PrimeBite Fishing Blue Marlin Research & Methodology

What Conditions Turn On Blue Marlin?

Last updated: 2026-09-26

This page explains the Atlantic Blue Marlin rating: what PrimeBite uses, what the research supports, and what the limits are.

How it works (overview)

The model is an all-day rating built from two drivers: Atlantic seasonality and sea-surface temperature (SST).

Season and sea-surface temperature are the only scored drivers. The highest displayed rating is 3.8 in the NW Atlantic and transition regions and 3.4 in the SW Atlantic. These ceilings are modeling choices informed by evidence, not maximum scores measured by the studies; the exact calibration stays private.

The model covers PrimeBite's approved NW Atlantic, SW Atlantic, and equatorial transition cohorts. Cape Verde, Madeira, the Canaries, the Azores, and locations outside the Atlantic show no Blue Marlin rating rather than borrowing evidence from another stock or ocean.

Research supports which drivers belong and the direction of their effect. PrimeBite chooses the exact curve shape, thresholds, weights, and rating ceilings within those evidence limits.

Variables (what's research-backed vs calibrated)

Research-backed means the driver, pattern, or mechanism is supported by open-access or peer-reviewed sources. Calibrated means we chose the exact curve shape, thresholds, weights, or caps to produce a stable 1.0–5.0 score and to avoid overclaiming from region- or gear-specific studies.

Why moon phase is not scored

PrimeBite does not use moon phase in the Atlantic Blue Marlin rating. The earlier page overstated what the evidence could support. The corrected reading does not justify a transferable moon-phase contribution to this day rating, so the model leaves moon phase unscored.

Andrzejaczek et al. (2025) does not cite Orbesen (2017). Its Blue Marlin chart shows two higher-catch results and three no-effect results, without identifying their fishing methods. Its reference to Shimose (2013) is a secondary citation, not another independent study.

Orbesen et al. (2017) analyzed 4,840 Gulf of Mexico longline sets. Its higher daytime Blue Marlin catch result is not a NW Atlantic full-moon peak. PrimeBite no longer uses that paper to support a moon contribution.

SST (sea-surface temperature)

Research-backed: Orbesen et al. 2017 modelled SST in five bands; the paper's published appendix shows the five-band model. The band-by-band coefficient table is not printed in the paper. The authors shared it with PrimeBite by email in May 2026 as unpublished analysis behind that paper, and it is not peer-reviewed. It shows catch rates roughly 18× higher in the two warmest bands (above about 27°C) than in the coldest (below about 21°C); the two warmest bands are statistically indistinguishable in that analysis.

What the printed Atlantic research adds:

• Kraus 2011, Brander 2003, Domingo & Pons: three separate studies agree blue marlin are rarely found below about 19–20°C. Cold water contributes little or nothing to the rating; the exact scoring threshold stays private.
• Goodyear 2002 (ICCAT): a basin-wide habitat-suitability model, not a catch rate. Optimum 26–27°C and little habitat above 31°C, which is why very hot water counts for less.
• Hazin et al. 2007: SST significant in the SW Atlantic GLM (p less than 0.0001), direction only.
• Martinez-Escauriaza et al. 2021 (Sustainability): the one Atlantic rod-and-line model. SST significant at p=0.0007, but its Madeira optimum is a cooler 22–23°C and month is the larger term in the same model. That conflict with the Gulf of Mexico peak is real and we name it rather than smooth it over.

A correction (15 September 2026): until this update the temperature effect was the biggest number in the Blue Marlin rating, and the reason written down for its size was that one study had a bigger sample than most. Our rules do not allow a sample size to set a rating, so we reduced the temperature effect, aligned the cold-water treatment with three studies, accounted for the warm bands the table cannot tell apart, and made very hot water count for less instead of more. Season now leads. The size of the temperature effect still rests on that one unpublished table, and this page says so.

Calibrated: the exact SST curve shape and thresholds are proprietary.

