middle and upper Basin fairly quickly.  Such blockages were surely common in the Columbia Gorge on the millennial time scale.

            Even before the Bridge of the Gods and similar landslides, the Bretz Floods scoured the mainstem Columbia multiple times 13,000 – 15,000 years ago.  The distribution and abundance of salmon in the Columbia River Basin would drastically change with each flood (and recovery).

            Among salmon scientists, the mid- or pre-1850s are typically selected as the baseline period to benchmark current salmon abundance.  However, the 1850s was toward the end of the Little Ice Age, a relatively cool period (1300s – 1800s) that was likely much more favorable to salmon than the Modern Warm Period (1900s – ).

            Complicating matters further is the biological reality that salmon and steelhead are fairly adaptable and thus survive in diverse habitats across broad geographic expanses.  As populations, they often successfully extend their distribution into different environments.  For example, salmon were successfully introduced and spread widely in New Zealand, Chile, Argentina, northwestern Russia, and the U.S./Canada Great Lakes.  They colonized other rivers from relatively limited initial introductions and often established self-perpetuating runs.  For a contemporary and active example, pink salmon (the species native to the Columbia River Basin), introduced into northwestern Russia, are now expanding into rivers in Norway and probably Sweden.

            Caveats aside for the Columbia River Basin, what was the “natural” abundance of salmon and steelhead in prior times?  Specifically, how many salmon and steelhead did the Basin support before the 1850s — before most human-driven alterations dramatically reduced salmon numbers?  After evaluating all available historical reconstruction data, the Columbia River Basin Independent Science Advisory Board that the Basin-wide pre-1850s aggregate run was likely in the range of 5 to 9 million adult fish per year, with the most likely level at around 6 million.  Some other estimates are higher, with the upper bound at 15-16 million.  Reconstructing historical run size nearly two centuries ago is challenging — and such assessments should be accepted cautiously.

            These days, fundamental technical questions about the influence of fishing, hatcheries, and other factors in controlling the number of salmon (wild and hatchery origin) remain difficult to answer, but have important policy and management implications.  For example, many advocates argue that fishing for Columbia River Basin stocks is only viable because hatchery releases sustain current runs.  Further and usually, only fin-clipped (hatchery-origin) salmon may be kept, which lessens fishing mortality on wild fish.  Without hatchery supplementation, the argument goes, wild runs would be too small to support sustainable harvests at levels demanded by fishing interests.

            Classifying a salmon as “wild” requires a degree of arbitrariness.  To some, a “native” species may be interchangeable with “wild.”  To others, “wild” might mean an individual fish whose ancestors were never in a hatchery (i.e., not “naturalized”).  However, when I use “wild” to classify a salmon, I use the definitionAny salmon spawned naturally and whose parents spawned naturally.”  This choice is subjective and used for communication clarity rather than implying a policy preference.  This definition means that the offspring of a hatchery fish that spawned naturally (without human intervention) in natural (non-hatchery) habitat are categorized as “wild.”  At the end of the day, what society defines as “wild” is a policy or political call, not a scientific one, but it is important to be clear about which boundaries are used to classify individual salmon as wild or hatchery origin.

            From another perspective, some say that the only reason there are any “wild” salmon in the Basin is that they (the so-called wild fish) are offspring from hatchery strays that spawned naturally in natural habitat and thus are legally “wild” fish.

            Conversely, others claim that most wild runs would flourish without hatcheries and reach sufficient levels to support significant fishing pressure.  Thus, hatchery operations are a cause of the low wild runs, and, therefore, they ought to be closed to allow wild runs to rebound.

            Of course, answers to these and similar questions involve some hard scientific “facts,” but such questions are far too complicated to answer with traditional scientific experiments and analysis.  Modeling might provide a tempting analytical tool, but understanding key ecological/mathematical relationships is limited.  Further, the Basin itself has changed in fundamental ways (i.e., dams and substantial changes in land and resource use) in the preceding two centuries.  Thus, the opinion of acknowledged experts likely provides the most credible answers to the Basin’s most pivotal and complex ecological questions.

Wild Salmon Recovery

            No matter how obvious the past may be to salmon experts, it is important to start salmon recovery discussions with a little historical context.  Nearly all salmon runs in California, Oregon, Washington, and Idaho are a shadow of their pre-1850s levels.  Further, most of the remaining runs are largely maintained by releases of hatchery-raised fish.  Wild salmon comprise only a small portion of most runs.  Their overall abundance is a sliver of historical levels.  The few relatively abundant current runs are mainly species (i.e., pink and chum salmon) that spend little time in freshwater before migrating to the sea.

            This decline has been known for 150 years, and concerted and substantial efforts have been implemented to recover salmon runs.  Especially during the past few decades, the extent and cost of formal recovery efforts for wild salmon have substantially increased — largely a response to lawsuits filed under the Endangered Species Act (ESA).

            Every watershed presents unique features and complications.  Further, the species are highly variable, and there are exceptions to almost every biological generality.  However, overall, the Columbia River Basin follows the typical downward trend in salmon abundance seen throughout California, Oregon, Washington, Idaho, and British Columbia.

            While using hatcheries to sustain relatively large salmon runs is plausible — although technically challenging — the requirements of the ESA relative to wild salmon have made the role of hatcheries in maintaining or increasing runs legally contentious.

