MIGRATION PATTERNS OF YELLOWFIN, SKIPJACK AND BIGEYE …

SCRS/2019/184 Collect. Vol. Sci. Pap. ICCAT, 76(6): 903-950 (2020)

903

MIGRATION PATTERNS OF YELLOWFIN, SKIPJACK

AND BIGEYE TUNAS IN THE TROPICAL ATLANTIC,

BASED ON RECENT TAGGING AND RECAPTURE DATA

I. Arregui1, Nicolas Goñi1, M. Chifflet1

SUMMARY

The present work is part of the WP2 of the CISEF consortium (ICCAT/AOTTP mortality and

movement analysis) and aims at describing migration patterns of the three main tropical tuna

species across the Atlantic, based on AOTTP tagging and recapture data. We defined 15 areas

according to the tagging performed and recoveries that occurred in each of them. We focused on

the following elements: (1) mapping the minimum distance travelled by recovered fish, by species

and bi-monthly of tagging; (2) plotting the latitudinal and longitudinal distance travelled by

species, tagging area and bi-monthly of recovery, and (3) plotting the distribution of distances

travelled on each azimuth range by species, size category, tagging area and tagging bi-monthly.

We identify for the three species a general year-round counter-clockwise movement pattern in

the northeast tropical Atlantic, broad-scale movements of bigeye and skipjack tagged in Azores,

Madeira, and Canaries, and transoceanic migrations of bigeye and yellowfin in the equatorial

region. For the three species we observe greater distances travelled by small individuals than by

large ones. Finally, a preliminary tag-attrition model applied to these data is presented.

RÉSUMÉ

Le présent travail fait partie du WP2 du consortium CISEF (analyse de la mortalité et des

déplacements de l’AOTTP) et vise à décrire les schémas de migration des trois principales

espèces de thonidé tropical dans l'Atlantique, sur la base de données de marquage et de recapture

de l’AOTTP. Nous avons défini 15 zones en fonction du marquage effectué et des récupérations

qui ont eu lieu dans chacune d'entre elles. Nous nous sommes concentrés sur les éléments

suivants : (1) cartographier la distance minimale parcourue par les poissons récupérés, par

espèce et par période de deux mois de marquage ; (2) tracer la distance latitudinale et

longitudinale parcourue par espèce, zone de marquage et période de deux mois de récupération

et (3) tracer la distribution des distances parcourues sur chaque bande azimut par espèce,

catégorie de taille, zone de marquage et période de deux mois de marquage. Nous identifions

pour les trois espèces un schéma général de déplacement dans le sens inverse des aiguilles d'une

montre toute l'année dans l'Atlantique tropical Nord-Est, des déplacements à grande échelle de

thon obèse et de listao marqués aux Açores, à Madère et aux Canaries, et des migrations

transocéaniques de thon obèse et d'albacore dans la région équatoriale. Pour les trois espèces,

nous observons de plus grandes distances parcourues par les petits spécimens que par les grands.

Enfin, un modèle préliminaire de perte de marques appliqué à ces données est présenté.

RESUMEN

El presente trabajo forma parte del WP2 del consorcio CISEF (análisis de mortalidad y

movimiento del AOTTP-ICCAT) y tiene como objetivo describir los patrones de migración de las

tres principales especies de túnidos tropicales en el Atlántico basándose en los datos de marcado

y recaptura del AOTTP. Hemos definido 15 áreas de acuerdo con el marcado realizado y las

recuperaciones realizadas en cada una de ellas. Nos hemos centrado en los siguientes elementos:

(1) cartografiar la distancia mínima viajada de los peces recuperados, por especies y por un

periodo de dos meses de marcado; (2) dibujar la distancia latitudinal y longitudinal viajada por

especie, zona de marcado y por un periodo de dos meses de recuperación y (3) dibujar la

distribución de las distancias viajadas en cada rango azimut por especies, categoría de tallas,

zona de marcado y por periodo de dos meses de marcado. Hemos identificado para las tres

especies un patrón general de movimiento contrario a las agujas del reloj para todo el año en el

Atlántico nororiental tropical, movimientos a gran escala del patudo y el listado marcado en

Azores, Madeira y Canarias y migraciones transoceánicas de patudo y rabil en la región

1 AZTI – Marine Research Division, Herrera kaia portualdea 20110 Pasaia (Spain). Corresponding author [email protected] +34667174479

904

ecuatorial. Para las tres especies, hemos observado que los ejemplares pequeños han viajado

mayores distancias que los grandes. Por último, se presenta un modelo preliminar de pérdida de

marcas aplicado a estos datos.

