The Fish We Never Killed, According to MRIP

At the end of August, NOAA republished 43 years of Marine Recreational Information Program (MRIP) data, the only quantitative tool ASMFC managers have to estimate recreational fishing effort and catch. Across the entire MRIP time series, 1982-2025, the revision cut recreational trips by 41%, harvest by 43%, and releases by 43%. 

This is huge, so settle in because the MRIP revision has far reaching consequences.  Some important numbers change and some crucial trends don’t. Here’s the short version:

Recreational trips, harvest, and releases were less than estimated. NOAA warned that 40% of the striped bass trips, harvest, and releases could disappear from ASMFC’s tables and charts. For managers, that warning is now a reality. 

It does not mean there are more fish in the sea. The stock assessment estimates the size of the population from the size of the catch. Cut the catch and the population estimate falls with it. Simply put, the fish we never caught never existed.

The fishery is still collapsing. Trips, harvest, releases, and the number of fish caught per trip are falling in the old MRIP numbers and in the new ones. Catch per trip, the most important indicator that the revision barely changed, is the lowest since 1984.

What MRIP is, and what NOAA changed

FES x APAIS = MRIP
Only FES changed because it overcounted trips, mostly in spring and fall

It can be intricate, but compared with most fisheries data, MRIP is relatively simple. MRIP has two parts. First, the Fishing Effort Survey (FES) asks and estimates how many trips shore and private-boat anglers take. Second, the Access Point Angler Intercept Survey (APAIS) interviews anglers and estimates what recreational anglers catch on those trips. Think of APAIS like an exit poll, which is conducted right after the angler returns from a trip. Combine the number of trips from FES with the corresponding catch from APAIS, and you get the coast-wide MRIP database.

NOAA recalibrated MRIP, because the FES survey overestimated recreational anglers’ trips. In 2024, NOAA ran the old FES survey alongside a newly redesigned version in 17 coastal states. The new survey changed the order of the questions.  Instead of asking how often you fished during a confusing two-month “wave” period, it asked how often you fished in two separate questions for each month in that “wave.” In the summer, the old and new surveys found nearly the same number of private-boat trips. In November and December, and again in March and April, the redesigned survey revealed the error: a large proportion of respondents who, in the old survey, claimed they took an early- or late-season trip now more accurately reported when that trip took place. 

We’re all guilty of misremembering dates. That’s why so many anglers rely on our logs and GotOne. Some years, my favorite tide falls in the last week of October; other years, it lines up in the first week of November. Sometimes the migratory fish arrive on the last moon of April; in other years, like 2026, that moon was the first of three moons in May. That’s why the FES correction and this MRIP calibration affect trip numbers during the peak months of the striper migration from NY to NC.  Waves 2 and 6 are the migration. Together they carry 36% of striped bass effort, and each was cut by 44%. A July fishery barely moved; a spring and fall fishery was cut nearly in half.

The next pattern we see also isn’t rocket science. The FES survey began in 2018. Estimates before 2018 came from the good old days of landline phone surveys. To keep one continuous record, NOAA had to run another calibration to translate those phone-survey estimates into mail-survey estimates. Now that the mail survey itself has been corrected, the phone-to-mail conversion had to be redone. That is why, as we’ll see, 1999 trips come down 59% and 2025 trips only come down 17%.  

You will hear that this proves MRIP was always garbage. If you don’t own a for-hire operation or commercial tags, this revision proves the opposite. The CIE independent peer review called NOAA’s calibration model “state of the art” and said it “fully addresses the complexity of the survey data.”^[i]  A survey that can be tested against a controlled comparison, corrected, and reviewed does its job. Every fishery data source has errors. An observer on a for-hire boat can’t see the entire deck. Neither can the for-hire captain, and no one is looking over their shoulder when they fill out their eVTR trip reports. Commercial landings are self-reported by commercial fisherman and their dealers, and they are notoriously unreliable.

Hate it or love it, MRIP isn’t good, bad, or evil. Like the bombshell this revision is, it doesn’t care about your feelings, its management implications, or what happens to ASMFC’s flagship species. MRIP isn’t perfect. It’s simply a sharper tool. This is exactly how the scientific method works: the tools and methodology get better at capturing reality.

What the revision shows

4 of every 10 trips since 1982 never happened.
49.9M fish were never harvested and 421.3M fish were never released

Our trips were overstated by more than 50% in the 1990s and early 2000s, by more than 40% through 2013, and by 20 to 30% over the last decade. More than half of the trip correction, 152.9M of 272.4M trips, falls in the fifteen years from 1995 through 2009, the years everyone points to as the good old days. 

The old series told a story of explosive growth from the moratorium into the mid-2000s, then a collapse. The revised series is compressed. It still climbs through 90s and 2000s, then trips plateau, but the revision barely softens the last decade. In 2025 anglers harvested 1.14 million striped bass, the fewest since the stock was declared rebuilt in 1995. We released 9.24 million striped bass, the fewest since 1994. In 2017, we released 28.35 million fish, three times as many. Since 2017, harvest is down 43% and releases are down 67%.