            In my interactions with professional colleagues, they agree — usually only when speaking unofficially — that current efforts will not successfully recover wild salmon to abundances that would assure self-sustainability and support sizable sport, commercial, and tribal harvests.  Arbitrarily, such a level of abundance would need to be at least a third of the typical pre-1850s run size to provide politically adequate fishing opportunities, but would still be far short of historical abundance.

            But when technical and scientific judgments go against “conventional wisdom,” are scientists likely to be candid publicly?  Perhaps, when professional scientists can keep their technical and scientific judgments anonymous (and avoid attacks from policy advocates), their “best guess” might be different.

The Thought Experiment

            The thought experiment involved identifying 100 nationally and internationally known experts on the Columbia salmon situation who have published widely in peer-reviewed scientific literature.  Based on their extensive scientific publications and other professional activities, these individuals are well known to those who work on salmon issues.  All have worked on Columbia River Basin (and elsewhere) salmon issues for decades.

            I promised anonymity to all respondents.  In the past, working with experts, I was surprised that even “big names” are reluctant to be quoted on such important ecological and policy questions.  Specifically, when evaluating salmon issues in the past, I have been amazed to learn that what experts say in public is sometimes not the same as what they say “off the record.”  Hence, I guaranteed anonymity to all participants to increase the probability of obtaining more candid responses.

            Here is the question sent to each:

            I hope you are willing to contemplate and answer with an “expert opinion” (i.e., “a guesstimate”) about the future of salmon in the Columbia River Basin.  Briefly, here is the context for the question . . . after someone watched a recent Trout Unlimited presentation of mine, he asked a question that required, on my part, pure speculation to answer.  Essentially the listener asked for my opinion about the role of hatcheries and fishing in controlling the current and future abundance of wild salmon in the Columbia River Basin.  I offered him a guesstimate along with the usual caveats and cautions.  Still, I am not sure that my answer reflects the thinking of the “scientific community,” specifically those most knowledgeable about the issue (that would be you).  So . . . let me describe the thought experiment and formulate the precise question.

            To set the context for this thought experiment (and for some difficult-to-imagine scenario), assume that the U.S. Government decides to immediately and permanently (1) eliminate all releases from hatcheries (i.e., close them all forever) and (2) stop all fishing directly targeting salmon and steelhead.  Now . . . allow for time for at least a half dozen generations of completely “wild” spawning (i.e., several generations of fish whose recent ancestors never were in a hatchery).  Further, allow for time to pass so enough years of run size data would somewhat dampen the noise from ocean variations and other climatic factors.  So . . .  the question . . . after 20-30 years or so of (1) zero hatchery releases;  and (2) zero targeted fishing . . . roughly what would be the overall run size in the Columbia Basin compared to current average runs?  Stated differently, the question is, ‘approximately how many wild salmon and steelhead would the Basin support without salmon fishing and without hatchery releases, and how would that abundance compare to the past several decades and the run levels of the mid-1800s?’  Be assured that your response will be entirely anonymous.  Because you have spent years immersed in this and similar aspects of salmon management and policy, I look forward to reading your “best guess” answer.”

            Although the participants are not listed here, they are all well-known in the world of professional salmon scientists, managers, and policy makers.  At least in the salmon world, they are “household names” that have been professionally active for years.  Very importantly, I thank them all for their candor in answering the question and for trusting me to maintain their anonymity.


            The Appendix includes the 58 complete answers reported and discussed here.  I reworded (with the author’s permission) some initial responses to provide some measure of formatting consistency.

            Of the 100 initially identified experts, I was able to send email requests to 96.  Of those, 60 agreed to respond to the question.  Two original “acceptors” eventually withdrew their answers because of a concern that the “opposition” or “other side” would use the results to undermine the policy choices that these two individuals strongly supported.  In the end, there were 58 of the 60 formally responded to the question.  Six of the 96 declined to answer the question, typically citing health or other personal reasons.  Three others agreed to answer, but never did submit one, even after receiving an email follow-up request.  Twenty-nine never acknowledged receiving the initial email request, nor a follow-up request.  Invalid email addresses were likely a cause for at least some of these non-responders, especially because some of these individuals were no longer formally employed by fisheries organizations.   A summary of the participation rates is presented in Table 1).

            As for the other 58 individuals who did provide estimates, how much their answers were affected by personal or employer policy preference is unknown.  However, based on some policy perspectives expressed in emails, policy preferences likely were a factor in some cases.   Most (60%) of the contacted salmon experts participated, probably providing a representative sample of prevailing opinion.

            I expected to receive many “back-of-the-envelope” and one-sentence guesses, and this was common, but so were detailed analyses and answers.  I was impressed by how many respondents took the time to provide logical backing for their key conclusions.  Some wrote pages of detailed analysis justifying every assumption used in answering the question.  Others bluntly acknowledged that their answer was a “shot in the dark,” and rigorous analysis would not change the reality that the answer provided was an “expert opinion” and nothing more.

            As expected with such a general and speculative question, even with light editing (with the author’s involvement and permission), there was considerable variation in how each person worded the answer.  To simplify analyzing the overall results, I arbitrarily classified each response into one of 5 generic answers (Table 2 and Appendix).  Not every prediction was precisely worded, so in those few cases, and after reading the totality of the background submission, I used my best judgment.


            Table 2 provides a digest of what the experts conclude about whether stopping hatchery releases and fishing will significantly affect the overall wild run size.  I assigned each prediction to one of five general responses (see Appendix).

