CoChat
Sign Up
Cited by 194Marine and fisheries researchFish Ecology and Management StudiesFish Biology and Ecology Studies

Global habitat preferences of commercially valuable tuna

Open access
Arrizabalaga, Haritz, Dufour, Florence, Kell, Laurence, Merino, Gorka, Ibaibarriaga, Leire, Chust, Guillem, Irigoien, Xabier, Santiago, Josu +10 more·2015·Deep Sea Research Part II: Topical Studies in Oceanography
DOI 10.1016/j.dsr2.2014.07.001
Open this paper in CoChat — free accountOpen PDF

AI snapshot

Written from the paper’s full text.

TL;DR

This study uses longline CPUE from 1958–2007 for six tuna species across the Atlantic, Indian, and Pacific Oceans to compare global habitat preferences with quotient analysis and Generalized Additive Models. It finds tropical tunas prefer warm, low-oxygen, stratified waters, while Atlantic and southern bluefin tunas prefer higher chlorophyll and tolerate the most extreme sea surface height anomalies and mixed layer depths.

Key findings

  • Full GAMs explained 45% (bigeye) to 64% (skipjack) of null deviance; all final models retained TEMP5, salinity, SSH, MLD, and DCHL.
  • Quotient analysis showed tropical tunas prefer warmer surface waters than temperate tunas: yellowfin preferred >25°C, while bigeye and skipjack preferred 20–28°C; albacore avoided waters below 14°C, and southern bluefin preference and Atlantic bluefin tolerance extended down to 4°C.
  • Atlantic bluefin tolerated the widest range of environmental conditions, including the widest temperature tolerance range (approximately 1–20°C), high salinity above 37 PSU, and negative SSH anomalies; southern bluefin preferred mixed layer depths beyond approximately 300 m.
  • Tropical tunas generally preferred lower oxygen environments than temperate tunas; yellowfin preferred relatively low oxygen waters (<0.2 mmol/l), bigeye tolerated them, and skipjack avoided waters below 0.1 mmol/l.

Methods

  • Longline catch and effort data were obtained from the five TRFMOs for 1958–2007, except southern bluefin from 1965 and eastern Pacific stocks only until 2004; no skipjack longline data were available from WCPFC. Data were aggregated by quarter and 5°×5° resolution, and nominal CPUE was computed as catch per number of hooks, using only positive CPUE records.
  • Environmental variables came from the PISCES biogeochemical model: surface temperature (TEMP5), temperature at 100 m (TEMP100), sea surface salinity, dissolved oxygen at 100 m, sea surface height, mixed layer depth, chlorophyll, and mesozooplankton averaged over 0–100 m.
  • Habitat preferences were assessed with quotient analysis using 400 re-sampling iterations to test preference/tolerance/avoidance, and with GAMs assuming lognormal positive CPUE, spline functions with smoothness restricted to 4, fixed factors for year, season, and stock, and stepwise backward selection by AIC; TEMP100, oxygen, and zooplankton were excluded from full GAMs due to collinearity.

Limitations

  • The paper states that CPUE may not accurately reflect population abundance due to changes in fishing practice, fleet efficiency, or environmental effects on catchability, although fixed factors were included and interpreting abundance trends was out of scope.
  • The paper states that collinearity prevented inclusion of oxygen and other correlated variables in the final GAMs, and that predictions outside the observed range were unrealistic for some species, such as skipjack in the Mediterranean and at extreme latitudes.
  • The paper states that longline data mostly catch adult tunas, so results mainly reflect adult habitat preferences; it also notes that spatio-temporally averaged chlorophyll may not capture frontal effects, prey field observations are generally lacking, and tagging studies are often local and short-term.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.dsr2.2014.07.001

Cites

99 of 99 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

22 of 79 cited sources checked · 13 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    Temperature and oxygen affect important biological processes and thus determine the spatial distribution of tunas.

