Psychology : Investigating Behaviour 1 Research Methods Investigating Behaviour.
Back to the future: Investigating historical data from ...€¦ · Investigating historical data...
Transcript of Back to the future: Investigating historical data from ...€¦ · Investigating historical data...
Jean-Marc Fromentin
Investigating historical data from Atlantic
bluefin tuna fisheries
Back to the future:
Exploitation since the first millennium before Christ
Generalcontext
Bluefin tuna catch did not remain stable but displayed conspicuous long-term fluctuations,
possibly in relation to changes in temperature (Ravier and Fromentin 2001; 2004)
0
20000
40000
1600 1700 1800 1900
0
5000
10000
1600 1700 1800 1900
0
5000
1600 1700 1800 1900
0
5000
10000
1600 1700 1800 1900
0
4000
8000
1600 1700 1800 1900
Medo das Casas,
Portugal
1600 1700 1800 1900
0
10000
20000Zahara,
Spain
SalineSardinia
Porto ScusoSardinia
Favignana,
Sicily
FormicaSicily
Sidi Daoud,
Tunisia
0
5000
10000
1600 1700 1800 1900
70,0%
75,9%
57,8%
82,2%
74,7%
78,8%
63,4%
Skagen 1950 Bergen 1955
What’s happen to BFT in the North
Sea and Norwegian Sea?
Changes in migration patterns
Restriction of the spatial distribution
Recruitment failure/overfishing
Eradication of a sub-population
However, the leading Nordic fisheries were the most spectacular
example of Atlantic bluefin tuna fisheries collapse (Tiews 1978)
These fisheries suddenly collapsed in 1963
without any warning
Land
ings
(ton
nes)
0
5000
10000
15000
20000
1920 1930 1940 1950 1960 1970 1980 1990 2000
TheNordicfisheriescase
TheNordicfisheriescase
Nordic catch-at-size displays an unusual pattern
NordicNordicNordicNordic Fisheries
FisheriesFisheriesFisheries Collapse
CollapseCollapseCollapse
Retrospective Analysis
Italian TP(1900-00 / 1956-79)
Spanish TP(1900-00 / 1956-00)
Jap. LL (Med.)(1963-00; 1974-00)
German PS(1947-62; 1951-62)
Norwegian PS
Jap. LL (E. Atl.)(1957-00; 1974-00)
US PS(1958-00; 1963-00)
US TP(1930-73; 1955-61)
Jap. LL (W. Atl.)(1957-00; 1971-00)
Spanish BB(1940-00 / 1966-00)
BB: Bait BoatLL: Long LinePS: Purse SeineTP: Trap
French & Italian PS(1950-00; 1971-00)
(1927-79; 1956-79)
Retrospective Analysis: yeild
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
10000
20000
Yie
ld(t
) NEA-TP
Yie
ld(t
)
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
10000
20000NEA-PS
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
2000
4000
Yie
ld(t
) NEA-LL
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
2000
4000
Yie
ld(t
)
NEA-BB
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
5000
10000
Yie
ld(t
)
NWA-LL
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
1000
2000
3000
Yie
ld(t
) MED-TP
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
10000
20000
30000
Yie
ld(t
) MED-PS
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
4000
8000
Yie
ld(t
) MED-LL
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
500
1000
Yie
ld(t
)
NWA-TP
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
2000
4000
6000
Yie
ld(t
) NWA-PS
-0.5 0 0.5
-0.5
0
0.5
MED-TP
MED-PSMED-LL
NEA-TP
NEA-PS
NEA-LL
+
NEA-BB+
NWA-PS NWA-LL
axis 1 (38%)
axis
2 (
22%
)
+
++
+
+ +
++
1950 1960 1970 19800
2000
4000
6000Group 3
1950 1960 1970 19800
2000
4000
6000
8000Group 1
1950 1960 1970 19800
5000
10000Group 2
00.5
11.5
2
NE
A.T
P
ME
D.T
P
NE
A.P
S
NW
A.T
P
NE
A.B
B
ME
D.P
S
NE
A.LL
ME
D.LL
NW
A.LL
NW
A.P
S
Group 1Group 3Group 2NWA-TP
Retrospective Analysis: yeild
Median Dispersion Dispersion=f(Median)
1955 1960 1965 1970 1975 1980180
220
260
1955 1960 1965 1970 1975 1980
40
80
180 200 220 240 260
40
80
1955 1960 1965 1970 1975 1980
200
240
1955 1960 1965 1970 1975 198050
100
170 180 190 200 210 220 230 24050
100
1955 1960 1965 1970 1975 1980 1985150
200
1955 1960 1965 1970 1975 1980 1985
50
100
160 180 200 220 240
50
100
1965 1970 1975 198050
100
150
1965 1970 1975 1980
40
80
120
50 100 150
40
80
120
1970 1975 1980 1985 1990
100
150
200
1970 1975 1980 1985 1990
100
150
100 150 200
100
150
1970 1975 1980 1985 1990160
