nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
Billfish are large oceanic and epipelagic organisms that can be found in all the tropical and temperate seas of the world. The 2 families comprising all billfish species are Istiophoridae (marlins, sailfishes, and spearfishes) and Xiphiidae (a single- species family comprising the swordfish, Xiphias gladius).
The distribution of billfish and other tropical pelagic fish is affected by the seasonal and interannual variations of sea surface temperature (SST). For example, Norton (1999) found that the preferred habitat of the dolphinfish (Coryphaena hippurus) expands northwards in the eastern Pacific Ocean during extreme interannual events like the 1997-1998 El Niño. Farrell et al. (2014) found that SST was one of the variables that most influenced dolphinfish catch in the western Atlantic. Su et al. (2008) suggested that blue marlin (Makaira nigricans) prefer waters with SST values between 27 and 30 ºC. Martinez-Rincon et al. (2015) reported that sailfish (Istiophorus platypterus) catch was higher when SST values were >26 ºC. Thermal preferences are well known for some billfish species but they have been less explored for spearfish (Boyce et al. 2008).
Although considered one of the largest predators in the tropical food web (Lehodey 2004), the shortbill spearfish (hereafter spearfish; Tetrapturus angustirostris Tanaka, 1915) is one of the smallest members of the Istiophoridae family (on average ~140 cm in fork-eye length and ~18 kg in weight; Nakamura 1985). This species inhabits mainly the oceanic tropical and temperate waters of the Pacific and Indian oceans, with rare sightings in the Atlantic Ocean (Nakamura 1985). These fish are incidentally caught by the tuna purse seine and longline fleets and very occasionally by the surface-trolling fleet (Nakamura 1985). Spearfish are believed to spawn during the winter months in offshore waters with SST close to 25 ºC (Nakamura 1985). Their feeding habits are different around the world. In the eastern Pacific Ocean, for example, they feed on a variety of cephalopods and fish (Gempylidae, Scombridae, Exocotidae, Stromatidae; Nakamura 1985), and in the western Indian Ocean this species is thought to prey on fish mainly in the mixed layer (Ménard et al. 2013).
After Nakamura’s (1985) seminal work on billfish, only few papers about the spearfish have been published, most focusing on topics unrelated to fisheries, such as food poisoning issues (e.g., Kaneko and Ralston 2007, Chen et al. 2010). According to phylogenetic analysis, Collette et al. (2006) suggested that spearfishes should be classified in their own genera (Tetrapturus), apart from all marlin species. Amandè et al. (2010) found that spearfish accounted for only 0.3% of the total bycatch made by the tuna purse seine fleet in the Atlantic Ocean from 2003 to 2007. Polovina et al. (2009) suggested that the ecosystem shift to increased relative abundance of mid-trophic, fast-growing predators was likely the result of a decrease in the catch rates of apex predators (including spearfish) in the subtropical North Pacific. Using data from longline surveys in the Northeast Pacific Ocean, Shimose et al. (2010) found that spearfish have specific habitat preferences; all spearfish catches occurred in the open ocean, between ~15ºN and 20ºN, where the species fed mainly on fish from the Molidae family.
To the best of our knowledge, no quantitative analysis including spearfish incidental catch, spatial distribution, temporal trends, relation with SST/climatic indexes, and affinity for certain fishing conditions (set types) in the Pacific Ocean has been published. Additionally, the International Union for the Conservation of Nature (IUCN) has listed this species in the red list of threatened species as “data deficient”. Regarding spearfish, the IUCN suggests that “given that it is taken with the same gear as the Blue Marlin and Stripped Marlin, […] the population of this species is declining as well, but there’s no data to quantify this” (Collette et al. 2011). Therefore, the main objective of this paper is to provide basic information on spearfish spatial distribution using a long-term database.
Materials and methods
Fishery data base
A 23-year (1993-2015) database of spearfish incidental catch was analyzed for this study. The data from this database was gathered by on-board scientific observers from the Inter-American Tropical Tuna Commission (IATTC). The IATTC requires that one observer be aboard every class-6 tuna purse seine vessel (>435 m3 storage capacity) in every fishing trip (IATTC 2009). The database included total number of spearfish, total number of purse seine sets, type of set (fishing indicator: dolphin, unassociated, and floating object), and geographical coordinates (longitude, latitude), all aggregated in monthly 1º × 1º quadrants.
Numerical procedures
Incidental catch was standardized using the total number of sets as fishing effort in the following equation:
where ICPUEi is the incidental catch per unit effort in number of fish per set, SSPi is the total number of caught spearfish, and NSi is the total number of purse seine sets, all for ith quadrant.