At the state level, the cut did not fall evenly across states, because states don’t concentrate effort in the same months. Virginia lost 60% of its trips and 71% of its harvest. Connecticut lost 23% and 22%, respectively. That is a 37-point spread between states on the same fish, in the same years, from the same survey redesign.

We were told the revision is a change in scale, not in trend. At the coastwide level, that’s mostly true. The coastwide series has a clean shape: the cut is deep in the 1990s and shallow in the last five years. That pattern does not survive state disaggregation. Massachusetts follows that coastwide pattern with a 48% cut over the full period, but only 8% from 2021 through 2025. Meanwhile, New York’s directed trips fall 29% over the long run, and 29% in the recent window, and its harvest and releases are cut harder in recent years than historically. 

For some states, the revision is a historical correction that fades towards the present. For others, it’s current. Others, like NJ, are stuck somewhere in between.  Thank the striper gods we don’t have to figure out how this mess would impact conservation equivalency. 

What it means for the stock

The fish we didn’t catch didn’t exist.
MRIP doesn’t spawn or create more fish

A survey can’t put fish in the ocean. The fish they thought we killed on the trips we never took only existed in the ASMFC’s Technical Committee (TC) computers. During the 2023 Emergency Action debates, when special interests implied this revision would mean more fish for them to harvest, TC members produced this oversimplified equation to explain how catch (F) scales the size of the stock. ⁽²⁾

The stock assessment treats the catch as known. It estimates the population that could have produced that catch while matching the trends in the survey indices. 

ASMFC has been through this once already. In 2018, NOAA’s first FES recalibration—from the telephone survey to the mail survey—raised the historical number of trips and therefor of catch. The benchmark assessment that year raised the estimate of the stock. The Commission’s own summary of that assessment explained why: “Indices of abundance track relative changes in the population over time while catch data provide information on the scale of the population size.” ⁽³⁾ Catch went up, so the stock went up. Catch is now going down. 

If we catch fewer fish, there are fewer fish in the ocean to catch.  If we catch more fish, there are more fish in the ocean to catch. It is really that simple.  Many stakeholders and even some managers don’t understand this basic stock assessment principle, but we hope it’s clear to you. 

The trends don’t change

Every measure of this fishery is falling: trips, fish harvested, fish released, and most importantly CPUE.

Every indicator of the quality of the striped bass fishery points in the same direction before and after the revision. Look at Figures 2 and 3 again and focus on the trend. Harvest peaks, releases peak, and then they fall together through the last decade. The revision moved the lines. It doesn’t change where they’re headed.

The rows that matter most in Table 3 are the share of catch released and fish caught per directed trip, AKA Catch Per Unit Effort (CPUE).  The FES revision barely changed the share of releases and CPUE, because those numbers come from APAIS. 

Back to the moratorium, through every regulatory change along the way, nine out of every ten striped bass caught by a recreational angler have been released. The evidence that striped bass had always been a catch-and-release fishery remains unchanged by the revision. That MRIP data still disproves the claim that restrictive size limits are why effort, harvest, and releases are all nosediving. The slot did not create the release rate; the release rate has been near nine in ten throughout the modern series, before and after this update.

On the other hand, CPUE fluctuates with abundance. When there are more striped bass, we catch more of them. In 2017, anglers averaged 2.48 striped bass per directed trip. In 2025. they averaged 0.89 per directed trip.  That 64% drop in eight years is striped bass abundance crashing and anglers only taking 5% fewer trips. It gets worse: .89 fish per trip is the lowest CPUE since the depths of the moratorium in 1984. Let’s repeat that, with all of today’s easily accessible information, modern technology, and tackle that would’ve been unrecognizable to the old timers, we’re catching fewer fish per trip than the last year before the harvest moratorium. 

We’ll leave it up to you to let that sink in. How many stripers do you think you could catch per trip if you only had what was available in 1984?  Most of us are too young to know what the tackle, gear, or compass bearings and RPMs would be.

Look out for Part II and Part III on how this MRIP revision and the corrected release mortality rate will reframe total removals and every striped bass management debate.

---

⁽¹⁾ NOAA Fisheries MRIP, 2026 Transition Plan to Implement the Improved Fishing Effort Survey (p. 4) https://www.fisheries.noaa.gov/s3/2026-01/FESTransitionPlan508.pdf

⁽²⁾ ASMFC. Review of the Interstate Fishery Management Plan for Atlantic Striped Bass, 2025 Fishing Year. Approved August 6, 2026. Section II, Status of the Stocks, p. 8.
https://asmfc.org/wp-content/uploads/2026/08/AtlStripedBassFMPReview_FY2025.pdf