The gill-oxygen limitation theory (GOLT) and its critics

See allHide authors and affiliations

Science Advances  06 Jan 2021:
Vol. 7, no. 2, eabc6050
DOI: 10.1126/sciadv.abc6050
  • Fig. 1 Different forms of the von Bertalanffy growth function.

    The VBGF fitted to bluefin tuna (T. thynnus) length-age data pairs (137). (A) Standard VBGF (Eq. 2), which assumes d = 2/3, and hence D = 1 (which can thus be omitted). (B) Same length-at-age data, fitted by Eq. 4, with b = 3 and D = 0.3, corresponding to d = 0.9 (46). (C) Two versions of the generalized VBGF for weight (Eq. 5), with D = 1 and D = 0.3, with weights converted from lengths using W = 0.0182·L3 (from FishBase;, where W is in g and (fork) length is in cm. Note the position of Lm [from (138)] relative to Lmax, L and Wm relative to Wmax and W, and that the weights at inflection points of the growth curves (Wi) are much higher than Wm, i.e., that bluefin tuna growth is still accelerating when they reach maturity.

  • Fig. 2 Schematic representation of water flows across the gills of a fish.

    Note that once water has flown between lamellae (which extracted most of the O2 it contained), there is no point for this water to flow through another set of lamellae. Hence, gills function as a surface, although their arrangement in 3D space may suggest otherwise.

  • Fig. 3 Two views of the relationships between size at first maturity and maximum size.

    (A) Traditional view, where “linear” growth slows down when length at first maturity (Lm; black star) is reached, with growth then continuing at a reduced pace, depending on circumstances [i.e., a, b, or c; redrawn from (77)]. (B) More appropriate, but uncommon, view, with growth expressed as change in body weight (in line with Eq. 1). This shows not only that weight at first maturity in females and males (Wm; black star) is reached when growth is still accelerating (i.e., Wm < Wi, the inflexion of the curve) but also that females grow faster and reach larger weights than the males despite investing more in reproduction (see also text and Table 6). Graph based on length growth parameters, a length-weight relationship, and length at first maturity for Alaska pollock (Gadus chalcogrammus) in FishBase (, which contains hundreds of similar datasets.

  • Fig. 4 Growing fish mature when their relative gill surface area reaches a threshold.

    (A) In the ontogeny of fish, when their relative gill surface area declines, their oxygen supply declines as well; when the latter reaches 1.3 to 1.4 times the oxygen supply required for maintenance and routine activities, i.e., as fish increasingly get “out of breath” (and suffer from hypercapnia), the hormonal cascade is initiated that leads to gonad maturation and spawning. (B) If the same fish are in a stressful, e.g., warmer environment, causing oxygen demand to be elevated, the same 1.3 to 1.4 threshold will cause them to mature and spawn at smaller sizes. (C) Plot, whose 56 points represent the 34 fish species, ranging from guppies to tuna (87) (see the Supplementary Materials) used to estimate the average threshold value of 1.36 (with 95% confidence interval of 1.218 to 1.534). (D) Same plot but for different populations of redband trout (Oncorhynchus mykiss). (E) Ditto for Yellowstone cutthroat trout (Oncorhynchus clarkii). (F) Ditto for mountain whitefish (Prosopium williamsoni).

  • Fig. 5 Fish, at higher temperatures, tend to grow fast toward smaller maximum sizes.

    (A) “Observed phenomenon” that needs to be explained [adapted from an insert in figure 1 of (102)]. (B) Simplified version of figure 1 in (25). (C) Atlantic cod (G. morhua) has wide geographic and temperature ranges; in Eastern Iceland (1° to 10°C), they reach much larger sizes than in French waters (8° to 18°C), based on data in (138, 139).

  • Fig. 6 Simplicity versus scope in explaining why higher temperatures lead to smaller sizes.

    The six explanatory models are adapted from (102) and were presented in two columns, as “intrinsic mechanisms” (A to D) and “extrinsic mechanisms” (E and F). Here, they are arranged according to the perceived complexity of the mechanism(s) they require (abscissa) and their generality or “scope” (ordinate).

  • Table 1 Some physiological and related differences between young and older WBE.

    Here, item (1) is the cause of all others. “Relative” stands for “per unit weight.”

