ArticlePDF Available

Abstract and Figures

Consumption rates are the foundation of trophic ecology, yet bioenergetics models used to estimate these rates can lack realism by not incorporating the ontogeny of diet. We constructed a bioenergetics model of a marine predatory fish (tailor, Pomatomus saltatrix) that incorporated high-resolution ontogenetic diet variation, and compared consumption estimates to those derived from typical bioenergetics models that do not consider ontogenetic diet variation. We found tailor consumption was over- or under-estimated by ~5-25% when only including the most common prey item. This error was due to a positive relationship between mean prey energy density and predator body size. Since high-resolution diet data isn't always available, we also simulated how increasing dietary information progressively influenced consumption rate estimates. The greatest improvement in consumption rate estimates occurred when diet variation of 2-3 stanzas (1-2 juvenile stanzas, and adults) was included, with at least 5-6 most common prey types per stanza. We recommend increased emphasis on incorporating the ontogeny of diet and prey energy density in consumption rate estimates, especially for species with spatially segregated life stages or variable diets. A small-moderate increase in the resolution of dietary information can greatly benefit the accuracy of estimated consumption rates. We present a method of incorporating variable prey energy density into bioenergetics models.
No caption available
Content may be subject to copyright.
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
The inuence of ontogenetic diet
variation on consumption rate
estimates: a marine example
Christopher L. Lawson
1,2, Iain M. Suthers1,2, James A. Smith1,2, Hayden T. Schilling
John Stewart3, Julian M. Hughes3 & Stephanie Brodie1,2
Consumption rates are the foundation of trophic ecology, yet bioenergetics models used to estimate
these rates can lack realism by not incorporating the ontogeny of diet. We constructed a bioenergetics
model of a marine predatory sh (tailor, Pomatomus saltatrix) that incorporated high-resolution
ontogenetic diet variation, and compared consumption estimates to those derived from typical
bioenergetics models that do not consider ontogenetic diet variation. We found tailor consumption
was over- or under-estimated by ~5–25% when only including the most common prey item. This error
was due to a positive relationship between mean prey energy density and predator body size. Since
high-resolution diet data isn’t always available, we also simulated how increasing dietary information
progressively inuenced consumption rate estimates. The greatest improvement in consumption
rate estimates occurred when diet variation of 2–3 stanzas (1–2 juvenile stanzas, and adults) was
included, with at least 5–6 most common prey types per stanza. We recommend increased emphasis on
incorporating the ontogeny of diet and prey energy density in consumption rate estimates, especially
for species with spatially segregated life stages or variable diets. A small-moderate increase in the
resolution of dietary information can greatly benet the accuracy of estimated consumption rates. We
present a method of incorporating variable prey energy density into bioenergetics models.
Consumption is the basis of trophic ecology, and measuring it accurately is essential for modelling the impact of
consumers and the trophodynamics of ecosystems1. Consumption rates change with ontogeny, and this change
is typically expressed in bioenergetics models using allometric scaling, which denes the change in consumption
rate with body size. However, allometric scaling is not the only factor inuencing the ontogeny of consumption,
and one frequently excluded factor is diet composition. Bioenergetics models can estimate a consumer’s energy
requirements (in joules) but converting this to a consumption rate (in grams) requires information on the prey
types consumed and their energy density. is data is lacking for many consumers and ecosystems, so bioenerget-
ics models oen use a single, common prey item of adults to represent the prey for all individuals of that species
regardless of body size216.
Many consumers have diet shis as they grow due to morphological changes and physiological needs, with
ontogenetic diet variation observed in many freshwater17,18, estuarine19,20, and marine systems21,22. Without
accounting for the ontogenetic variation in diet and its changing energy density, models that accurately estimate
energy requirements (in joules) may be poor at estimating the consumption of prey (in grams) over the consum-
er’s lifetime. Incorporating diet variability into bioenergetics models and estimates of consumption is rare but it
is important to acknowledge this as a source of uncertainty. is is dicult due to a lack of studies measuring the
inuence of ontogenetic diet variation on estimates of consumption.
Measuring the inuence of diet variability is especially important for species with substantial ontogenetic
variation, such as generalist predators, in which juveniles oen target smaller and lower trophic level prey groups
than adults23. Similarly, it is important to consider ontogenetic diet shis for species with a spatially segregated
juvenile phase. Along with higher consumption rates24, spatially segregated juveniles may consume dierent prey
and reside in dierent habitats25,26, and therefore the predatory impact of juveniles may be directed elsewhere and
1Evolution and Ecology Research Centre, School of Biological, Earth and Environmental Sciences, University of New
South Wales, Sydney, NSW, 2052, Australia. 2Sydney Institute of Marine Science, Chowder Bay Road, Mosman,
NSW, 2088, Australia. 3New South Wales Department of Primary Industries, Sydney Institute of Marine Science,
Chowder Bay Road, Mosman, NSW, 2088, Australia. Correspondence and requests for materials should be addressed
to C.L.L. (email:
Received: 30 November 2017
Accepted: 14 June 2018
Published: xx xx xxxx
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
in greater proportions than adults. To understand how consumption by multiple age groups of the same species
may impact their respective ecosystems, size-specic consumption rates are needed that incorporate not just
metabolic scaling, but the ontogeny of diet composition and its energy density.
e goal of this study was to quantify the change in consumption rate estimates when including ontogenetic
diet and prey energy density variation in bioenergetics models. To achieve this, a preliminary goal was to develop
a bioenergetics model parametrised by respirometry experiments to estimate consumption rates of a marine
predatory sh (tailor, Pomatomus saltatrix). We used a high resolution dietary analysis of this species27, plus
measured energy density of numerous prey types, to incorporate the predator’s ontogenetic diet variation into
calculations of its consumption. is full model was compared to models using prey energy density from an indi-
vidual prey item (a typical approach). Comprehensive diet data is not oen available, so we also simulated how
increasing dietary information (i.e. number of prey measured, number of consumer life stages or ‘stanzas’) pro-
gressively inuenced consumption rate estimates, to identify the approximate resolution of dietary information
required to achieve acceptable accuracy in consumption rate estimates.
Metabolic rate. Respirometry experiments determined the mass- and temperature-dependent resting met-
abolic rate (RMR; gO2 g1 d1) of tailor. e mass-dependent respirometry experiment showed tailor RMR at
24 °C decreased with increasing body mass and was best described by a negative power curve (R2 = 0.54, n = 24;
Fig.1A). e temperature-dependent respirometry experiment showed tailor RMR increased exponentially with
water temperature (R2 = 0.79, n = 61; Fig.1B).
Consumption rate estimates using ontogenetic diet information. Bomb calorimetry measured the
energetic density of tailor and ten common prey items. Sardinops sagax (sardine) was the most energy-dense
prey item (6.84 kJ g1), while Hyperlophus vittatus (sandy sprat) was the least energy-dense sh prey (4.42 kJ g1;
Table1). e only invertebrate tested, Metapenaeus macleayi (school prawn), had the lowest energy density of all
prey items measured (3.84 kJ g1; Table1).
e diet of tailor changed with ontogeny from a predominately invertebrate-based diet to a largely piscivorous
diet (Fig.2)27, with the mean energy density of typical prey species increasing asymptotically with predator size
(Fig.3). Juvenile tailor exhibited signicantly higher mass-specic consumption than adults (Fig.4). At one year
Figure 1. (A) Mass-specic resting metabolic rate (RMR) of tailor (n = 24) with increasing sh mass, at 24 °C.
A linear regression of these data with both axes logged denes the parameter RA (the intercept) and parameter
RB (the slope) from equation (3) (Table2). (B) Mass-specic resting metabolic rate (RMR) of tailor (n = 12 per
treatment) with increasing water temperature. e slope of logged values is the parameter RQ (equation (4),
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
of age, tailor consumed 5.7% of their body weight in prey daily, or an annual Q:B of 20.6 (Fig.4). Aer 4 years, the
daily consumption stabilised at 1.5–2.8% body weight (mean 2.1%), or a Q:B of 5.5–10.1 (mean 7.7), with cyclical
variation caused by seasonal uctuations in the mean daily water temperature of the study area, Sydney Harbour
Australia (Fig.4).
Inuence of ontogenetic diet information. In comparison with our bioenergetics model that included
size-structured diet information, consumption rate estimates that included only the energy density of the most
common prey (sardine, high energy density) underestimated tailor consumption (g) by 26.7% in their rst year,
and by 8.4% over their lifetime (7 years). Using the second most common prey (anchovy, low energy density)
underestimated tailor consumption in their rst year by 3.6% but overestimated lifetime consumption by 20.4%.
In terms of mass-specic consumption rates, only including the energy density of sardines underestimated the
Q:B of a 1 year old sh by 20.1% and of a 4 year old sh (when Q:B stabilises) by 6.6%, while only including
anchovy energy density overestimated the Q:B of a 1 year old sh by 5.1% and of a 4 year old sh by 22.8% (Fig.4).
Parameter description Symbol Va lue s.d Units Equation Source
Proportion of ingested energy lost to
egestion, excretion, and digestion A0.319 0.068 — 1 Derived from
Hartman and Brandt5
Energy density of P. saltatrix Fj 7057.1 1316 J g12Measured
*Von Bertalany g rowth curve parameter t00.119 0.07 69,71
Von Bertalany g rowth curve parameter k0.31 0.015 — 69
Von Bertalany g rowth curve parameter L81.5 0.75 cm FL 69
Mass-dependent intercept of metabolic rate RA0.0047 0.0001 — 3 Derived
Mass-dependent gradient of metabolic rate RB0.2406 0.047 3 Derived
Activity multiplier ACT 1.881 0.502 — 3 5
Oxy-caloric coecient oxy 14140 0.135 J gO21380
Temperature-dependent gradient of
metabolic rate RQ0.091 0.009 — 4 Derived
Energy density function coecient EWA1.293 — 6 Derived
Energy density function constant EWB1.361 — 6 Derived
Table 2. Summary of parameter mean values and standard deviations (s.d) used in the bioenergetics model.
*e von Bertalany growth equation was taken from USA P. saltatrix69, and the t0 parameter was modied
here (originally t0 = 0.3) to better represent the juvenile phase of Australian P. saltatrix in the estuary71.
