A red snapper that disappears below the boat may survive, die later or become prey. Watching it swim away answers only the first few seconds of a longer question. That is why release-method advice needs more than a video of a fish going down.
Our conclusion: prepare a suitable descending device before fishing for red snapper at depth. For barotrauma-affected, lip-hooked fish captured at 30–50 metres in the eastern Gulf study examined here, release at 20–30 metres had stronger relative tag-return evidence than venting at the surface. A 10-metre descent did not show the same benefit. Other Gulf telemetry evidence found similar survival with venting and descending, so this is a conditional recommendation, not proof of universal superiority or a Louisiana survival guarantee.
Source depths are in metres: 10 m is about 33 ft, 20 m about 66 ft, 30 m about 98 ft and 50 m about 164 ft. These are measured study depths, not universal device settings.
DR06 · PUBLISHED RESEARCH CASE STUDY · v1.0.0
A prepared descender has the stronger depth-specific case
Three primary papers distinguish relative tag returns, telemetry-inferred survival and a modeled fishery scenario. These are separate outcomes, rather than one pooled survival score.
Vented · Florida Gulf
6.4%reported tag returns
Descended 20 m · same study
16.5%reported tag returns
Absolute survival gain
Unresolvedfor Louisiana trips
Conclusion: for red snapper showing barotrauma after capture at 30–50 m, the Florida Gulf evidence favors descent to 20–30 m over venting at the surface. A 10 m descent did not show that advantage. Prepare a suitable descender before fishing; the evidence does not establish that descending always beats skilled venting, at every capture depth, season or location.
Original figure from Stallings et al. (2023), Results p. 5. Fish were lip-hooked without hook injury and captured at 30–50 m off western Florida during July–October trips in 2014/2016. No new error bars are inferred from rounded rates.
Stallings et al. · sample sizes and rates as separately published
Method
Table 1 n
Reported returns
Vented at surface
344
6.4%
Descended 10 m
262
6.9%
Descended 20 m
202
16.5%
Descended 30 m
175
15.5%
Source discrepancy: 16.5% with n=202 and 15.5% with n=175 do not reconcile to integer returned-fish counts at the displayed precision. We retain Table 1’s sample sizes and the Results section’s rates separately. We do not invent counts, recalculate confidence limits or silently correct the paper. This prevents an independent count-level reproduction of its comparison.
Inspect the published odds ratios and our limited arithmetic
Source-reported odds of a tag return versus venting · not survival ratios
Descent
Odds ratio
95% CI
P
10 m
1.08
0.53–2.16
0.87
20 m
2.88
1.56–5.41
0.0004
30 m
2.68
1.42–5.12
0.001
The paper reports higher return odds at 20 m and 30 m; the 10 m interval includes 1. Its 20 m versus 30 m comparison did not detect a difference (P=0.89), which does not prove equivalence. The 40 m arm contained only three red snapper and was excluded from comparisons.
Our descriptive arithmetic on the rounded rates gives a +10.1 percentage-point difference at 20 m and +9.1 points at 30 m versus venting. The corresponding rate ratios are approximately 2.58 and 2.42. Those are ratios of reported tag-return percentages; they are neither the published odds ratios nor a measured multiplication of survival. The original tests and intervals are attributed, not independently reproduced or newly adjusted for multiple comparisons.
A second Gulf case checks what happened after fish left the surface
Curtis et al. (2015) tracked fish near Texas platforms using depth and acceleration sensors. The study’s overall known-fate sample included 62 survivors, 13 surface deaths and 11 delayed deaths among 86 fish, with another 25 of 111 tagged fish unclassified. All detected delayed deaths occurred within three days. That total includes controls and several treatments; it is not a descending-method success rate.
Original count-based figure from Curtis et al. (2015), Table 1, p. 440. The denominator here is all tagged fish, keeping unknown fates visible. The paper’s survival percentages instead exclude unknowns. A missing venting arm in winter is not a zero-survival result.Inspect known-fate survival and missing-fate sensitivity
Bounds put every unknown fish first among deaths, then among survivors. These are extreme missing-fate scenarios, not confidence intervals. They show why dropping unknowns can make tiny cohorts appear more decisive than the records support. We do not infer statistical superiority from a higher bar or combine different seasons/depths into a new treatment effect.
Fish were descended to the bottom, rather than assigned the 10–30 m release treatments of the Florida study. Surface releases were initially inside a protective open-bottom cage. Venting was added during the summer trial and absent in winter; controls had recovered in surface-pressure tanks. Those controls are not freshly caught untreated fish. These design differences matter when comparing the papers.