            About 57% (A and B) predicted that wild runs would end up the same (or decrease) in the absence of hatchery stocking and fishing.   Such a conclusion must be based on the presumption that hatchery releases somehow “support” the so-called “wild” run, perhaps a case of hatchery strays spawning in the wild and their offspring “replenishing” wild runs.  Regardless of the reasoning, most experts did not expect any improvement in overall wild salmon abundance.

            About 83% (A, B, and C) predicted that current (wild plus hatchery) salmon abundance (overall Columbia Basin run) would decline without hatchery stocking and fishing.

            About 43% (C, D, and E) predicted that wild runs would eventually increase, but most of these predictions were for less than the current wild plus hatchery level.  The implicit assumption here is that either fishing or hatcheries (or both) inhibit wild salmon abundance.

            About 5% (E) of the respondents predicted an overall abundance greater than the current total (wild plus hatchery) after salmon fishing was stopped and hatcheries were closed.

            Many respondents provided detailed explanations of how they arrived at their bottom line.  The debate among scientists about the causal mechanisms will surely go on for decades, but resolving that debate was beyond the scope of this exercise.  What is important here is to tease out a credible, consensus prediction that might provide policy makers with useful scientific input when addressing the unpleasant choices about the future of Columbia River Basin salmon (vs. all the other competing and popular public priorities).

            To reinforce a prior explanation about the source of these predictions, who are the experts I selected for this survey?  Almost all were practicing scientists with advanced degrees and publications (not hatchery or fisheries managers).  Although only an educated guess, based on my past professional interactions with them, I guess that most probably tend toward the “conservation” side of the salmon policy debate, rather than the “harvest” side.

            A final caveat is important to underscore:  these scientific predictions are judgments (i.e., expert opinions).  Presumably, the ones making the scientific judgments are the most knowledgeable about the topic, but these predictions involve assessing complicated scientific issues and will always involve considerable scientific controversy. 

            More broadly, and perhaps more pertinent here, such information is only one input into fisheries management and policy.  For example, how important to society is restoring runs of wild salmon to support genetic diversity priorities vs. increasing hatchery-supported runs that would sustain higher harvest rates.  Scientific input is important for settling such tradeoff questions, of course, but it is only one element driving public policy choices (i.e., the competing values and priorities that play out during the political process).


            The experts who participated in this thought experiment did not seem to believe that closing hatcheries and stopping fishing are currently key limitations to increasing the abundance of wild salmon in the Columbia River Basin.  In addition, based on many narrative comments accompanying the predictions, many participants voluntarily noted that a wild salmon recovery strategy that has any chance of restoring wild runs to anything approaching pre-1850s numbers must address several overarching and undisputed realities.  They were not arguing that society ought to change these constraining realities, just acknowledging that it is unrealistic to expect much change in wild salmon runs without doing so.

            So, according to the surveyed experts, if closing hatcheries and stopping fishing do not appear to be the key to restoring wild runs, what is?  Regardless, what are these other realities and how must they be changed to recover wild salmon to even a third of their historical level?   Let’s look at some key ones.

            Anyone even moderately familiar with the history of North American West Coast salmon is well aware of the main causes of the dire state of salmon runs along the West Coast.  These causes are well documented scientifically and include fluctuating and long-term climatic and ocean conditions, dams, mining, water pollution, habitat alteration, over-fishing, irrigation water withdrawals, predation on salmon by many species (often non-native fish species), and many other influences.

            Anywhere wild salmon were once plentiful (Europe, Asian Far East, Eastern North America), the decline in their abundance is roughly inversely proportional to the area’s growth in the human population.  Over decades and centuries, as the human population expanded in these regions, the size of salmon runs declined to minuscule levels.

            Since the mid-1800s, the West Coast (including the Columbia River Basin) is playing out similarly for wild salmon.  For example, from a pre-1850 human population level of a few hundred thousand, California, Oregon, Washington, and Idaho are now home to 54 million people.  Over the same period, wild salmon abundance in the four States has declined from roughly 50 million to a few million.  And the future?  Assuming expected human population growth in these four States, by 2100, they will be home to somewhere between 150 and 200 million people — a tripling or quadrupling by the end of this century — less than 80 years from now.

            For Oregon, Washington, and Idaho, the human population in 2100 would be roughly 40 – 55 million.  Even though many of these people will not reside within the Basin, they will directly affect the Basin’s future.  So, what about wild salmon in 2100?

            Without dramatic changes in economic policies and lifestyles, future options for restoring salmon runs to significant, sustainable levels will be greatly constrained.  Consider that by 2100, Oregon, Washington, and Idaho will have millions more people, including their additional demand for houses, roads, stores, food, coffee shops, air conditioning, drinking water, office buildings, and golf courses — the list is very long.

            To illustrate these distasteful policy tradeoffs with one well-known example, how does society choose to balance using water to grow food (e.g., irrigation) vs. using the same water to sustain wild salmon?

            Returning to the beginning of this article, not surprisingly, many fish species thrive in the highly altered Columbia River Basin.  Although unappealing to many salmon advocates, perhaps any future analysis should address what is realistically plausible?  From a fisheries management and policy perspective, it would surely be easier to maintain sustainable populations of many highly valued non-native West Coast fish species (e.g., American shad, walleye, bluegill, smallmouth bass, largemouth bass, channel catfish, brown trout, brook trout, and striped bass).  Unlike salmon, these species flourish in the even more altered Sacramento–San Joaquin river system.

