Embed Size (px)
Citation preview
ORIGINAL PAPER
Variability of albacore (Thunnus alalunga) diet in the NortheastAtlantic and Mediterranean Sea
Nicolas Goni • John Logan • Haritz Arrizabalaga •
Marc Jarry • Molly Lutcavage
Received: 6 April 2010 / Accepted: 14 January 2011 / Published online: 3 February 2011
� Springer-Verlag 2011
Abstract This study aims to describe the variability of
albacore (Thunnus alalunga) diet in the Northeast Atlantic
and Mediterranean Sea and to identify possible relation-
ships between this variability and the features of different
feeding areas, the behavior, and the energetic needs of
albacore. Stomach contents from albacore caught in five
zones of the Bay of Biscay and surrounding waters
(n = 654) and three zones of the Mediterranean Sea
(n = 152) were analyzed in terms of diet composition and
stomach fullness. Carbon and nitrogen stable isotope and
C/N ratios were measured for white muscle and liver from
albacore in the Bay of Biscay (n = 41) and Mediterranean
Sea (n = 60). Our results showed a spatial, seasonal, inter-
annual, and size-related variability in the diet of albacore.
Albacore diet varied by location in the Mediterranean Sea,
with a particularly high proportion of cephalopods, and low
d15N values in the Tyrrhenian Sea. In the Northeast
Atlantic, albacore consumed a higher proportion of crus-
taceans and a lower proportion of fishes in the most off-
shore sampling zone than inshore. The digestion states of
the major prey reflected a diurnal feeding activity, indica-
tive of feeding in deeper waters offshore, whereas on the
continental slope, feeding probably occurred in surface
waters at night. Important seasonal and inter-annual diet
variability was observed in the southeast of the Bay of
Biscay, where preferred albacore prey appeared to be
anchovy (Engraulis encrasicolus). Stomach fullness was
inversely related to body size, probably reflecting higher
energetic needs for smaller individuals. Albacore from the
Bay of Biscay had significantly lower d13C and higher
d15N values compared with albacore from the Mediterra-
nean Sea, indicative of regional baseline shifts, and trophic
position and muscle lipid stores in albacore increased with
body size.
Introduction
Albacore (Thunnus alalunga) is a highly migratory tuna
species found in both tropical and temperate areas of the
Atlantic, Indian, and Pacific oceans, and in the Mediter-
ranean Sea. In the North Atlantic, adult albacore spawn in
summer months in tropical regions. Juveniles (age-0 to
age-4 individuals) make a feeding migration between
subtropical and temperate regions (Bard 1981; Santiago
2004) in summer months, when they show highest growth
rates (Santiago and Arrizabalaga 2005). Some adult indi-
viduals also appear in temperate regions in late summer
and autumn. The central North Atlantic is considered the
wintering zone for juveniles and adults. Mediterranean
adult albacore spawn in 10 different regions of the
Communicated by C. Harrod.
N. Goni (&) � H. Arrizabalaga
AZTI-Tecnalia, Herrera kaia portualdea z/g,
20110 Pasaia (Gipuzkoa), Spain
e-mail: [email protected]
J. Logan
Massachusetts Division of Marine Fisheries,
1213 Purchase Street, New Bedford, MA 02740-6698, USA
M. Jarry
INRA, UMR 1224 Ecobiop, 64310 Saint Pee sur Nivelle, France
M. Jarry
UPPA, UMR 1224 Ecobiop, 64013 Pau, France
M. Lutcavage
Large Pelagics Research Center, Department of Natural
Resources Conservation, University of Massachusetts Amherst,
108 East Main Street, Gloucester, MA 01930-3846, USA
123
Mar Biol (2011) 158:1057–1073
DOI 10.1007/s00227-011-1630-x
Mediterranean Sea (Marano et al. 1999). According to
conventional tagging experiments (Arrizabalaga et al.
2002), they appear to have a more constrained range with
no clear long-range migration. Albacore migration rates
between the Eastern North Atlantic and the Mediterranean
Sea are less than 1%, in both directions (ibid.). In the
Northeast Atlantic, juvenile albacore are exploited from
June to October by surface baited gears (mainly Spanish
trolling and baitboat fleets) and pelagic trawling (mainly
French and Irish fleets), in the Bay of Biscay and sur-
rounding seas (ICCAT 2008). Juvenile and adult Mediter-
ranean albacore are exploited mainly by longline in the
Balearic, Tyrrhenian, Ionian, Adriatic, and Aegean Seas
(Marano et al. 1999; de la Serna et al. 2003).
Several studies have been performed to date on albacore
diet in the Northeast Atlantic, through stomach contents
(Aloncle and Delaporte 1974; Ortiz de Zarate 1987; Pusi-
neri et al. 2005) and through carbon and nitrogen stable
isotope ratios (Das et al. 2000). However, since the study
by Aloncle and Delaporte (1974), geographic, seasonal,
and size-related variabilities in diet have not been inves-
tigated in albacore caught in this area. Even fewer studies
have been performed on albacore diet in the Mediterranean
Sea (Bello 1999; Consoli et al. 2008, Salman and Karakulak
2009), none of which included comparisons among areas
or stable isotope analysis.
Stable isotopes of carbon and nitrogen provide inte-
grated records of assimilated diet (Peterson and Fry 1987).
Nitrogen stable isotope ratios (15N/14N; d15N) generally
increase across trophic levels to a greater degree than
carbon isotopes (13C/12C; d13C) and can be used as a proxy
for trophic position (Post 2002), i.e., relative place in the
food web. In fishes, consumer tissue isotope values reflect
recent diet more rapidly in liver (weeks) than in white
muscle (months) (Logan et al. 2006). Analysis of both
tissues provides dietary information over different time-
scales. Comparisons of isotope values for multiple tissues
from individual fish can provide an estimate of residency in
the sampling region (Phillips and Eldridge 2006). Tissue
C/N values increase linearly with lipid content (Logan
et al. 2008) and can be used as a proxy for body condition
(Estrada et al. 2005).
Stomach content analyses provide detailed information
on recent diet composition, whereas stable isotope data
track diet over longer timescales. The combined use of
both techniques can help us to have a broader view of the
variability of albacore diet in several periods and locations.
Tunas have high standard metabolic rates compared
with strictly poikilothermic fish species (Korsmeyer and
Dewar 2001). This metabolic rate may be particularly high
for populations that perform long-distance seasonal
migrations (such as Atlantic albacore) and for juvenile
individuals, i.e. in rapid growth phase with possible
variations or shifts in their physiology (Goni and Arriza-
balaga 2010). Therefore, their feeding ecology has critical
implications for life history features of growth and sur-
vival. Moreover, their wide geographic range—especially
in the North Atlantic—suggests that albacore target a
variety of prey and have flexible feeding behaviors,
resulting in variable prey energy inputs and in variable top-
down effects on different prey communities. Knowledge of
the variability in albacore diet is necessary to assess these
relationships.
The aims of this study were to identify spatio-temporal
and size-related variations in albacore diet in the Northeast
Atlantic and Mediterranean Sea through stomach content
and stable isotope analysis and to examine these variations
against behavior and energetic needs. Are some areas/
periods more energetically suitable for albacore? Does the
energy input of diet vary along with albacore growth? The
goals of this study are also to assess the plasticity of
albacore feeding habits and to provide basic data on alba-
core trophic relationships that can be used in ecosystem
studies. Detailed dietary records were derived mainly
through stomach content analysis. Stable isotope analysis
was used to assess general spatial and size-related trends in
albacore diet and trophic position.
Materials and methods
Samples
Stomachs were sampled from albacore caught in the
Northeast Atlantic (n = 654), and in the Mediterranean
Sea (n = 152). Individuals from the Northeast Atlantic
were sampled in five geographic zones (Fig. 1), from 2004
to 2007, using a range of gears from different depths
(Table 1). In the Mediterranean Sea, the individuals were
sampled in the Balearic Sea, the South Adriatic Sea, and
the Tyrrhenian Sea (Fig. 1), in 2005, 2006, and 2008,
respectively. Predator length was converted to age
according to the growth equations for Atlantic and Medi-
terranean albacore (Santiago and Arrizabalaga 2005;
Megalofonou 2000). Size-range of Atlantic albacore was
39.6–112.0 cm fork-length (FL); size-range of Mediterra-
nean albacore was 58.5–83.2 cm FL.
The fork-length of all albacore was measured, and
individuals sampled after landing were weighed. In the
case of individuals sampled onboard in the Northeast
Atlantic, an estimated mass was calculated using the
length–mass relationship by Santiago (1993). Individual
catch dates and locations are known for fish caught in zone
1 and zone 2 (Fig. 1) in 2005. Individual catch dates are
known for all fish caught in zone 5 and individual catch
locations for 60% of them. Individual catch time is known
1058 Mar Biol (2011) 158:1057–1073
123
Fig. 1 Locations and
corresponding sample sizes of
the five albacore sampling zones
in the Bay of Biscay and
surrounding waters (upperpanel), and of the three
sampling zones in the
Mediterranean Sea
(lower panel)
Table 1 Area, catch period, catch time, catch depth, gear type (active vs. baited), and number by age-group of 654 albacore sampled in the Bay
of Biscay (zones 1–5) and 152 albacore sampled in the Balearic Sea, the South Adriatic Sea, and the Tyrrhenian Sea
Area Catch period Catch time Catch depth Gear type Number by age-group Total by subset
Age 0 Age 1 Age 2 Age 3 Age 4 Age 5?