Why “8% mortality” is a best-case estimate, not a Louisiana promise
Runde et al. (2021) reported 8% mortality (95% CI 0–17%) among 36 eligible red snapper successfully descended once with no deep hooking, at about 37 m off North Carolina. The device was set to open at 31 m. Their Kaplan–Meier estimate accounts for censoring; it is not a simple mortality fraction for all 44 live-tagged fish. Self-descending, repeated-descent and deeply hooked fish were excluded from that baseline.
A separate simulation estimated median mortality of 13%, with 2.5th–97.5th percentiles of 10–17%, under a fishery-wide scenario of universal descending and no device failures, incorporating hooking evidence. Its percentile range is not the empirical cohort confidence interval. There was no venting comparison in this experiment. Neither result supplies a Gulf-wide effect or a Louisiana survival percentage.
Save evidence.json and reproduce.mjs together; run node reproduce.mjs with Node 24. No packages or internet are required. This regenerates the JSON and four CSVs from attributed aggregate facts, including missing-fate bounds. It does not refit tag-return tests, telemetry models or the South Atlantic simulation. Source snapshots and original plotting code are retained in the repository.
01The question we can actually answer
The useful question is: under which studied conditions did descending improve the measured release outcome compared with venting? A descending device returns a fish under weight to greater pressure before detaching. Venting uses a purpose-made tool to release expanded gas. Both address pressure-related effects of capture; neither repairs every hooking injury, erases handling stress or prevents all predation.
Three endpoints need separate interpretations. Conventional tag returns require a fish to survive, be caught again and have its tag reported. Acoustic telemetry can infer survival or death from movement and depth over time, subject to detection and fate-classification assumptions. Video can document recovery of swimming behavior. These observations are useful, but they are not interchangeable survival percentages.
Our case review uses full primary papers by Stallings and colleagues (2023), Curtis and colleagues (2015) and Runde and colleagues (2021). We checked the principal numerical tables visually and retained source snapshots. The original contribution here is an evidence comparison, original figures and reproducible arithmetic on attributed aggregate facts. We did not collect fish, acquire raw telemetry records or conduct a comprehensive systematic review. No single pooled mortality estimate is justified by these three designs.
02Case one: release depth changed the comparison
The Florida study caught red snapper at 30–50 metres on charter-vessel research trips in the eastern Gulf. Only lip-hooked fish without hooking injury entered the comparison. Fish with visible barotrauma were vented and released at the surface or recompressed to assigned depths. Fishing occurred in July–October, during warm-water conditions, on trips in 2014 and 2016.
The initial treatment assignment was random, followed by an attempted rotating sequence. Device availability sometimes changed assignment: if tools were occupied returning other fish, researchers vented a fish rather than hold it longer. This practical design kept air exposure down but was not a perfectly balanced randomized trial. The authors also pooled two recompression tools after comparing their return rates. Stallings et al. (2023), methods and results.
At 10 metres, reported returns were close to the vented group. At 20 and 30 metres, reported return rates were substantially higher and the published comparisons supported higher odds of return. This supports paying attention to release depth, rather than treating any trip below the surface as the same intervention. It does not establish that every fish caught at any depth should be released at one fixed setting. The authors explicitly limit extension beyond their 50-metre maximum capture depth.
The rate/sample-size discrepancy described above limits independent numerical reproduction. We show it because an evidence-based page should disclose a problem in the source rather than manufacture precision. The published comparisons remain the authors’ results. Our arithmetic illustrates the rounded reported rates and adds no new count-level statistical test.
03Case two: telemetry makes delayed mortality visible
The Texas platform study adds a different view. Fish carried sensors recording depth and acceleration after release, allowing researchers to distinguish active movement from patterns consistent with death. All detected delayed deaths happened within three days, including fish that initially left the surface. The outcome shows why immediate disappearance should not be reported as a complete successful release.
Vented and descended fish had similar survival in the authors’ overall interpretation, while unvented surface releases fared worse, especially in the summer trial. Those results do not cancel the Florida study. The studies differ in location, year, assignment, release depth, fish condition, sample size and endpoint. Texas descents returned fish to the bottom; Florida compared several intermediate depths. Unknown telemetry fates also reduce certainty in the small Texas cohorts. Curtis et al. (2015), Table 1 and methods.
Our chart keeps unknowns visible among all tagged fish. The expanded table also shows survival among classifiable fates and extreme bounds if every unknown died or survived. These bounds are a transparency check, not a substitute for a time-to-event analysis or evidence that the methods are equivalent. An apparently better percentage based on four or six known fish is not a reliable universal ranking.
The two Gulf cases support a practical distinction: effective mitigation can matter, while the relative benefit of a particular method depends on how it is carried out and measured. We should teach release preparation and the limits of each comparison together.