Prey item Energy Density
(kJ g1) s.e.m Source
Pomatomus saltatrix 7.06 0.42 is study
Sardinops sagax 6.84 0.33 is study
Hyporhamphus regu larisa5.81 0.08 is study
Trachurus novaezelandiae 5.78 0.59 is study
Acanthopagrus au stralis 5.75 0.36 is study
Scomber australasicus 5.43 0.34 is study
Liza argenteab5.39 0.19 is study
Gerres subfasciatus 5.20 0.20 is study
Sillago ciliatac5.02 0.10 is study
Hyperlophus vittatusd4.24 0.19 is study
Metapenaeus macleayie3.84 0.12 is study
Mysida 3.26 83,84
Gobiidae 4.26 85
Atherinidae 4.23 86
Engraulidae 5.20 87
Polychaeta 3.06 88
Cephalopoda 3.90 8992
Decapoda (crabs) 2.63 86
Larval sh 4.18 93
Table 1. Summary of mean values and standard error of the mean (s.e.m.) for energy density of representative
tailor prey items measured by bomb calorimetry and used to calculate length-dependent mean energy
density (E) of tailor prey. N = 10 for all species measured in this study. aUsed as a proxy for all members
of Hemiramphidae. bUsed as a proxy for all members of Mugilidae. cUsed as a proxy for all members of
Sillaginidae. dUsed as a proxy for all members of Hyperlophus. eUsed as a proxy for all members of Penaeidae.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
How many prey types, how many stanzas. We conducted two simulations to identify the approximate
resolution of dietary information required to achieve acceptable accuracy in consumption rate estimates. e rst
simulation revealed the number of most common prey items to include in our bioenergetics model to reach a rea-
sonable estimate of true energy density. e simulation showed that for both juveniles and adults, 5–6 prey items
were needed for the mean energy density to be within 5% of the ‘true’ mean (measured from all prey items) and to
Figure 2. e proportion of common prey types consumed by tailor throughout their ontogeny n = 1437;
adapted from27. e “Other Fish” category contains ~30 species of teleosts.
Figure 3. e mean energy density (E) of prey typically consumed by tailor (E, n = 1437 tailor stomach
contents) throughout ontogeny (black dots represent mean values of 1 cm tailor size classes). e solid black line
is the tted curve described by equation (6).
Figure 4. Modelled consumption: biomass ratios (Q:B) of tailor calculated using dierent prey compositions.
A variable prey energy density (solid black), a constant prey energy density based on a 100% sardine diet (high
energy content; dotted red), and a constant prey energy density based on a 100% anchovy diet (low energy
content; dashed blue). Environmental water temperature is overlayed (dash-dot green).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
approach it monotonically (Fig.5A). In both juveniles and adults, these ve most common prey items accounted
for 65% of the total diet (by mass; Fig.5B).
e second simulation showed the inuence of partitioning lifetime consumption into discrete stanzas. In
general, the consumption rate error decreased and approached the full model when more stanzas were used.
Compared to our full model, a single stanza using the weighted mean energy density of the ve most common
prey types underestimated mean juvenile Q:B by 28.9%, but this error declined when two (14.5%) and three
(8.5%) stanzas were used (Fig.5C). Mean lifetime Q:B was underestimated by 10.9% using one stanza, 4.0% using
two stanzas, and 1.1% using three stanzas (Fig.5C). In terms of absolute consumption (as opposed to relative
consumption measured as Q:B), a single stanza using the weighted mean energy density of the ve most common
prey types overestimated total lifetime consumption by 7.7%, whereas total consumption was overestimated by
8.0% when two or three stanzas were used. Total juvenile consumption was underestimated by 13.7% when a
single stanza was used but overestimated when two (3.6%) and three (5.3%) stanzas were used.
Robust consumption rate estimates of consumers are vital for understanding trophic ecology and predator-prey
interactions. is is especially true for managing exploited sh populations or species that migrate across dier-
ent habitats or ecosystems. By constructing a bioenergetics model that included ontogenetic diet variation of a
marine predator, we demonstrated that variation in energy density of prey sources is an important component
that can substantially alter estimated consumption rates. By simplifying diet information when estimating con-
sumption (including mean ‘lifetime’ consumption), acceptable levels of accuracy were achieved (within 1–4%
of full model) when mean energy density was calculated for 2–3 age stanzas using information for the ve most
common prey types per stanza. Although the resolution required will be somewhat species-specic, it is likely this
level is an appropriate starting point for many predatory sh.
The importance of ontogenetic diet information. Assuming a constant prey energy density through-
out the life of a consumer overlooks key aspects of ontogeny, yet a lack of comprehensive dietary information for
Figure 5. (A) e variation in mean prey energy density (kJ g1) against prey number. e weighted mean
energy density for juveniles and adults was calculated with the most common prey items added sequentially.
Red triangles show juveniles (0–1 y), blue circles show adults (>1 y). Horizontal lines represent the actual
weighted mean for each stage (i.e. the weighted mean when all prey items are included). Grey bands show ±5%
from actual weighted mean for each stage. (B) e contribution (% by mass) of prey types to adult (full line) and
juvenile (dashed line) tailor diet, illustrated as accumulation plots. C) e calculated relative consumption (Q:B,
y1) as the number of prey stanzas is increased (1–3), compared to the full model. Each stanza represents the age
class for which a constant prey energy density is calculated (from the 5 most common prey types per stanza).
e lines of the 1, 2, and 3 stanza simulations >1-year overlay. Compared with our full model, mean lifetime
Q:B was underestimated by 11% using one stanza, 4% using two stanzas, and 1% using three stanzas. e x-axis
is truncated at 0.25 and 3 years for clarity.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
many species means this assumption must be frequently made in bioenergetics models. e models we present
here show that incorporating consumer size-dependent prey energy density (based on detailed dietary informa-
tion) is important for improving the accuracy of consumption estimates. It is common in bioenergetics models to
use the energy density of only the most common prey item of adults, and our results showed that doing this (using
sardine) underestimated the consumption rate of juvenile tailor by ~27%, and by ~8% over their lifetime (7 years),
compared with our full model that incorporated predator size-dependent prey energy density. When the second
most common prey item of adults (anchovy) was used, the consumption by juveniles was coincidentally similar to
the full model but the consumption: biomass ratio (Q:B) of adult tailor was even further overestimated (by ~23%).
Without a size-dependant diet analysis, it is reasonable to assume that either of the two most common prey
species could be used as a “one size ts all” value of prey energy density. Yet a large amount of variation in energy
density can exist between the most common prey items and selecting a specic prey from numerous common
items can introduce uncertainty into consumption estimates. Here, the tailor prey with the highest energy density
(7.06 kJ g1) was 268% greater than the prey with the minimum energy density (2.63 kJ g1). e energy density
of prey items is a particularly important parameter in bioenergetics modelling because it acts as a scaler for the
nal consumption estimate (in grams of prey) produced by the model. If the value of energy density is under or
overestimated by 50%, the nal consumption estimate will also be under or overestimated by 50%. ere are few
other parameters in bioenergetics models with such a large impact on nal consumption estimates. For exam-
ple, the growth parameter may comprise up to 30% of total energy expenditure28, but an error in growth rate of
50% will only result in a 15% error in a consumption estimate. With recent advancements in electronic tagging
technology, there is a current focus on gathering accurate eld-based energetic data for use in bioenergetics mod-
els29,30. ese advancements are important for learning about free-ranging animals and could improve the model
presented here by accurately estimating activity costs rather than using a constant activity multiplier that assumes
no dierence in activity levels through ontogeny. However, models may poorly convert highly accurate energy
requirements (in joules) to a useable prey consumption (in grams) if diet variation is not considered.
How many prey types, how many stanzas. Not all models contain a resolved age- or size-structure,
but all estimates of animal consumption (be it adult, lifetime, or size-structure consumption) will benet from
incorporating dietary ontogeny is some way. It will always be better to estimate consumption using averages of
mass-weighted energy densities for multiple prey items, and this can be achieved in some cases with only a few of
the most common prey species (Fig.5A). Great benet will also come from splitting a species into stanzas when
possible31,32 to account for dierences in juvenile and adult prey composition, and, depending on the pattern of
changing prey composition, a great improvement in accuracy of composition may be achieved with only 2–3
stanzas (Fig.5C). is is especially true for juveniles, which can have vastly dierent diets to adults, and improv-
ing the accuracy of juvenile consumption then benets the accuracy of lifetime mean consumption estimates.
Ecosystem models rely on consumption rates; for example, in the Ecopath with Ecosim framework33,34 bio-
mass ux among trophic groups is dened by each group’s Q:B. Ontogenetic dierences are of course acknowl-
edged in these types of models, as are the challenges in representing all trophic impacts from diet ontogeny35.
Although using a single Q:B value for a species or group is most common, a single group can also be split into
stanzas. But even if a species is split into stanzas, the Q:B values for the dierent stanzas are necessarily based
only on a standard allometric scaling relationship34. Given the potential inuence of ontogenetic diet variation on
Q:B, we recommend more emphasis be placed on determining prey energy densities to inform the Q:B values for
groups and stanzas in ecosystem models. Even if a model is structured so that there must be a single Q:B value for
a species or group, this can still be derived from a bioenergetics model that accounts for diet ontogeny by calcu-
lating a biomass-weighted Q:B from a size- or age-based bioenergetics model (e.g. Fig.4).
e degree to which consumption estimates can be improved by using multiple prey items and stanzas will
vary between species and likely depends on two factors. First, the number of prey items required to account
for a sucient amount of the variation in prey energy density will depend on the diversity of a predator’s diet;
e.g. for generalist predators more prey items will need to be sampled to provide an accurate estimate of mean
energy density, compared to consumers with specialist diets. For tailor in this study, reasonable estimates of prey
energy density are achieved with ~5 prey items (Fig.5A), which made up 65% of the total diet diversity (by mass;
Fig.5B). Species with more specialised diets (and less variability in energy density) will need fewer prey items to
account for similar proportions; e.g. copepods alone comprise 87% of the diet of anchovy Engraulis encrasicolus36,
so it would likely be sucient for anchovy to measure only the energy densities of representative copepod species.