            Conversely, perhaps restoring (and even maintaining) wild salmon and steelhead runs is so politically popular that nearly any cost or sacrifice is warranted.  Such questions are adjudicated in the larger political debates — the events all salmon technocrats read about daily.

            In conclusion, based on these survey results, I infer that the current overall abundance of wild salmon in the Columbia River Basin (roughly 2-4% of the pre-1850s level) is within the expected range, given the amount and accessibility of high-quality salmon habitat, past and current ecological changes, and overarching trends in oceanic and climatic conditions.

            Thus, considering the policy implications of the survey results (summarized in Table 2), eliminating hatcheries and fishing is very unlikely to increase the current overall total (hatchery and wild) run size.  In fact, it is possible (even likely according to the experts) that current wild salmon runs (wild) would decrease from their already low levels.

            Policy-wise, perhaps society is prepared to double down and support the governmentally and individually painful choice to restore wild salmon despite a warming climate, greatly altered aquatic habitats, and opposing demands for flood control, food production, electricity, and other lifestyle desires.  Such questions are only partially answered by scientific and technical experts (such as those I surveyed).  Rather, these choices primarily depend on competing, often mutually exclusive, individual and societal priorities.  These tradeoffs are wholly unpleasant politically;   thus, it should not be surprising that options that would actually recover wild salmon are rarely implemented.

* * * * * * * * * * * *


            Categorization of responses (58) to the question “What Would Happen to Wild Salmon Runs if Hatchery Stocking and Fishing Were Ended in the Columbia River Basin?” with the number (and percentage) of respondents agreeing with each prediction.

A – Predicted wild-only Basin-wide run would be less than the current wild-only Basin-wide run.   22 (38%)

  1. “Current total aggregate Basin run of salmon would drop by roughly 80% with the resulting wild-only runs collectively being less than 1% of pre-1850s runs on average.”  A
  2. “When things settle out after closing fishing and eliminating hatcheries, I expect the net effect in the Basin to be a decline in the overall abundance of wild salmon because of diverse other adverse factors.”  A
  3. “Stopping all additions of hatchery fish and eliminating all salmon fishing in the Basin, after a couple of decades, will result in a decline in natural-origin fish (i.e., wild salmon and steelhead) compared to the current overall abundance level of wild salmon.”  A
  4. “I would expect that most of the small populations in the Snake and Upper Columbia would be extinct or functionally extinct and, for the larger populations elsewhere in the Basin, I would expect that they would be at about 10% of their current abundance and declining towards extinction.”  A
  5. “If all salmon fishing and salmon hatchery production in the Columbia was terminated, I would expect salmon remnant runs to survive for a while, but over the following 30 years, climate change will ultimately make the Columbia and the associated marine environment uninhabitable for salmon.”  A
  6. “Carrying capacity is the key to answering this question because almost none of the current aquatic environments in the Basin has the same capacity as in the pre-1850s and, therefore, a few decades after closing all hatcheries and stopping all fishing, some of the fully wild runs could be extirpated, and others could be slightly larger than the wild portion of current runs.”  A
  7. “Over the following few decades, the current average total of all hatchery and wild salmon and steelhead in the Basin would be reduced 90% from current levels to approximately 1% of the pre-1850s level.”  A
  8. “Without salmon fishing and with a total closure of all salmon hatcheries in the Basin and with all other policy drivers remaining in place,  the overall total run in the Columbia Basin (wild only) would decrease and end up close to zero after a few decades.”  A
  9. “After several decades of fluctuations, after stopping fishing and closing hatcheries, the most likely aggregate Columbia run size of wild adult salmon would be approximately 10-20% of current run size (both hatchery and wild).”  A
  10. “. . . the end of hatchery production in the Columbia will lead to the extinction of naturally produced salmon.”  A
  11. “My gut response to the bottom line of the question —  with no fishing and no hatchery releases — I suspect Columbia River Basin salmon will substantially decline from current wild/hatchery levels, and some local populations would be heading towards extinction.”  A
  12. “After several decades of no fishing and no hatcheries, wild salmon and steelhead populations in the Columbia would likely drop to 80% overall of the current aggregate level of wild runs.”  A
  13. “ . . . there would be less than a quarter of the wild salmon in existence today in the Columbia Basin.”  A
  14. “Without fishing and hatchery operations, the overall abundance of wild salmon and steelhead would decrease substantially  (to about 50%) due to loss of hatchery contributions to overall abundance.”  A
  15. “With no fishing and no hatcheries, after 30 years or so, the number of wild salmon and steelhead in the Columbia Basin would be roughly the same as the current overall number of wild salmon and steelhead;  that is, it would be substantially smaller than the current total salmon run size.”  A
  16. “Even if all remaining harvest fisheries were eliminated and anadromous hatcheries were mothballed, unless the Snake River migration corridor is restored, all wild Snake River stocks (currently < 3% and < 1% of the 1950s – 1960s and pre-1850 abundances, respectively) will very likely be extirpated within 20 – 30 years.”  A
  17. “After a few decades of zero salmon fishing and zero hatchery releases, and everything else continuing along, the abundance of wild runs would end up being less than the current wild run abundance.”  A
  18. “With wild fish only, chinook, sockeye, and steelhead (and possibly coho) will likely be near extinction or gone from many, if not most areas of the basin, and wild chum and pink salmon runs may continue at or above current levels.”  A
  19. “After several decades without salmon fishing and without hatcheries, the overall net effect in the Basin would be a reduction in the current overall wild-only run size.”  A
  20. “Even with the complete closure of all salmon fishing and all releases from hatcheries, the long-term downward trend in wild salmon abundance would continue without any other changes.”  A
  21. “Even with closing all salmon fishing and eliminating all salmon hatcheries, the current downward trend in wild salmon runs would continue due to environmental degradation and global warming.”  A
  22. “After several decades of ecological adjustment, the resulting aggregate runs of wild salmon (and steelhead) would be less than current runs of the wild component of salmon/steelhead runs and overall run size would, of course, be much lower than it is currently.”  A