    “Temperature and oxygen affect important biological processes and thus determine the spatial distribution of tunas (Barkley et al., 1978;Boyce et al., 2008;Brill, 1994;Strama et al., 2012).”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Barkley et al., 1978

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Boyce et al., 2008 source

      The Introduction links physiological temperature tolerance to tuna range boundaries and identifies temperature as a strong predictor of distribution. It mentions oxygen content only as a factor hypothesized to correlate with abundance and distribution; it does not establish that oxygen affects biological processes and thereby determines tuna distribution. The cited claim states that relationship more definitively than this paper does. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Brill, 1994 source
  • Premisecitation

    Salinity can influence large-scale spatial distribution of tunas.

    “Salinity can influence large scale spatial distribution (e.g.Fromentin et al., 2014;Maury et al., 2001;Reygondeau et al., 2012).”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites Fromentin et al., 2014 source
    • cites Maury et al., 2001 source
    • cites Reygondeau et al., 2012 source
  • Premisecitation

    Habitat studies using fishery or survey data are often spatially limited and focused on single species.

    “Habitat studies using fishery or survey data are often spatially limited and focused on single species (e.g.Chen et al., 2005;Sagarminaga and Arrizabalaga, 2010).”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites Chen et al., 2005 source

      The Introduction describes limitations in the spatial coverage of past fishery-oceanography studies, and this paper’s own habitat analysis focuses on one species, albacore. But it uses fishery and environmental data across the whole Indian Ocean, so Chen et al. is not itself an example of a spatially limited study. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites Sagarminaga and Arrizabalaga, 2010 source
  • Premisereasoning

    There are few studies where the habitat preferences of a group of species are analysed using similar datasets and methods at broad spatial scales.

    “In fact, there are few studies where the habitat preferences of a group of species are analysed using similar datasets and methods at broad spatial scales (but see Reygondeau et al. 2012).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Reygondeau et al. 2012 source
  • Premisecitation

    The distribution of prey is one of the main drivers of the spatial distribution and behaviour of tunas.

    “In addition to the physical environment, the distribution of prey is also one of the main drivers of the spatial distribution and behaviour of tunas (Bertignac et al., 1998;Schick and Lutcavage, 2009, tunas, but data on tuna prey distribution and abundance is scarce and is mostly available from models (Lehodey et al., 1998).”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Bertignac et al., 1998 source
    • cites Schick and Lutcavage, 2009 source

      The Results report a significant association between herring density and bluefin tuna presence, and the Discussion says prey distribution predicted bluefin presence better than SST fronts. The study covers a one-week snapshot in the Gulf of Maine; it does not establish prey distribution as a main driver of behaviour or of tuna distribution generally. (LLM, conf=0.88)

      Experimental — accuracy not yet measured
    • cites Lehodey et al., 1998 source
  • May combine two claims. 41
  • Classified as probably not a substantial claim. 16

  • May combine two claims.

    Premisecitation

    PISCES has been extensively validated and can realistically simulate seasonal, interannual and decadal physical and biogeochemical ocean variability.

    “PISCES has been extensively validated and has demonstrated the ability to realistically simulate seasonal, interannual and decadal physical and biogeochemical ocean variability.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Aumont and Bopp, 2006 source

      Section 3.2 (Chlorophyll; page unavailable) reports that PISCES qualitatively reproduces observed chlorophyll patterns for May and November, supporting a limited claim about seasonal biogeochemical simulation. Section 3 (Model Mean State; page unavailable) explicitly says its comparisons are not an exhaustive validation. The paper does not establish realistic interannual or decadal physical and biogeochemical variability, so the broader claim overstates its evidence. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    GAMs have been previously used to study habitat preferences of many tuna species using fishery data, larval survey data, and electronic tagging data.

    “GAMs have been previously used to study habitat preferences of many tuna species using fishery data (Lan et al., 2013;Maury et al., 2001;Mugo et al., 2010;Sagarminaga and Arrizabalaga 2010), larval survey data (Rooker et al., 2013) as well as electronic tagging data (Hazen et al., 2013).”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Lan et al., 2013 source
    • cites Maury et al., 2001 source
    • cites Mugo et al., 2010 source

  • May combine two claims.

    Premisecitation

    Longlines resemble long baited transects and catch a wide range of species in a consistent way over a vast spatial scale.