Second, splitting consumption into multiple stanzas will improve consumption estimates most for species with
high ontogenetic diet variation, particularly where diet shis are observed (e.g. shiing from invertebrates to
sh in the present study). Here, the Q:B of simulations improved when consumption was calculated from three
stanzas, which aligns with diet information on tailor whereby diet shis cause three ‘clusters’ of prey types27. Diet
shis are commonly seen in marine predators22,37,38, but this is species-specic and the appropriate number of
stanzas useful for estimating consumption will vary between species.
e combination of diet diversity and ontogenetic diet shis will determine the importance of considering a
model like our full model presented here. For example, Australian bass Macquaria novemaculeata have high diet
diversity throughout life but no major diet shis17, mulloway Argyrosomus japonicus have moderate diet diversity
but experience a strong diet shi from mysid shrimp to sh37, and tailor exhibit both high diet diversity through-
out life and two ontogenetic diet shis27. Similarly, the importance of accurate consumption estimates may need
to be emphasised under certain research questions, e.g. contaminate uptake or changes in prey/habitat availability.
Regardless, using multiple prey types in multiple stanzas should be the starting point for calculating prey energy
densities when estimating consumption.
Interestingly, we saw Q:B improve incrementally as consumption was calculated from more stanzas, however
total consumption by mass showed a less clear trend. Total consumption was less accurate when calculated from
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
the three-stanza simulation than the two-stanza simulation. is is because the prey energy density using two
stanzas was more accurate for large juveniles, which account for most of the consumption (by mass) in the rst
year. Conversely, using three stanzas gives a better t of the mean consumption rate for juveniles, and hence the
mean Q:B will be more accurate. As a result, these two metrics provide dierent methods of measuring error. If
measuring total consumption by an individual sh is the primary concern, then the prey energy density used
should aim to be closest to the life stage where the highest absolute consumption occurs. However if population
level consumption is required, an accurate Q:B should be emphasised, because although juveniles have a smaller
individual consumption by mass, they have the highest abundance and total population consumption can peak
in the juvenile phase39,40.
Juvenile diet and consumption rates. e juvenile phase is usually when the greatest change in size,
speed, morphology, hunting ability, and prey selection occurs in predators19,37,41. Population consumption esti-
mates are oen based on adult consumption rates4245 and are thus likely to be underestimating consumption by
juveniles. e simulation presented here illustrates the impact of juvenile predators on the consumption of prey,
with juvenile tailor in the estuary consuming at least 3–10 times that of adults per unit biomass. is heightened
consumption rate of juveniles is a result of consuming less energy-dense prey items, as well as the eect of body
mass on metabolic rate, and increased energy requirements associated with more rapid growth rates24,46.
Including aspects of ontogeny in consumption estimates is particularly important for consumers with spatially
separated life history stages, such as tailor. Consumption rate estimates that do not consider spatially explicit
ontogenetic life history stages may result in errors in consumption rate estimates that are not evenly distributed
throughout the consumer’s entire habitat, but rather will be focused on certain locations. For example, if juvenile
tailor consumption rates are not accounted for in overall models of the species’ consumption, the errors will be
primarily concentrated in estuaries where the juveniles are distributed. Because we know that adult tailor rarely
enter estuaries47, we can quantify the consumption by tailor that is underestimated specically in estuaries if
ontogeny is not considered in models (i.e. the dierence between juvenile and adult consumption rates).
Future directions and conclusions. Including detailed ontogenetic diet and prey energy density informa-
tion in biogenetics models can improve consumption estimates. Even if detailed prey information is lacking for a
consumer, bioenergetics models could still provide multiple outputs of prey consumption corresponding to any
available prey energy density data, as a way of communicating a possible range of prey consumption estimates.
While prey energy density is an important parameter in estimating consumption from a modelling perspective,
its real impact on free ranging animals should not be overlooked. For example, yellown tuna unnus albacares
need to consume 66% more prey to maintain growth rates if feeding solely on cephalopods rather than sh48,
which, in a world of increasing cephalopod abundance49, is important for understanding the trophodynamics of
such a commercially valuable species. Similarly, potential changes to prey items should be considered with respect
to climate change50,51, as many range shiing sh species may move to areas with altered prey availability52.
Ecosystem based sheries management is designed to ensure that human harvest of prey species does not
negatively aect the sustainability of sh populations or compromise healthy ecosystem function53,54. To most
eectively inform management, the consumption requirements of entire cohorts (animals of the same age) or
populations are needed. Individual sh in the simulations presented here are representative of these cohorts, in
that all other sh in the cohort will experience similar shis in diet and Q:B with ontogeny. Consumption require-
ments, such as those calculated here, can be extrapolated to the population level if accurate biomass and mortality
estimates are available. In this way, more realistic bioenergetics models that include ontogenetic variation in diet
can lead to more informed and eective management of exploited sh populations. Hughes et al.2 estimated a
10,000 tonne population of Australian salmon in south-eastern Australia would consume 36,296–48,190 tonnes
of prey annually; removing ~15% of the biomass of its main prey species. Using the error range presented here by
using a single prey energy density, the cohort consumption estimate of Australian salmon becomes 29,557–60,313
tonnes (12–19% of prey biomass).
Other properties of prey items can have lesser impacts on consumption models, including the assimilation
eciency between dierent prey sources depending on nutritional composition55,56. Including this variation in
bioenergetics models would result in the mean energy density being weighted not only by the dietary proportions
of prey but also by their specic assimilation eciencies. Most consumption models2,28,57 only present a mean
assimilation value for each consumer (including our study, in parameter A). e assimilation eciency of dier-
ent prey items can be determined with feeding experiments58, with the greatest dierences likely to occur between
major food types (e.g. animals versus plants)59,60. It should be noted that a predator may be able to obtain a similar
energy content from its diet while changing the proportions of macronutrients ingested61. Furthermore, dierent
sources of the same macronutrient (e.g. carbohydrates) may represent diering levels of assimilation (e.g. starch
vs cellulose) and this can impact energy ingestion61. Food assimilation in sh is generally considered to account
for 5–20% of total energetic costs62,63, although it has been measured for tunas at 35%64. erefore, variation in
assimilation eciency may represent an aspect of bioenergetics models that can also be improved to increase
accuracy of consumption estimates.
Ontogenetic diet shis may also include consumption of larger individuals within prey species, and energy
density may increase with body size within species65,66. erefore, the dierence in energy density between the
prey of juvenile and adult predators may be greater than presented here and would lead to greater dierences in
consumption estimates between models that include detailed ontogenetic prey energy density and those that
don’t. Similarly, predator energy density may increase with size, meaning that the cost of growth may be lower or
higher than expected for juveniles or adults, respectively. Accounting for size-dependent predator energy density
may inuence absolute consumption estimates for juveniles in bioenergetic models and is an avenue that could
be explored in future studies.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
Although the consumption rate estimates presented here were not validated using laboratory tests, we assume
more detailed data would produce a model that more accurately represents the real world. Additionally, integrat-
ing data from our respirometer experiment further bounds our modelled consumption rates to reality. Our esti-
mates of tailor consumption rates are similar to other similar predators in the ecosystem. Specically, the mean
adult tailor Q:B presented here (7.7) is intermediate compared to other local mesopredators; Australian salmon
Arripus trutta (Q:B 3.2–5.252) and Australian bonito Sarda australis (15.667). e similarities in these Q:B ratios
are related to similarities in sh growth rates, activity levels, and morphologies. Australian salmon is similar to
tailor in diet, size, behaviour and habitat, but is a slower growing species68,69 and Australian bonito is an active,
medium-sized scombrid that is expected to have high energy requirements based on caudal aspect ratio70.
e energy density of prey, and how it changes throughout a consumer’s ontogeny, appears to be an under-
appreciated parameter in bioenergetics modelling. ere are few parameters of a bioenergetics model to which
an estimate of prey consumption would be more sensitive, possibly with the exceptions of allometric metabolic
scaling and temperature (which both have exponential relationships with metabolism). Despite this, consump-
tion models oen overlook variation in prey energy density that arises from consumer ontogeny and habitats,
primarily because this information if oen lacking for marine consumers. Frequently studied species, such as
those of commercial importance, may oer opportunities to further sample diet data if existing diet records
are unavailable. In cases where detailed diet data is unavailable, we suggest considering juvenile and adult con-
sumption separately, and to use what diet data is available to guide a prey composition that will be more repre-
sentative of free-ranging animals than simply choosing a single, common prey item. We recommend developing
species-specic versions of equation (6) presented here to incorporate ontogenetic diet and prey energy density
in consumption models, especially for species with spatially segregated life stages and for species with variable
diets, as the consumption estimates of these species are those most likely to improve with this additional trophic
Materials and Methods
Tailor (Pomatomus saltatrix Linneaus 1766) are a globally distributed generalist predator71, with spatially seg-
regated juveniles which may reside in estuaries for approximately one year before migrating to coastal waters as
adults72. is ontogenetic migration is accompanied by a diet shi resulting from altered prey availability and
increased foraging ability. Tailor can prey on a range of sh and invertebrates and a large amount of variation
exists in their diet27,73.
Measuring metabolic rate. Metabolic rates are the foundation for quantifying the energy requirements of
consumers, and are usually measured using respirometry. We used two respirometry experiments to measure the
eect of body mass and temperature on the resting mass-specic metabolic rate (RMR; g O2 g sh1 d1) of tailor
to parameterise a bioenergetics model. e mass-dependent respirometry experiment determined the relation-
ship between body mass and RMR at 24 °C using 24 tailor consisting of 16 juveniles (11–25 cm fork length (FL),
18–275 g) and 8 adults (33–42 cm FL, 500–1035 g). e temperature-dependent respirometry experiment tested
the eect of temperature on RMR at 18, 21, 24, 27 and 30 °C (±1 °C), with 12 juvenile tailor in each temperature
treatment (10–25 cm FL, mean mass 68.8 ± s.e.m. 6.0 g). Juveniles were used to allow the greatest breadth of trials;
both in terms of temperature range and individuals within temperature treatments. We assumed juveniles and
adults have the same response to temperature, and that any dierence would not change the conclusions regard-
ing dierences in consumption between models with varying levels of prey energy density information.