B – Predicted wild-only Basin-wide run would be roughly equal to the current wild-only Basin-wide run.   11 (19%)

  1. “The current abundance of salmon produced in the Columbia Basin is mostly influenced by coastal ocean carrying capacity.”  B
  2. “With the total closure of hatcheries and fishing, over the multiple decade perspective, I expect the change in wild run size to be highly variable, but the overall run size (all wild fish) not to be greatly different than current wild run size, and thus much less than the current total (wild and hatchery) run size, given that habitat does not change.”  B
  3. “With no hatchery additions and no fishing, I’m not sure that, even after several decades, there would be much change in wild salmon total numbers in the Basin.”  B
  4. “Over the 20-30 yr. period following the cessation of fishing and the closure of hatcheries, the resulting overall level of Columbia Basin wild salmon and steelhead runs would roughly be the same as current wild runs, recognizing that there is a lot of variation in how individual runs would change.”  B
  5. “After a few decades without fishing or hatchery stocking in the Columbia, the effects on wild salmon abundance, compared to current levels, will be small.”  B
  6. “We will likely not see dramatic changes in wild salmon and steelhead numbers in the Columbia River Basin even if harvest and hatchery production is eliminated, but the number of wild salmon will probably end up slightly higher than it is currently.”  B
  7. Zero change in [wild] adult return rates in the medium to long term.”  B
  8. “. . . I do not see great increases in run sizes.”  B
  9. “I wouldn’t expect a hypothetical scenario of zero harvest and zero hatcheries would lead to much, if any, improvement to the current overall abundance of wild salmonids in the Columbia Basin unless additional factors were also addressed (habitat, hydro, climate change).”  B
  10. “After a few decades of adjusting to the removal of hatchery fish and ceasing fishing (and assuming no other changes),  the resulting change in the number of wild salmon and steelhead would be small, if any.”  B
  11. “If the U.S. Government decides to immediately and permanently (1) eliminate all releases from hatcheries (i.e., close them all forever) and (2) stop all fishing directly targeting salmon and steelhead, then because Snake/Columbia River Basin salmon and steelhead face persistent constraints to recovery due to development and operation of the Columbia River hydrosysytem, we would continue to see overall rapid declines (although variable) for salmon and steelhead in the Columbia/Snake River Basin and the overall decline would be more dramatic for hatchery supplemented Snake River fall Chinook and Sockeye.”  B

C – Predicted wild-only Basin-wide run would be greater than the current wild-only run Basin-wide run, but less than the current wild plus hatchery overall run.   15 (26%)

  1. “After several decades, the abundance of wild salmon and steelhead in the Basin would be well below the abundance in the pre-1850’s, but also below current levels (wild and hatchery) of abundance.”   C
  2. “After all the runs adapt to stopping all harvest and closing all hatcheries, the aggregate run size (salmon and steelhead) would average over the long-term roughly 7% of the historical run size (assuming a pre-1850 average of 10 million).”  C
  3. “Without fishing and hatcheries, Columbia Basin salmon and steelhead would increase 10-15% in total abundance (compared to total wild fish abundance now).”  C
  4. “Cessation of harvest and hatchery production will not lead to a resurgence in the abundance of natural-origin (aka, “wild”) salmon/steelhead in the Columbia River Basin, but will result in maybe a doubling (and this could be optimistic) of the current paltry number of wild adult returns — to a number a bit less paltry.”  C
  5. “In the absence of hatcheries and fishing, wild salmon would do better than they currently are, but the resulting wild-only abundance would be somewhat less than the current wild/hatchery abundance.”  C
  6. “For the Columbia River Basin, eliminating all harvest and hatcheries will likely result in wild-only runs being somewhat higher in abundance than their current low level, thus substantially below the current hatchery-wild total run, and far below the pre-1850 total run, but also more genetically diverse, fit and resilient, thus better able to deal with and adapt to stressors such as climate change.”  C
  7. “As for total run abundance in the Basin, given current habitat conditions, I wouldn’t be surprised to see wild stocks replace much of the lost hatchery production, but there would surely be more big Chinook without an ocean fishery.”  C
  8. “Three decades after no salmon/steelhead fishing and no salmon/steelhead hatcheries, there would be substantially fewer salmon and steelhead (wild only) present in the Columbia Basin than are currently present (combined wild and hatchery).”  C
  9. Assuming that only fishing and hatcheries are stopped, but the quality of all other fresh- and saltwater habitats continue on their current trajectories, then there would be about a 10 to 20% increase in natural spawners (wild) in 30 years in the Columbia River Basin.”  C
  10. “With no fishing and no hatcheries, the resulting wild-only salmon run in the Columbia Basin would be about twice the current wild-only run, but the current wild run is only a small fraction of the present (wild/hatchery) run.”  C
  11. “After several decades without fishing and hatcheries, the overall wild fish run size in the Basin would increase, but, without hatchery salmon, this wild-fish only run would be much less than the current hatchery/wild total, and this resulting aggregate wild-only run size also would be lower than the pre-1850s overall salmon abundance in the Basin.”  C
  12. “The resulting wild run only would be 70% of the current adult [wild/hatchery] return.”   C
  13. “. . . 1 to 4 fold increase range seems like a reasonable guesstimate [in wild-only overall run size].”  C
  14. “The sustainable population size for a wild fish only scenario (no fishing) generally would be quite small.  Overall, the total wild Columbia Basin run would eventually settle out at approximately 5% of the pre-1850s run sizes.”  C
  15. “Several decades later, the total run in the Columbia Basin (wild only) would be from 5 times to 10 times the current wild fish only run overall, but less than the current total (wild and hatchery) run.”  C