    No.Young/small WBEOlder/larger WBESource(s)
    1Relative gill surface area is high.Relative gill surface area is low.See text
    2Thus, relative O2 supply is high.Thus, relative O2 supply is low.See text
    3Growth in weight is accelerating.Growth in weight is decelerating.See Figs. 1 and 3
    4High temperatures and thus
    shallower habitats are preferred.
    Low temperatures and thus deeper
    habitats are preferred.
    (38, 140)
    In Cyprinodon macularius, a
    temperature of 30°C is “optimal
    only up to an age of 22-28 weeks.”
    In older C. macularius, “the
    temperature optimum shifts
    gradually to 22-26°C.”
    (92); see text for other species
    5Relative food consumption is
    Relative food consumption is low.(141)
    6Food conversion efficiency is high.Food conversion efficiency is low,
    trending toward zero.
    (108, 109, 142)
    7Young adult fish may skip spawning,
    but spawn during the next season.
    Adults do not skip spawning; large
    adults may spawn repeatedly in a
    spawning season.
    (16, 143145)
    8Enzymes in tissues are mainly
    Enzymes in tissues are mainly glycolytic.(53); also see text
    9Fish otoliths contain proteinaceous
    The external layers of fish otoliths are
    purely crystalline.
    (146, 147)
    10Clear daily “rings” are formed in
    otoliths of fish and statoliths of
    Daily “rings” in otoliths or statoliths
    are blurred and, later, disappear
    (16, 148, 150)
    11Extensive seasonal migrations are
    not undertaken.
    Extensive, often temperature-driven
    seasonal migrations are
    12Fat content is low.In fish, fat content is high, particularly
    when seasonal temperature
    oscillations are high.
    (16, 97)
  • Table 2 Arguments raised against the GOLT: Claims by Lefevre and associates.

    2.1Fish could, if they needed it, grow new gill
    lamellae to maintain the ratio gill surface area/
    body weight constant, but they do not need to,
    i.e., “gill surface area can scale proportionally
    with body mass, and if it does not do so, it is
    because oxygen demands are reduced with
    body size…” (42)
    It is actually impossible, for gill lamellae, which
    must function as 2D surfaces (Fig. 2), to keep up
    with the growth of the 3D bodies they supply
    with oxygen (152). In addition, the suggestion
    that large fish could increase their gill surface if
    they wanted, but somehow do not, makes
    these claims effectively unfalsifiable.
    2.2“Weatherley and Gill (153) […] had already
    concluded that there was no evidence that
    capacity for gas exchange or gill surface area
    could limit growth performance in fishes…”
    The quote in question (153) was actually “[t]here is
    little doubt that the relative size of the gills may
    be important in influencing growth and size of
    fish, but Pauly’s claim that his hypothesis ‘offers
    a single, simple explanation to a whole set of
    growth related phenomena…’ seems
    extravagant.” Thus, it is the scope of gill
    limitation that was disputed, not the idea itself.
    2.3Here is another version of the above citation: “…
    Blier et al. (154) had already concluded that
    there was no evidence that capacity for gas
    exchange or gill surface area could limit growth
    performance in fishes, and their analysis remain
    valid today”.
    No, it is no longer valid. Following an exchange
    with P. U. Blier, he conceded that “under natural
    conditions, particularly when fish have to move
    at the same time as they feed or digest, it is very
    probable that aerobic scope, i.e., the oxygen
    supply through the gills, acts as a limiting
    factor” (pers. comm., 16 March 1998, translated
    from French) (15).
    2.4Lefevre et al. (155) asserted that “Pauly and
    Cheung (17) seem to suggest that when the gill
    area grows, it will eventually deplete the water
    of oxygen, and more surface area will be
    useless. However, an increase in body and gill
    size will of course coincide with a proportional
    increase in water and oxygen movement, so a
    doubling of surface area effectively doubles the
    capacity for oxygen uptake.”
    They did not. What was suggested (17) is that the
    growth of gill surface area can proceed only by
    making the gill “sieve” higher and wider (2D)
    but not deeper (3D), as this would put the new
    gill lamellae behind the first layer of lamellae.
    Lamellae that were so placed would be
    “useless,” as the first layer of lamellae would
    reduce the water flowing across the gills of O2,
    leaving the second layer (and any subsequent
    layer) with little to nothing to do (see Fig. 2).
    2.5“…a fundamental pillar of the GOLT – that
    geometrical constraints hinder the gills and
    their surface from growing at the same pace as
    the fish body – is not supported by existing
    data and knowledge” (155)
    On the contrary, the GOLT has the support of an
    immense amount of data, stemming from
    numerous anatomical studies, physiological
    experiments, and ecological surveys. The points
    are that this evidence had never been
    assembled into the coherent picture that the
    GOLT provides and that this picture requires a
    rethink of old assumptions.
  • Table 3 Arguments raised against the GOLT: Issues regarding gill surface areas.