Tailor were caught in Sydney Harbour, Australia, and transported to aquarium facilities at the Sydney Institute
of Marine Science (SIMS), Chowder Bay (33°5031.6S; 151°1451.18E), and fed daily a diet of frozen sh, ceph-
alopods, and crustaceans. Holding tanks were maintained at the treatment temperatures using bar heaters or an
Oasis EC9bp water chiller (Oasis Heat Pumps Knoxeld, Victoria, Australia). Holding tanks were adjusted 1 °C
each day until the treatment temperature was reached74, and held at this temperature for one week before exper-
iments began75.
e RMR of individual tailor was determined by measuring oxygen consumption in darkened respirometry
chambers. Rectangular chambers of varying size were used, with sh mass to water volume ratios (g:mL) between
1:20 and 1:10076. Prior to the respirometry trials, tailor were fasted for 24 hours and acclimated for 3 hours in the
respirometry chamber77,78. e chamber was sealed from atmospheric air at the beginning of each respirometry
trial. Oxygen concentration was measured by an oxygen meter (Hach HQ40d Loveland, Colorado, USA), and
each respirometry trial ran until the dissolved oxygen concentration reached 80%. Microbial respiration in the
respirometer was measured aer each sh was removed from the chamber and subtracted from the total oxygen
consumed to calculate oxygen consumption by the sh only79.
Two relationships were derived to describe the metabolic rate of tailor. First, the relationship between
body mass (g) and the mass-specic RMR (gO2 g1 d1) of tailor was determined using linear regression of
log-transformed values. is relationship was used to parameterise the bioenergetics model (RA and RB in equa-
tion3 below) and was isolated from the eect of temperature on RMR (RQ in equation4 below), which was
examined separately. Second, a linear regression of log-transformed values was used to determine the eect of
temperature on the mass-specic RMR (gO2 g0.76 d1) of tailor. To isolate the eect of temperature and account
for the eect of body mass on metabolic rate, a scaling exponent (0.76) was used in estimating the relationship
between RMR and temperature. is scaling exponent was informed from the rst, mass-dependent respirometry
experiment and subsequent relationship between body mass and RMR. All statistical models were done using R
statistical computing (v3.3.1; R Core Development Team 2016). All experiments were performed in accordance
with relevant guidelines and regulations under approval by the University of New South Wales Animal Care and
Ethics Committee (No. 15/152B).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
Estimating energy requirements. To demonstrate the dierence in consumption (in grams) given a con-
sumer’s ontogenetic diet variation, the energy requirements (in joules) of the consumer at any one time must
rst be quantied. e energy requirement (C; J d1) of individual tailor was estimated using a bioenergetics
model that incorporated the mass- and temperature-dependant metabolic rates determined from the above
respirometry experiments. e bioenergetics model used was based on the energy balance model of Kitchell, et
al.28 (Supplementary Note):
where C is energy requirement (J d1), R is energy required for metabolism (J d1), G is energy allocated to daily
sh growth (J d1), and A is the proportion of energy that is lost to food digestion, egestion, and excretion (J d1;
Table2). Energy required for daily growth (G; J d1) was determined by:
Δ=GWFj (2)
where ΔW is the daily change in sh mass (g d1) and Fj is the energy density of somatic tissue of tailor (J
g1; Table2). ΔW (for an individual at a given weight) was calculated using the von Bertalany growth equa-
tion69 (Table2), converted to mass from sh length using a length-weight relationship (Weight = 0.0104 Fork
Length3.0824; H. T. Schilling, unpublished data). Fj was estimated from bomb calorimetry analysis (see details
below). Spawning losses for adults were not accounted for in the simulations due to a lack of baseline informa-
tion, however while their inclusion would benet the model in general, it does not aect comparisons between
our models.
e energy required for metabolism (R; J d1) was calculated, using parameters derived from our respirometry
experiments, as:
=RRWfTACT oxy()()
where RA and RB are the intercept and slope, respectively, of the logged mass-dependant function of mass-specic
RMR, W is age-specic sh mass (g), f(T) is a temperature-dependence function, ACT is a constant multiplier
which accounts for active metabolic rate, and oxy is the caloric coecient of oxygen (J gO21) which is com-
monly used in animal energetics80 to convert the units of metabolic rate from oxygen (gO2 d1) to energy (J d1;
Table2). RA and RB were derived from the mass-dependant laboratory respirometry experiment (Table2). W
was derived from the von Bertalany growth curve as detailed above. ACT accounted for the heightened energy
use in active metabolic rate associated with periods of activity by the sh, and is a constant multiplier of R5.
e temperature-dependence function (f(T)) accounted for variation in metabolic rate due to changes in water
fT e() (4)
where RQ is the slope of the logged temperature-dependant function of resting metabolic rate, and T is water tem-
perature (°C). RQ was derived from the temperature-dependant laboratory respirometry experiments (Table2).
Integrating ontogenetic diet information. To demonstrate the inuence of including ontogenetic diet
information in a bioenergetics model, we tracked the energy requirement (J d1) of an individual tailor through-
out its life and converted this to a consumption rate (g d1) dependent on diet composition. A model was used
to track the energy requirement (J d1) through time of a typical tailor recruiting to an estuary: Sydney Harbour,
Australia. As the above energy requirement (C) is temperature-dependent, temperature data from Sydney
Harbour was used to estimate C for the model. Daily water temperature measured over 3 years (2013–2016) at
Chowder Bay was used to generate a seasonal sine function of daily mean temperature for Sydney Harbour (range
15.6–22.1 °C). e primary spawning event for tailor occurs in late winter and spring47, and so the tailor was
assumed to enter the estuary on the rst day of spring (September 1).
e energy required by tailor (C, J d1, equation (1)) was converted to a consumption rate of prey (g d1),
by dividing C by the mean energy density (E, J g prey1) of the prey items. To demonstrate the importance of
including consumer length-dependent prey energy densities in bioenergetics models, we required a relationship
between E and tailor body length. is was done by calculating the mean energy density of prey (E) for every
1 cm tailor size class using detailed length-dependant tailor gut content data (FL 3–76 cm, n = 1437)27. e mean
energy density of tailor prey for each tailor size class was calculated as the average prey energy density weighted
by each prey’s proportional contribution (by mass) to the diet at that 1 cm size class:
where Ex is the weighted mean energy density of prey at predator size x (J g1), PEi is the energy density of prey
item i (J g1), and PPi is the proportion of diet comprised by prey item i at size class x.
Bomb calorimetry (6400, Parr Instrument Company, Illinois, USA) was used to measure the energy density of
the prey, and was done for ten common prey types found in tailor diets. ese ten prey types were used alongside
published sources to estimate the energy density of all prey items (Table1). Samples of prey for bomb calorimetry
were obtained from the Sydney Fish Markets (Sydney, Australia), which were sourced from commercial sheries
close to Sydney. Preparation of prey samples for bomb calorimetry was done as described by the “subsample
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
method” outlined by Glover, et al.81, whereby whole sh samples are homogenized and a small subsample is
burned to determine the mean energy density of the sh.
e mean energy density of prey (Ex, J g1) for the tailor size classes were tted against tailor fork length using
a logarithmic function to create a continuous relationship for use in the model:
=+EEWFLEWln() (6)
where FL is fork length of tailor (cm), and EWA and EWB are constants (Table2). Using this equation, E estimates
the mean energy density of prey which is typically consumed when tailor are a specic size. Equation (6) was used
to determine E for tailor with FL 42 cm. However E was xed at 6300 J g1 for tailor with FL > 42 cm, as this is
the maximum mean weighted energy density of prey items consumed by free ranging tailor.
e ‘consumption: biomass ratio’ (Q:B) metric was used to communicate how consumption estimates change
when using high quality diet and energy density information. Q:B indicates annual consumption by an individual
relative to its biomass, and is a common metric to indicate the trophic impact of a consumer33,82. Q:B was calcu-
lated by expressing consumption rate (g d1) as a proportion of sh mass and multiplying by 365 days to reach
annual consumption relative to body mass (g g1 y1).
To quantify the eect of integrating variable prey energy density with consumer ontogeny in the model, the
consumption calculated from the full model (using consumer-length dependant E data, equation (6)) was com-
pared with two alternative models. Each alternative model used only a single value for E derived from one of
the two most common prey items for adult tailor (Australian sardine Sardinops sagax and Australian anchovy
Engraulis australis). is approach of estimating E from only the most common prey types is typical of numerous
bioenergetics models24. To compare between our model and these two alternative models, consumption was
expressed as the Q:B at 1 and 4 years of age (Q:B stabilised at ~4 years), as well as the total juvenile and lifetime
consumption (in grams) for an individual sh. e energy density value for sardine was derived from bomb cal-
orimetry, and the value for anchovy was taken from published sources (Table1).
How many prey types, how many stanzas. Acknowledging that highly detailed size-structured diet
data is not always available, we also performed two simulations to explore how incremental increases in dietary
information can improve the accuracy of consumption estimates. e rst simulation examined the number of
most common prey items that should be included in a model to reach a reasonable estimate of the true energy
density of prey consumed by tailor. is was done by adding prey items one by one, weighted by proportion in
diet, for juvenile (<1 y) and adult (>1 y) sh separately, and comparing the resulting prey energy density to the
true mean prey energy density consumed (i.e. the weighted mean of all prey items consumed by juveniles or
adults). We also determined the percentage (by mass) of total diet that was accounted for as prey items were
added sequentially for both juveniles and adults of our predatory sh.
e second simulation using less detailed diet information examined the eect of measuring lifetime con-
sumption from multiple stanzas (life history stages), and compared results to our full model. We ran simulations
that used one, two (juveniles and adults; 0–1 y, >1 y), or three (0–0.5 y, 0.5–1 y, >1 y) stanzas, with each stanza
having a single value for prey energy density measured as the weighted mean of the ve most common prey types
of that stanza. Five prey items were used based on the results of the previous simulation that added prey items
sequentially to compare resulting prey energy density with our full model. e resulting estimated consumption
rates, measured as total consumption in grams and mean Q:B, were then compared to results from our full model
where prey energy density was a function of predator size, for both juvenile (rst year) and lifetime consumption.
Both simulations testing the eect of adding stanzas and prey items sequentially were truncated at 40 cm FL, as
inconsistent sampling of gut contents at lengths greater than this introduced variation in the data that was not
indicative of actual consumption.
Data availability. e datasets generated during and/or analysed during the current study are available from
the corresponding author on reasonable request.
1. leiber, M. e re of life. An introduction to animal energetics (second ed.). (.E. rieger Publishing Company, 1975).
2. Hughes, J. M., Stewart, J., Lyle, J. M. & Suthers, I. M. Top-down pressure on small pelagic sh by eastern Australian salmon Arripis
trutta; estimation of daily ration and annual prey consumption using multiple techniques. Journal of Experimental Marine Biology
and Ecology 459, 190–198 (2014).