Seasonality (hemisphere-aware)

NW Atlantic

• Hammond 2001 (SCDNR Carolinas, n=1,610): May–July accounted for 87% of captures, with June–July as the apex.
• Martinez-Escauriaza et al. 2021 (Madeira): Month explained 78% of CPUE deviance. June–July peak.

SW Atlantic

• Amorim 2009 (ICCAT, SE Brazil): Q1 + Q4 austral summer peak from yacht-club tournament data.
• Mayer 2019 (USP thesis): Q1 + Q4 weak positive seasonal signal from Bayesian modeling of Japanese longline data 1970–2009.

Calibrated: the exact monthly proportions for each cohort are proprietary.

What's NOT included (and why)

• Chlorophyll / fronts / eddies: zero direct Atlantic Blue Marlin CPUE evidence links these to catch rates. Pacific chlorophyll results exist but are excluded under PrimeBite's Atlantic-stock-only policy for this species.
• Wind / pressure / rain / NAO: not significant in the Atlantic Blue Marlin literature. Wind was explicitly NULL in Ortiz 2001; pressure, rain, and NAO were not significant in Martinez 2021's adjusted GAM.
• Moon phase: see the moon section above. The reviewed evidence does not justify a transferable moon-phase contribution to this Atlantic day rating, so moon remains unscored.
• Major/Minor solunar windows: Blue Marlin is rated as a day-level value. No published evidence supports sub-daily window effects for this species.

Limitations / caveats

• Moon phase is not a driver for this model. This is a model-evidence decision, not a claim that moon phase can never affect Blue Marlin behavior or fishing.
• Cape Verde, Madeira, the Canaries, the Azores, and locations outside the Atlantic receive no rating instead of borrowing another region's evidence.
• SST evidence is primarily from the Gulf of Mexico (Orbesen 2017) and Madeira (Martinez 2021), and the two disagree about where the best water is. PrimeBite applies a single SST curve basin-wide as a modeling choice.
• The size of the temperature effect rests on one source: an unpublished coefficient table from the Orbesen study, shared by its author. Commercial longline catch per set, not rod-and-line. We would rather it were more than one, and we sized the effect with that in mind.
• The highest displayed ratings are 3.8 in the NW Atlantic and transition regions and 3.4 in the SW Atlantic. These are modeling choices informed by evidence, not maximum catch scores measured by the studies.

Sources

• Orbesen et al. 2017 — Bull. Mar. Sci. 93(2):573–589
• Hazin et al. 2007 — Coll. Vol. Sci. Pap. ICCAT 60(5):1652–1662
• Afonso et al. 2021 — Fisheries Research 235:105822
• Goodyear 2002 — ICCAT Coll. Vol. Sci. Pap. 54(3):834–846
• Martinez-Escauriaza et al. 2021 — Sustainability 13:8975
• Hammond 2001 — SCDNR (South Carolina Department of Natural Resources)
• Jones, Judge & Ortiz 1998 — ICCAT Coll. Vol. Sci. Pap.
• Amorim 2009 — ICCAT Coll. Vol. Sci. Pap.
• Mayer 2019 — USP (Universidade de São Paulo) thesis
• Andrzejaczek et al. 2025 — Springer Nature (cross-billfish review)

Changelog

• 2026-09-15: Sea-surface temperature resized to its evidence. The effect is smaller, cold water contributes little or nothing, the source cannot tell its two warmest bands apart, and very hot water counts for less. Season now leads. The best possible rating is 3.8 out of 5 in the NW Atlantic and 3.4 off Brazil. We also corrected the sample size we had been quoting for the Gulf of Mexico study (4,840 sets, not 4,805) and this page's wording on what that study published.
• 2026-09-14: Moon phase was removed as a driver because the reviewed evidence did not justify transferring a dependable moon contribution into this Atlantic day rating. The earlier NW Atlantic full-moon claim was retired.
• 2026-05-25: Blue Marlin v1 launched with 2 active cohorts (NW + SW Atlantic). NE Atlantic (Macaronesia) gated pending local evidence.