KEYWORDS

Tagging, Migration, Bigeye, Skipjack, Yellowfin, Atlantic, AOTTP

1. Introduction

The overall objective of the Atlantic Ocean tropical tuna tagging program (ICCAT/AOTTP) is to contribute to

food security and economic growth of the coastal states of the Atlantic Ocean, ensuring sustainable management

of their tropical tuna resources. Specifically, AOTTP will use the information it collects on tuna stocks in the

Atlantic to provide improved and updated scientific advice to developing coastal states, and other Contracting

Parties, enabling them to adopt appropriate conservation and management measures for tunas in the framework of

the International Commission for the Conservation of Atlantic Tunas (ICCAT).

Tagging operations started in 2016, are still ongoing and led to more than 15 000 recoveries so far, with an

important coverage of the Atlantic regions from Azores to Uruguay and from South Arica to the US east coast.

Data analyses started in 2019, with a consortium (CISEF) in charge of mortality and movement analyses. The

present work is a preliminary analysis of tuna migrations in the framework of the 2nd working package of the

CISEF consortium.

As a preliminary analysis before modelling displacements through a tag-attrition model, we led an exploration of

movements patterns of tunas tagged by area (15 regions defined), species, size-category and two-month period.

2. Material and Methods

2.1. Database: structure and version used

The database used was provided by ICCAT/AOTTP on April 22nd 2019. It contains for each deployed tag the

information on species, size, release date and location, type of school, fishing gear; and similar information

regarding the recoveries.

2.2. Definition of areas

The areas were defined following a partially empirical approach, according to the tagging activity performed and

recoveries occurred in each of them (Table 1, Fig. 1 and 2). In particular, Senegal and Mauritania were merged in

a single area, where an important fishing effort occurs in the fourth bimonth of year (July-August). South from

Senegal and Cabo Verde, an important tagging effort occurred on the Sierra Leone Rise region, which we separated

from the Guinean shelf region. Within the Gulf of Guinea, we also created a separated region corresponding to the

coastal tagging done in Côte d’Ivoire

2.3. Variables described

Due to the magnitude of the database, an important amount of combinations can be explored: i.e. the respective

distributions of time at liberty, azimuth, longitudinal and latitudinal displacement of recovered fish, those being

grouped by species, size category, tagging region, tagging period (in quarter or bimonth i.e. pair of months, here

after referred to as S1 to S6) and recovery period.

We focused this first descriptive approach on the following elements:

- mapping the straight displacements (minimum distance travelled) of recovered fish, by species and

bimonth (or “sixth”) of tagging

- plotting the latitudinal and longitudinal distance travelled by species, tagging area and bimonth of

recovery

905

- plotting the distribution of distances travelled on each azimuth range by species, size category, tagging

area and tagging bimonth

Broad size categories were defined as follows: Fork length <45cm, 45 to 55 cm, >55cm, so all categories are

similarly represented.

Number days at liberty are divided in period categories (ndayClas) deined as: “short” between 15-150 days and

“long” >150 days

The fish recaptured prior to 15 days at sea are excluded in the descriptive analysis. In the case of the tag attrition

migration model the fish recaptured with less than 30 days at sea are considered unmixed and then excluded.

2.4. Movement rates estimation

A migration tag attrition model based on previous framework (Xiao, 1996; Arregui et al., 2006)was adapted for

tropical tuna, with inclusion of the migration between areas during the release year. The Eastern Atlantic area is

almost a closed system with very low evidences of trans-oceanic migrations. The SKJ migration rates between

main eastern areas and probability of capture (capturability*fishing effort) for each are estimated. The model is

age structured for tagged population dynamics, using the mean estimates of natural mortality rates from Gaertner

et al. (2015), while recoveries are age aggregated. The von Bertalanffy-Fabens growth parameters estimated by

Hallier and Gaertner (2006) for the eastern SKJ are used to aggregate the length in age classes and assign natural

mortality for each age below four years old. The age plus, considered as individuals older than four years old

(ICCAT, 2014) is considered aggregated.