Zone 1 29/07/05–02/08/05 Day Surface Baited – 7 33 9 2 – 51
Zone 2 24/08/04–04/10/04 Night 0–100 m Active – – 5 6 13 3 27
27/07/05–28/07/05 Day Surface Baited – 9 7 2 – – 18
28/07/06–30/07/06 Day Surface Baited – 30 16 3 – – 49
Zone 3 07/08/05–15/08/05 Night 0–100 m Active – 4 61 1 – – 66
27/09/05–09/10/06 – – 8 33 1 – 42
Zone 4 18/07/06–19/09/06 Night 0–100 m Active – 1 1 23 12 – 37
Zone 5 15/06/05–24/10/05 Day Surface Baited – 84 98 29 26 4 259
10/08/06–31/10/06 21 13 17 4 10 3 68
04/08/07–22/10/07 – 13 6 8 5 3 37
Balearic October 2005 Night 0–50 m Baited – – 39 13 – – 52
S. Adriatic November 2006 Night 0–50 m Baited – – 22 28 – – 50
Tyrrhenian January 2008 Night 0–50 m Baited – – – 8 30 12 50
Mar Biol (2011) 158:1057–1073 1059
123
for all fish caught in zone 1 and zone 2 in 2005 and for 203
of the fish caught in zone 5. For fish caught in zone 2
(except 2005), zone 3, and zone 4, individual catch dates
and locations are not known, but the dates and latitude/
longitude ranges of the catches corresponding to each
landing are known. In the case of fish caught by longline,
only the broad area and catch month are known.
Fish caught by pelagic trawl and longline were com-
mercial fish. They were kept on ice onboard and generally
landed 1–6 days after catch. Their stomachs were sampled
after landing and kept frozen. Fish caught by rod-and-reel
and trolling were caught during daytime, their stomachs
were sampled onboard immediately after catch, and frozen.
Stomachs from zone 5 were sampled from albacore caught
during scientific surveys (n = 97, Goni et al. 2009) and
from recreational fishery catches (n = 267). For stable
isotope analysis, white muscle from the pectoral region and
liver were sampled from albacore in zone 5 in the Bay of
Biscay (n = 41, 53.1–81.0 cm FL), the South Adriatic Sea
(n = 30, 58.5–69.9 cm FL), and the Tyrrhenian Sea
(n = 30, 67.6–83.2 cm FL). Tissue samples were stored
frozen prior to analysis.
Stomach content analyses
Whole stomachs were weighed before content analysis. All
contents were then removed, and the stomach lining was
weighed after being rinsed and blotted dry. The difference
between both masses was considered the total content
mass, including gastric liquid, which often contained
remains of crustaceans and fish in the most advanced
digestion state. Stomach fullness was defined for each
sampled albacore as the ratio between the mass of the
stomach content (g) and the mass—measured or calcu-
lated—of the individuals (kg).
Each prey item was identified to the lowest possible
taxon. Fishes were identified using the identification keys—
based on morphological characteristics—by Ibanez Artica
et al. (1989) and the online database http://www.fishbase.org
(Froese and Pauly 2010). Crustaceans were identified using
the manual by Todd et al. (1996), which comprises mor-
phological descriptions of crustacean species. Cephalopods
were identified by the morphological characteristics of their
beaks, according to the handbook by Clarke (1986). All
thaliacea were barrels associated with the amphipod
Phronima sedentaria (Laval 1978). For all stomachs, the
mass, number of individuals, length, and digestion state of
each identified prey item were recorded. The digestion state
of each prey item was considered, following the four states
defined in the case of albacore by Aloncle and Delaporte
(1974) for crustaceans and fish prey and the six states defined
by Bertrand (1999) for cephalopod prey. No digestion state
was recorded for thaliacea prey.
In 57% of the stomachs containing euphausiids, several
individuals were not countable and lacked measurable parts
(e.g. telson) due to partial digestion. Fresh mass could not
be estimated for these individuals, so reconstituted prey
mass was not taken into account in this study.
Species for which only hard parts (otoliths, bones, and
cephalopod beaks) were found were not considered for
qualitative analysis, as they are likely to be remains of prey
ingested several days earlier, but their mass was taken into
account for calculating stomach fullness. Longline bait
(sardines (Sardina pilchardus)) was not considered in any
analysis and its mass was subtracted from stomach content
mass. The mass percentage of a given prey in a stomach
was defined as the mass of the prey divided by the total
mass of all identified prey. The mean mass percentage
(MW%) of a given prey in a subset was defined as the
average of its mass percentages in the stomachs of albacore
caught in a particular area and period. The frequency (F%)
of a given prey in a subset was defined as the percentage of
predators that contained this prey in this subset.
Geographic variations in the prey composition of alba-
core diet were assessed through a correspondence analysis
(CA), a multivariate technique based on decomposing the
Chi-square statistic associated with a pivot table into
orthogonal factors (Benzecri 1973). Prey with MW%[5%
in at least one of our data subsets were selected for this
analysis. This CA was performed for all non-empty stom-
achs, using the mass percentages of the different prey
ingested for each albacore. An additional analysis was
performed on albacore from zone 5 to look for seasonal and
inter-annual variations in prey composition. Stomachs with
only liquid remains or unidentified prey in digestion state 4
(the most advanced state) were not used in these CAs.
Apart from the correspondence analyses, the MW% of prey
(or prey types) was also directly compared between dif-
ferent subsets, through Friedman rank sum tests.
Year, zone, age-group, and catch date are not indepen-
dent in our samples (Table 1). In the larger data subset
(zone 5, year 2005), albacore size and catch date are also
not independent. Consequently, stomach fullness cannot be
directly modeled as a function of these different factors and
variables. A multiple correspondence analysis (MCA) was
therefore performed beforehand on stomach fullness, year,
zone, length, and catch period. Seven length classes (\53,
53–56, 56–62, 62–65, 65–70, 70–78, and [78 cm FL),
seven levels of stomach fullness (\0.7, 0.7–1.1, 1.1–1.65,
1.65–2.5, 2.5–4.5, 4.5–10, and [10 g/kg), and six periods
(June, 1st half of July, 2nd half of July, August, September,
and October) were set, according to arbitrary divisions
allowing classes to have balanced numbers of individuals
(Benzecri 1973). Based on the results of this preliminary
analysis, the variability of stomach fullness was more
thoroughly analyzed through a generalized additive model
1060 Mar Biol (2011) 158:1057–1073
123
(GAMs, Hastie 1992), assuming a Gamma distribution for
stomach fullness.
Circadian patterns in diet were analyzed in zones 1 and
5, for which catch time (GMT) is known. Stomach fullness
was analyzed as a function of catch time through a GAM,
and the distribution of the four digestion states was
described as a function of catch time.
Stable isotope analysis
Albacore liver and white muscle samples were prepared
and analyzed for d13C, d15N, % carbon, and % nitrogen
according to the method used by Logan et al. (2008). All
sample d13C values were corrected for lipid content a pos-
teriori using mathematical approaches based on sample
C/N with parameters derived from data sets of Atlantic
bluefin tuna (Thunnus thynnus) liver and white muscle
(ibid.). All carbon and nitrogen isotope data are reported
in d notation (see review by Peterson and Fry (1987)).
All analyses were performed at the University of New
Hampshire Stable Isotope Laboratory. Precision is *0.2%for d13C and d15N. All d13C and d15N values were nor-
malized on the VPDB and AIR scales with IAEA CH6
(-10.4%), CH7 (-31.8%), N1 (0.4%), and N2 (20.3%).
Pairwise t tests with Holm-adjusted P values (Holm
1979) were performed to determine which regions and/or
years significantly differed. To further explore the effects
of size on condition (C/N as a proxy for lipid content) and
trophic position (d15N), simple linear regressions were
performed for d15N and C/N in relation to log10 length.
Differences in isotope values between muscle and liver
were calculated for individual fish as dXMuscle–dXLiver
where X is 13C or 15N. These values were compared among
the three sampling regions (zone 5 of the Bay of Biscay,
South Adriatic Sea, Tyrrhenian Sea) using pairwise Student
t tests. All values are reported as mean ± standard devia-
tion (SD) unless otherwise noted.
All statistical analyses were performed using the R v.
2.9.2. (R Development Core Team 2008) statistical
software (available online at http://www.r-project.org/).
Correspondence analyses were performed using the Fac-
toMineR 1.12 package. GAMs were performed with the
mgcv 1.5-5 package. Other statistical analyses were per-
formed using the stats 2.9.2 package.
Results
Geographic diet variations
The most ubiquitous prey of albacore in the Northeast
Atlantic and Mediterranean Sea is krill (Meganyctiphanes
norvegica), which is present in all our data subsets
(Tables 2, 3, 4, 5). In albacore caught in the Bay of Biscay,
the other prey with MW%[5% in one or more of our data
subsets were blue whiting (Micromesistius poutassou),
saury (Scomberesox saurus), garpike (Belone belone),
snake pipefish (Entelurus aequoreus), anchovy (Engraulis
encrasicolus), lanternfish (Notoscopelus elongatus kroy-
eri), pink glass shrimp (Pasiphaea sp.), European locust
lobster (Scyllarus arctus) in phyllosoma stage, hyperiid
amphipod (Themisto gaudichaudii), decapodiform cepha-
lopods, and salps. In the Mediterranean Sea, albacore diet
was dominated by krill and anchovy in the Balearic Sea, by
histioteuthid squids in the Tyrrhenian Sea, and by krill and
paralepidid fishes in the South Adriatic Sea (Table 5). The
proportion of the hyperiid amphipod Phrosina semilunata
was also [5% in the South Adriatic and Tyrrhenian Seas.
