The status of the North Atlantic swordfish stocks is based on the most recent analyses conducted in July 2017 by means of applying statistical modelling to the available data up to 2015. Complete information on the assessment can be found in the Report of the 2017 ICCAT Atlantic Swordfish Stock Assessment Session.

Swordfish (Xiphias gladius) are members of the family Xiphiidae and are in the suborder Scombroidei. They can reach a maximum weight in excess of 500 kg. They are distributed widely in the Atlantic Ocean and Mediterranean Sea. In the ICCAT Convention area, the management units of swordfish for assessment purposes are a separate Mediterranean group, and North and South Atlantic groups separated at 5°N. New genetic information was reviewed that indicated that the existing stock boundaries should be refined for the Atlantic and Mediterranean stocks. Swordfish feed on a wide variety of prey including groundfish, pelagic fish, deep-water fish, and invertebrates. They are believed to feed throughout the water column, and from electronic tagging studies, undertake extensive diel vertical migrations.

Swordfish mostly spawn in the western warm tropical and subtropical waters throughout the year, although seasonality has been reported in some of these areas. They are found in the colder temperate waters during summer and fall months. Young swordfish grow very rapidly, reaching about 140 cm LJFL (lower-jaw fork length) by age three, but grow slowly thereafter. Females grow faster than males and reach a larger maximum size. Tagging studies have shown that some swordfish can live up to 15 years. Swordfish are difficult to age, but about 50% of females were considered to be mature by age five, at a length of about 180 cm. However, the most recent information indicates a smaller length and age at maturity.

The analysis of the horizontal movements evidences seasonal patterns, with fish generally moving south by winter and returning to the temperate foraging grounds in spring. Broader areas of mixing between some eastern and western areas were also suggested. These new results obtained by pop-up satellite tags also fully confirm the previous knowledge that was available from fishery data: deep longline settings catch swordfish during the day-time as a by-catch, while shallow setting longliners target swordfish at night closer to the surface.

More information on swordfish biology and ecology is published in the ICCAT Manual.

Swordfish is a temperate tuna widely distributed throughout the Atlantic Ocean and Mediterranean Sea, in coastal and offshore areas, mostly ranging from 50°N to 45°S.

On the basis of the biological information available for assessment purposes, the existence of three stocks is assumed: northern and southern Atlantic stocks (separated at 5ºN) and Mediterranean stock (Figure 1).

Figure 1: Geographic distribution of swordfish cumulative catch (t) by gear, in the Convention area, shown on a decadal scale. The maps are scaled to the maximum catch observed during 1960-2016 (the last decade only covers 7 years).

Description of fisheries

North Atlantic

Due to the broad geographical distribution of Atlantic swordfish (Figure 1) in coastal and off-shore areas (mostly ranging from 50ºN to 45ºS), this species is available to a large number of fishing countries. Directed longline fisheries from Canada, EU-Spain, and the United States have operated since the late 1950s or early 1960s, and harpoon fisheries have existed at least since the late 1800s. Other directed swordfish fisheries include fleets from Brazil, Morocco, Namibia, EU-Portugal, South Africa, Uruguay, and Venezuela. The primary by-catch or opportunistic fisheries that take swordfish are tuna fleets from Chinese Taipei, Japan, Korea and EU-France. The tuna longline fishery started in 1956 and has operated throughout the Atlantic since then, with substantial catches of swordfish that are produced as a by-catch of tuna fisheries. The largest proportion of the Atlantic catches is made using surface-drifting longline. However, many additional gears are used, including traditional gillnets off the coast of western Africa. The trends in mean fish weight taken in the North Atlantic fisheries are shown in Figure 3.