    3.1It was asserted (42) that “in morphometric studies
    where both total lamellae area and gill mass
    have been measured, a linear scaling
    relationship (scaling exponent of 1.0) has been
    found in fishes (43) as well as bivalves (44).
    Consequently, there is no geometric constraint
    that prevents an increase in body size (mass or
    volume) from being accompanied by a
    corresponding increase in gill mass and hence
    respiratory surface area. In other words, gill
    surface area can scale proportionally with body
    mass and, if it does not do so, it is because
    oxygen demands are reduced with body size.”
    Several meta-analyses of gill surface area, covering
    hundreds of fish species exist; they report
    scaling exponents ranging overwhelmingly
    from 0.7 to 0.9 (27, 58) and mention the
    difficulties in obtaining accurate values when a
    small range of body sizes are included (156).
    Thus, the value of 1.04 mentioned here is not
    representative of fish in general and a likely
    overestimate, due to the largest specimen
    considered being only 12% of the maximum
    weight reported in L. unicolor (see The scaling exponent between
    gill surface area and bivalve body weight
    appears to range from 0.51 to 0.80 (58, 157),
    with 0.85 in S. velum (44). The scaling exponent
    of 1.0 linking gill surface area to gill mass in
    S. velum is irrelevant to the O2 supply to its
    body. Also note that the last sentence of the
    argument precludes falsification.
    3.2The presence of very large fish in warm tropical
    waters, e.g., Goliath groupers (Epinephelus
    itajara and Epinephelus quinquefasciatus),
    sunfishes (Mola mola), billfishes and other
    scombroids, giant manta ray (Manta birostris),
    and especially the largest extant fish, the whale
    shark (Rhincodon typus), refutes the GOLT,
    which postulates that high temperatures tend
    to reduce the size of fish (42). [This issue was a
    genuine challenge to the GOLT, and its
    successful resolution (see adjacent column and
    main text) widened its scope.]
    Following an extensive review of the biology of
    the species in question (16), it concluded that
    rather than being invalidated by large fishes
    occurring in the tropics, the GOLT can be used
    to classify their response to the challenge that
    high temperatures pose to their metabolism.
    Thus, in addition to breathing air, as often
    occurs in tropical freshwater fishes, three types
    of increasingly complex adaptations occur,
    none mutually exclusive: (i) placid behavior,
    combined with ambush predation (e.g.,
    groupers) or filter-feeding (e.g., whale shark); (ii)
    yo-yo–type swimming between the warm
    surface and colder, deeper water layers and
    feeding mainly near the surface (bluefin tuna
    and whale shark) or at depth (swordfish and
    billfish), the latter cases involving heating
    systems to keep their huge eyes and brain
    warm; and (iii) huge anatomical changes from
    the ancestral fusiform shape, turning the body
    into a shell around a cavernous mouth and
    oversized gills (giant manta ray) or a mass of
    inert jelly surrounding specialized locomotory
    muscles (M. mola).
    3.3Squid respire through their skin; moreover, by
    having tubular bodies, squid have such large
    respiratory area that they cannot be O2-limited
    (158). In addition, their changed shape as they
    grow increases the surface area of their body
    Squid do not breathe though their skin (159), and
    even if they did, it would not matter because
    their body surface (even when multiplied by 2
    because of their tubular nature and even after
    changing from roundish to lanceolate in the
    course of their ontogeny) is much smaller than
    that of their gill surface area.
    3.4The demonstrably asymptotic growth of Growing
    Sealife plastic squids implies that asymptotic
    growth does not require a limiting surface (160).
    A detailed analysis of what occurs in plastic squids
    that “grow” when placed in water shows that,
    actually (and surprisingly), it is a surface that
    limits their growth (16, 33).
  • Table 4 Arguments raised against the GOLT: Mistaking cause and effect.