Due to the heterogeneous distribution of the tagging effort (lower in the Wetsern Atlantic) on one hand, and based

on the movement patterns described in the descriptive sections, migration between main separable areas was

proposed: One involving a spatial continuity of releases and recoveries of the eastern central tropical Atlantic (i.e.

Eastern and Central Gulf of Guinea, Guinean shelf, Sierra Leone Rise, Cabo Verde, Senegal and Mauritania) and

northern island systems (Azores, Madeira, Canary Islands) were the area of the Azores archipelago is considered

independently form the areas surrounding Canary and Madeira archipelagos (Table 2). The annual migrations are

assumed to happen between the adjacent established areas (Fig. 41).

3. Results and Discussion

3.1. Overall displacements by species and period

For bigeye tuna, the tagging effort was heterogeneous during the year, with January through April and July through

October being the most tagged periods (Figures 3 to 8), and May-June the least tagged one (providing only 20

useful recoveries with mostly local migrations, except from Madeira).

The individuals tagged in the first bimonth were mostly released in the Central Gulf of Guinea and Sierra Leone

Rise regions. Their recoveries describe an anticlockwise pattern: they were mostly recovered south from equator

in S1, in the Gulf of Guinea in S2, in the Sierra Leone Rise in S3, in Mauritania in S4 and in Cabo Verde in S5.

A similar pattern was observed for bigeye tunas tagged in S2 (mostly in the Sierra Leone Rise), except regarding

the migration to the Gulf of Guinea.

Bigeye tunas tagged in S4 were mostly recovered off West Africa during S5, mostly in South Cabo Verde and

Senegal, and a significant number of individuals performed long distance migrations, from Madeira and Canary

islands to West Africa, from south Cabo Verde to the central Gulf of Guinea, and from North Brazil to central

Atlantic and Gulf of Guinea.

Bigeye tunas tagged in S5 in Canary islands, Cabo Verde and Sierra Leone Rise were mostly recovered in West

Africa during S1, in the Gulf of Guinea in S2, in the Sierra Leone Rise and Madeira in S3 (the ones recovered in

Madeira having been tagged in Canary islands) and Mauritania in S4, confirming the same year-round pattern of

movement

Finally, bigeye tunas tagged during S6 also followed a similar year-round pattern movement.

906

For skipjack tuna, the tagging effort was similarly heterogeneous as for bigeye (Fig. 9 to 14), with a particularly

low number of useful recoveries in S3.

Skipjack tunas tagged in S1 were mostly released in the Central Gulf of Guinea and Sierra Leone Rise regions.

They were mostly recovered in the Gulf of Guinea in S2 and S3, in the Sierra Leone Rise and Guinean shelf in S3,

in Mauritania in S4 and in Cabo Verde in S5, following a similar pattern as the one observed for bigeye tunas.

Skipjack tunas tagged in S2 were mostly released in the Sierra Leone Rise area, Ivory Coast, and Brazil (north,

central and south). Recoveries occurred in S3 in the Sierra Leone Rise and Guinean shelf, in S4 and S5 in

Mauritania and Cabo Verde, and near the equator in S6. Skipjack tagged in Brazil in S2 were mainly recovered

locally.

Release areas of skipjack tunas tagged in S4 and S5 included Azores and Canary islands, from which the longest

distances were registered: significant amounts of skipjack tunas tagged in Azores and Canary Islands in S4 were

recovered in S4 and S5 in Mauritania and Senegal, and in lesser amounts between Cabo Verde and the central Gulf

of Guinea; they were also present in S1 in the Guinean shelf area. Skipjack tunas tagged in S5 in Azores and

Canary Islands were not so densely recovered in Senegal and Mauritania, but migrated to Canary Islands and Sierra

Leone Rise (where they were present in S6) and Guinean shelf (where they were present in S1).

Skipjack tunas tagged in the same period but in the Sierra Leone Rise migrated to a wide array of directions,

although their major area of presence in S1 was the Guinean shelf.

Skipjack tagged in S6 were released mainly in the Guinean shelf region, and also migrated to all directions,

although to lower distances than the ones tagged in the Sierra Leone Rise in S5.

For yellowfin tuna, the tagging effort was similarly heterogeneous as for bigeye and skipjack (Fig. 15 to 20), but

the number of useful recoveries in S3 was higher than for the former two species.