D – Predicted wild-only Basin-wide run would be roughly equal to the current wild plus hatchery Basin-wide overall run.   7 (12%)

  1. “Without hatcheries and fishing, there would be about the same production as currently exists for combined wild and hatchery fish in the Columbia River Basin.”  D
  2. “Given a few decades to adjust, the wild-only aggregate abundance would increase compared to the present aggregate return of wild/hatchery salmon to the Basin.”  D
  3. “My best guess for the Basin population response is that (after stopping fishing and closing hatcheries) the total number of returning fish (composed entirely of naturally produced adults) would equal or increase slightly over the present total return (composed of mostly hatchery and some naturally produced adults).”  D
  4. “Under the hypothetical scenario of zero hatchery releases, zero salmon harvest, and everything else continuing, after a few decades, I would predict that many (but perhaps not all) salmonid populations (they would be wild only by then)  would level out to roughly 20% of the pre-1850s levels.”  D
  5. “With no fishing and no hatcheries and allowing for several decades for adjustment, the collective run size would be about 1-2 million, which is roughly what the total (hatchery and wild) is currently.”  D
  6. “Without hatchery supplementation, wild fish runs alone would be about a fifth of the historical size, approximately the current level with both wild and hatchery fish.”  D
  7. “After sufficient time for salmon and steelhead runs to adjust to no fishing and the complete elimination of hatchery operations, roughly speaking, total salmon abundance in the Basin would be largely the same, but there likely would be differences in how different species respond.”  D

E – Predicted wild-only Basin-wide run would be greater than the current wild plus hatchery Basin-wide overall run.   3 (5%)

  1. “Eventually, salmon and steelhead runs would return to about 20-30% of their pre-1850s abundance with the now totally wild runs fluctuating between about 2,000,000 to 3,500,000.”  E
  2. The remaining wild salmon and steelhead populations would rebuild to the limit of today’s Basin habitat capacity, which roughly would result in an all-species total run of 50% of historical run size.”  E
  3. The aggregate run size (wild fish only) in the Basin would rebound to approximately 30% of the pre-1850s run size.”  E


Robert T. Lackey

Many of today’s ecological policy issues are politically contentious, socially wrenching, and replete with scientific uncertainty.  They are often described as wicked, messy policy problems (e.g., reversing the decline of salmon;  deciding on the proper role of wildfire on public lands;  what to do, if anything, about climate change;  worries about the consequences of declining biological diversity;  making sense about the confusing policy choices surrounding notions of sustainability).

Wicked, messy ecological policy problems share several qualities:  (1) complexity —  innumerable options and trade-offs;  (2) polarization — clashes between competing values;  (3) winners and losers — for each policy choice, some will clearly benefit, some will be harmed, and the consequences for others is uncertain;  (4) delayed consequences — no immediate “fix” and the benefits, if any, of painful concessions will often not be evident for decades;  (5) decision distortion — advocates often appeal to strongly held values and distort or hide the real policy choices and their consequences;  (6) national vs. regional conflict — national (or international) priorities often differ substantially from those at the local or regional level;  and (7) ambiguous role for science — science is often not pivotal in evaluating policy options, but science often ends up serving inappropriately as a surrogate for debates over values and preferences.

As if they are not messy enough, ecological policy issues may become further clouded by skepticism about the independence of scientists and scientific information.  Much of the available science is tendered by government agencies, companies and corporations, and public and private organizations, as well as myriad public and private interest and advocacy groups.  Each arguably has a vested interest in the outcome of the debate and often promulgates “science” that supports its favored position.

All ecological policy problems have unique features, thus there are exceptions to every generality, but are there lessons learned that can be broadly applied?  Like all axioms, mine are not universally true, but are applicable in most situations.

  • Ecological Policy Axiom 1 — The policy and political dynamic is a zero-sum game

Probably the most sobering reality for the uninitiated is that selecting any proposed policy choice results in winners and losers.  The search for a “win-win” choice, which sounds so tantalizing to decision makers, is hopeless with even superficial policy analysis.  There are always winners and losers even though people running for office may try to convince the voters otherwise.  This axiom is why policy making is sometimes described as “the political process of picking winners and losers.”

  • Ecological Policy Axiom 2 — The distribution of benefits and costs is more important than the ratio of total benefits to total costs

Benefits are the consequences of a policy option or decision that are categorized as good outcomes.  Benefits are sometimes measured solely in terms of money, but are more broadly encompassed by all the desirable things that are most likely to happen.  Conversely, the costs are the undesirable outcomes that are likely to happen (often, but not always, measured in monetary terms).