200
240
1970 1975 1980 1985 1990
60
100
140
160 180 200 220 240
60
100
140
through time…
NEA PS
NEA TP
MED TP
NWA LL
NWA PS
NEA BB
Retrospective Analysis: Size
Retrospective Analysis: Size
Yield vs Median size
NEA-PS
NEA-TP
MED-TP
NWA-PS
Yield vs Dispersion size
0 2000 4000 6000 8000 10000180
220
260 p < 0.001
0 2000 4000 6000 8000 1000020
40
60
80
p < 0.001
0 4000 8000 12000 16000150
200
250p < 0.001
0 4000 8000 12000 1600040
80
120
p > 0.1
0 500 1000 1500 2000 2500 3000
160
200
240p < 0.001
0 500 1000 1500 2000 2500 3000
60
80
100p < 0.001
0 1000 2000 3000 4000 5000 600050
100
150p > 0.1
0 1000 2000 3000 4000 5000 6000
40
80
120p > 0.1
Yield (ton) Yield (ton)
Med
ian
leng
th(c
m)
Inte
r-qu
artil
e-ra
nge
(cm
)
-7 -6 -5 -4 -3 -2 -1 0 1-3
-2
-1
0
1
2
19361937193819391940194119421943
1944
1945194619471948
1949
1950
1951
1952 1953
1954195519561957195819591960196119621963196419651966196719681969197019711972197319741975197619771978
NEA-PS (1956-1979)
axis1, 76%
axis
2, 1
0%
Year-class 1954-1974
Retrospective Analysis: Cohort
PCA…
Same strong year-classes in both fisheries
Include several continuous strong year-classes (here 1945 to 1952)
There is no sudden stop, but a kind of “ebb and flow”
-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1-2
-1
0
1
2
3
193619371938193919401941194219431944
1945
1946
1947
1948
1949
1950
19511952
1953
1954
19551956 1957195819591960196119621963196419651966196719681969197019711972197319741975197619771978
axis 1, 63%
axis
2, 2
1%
NEA-TP (1956-1979)
axis 1, 32%-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5-2
-1.5
-1
-0.5
0
0.5193619371938193919401941
19421943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
195519561957
1958
1959
1960
1961
19621963
19641965196619671968196919701971197219731974197519761977197819791980198119821983
MED-TP (1956-1979)
axis
2, 2
0%
axis
2, 1
3%Other fisheries display different patterns, but also some similarities
Also include several continuous key year-classes
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2-4
-2
0
2
4
6
193519361937193819391940194119421943194419451946194719481949
1950
1951
19521953
1954
1955
1956
1957195819591960
NWA-TP (1955-1961)
axis 1, 74%
PCA…
Retrospective Analysis: Cohort
clustering…
Retrospective Analysis: Cohort
NE
A-B
B
NW
A-P
S
ME
D-T
P
NE
A-T
P
NE
A-P
S
1966-1979
The strong link between Norway PS and Spain trap vanished during the 1970s
ME
D-T
P
NE
A-T
P
NE
A-LL
NE
A-P
S
NW
A-LL
NW
A-P
S
NE
A-B
B
1971-1981
Strong connection between Gibraltar
and the Northeast Atlantic fisheries
during the 1950s and 1960s
Possible connections between the
Mediterranean Sea / Northwest
Atlantic and the Northeast Atlantic
Discussion –Part 1
Key feeding ground in the Northeast Atlantic
Main migration routes of Atlantic bluefin
tuna were from the Mediterranean
spawning grounds and from the West
Atlantic coasts to the North Sea
In agreement with conventional tagging experiments from the 1950s and 1960s (Aloncle et al. 1974; Tiews 1978; Mather et al. 1995)
Bluefin tuna primarily fed on juveniles
herring (up to 400,000 tons/year, Tiews 1978)
BFT collapse
Norwegian spring-spawning herring (Torensen & Ostved 2000)The Nordic bluefin tuna fisheries collapse
might be due to the failure of the major
Northeast Atlantic herring stocks
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
-0.4
-0.2
0
0.2
0.4
-0.6
0.6
Cooling Warming Cooling Warming
SS
T a
nom
alie
s (°C
)
SST anomalies from the Northeast AtlanticA significant cooling rapidly took place in the
Northeast Atlantic in the early 1960s
Atlantic herring stocks displayed large
temporal and spatial variations being related
to climate changes (Hjort 1914; 1926; Alheit & Hagen
1997; Corten 2001…)
Discussion –Part 1
Norwegian spring-spawning herring (Torensen & Ostved 2000)
The rapid cooling of the Northeast Atlantic could
have affected herring recruitment that has
indeed dropped since the mid-1950s and more
obviously after 1962, so that this area could not
support the voracious appetite of bluefin tuna
