Page 2 - 10.1.1.498.8230

1300 C. Ravier and J.-M. Fromentin

50 32 to 38 49
30 39 to 45 46 47
51
52 31
53
13°W 54
49°N

b 48
c

40°N Spain Italy
Portugal
b

a Tunisia c 25 26
d
Morocco

30°N 27
30°E 28
29

1 to 12 13 to 21

23 d
24 3000 tunas/year

a 22

Figure 1. Map of the western Mediterranean and adjacent North Atlantic. Arrows indicate spawning migration routes of eastern
Atlantic bluefin tuna and stippled areas its two known main spawning sites, around the Balearic Islands and in the South
Tyrrhenian Sea. Insets show the location of traps around (a) Portugal, Spain, and Morocco, (b) Sardinia, (c) Sicily and the South
coast of Italy, and (d) Tunisia. Numbers refer to trap names given in Table 1. The diameter of the circle is proportional to the
median production of the trap.

large, pelagic, migratory fish that are known to migrate fluctuations in trap catches in different areas. At the
for spawning purposes in the same manner every year. beginning of the century, Sella (1929) mentioned poss-
In May, they enter the Mediterranean from the North ible synchronous fluctuations in catches of Sardinian,
Atlantic and head for two main spawning areas, north of Sicilian, and Tunisian traps. Adding Portuguese traps,
Sicily and around the Balearic islands (Figure 1; Mather Fromentin et al. (2000a) analysed five long time-series
et al., 1995). For centuries, fishermen have used their by correlation analysis, and supported the hypothesis of
knowledge of these spawning migrations to set traps synchrony in long-term fluctuations.
along the routes. Fixed and passive gear hardly changed
until the middle of the 20th century (Thomazi, 1947; Those studies, however, were confined to a restricted
Doumenge, 1998), suggesting that fishing effort may data set. To validate the hypothesis of synchrony, it was
have remained unchanged over centuries. Therefore, a necessary to look for additional time-series, and then to
trap can be considered to be a sampler that catches the carry out more extensive analyses. A data set of 54
same proportion of the migrating bluefin tuna popu- time-series from the western Mediterranean and the
lation each year. To investigate whether fluctuations in Atlantic coasts of Portugal, Spain, and Morocco is
catches could be considered as a proxy for true abun- therefore presented and investigated here. Using time-
dance, we tested whether long-term fluctuations in trap series analyses, we decomposed the temporal signal and
catches varied simultaneously around the western identified trends and cycles in catch, and we further
Mediterranean and along the Atlantic coasts of Spain, tested whether the fluctuations were synchronous. The
Morocco, and Portugal. Two studies had already aim was to answer the following questions. (1) What are
documented possible synchrony between long-term the patterns and the magnitude of long- and short-term
fluctuations in catches? (2) Is there evidence of periodic