Yellowfin tunas tagged in S1 were mostly released in the Central Gulf of Guinea and Sierra Leone Rise regions.

They were recovered in the Gulf of Guinea in S2, Sierra Leone Rise and Guinean shelf in S3, Mauritania in S4,

Cabo Verde in S4 and S5, and Equatorial area in S6. A significant amount of yellowfin tunas tagged in the Gulf

of Guinea in S1 displayed particularly long migrations to Brazil, where they were recovered in S3 to S6.

Yellowfin tunas tagged in S2 were mainly released in the Sierra Leone Rise (to a lesser extent in north Brazil),

and were recovered in this same region and the Guinean shelf in S3, Mauritania in S4, Cabo Verde in S5 and

central/western Gulf of Guinea in S6, thus following a similar pattern as bigeye and skipjack tunas.

Yellowfin tunas released during S3, although fewer, displayed important longitudinal displacements along the

equator.

Yellowfin tunas tagged in S4 were released mainly in Senegal and Mauritania, North Brazil, Eastern Gulf of

Guinea. They were recovered in important amounts in North Mauritania in S4, south Mauritania, Senegal and

Cabo Verde in S5, Cabo Verde and Guinea Bissau in S6, and along the 17º30W meridian between Dakar cape and

latitude 10 in S1, this last area being an important winter fishing ground for Dakar-based baitboats.

Yellowfin tunas tagged in S5 were released in Canary Islands, Sierra Leone Rise, Ivory Coat, Eastern Gulf of

Guinea, St Helena, North and center Brazil. Recoveries took place in the Sierra Leone Rise in S6 (particularly

around the Machucambo seamount), on the Guinean shelf in S6, S1, S2 and S3. Recoveries in S3 also occurred

around the Machucambo seamount, in Senegal and Mauritania. Yellowfin tunas tagged in Canary Islands during

S5 were recovered locally during S6 or in Mauritania in S4.

Finally, yellowfin tunas tagged in S6 were released mostly in the Sierra Leone Rise and Guinean shelf areas, Ivory

Coast, St Helena and northern Brazil. Movement patterns from the Guinean shelf were directed southeast, with

recoveries in western and central Gulf of Guinea and northwest, with recoveries between Guinea and Mauritania

in both cases in S1, S2 and S3.

Overall, we identify for the three species a seasonal movement pattern in the northeastern tropical Atlantic,

involving a presence in the Gulf of Guinea at S1 and S2, in the Sierra Leone Rise and Guinean shelf in S3, in

Mauritania in S4, Cabo Verde in S5, and back to the equatorial zone in S6. Bigeye and skipjack tunas tagged in

Madeira and Canary Islands displayed movements to Mauritania, Senegal, Cabo Verde and the Sierra Leone Rise,

and bigeye and skipjack tunas tagged in Azores displayed movements to all the aforementioned areas. Bigeye and

907

yellowfin tunas appear to migrate between northern Brazil and western Gulf of Guinea, but this pattern was not

displayed by skipjack tunas. Tunas tagged in coastal areas of Ivory Coast appear to be more resident than those of

other regions, with limited displacements throughout the year. Yellowfin tunas appear to move on the broadest

longitudinal range.

3.2. Latitudinal and longitudinal displacements by species and areas

Consistently with patterns observed on the aforementioned maps, skipjack tunas tagged in Azores displayed

important eastward and southward displacements, the orders of magnitude being 500nm eastwards and 1500nm

southwards (fig. 21), except for individuals recovered during S4 which is also the main fishing season in Azores.

For bigeye, the patterns are not as clear as for skipjack tunas, possibly due to a lower number of tagged individuals.

Bigeye tunas tagged in Madeira area display important southward migrations in late autumn and are recaptured in

areas an order of magnitude of 1000 nm south from Madeira in S6 and S1 (Fig. 22), then recoveries in S3 and S4

are closer to tagging locations. Skipjack tunas tagged in Madeira also display southwards migrations in S6,

although in a lesser extent compared to BET.

Bigeye and skipjack tunas tagged in North Canaries display important south-westwards displacements in winter

months (S6 to S2) and are present in the tagging area in S3 (Fig. 23). Fish tagged in south Canary Islands displayed

different patterns (Fig. 24), with more dispersion for bigeye tunas and a presence in more southern areas from S2

to S5 for skipjack tunas.