    “Longlines resemble long baited transects and catch a wide range of species in a consistent way over a vast spatial scale and longline catch data have been previously used to analyse changes in abundance (Myers and Worm, 2003), diversity (Worm et al., 2005), range contraction (Worm and Tittensor, 2011) and biogeography (Reygondeau et al., 2012).”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Myers and Worm, 2003 source
    • cites Worm et al., 2005 source
    • cites Worm and Tittensor, 2011 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Extensive longline datasets with broad spatio-temporal coverage have supported a number of studies on tuna distribution and habitat preferences.

    “In fact, extensive longline datasets with broad spatio-temporal coverage have supported a number of studies in relation to tuna distribution and habitat preferences (e.g.Fromentin et al., 2014;Reygondeau et al., 2012;Worm and Tittensor, 2011).”
    Quote in the paperNot a substantial claim3 of 3 sources checked
    • cites Fromentin et al., 2014 source
    • cites Reygondeau et al., 2012 source
    • cites Worm and Tittensor, 2011 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Reygondeau et al. (2012) proposed a conceptual scheme where yellowfin, bigeye, albacore and bluefin species dominated different ecoregions as latitude increased.

    “Reygondeau et al. (2012) proposed a conceptual scheme where yellowfin, bigeye, albacore and bluefin species dominated different ecoregions as latitude increased.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Reygondeau et al. (2012) source

  • May combine two claims.

    Findingcitation

    For Atlantic bluefin, the model identifies the Mediterranean and northern Atlantic (>40oN) as most suitable, with habitat suitability increasing further north beyond the northern range of observations and decreasing towards the equator and southern hemisphere.

    “In the case of Atlantic bluefin, the model identifies the Mediterranean and the northern Atlantic (>40ºN) as most suitable, habitat suitability increasing further north beyond the northern range of the observations and decreasing towards the equator and southern hemisphere in spite of important catches that occurred in the past (Fromentin et al., 2014).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Fromentin et al., 2014 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    In the case of albacore, our quotient analysis results for temperature preference (up to and beyond 20oC) are comparable to those of Hazen et al. (2013), while our GAM results suggest somewhat lower preferred values (around 15oC), more in line with Boyce et al. (2008) and Lan et al. (2012).

    “In the case of albacore, our quotient analysis results for temperature preference (up to and beyond 20ºC) are comparable to those of Hazen et al. (2013), while our GAM results suggest somewhat lower preferred values (around 15ºC), more in line with Boyce et al. (2008) and Lan et al. (2012).”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Hazen et al. (2013) source
    • cites Boyce et al. (2008) source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites Lan et al. (2012) source
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Interpretationreasoning

    Our predicted distributional ranges for albacore, bigeye and yellowfin were comparable to those in Worm and Tittensor (2011), but were more extensive for both bluefin species and skipjack.

    “Our predicted distributional ranges for albacore, bigeye and yellowfin were also comparable to those in Worm and Tittensor (2011), but were more extensive for both bluefin species and skipjack.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Worm and Tittensor (2011) source
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    In the univariate GAMs, the stock factor explained a relatively high proportion of the null deviance for skipjack (23%), bigeye (20%) and albacore (12%), but less than 1% for other species.

    “In the univariate GAMs, the”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingown data

    Scatterplots between environmental variables suggested high collinearity (R2 > abs(0.6)) between TEMP5 and TEMP100, TEMP5 and O2, and DCHL and ZOO.

    “Scatterplots between environmental variables suggested high collinearity (R 2 >abs(0.6))between TEMP5 and TEMP100, TEMP5 and O2, and DCHL and ZOO (Figure 1).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The full GAMs generally included the three-way stock:year:season interaction for fixed factors, except for albacore (stock:year and stock:season) and southern bluefin (year:season).

    “They also generally included the three way stock:year:season interaction for fixed factors, except in the case of albacore where the interactions stock:year and stock:season were kept in the final model, and in the case of southern bluefin, with a single stock, where the year:season interaction was included.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingown data

    Bluefin tunas can tolerate a wide range of DCHL and ZOO values, although their preference and tolerance ranges are not very well defined.