A global CA was based on these prey groups, with the
exception of salps—in a previous analysis including this
prey, the variability of the data was artificially under-rep-
resented and focused on the presence/absence of salps. The
first (F1) and second (F2) component of this CA represent
33 and 25% of the total inertia, respectively. On this CA,
the majority of Mediterranean individuals have positive
values on F1 and the majority of Bay of Biscay individuals
have negative values (Fig. 2). Prey and predators from the
Bay of Biscay and from the Mediterranean Sea appear
distributed along two distinct gradients. The first one
opposes blue whiting to a group formed by lanternfish,
krill, and locust lobster. The second one, along which
Mediterranean albacore and prey are distributed, is more
pronounced and opposes krill to histioteuthid cephalopods.
For a better representation of the geographic prey vari-
ability within each basin, two other separate CAs were per-
formed on data from the Bay of Biscay and from the
Mediterranean Sea, respectively. For each of the five zones
in the Bay of Biscay and of the three zones in the Mediter-
ranean Sea, mean and standard deviation of the factorial
coordinates (FC) of predator individuals on F1 and F2 were
calculated and are represented on Fig. 3. In the Bay of Biscay
CA, F1 and F2 represent 37 and 28% of the total inertia,
respectively. The relative distribution of prey on this CA is
similar to the one observed for the same prey in the previous
CA (left part of Fig. 2). Among the most eastern zones, F1
appears to represent a gradient on which decreasing values
from zone 3 to zone 5 reflect an increasing proportion of blue
whiting and anchovy and a decreasing proportion of krill. F2
appears to represent a separation between the most north-
western zones—1 and 2—and the most southeastern ones—
3, 4, and 5 (Fig. 3)—in which garpike, snake pipefish, saury
and T. gaudichaudii are scarce or absent (Tables 3, 4). The
FCs of individuals from zone 1, zone 2, and zone 3 appear
associated with T. gaudichaudii, saury, and krill, respec-
tively (Fig. 3). These zones correspond to the respective
highest MW% of these three prey (Tables 2, 3).
Mar Biol (2011) 158:1057–1073 1061
123
In the Mediterranean CA, F1 and F2 represent 41 and
33% of the total inertia, respectively. This CA shows a
marked Guttman (or ‘‘arch’’) effect, i.e., F2 appears to be
an arched function of F1, which is caused by the unimodal
distribution of species along gradients (Benzecri 1980).
Along this ‘‘arch’’, albacore appear positioned on a gradi-
ent of successive high proportions of anchovy, krill,
P. semilunata, paralepidid fishes, and histioteuthid squids
in their diet (Fig. 3). Except for anchovy (also present in
the Bay of Biscay), the successive positions of these prey
are similar to the one observed in the previous CA (right
part of Fig. 2). On this gradient, individuals from the
Balearic Sea, South Adriatic Sea, and Tyrrhenian Sea
appear successively, overlapping on krill, P. semilunata
and paralepidid fishes. Anchovies are found in samples
from the Balearic Sea only and histioteuthid squids in
samples from the Tyrrhenian Sea only (Table 5).
We identified different circadian patterns in the
digestion states of the main prey in the southeastern Bay
of Biscay (zone 5) and offshore (zone 1). In zone 1, krill
T. gaudichaudii were present in all digestion states all
day, despite the small number of albacore sampled there.
It is also the case for saury and snake pipefish in alba-
core sampled in this zone. In contrast, in zone 5, blue
whiting in early digestion and fresh or early digested
krill occurred in stomachs only in morning hours (before
10:00 GMT for blue whiting, before 11:30 GMT for
krill). In the stomachs of albacore caught during after-
noon or evening, these prey were present in advanced
digestion states only.
Albacore caught at night have higher proportions of
crustaceans and lower proportions of fishes in their stom-
achs than those caught during daytime (Table 6). Most of
these crustaceans are krill (Tables 2, 3, 5). However, in the
absence of subsets of albacore caught at night and during
daytime in the same area and year, any comparison is
limited. Albacore from the Tyrrhenian Sea differed
from other regions, with a particularly high MW% of
Table 2 Prey taxa encountered in non-empty stomachs of albacore caught in Zones 1 and 2 of the Bay of Biscay
Area Zone 1 Zone 2 Zone 2 Zone 2
Catch period 29/07/05–02/08/05 24/08/04–04/10/04 27/07/05–28/07/05 28/07/06–30/07/06
Gear Trolling line Pelagic trawl Trolling line Trolling line
Sample size (non-empty) n = 51 (46) n = 27 (19) n = 18 (14) n = 49 (39)
Fishes (teleosts)
Belone belone 5.8 – 20.9 (8.7) – 7.1 – 26.7 (7.1) 2.6 ± 16.0 (2.6)
Entelurus aequoreus 9.0 ± 25.2 (21.7) – – 35.0 ± 46.5 (41.0)
Micromesistius poutassou – 4.4 ± 19.1 (5.3) 7.7 – 19.7 (28.6) 23.8 ± 41.7 (28.2)
Notoscopelus elongatus kroyeri – 0.8 ± 2.5 (15.8) – –
Scomberesox saurus 16.5 ± 32.2 (26.1) 18.9 – 37.5 (26.3) 26.3 – 39.7 (35.7) 15.1 ± 34.2 (20.5)
Fish unid. 6.1 ± 19.7 (21.7) 8.2 ± 25.7 (10.5) 40.0 ± 48.5 (28.6) 5.5 ± 19.5 (10.3)
Crustaceans
Amphipoda
Phronima sedentaria 0.1 ± 0.7 (4.3) – – –
Themisto gaudichaudi 27.3 ± 38.7 (54.3) \0.1 ± 0.1 (10.5) 14.7 – 34.1 (21.4) 1.0 ± 6.0 (2.6)
Decapoda
Pasiphaea sp. 0.1 ± 0.6 (6.5) – – –
Scyllarus arctus (larvae) – – 2.5 ± 9.5 (7.1) –
Euphausiacea
Meganyctiphanes norvegica 32.9 ± 42.5 (50.0) 66.7 – 46.7 (73.7) 1.5 ± 3.9 (14.3) 14.1 ± 34.2 (15.4)
Nematoscelis megalops 0.3 ± 2.0 (4.3) – – –
Euphausiacea unid. 0.1 ± 0.3 (6.5) – – –
Isopoda (unid.) – – – 0.1 ± 0.9 (2.6)
Cephalopods
Decapodiform (unid.) – 0.1 ± 0.6 (5.3) – 2.9 ± 16.1 (7.7)
Thaliacea
‘‘barrels’’ of P. sedentaria 1.8 ± 9.9 (6.5) – – –
Mean ± standard deviation of prey mass percentage and—italic, in parentheses—frequency of occurrence. Mean mass percentages[5% in bold
1062 Mar Biol (2011) 158:1057–1073
123
cephalopods (52%), relative to \5% in albacore from all
other sampled zones (Table 6).
Among albacore caught during the day (zones 1, 2, and
5 of the Bay of Biscay), those caught in zone 1 had the
highest MW% of crustaceans (60.8%) and the lowest of
fishes (37.4%), whereas individuals caught in zones 2 and 5
(shelf-break zones) had MW% of crustaceans of 10–30%
and MW% of fishes of 66–88%, among the different sub-
sets (Table 6). This difference between albacore from zone
1 and albacore from zone 2 (troll-caught) and zone 5 is
significant (P = 0.0339 for crustaceans and P = 0.0133
for fishes in Friedman rank sum test). The MW% of fishes
in the diet of longline-caught Mediterranean albacore is
similar to the MW% of fishes in trawl-caught Bay of
Biscay albacore (Table 6).
Prey species diversity by predator was low, with 73% of
the stomachs containing only one prey group, and 22% of
them containing only two prey groups. For the major prey
groups except krill, mass percentage in stomach contents
tends to take extreme values, with values [95 or \5% in
more than 95% of their respective occurrences. In the case
of krill, mass percentages[95 or\5% represent 88% of its
occurrences.
Diet variability in the southeastern Bay of Biscay
Krill, anchovy, and blue whiting were considered for the
CA performed on the diet of albacore from zone 5.