Among fish tagged in Mauritania-Senegal area, bigeye tunas appear to be relatively resident (Fig. 25), whereas

skipjack and yellowfin tunas are recovered further south (average circa 500 nm) in the first half of the year. In the

second half of the year they are recovered and in latitudes of a similar average as their tagging locations but with

an important southward dispersion.

Fish tagged in Cabo Verde do not display a clear seasonal pattern of movement (Fig. 26), but bigeye tunas display

an important dispersion, both in terms of longitude and latitude.

Fish tagged in the Sierra Leone Rise display a limited seasonal pattern of movement (a northwards displacement

by an order of magnitude of 250 to 500nm) but the most important dispersion of all regions, for all species and

months of recovery (Fig. 27).

Fish tagged in the Guinean shelf region display different patterns among species. Bigeye and yellowfin tunas are

recovered an order of magnitude of 500 nm north in S4, yellowfin tunas an order of magnitude of 500nm east and

south during S1, but no clear pattern for other months and for skipjack tunas (Fig. 28).

Bigeye tunas tagged in the central Gulf of Guinea are also recovered an order of magnitude of 500 nm north and

west in S4, skipjack tunas also move north to a similar extent in S4 as well as S5, whereas yellowfin tunas display

a different pattern of westward migration during S4 to S6 (Fig. 29), pattern also observed for yellowfin tunas

tagged in the Eastern Gulf of Guinea and recovered in S6 (Fig. 31), whereas fish tagged in Ivory Coast appear to

be residing in the area (Fig. 30), as the bigeye tunas tagged in São Pedro and São Paulo islands (Fig. 32).

The westwards movement in S6 of yellowfin tunas tagged in the Gulf of Guinea has a symmetric one in the

yellowfins tagged in north Brazil, which are spreading eastwards during S5 and S6 (Fig. 33). Skipjack tagged in

this area also tend to spread eastwards in the same period, and southwards during S3. Finally, fish tagged in central

and southern Brazil appear to be resident in these respective areas (Fig. 34 and 35).

Overall, these observations confirm and deepen the ones provided by the maps (section 3.1.) and show a division

between areas where important advective movements start (Azores, Madeira and Canary Islands) and areas with

more dispersal and a higher rate of residence (West African regions from Mauritania to the Eastern Gulf of

Guinea), displaying cyclical seasonal displacements of a lower magnitude as the ones from northern island systems

(Azores, Madeira and Canary Islands).

3.3. Distances travelled by azimuth

The distances travelled by the tagged tunas were the highest for skipjack tunas tagged in Azores, for all quarters

of recovery (Fig. 36). In comparison, skipjack tunas tagged in Madeira travelled lower distances (Fig. 37). Among

the fish tagged in the Sierra Leone Rise and Guinean shelf, skipjack tunas were also the ones traveling the longest

908

distances, mostly northwards in both cases. Bigeye and yellowfin tunas had similar distribution of travel directions,

with lower distances (Fig. 38 and 39).

Regarding the distances travelled as a function of tuna size, it appears that large individuals displayed the lowest

travelled distances, compared to small and medium-sized individuals (Fig. 40)

3.4. Tag-attrition model

The estimated migration rates given by the model for the SKJ in the Eastern Atlantic show very important

migration from the Eastern Tropical Atlantic to Madeira and Canary Islands (over 58%, Table 3), and important

from Azores area to the Madeira and Canary Islands (over 23%), but very little from Madeira and Canary Islands

to the other two locations. This case is a simple example of the multiple plausible options testable by means of the

general tag attrition model framework and the available data from the ICCAT/AOTTP program. The data

assimilation from previous tagging information, overviewed in Sculley (2016) would require the use of fishing

effort (due to important changes in successive years). Nevertheless the results from Sculley (2016) using a

Bayesian approach for SKJ were quite consistent with present estimates (Table 3). For example a the migration to

the area around Canary and Madeira, “CanMad” archipelagos from the rest of the areas estimated by Sculley

(2016) was quite high (5, 11 and 18%) and very low from “CanMad” to the rest (Azores is not considered in the

analysis).