    4.1It was suggested (42)
    that because “the
    activity of
    oxidative enzyme
    falls with body
    mass in fishes (53),”
    larger/older fish
    need less oxygen
    anyway. Thus, it is
    not necessary to
    maintain a high O2
    This is mistaking
    cause and effect:
    Fish shift from
    oxidative to
    glycolytic enzymes
    because their
    relative O2 supply
    declines. This was
    well understood by
    earlier authors (57),
    who attributed the
    shift from oxidative
    to glycolytic
    enzymes, if
    tentatively, to “[l]
    imitation on
    [which] may derive
    from surface-
    4.2Lefevre et al. (42)
    wrote “In our field,
    it is generally
    accepted that a
    species’ oxygen
    determines the
    size of their [sic]
    respiratory surface
    area, not the other
    way around.”
    Something being
    accepted” within
    one’s field is not
    evidence of its
    validity. Thus, e.g.,
    plate tectonics was
    not mentioned in
    geology textbooks
    and generally not
    accepted by
    geologists… until
    it was (163).
    4.3There is “a large body
    of evidence
    that respiratory
    surface areas in
    fishes reflect
    metabolic needs,
    not vice versa,
    which explains the
    large interspecific
    variation in scaling
    of gill surface
    areas” (42).
    There is no such body
    of evidence.
    Rather, the O2
    consumption of
    fish is generally
    assumed to reflect
    their “needs.” What
    is missing are tests
    of whether the
    supply of O2 by the
    gills to the body
    (always) satisfies
    the O2 demand of
    the fish tissues. It
    does not in large
    adult fish, which is
    the reason why
    they switch from
    oxidative to
    glycolytic enzymes
  • Table 5 Arguments raised against the GOLT: Different definitions of anabolism and catabolism.

    5.1Von Bertalanffy’s hypothesis of a surface-limiting fish
    growth (which is a key element of the GOLT) is wrong
    because the absorptive surface area of the gut is not in
    permanent contact with food (162, 163).
    Von Bertalanffy (1924) did not commit himself to stating
    that the surface-limiting growth was that of the gut. He
    thought that “the actual surface responsible for growth
    of an organism is in general unknown” (20). However, he
    clearly favored a link to respiration (albeit without
    explicitly mentioning gill surface area).
    5.2The claim was also made that “apparently, it was overlooked
    that although catabolic processes are going on all over
    the body, the necessary oxygen supply has to be
    introduced through some surface or the other, mainly
    the gills. With our basic assumption of isometric growth,
    this 2/3 means that catabolism is proportional to w2/3
    This was not overlooked. In the GOLT, the catabolic
    processes “going on all over the body” do not require
    oxygen. They consist of the (temperature dependent)
    spontaneous denaturation (equal to loss of the
    quaternary structure) of protein molecules. This process
    is proportional to weight; the denatured proteins must
    be resynthesized, which requires ATP and hence O2.
    However, this is part of anabolism, not catabolism.
    5.3Another claim (164) was “…anabolism is proportional to the
    area of the circulatory network rather than to gill surface
    area (35).”
    If this were correct, then the scaling factor of anabolism to
    weight in fish and invertebrates would always be 0.75.
    This, however, is emphatically not the case (15, 16, 165).
    5.4A critique (166) of (7) included “Methodological
    shortcomings include (i) assimilated consumption (the
    ‘anabolic’ part of the growth equation) is assumed to be
    proportional to oxygen, but oxygen is only a limiting
    factor for growth not a controlling factor, i.e. it only affect
    growth if the oxygen concentration is below a critical
    value (167).”
    The response (7) was that “[w]hile Brander et al. cite Brett
    (167) to suggest that oxygen is a limiting factor for
    growth, and not a controlling factor, there is abundant
    theoretical and empirical support in the peer- reviewed
    literature for oxygen being both a limiting and
    controlling factor for the growth of fish and aquatic
    invertebrates.” (14, 93, 168172).
  • Table 6 Arguments raised against the GOLT: Spawning versus growth and vice versa.