Although the sampling periods differed for the 3 years
considered (2005, 2006, and 2007), we observed different
Table 3 Prey taxa encountered in non-empty stomachs of albacore caught in Zones 3 and 4 of the Bay of Biscay
Area Zone 3 Zone 3 Zone 4
Catch period 07/08/05–15/08/05 27/09/06–09/10/06 18/07/06–19/09/06
Gear Pelagic trawl Pelagic trawl Pelagic trawl
Sample size (non-empty) n = 66 (29) n = 42 (32) n = 37 (22)
Fishes (teleosts)
Arctozenus risso – 0.5 ± 3.0 (3.4) –
Belone belone 3.2 ± 17.4 (3.4) – –
Benthosema glaciale 0.7 ± 2.5 (10.3) 0.6 ± 3.3 (3.4) –
Entelurus aequoreus 3.4 ± 18.7 (3.4) – –
Micromesistius poutassou 2.6 ± 13.6 (10.3) 2.5 ± 13.5 (3.4) 29.4 ± 43.5 (35.3)
Notoscopelus elongatus kroyeri 6.3 ± 23.7 (6.9) – –
Paralepidae unid. – 0.1 ± 0.6 (3.4) –
Scomberesox saurus – – 11.8 ± 28.1 (17.6)
Fish unid. 0.4 ± 2.0 (3.4) 0.3 ± 1.2 (19.4) \0.1 ± 0.1 (5.9)
Crustaceans
Amphipoda
Phronima sedentaria 0.6 ± 3.3 (3.4) 0.2 ± 0.9 (13.8) –
Themisto gaudichaudi 7.8 – 26.0 (10.3) – –
Decapoda
Pasiphaea sp. 6.9 ± 25.8 (6.9) \0.1 ± 0.2 (3.4) –
Scyllarus arctus (larvae) 4.1 ± 18.7 (6.9) 0.9 ± 3.9 (20.7) –
Euphausiacea
Meganyctiphanes norvegica 47.2 – 48.2 (65.5) 84.3 – 33.6 (89.7) 57.4 – 46.5 (70.6)
Mysidae unid. 3.4 ± 18.6 (3.4) 3.9 ± 18.7 (6.9) \0.1 ± 0.1 (5.9)
Crustacean unid. 3.4 ± 18.6 (3.4) – 0.1 ± 0.5 (5.9)
Cephalopods
Gonatus steenstrupi – 0.3 ± 1.6 (3.4) –
Todarodes sagittatus – 0.2 ± 1.2 (6.9) –
Decapodiform unid. 3.4 ± 18.6 (3.4) 3.8 ± 18.9 (6.9) 1.3 ± 5.5 (5.9)
Thaliacea
‘‘barrels’’ of P. sedentaria 3.6 ± 15.7 (6.9) 2.2 ± 12.0 (10.3) –
Mean ± standard deviation of prey mass percentage and—italic, in parentheses—frequency of occurrence. Mean mass percentages[5% in bold
Mar Biol (2011) 158:1057–1073 1063
123
seasonal patterns in 2005 and in the following 2 years
(Fig. 4). In 2005, blue whiting was the dominant diet item
in June. The proportion of this prey then progressively
decreased, with a higher proportion of krill in early August,
then a higher proportion of anchovy, which was the dom-
inant prey in late September and October. A different
pattern was observed in 2006 and 2007, when the propor-
tion of krill was relatively important in August, and blue
whiting proportion increased then until October, when it
became the dominant prey. Anchovy was absent from the
diet in October in both years. In 2005, the MW% of
anchovy was actually 33.2% in August and 93.7% in late
September and October. In 2006 and 2007, the MW% of
anchovy in the diet was much lower: 19 and 17.5%,
respectively, in late August and late September 2006; 10
and 12.7%, respectively, in early and late August 2007.
No clear relationship was found between the proportion
of krill, anchovy, and blue whiting, and the estimated age
of albacore in the southeastern Bay of Biscay. At a higher
taxonomic level, the proportion of crustaceans, fishes,
cephalopods, and thaliacea in the stomach content of
albacore does not display significant variation among the
different age-groups. All anchovy and blue whiting found
in stomachs were age-0 individuals, according to the
respective age–length relationships of these species
(Aldanondo et al. 2010; Bailey 1974).
Stomach fullness variability
The MCA performed on stomach fullness, length class,
zone, year, and period of catch showed a negative rela-
tionship between length and stomach fullness (Fig. 5), with
Table 4 Prey taxa encountered in non-empty stomachs of albacore caught in Zone 5 of the Bay of Biscay
Catch period 15/06/05–24/10/05 10/08/06–21/10/06 04/08/07–22/10/07
Gear Rod-and-reel Rod-and-reel Rod-and-reel
Sample size (non-empty) n = 259 (186) n = 68 (58) n = 37 (31)
Fishes (teleosts)
Belone belone – – 3.1 ± 17.3 (6.3)
Benthosema glaciale 0.5 ± 7.3 (0.5) – –
Engraulis encrasicolus 19.9 ± 39.1 (21.8) 7.6 – 23.5 (12.1) 7.8 – 25.1 (12.5)
Micromesistius poutassou 22.1 ± 40.7 (24.5) 68.5 – 44.3 (72.4) 57.7 – 48.3 (62.5)
Paralepididae unid. 0.6 ± 7.2 (1.1) – –
Sardina pilchardus – 2.2 ± 13.5 (3.4) –
Scomber scombrus 0.2 ± 1.8 (1.1) – –
Scomberesox saurus – 0.9 ± 6.5 (1.7) 3.2 ± 18.0 (6.3)
Trachurus trachurus 7.8 – 26.5 (8.5) – –
Fish unid. 16.8 – 36.4 (20.2) 8.1 – 24.8 (12.1) 3.3 ± 17.6 (9.4)
Crustaceans
Amphipoda
Phronima sedentaria – \0.1 ± 0.1 (1.7) –
Themisto gaudichaudi \0.1 ± 0.1 (0.5) – –
Decapoda
Pasiphaea sp. 3.7 ± 17.1 (6.4) 1.8 ± 8.1 (12.1) 2.9 ± 16.4 (6.3)
Polybius henslowii 0.1 ± 1.2 (1.1) – –
Scyllarus arctus (larvae) 2.6 ± 14.0 (6.4) 1.8 ± 13.1 (6.9) \0.1 ± 0.1 (6.3)
Euphausiacea
Meganyctiphanes norvegica 22.7 – 40.8 (35.6) 3.4 ± 15.2 (25.9) 21.9 – 39.7 (56.3)
Squillidae unid. 1.3 ± 10.6 (2.7) – –
Mysidae unid. – 1.5 ± 11.1 (1.7) –
Crustacean unid. – 1.7 ± 12.6 (1.7) –
Cephalopods
Todarodes sagittatus 0.2 ± 2.1 (0.6) 0.3 ± 2.1 (1.7) –
Decapodiform unid. 1.8 ± 8.9 (5.2) 2.4 ± 12.3 (10.3) –
Mean ± standard deviation of prey mass percentage and—italic, in parentheses—frequency of occurrence. Mean mass percentages[5% in bold
1064 Mar Biol (2011) 158:1057–1073
123
length classes 1 and 2 associated with high fullness levels,
length classes 3–5 associated with intermediate fullness
levels, and length classes 6 and 7 associated with low
fullness levels. Zones 3 and 4 were associated with length
classes 6 and 7 and with low stomach fullness levels. No
clear relationship between stomach fullness and year or
month was detected.
Given this apparent negative relationship between
stomach fullness and length, albacore fork-length was used
as an explanatory variable in a GAM of stomach fullness.
This GAM showed a negative relationship between stom-
ach fullness and length (P = 2.0 9 10-13), confirming the
relationship demonstrated by the MCAs. The fork-length of
albacore explained 9.32% of the variability of stomach
fullness, which displays important inter-individual vari-
ability (Fig. 6). In zone 1 and zone 5, no significant effect
of catch time on stomach fullness was shown by GAMs.
A linear regression showed a significant relationship
between albacore empty stomach mass and body mass
(P \ 2.10-16, R2 = 0.75 with normally distributed resid-
uals, Shapiro–Wilk P = 0.62), indicative of a proportion-
ality between stomach size and body size of albacore. This
linear relationship is also characterized by an allometry
with slope of 0.874, i.e. a proportionally smaller stomach
for large individuals. In order to take this allometry into
account when relating stomach fullness to body size, we
used a corrected body mass that was defined as the empty
stomach mass estimated by this correlation and scaled
using the mean body mass and the mean empty stomach
mass of our data set. Stomach fullness calculated using the
corrected body mass was modeled as a function of fork-
length through a GAM. This also showed a significant
(P = 1.6 9 10-3) negative relationship between stomach
fullness and fork-length.
Table 5 Prey taxa encountered in non-empty stomachs of albacore caught in the Mediterranean Sea
Area Balearic Sea South Adriatic Sea Tyrrhenian Sea
Catch month Oct.–Nov. 2005 Nov. 2006 Jan. 2008
Gear Longline Longline Longline
Sample size (non-empty) n = 52 (24) n = 50 (24) n = 50 (29)
Fishes (teleosts)
Engraulis encrasicolus 20.0 ± 35.3 (29.2) – –
Paralepis sp. – 21.3 ± 38.3 (29.2) 12.0 ± 25.0 (31.0)
Fish unid. 9.1 ± 26.2 (12.5) 17.6 ± 36.0 (25.0) 17.3 ± 33.8 (27.6)
Crustaceans
Amphipoda
Anchylomera blossevillei – 0.4 ± 1.9 (4.2) 3.3 ± 18.0 (3.4)
Brachyscelus crusculum – 0.4 ± 1.6 (12.5) 0.9 ± 3.6 (6.9)
Platyscelus sp. \0.1 ± 0.1 (4.2) – 0.4 ± 1.6 (6.9)
Phronima sedentaria 1.4 ± 3.8 (20.8) 4.5 ± 11.1 (29.2) 1.0 ± 5.1 (6.9)
Phrosina semilunata 1.8 ± 4.4 (33.3) 5.4 ± 13.2 (33.3) 6.5 – 19.9 (27.6)
Euphausiacea
Meganyctiphanes norvegica 48.9 ± 45.1 (75.0) 35.2 ± 42.2 (79.0) 0.1 ± 0.5 (3.4)
Nematoscelis megalops 4.2 ± 20.4 (4.2) – –
Decapoda
Pasiphaea sp. 4.5 ± 20.4 (12.5) 4.3 ± 17.8 (12.5) 2.0 ± 10.9 (3.4)
Isopoda unid. 0.3 ± 1.4 (4.2) – –
Crustacean unid. 3.8 ± 11.4 (12.5) 7.4 ± 26.7 (8.3) 1.7 ± 7.1 (6.9)
Cephalopods (decapodiform)
Histioteuthis sp. – – 41.7 – 43.5 (55.2)
Illex coindetti – – 3.3 ± 10.0 (10.3)
Decapodiform unid. 0.4 ± 2.0 (8.3) \0.1 ± 0.2 (4.2) 6.6 ± 21.6 (17.2)
Thaliacea
‘‘barrels’’ of P. sedentaria 5.3 ± 19.4 (16.7) 3.5 ± 10.7 (16.7) –
Mean ± standard deviation of prey mass percentage and—italic, in parentheses—frequency of occurrence. Mean mass percentages[5% in bold
Mar Biol (2011) 158:1057–1073 1065
123
Given the strong heterogeneity of the Mediterranean
subset regarding year, zone, and size-range, neither MCA
nor GAM was performed to analyze the variability of
stomach fullness values. Friedman rank sum test results
showed a significantly lower stomach fullness in large
albacore from the Tyrrhenian Sea than in smaller ones from
the Balearic and Adriatic Sea (P = 0.0121) and a signifi-
cantly lower stomach fullness in Mediterranean relative to
Atlantic albacore (P = 0.0074).