3. Olson, . J. & Boggs, C. H. Apex predation by yellown tuna (unnus albacares): independent estimates from gastric evacuation
and stomach contents, bioenergetics, and cesium concentrations. Canadian Journal of Fisheries and Aquatic Sciences 43, 1760–1775
4. itchell, J. F., Neill, W. H., Dizon, A. E. & Magnuson, J. J. Bioenergetic spectra of sipjac and yellown tunas. e Physiological
Ecology of Tunas, 357–368 (1978).
5. Hartman, . J. & Brandt, S. B. Comparative energetics and the development of bioenergetics models for sympatric estuarine
piscivores. Canadian Journal of Fisheries and Aquatic Sciences 52, 1647–1666 (1995).
6. Essington, T. E. Development and sensitivity analysis of bioenergetics models for sipjac tuna and albacore: a comparison of
alternative life histories. Transactions of the American Fisheries Society 132, 759–770 (2003).
7. Neer, J. A., ose, . A. & Cortés, E. Simulating the eects of temperature on individual and population growth of hinoptera
bonasus: a coupled bioenergetics and matrix modeling approach. Marine Ecology Progress Series 329, 211–223 (2007).
8. Hartman, . J. & Jensen, O. P. Anticipating climate change impacts on Mongolian salmonids: bioenergetics models for Leno and
Baial grayling. Ecology of Freshwater Fish 26, 383–396 (2017).
9. Beltran, . S., Testa, J. W. & Burns, J. M. An agent-based bioenergetics model for predicting impacts of environmental change on a
top marine predator, the Weddell seal. Ecological Modelling 351, 36–50 (2017).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
10. Borin, J. M. et al. Energetic requirements of green sturgeon (Acipenser medirostris) feeding on burrowing shrimp (Neotrypaea
californiensis) in estuaries: importance of temperature, reproductive investment, and residence time. Environmental Biology of Fishes
100, 1561–1573 (2017).
11. Yoon, S., Watanabe, E., Ueno, H. & ishi, M. J. Potential habitat for chum salmon (Oncorhynchus eta) in the Western Arctic based
on a bioenergetics model coupled with a three-dimensional lower trophic ecosystem model. Progress in Oceanography 131, 146–158
12. Deslauriers, D., Heironimus, L. B. & Chipps, S. . Test of a foraging-bioenergetics model to evaluate growth dynamics of endangered
pallid sturgeon (Scaphirhynchus albus). Ecological modelling 336, 1–12 (2016).
13. Hovel, . A., Beauchamp, D. A., Hansen, A. G. & Sorel, M. H. Development of a bioenergetics model for the threespine sticlebac.
Transactions of the American Fisheries Society 144, 1311–1321 (2015).
14. Plumb, J. M., Blancheld, P. J. & Abrahams, M. V. A dynamic-bioenergetics model to assess depth selection and reproductive growth
by lae trout (Salvelinus namaycush). Oecologia 175, 549–563 (2014).
15. Anderson, . ., Chapman, D. C., Wynne, T. T., Masagounder, . & Pauert, C. P. Suitability of Lae Erie for bigheaded carps based
on bioenergetic models and remote sensing. Journal of Great Laes esearch 41, 358–366 (2015).
16. Armstrong, A. O. et al. Prey density threshold and tidal inuence on reef manta ray foraging at an aggregation site on the Great
Barrier eef. PloS one 11, e0153393 (2016).
17. Smith, J., Baumgartner, L., Suthers, I. & Taylor, M. Generalist niche, specialist strategy : the diet of an Australian percichthyid. Journal
of sh Biology 78, 1183–1199 (2011).
18. García-Berthou, E. Food of introduced mosquitosh: ontogenetic diet shi and prey selection. Journal of Fish Biology 55, 135–147, (1999).
19. Hartman, . J. & Brandt, S. B. Trophic resource partitioning, diets, and growth of sympatric estuarine predators. Transactions of the
American Fisheries Society 124, 520–537 (1995).
20. Baer, . & Sheaves, M. edening the piscivore assemblage of shallow estuarine nursery habitats. Marine Ecology Progress Series
291, 197–213 (2005).
21. Bethea, D. M. et al. Geographic and ontogenetic variation in the diet and daily ration of the bonnethead shar, Sphyrna tiburo, f rom
the eastern Gulf of Mexico. Marine Biology 152, 1009–1020, (2007).
22. Lowe, C. G., Wetherbee, B. M., Crow, G. L. & Tester, A. L. Ontogenetic dietary shis and feeding behavior of the tiger shar,
Galeocerdo cuvier, in Hawaiian waters. Environmental Biology of Fishes 47, 203–211 (1996).
23. Graeb, B. D., Galarowicz, T., Wahl, D. H., Dettmers, J. M. & Simpson, M. J. Foraging behavior, morphology, and life history variation
determine the ontogeny of piscivory in two closely related predators. Canadian Journal of Fisheries and Aquatic Sciences 62,
2010–2020 (2005).
24. Bartell, S., Brec, J., Gardner, . & Brenert, A. Individual parameter perturbation and error analysis of sh bioenergetics models.
Canadian Journal of Fisheries and Aquatic Sciences 43, 160–168 (1986).
25. Bec, M. W. et al. e identication, conservation, and management of estuarine and marine nurseries for sh and invertebrates: a
better understanding of the habitats that serve as nurseries for marine species and the factors that create site-specic variability in
nursery quality will improve conservation and management of these areas. Bioscience 51, 633–641 (2001).
26. Parrish, J. D. Fish communities of interacting shallow-water habitats in tropical oceanic regions. Marine Ecology Progress Series 58,
143–160 (1989).
27. Schilling, H. T. et al. Latitudinal and ontogenetic variation in the diet of a pelagic mesopredator (Pomatomus saltatrix), assessed with
a classication tree analysis. Marine Biology 164, 75 (2017).
28. itchell, J. F., Stewart, D. J. & Weininger, D. Applications of a bioenergetics model to yellow perch (Perca avescens) and walleye
(Stizostedion vitreum vitreum). Journal of the Fisheries Board of Canada 34, 1922–1935 (1977).
29. Cooe, S. J. et al. emote bioenergetics measurements in wild sh: opportunities and challenges. Comparative Biochemistry and
Physiology Part A: Molecular & Integrative Physiology 202, 23–37 (2016).
30. Brodie, S. et al. Improving consumption rate estimates by incorporating wild activity into a bioenergetics model. Ecology and
Evolution 6, 2262–2274 (2016).
31. Fris, M., Miller, T., Latour, . & Martell, S. Assessing biomass gains from marsh restoration in Delaware Bay using Ecopath with
Ecosim. Ecological modelling 222, 190–200 (2011).
32. Walters, C., Martell, S. J., Christensen, V. & Mahmoudi, B. An Ecosim model for exploring Gulf of Mexico ecosystem management
options: implications of including multistanza life-history models for policy predictions. Bulletin of Marine Science 83, 251–271
33. Christensen, V. & Pauly, D. ECOPATH II—a software for balancing steady-state ecosystem models and calculating networ
characteristics. Ecological Modelling 61, 169–185 (1992).
34. Christensen, V. & Walters, C. J. Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling 172, 109–139
35. Pauly, D., Christensen, V. & Walters, C. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of sheries. ICES
Journal of Marine Science: Journal du Conseil 57, 697–706 (2000).
36. B acha, M. & Amara, . Spatial, temporal and ontogenetic variation in diet of anchovy (Engraulis encrasicolus) on the Algerian coast
(SW Mediterranean). Estuarine, Coastal and Shelf Science 85, 257–264 (2009).
37. Taylor, M. D., Fielder, D. S. & Suthers, I. M. Spatial and ontogenetic variation in the diet of wild and stoced mulloway (Argyrosomus
japonicus, Sciaenidae) in Australian estuaries. Estuaries and Coasts 29, 785–793 (2006).
38. Field, I. C., Bradshaw, C. J., van den Ho, J., Burton, H. . & Hindell, M. A. Age-related shis in the diet composition of southern
elephant seals expand overall foraging niche. Marine Bi ology 150, 1441 (2007).
39. Herwig, B. & Zimmer, . Population ecology and prey consumption by fathead minnows in prairie wetlands: importance of detritus
and larval sh. Ecology of Freshwater Fish 16, 282–294 (2007).
40. udstam, L. G. et al. Prey consumption by the burbot (Lota lota) population in Green Bay, Lae Michigan, based on a bioenergetics
model. Canadian Journal of Fisheries and Aquatic Sciences 52, 1074–1082 (1995).
41. Shimose, T., Watanabe, H., Tanabe, T. & ubodera, T. Ontogenetic diet shi of age-0 year Pacic bluen tuna unnus orientalis.
Journal of Fish Biology 82, 263–276, (2013).
42. itchell, J. F., Essington, T. E., Boggs, C. H., Schindler, D. E. & Walters, C. J. e role of shars and longline sheries in a pelagic
ecosystem of the central Pacic. Ecosystems 5, 202–216 (2002).
43. Heymans, J. J., Shannon, L. J. & Jarre, A. Changes in the northern Benguela ecosystem over three decades: 1970s, 1980s, and 1990s.
Ecological Modelling 172, 175–195 (2004).
44. Coll, M., Navarro, J. & Palomera, I. Ecological role, shing impact, and management options for the recovery of a Mediterranean
endemic sate by means of food web models. Biological Conservation 157, 108–120 (2013).
45. Whitehouse, G. A. & Aydin, . Y. Trophic Structure of the Eastern Chuchi Sea: An Updated Mass Balance Food Web Model. (National
Oceanic and Atmospheric Administration, Alasa, 2016).
46. Beauchamp, D. A., Stewart, D. J. & omas, G. Corroboration of a bioenergetics model for soceye salmon. Transactions of the
American Fisheries Society 118, 597–607 (1989).
47. Zeller, B., Polloc, B. & Williams, L. Aspects of Life History and Management of Tailor (Pomatomus saltatrix) in Queensland. Marine
and Freshwater esearch 47, 323–329 (1996).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
48. Wexler, J. B. et al. Tan culture of yellown tuna, unnus albacares: developing a spawning population for research purposes.
Aquaculture 220, 327–353 (2003).
49. Doubleday, Z. A. et al. Global proliferation of cephalopods. Current Biology 26, 406–407,
cub.2016.04.002 (2016).
50. Hughes, J. M. et al. Latitudinal, ontogenetic, and historical shis in the diet of a carnivorous teleost, Arripis trutta, in a coastal pelagic
ecosystem altered by climate change. Canadian Journal of Fisheries and Aquatic Sciences 70, 1209–1230 (2013).