    “In general, relatively low DCHL and ZOO waters are preferred by tropical tunas and albacore, compared to both bluefin tunas, for which the preference and tolerance ranges are not very well defined, but results suggest that they can tolerate a wide range of DCHL and ZOO values.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Albacore avoids waters below 14°C, which are preferred or tolerated by both Atlantic bluefin and southern bluefin.

    “Albacore avoids waters below 14ºC, which are preferred or tolerated by both Atlantic bluefin and southern bluefin.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Albacore preference and tolerance ranges extend to about 200 m, while southern bluefin tuna preference range extends to highly mixed waters with mixed layer depths of several hundred meters that Atlantic bluefin tuna tolerates.

    “Albacore preference and tolerance ranges extend to about 200 m, while southern bluefin tuna preference range extends to highly mixed waters with mixed layer depths of several hundred meters that Atlantic bluefin tuna tolerates.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    SSH was the environmental variable that explained the highest percentage of deviance (22%) for Atlantic bluefin, and between 28% and 5% for the rest of the species, except for albacore (<1%).

    “SSH was the environmental variable that explained the highest percentage of deviance (22%) in the case of Atlantic bluefin, and between 28% and 5% for the rest of the species, except for albacore (<1%).”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingown data

    Oxygen, highly correlated with temperature, explained similar proportions of null deviance.

    “Oxygen, highly correlated with temperature, explained similar proportions.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingown data

    TEMP5 was the most important variable for yellowfin tuna, as TEMP100 was for southern bluefin.

    “TEMP5 was the most important variable for yellowfin tuna, as TEMP100 was for southern bluefin.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingnone given

    Among temperate tunas, albacore showed a relatively narrow preference-tolerance range compared to southern bluefin, which showed the largest oxygen preference values.

    “Among the temperate tunas, albacore showed a relatively narrow preference-tolerance range compared to southern bluefin that showed the largest oxygen preference values.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The percent deviance explained by the year factor varied between 2% (bigeye) and 21% (skipjack), while the season factor explained minor deviance (<2% in all cases except for bluefin).

    “The percent deviance explained by the”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Positive CPUE observations follow a lognormal error distribution.
    • The authors state this as an assumption: The niche model is unimodal.
    • Rests on an assumption: Changes in abundance and catchability are likely to occur at the stock level.

  • May combine two claims.

    Findingnone given

    Yellowfin tuna preferred a wide range of relatively low oxygen waters (<0.2 mmol/l) while bigeye tolerated them.

    “Yellowfin tuna preferred a wide range of relatively low oxygen waters (<0.2mmol/l) while bigeye tolerated them.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Skipjack tolerance range was more restricted and avoided waters below 0.1 mmol/l.

    “Skipjack tolerance range was more restricted and avoided waters below 0.1 mmol/l.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The quotient analysis suggests that Atlantic bluefin tuna has no well-established temperature preference but the widest temperature tolerance range (approximately between 1°C and 20°C).

    “It is remarkable that this analysis suggests that there is no well-established temperature preference, but the widest temperature tolerance range (approximately between 1ºC and 20ºC) for Atlantic bluefin tuna.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The preference range for southern bluefin and the tolerance range for Atlantic bluefin extend down to 4°C.

    “The preference range for southern bluefin and the tolerance range for Atlantic bluefin extend down to 4ºC (Figure 3, Figure S1).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    Bluefin tunas showed preference and tolerance for negative SSH values, with Atlantic bluefin preferring the most negative values.

    “On the contrary, bluefin tunas showed preference and tolerance for negative SSH values, Atlantic bluefin preferring most negative values.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Bigeye and yellowfin showed decreasing salinity trends beyond 34 and 35 PSU respectively, while skipjack showed increasing trends.

    “Bigeye and yellowfin showed decreasing trends beyond 34 and 35 PSU respectively, but increasing trends were seen for skipjack.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    SSH response curves suggested positive SSH anomalies were preferred by southern bluefin, albacore and skipjack, and negative SSH anomalies were preferred by Atlantic bluefin, bigeye, yellowfin and also by skipjack.