Stable isotope analysis
Albacore C/N, d13C, and d15N values significantly differed
among regions for both liver and muscle (Table 7, Fig. 7).
Liver C/N values were significantly lower, and muscle C/N
values higher, for albacore sampled from the Bay of Biscay
in 2005 than for albacore sampled from the South Adriatic
Sea (Table 7). Albacore sampled from the Bay of Biscay in
2006 and from the Tyrrhenian Sea had intermediate values,
with no significant difference. For both tissues, d13C values
were lower and d15N values were higher for albacore from
the Bay of Biscay relative to the two Mediterranean Sea
regions. Albacore from the Tyrrhenian Sea, which com-
prised a larger size class than fish from the other two
regions, had the lowest d15N values. Muscle–liver d15N
differences were consistent across sampling years for
albacore from the Bay of Biscay (*1.2%) and lower than
differences for albacore from the Mediterranean Sea
(*2.0%) (Table 7).
Liver C/N increased with log10length for albacore from
the Adriatic Sea, but not for the other two sampling regions
(Table 7, Fig. 7). Muscle C/N increased with length for
albacore from the Bay of Biscay and Adriatic Sea, but not
for larger individuals from the Tyrrhenian Sea. The rela-
tionship between log10 length and d15N was significant for
both tissue types for all three regions (Table 7).
Fig. 2 Correspondence analysis of the weight percentages of the
main albacore prey by predator, in the Bay of Biscay (predators
represented by open circles) and in the Mediterranean Sea (predators
represented by crosses). Eaeq: Entelurus aequoreus; Bbel: Belonebelone; Ssau: Scomberesox saurus; Tgau: Themisto gaudichaudii;Mpou: Micromesistius poutassou; ceph: cephalopods; Pasi: Pasip-haea sp.; Eenc: Engraulis encrasicolus; Mnor: Meganyctiphanesnorvegica; Sarc: Scyllarus arctus; Nkro: Notoscopelus elongatuskroyeri; Psem: Phrosina semilunata; Para: Paralepis sp.; Hist:
Histioteuthis sp.; Adr: Adriatic Sea; Bal: Balearic Sea; Tyr: Tyrrhe-
nian Sea. Note: predators with identical factorial coordinates are
represented by the same point
Fig. 3 Correspondence analysis of the weight percentages of the
main albacore prey by predator in the Bay of Biscay (left panel) and
in the Mediterranean Sea (right panel). Mean (squares) and standard
errors (bars) of the factorial coordinates on the two first axes for the
different sampling zones. Factorial coordinates of albacore from the
Bay of Biscay (circles, left panel) and from the Mediterranean Sea
(crosses, right panel). See Fig. 2 for prey legends. Bal: Balearic Sea;
Adr: South Adriatic Sea; Tyr: Tyrrhenian Sea. Note: predators with
identical factorial coordinates are represented by the same point
1066 Mar Biol (2011) 158:1057–1073
123
Discussion
Albacore diet and feeding behavior varied spatially and
temporally in the Northeast Atlantic and Mediterranean
Sea. In the Northeast Atlantic, albacore consumed a higher
proportion of crustaceans and a lower proportion of fishes
in the most offshore sampling zone than in shelf-break
zones. The digestion states of the major prey from the
offshore zone reflected a diurnal feeding activity, whereas
in shelf-break zones, feeding appeared to be mainly noc-
turnal. Mediterranean albacore diet displayed a striking
spatial variability, with a particularly high proportion of
cephalopods in the Tyrrhenian Sea. Important seasonal and
inter-annual diet variability was observed in the south-
eastern Bay of Biscay. Stomach fullness decreased with
body size, but in general was higher in albacore caught in
the Northeast Atlantic than in the Mediterranean Sea.
Albacore from the Bay of Biscay had significantly higher
d15N and lower d13C values relative to those from the
Mediterranean Sea. Trophic position and muscle lipid
stores also increased with size.
Comparison with past studies
Important geographic and inter-annual variability in prey
composition was also shown for Atlantic albacore by
Aloncle and Delaporte (1974) on 1756 albacore (39–83 cm
FL) caught by trolling during daytime. Most of the prey
encountered in the present study were identified in previous
works on albacore diet in the Bay of Biscay and sur-
rounding seas, but our results differ regarding several
points. Sternoptychidae (Maurolicus muellerii) and paral-
epidid fish were important prey according to Aloncle and
Delaporte (1974) and to Pusineri et al. (2005) in a study
based on 78 albacore (53–93 cm FL) caught by driftnet at
night off the Bay of Biscay. These species are scarce—in
the case of paralepidid fish—or absent—in the case of
Sternoptychidae—in the present work. Aloncle and Dela-
porte (1974) did not observe blue whiting or anchovy in
albacore diet in the southeastern Bay of Biscay. Both
species were identified in a more recent study (Ortiz de
Zarate 1987) based on 97 albacore—52–90 cm FL and
Table 6 Mean percentage of fishes, cephalopods, crustaceans and thaliacea in the prey weight of 654 albacore sampled in the Bay of Biscay
(zones 1–5) and 152 in the Mediterranean Sea (Balearic Sea, South Adriatic Sea, Tyrrhenian Sea)
Zone Year Catch time Sample size (non-empty) Fishes (%) Cephalopods (%) Crustaceans (%) Thaliacea (%)
Zone 1 2005 Day 51 (46) 37.4 – 60.8 1.8
Zone 2 2004 Night 27 (19) 32.3 0.1 67.6 –
Zone 2 2005 Day 18 (14) 81.2 – 18.8 –
Zone 2 2006 Day 49 (39) 81.9 2.9 15.2 –
Zone 3 2005 Night 66 (29) 19.5 3.4 73.4 3.6
Zone 3 2006 Night 42 (32) 4.1 4.3 89.3 2.2
Zone 4 2006 Night 37 (22) 41.2 1.3 57.5 –
Zone 5 2005 Day 259 (186) 66.3 2.0 29.7 –
Zone 5 2006 Day 68 (58) 87.3 2.6 10.1 –
Zone 5 2007 Day 37 (31) 75.1 – 24.9 –
Balearic 2005 Night 52 (24) 29.1 0.4 65.2 5.3
S. Adriatic 2006 Night 50 (24) 38.8 \0.1 57.6 3.5
Tyrrhenian 2008 Night 50 (29) 29.4 51.6 19.0 –
Fig. 4 Correspondence analysis of the weight percentages of blue
whiting (Mpou), krill (Mnor), and anchovy (Eenc) in the diet of
albacore caught in zone 5, barycenters of the factorial coordinates by
periods for 2005 (squares, solid line), 2006 (circles, solid line), and
2007 (triangles, dotted line, italic charcters). jn: June; jl1: July 1–15;
jl2: July 16–31; au1: August 1–15; se2: September 16–30; oc:
October
Mar Biol (2011) 158:1057–1073 1067
123
troll-caught in the Bay of Biscay during daytime— but the
main prey of albacore according to this study (i.e. horse
mackerel, Trachurus trachurus) is poorly represented in
our observations. In the absence of long-term information
on biomass of these different prey in the Northeast
Atlantic, it is difficult to relate differences in albacore diet
to possible variations in prey abundance. Overall, the
proportion of cephalopods was low in albacore caught in
the Bay of Biscay.
Compared with past isotope studies (Das et al. 2000),
Atlantic albacore in our study had lower mean d15N values
for both liver and muscle. Differences may reflect a lower
trophic position for albacore sampled in our study relative
to fish sampled farther offshore in the early 1990s. Dif-
ferences between studies could also be influenced by lipid
correction procedures as the method employed by Das
et al. (2000) causes increases in d15N values in tuna muscle
(Logan et al. 2008). Das et al. (2000) observed two liver
isotope groups, which they attributed to dietary segregation
with one group feeding predominantly on squids and the
other feeding instead on small fishes. Our liver values
closely matched values observed by Das et al. (2000) for
the latter group. The group that fed mainly on cephalopods
was not reflected in either our stomach content or our
isotope results.
In the present work, paralepidid fishes, squids Histio-
teuthis sp. and Illex coindetti, and all the crustaceans
(except Pasiphaea sp.) found in Tyrrhenian Sea albacore
were also identified in a study by Consoli et al. (2008).
Higher proportions of cephalopod prey, specifically from
the families Histioteuthidae, Ommastrephidae, and Sepio-
lidae, were found in albacore sampled in September and
October from the South Adriatic Sea (Bello 1999). Given
observed inter-annual variability in albacore from the Bay
of Biscay, the lower proportions in albacore sampled from
the South Adriatic Sea in this study could be due to inter-
annual dietary shifts.
Comparison with another local predator: bluefin tuna
Compared with juvenile bluefin tuna (Thunnus thynnus)
sampled in the Mediterranean Sea (Sara and Sara 2007),
albacore isotope values fall between bluefin size classes of
1–2.2 kg and 15–30 kg, suggesting that they probably were
feeding at the same trophic position as bluefin of similar
size in this region.