    6.1Old/large adult fish stop growing because all their
    energy goes to reproduction (7077)
    Well-fed, non-reproducing fish (e.g., in aquaria)
    stop growing at some point. In addition, the
    females of >80% of fish species grow to be
    larger than the males (see
    and section on “Fish growth vs. reproduction”).
    6.2“Pauly’s assumption that female fish have higher
    reproductive output than male fish is
    unsupported by data. There is no pattern of
    female fish investing more in reproduction than
    males in fish (or other water-breathing
    ectotherm Parker et al. (85). Indeed, for the
    species given by Pauly (84), females invest
    relatively less in reproduction than males as a
    proportion of body mass (see figure 5.5 in
    Sarre’s doctoral dissertation (173)” (83). Note
    that “figure 5.5” is a plot of ova stages versus
    body weight in female (only) black bream
    (Acanthopagrus butcheri), which does not deal
    with the female-to-male comparison at hand; it
    is likely that the authors meant figure 5.6, which
    compare the gonosomatic index (GSI) of
    females and male black breams. In addition, in a
    context similar to that above, an author (174)
    proposed the ad hoc hypothesis that the
    greater reproductive investment of the female
    is more apparent than real, i.e., “[t]he male
    gonad often weighs less than the female
    gonad. This does not mean smaller spawning
    loss in males because sperm, consisting almost
    entirely of DNA, RNA and lipoids, is likely to be
    the most expensive substance in the fish body.”
    A review of 168 mammal, 97 bird, 3 reptile, 100
    amphibian, 98 fish, and 16 invertebrate species
    (175) concluded that, overall, the cost of
    reproduction, in female was up to three orders
    of magnitude higher than for males. This
    confirms Gould (176), who wrote “[s]perm is
    small and cheap, easily manufactured in large
    quantities by little creatures. A sperm cell is
    little more than a nucleus of naked DNA with a
    delivery system. Eggs, on the other hand, must
    be large, for they provide the cytoplasm (all the
    rest of the cell) with mitochondria […]), and all
    other parts that a zygote needs to begin the
    process of embryonic growth….” Parker et al.
    (85) state in their abstract, that sessile
    invertebrates (not “fish”) are “subject mainly to
    selection on gamete production and gamete
    success and so high gonad expenditure is
    expected in both sexes. […]When GSI is
    asymmetric, female GSI usually exceeds male
    GSI, as least in echinoderms. […] Intriguingly,
    higher male GSI also occur in some species […]
    of gastropod molluscs”. If these authors had
    found that male GSI routinely matches that of
    females, they would not have used the word
    ‘intriguingly’. They also note that their “limited
    data also suggest that higher male GSI may be
    the prevalent pattern in sperm casters (where
    only males release gametes).” As for figure 5.6 in
    Sarre’s unpublished thesis, it shows male GSI to
    be occasionally higher than female GSI, but GSI
    is an index relating gonad weight a given time
    to the weight of the body, not the rate of
    production of gonad tissue, which alone relates
    to reproductive costs.
    6.3A critique (166) of (7) included “the bioenergetic
    model assumes that the term scaling directly
    with weight is due to catabolism, but the there
    is a strong case that reproductive investment is
    the principal factor (75, 177, 178).”
    The answer to this (7) was that “Brander et al.
    argue that fish growth is inversely proportional
    to reproductive investment. However, this […]
    cannot explain why female fish (which have a
    much larger reproductive investment than
    male fish) reach larger sizes than male in the
    majority of fish species, and why sterile fish […]
    grow asymptotically. Moreover […] diploid
    (reproductively active) and triploid (sterile) fish
    show very similar growth patterns despite large
    differences in reproductive investment (80).”
  • Table 7 Theoretical versus empirical predictions of weight at fish maturity.

    The “theoretical” predictions of Wm based on the GOLT (Eq. 9) match the empirical estimates based on Eq. 10 (90); the relationship of Wm to the inflexion (Wi; Eq. 7) of weight growth curves is also as predicted (see text).

    W (g)*dWm (% of W)
    Eq. 10
    Wm (% of W)‡
    Eq. 9
    Wi (% of W)
    Eq. 7
    2≈20.080.64746 (3460)28
    2106.6100.72935 (2451)30
    3100521040.81421 (1137)33
    410004121070.974 (1–14)34

    *Assuming the length-weight relationship W = 0.01·L3, corresponding to a trout-shaped (i.e., “average”) fish when in cm and g and used for L and W, respectively.

    †Estimated from W (g) and d ≈ 0.6742 + 0.03574·logWmax in (14, 16), with WWmax.

    ‡The range (in brackets) corresponds to the 95% confidence interval of A = 1.365, i.e., 1.218 to 1.534.

    §The first row values in italics are meant only to illustrate the behavior of Eqs. 7, 9, and 10 for very small sizes. Such fishes are usually iteroparous, and hence, their LmL and their WmW (see text).

    • Table 8 Arguments raised against the GOLT: Miscellaneous, mainly normative arguments.