    “Response curves for SSH suggested that positive SSH anomalies were preferred by southern bluefin, albacore and skipjack, and negative SSH anomalies were preferred by Atlantic bluefin, bigeye, yellowfin and also by skipjack.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    Southern bluefin tuna preferred mixed layer depths beyond approximately 300 m, while Atlantic bluefin showed a clear preference for increasing mixed layer depths.

    “Southern bluefin tuna preferred mixed layer depths beyond approximately 300 m, while Atlantic bluefin showed a clear preference for increasing mixed layer depths.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    For southern bluefin, models predicted suitable habitat in the northern fringe (>40oN) where there are no observations, but middle latitudes would act as a barrier to potential expansion from south to north.

    “In the case of southern bluefin, the models predicted suitable habitat also in the northern fringe (>40ºN) of the planet (where there are no observations of southern bluefin), but middle latitudes would act as a barrier to potential expansion from south to north.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Fitted smooth curves show preference for warm waters (20-30oC) for tropical tunas, temperate waters (around 15oC) for albacore and Atlantic bluefin tuna, and colder waters (<10oC) for southern bluefin.

    “The fitted smooth curves show the preference for warm waters (between 20ºC and 30ºC) in the case of tropical tunas, for temperate waters (around 15ºC) in the case of albacore and Atlantic bluefin tuna, and for colder waters (< 10ºC) in the case of southern bluefin.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    Predictions for albacore, yellowfin, and bigeye were reasonable, with poor habitat predictions out of the observed range (e.g., north of 50oN and south of 50oS).

    “The predictions for the species with most extensive data coverage (albacore, yellowfin, bigeye) were reasonable, with poor habitat predictions out of the observed range (e.g.north of 50ºN and south of 50ºS).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Albacore preferred low DCHL waters, while southern bluefin, Atlantic bluefin and yellowfin showed an optimum at relatively high DCHL concentrations.

    “Albacore preferred low DCHL waters, while southern bluefin, Atlantic bluefin and yellowfin showed an optimum at relatively high concentrations.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    Salinity response curves showed a peak slightly above 35 PSU for southern bluefin, remained high above 36 PSU for Atlantic bluefin, and showed an increasing trend beyond 36 PSU for albacore.

    “Salinity response curves showed a peak slightly above 35 PSU for southern bluefin, remained high above 36 PSU for Atlantic bluefin and showed an increasing trend beyond 36 PSU for albacore.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Findingown data

    Albacore is a temperate to subtropical species with highest CPUEs between 20o-40o of each hemisphere.

    “Albacore is a temperate to subtropical species with highest CPUEs between 20º-40º of each hemisphere.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Longline catch per unit effort (CPUE) is a valid proxy for tuna habitat preferences.

  • May combine two claims.

    Findingown data

    Skipjack's northernmost distribution is limited to around 40oN, and highest CPUEs are observed around 20o latitude in each hemisphere.

    “In the case of skipjack, whose northernmost distribution is limited to around 40ºN, highest CPUEs are observed around 20º latitude in each hemisphere.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Longline catch per unit effort (CPUE) is a valid proxy for tuna habitat preferences.

  • May combine two claims.

    Premisecitation

    Most tagging studies are local with short deployment periods and provide information about habitat characteristics only around deployment areas.

    “most tagging studies are local with short deployment periods (e.g.Cosgrove et al., 2014), and provide information about habitat characteristics only around deployment areas (Hobday and Evans, 2013).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Cosgrove et al., 2014 source
    • cites Hobday and Evans, 2013 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Missing values were omitted before the stepwise selection, so that model comparisons were based on the same dataset.

    “Missing values were omitted before the stepwise selection, so that model comparisons were based on the same dataset.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    A GAM enables fitting non-linear models for a wide family of statistical distributions.

    “A GAM enables fitting non-linear models for a wide family of statistical distributions, and thus is preferred to some other habitat modelling approaches (Guisan and Zimmermann, 2000).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Hastie and Tibshirani, 1990

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Guisan and Zimmermann, 2000 source

  • May combine two claims.