Liver and muscle isotope values were similar to juvenile
bluefin tuna ranging from 60 to 95 cm FL sampled from
the Bay of Biscay (Logan et al. 2011), suggesting that they
feed at a similar trophic position. Liver C/N values were
high and similar to bluefin tuna from the area, indicating
high lipid content in this tissue. White muscle had high
lipid stores compared with bluefin tuna from the area, and
muscle-liver isotope difference was also similar to values
measured for bluefin tuna. However, this is discrepant with
stomach content observations, which display important
differences between juvenile albacore and age-1 bluefin
tuna in the southeastern Bay of Biscay, fish MW% being
significantly higher and crustacean MW% significantly
Fig. 5 Multiple
correspondence analysis of
length (l1–l7), stomach
repletion (r0–r6), year
(2005–2007), month (see
Fig. 3), and sampling zone (z1–
z5) of albacore caught in the
Bay of Biscay. For an easier
reading, coordinates of the
modalities of the two first
factors are represented on the
left panel, and coordinates of
the modalities of the other
factors on the right panel
Fig. 6 Observed (gray crosses) and GAM fitted (solid line) stomach
fullness of albacore in the Bay of Biscay as a function of fork-length;
mean values (squares) and standard deviation (bars) of stomach
fullness and fork-length by age-group (0 to 5?)
1068 Mar Biol (2011) 158:1057–1073
123
lower, in stomach contents of age-1 bluefin than in alba-
core for a given period (Goni 2008). These differences in
stomach content, in spite of a similar trophic position
indicated by d15N ratios, suggest these two species might
have different ecological niches for the size-ranges
considered.
Table 7 Stable isotope d13C (%) and d15N (%) values, and carbon/nitrogen ratios (C/N) in liver and white muscle of albacore from the Bay of
Biscay, South Adriatic Sea, and Tyrrhenian Sea (Mean ± standard deviation)
Region Bay of Biscay (zone 5) South Adriatic Sea Tyrrhenian Sea
Year (sample size) 2005 (n = 31) 2006 (n = 10) 2006 (n = 30) 2008 (n = 30)
Fork-length 63.2 ± 8.5a,b 61.1 ± 2.3b 64.4 ± 3.9a 75.0 ± 3.0c
Liver d13C (%) -18.3 ± 0.3a -18.2 ± 0.3a -17.5 ± 0.5b -17.5 ± 0.3b
d15N (%) 9.8 ± 0.3a 9.7 ± 0.3a 7.4 ± 0.3b 6.8 ± 0.4c
C/N 6.3 ± 1.1a 6.3 ± 1.2a.b 8.1 ± 3.0b 7.2 ± 2.0a,b
Muscle d13C (%) -18.4 ± 0.2a -18.6 ± 0.2b -17.7 ± 0.1c -17.7 ± 0.3c
d15N (%) 11.0 ± 0.4a 10.9 ± 0.3a 9.4 ± 0.3b 8.5 ± 0.5c
C/N 4.1 ± 0.6a 3.8 ± 0.4a,b 3.7 ± 0.5b 4.0 ± 0.6a,b
Muscle-Liver difference d13C (%) -0.1 ± 0.2a -0.4 ± 0.3b -0.3 ± 0.4a,b -0.2 ± 0.4a,b
d15N (%) 1.2 ± 0.4a 1.2 ± 0.4a 2.0 ± 0.3b 1.8 ± 0.3c
r2 F1,39 P r2 F1,28 P r2 F1,28 P
Liver
d15N*log10 (FL) 0.16 7.60 0.009 0.14 4.61 0.041 0.23 8.25 0.008
C/N*log10 (FL) \0.01 0.18 0.671 0.32 13.33 0.001 0.02 0.65 0.426
r2 F1,38 P r2 F1,28 P r2 F1,28 P
Muscle
d15N*log10 (FL) 0.14 6.25 0.017 0.43 20.78 <0.001 0.17 5.81 0.023
C/N*log10 (FL) 0.29 15.61 <0.001 0.32 13.33 0.001 \0.01 0.03 0.868
Row values with different letter superscripts (a, b, or c) are significantly different (P \ 0.05 in Student test). Row values with two letter
superscript (a, b) indicate no significant difference between the corresponding two groups. Linear regressions of d15N values and carbon/nitrogen
ratios in function of log10 of fork-length (d15N*log10 (FL) and C/N*log10 (FL), respectively). Significant relationships indicated in bold
Fig. 7 d15N and carbon/
nitrogen ratios in liver (leftpanels) and white muscle (rightpanels) plotted versus log [fork-
length] of albacore sampled in
the Bay of Biscay (circles, solidline), in the South Adriatic Sea
(squares, dotted line) and in the
Tyrrhenian Sea (triangles,
dashed line)
Mar Biol (2011) 158:1057–1073 1069
123
Possible influences of fishing and sampling time
on stomach content
Fishing gear is a possible source of variability of stomach
contents (Bertrand et al. 2002). In our case, geographic
variations in stomach contents are difficult to interpret
when they involve fishing gears operating during the day
(e.g. trolling, rod-and-reel) and at night (e.g. pelagic
trawling, longline). If we assume that krill and small
crustaceans are rapidly digested (Aloncle and Delaporte
1974), as small prey with a low lipid content (Olson and
Boggs 1986), the higher proportion of krill in stomachs of
albacore caught at night is more likely to be related to a
higher availability of this prey to albacore at night, rather
than to dramatic spatial (zones 3 and 4 vs. 1, 2 and 5) or
temporal (2004 vs. 2005 and 2006 for zone 2) changes in
krill abundance. This circadian variation in krill availabil-
ity can therefore induce a bias if albacore are not caught at
all hours in a given area. Krill will then be overrepresented
in the stomachs of albacore caught at night and underrep-
resented in the stomachs of albacore caught during day-
time. Moreover, differences in sampling times can be
another source of bias in observed stomach content. In our
case, their effect cannot be distinguished from the possible
effect of catch time, as fish caught at night (by pelagic
trawl or longline) were commercial fish that were sampled
1–6 days after catch, whereas fish caught by day were
sampled onboard the fishing vessels a few minutes after
catch.
Inter-basin comparison
Comparison of albacore stomach contents between the
Bay of Biscay and the Mediterranean Sea is limited, and
the differences in fishing gears, catch times, and sampling
periods do not allow a thorough interpretation of the
observed differences. The difference in stomach fullness
between Atlantic and Mediterranean albacore could result
from higher energy needs in North Atlantic albacore,
possibly related to a higher growth rate (Santiago and
Arrizabalaga 2005; Megalofonou 2000), to lower tem-
peratures, or to the energetic cost of a long-range seasonal
migration. From another point of view, this higher growth
and these long-range migrations in North Atlantic alba-
core could also result from better feeding possibilities in
the North Atlantic. However, region and fishing gear are
not independent (longline in the Mediterranean Sea versus
pelagic trawling and surface baited gears in the Bay of
Biscay), so our interpretation of this result remains lim-
ited. It is also the case for comparisons within the Med-
iterranean Sea, where the lower stomach fullness in large
albacore from the Tyrrhenian Sea than in smaller albacore
from the Balearic and South Adriatic Seas could either be
related to the size of the individuals or to differences
in prey abundance and availability between these zones
or years. The non-independence of size, sampling zone,
and year do not allow broader interpretation of this
observation.
The lower d13C values and higher d15N values in Bay of
Biscay fish than in Mediterranean fish likely reflect regio-
nal baseline shifts, rather than trophic shifts (Sara and Sara
2007). Baseline d15N values appear lower for regions
within the Mediterranean Sea than Bay of Biscay waters.
Tuna prey items sampled in both regions show higher d15N
values for the Biscay region relative to the Tyrrhenian Sea
region (e.g. horse mackerel *1.8% higher d15N value for
the Bay of Biscay (Bode et al. 2007) relative to the Tyr-
rhenian Sea (Pinnegar et al. 2003), while sardines are
*2.8% higher across regions). For d13C, the baseline shift
appears to occur in the opposite direction, with lower
values in Atlantic waters than in the Mediterranean Sea.
These baseline shifts are consistent with the very low
mixing rate between Atlantic and Mediterranean albacore
populations, revealed by tagging (Arrizabalaga et al. 2002).
Similar spatial shifts have been observed in bluefin tuna
and other large pelagic fishes between coastal feeding areas
in the NW Atlantic and offshore feeding areas in the central
Atlantic, with fish from coastal areas showing higher d15N
but lower d13C values than offshore conspecifics (Logan
2009). Albacore from both Mediterranean feeding areas
had higher d15N difference values relative to albacore from
the Bay of Biscay. These greater differences in d15N
between liver and muscle could reflect recent movements
of Mediterranean albacore from more coastal feeding areas
to more offshore regions where they were sampled during
autumn and winter. Alternatively, these differences could
also be due to recent seasonal changes in prey availability
altering diet composition.
The lower d15N values in large albacore from the
Tyrrhenian Sea relative to smaller albacore from the
South Adriatic Sea could reflect a lower trophic position
prey base for the Tyrrhenian Sea or alternatively may
reflect regional baseline shifts within the Mediterranean
Sea. Given that stomach contents contained mainly krill
in the Adriatic Sea and histioteuthid squids in the Tyr-
rhenian Sea and given that d15N increases with size in
both regions, a baseline shift is a more likely explanation
for these regional differences. Spatial isotope differences
between these two feeding areas suggest a separation
between these two assemblages and a reduced range rel-
ative to North Atlantic albacore. This is consistent with
the reduced movements of Mediterranean albacore,
observed by tagging and recapture (Arrizabalaga et al.