51. Potts, W., Bealey, . & Childs, A. Assessing trophic adaptability is critical for understanding the response of predatory shes to
climate change: a case study of Pomatomus saltatrix in a global hotspot. African Journal of Marine Science 38, 539–547 (2016).
52. Blanchard, J. L. Climate change: A rewired food web. Nature 527, 173–174, (2015).
53. Scandol, J. P., Holloway, M. G., Gibbs, P. J. & Astles, . L. Ecosystem-based sheries management: an Australian perspective. Aquatic
Living esources 18, 261–273 (2005).
54. Hall, S. J. & Mainprize, B. Towards ecosystembased sheries management. Fish and Fisheries 5, 1–20 (2004).
55. Bree, B. & Gabrielsen, G. W. Assimilation eciency of adult ittiwaes and Brünnich’s Guillemots fed Capelin and Arctic Cod.
Polar Biology 14, 279–284 (1994).
56. Lawson, J. W., Miller, E. H. & Noseworthy, E. Variation in assimilation eciency and digestive eciency of captive harp seals (Phoca
groenlandica) on dierent diets. Canadian Journal of Zoology 75, 1285–1291 (1997).
57. ice, J. A., Brec, J. E., Bartell, S. M. & itchell, J. F. Evaluating the constraints of temperature, activity and consumption on growth
of largemouth bass. Environmental Biology of Fishes 9, 263–275 (1983).
58. elso, J. . Conversion, maintenance, and assimilation for walleye, Stizostedion vitreum vitreum, as aected by size, diet, and
temperature. Journal of the Fisheries Board of Canada 29, 1181–1192 (1972).
59.  osen, D. A. & Trites, A. Digestive eciency and dry-matter digestibility in Steller sea lions fed herring, polloc, squid, and salmon.
Canadian Journal of Zoology 78, 234–239 (2000).
60. Castro, G., Stoyan, N. & Myers, J. P. Assimilation eciency in birds: A function of taxon or food type? Comparative Biochemistr y and
Physiology Part A: Physiology 92, 271–278, (1989).
61. Machovsy-Capusa, G. E., Senior, A. M., Simpson, S. J. & aubenheimer, D. e multidimensional nutritional niche. Trends in
Ecology & Evolution 31, 355–365 (2016).
62. Pec, M. A., Bucley, L. J. & Bengtson, D. A. Energy losses due to routine and feeding metabolism in young-of-the-year juvenile
Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 60, 929–937 (2003).
63. Fu, S. J., Xie, X. J. & Cao, Z. D. Eect of meal size on postprandial metabolic response in southern catsh (Silurus meridionalis).
Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 140, 445–451,
cbpb.2005.02.008 (2005).
64. Fitzgibbon, Q., Seymour, ., Ellis, D. & Buchanan, J. e energetic consequence of specic dynamic action in southern bluen tuna
unnus maccoyii. Journal of Experimental Biology 210, 290–298 (2007).
65. Pothoven, S. A., Bunnell, D. B., Madenjian, C. P., Gorman, O. T. & oseman, E. F. Energy density of bloaters in the Upper Great
Laes. Transactions of the American Fisheries Society 141, 772–780 (2012).
66. Canale, . P. & Brec, J. E. Comments on proper (and improper) solutions of bioenergetic equations for modeling sh growth.
Aquaculture 404, 41–46 (2013).
67. Griths, S. P. et al. Ecological eects of longline shing and climate change on the pelagic ecosystem o eastern Australia. eviews
in Fish Biology and Fisheries 20, 239–272 (2010).
68. Hughes, J. e biology and population structure of eastern Australian salmon (Arripis trutta) in south-eastern Australia, Ph. D. esis.
e University of New South Wales, Sydney, (2012).
69. obillard, E., eiss, C. S. & Jones, C. M. Age-validation and growth of bluesh (Pomatomus saltatrix) along the East Coast of the
United States. Fisheries esearch 95, 65–75 (2009).
70. Palomares, M. L. D. & Pauly, D. Predicting food consumption of fish populations as functions of mortality, food type,
morphometrics, temperature and salinity. Marine and freshwater research 49, 447–453 (1998).
71. Juanes, F., Hare, J. A. & Misiewicz, A. G. Comparing early life history strategies of Pomatomus saltatrix: a global approach. Marine
and Freshwater esearch 47, 365–379 (1996).
72. Polloc, B. e tailor (Pomatomus saltatrix) shery at Fraser Island and its relation to the life-history of the sh. Proceedings of the
oyal Society of Queensland 95, 23–28 (1985).
73. Bucel, J., Fogarty, M. & Conover, D. Foraging habits of bluesh. Pomatomus saltatrix, on the US east coast continental shelf. Fishery
Bulletin - National Oceanic and Atmospheric Administration 97, 758–775 (1999).
74. Mora, C. & Maya, M. F. Eect of the rate of temperature increase of the dynamic method on the heat tolerance of shes. Journal of
ermal Biology 31, 337–341 (2006).
75. Barrionuevo, W. & Fernandes, M. Time-course of respiratory metabolic adjustments of a South American sh, Prochilodus scrofa,
exposed to low and high temperatures. Journal of Applied Ichthyology 14, 37–41 (1998).
76. Clar, T. D., Sandblom, E. & Jutfelt, F. Aerobic scope measurements of shes in an era of climate change: respirometry, relevance and
recommendations. Journal of Experimental Biology 216, 2771–2782 (2013).
77. Plaut, I. esting metabolic rate, critical swimming speed, and routine activity of the euryhaline cyprinodontid, Aphanius dispar,
acclimated to a wide range of salinities. Physiological and Biochemical Zoology 73, 590–596 (2000).
78. McDonnell, L. H. & Chapman, L. J. At the edge of the thermal window: eects of elevated temperature on the resting metabolism,
hypoxia tolerance and upper critical thermal limit of a widespread African cichlid. Conservation Physiology 3, 50–53 (2015).
79. Ohlberger, J., Staas, G. & Höler, F. Eects of temperature, swimming speed and body mass on standard and active metabolic rate
in vendace (Coregonus albula). Journal of Comparative Physiology B 177, 905–916 (2007).
80. Elliott, J. & Davison, W. Energy equivalents of oxygen consumption in animal energetics. Oecologia 19, 195–201 (1975).
81. Glover, D. C., DeVries, D. ., Wright, . A. & Davis, D. A. Sample preparation techniques for determination of sh energy density
via bomb calorimetry: an evaluation using largemouth bass. Transactions of the American Fisheries Society 139, 671–675, https://doi.
org/10.1577/T09-110.1 (2010).
82. Polovina, J. J. Model of a coral reef ecosystem. Coral eefs 3, 1–11 (1984).
83. Crowl, T., Bouwes, N., Townsend, M., Covich, A. & Scatena, F. Estimating the potential role of freshwater shrimp on an aquatic
insect assemblage in a tropical headwater stream: a bioenergetics approach. Internationale Vereinigung fur Theoretische und
Angewandte Limnologie Verhandlungen 27, 2403–2407 (2001).
84. Chipps, S. . & Bennett, D. H. Evaluation of a Mysis bioenergetics model. Journal of Planton esearch 24, 77–82 (2002).
85. Healey, M. Bioenergetics of a sand goby (Gobius minutus) population. Journal of the Fisheries Board of Canada 29, 187–194 (1972).
86. McClusey, S. M., Bejder, L. & Loneragan, N. . Dolphin prey availability and calorific value in an estuarine and coastal
environment. Frontiers in Marine Science 3, 30 (2016).
87. Bunce, A. Prey consumption of Australasian gannets (Morus serrator) breeding in Port Phillip Bay, southeast Australia, and potential
overlap with commercial sheries. ICES Journal of Marine Science: Journal du Conseil 58, 904–915 (2001).
88. Vedel, A. & iisgaard, H. U. Filter-feeding in the polychaete Nereis diversicolor: growth and bioenergetics. Marine Ecology Progress
Series 100, 145–145 (1993).
89. Van Heuelem, W. F. Growth, bioenergetics and life-span of Octopus cyanea and Octopus maya. PhD thesis, University of Hawaii,
Manoa (1976).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
SCIENtIfIC REPORts | (2018) 8:10725 | DOI:10.1038/s41598-018-28479-7
90. Clare, A., Clare, M., Holmes, L. J. & Waters, T. Calorific values and elemental analysis of eleven species of oceanic squids
(Mollusca: Cephalopoda). Journal of the Marine Biological Association of the United ingdom 65, 983–986 (1985).
91. Perez, M. A. Calorimetry measurements of energy value of some Alasan shes and squids. (US Department of Commerce, National
Oceanic and Atmospheric Administration, National Marine Fisheries Service, Alasa Fisheries Science Center, 1994).
92. Benoit-Bird, . Prey caloric value and predator energy needs: foraging predictions for wild spinner dolphins. Marine Biology 145,
435–444 (2004).
93. arjalainen, J., Miserque, D. & Huusonen, H. e estimation of food consumption in larval and juvenile sh: experimental
evaluation of bioenergetics models. Journal of Fish Biology 51, 39–51 (1997).
Samples for this project were collected under New South Wales Department of Primary Industries Scientic
Collection Permit No. P03/0086(F)-8.1. Experiments were conducted under approval of the University of
New South Wales Animal Care and Ethics Committee (No. 15/152B), and all experiments were performed
in accordance with relevant guidelines and regulations of this ethics approval. is work was funded by the
Australian Research Council (Linkage Project LP150100923). is is publication number 228 from the Sydney
Institute of Marine Science. anks to all volunteers who helped with sh collection.
Author Contributions
C.L.L., I.M.S., J.A.S. and S.B. conceived the ideas and designed methodology; C.L.L., S.B. and H.T.S. collected the
data; C.L.L., J.A.S. and S.B. analysed the data; C.L.L. led the writing of the manuscript. C.L.L., I.M.S., J.A.S., H.T.S.,
J.S., J.M.H. and S.B. contributed critically to the dras and gave nal approval for publication.
Additional Information
Supplementary information accompanies this paper at
Competing Interests: e authors declare no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit
© e Author(s) 2018
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
... There will be differences between the assumed diet and actual diet specific to a location or habitat, and these differences may impact estimates of consumption rate and other energy-dependent parameters, such as growth rate. While earlier work has shown the importance of using multiple prey items in bioenergetics models to estimate consumption (Lawson et al., 2018), there are no studies to the author's knowledge that use detailed diet data (i.e., multiple prey items) to quantify the uncertainty of consumption and growth estimates introduced by supplementing diet information from different spatiotemporal regions. ...