909

References

Arregui, I., Arrizabalaga, H., and de la Serna, J. M. 2006. Preliminary approach to the experimental design of

tagging campaigns for movement rates estimation of East Atlantic bluefin tuna. ICCAT, Col. Vol. Sci.

Pap., 59: 769-788.

Gaertner, D. 2015. Indirect estimates of natural mortality rates for Atlantic skipjack (Katsuwonus pelamis), using

life history parameters. Collect. Vol. Sci. Pap. ICCAT, 71: 189-204.

Hallier, J. P., and Gaertner, D. 2006. Estimated growth rate of the skipjack tuna (Katsuwonus pelamis) from

tagging surveys conducted in the Senegalese area (1996-1999) within a meta-analysis framework. Collec.

Vol. Sci. Pap. ICCAT, 59: 411-420.

ICCAT 2014. Informe de la reunión de ICCAT de 2014 de evaluación de los stocks de listado del Atlántico este y

oeste. In Evaluación de los stocks de listado este y oeste, p. 38. Ed. by P. Pallarés. Dakar, Senegal.

Sculley, M. L. 2016. Estimating movement rates of Atlantic Ocean tropical tunas, Katsuwonus pelamis, Thunnus

albacares, and T. obesus, from tagging data. p. 192. University of Miami.

Xiao, Y. 1996. A framework for evaluating experimental designs for estimating rates of fish movement from tag

recoveries. Can. J. Fish. Aquat. Sci., 53: 1272-1280.

910

Table 1: spatial definition of the 15 regions used in the descriptive analysis

Area number Area name Area boundaries

1 Azores Lat: 35º to 45ºN

Lon: 35º to 20ºW

2 Brazil centre Lat: 25º to 15ºS

Lon: 45º to 35ºW

3 Brazil North Lat: 10ºS to 4ºN

Lon: 40º to 30ºW

4 Brazil South Lat: 40º to 25ºS

Lon: 55º to 45ºW

5 Cabo Verde 24ºN 30ºW; 9ºN 30ºW; 11ºN 19ºW; 17ºN 20ºW;

24ºN 23ºW

6 Canaries North Lat: islands north shore to 32ºN

Lon: 25º to 10ºW

7 Canaries South Lat: 24ºN to islands south shore

Lon: 25º to 10ºW

8 Central Gulf of

Guinea

Lat: 5ºS to 5ºN except Côte d’Ivoire area

Lon: 23ºW to 1ºE except Côte d’Ivoire area

9 Côte d’Ivoire Lat: 4ºN to 5ºN

Lon: 7º30 to 2ºW

10 Eastern Gulf of

Guinea

Lat: 5ºS to 5ºN

Lon: 1º to 12ºE

11 Guinean shelf 11ºN 19ºW; 9ºN 18ºW; 5ºN 18ºW; 5ºN 7º30W; 11ºN

13ºW

12 Madeira Lat: 32º to 35ºN

Lon: 25ºW to 10ºW

13 Senegal and

Mauritania

24ºN 23ºW; 17ºN 20ºW; 11ºN 19ºW; 11ºN 16ºW,

24ºN 16ºW

14 Sierra Leone Rise 9ºN 30ºW; 4ºN 30ºW; 4ºN 18ºW; 9ºN 18ºW; 11ºN

19ºW

15 São Pedro e São

Paulo islands

Lat: 4ºN to 10ºN

Lon:40º to 30ºW

Table 2: definition of the broad areas used in the tag-attrition model

Number Name AreaKey boundaries

1 Azores Azores and temperate Atlantic = 1 35º to 45ºN, 35ºW to European coast

2 CanMad Canary and Madeira islands = 2 24º to 35ºN, 25º to 10ºW

3 Etrop Eastern Tropical Atlantic = 3 20ºS to 20ºN, 30ºW to African coast

Table 3: model estimates

Migration Concerned areas rates (%)

"p(1->2)"; Azores -> CanMad 23.385

"p(2->1)"; CanMad -> Azores 2e-05

"p(2->3)"; CanMad -> Etrop 1.584

"p(3->2)"; Etrop -> CanMad 58.279

Probability of capture

"pc[1]" Azores 1.099e-01

"pc[2]" CanMad 4.276e-02

"pc[3]" Etrop 5.166e-01

911

Figure 1: releases of tagged tunas by area

Figure 2: recoveries of tagged tunas by area

912

Figure 3: minimum distances travelled by the bigeye tunas tagged in the months of January and February