    Premisereasoning

    Pelagic longlines show the largest spatio-temporal coverage among fishing methods, having operated for several decades across all oceans and targeting all main commercial tuna species except skipjack.

    “Among the fishing methods, pelagic longlines show the largest spatio-temporal coverage, since they have operated for several decades across all oceans, targeting all main commercial tuna species except skipjack.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Some environmental variables (e.g., temperature, DCHL) have been used in previous studies, but description of habitat preferences based on other variables (e.g., MLD, ZOO) is rarely found in the literature.

    “Some of these variables (e.g.temperature, DCHL) have been used in previous studies (e.g.Sagarminaga and Arrizabalaga 2010;Mugo et al., 2010), but description of the habitat preferences based on other variables (e.g.MLD, ZOO) is rarely found in the literature.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Sagarminaga and Arrizabalaga 2010 source
    • cites Mugo et al., 2010 source

  • May combine two claims.

    Interpretationown data

    Our habitat predictions are generally consistent with Reygondeau et al.'s conceptual scheme, although latitudinal ranges of suitable habitat for bigeye tuna were comparable to those of albacore.

    “Our habitat predictions are generally consistent with that view, although latitudinal ranges of suitable habitat for bigeye tuna were comparable to those of albacore.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Interpretationown data

    Hazen et al. (2013) estimated an optimum of around 20oC for yellowfin, while our results of both quotient analysis and GAM suggest a preference for waters of up to 30oC.

    “However, Hazen et al. (2013), using an extensive and detailed tagging dataset, estimated an optimum of around 20ºC for yellowfin, while our results of both the quotient analysis and the GAM suggest a preference for waters of up to 30ºC.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hazen et al. (2013) source
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Fishery data have been criticised for inducing biased conclusions about stock abundance trends because CPUE series may not accurately reflect trends of population abundance due to changes in fishing practice, efficiency of fleets or environmental effects on catchability.

    “Fishery data have been criticised for inducing biased conclusions about stock abundance trends because CPUE series may not accurately reflect trends of population abundance due to changes in fishing practice, efficiency of the fleets or environmental effects on catchability, among others (see Maunder and Punt, 2004;Maunder et al., 2006;Polacheck, 2006).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Maunder and Punt, 2004 source
    • cites Maunder et al., 2006 source
    • cites Polacheck, 2006 source

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning

    Considering differences in the nature of the data and the approach used, our results are generally consistent with several other studies.

    “Considering the differences in the nature of the data and the approach used, our results are generally consistent with several other studies.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In the GAMs, fixed factors were included to accommodate changes in abundance and catchability.

    “In our case, we included fixed factors in the GAMs to accommodate changes in abundance and catchability.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The environmental variables in areas where skipjack suitable habitat was predicted outside observations are out of the ranges used in the skipjack model, and thus the GAM predictions there are unrealistic.

    “The environmental variables in these areas are out of the ranges used in the skipjack model and thus the GAM predictions are unrealistic.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The 25%-75% interquartile range of the temperature preference profile in Cheung et al. (2013) is consistent with our estimates, but for temperate tunas the tolerance we estimated extended to colder waters.

    “The 25%-75% interquartile range of the temperature preference profile in Cheung et al. (2013) is also consistent with our estimates, but in the case of temperate tunas the tolerance we estimated extended to colder waters.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Cheung et al. (2013) source
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • May combine two claims.

    Interpretationreasoning

    Our habitat predictions showed a slightly higher latitudinal range compared to those predicted by coupled biophysical models for skipjack in the Pacific, Indian, and worldwide, and for Pacific bigeye tuna.

    “However, our habitat predictions showed a slightly higher latitudinal range compared to those predicted by coupled biophysical models in the case of skipjack in the Pacific Ocean (Bell et al., 2013;Lehodey et al., 2013), Indian Ocean (Dueri, 2012), and worldwide (Dueri, 2014), as well as in the case of Pacific bigeye tuna (Lehodey et al., 2010).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Bell et al., 2013 source
    • cites Lehodey et al., 2013 source
    • cites Dueri, 2012 source
    • Rests on an assumption: The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Briand et al. (2011) suggested DCHL is important for Pacific albacore.