... Consumption rate estimates for Bream were estimated using a bioenergetic model, based on the format of Kitchell et al. (1977) and modified as in Lawson et al. (2018): ...
... To convert an energy requirement (in joules) accurately to a food requirement (in grams), information on the ED of prey consumed is required. Based on the diet reported here and associated prey ED values found in the literature Cleveland, 1979) was fitted to determine the relationship between prey energy density and Bream length (Lawson et al., 2018). This indicates the mean prey energy density (ED, J g −1 ) throughout ontogeny of Bream. ...
Full-text available
Consumption is the primary trophic interaction in ecosystems and its accurate estimation is required for reliable ecosystem modeling. When estimating consumption, species’ diets are commonly assumed to be the average of those that occur among habitats, seasons, and life stages which introduces uncertainty and error into consumption rate estimates. We present a case study of a teleost (Yellowfin Bream Acanthopagrus australis ) that quantifies the potential error in consumption (in mass) and growth rate estimates when using diet data from different regions and times and ignoring ontogenetic variability. Ontogenetic diet trends were examined through gut content analysis ( n = 1,130 fish) and incorporated into a bioenergetic model (the “primary” model) that included diet variability ( n = 144 prey sources) and ontogenetic changes in metabolism (1–7 year) to estimate lifetime consumption. We quantified error by building nine model scenarios that each incorporated different spatiotemporal diet data of four published studies. The model scenarios produced individual lifetime consumption estimates that were between 25% lower and 15% higher than the primary model (maximum difference was 53%, range 11.7–17.8 kg). When consumption (in mass) was held constant, differences in diet quality among models caused a several-fold range in growth rate (0.04–1.07 g day –1 ). Our findings showcase the large uncertainty in consumption rate estimates due to diet diversity, and illustrate that caution is required when considering bioenergetic results among locations, times, and ontogeny.
... On rare occasions, in the absence of fish it consumes bottom invertebrates. Bluefish juveniles feed on zooplankton, such as molluscan larvae, copepods and mysids, but when the individuals reach 6-8 cm of body length, they switch to a diet dominated by fish (Stoyanov et al. 1963, Schilling et al. 2017, Lawson et al. 2018. Bluefish is reported to exert a pronounced selectivity of prey and their sizes (Juanes & Conover 1994). ...
... Bluefish is reported to exert a pronounced selectivity of prey and their sizes (Juanes & Conover 1994). During summer and early autumn, the juvenile bluefish grow faster, due to high piscivore feeding rates (Juanes & Conover 1994), and may have significant impact on prey fish populations (Buckel et al. 1999, Lawson et al. 2018). In the Black Sea, the predatory impact of bluefish strongly affects other commercial stocks (Prodanov et al. 1991) and is thought to have contributed to the expiration of the Black Sea mackerel (Ivanov & Beverton 1985, Prodanov et al. 1991. ...
... Juvenile fish has higher feeding intensity but a smaller individual consumption in weight (Lawson et al. 2018). The high consumption rate of juveniles is a result of consuming prey items containing less energy (Bartell et al. 1986, Beauchamp et al. 1989, Lawson et al. 2018. ...
... However, the energetic cost of reproduction is often estimated with little certainty or excluded from energetics studies (e.g. Lawson et al., 2018). Nevertheless, variability in reproductive costs provides insights into the potential for species to respond to stressors (Hammerschlag et al., 2018). ...
... This model used parameters measured and derived here, as well as some from the literature. The model followed the format of Kitchell et al. (1977), modified as in Lawson et al. (2018) to estimate consumption (J d -1 ): ...
Here we investigated measurements of energy density and bioenergetic modelling for a pelagic ray, Mobula eregoodoo, to estimate its relative allocation to various bodily processes, and especially reproduction. The data revealed M. eregoodoo uses up to 21.0 and 2.5% of its annual energy budget on growth and reproduction, respectively. During pregnancy, females depleted energy reserves in the liver which, along with their biennial reproductive cycle, aligns with general theory that ectotherms are capital breeders, and thus build energy reserves prior to reproduction. However, the reduction in energy reserves did not account for all reproductive costs, and hence gravid females supplement reproductive costs through energy derived from the diet; as per an income breeding strategy. These characteristics imply M. eregoodoo exhibits some flexibility in fuelling reproduction depending on energy availability throughout the reproductive cycle, which may be prevalent in other elasmobranchs. The data represent the first estimates of both the metabolic costs of gestation in elasmobranchs, and the relative cost of reproduction in rays. Energy costs and plasticity associated with highly variable reproductive strategies in elasmobranchs may influence long-term population viability under a rapidly changing environment. This article is protected by copyright. All rights reserved.
... The Metabolic Theory of Ecology proposes that metabolic rate is a fundamental biological rate, controlling the majority of observed ecological patterns. Relationships between environmental variables and metabolic rates are often tested within a controlled laboratory setting, which generates a metabolic performance curve (Claireaux & Lagardère 1999, Sinclair et al. 2016, Lawson et al. 2018. Considering fishes' ectothermic physiology, such studies commonly focus on metabolic response to temperature changes, which can reveal a thermal optimum -the temperature above which an organism's metabolic function and linked physiological performance such as growth begin to decline (Pirozzi & Booth 2009, Vinagre et al. 2015. ...
... The main diet variation occurred in the larger and older individuals, especially in females, with a marked increase in the intake of polychaetes. This shift in consumption appears related to the greater energy requirements for reproductive processes, because polychaetes have high nutritional value (Olive 1999;Shucksmith et al. 2006;Dorgham et al. 2015) and higher energy density values than other invertebrates (James et al. 2012;Lawson et al. 2018). In addition, polychaetes represent an important vector to transfer essential fatty acids to fish and crustaceans (Bischoff et al. 2009;Palmer et al. 2014). ...
The diet of the diamond stingray (Hypanus dipterurus) was quantified based on individuals collected monthly from October 2013 to December 2015 on Espiritu Santo Island in the Bahı´a de La Paz, Me´xico. Of the 473 stomachs analysed, 211 (44.6%) contained food. Analysis of the prey-specific index of relative importance indicated that the diet of H. dipterurus was based on infaunal and epibenthic invertebrates, especially the bivalve Solemya spp. (42.2%) and the stomatopod Nannosquilla raymanningi (7.7%). Statistical analyses of the diet considering sex, age, maturity stage and interactions between sex, age and season indicated that older females consumed more polychaetes in the cold (November–April) season and that younger females consumed more stomatopods and bivalves in the warm (May–October) season. The feeding strategy of the species is specialised on three prey categories (bivalves, polychaetes and stomatopods), with low trophic niche breadth values. The calculated trophic level indicated that the diamond stingray is a secondary consumer and mesopredator. We conclude that the diamond stingray does not maintain a redundant trophic function with sympatric elasmobranchs in the study area, and is therefore likely a key prey density-regulator functioning to link energy derived from lower to upper trophic levels within the southern Gulf of California
... It is a key parameter in a wide variety of bioenergetics and growth models (e.g., Benoit-Bird, 2004;Deslauriers, Heironimus, & Chipps, 2016;Maino & Kearney, 2015) but can be quite sensitive to error (Bartell, Breck, Gardner, & Brenkert, 1986). This error remains an important source of variability as somatic energy content and composition can vary substantially depending on spatial context (Dessier et al., 2018;Ruck, Steinberg, & Canuel, 2014;Schultz & Conover, 1997), season sampled (Chen et al., 2004;Pedersen & Hislop, 2001), or ontogenetic stage (Lawson et al., 2018;Woodland, Hall, & Calder, 1968). Thus, accurate measurements of ED are imperative to the validity of predictions made from these models. ...
Full-text available
Abstract Determining how energy flows through ecosystems reveals underlying ecological patterns that drive processes such as growth and food web dynamics. Models that assess the transfer of energy from producers to consumers require information on the energy content or energy density (ED) of prey species. ED is most accurately measured through bomb calorimetry, but this method suffers from limitations of cost, time, and sample requirements that often make it unrealistic for many studies. Percent dry weight (DW) is typically used as a proxy for ED, but this measure includes an indigestible portion (e.g., bones, shell, salt) that can vary widely among organisms. Further, several distinct models exist for various taxonomic groups, yet none can accurately estimate invertebrate, vertebrate and plant ED with a single equation. Here, we present a novel method to estimate the ED of organisms using percent ash‐free dry weight (AFDW). Using data obtained from 11 studies diverse in geographic, temporal and taxonomic scope, AFDW, DW as well as percent protein and percent lipid were compared as predictors of ED. Linear models were produced on a logarithmic scale, including dummy variables for broad taxonomic groups. AFDW was the superior predictor of ED compared to DW, percent protein content and percent lipid content. Model selection revealed that using correction factors (dummy variables) for aquatic animals (AA) and terrestrial invertebrates (TI) produced the best‐supported model—log10(ED) = 1.07*log10(AFDW) − 0.80 (R2 = 0.978, p 0.97), suggesting that AFDW can be used with high degrees of certainty to predict the ED of taxonomically diverse organisms. Our AFDW model will allow ED to be determined with minimal cost and time requirements and excludes ash‐weight from estimates of digestible mass. Its ease of use will allow for ED to be more readily and accurately determined for diverse taxa across different ecosystems.
Full-text available
The yellowfin notothen Patagonotothen guntheri is an important fish species in the Marine Protected Area Namuncurá-Burdwood Bank (MPAN-BB) due to its high abundance and significant ecological role. Thus, diet composition and feeding habits were analyzed. To fulfill this purpose, percentage frequency of occurrence, prey-specific abundance, and prey-specific index of relative importance (%PSIRI) were calculated. Generalized linear models were built for the defined prey categories to assess changes in feeding associated with total length, sex, depth, and zone of capture. Two hundred fifty-two specimens were sampled (69–256 mm total length), of which 186 contained food (73.41%). According to PSIRI, ophiuroids (23.28%), unidentified polychaetes (17.92%), and Munida gregaria (12.73%) were the main prey items. However, yellowfin notothen showed variations in diet composition according to their length and the depth of capture. The consumption of decapods and other invertebrates increased with depth, while the ingestion of ophiuroids decreased. Also, the number of polychaetes was higher in the stomach content of larger specimens. The present work provides the first description and analysis of P. guntheri trophic ecology in the MPAN-BB, concluding that it is a species with a generalist diet, which feeds on a great diversity of both benthic and pelagic organisms depending on availability.