  • Ecological Policy Axiom 3 — The most politically viable policy choice spreads the benefits to a broad majority with the costs limited to a narrow minority of the population

Democracies operate on delegated compromise validated by periodic voting.  To gain sufficient political support (votes) for a proposed policy, it is prudent for the decision maker to spread the benefits across a sufficiently large number of people to garner majority support.  The corollary is that those (including future generations) who bear the costs should be a minority and the smaller the better.

  • Ecological Policy Axiom 4 — Potential losers are usually more assertive and vocal than potential winners and are, therefore, disproportionately important in decision making

With many ecological policy questions, those who bear the costs, the losers, have a disproportionately greater influence on the decision making process.  While policy analysis tends to evaluate the rationality of competing policy arguments, the political process tends to weigh breath and vigor in support of each competing policy option.  Issues of perceived fairness are important in the political process, but difficult to quantify in policy analysis.

  • Ecological Policy Axiom 5 — Many advocates will cloak their arguments as science to mask their personal policy preferences

Technocrats, as I apply the label, are individuals with scientific training who are responsible for implementing law or ecological policy.  There is an understandable impulse by technocrats to insert what they think is or should be the appropriate public policy goal or option.  For example, should ecological restoration be aimed at recreating the ecological condition that existed at the beginning of the Holocene, just prior to 1492, or at the end of last week?  The answer requires making a value judgment — a policy choice that is necessarily a political judgment — and it is not a scientifically derived decision.  Ecologists and other scientists should assess the feasibility and ecological consequences of achieving each possible restoration target.  Selecting from among the choices, however, is a societal enterprise.

  • Ecological Policy Axiom 6 — Even with complete and accurate scientific information, most policy issues remain divisive

The lament that “if we just had some better science, we could resolve this policy question” is common among both scientists and decision makers.  Calls for more research are ubiquitous in ecological policy debates.  In most policy cases, even if we had complete scientific knowledge about all aspects of an issue, the same rancorous debate would emerge.  Root policy differences are invariably over values and preferences, not science and facts.

  • Ecological Policy Axiom 7 — Demonizing policy advocates supporting competing policy options is often more effective than presenting rigorous analytical arguments

Scientists and policy analysts become frustrated when they fail to recognize that political debates are partly logical argument and partly image.  Negative images are often considered more effective in swaying people than positive ones.  In fractious ecological policy debates, proponents often spend more energy demonizing their opponents than sticking to rational policy analysis.  My experience is that such tactics are often effective in policy debates;  many people are moved by negative arguments.

  • Ecological Policy Axiom 8 — If something can be measured accurately and with confidence, it is probably not particularly relevant in decision making

In my experience, most scientists prefer to talk about things that they can measure with some degree of confidence.  Fish population abundance, recruitment rates, optimal habitat, toxicity levels, and field surveys are within our comfort zone.  We can put confidence limits on these numbers;  we can duplicate the data gathering year after year;  we can often forecast future conditions with some degree of confidence.

  • Ecological Policy Axiom 9 — The meaning of words matters greatly and arguments over their precise meaning are often surrogates for debates over values

In my experience, many citizens get frustrated in ecological policy debates because the advocates of various competing choices often seem to argue over semantic nuances rather than getting on with making decisions.  The precise meaning ascribed to key words is important and is often the battleground over what policy option is ultimately selected.  The debate over definitions is really a policy debate.  How should pivotal words such as “ecosystem health,” “sustainability,” “degraded,” “biological integrity,” “endangered,” “wild,” and “impaired” be defined?  Definitions chosen will lead (at least in the mind of the uninformed) to a particular policy option.  Thus, the debate over what might appear to be semantic nuances is really a surrogate debate over values and policy preferences.


Many of today’s ecological policy issues are contentious, socially divisive, and full of conundrums.  They are, however, typical of those that professional natural resource and environmental scientists will confront, both now and for the foreseeable future.  Those of us who provide information to help inform the participants involved in ecological policy debates need to be cognizant of and appreciate the importance of scientific information, but we also must recognize the reality that scientific information is just one element in complex political deliberations in a democracy.



Robert T. Lackey

More than two decades ago, while Deputy Director of EPA’s national research laboratory in Corvallis, Oregon, I presented a talk to a group of community activists about why salmon populations along the West Coast have dropped to less than 5% of their historical levels.  I’ve given such talks many times so I was confident that I had heard just about every question that might be asked.  I was wrong.

The opening question was asked by a well-known political activist.  He was direct, pointed, and bursting with hostility:  “You scientists always talk about our choices, but when will you finally tell us what we SHOULD do about the dramatic decline of West Coast salmon?  Quit talking about the science and your research and tell us what we should do!  Let’s get on with it!”

From the nods of approval offered by many in the audience, his impatience with science and scientists was broadly shared.

What does the public expect from scientists regarding today’s ecological policy issues? Some examples of such policy challenges include the decline of salmon;  deciding on the proper role of wildfire on public lands;  what to do, if anything, about climate change;  the consequences of declining biological diversity;  and making sense of the confusing policy choices surrounding “sustainability.”

The lament “if we just had some better science, a little more data, we could resolve this policy question” is common among both scientists and decision makers.  Calls for more research are everywhere in ecological policy debates.