    “The importance of DCHL was also suggested by Briand et al. (2011), using a detailed fishery dataset, for Pacific albacore.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Briand et al., 2011 source

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Fishery data remain a valuable source of information to analyse global habitat of tunas at broad spatial and temporal scales.

    “In the meantime, fishery data remain a valuable source of information to analyse global habitat of tunas at broad spatial and temporal scales.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    Trophodynamics of tunas are complex and dynamic.

    “trophodynamics of tunas are complex and dynamic (Young et al., 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Young et al., 2014

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Interpretationnone given

    The mechanisms through which salinity could affect tuna distribution remain unclear and could be a proxy of other underlying processes.

    “However, the mechanisms through which salinity could affect tuna distribution remain unclear and could be a proxy of other underlying processes.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Maury et al. (2001) suggested that low salinities could indicate favourable trophic areas induced by fluvial water supplies.

    “Maury et al. (2001) suggested that low salinities could indicate favourable trophic areas induced by fluvial water supplies.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Maury et al., 2001 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No data on skipjack caught by longline was available from the Western and Central Pacific.

    “No data on skipjack caught by longline was available from the Western and Central Pacific.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The importance of salinity for tuna distribution remains controversial in the literature.

    “The importance of salinity for tuna distribution remains controversial in the literature, with some suggesting it is not important in influencing oceanic distribution (Sund et al., 1981), while others have found it to be influential (e.g.Bernal et al., 2009;Maury et al., 2001;Reygondeau et al., 2012).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Sund et al., 1981

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Bernal et al., 2009

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Maury et al., 2001 source

  • May combine two claims.

    Interpretationown data

    Salinity is important for tuna distribution, as supported by GAM results showing variable response curves for all species.

    “Our GAM results tend to support the importance of salinity since this variable showed variable response curves for all the species.”
    Quote in the paperSubstantial claim
55 of 55

  • The authors state this as an assumption.

    CPUE is a proxy for fish abundance.

    “the CPUE was assumed to be a proxy for fish abundance.”

  • The paper relies on this without support.

    Tunas show a wide vertical distribution.

    “For this study, and because tunas show a wide vertical distribution, we selected the following variables to characterise the surface and subsurface environment:”

  • The authors state this as an assumption.

    Positive CPUE observations follow a lognormal error distribution.

    “Positive CPUE observations were assumed to follow a lognormal error distribution and they were modelled as a function of the environmental variables, as well as fixed factors for”

  • The authors state this as an assumption.

    The niche model is unimodal.

    “The degree of smoothness of model terms was restricted to 4 in order to assume a unimodal niche model (sensu Hutchinson, 1957).”

  • The paper relies on this without support.

    Changes in abundance and catchability are likely to occur at the stock level.

    “Changes in abundance and catchability are likely to occur at the stock level and so three way interactions among fixed factors were considered in the full GAMs.”

  • The paper relies on this without support.

    The response curves of the final GAMs can be interpreted as habitat suitability for the tuna species.

    “The final GAMs were kept as those describing the habitat of the tuna species, and their response curves for environmental variables were interpreted in terms of habitat suitability (Figure 4).”

  • The paper relies on this without support.

    Longline catch per unit effort (CPUE) is a valid proxy for tuna habitat preferences.

    “our study reflects mostly habitat preferences for adult tunas (since it is based on longline data that mostly catches adult fish)”

  • The paper relies on this without support.

    The quotient analysis's preference, tolerance and avoidance ranges validly indicate species-specific habitat preferences.

    “Summary of quotient analyses, with species specific preference (green), tolerance (yellow) and avoidance (red) ranges for each environmental variable.”
8 of 8

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Skipjack performs migratory movements conditioned by SST and prey availability.

    “It typically preys on small (2-5 cms) pelagic fishes, crustaceans and cephalopods (Magnuson, 1969;Roger, 1994;Druon et al., 2017), and performs migratory movements conditioned by SST and preys availability (Arrizabalaga et al., 2015;Druon et al., 2017).”