913

Figure 4: minimum distances travelled by the bigeye tunas tagged in the months of March and April

914

Figure 5: minimum distances travelled by the bigeye tunas tagged in the months of May and June

915

Figure 6: minimum distances travelled by the bigeye tunas tagged in the months of July and August

916

Figure 7: minimum distances travelled by the bigeye tunas tagged in the months of September and October

917

Figure 8: minimum distances travelled by the bigeye tunas tagged in the months of November and December

918

Figure 9: minimum distances travelled by the skipjack tunas tagged in the months of January and February

919

Figure 10: minimum distances travelled by the skipjack tunas tagged in the months of March and April

920

Figure 11: minimum distances travelled by the skipjack tunas tagged in the months of May and June

921

Figure 12: minimum distances travelled by the skipjack tunas tagged in the months of July and August

922

Figure 13: minimum distances travelled by the skipjack tunas tagged in the months of September and October

923

Figure 14: minimum distances travelled by the skipjack tunas tagged in the months of November and December

924

Figure 15: minimum distances travelled by the yellowfin tunas tagged in the months of January and February

925

Figure 16: minimum distances travelled by the yellowfin tunas tagged in the months of March and April

926

Figure 17: minimum distances travelled by the yellowfin tunas tagged in the months of May and June

927

Figure 18: minimum distances travelled by the yellowfin tunas tagged in the months of July and August

928

Figure 19: minimum distances travelled by the yellowfin tunas tagged in the months of September and October

929

Figure 20: minimum distances travelled by the yellowfin tunas tagged in the months of November and December

930

Figure 21: longitudinal (upper) and latitudinal (lower) displacements of bigeye and skipjack tunas tagged in

Azores, by bimonth of recovery and type of time at liberty (short vs long)

931

Figure 22: longitudinal (upper) and latitudinal (lower) displacements of bigeye and skipjack tunas tagged in

Madeira, by bimonth of recovery and type of time at liberty (short vs long)

932

Figure 23: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in north Canaries, by bimonth of recovery and type of time at liberty (short vs long)

933

Figure 24: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in south Canaries, by bimonth of recovery

934

Figure 25: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in Mauritania and Senegal, by bimonth of recovery

935

Figure 26: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in Cabo Verde, by bimonth of recovery

936

Figure 27: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in Sierra Leone Rise, by bimonth of recovery

937

Figure 28: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in the Guinean shelf, by bimonth of recovery

938

Figure 29: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in the Central Gulf of Guinea, by bimonth of recovery

939

Figure 30: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in coastal areas of Ivory Coast, by bimonth of recovery

940

Figure 31: longitudinal (upper) and latitudinal (lower) displacements of skipjack and yellowfin tunas tagged in

the Eastern Gulf of Guinea, by bimonth of recovery

941

Figure 32: longitudinal (upper) and latitudinal (lower) displacements of bigeye tunas tagged in São Pedro e São

Paulo islands, by bimonth of recovery

942

Figure 33: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in North Brazil, by bimonth of recovery

943

Figure 34: longitudinal (upper) and latitudinal (lower) displacements of bigeye, skipjack and yellowfin tunas

tagged in central Brazil, by bimonth of recovery

944

Figure 35: longitudinal (upper) and latitudinal (lower) displacements of skipjack tunas tagged in south Brazil, by

bimonth of recovery

945

Figure 36: frequency distribution of the travelled distances of tunas tagged in Azores in each azimuth sector, by

quarter of recapture, size category and species

946

Figure 37: frequency distribution of the travelled distances of tunas tagged in Madeira in each azimuth sector,

by quarter of recapture, size category and species

947

Figure 38: frequency distribution of the travelled distances of tunas tagged in Sierra Leone Rise in each azimuth

sector, by quarter of recapture, size category and species

948

Figure 39: frequency distribution of the travelled distances of tunas tagged in the Guinean shelf in each azimuth

sector, by quarter of recapture, size category and species

949

Figure 40: frequency distribution of the travelled distances of tunas tagged (all species, quarters and areas merged)

in each azimuth sector, by size category

950

Figure 41: definition of broad areas for the application of a preliminar tag-attrition model to the tag and recovery

data.