For the past decade, the North Atlantic estimated catch (landings plus dead discards) has averaged about 12,000 t per year (Table 1 and Figure 2). The catch in 2017 (10,046 t) represents a 50.4% decrease since the 1987 peak in North Atlantic landings (20,238 t). These reduced landings have been attributed to ICCAT regulatory recommendations and shifts in fleet distributions, including the movement of some vessels in certain years to the South Atlantic or out of the Atlantic. In addition, some fleets, including at least the United States, EU-Spain and EU-Portugal have changed operating procedures to opportunistically target tuna and/or sharks, taking advantage of market conditions and higher relative catch rates of these species previously considered as by-catch in some fleets. Recently, socio-economic factors may have also contributed to the decline in catch.

Trends by area (NE vs. NW Atlantic) in the CPUE indexes were consistent with the seasonal movement patterns observed in the electronic tagging data, as well as in the catches and sex-ratio distributions. Relationships observed for the eastern Atlantic were opposite to those in the western Atlantic. This pattern was correlated with the decadal cycling of the AMO as well as that of the North Atlantic Oscillation (NAO). Including the AMO as a covariate to area specific catchability within the assessment model helped reduce the conflicting directions of the various CPUE trends. Further analysis and hypothesis testing were recommended to determine if the relationship was due to a swordfish temperature preference, a change in prey distribution, or perhaps both. For the North Atlantic stock some of the indices of abundance were affected by changes in gear technology and management that could not be accounted for in the CPUE standardization, and therefore had to be split.

Available catch per unit effort (CPUE) series were evaluated by the Committee and certain indices were identified as suitable for use in the assessment models (Canada, EU-Portugal, EU-Spain, Japan, Morocco, and USA). Trends in standardized CPUE series by fleets contributing to the stock assessment models are shown in Figure 4. Most of the series have an increasing trend since the late 1990s, but show a decrease in the more recent years. There have been some recent changes in United States regulations that may have impacted catch rates. The combined index used as the continuity model from the previous assessment is shown in Figure 5.

Figure 2: Total swordfish catches reported to ICCAT (Task I) for the northern Atlantic stock including TAC.

Figure 3: Trends in mean weight (kg) for the North and South Atlantic swordfish stocks.

Effects of current regulations

In 2017, the Committee provided information on the effectiveness of existing minimum size regulations. New catch regulations were implemented on the basis of Rec. 06-02, which entered into effect in 2007 (Rec. 08-02 extended the provisions of Rec. 06-02 to include 2009). Rec. 09-02 came into effect in 2010 and extended most of the provisions of Rec. 06-02 for one year only. Rec. 10-02 came into effect in 2011, and again extended those provisions for one year only, but with a slight reduction in total allowable catch (TAC). For the North Atlantic swordfish the most recent recommendations can be found in Rec. 17-02.

The total allowable catch in the North Atlantic during the 2007 to 2009 period was 14,000 t per year. The reported catch during that period averaged 11,811 t and did not exceed the TAC in any year. In 2010, the TAC was reduced to 13,700 t. The reported catch since then averaged 11,472 t and exceeded the TAC in one year (2012, 13,868 t).

There are two minimum size options that are applied to the entire Atlantic: 125 cm LJFL with a 15% tolerance, or 119 cm LJFL with zero tolerance and evaluation of the discards. Since the implementation of the minimum landing sizes in 2000, the estimate of percentage of swordfish less than 125 cm LJFL reported landed (in number) has been generally decreasing in the North Atlantic, the estimate was 33% in 2000 and decreased to 23% in 2015. The Committee notes that these estimations have high levels of substitutions for a significant portion of the total catch and are highly unreliable and biased unless CPCs fully report size samples from the entire catch.

The Committee also noted high values of hooking mortality (ranging between 78-88%) on small swordfish (<125 cm LJFL) on some surface longline fisheries targeting swordfish, with the post-release mortality of specimens discarded alive unknown. Recommend and evaluating other strategies to protect juvenile swordfish will need completeness of datasets on fishing effort and size data over the entire Atlantic and should take into account the effects on other species. In view of the Commission objective to protect small swordfish, the Committee therefore recommends that future work should be carried out to determine more precisely the spatial distribution and magnitude of fishing effort, size and sex distribution of undersized swordfish in the Atlantic, using high resolution observer data.