In most cases, even if we had complete scientific knowledge about all aspects of an issue, the same rancorous debate would emerge.  Root policy differences are invariably over values and preferences, not science, data, and facts.

In a pluralistic society, with a wide array of values and preferences competing for dominance, the ecological policy debate is usually centered around whose values and preferences will carry the day rather than over scientific information.

So what was my answer to the emotionally charged question from the political activist?

It was: “Science, although an important part of policy debates, remains but one element, and often a minor one, in the decision-making process.  We scientists can assess the ecological consequences of various policy options, but in the end, it is up to society to prioritize those options and make their choices accordingly.”

He wasn’t pleased.



Robert T. Lackey

In science, when you see the words “natural,” “healthy,” “degraded,” and “biological integrity,”  all these terms, and many others, have embedded assumptions about what someone or some organization regards as a desirable value choice, a preferred policy choice.

These and similar words have no place in science.  They are classic examples of normative science.  Their use in scientific publications is simply policy advocacy disguised as science.

The words are fine for management, expected in policy advocacy, but not OK in science.

Here is a test:  first, put on your science hat.  Now imagine that the public owns a 5,000-acre stand of old-growth (never logged) forest which is being considered by a government agency for an alternative use.  Scientifically, is it preferable to (1) preserve this landscape as is, or (2) remove the trees and build a wind farm?

Neither ecological state is preferable scientifically!   At least not without assuming, perhaps unwittingly, a policy preference, a value choice.  If the science relevant to this policy question is presented in such a way to subtly favor either policy option, it is a classic example of normative science.

It may look like a scientific statement.  It may sound like a scientific statement.  It is often presented by people who we assume to be operating as scientists.   But, such statements in science are nothing more than “policy advocacy masquerading as science.”

Anyone following basic scientific principles should say:

            “Pristine ecosystems are neither superior, nor inferior, to human-altered ones.  Different, for sure, but not better or worse.”

Let me wrap up by offering Charles Darwin’s advice to scientists.  Remember that he was under a lot of social pressure to make his scientific findings conform to the dominant political and theological views of the time.  He offered pithy guidance to scientists:

          “A scientific man ought to have no wishes, no affections, a mere heart of stone.“

Strict, uncompromising, and unequivocal advice, but spot-on for scientists both then and now.



Robert T. Lackey

Several years ago, toward the end of my career with the U.S. Environmental Protection Agency, one of the “outreach” staff in Washington, DC, telephoned me in Corvallis and posed a question:  “In science, why isn’t the notion of ecosystem health a useful metaphor to convey scientific information?  I see the metaphor used all the time.”

When I provided a long-winded, complicated, technically rigorous answer, the staffer was obviously disappointed. Evidently, he was looking for something simple and clear-cut.  I then asked, ever so diplomatically, to provide a brief written answer that could be published in the EPA internal blog.  What follows is an edited version of what was submitted.

Very young children have a habit of asking innocent, but thorny questions.   My grandson, however, has reached an age where innocence no longer passes for an excuse for his questions;   he knows enough now that his questions reflect the traits of a budding intellectual troublemaker.

A case in point:  here is my answer to his question about the increasingly popular term:  ecosystem health.

          “Grandpa, in school today in my science class, we talked about healthy ecosystems. My teacher says that when we are not feeling well, we go to a doctor to find out how to get healthy. If I have a sick ecosystem, she says that I should go to a scientist to find out how to make the ecosystem healthy. Dad says you are a scientist, so what is a healthy ecosystem?”

It is a good question and one that I, as a research scientist who has worked on such issues for over 40 years, should be able to answer with ease.

This seemingly straightforward question, however, does not have a simple answer. Further, the answer requires a clear understanding of the proper role of science in a democracy.

First, how is a person to recognize a healthy ecosystem?  Many might identify the healthiest ecosystems as those that are pristine. But what is the pristine state of an ecosystem? Is it the condition of North America prior to alterations caused by European immigrants, say 1491?  Or perhaps it is the condition of the land sometime well after the arrival of immigrants who came by way of the Bering land bridge, say 1,000 years ago? Or maybe it is the state of North America prior to the arrival of any humans, say more than 15,000 years ago?

Ultimately it is a policy decision that will specify the desired state of an ecosystem. It is a choice, a preference, a goal.

Scientists can provide options, alternatives, and possibilities, but ultimately in a democracy, it is society that chooses from among the possible goals.

For example, a malarial-infested swamp in its natural state could be defined as a healthy ecosystem, as could the same land converted to an intensively managed rice paddy.  Neither the swamp nor the rice paddy can be seen as a “healthy” ecosystem except through the lens of a person’s values or policy goals.

Once the desired state of an ecosystem is specified by someone, or by society overall through laws and regulation, scientists can determine how close we are to achieving that goal. They might even offer some approaches that might better achieve the goal.  Ultimately, though, it is society that defines the goal, not scientists. One person’s sick ecosystem is another person’s healthy ecosystem.

So, the answer to my grandson’s provocative question is that human health is not an appropriate metaphor for ecosystem health.  There is no inherently “healthy” state of ecosystems except when viewed from the perspective of societal values.

Pristine ecosystems (e.g., wilderness watersheds, Antarctica, uninhabited tundra) are certainly very different than highly altered ecosystems (e.g., farms, city parks, harbors) but neither a pristine ecosystem nor a highly altered ecosystem is scientifically better or worse — just different.