2002), and with the existence of separate spawning
grounds in the Tyrrhenian Sea and in the South Adriatic
Sea (Marano et al. 1999).
1070 Mar Biol (2011) 158:1057–1073
123
Size-related patterns
The higher stomach fullness in small individuals in the Bay
of Biscay may correspond to higher energetic needs (Goni
and Arrizabalaga 2010), influencing feeding rates and
behavior. These energetic needs could be explained by a
higher growth rate for small albacore (Santiago and Arri-
zabalaga 2005). They could also be related to a higher
energetic cost of maintaining neutral buoyancy, where
muscular activity might be required to compensate for the
lack of a fully functional swim bladder, which is not
developed until they reach 85 cm FL (Bard 1981). Another
possible explanation for higher apparent feeding rates
could be a higher energetic cost of thermoregulation for
small individuals, due to a lower surface–volume ratio that
favors heat losses (Schmidt-Nielsen 1984). The allometry
observed between empty stomach mass and body mass
could also be a cause of a lower fullness for large indi-
viduals. However, when taking this allometry into account,
the relationship between stomach fullness and fork-length
is still significant (although with a lower significance
level), so stomach fullness is also determined by feeding
behavior.
Graham et al. (2007) did not identify any clear rela-
tionship between body size and stomach fullness for
juvenile yellowfin (Thunnus albacares) near Hawaii.
However, they identified diet variations as a function of
body size, although no clear variation was observed
between 50 and 100 cm FL, which corresponds to the size-
range of our sample. South Pacific albacore undergo diet
shifts at around 60–69 cm FL, the main prey of small
individuals being cephalopods and small crustaceans, and
teleost fishes for larger individuals (Bailey and Habib
1982). No similar diet change with body size was observed
in the present study or in a recent study on albacore diet in
the Tyrrhenian Sea (Consoli et al. 2008). Our isotope
results did indicate an average trophic position increase
with size, not reflected by stomach content observations,
both in the Bay of Biscay and in the Mediterranean Sea.
However, the relationship between d15N and log10 length,
although significant in all subsets for both liver and white
muscle, has a particularly weak slope compared with
bluefin tuna (Sara and Sara 2007) or with a tropical tuna
species such as yellowfin (Graham et al. 2007). This weak
increase in trophic position (indicated by d15N) with body
size may explain the absence of diet change—as observed
through stomach contents—with body size.
The increase of C/N ratio in white muscle along with
albacore size for the South Adriatic and Bay of Biscay
individuals and the absence of relationship between C/N
ratio and length for larger Tyrrhenian albacore are con-
sistent with the increase in fat content along with albacore
size observed by Goni and Arizabalaga (2010). According
to this study, this increase is weaker or absent for indi-
viduals larger than 80 cm.
Trophic features of offshore versus continental slope
areas in the Bay of Biscay
The respective circadian patterns in digestion states of the
main prey offshore (zone 1) and in the southeastern Bay of
Biscay (zone 5) suggest that in zone 5, krill, blue whiting,
and anchovy are consumed at nighttime and/or dawn,
consistent with patterns observed by Aloncle and Delaporte
(1974). In zone 1, krill and T. gaudichaudii may be con-
sumed during daytime. Due to the circadian vertical
movements of krill (Mauchline 1980) and of small pelagic
fishes (Freon and Misund 1999), it is therefore likely that
krill and T. gaudichaudii are consumed during deeper
vertical movements offshore (zone 1), whereas in the slope
area (zone 5), albacore likely feed on krill and blue whiting
in surface waters, as proposed by Goni et al. (2009).
Among the areas where albacore were caught during the
daytime, the most offshore region (zone 1) had the highest
proportion of crustaceans and the lowest proportion of
fishes in stomach contents. In the California current eco-
system, crustaceans in albacore diet have a much lower
caloric content than fishes (Glaser 2009). If we assume a
similar difference between crustaceans and fishes in alba-
core diet in the Northeast Atlantic, and considering that
feeding in zone 1 requires feeding on prey located at
greater depths, offshore areas may be less energetically
suitable for albacore (i.e. lower energy input from diet to
higher energetic cost of feeding) than inshore areas (e.g.
zone 5). Different feeding strategies between inshore and
offshore can also affect albacore catchability by surface
gears, particularly in the case of baited gears like trolling
and baitboat.
Seasonal diet variability in the southeastern
Bay of Biscay
In the southeastern Bay of Biscay (zone 5), blue whiting
was the main albacore diet component in July 2005, and in
September and October 2006 and 2007. Anchovy was
present in 2005 in albacore diet in its earliest juvenile stage
(body length between 40 and 60 mm) in late July, then
became the dominant prey at the end of the fishing season.
Before late July, age-0 anchovy occur inshore, still mainly
in the larval phase, associated with plankton (Irigoien et al.
2007). In 2005, the biomass of age-0 anchovy was around
134 000 tons, whereas in 2006 and 2007, the respective
abundances were around 78 000 tons and 13 000 tons
(Boyra et al. 2008). During 2006, anchovy were mainly
distributed on the shelf, close to the coast, where they were
probably less available to albacore. These elements of the
Mar Biol (2011) 158:1057–1073 1071
123
anchovy life cycle in relation to the biomass and spatial
distribution of juvenile anchovy suggest that in the south-
eastern Bay of Biscay, albacore select anchovy in prefer-
ence to blue whiting. This selectivity is probably due to a
higher caloric content in anchovy than in blue whiting
(Soriguer et al. 1997). However, the partial substitution of
anchovy by blue whiting in albacore diet in the south-
eastern Bay of Biscay from 2005 to 2007 did not seem to
affect albacore energetics, at least in terms of lipid storage
(Goni and Arrizabalaga 2010). Moreover, this selectivity
is lower for albacore than for age-1 bluefin in the same
area (Goni 2008). The seasonal and interannual patterns
observed in the diet of albacore caught in zone 5 tend to
question the representativeness of their diet as observed in
other areas, in which the temporal distribution of samples
is narrower. It is particularly critical in the case of zone 3,
in which in both years, the whole data subsets come from
single fishing trips. Our results regarding the geographic
diet variability within the Bay of Biscay should therefore
be considered cautiously.
Acknowledgments We are very grateful to Deirdre Brophy and
GMIT samplers, Jean-Pierre Esain, Jean-Hilaire de Bailliencourt,
Luis Alberto ‘‘Luxia’’ Martın, Luis Arregi, Inigo Onandia, Jose-Angel
Fernandez, Peio Olazabal, Enrique Keler, and collaborating recreative
fishermen for providing stomach samples. We thank Irene Gomez,
Deniz Kukul, Maite Cuesta, and Inaki Rico for their help in stomach
content observations, Lucıa Zarauz and Aitor Albaina for their help in
crustacean identifications, David Milly for the information on trawler
fishing zones, and Andrew Ouimette for assistance with stable isotope
analyses. We finally thank both anonymous reviewers and associate
editor for their comments on an earlier version of this manuscript.
This work was partly funded by a PhD grant from the Fundacion
Centros Tecnologicos to N. Goni and by NOAA grant no.
NA04NMF4550391 to M. Lutcavage. This paper is contribution
number 524 from AZTI-Tecnalia (Marine Research). This paper is a
contribution to the CLIOTOP (Climate Impacts on Oceanic Top
Predators) project. The experiments complied with the current laws of
the countries in which they were performed.
References
Aldanondo N, Cotano U, Tiepolo M, Boyra G, Irigoien X (2010)
Growth and movement patterns of early juvenile European
anchovy (Engraulis encrasicolus L.) in the Bay of Biscay based
on otolith microstructure and chemistry. Fish Oceanogr
19(13):196–208
Aloncle H, Delaporte F (1974) Donnees nouvelles sur le germon
Atlantique Thunnus alalunga Bonnaterre 1788 dans le Nord-Est
Atlantique. 1ere Partie—Rythmes alimentaires et circadiens.
