Thursday, July 23, 2020

NMFS SHEDS SOME LIGHT ON COVID-RELATED DATA ISSUES



At the time, I expressed real concern that managers would have little or no 2020 data to guide their fisheries decisions, and that management, and fish stocks, could suffer as a result.

We now have a little more information on how the National Marine Fisheries Service hopes to move forward.  While the situation still is not good, it appears that we will not be facing as worst case scenario, and NMFS staff is working to find a rational approach to what will still be a difficult season.


Dr. Cody and Ms. Kelly combined to provide a very detailed and very candid description of the challenges that NMFS has faced, and will continue to face, this year in collecting recreational fishing data, filling in the gaps where data is missing, and crafting recreational regulations that meet the legal standards for avoiding overfishing and rebuilding overfished stocks, without unnecessarily burdening anglers.

Dr. Cody began the webinar, briefly describing the various surveys included in MRIP, and explaining that in New England and the Mid-Atlantic, COVID’s biggest impact is on the so-called Access Point Angler Intercept Survey (APAIS), in which surveyors conduct in-person interviews with anglers, ask a number of questions about their fishing day, and actually count and measure the fish that anglers harvest.

The COVID impacts have been about what one might expect.  State rules about “phased reopening” determined whether the APAIS survey could be conducted at all.  In places where surveying was still possible, each state provided its own protocols with respect to how it could be conducted.  As all state reopening plans were and are subject to change as the rate of infection ebbs and flows, the APAIS process remains somewhat uncertain and subject to local interruption.

As a result, catch sampling has been incomplete, and the process remains challenging.  COVID’s first impacts manifested themselves in late March and early April, the heart of “Wave 2;” sampling was suspended for a part of late April.  As a result, no catch estimates for Wave 2 will be released.

Sampling coverage was also limited in Wave 3 (May-June) and will inevitably be limited in Wave 4 (July-August) as well; so far, for all of this year, Dr. Cody described both the geographical and temporal coverage as “spotty.”  “Social distancing” also makes it much more difficult for surveyors to interview anglers and check their catch, so there will be fewer interviews conducted than there would be in a typical year.  

As a result, NMFS will have less, and less reliable, data with respect to catch per unit effort (how many fish anglers catch in a given amount of time, which is generally a good indication of relative abundance), the species that compose anglers’ catch, and the size of the fish being landed.

Because of such incomplete sampling, Dr. Cody said that there was

“the potential for some serious bias”
in the APAIS data.  

That is troubling, because the recreational catch rate information provided by APAIS is a key input used to calculate recreational catch and landings.  Since recreational catch and landings estimates are then used in biologists’ estimates of stock health, biased APAIS data has the potential to create real problems for managers.

The good news is that, although the APAIS data is going to be iffy this year, the Fishing Effort Survey, which is conducted by mail, along with the For-Hire Telephone Survey, have been largely unaffected by COVID so far.  That could change, if the contractor conducting the surveys is impacted by the virus and is unable to perform at current levels, but to date, the effort surveys seem to be on track.

That’s important, because with many people unemployed or working from home, fishing effort seems to have increased substantially.  There has been a lot of anecdotal comments to that effect, and if those comments are, in fact, true, they could foreshadow a big increase in recreational harvest. 

Once NMFS has the effort figures in hand, it will be better able to determine whether or not such an increase has actually occurred.  However, it will still be very difficult to come up with an accurate estimate of 2020 recreational landings.

Dr. Cody noted that the approaches to deal with the missing data are “very limited,” and that there will be “significant data gaps.”  NMFS is looking at two basic options:  It can create a model that addresses known gaps in the data, but creating such model will take a lot of time.  Any such model probably wouldn’t be completed by the end of the year, and would still require NMFS to make some assumptions about recreational catch and effort.

NMFS’ other option is to use “imputation,” which means taking data from other years and using it to plug holes in the information it has for 2020.  Such imputation has been used before to fill gaps in fisheries data, but never on the scale that would be required this year; if NMFS used imputed data and fishermen didn’t like the results (i.e., it resulted in additional restrictions on landings), there is a very good chance that the validity of such data would be challenged.

It’s possible that NMFS may also use what Dr. Cody called “auxiliary data sets,” such as the vessel trip reports (VTRs) filed by for-hire vessels or other sources of recreational fishing information, to color its decisions.  Such data sets aren’t perfect, but Dr. Cody felt comfortable that VTRs landings information, for example, could be somewhat ground-truthed by comparing them to APAIS data, although the limited number of interviews will limit the efficacy of that approach. 

However, absent video in a vessel’s cockpit, or an observer on board, it is impossible to know for certain whether VTR information on the number of fish released, and whether they were returned to the water alive, dead or dying.  He noted that the question of accurate VTR information has become “a growing issue” as regulations become more restrictive.

But in the end, both NMFS and the Atlantic States Marine Fisheries Commission will need some sort of estimates to guide them is setting 2020 regulations. 

Ms. Kelly took over the webinar at that point to describe how the regulation-setting process should work.

She explained that the regulatory process has three basic steps.  There is data gathering and compilation, there is the ASMFC and the regional fishery management councils deciding what to do with such data, and there is the rulemaking process itself.

The data and collection process involves all of the issues that Dr. Cody described, including potentially biased APAIS data, imputation, the use of auxiliary information, etc.  Scientists must decide, for each managed species, whether they can safely assume that past catch rates might also be used for 2020 harvest estimates, or whether intervening factors make such assumption untenable.  

Ms. Kelly noted that many species are managed with multi-year specifications, which eases the data-crunching burden somewhat.  But despite such multi-year programs, NMFS still has a legal obligation to ensure that overfishing does not occur.

At some point, the scientists’ estimates and advice will be handed over to the ASMFC and the regional fishery management councils.  It is up to those bodies to decide whether, and how, management measures should be adjusted to account for the 2020 data.  

Should the annual catch limit be adjusted downward to account for management uncertainty?  If states each have a separate recreational quota, does the uncertainty in the data impact each state equally, or do some potentially face in inequitable result?  Should management bodies adopt more precautionary measures in the case of overfished stocks or those that have been subject to overfishing?

In the case of ASMFC-managed fisheries, where managers need not worry about meeting legal standards for conservation and management, the process can end there, with each affected state adopting regulations that comply with the ASMFC’s decision.  But in the case of federally-managed fisheries, where the Magnuson-Stevens Fishery Conservation and Management Act applies, the rulemaking process is significantly more rigorous.

In those cases, federal managers must first determine the minimum level of data that is necessary to adequately quantify the risk of overfishing, for in all cases, whatever regulations are adopted must have at least a 50 percent probability of preventing overfishing from occurring.  Then the NMFS must be able to justify, as part of its regulatory package, why the proposed management measures meet that and other legal standards, and put its proposed measures out for public comment.  Only after the public comments have been addressed may the NMFS issue a final rule.

Which all means that fisheries managers have a difficult road ahead of them.  Thanks to the impacts of COVID-19, they will be forced to conserve and manage the nation’s fisheries resources without having all the needed information on hand.

Such situation calls out for additional precaution on the part of fisheries managers, especially in the case of overfished stocks such as striped bass and bluefish, to assure that, by failing to adopt sufficiently restrictive regulations in the face of what appears to be increased fishing effort, they make the situation worse than it is today.

But it’s clear from Dr. Cody’s and Ms. Kelly’s comments that the NMFS is working hard to develop the best information available, given the trying circumstances.

At the worst, the data situation looks somewhat better today than it did two or three months ago.

No comments:

Post a Comment