BFT collapse
The Nordic bluefin tuna fisheries collapse
might be due to the failure of the major
Northeast Atlantic herring stocks
The persistence of a high fishing capacity on
herring has induced a huge increase in the
fishing mortality and the crash of these
populations (another example of synergistic
effects of climate and overfishing)
Discussion –Part 1
Conclusions –Part 1
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
-0.4
-0.2
0
0.2
0.4
-0.6
0.6
Cooling Warming Cooling Warming
SS
T a
nom
alie
s (°C
)
BUT the close match between the
cooling and the Nordic fisheries
collapse may indicate that
changes in temperature might
have also affected BFT behaviour
Nordic fisheries
Yie
lds
(t)
0
10000
20000
1920 1930 1940 1950 1960 1970 1980 1990 2000
The Nordic bluefin tuna fisheries collapse
might be due to the failure of the major
Northeast Atlantic herring stocks
Why the strong connection between the Nordic fisheries and the Spanish traps did not
vanish before the mid-1970s???
No changes in BFT condition before and after the crash (Tiews 1978)
No indication of recruitment failure in the French and Spanish bait boat fisheries
The Nordic fisheries collapse fisheries might result from changes in BFT migration
patterns in relation to the collapse of the herring stocks and the cooling of the
Northeast Atlantic during the early 1960s
BFT could have stayed in the Northwest
Atlantic, as the high catches offshore of
Brazil and along the US Northeast coasts
during the 1960s could advocate for it
1960-1969 1970-1979
1980-1989 1990-1999
Discussion –Part 1
Changes in migration patterns from decade-
to-decade is consistent with past & current
knowledge as well as past & present tagging
information (conventional and electronic)
Would there be a nowadays a new key feeding
ground in the Northwest Atlantic??
Discussion –Part 1
Strong differences in yields,
size and year-classes between
the Northeast Atlantic and
Mediterranean traps
This can hardly be explained by changes in migration solely and underlying
processes of collapse/decline of these two fisheries are likely to be different
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
50000
100000
Cat
ch (
nb)
NEA-TP
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000
50000
100000
Cat
ch (
nb)
MED-TPDifferences in catches all over
the 20th century
Discussion –Part 2
21
3
During the 1950-1960s, sub-
population (1) would have been
dominant and could have spatially
mixed with the 2 others
Atlantic BFT could be constituted by 3 sub-populations : (1) a highly migratory one over
all the North Atlantic and Mediterranean, (2) a resident one in the Mediterranean and (3) a
resident one in the West Atlantic
Discussion –Part 2
2
1
3
Since the 1980s, the sub-population (2) would be dominant while sub-
populations (1) and (3) would have
strongly declined
The declining catch in the North Atlantic and the increasing catch in the Mediterranean
since the early 1980s may be interpreted as a change in sub-population size in response
to environmental modifications and possibly overexploitation
Overexploitation in the Western
Mediterranean Sea would prevent therebuilding of sub-population(1) and
could explain why bluefin tuna did not
come back massively in theNortheast Atlantic since the 1990s
Discussion –Part 2
Conclusion
The history of BFT fisheries is likely to reflect interactions between these
biological, trophic, environmental and fishing processes
Bluefin tuna migration patterns have probably changed through time (in relation to
abiotic and trophic factors)
Fisheries are also dynamics in both time AND space (mostly increasing/expanding)
Bluefin tuna population structure that remains largely unknown is probably more
complex than currently thought
Abiotic factors (e.g. SST) that affect fish population dynamics and structure are
also varying through time and space
Thanks for your attention
Bluefin tuna fisheries have a tremendous history, but
what about their future???