Rev Travaux de l’Institut Peches Maritimes 37 (4):475–572
Arrizabalaga H, Lopez Rodas V, Ortiz de Zarate V, Costas E,
Gonzalez-Garces A (2002) Study on the migrations and stock
structure of albacore (Thunnus alalunga) from the Atlantic
Ocean and the Mediterranean Sea based on conventional tag
release-recapture experiences. Col Vol Sci Pap ICCAT
54(5):1479–1494
Bailey RS (1974) The life-history and biology of blue whiting in the
Northeast Atlantic. Mar Res 1:29
Bailey KN, Habib G (1982) Food of incidental fish species taken in
the purse seine skipjack fishery, 1976–1981. Fisheries Research
Division Occasional Publication (N.Z. Ministry of Agriculture
and Fisheries). Data Series 6, 24 p
Bard FX (1981) Le thon germon (Thunnus alalunga Bonaterre 1788)
de l’Ocean Atlantique. De la dynamique des populations a la
strategie demographique. These de Doctorat d’Etat. Universite
Paris VI:333 pp
Bello G (1999) Cephalopods in the diet of albacore, Thunnusalalunga, from the Adriatic Sea. J Molluscan Stud 65:233–240
Benzecri JP (1973) L’analyse des donnees, Tome 2: L’analyse des
correspondances. Bordas, Paris
Benzecri JP (1980) Pratique de l’Analyse de Donnees. Dunod, Paris
Bertrand A (1999) Le systeme thon-environnement en Polynese
Francaise: caracterisation de l’habitat pelagique, etude de la
distribution et de la capturabilite des thons, par methodes
acoustiques et halieutiques. These de l’Ecole Nationale Superi-
eure Agronomique de Rennes, 295 p
Bertrand A, Bard FX, Josse E (2002) Tuna food habits related to the
micronekton distribution in French Polynesia. Mar Biol
140:1023–1037
Bode A, Alvarez-Ossorio M, Cunha M, Garrido S, Peleteiro J,
Porteiro C, Valdes L, Varela M (2007) Stable nitrogen isotope
studies of the pelagic food web on the Atlantic shelf of the
Iberian Peninsula. Prog Oceanogr 74:115–131
Boyra G, Cotano U, Martinez U, Pena M, Uriarte A (2008) JUVENA
series review of the spatial distribution of anchovy juveniles in
the Bay of Biscay. XI International Symposium on Oceanogra-
phy of the Bay of Biscay, San Sebastian (Spain), April 2008
Clarke MR (1986) A handbook for the identification of cephalopod
beaks. Clarendon Press, Oxford
Consoli P, Romeo T, Battaglia P, Castriota L, Esposito V, Andaloro F
(2008) Feeding habits of the albacore tuna Thunnus alalunga(Perciformes, Scombridae) from central Mediterranean Sea. Mar
Biol 155:113–120
Das K, Lepoint G, Loizeau V, Debacker V, Dauby P, Bouquegneau
JM (2000) Tuna and dolphin associations in the Northeast
Atlantic: evidence of different ecological niches from stable
isotope and heavy metal measurements. Mar Pollut Bull
40(2):102–109
de la Serna JM, Valeiras J, Alot E, Godoy D (2003) El atun blanco
(Thunnus alalunga) del Mediterraneo Occidental. Col Vol Sci
Pap ICCAT 55(1):160–165
Estrada JA, Lutcavage M, Thorrold SR (2005) Diet and trophic
position of Atlantic bluefin tuna (Thunnus thynnus) inferred from
stable carbon and nitrogen isotope analysis. Mar Biol 147:37–45
Freon P, Misund OA (1999) Dynamics of pelagic fish distribution and
behaviour: effects on fisheries and stock assessment. Blackwell
Science, London
Froese R, Pauly D (eds) (2010) FishBase. World Wide Web electronic
publication. http://www.fishbase.org, version (05/2010)
Glaser SM (2009) Foraging ecology of North Pacific Albacore in the
California Current System (CCS). California Sea Grant College
Program, UC San Diego. Retrieved from: http://escholarship.org/
uc/item/7130n4r0
Goni N (2008) Habitat et ecologie trophique du germon (Thunnusalalunga) dans l’Atlantique Nord-Est: variabilite, implications
sur la dynamique de la population. These de Doctorat, Universite
de Pau et des Pays de l’Adour, 198 pp
Goni N, Arrizabalaga H (2010) Seasonal and interannual variability
of fat content of juvenile albacore (Thunnus alalunga) and
bluefin (Thunnus thynnus) tuna during their feeding migration to
the Bay of Biscay. Prog Oceanogr 86(1–2):115–123
Goni N, Arregui I, Lezama A, Arrizabalaga H, Moreno G (2009)
Small scale vertical behaviour of juvenile albacore in relation to
their biotic environment in the Bay of Biscay. In: Nielsen J,
1072 Mar Biol (2011) 158:1057–1073
123
Arrizabalaga H, Fragoso N, Hobday A, Lutcavage M, Sibert J
(eds) Tagging and tracking of marine animals with electronic
devices (Reviews: Methods and Technologies in Fish Biology
and Fisheries), vol 2. Springer, Berlin, pp 51–73
Graham BS, Grubbs D, Holland K Popp BN (2007) A rapid
ontogenetic shift in the diet of juvenile yellowfin tuna from
Hawaii. Mar Biol 150(4):647–658
Hastie TJ (1992) Generalized additive models. In: Chambers JM,
Hastie TJ (eds) Statistical models in S. Chapman & Hall,
London, pp 249–307
Holm S (1979) A simple sequentially rejective multiple test
procedure. Scand J Stat 6:65–70
Ibanez Artica M, Menendez de la Hoz M, Matallanas J, Ramos A,
Sanchez F, San Millan MD (1989) Euskal Herriko arrainak.
ISBN: 978-84-7728-128-3, Editorial Kriselu SA, Donostia
(Spain)
ICCAT (2008) Report of the 2007 ICCAT albacore stock assessment
session (Madrid, Spain, July 5 to 12, 2007). Collect Vol Sci Pap
ICCAT 62(3):697–815
Irigoien X, Fiksen Ø, Cotano U, Uriarte A, Alvarez P, Arrizabalaga
H, Boyra G, Santos M, Sagarminaga Y, Otheguy P, Etxebeste E,
Zarauz L, Artetxe I, Motos L (2007) Could Biscay Bay anchovy
recruit through a spatial loophole? Prog Oceanogr 74(2–3):
132–148
Korsmeyer KE, Dewar H (2001) Tuna metabolism and energetics. In:
Block BA, Stevens ED (eds) Tunas: physiology, ecology and
evolution. Academic Press, San Diego, pp 35–78
Laval P (1978) The barrel of the pelagic amphipod Phronimasedentaria (Forsk.) (crustacea: hyperiidea). J Exp Mar Biol Ecol
33:187–211
Logan J (2009) Tracking diet and movement of Atlantic bluefin tuna
(Thunnus thynnus) using carbon and nitrogen stable isotopes.
PhD Thesis. University of New Hampshire, Durham, NH
Logan J, Haas H, Deegan L, Gaines E (2006) Turnover rates of
nitrogen stable isotopes in the salt marsh mummichog, Fundulusheteroclitus, following a laboratory diet switch. Oecologia
147:391–395
Logan J, Jardine T, Miller T, Bunn S, Cunjak R, Lutcavage M (2008)
Lipid corrections in carbon and nitrogen stable isotope analyses:
comparison of chemical extraction and modelling methods.
J Anim Ecol 77:838–846
Logan JM, Rodriguez-Marın E, Goni N, Barreiro S, Arrizabalaga H,
Golet W, Lutcavage ME (2011) Diet of young Atlantic bluefin
tuna (Thunnus thynnus) in eastern and western Atlantic foraging
grounds. Mar Biol 158:73–85
Marano G, De Zio V, Pastorelli AM, Rositani L, Ungaro N, Vlora A
(1999) Synopsis on the biology and fisheries on Thunnus
alalunga (Bonnaterre, 1788). Biol Mar Medit 6(2):192–214
Mauchline J (1980) The biology of mysids and euphausiids. Adv Mar
Biol 18:371–678
Megalofonou P (2000) Age and growth of Mediterranean albacore.
J Fish Biol 57:700–715. doi:10.1111/j.1095-8649.2000.tb00269.x
Olson RJ, Boggs CH (1986) Apex predation by yellowfin tuna
(Thunnus albacares): independent estimates from gastric evac-
uation and stomach contents, bioenergetics, and cesium concen-
trations. Can J Fish Aquat Sci 43:1760–1775
Ortiz de Zarate V (1987) Datos sobre la alimentacion del atun blanco
(Thunnus alalunga B.) juvenil capturado en el Golfo de Vizcaya.
Col Vol Sci Pap ICCAT 26(2):243–247
Peterson BJ, Fry B (1987) Stable isotopes in ecosystem studies. Annu
Rev Ecol Syst 18:293–320
Phillips DL, Eldridge PM (2006) Estimating the timing of diet shifts
using stable isotopes. Oecologia 147:195–203
Pinnegar JK, Polunin NVC, Badalamenti F (2003) Long-term changes
in the trophic level of western Mediterranean fishery and
aquaculture landings. Can J Fish Aquat Sci 60:222–235
Post DM (2002) Using stable isotopes to estimate trophic position:
models, methods, and assumptions. Ecology 83:703–718
Pusineri C, Vasseur Y, Hassani S, Meynier L, Spitz J, Ridoux V
(2005) Food and feeding ecology of juvenile albacore, Thunnusalalunga, off the Bay of Biscay: a case study. ICES J Mar Sci
62(1):116–122
Salman A, Karakulak S (2009) Cephalopods in the diet of albacore,
Thunnus alalunga, from the Eastern Mediterranean. J Mar Biol
Assoc UK 89(3):635–640
Santiago J (1993) A new length-mass relationship for the North
Atlantic albacore. Col Vol Sci Pap ICCAT 40(2):316–319
Santiago J (2004) Dinamica de la poblacion de atun blanco (Thunnus
alalunga Bonaterre 1788) del Atlantico Norte. PhD Thesis,
Euskal Herriko Unibertsitatea, Bilbao, 320 pp
Santiago J, Arrizabalaga H (2005) An integrated growth study for
North Atlantic albacore (Thunnus alalunga Bonn. 1788). ICES J
Mar Sci 62(4):740–749
Sara G, Sara R (2007) Feeding habits and trophic levels of bluefin
tuna Thunnus thynnus of different size classes in the Mediter-
ranean Sea. J Appl Ichthyol 23:122–127
Schmidt-Nielsen K (1984) Scaling: why is animal size so important?
Cambridge University Press, Cambridge, 242 pp
Soriguer F, Serna S, Valverde E, Hernando J, Martın-Reyes A,
Soriguer M, Pareja A, Tinahones F, Esteva I (1997) Lipid,
protein, and calorie content of different Atlantic and Mediter-
ranean fish, shellfish, and molluscs commonly eaten in the South
of Spain. Eur J Epidemiol 13:451–463
R Development Core Team (2008) R: a language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. ISBN 3-900051-07-0, URL:http://www.
R-project.org
Todd CD, Laverack MS, Boxshall G (1996) Coastal Marine
Zooplankton: a practical manual for students, 2nd edn. Cam-
bridge University Press, Cambridge, 106 pp
Mar Biol (2011) 158:1057–1073 1073
123