    From African desert dust influences migrations and fisheries of the Atlantic skipjack-tuna (2023) · cites this paper as Arrizabalaga et al., 2015

    Section 2.1 (page unavailable) includes skipjack among highly migratory tunas, and the Discussion (page unavailable) says tunas make trophic migrations to exploit productive areas. The paper also models skipjack habitat in relation to surface temperature, but it does not establish that skipjack’s migratory movements specifically are conditioned by both SST and prey availability. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • Skipjack tuna inhabits warm waters, preferably those with a sea surface temperature (SST) between 20 and 28°C.

    “Skipjack tuna inhabits warm waters, preferably those with a sea surface temperature (SST) between 20 and 28 • C (Arrizabalaga et al., 2015).”

    From African desert dust influences migrations and fisheries of the Atlantic skipjack-tuna (2023) · cites this paper as Arrizabalaga et al., 2015

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Block, 2001, vol. 19
  2. FAO, 2011. Tuna and tuna-like species. In: Review of the State of World Marine Fishery Resources. FAO Fisheries and Aquaculture Technical Paper No. 569. Rome, pp. 227–243.

Cited by

Sorted by citations

Related

Authors
Arrizabalaga, H., Dufour, F., Kell, L., Merino, G., Ibaibarriaga, L., Chust, G., Irigoien, X., Santiago, J., Murua, H., Fraile, I., Chifflet, M., Goikoetxea, N., Sagarminaga, Y., Aumont, O., Bopp, L., Herrera, M., Marc Fromentin, J., & Bonhomeau, S.
Affiliations
  • AZTI (ES)
  • International Commission for the Conservation of Atlantic Tunas (ES)
Venue
Deep Sea Research Part II: Topical Studies in Oceanography
Publisher
Elsevier BV
Published
2015-03
Volume
113
Pages
102-112
ISSN
0967-0645 (ISSN-L)
DOI
10.1016/j.dsr2.2014.07.001
OpenAlex
W2037723922
Semantic Scholar
8593a7cd0965d2a821fa75e8bbd894edf02396e7
Cited by
194
References
98
Citations by year
2026: 26 · 2025: 17 · 2024: 25 · 2023: 24 · 2022: 16 · 2021: 18 · 2020: 12 · 2019: 15 · 2018: 11 · 2017: 8 · 2016: 12 · 2015: 8 · 2014: 1
FWCI
7.87
Type
journal article
Language
en
Open access
Licence
www.elsevier.com/tdm/userlicense/1.0/ (tdm)
Locations
  • Deep Sea Research Part II Topical Studies in Oceanography (publishedVersion)
  • HAL (Le Centre pour la Communication Scientifique Directe) (submittedVersion)
  • Institutional Archive of Ifremer (French Research Institute for Exploitation of the Sea) (submittedVersion)
  • AZTI Institutional Repository (submittedVersion, cc-by-nc-sa)
  • King Abdullah University of Science and Technology Repository (King Abdullah University of Science and Technology) (submittedVersion, cc-by-nc-nd)
  • INFM-OAR (INFN Catania) (submittedVersion, other-oa)
Funders
  • EU Marie Curie EST
  • METAOCEANS (MEST-CT-2005–019678)
  • European Commission (264933)
  • EURO-BASIN: European Union Basin-scale Analysis, Synthesis and Integration
Topics
  • Physical Sciences › Environmental Science › Global and Planetary Change › Marine and fisheries research
  • Physical Sciences › Environmental Science › Nature and Landscape Conservation › Fish Ecology and Management Studies
  • Life Sciences › Agricultural and Biological Sciences › Aquatic Science › Fish Biology and Ecology Studies
Keywords
tuna, habitat preference, ecosystem approach to fisheries, Thunnus thynnus, southern bluefin tuna, albacore tuna, bigeye tuna, yellowfin tuna, skipjack tuna, chlorophyll, sea level anomalies, mixed layer depth, water column stratification, anoxic waters

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-16.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-16.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review