      7.1O’Dor and Hoar (158) claimed that “There is a
      fundamental flaw in examining Pauly’s surface
      area limited growth scheme by plotting two
      different sets of units (m2 and m3) on the same
      graph and then making quantitative conclusions.
      Not only is [the resulting figure] messy, it violates
      a rule of physics and engineering (179).” The rule
      alluded to here is probably “For an equation to
      have any applicability to the real world, not only
      must the two sides by numerically equal, but they
      must also be dimensionally equal” (179).
      The GOLT involves no equation with
      dimensionally unequal sides. Its
      presentation, however, may include
      graphs with two ordinates axes with
      different units, as illustrated on figure 6.8,
      p. 96 of the reference cited here (179). This
      reference is therefore not likely to have
      suggested that such figures violate the
      rules of physics and engineering. In fact,
      plots with two (or more) ordinate scales
      are common in science (180). The key
      issue, in any case, is that anything
      proportional to the third power of length
      will outgrow anything that remains
      proportional to a lower power of length,
      whatever the units and the starting values.
      7.2It was claimed (42) that in in the contribution of
      Cheung et al. (7), the GOLT predicted a strong size
      reduction of fish with temperature because a key
      parameter was deliberately set too low (d = 0.7)
      When the parameter in question was set at
      higher values (d = 0.8 to 0.9), the size
      reduction caused by increasing
      temperature actually increased (54).
      7.3That ecophysiological consideration should not be
      used to explain physiological processes was
      asserted in a contribution (181) that criticizes
      Pörtner et al.’s “oxygen and capacity limited
      temperature tolerance” (OCLTT) hypothesis, which
      partly overlaps with the GOLT (169, 182, 183).
      No biological subdiscipline can assume a
      priori a monopoly in answering a specific
      scientific question. In fact, scientific
      problems are nowadays best tackled using
      interdisciplinary approaches (184). Pörtner
      et al. (183) suggest that “to connect closely
      to ecological changes, studies need to
      consider the long-term consequences of
      subtle functional constraints. […] Indeed,
      such requirements are rarely met in purely
      physiological studies.”
      7.4Jutfelt et al. (181) suggest that Pörtner et al.’s OCLTT
      hypothesis “incorrectly [considers] aerobic scope
      or oxygen delivery capacity as the ‘energy’
      available to animals, when in fact it is only a
      permissive factor compared with other constraints
      (e.g., food availability).”
      Animals, including fish, deprived of oxygen
      die within minutes. In addition, the
      chemical energy embodied in their food
      becomes available to them only when that
      food is combined with oxygen, i.e., burnt.
      Thus, considering oxygen to be one of
      several “permissive” factors of metabolism
      to score a few points against a colleague
      takes us back to the times before the
      discoveries of Lavoisier (1743–1784).
      7.5Here is another argument against Pörtner et al.’s
      OCLTT hypothesis “it is hard to imagine why
      animals would allow tissue hypoxia to become
      severe enough to inflict performance decline at
      moderate levels of activity when possessing the
      functional capacity to significantly increase
      oxygen delivery to tissues” (181).
      That none of the 28 authors of that
      contribution could imagine why animal
      cannot operate all the time at peak
      performance is itself hard to imagine, but
      it bears repeating here: Peak performance
      extracts a massive toll on all organ systems
      and is used only to escape predators or
      life-threatening situations (17, 185).
      Repeated peak performance, as forced in
      experiments, renders the tested animals
      unfit for life in the wild.
      7.6The closing argument (42): “The idea that
      insurmountable geometric constraints on the size
      of the gills could determine the metabolic rate of
      fishes has never, as far as we know, been pursued
      as a valid hypothesis among respiratory
      physiologists. It is for example not mentioned in
      Schmidt-Nielsen or in Evans and Clairborne, two
      sources for overviews of animals and fish
      This meta-argument about the authority of
      textbooks (186, 187) is a strange one to
      make in the 21st century, although it could
      have been made in the Middle Ages with
      reference to species not mentioned in
      Aristotle’s Historia Animalium (188) or in
      the writings of Plinius the Elder (189).

    Supplementary Materials

    • Supplementary Materials

      The gill-oxygen limitation theory (GOLT) and its critics

      Daniel Pauly

      Download Supplement

      This PDF file includes:

      • Supplementary Text
      • Table S1
      • References

      Files in this Data Supplement:

    Stay Connected to Science Advances

    Navigate This Article