04Case three: a promising estimate with important exclusions
The North Carolina telemetry experiment is useful for understanding what successful descending can achieve under a defined set of conditions. It is less useful as a head-to-head answer because it contained no venting treatment. Its best-case estimate excluded deep hooking and unsuccessful first descents, and it was obtained at one approximately 37-metre site outside the Gulf.
The separate fishery simulation brings hooking evidence back into a broader scenario, but assumes universal descending and no device failure. Those assumptions describe a modeled possibility. They are not a measured account of present fishing practice, and they cannot be carried across to a Louisiana boat as a guaranteed outcome. Runde et al. (2021).
The lesson is valuable even without importing its percentage: handling and hook injury belong beside barotrauma in release planning. A descender can address pressure-related impairment while another injury continues to affect survival. Our redfish hook study explains why deep hooking and mortality must also remain separate endpoints; its inshore findings are not snapper survival estimates.
05Does descending simply feed sharks?
Predation is a real release hazard, and a recovered fish can still be eaten. These three cases do not isolate the causal effect of descending on shark attraction or provide a Louisiana predator-density threshold. Neither an alarming video nor a best-case survival estimate resolves that question across all fishing locations.
The practical response is to include predator activity in the fishing plan. Return ‘Em Right recommends moving before predators concentrate, using tackle that reduces drawn-out fights and preparing release gear in advance. If repeated interactions develop, reassess the location and method rather than continuing the same sequence while treating release equipment as a guarantee. Program best practices.
06Turn the evidence into a release plan
Before the first drop, identify the target species and likely capture depth, have the appropriate dehooking and release tools accessible, and know how the chosen descender attaches and releases under its manufacturer’s instructions. Check that the complete setup can return the fish effectively under the expected conditions. A tool stored away while a fish waits on deck does not deliver the intervention tested in these studies.
Keep handling and air exposure brief. Prepare the camera before lifting a fish, support its weight horizontally and protect its gills. Recognize that the planned photograph may need to be skipped. For deep-hooked fish, the program guidance favors cutting the line rather than prolonging extraction; release decisions must account for injury as well as buoyancy. These steps complement appropriate tackle and a prompt landing.
Venting requires suitable equipment and demonstrated species-specific technique. An exposed stomach is not a swim bladder to puncture, and this evidence review is not a needle-placement tutorial. Use the program’s training and qualified instruction before attempting an invasive procedure. Descent avoids that puncture, but still requires a prepared, correctly operated setup.
07Keep conservation advice separate from current law
Checked October 21, 2026: NOAA’s July 30, 2026 continuation-action page states that the original DESCEND Act equipment provisions expired in January 2026 and describes a proposed continuation. Its displayed status was Information Gathering. An old bulletin describing the original mandate is therefore not proof of current legal applicability to a particular trip. Verify the current rules for the water and vessel involved. Official action record.
This legal check does not replace release preparation. The evidence-based reason to carry and know how to operate suitable equipment remains useful even when the legal question requires a separate check.
08Final conclusion and what would change it
A prepared descender is a well-supported choice for depth-caught red snapper requiring barotrauma mitigation. The strongest direct Gulf comparison here favors 20–30-metre recompression over venting after capture at 30–50 metres; it supplies relative-return evidence, not a transferable absolute survival gain. The Texas telemetry case supports effective mitigation while showing both delayed deaths and uncertainty. The North Carolina case illustrates a carefully bounded best-case estimate.
A stronger Louisiana-specific answer would need comparable venting and descending cohorts across capture depths and seasons, with handling time, hooking injury, device failures, predator activity and delayed fates recorded. Raw records and a design accounting for trip-level differences would permit independent estimates rather than a claim assembled from incompatible percentages. Until then, prepare the release process, use a depth-appropriate method and report the outcome no more strongly than the measurements allow.
This is a local review draft. Independent fisheries/statistical review is pending, including the Florida source discrepancy. We will check source corrections and release guidance before production publication and review this evidence at least every six months. Our artificial-reef study connects the release question to how structure concentrates fish and fishing activity; it does not supply additional release-survival data.
Sources & editorial notes
Focused secondary review of Stallings et al. (2023), Curtis et al. (2015) and Runde et al. (2021), with attributed aggregate facts and original descriptive arithmetic. No pooled effect, original field experiment or Louisiana survival estimate. Published rate/sample-size discrepancy retained visibly; no invented return counts. Independent fisheries/statistical review pending. Local review draft.
Two Gulf cases separate local red-snapper catch rates, juvenile habitat value and concentrated fishing pressure from claims of net population production.