Full-text available
The composition and abundance of Mugil cephalusdiet in La Libertad region during 2016 was determined. A sample of 962 stomachs was analyzed.The composition of the diet was estimated using the numerical and volumetric (wet biomass of prey) methods. The diet consisted of 32 preys grouped into four preyitems. The pelagic diatoms were the most numerous, however, in terms of the percentage of wet biomass, it was higher in copepod item. The results constitute a significant contribution to the knowledge of the M. cephalusfood regime, however, studies on a larger spatial and temporal scale are needed, which allow an analysis at the trophic habitat level of planctophagous fish.
Stock enhancement is a contemporary management method employed to support fisheries productivity. Blue swimmer crab (Portunus armatus) is a widely distributed species that has been identified as a candidate for stock enhancement; however, the release strategy and ecological impact of releases have not yet been assessed. Here, we (1) quantify the bioenergetics of blue swimmer crab, (2) estimate consumption rates, and (3) develop these relationships in a stocking model to assess appropriate release densities and associated trophic impacts. Static respirometry was used to measure the resting metabolic rate (RMR) of blue swimmer crab at three temperatures (19, 24, and 298C). RMR was found to increase with temperature (Q10 ¼ 2.32), and was 73% higher when crabs were at a premoult or moulting stage. Parameters derived from respirometry experiments were applied to estimate blue swimmer crab stocking density in a south-eastern Australian estuary, by adapting an existing production-based simulation model. The model estimated a median stocking density of,1.2 crabs per 100 m². A sensitivity analysis showed that the growth rate was the most influential parameter in this model, showing the importance of this parameter when assessing stocking scenarios. Journal compilation
Forage fish are a vital trophic group in marine ecosystems and numerical models, linking plankton with higher trophic levels. The bioenergetics of a key forage fish in eastern Australia, yellowtail scad Trachurus novaezelandiae, was measured using static respirometry and bomb calorimetry to assess their trophic contribution as both predator and prey. The temperature-dependent standard metabolic rate (SMR) of yellowtail scad was 0.62 mgO2 g−0.79 h⁻¹ and Q10 of 1.98. The SMR was used with dietary information to calculate a minimum annual prey consumption of 17.8 g g−0.79 y⁻¹ (at 20 °C), which is equivalent to an SMR-specific consumption:biomass ratio (Q:BSMR) of 6.77 (100 g adult). This was incorporated into a standard energy balance model to estimate a total Q:B of at least 10.6 at 20 °C (or 7.2 at 14 °C), which is 2 to 3-fold larger than most values used to represent this species in ecosystem models. Implications of underestimating forage fish consumption could include errors in estimated prey biomass, ecotrophic efficiencies, or strength of top-down and bottom-up control. Yellowtail scad were a moderate-high value prey, with a mean energy density of 6 kJ g⁻¹ (±0.97 s.d.). Energy density declined with body size and showed considerable inter-individual and spatial-temporal variation, indicating the potential to influence predator consumption rates at seasonal and fine spatial scales. This research highlights the value of measuring species-specific bioenergetics information for improving our understanding of trophic dynamics in marine ecosystems and models.
Full-text available
Habitat use can be complex, as tradeoffs among physiology, resource abundance, and predator avoidance affect the suitability of different environments for different species. Green sturgeon (Acipenser medirostris), an imperiled species along the west coast of North America, undertake extensive coastal migrations and occupy estuaries during the summer and early fall. Warm water and abundant prey in estuaries may afford a growth opportunity. We applied a bioenergetics model to investigate how variation in estuarine temperature, spawning frequency, and duration of estuarine residence affect consumption and growth potential for individual green sturgeon. We assumed that green sturgeon achieve observed annual growth by feeding solely in conditions represented by Willapa Bay, Washington, an estuary annually frequented by green sturgeon and containing extensive tidal flats that harbor a major prey source (burrowing shrimp, Neotrypaea californiensis). Modeled consumption rates increased little with reproductive investment (<0.4%), but responded strongly (10–50%) to water temperature and duration of residence, as higher temperatures and longer residence required greater consumption to achieve equivalent growth. Accordingly, although green sturgeon occupy Willapa Bay from May through September, acoustically-tagged individuals are observed over much shorter durations (34 d + 41 d SD, N = 89). Simulations of <34 d estuarine residence required unrealistically high consumption rates to achieve observed growth, whereas longer durations required sustained feeding, and therefore higher total intake, to compensate for prolonged exposure to warm temperatures. Model results provide a range of per capita consumption rates by green sturgeon feeding in estuaries to inform management decisions regarding resource and habitat protection for this protected species.
Full-text available
Pelagic mesopredators are abundant in many marine ecosystems and exert strong top-down influence on food webs. We explored the dietary niche of Pomatomus saltatrix in eastern Australia, using a classification tree analysis to identify key factors driving diet variation. P. saltatrix was shown to be an opportunistic generalist predator which exhibited increased baitfish consumption, and decreased crustacean consumption, with increasing size. The classification tree analysis showed that body size and latitude had the greatest influence on the diet of P. saltatrix, with significant ontogenetic diet shifts occurring at 8 and 30 cm fork length (FL). While piscivory is evident in the majority of P. saltatrix diets by ~8 cm FL, crustaceans are almost entirely absent from the diet after ~30 cm FL. The importance of latitude was likely related to the broad-scale oceanography in the study region, including the East Australian Current and its separation from the continental shelf. The classification tree analysis is a powerful framework for identifying important variables in diet composition.
Full-text available
There is a growing need to incorporate biotic interactions, particularly those between predators and their prey, when predicting climate-driven shifts in marine fishes. Predators dependent on a narrow range of prey species should respond rapidly to shifts in the distribution of their prey, whereas those with broad trophic adaptability may respond to shifts in their prey by altering their diet. Small pelagic fishes are an extremely important component of the diet of many marine predators. However, their populations are expected to shift in distribution and fluctuate in abundance as the climate changes. We conducted a comparative study of the seasonal diet of adult Pomatomus saltatrix over two periods (June–December 2006 and 2012) and examined the available data on small pelagic fishes biomass in a global hotspot (the coastal region of southern Angola, southern Africa) to gain an understanding of the tropic adaptability of the species. Despite a drop (630 000 t to 353 000 t) in the abundance of their dominant prey (Sardinella aurita) in the region, it remained the most important prey item during both study periods (Period 1 = 99.3% RI, Period 2 = 85.3% RI, where %RI is a ranking index of relative importance). However, the diet during Period 2 was supplemented with prey typically associated with the nearshore zone. The seasonal data showed that P. saltatrix were capable not only of switching their diet from S. aurita to other prey items, but also of switching their trophic habitat from the pelagic to the nearshore zone. These findings suggest that P. saltatrix will not necessarily co-migrate if there is a climate-driven shift in the distribution of small pelagic fishes (their dominant prey). Accordingly, understanding the trophic adaptability of predators is critical for understanding their response to the impacts of climate change.
Full-text available
We constructed a bioenergetics model for sockeye salmon Oncorhynchus nerka and evaluated its sensitivity to parameter error. When used to predict annual growth, the model was most sensitive, in declining order of importance. to changes in the intercept of the dependence of consumption on body weight, the proportion of maximum consumption, the energy density of prey, low temperature and its associated proportion of maximum consumption in the temperaturedependence function, the intercept of the energy density relationship to predator weight, and the intercept of the relationship between body weight and respiration. Estimates of consumption from the model, when consumption was constrained by fixed growth, were quite insensitive to perturbation of all parameters except the energy density of prey. We computed consumption rates and energy budgets with the model and compared these with independently derived estimates for populations in Lake Dalnee, USSR; Lake Washington, USA; and Babine Lake, Canada. The close agreement of estimates from the model to independent estimates of prey consumption and energy budgets for three different populations indicated that the model may be widely applicable to other populations. Performance of the model can be enhanced further if the frequency with which water temperature and mean body weight data are collected is increased for each cohort of interest.
Full-text available
Human activities have substantially changed the world's oceans in recent decades, altering marine food webs, habitats and biogeochemical processes [1]. Cephalopods (squid, cuttlefish and octopuses) have a unique set of biological traits, including rapid growth, short lifespans and strong life-history plasticity, allowing them to adapt quickly to changing environmental conditions [2-4]. There has been growing speculation that cephalopod populations are proliferating in response to a changing environment, a perception fuelled by increasing trends in cephalopod fisheries catch [4,5]. To investigate long-term trends in cephalopod abundance, we assembled global time-series of cephalopod catch rates (catch per unit of fishing or sampling effort). We show that cephalopod populations have increased over the last six decades, a result that was remarkably consistent across a highly diverse set of cephalopod taxa. Positive trends were also evident for both fisheries-dependent and fisheries-independent time-series, suggesting that trends are not solely due to factors associated with developing fisheries. Our results suggest that large-scale, directional processes, common to a range of coastal and oceanic environments, are responsible. This study presents the first evidence that cephalopod populations have increased globally, indicating that these ecologically and commercially important invertebrates may have benefited from a changing ocean environment.
Full-text available
Large tropical and sub-tropical marine animals must meet their energetic requirements in a largely oligotrophic environment. Many planktivorous elasmobranchs, whose thermal ecologies prevent foraging in nutrient-rich polar waters, aggregate seasonally at predictable locations throughout tropical oceans where they are observed feeding. Here we investigate the foraging and oceanographic environment around Lady Elliot Island, a known aggregation site for reef manta rays Manta alfredi in the southern Great Barrier Reef. The foraging behaviour of reef manta rays was analysed in relation to zooplankton populations and local oceanography, and compared to long-term sighting records of reef manta rays from the dive operator on the island. Reef manta rays fed at Lady Elliot Island when zooplankton biomass and abundance were significantly higher than other times. The critical prey density threshold that triggered feeding was 11.2 mg m-3 while zooplankton size had no significant effect on feeding. The community composition and size structure of the zooplankton was similar when reef manta rays were feeding or not, with only the density of zooplankton changing. Higher zooplankton biomass was observed prior to low tide, and long-term (~5 years) sighting data confirmed that more reef manta rays are also observed feeding during this tidal phase than other times. This is the first study to examine prey availability at an aggregation site for reef manta rays and it indicates that they feed in locations and at times of higher zooplankton biomass.