In order to evaluate the seasonal variation of the spatial distribution of ICPUE, a quarterly approach was used. In this approach, months from January to March were considered the winter season, from April to June spring, from July to September summer, and from October to December autumn. Because the incidental catch vector was notoriously skewed (non-normal appearance), a Kruskal-Wallis test was applied to assess any possible seasonal variation. Quarterly maps were created by aggregating the number of fish per set per quadrant (latitude, longitude) for each of the 4 seasons.
To assess the possible temporal trends in ICPUE, a time series was built by calculating the mean for each month in the 23 years. To enhance visual interpretation of the incidental catch time series, raw data was smoothed using both the smooth and the smooth.spline functions in R’s base stats package (R Core Team 2016) and the default arguments. While smooth uses Tukey’s running median method, smooth.spline uses a more complex method, a cubic spline smoother. Smooth produced a curve with blunt ridges and smooth.spline produced a rather flat curve, so we decided to use a combination of booth smoothing techniques. The possible trends in the incidental catch time series were explored using a simple linear regression analysis, where the hypothesis that parameter b of the fitted regression model equals zero was tested (positive and different from zero b values indicating a positive trend, and vice versa; Polovina et al. 2009). The linear regression model was applied using R’s base stats library (R Core Team 2016). For all statistical tests, we considered a confidence level of 95%. Annual percent change rate in spearfish ICPUE was calculated by dividing the slope by the intercept of the fitted linear model and by multiplying the result by 12 to convert from monthly rate to yearly rate and then by 100 to convert to percentage (Polovina et al. 2009). To evaluate the possible effect of SST variations on spearfish catch, we explored the relation between the incidental catch time series and the SST monthly mean values extracted from the Niño 1+2, 3, and 3.4 regions, the Oceanic Niño Index (ONI), and the Pacific Decadal Oscillation (PDO) index by performing a cross-correlation analysis using the ccf function of the R stats package (R Core Team 2016). We decided not to include SST values extracted from the Niño 4 region, because the portion of this area that is not included in the 3.4 region falls outside our study area. The same approach that was applied to the incidental catch data was applied here to detect possible trends in the environmental time-series data. Information on the SST in the Niño regions and ONI can be obtained from the National Centers for Environmental Information website (https://www.ncdc.noaa.gov/teleconnections/enso/indicators/sst.php) of the National Oceanic and Atmospheric Administration (USA). A brief description of the PDO and corresponding data can be obtained from the same website at https://www.ncdc.noaa.gov/teleconnections/pdo/.
Results
General picture
Over the 23 years, a total of 422,970 purse seine sets were carried out in the study area and 4,333 sets were carried out in cells where at least one spearfish was caught. A total of 687 individuals were caught, and 292.95 fish were caught per set (mean = 0.002 fish per set; SD = 0.05). Maximum values of 4 fish per set were found for February 1995, June 2002, July 2013, and October 2015. There were small seasonal differences in the amount of fish caught during summer, spring, autumn, and winter, with 29%, 28%, 24%, and 19% of total fish per set (85.26, 81.76, 71.36, and 54.56 fish per set), respectively; these differences were not significant (K(3,136335) = 7.42; P = 0.06).
Spatially, most of the incidental catch occurred in a zone near the equator, from 10ºS to 10ºN, with some quadrants displaying “high” catch in the Southern Hemisphere during the summer months. One maximum (4 fish per set) occurred in each of the 4 seasons: one in the Central American coast during winter, one in the open ocean (~3ºN, 130ºW) during spring, one at the same latitude but closer to the coast (approximately 102ºW) during summer, and one near the South American coast during autumn (Fig. 1).
Temporal trend
The smoothed time series showed 4 peaks of incidental catch during 1996, 2007, 2013, and 2015, with one peak per year (Fig. 2). The linear regression model fit showed a positive trend in the incidental catch time series (a = 1.39 × 10-3, P < 0.05; b = 4.78 × 10-6, P < 0.05). The incidental catch of spearfish increased at a rate of ~4.15% per year. Cross- correlation analyses showed an inverse relation between spearfish catch and environmental data; however, only the correlation with SST from the Niño 3 and 3.4 regions and the PDO index, found with lags of 11, 9, and 2 months, was significant (Table 1). The PDO was the environmental variable that showed the highest absolute correlation value with ICPUE (Table 1), and the only variable to show a significant (negative) temporal trend (Fig. 3, Table 2).
Table 1 Summary of the cross-correlation between the incidental catch of spearfish per unit effort and environmental data. ONI, Oceanic Niño Index; PDO, Pacific Decadal Oscillation.
| Correlation (ρ) | Lag (months) | P (ρ = 0) | |
| Niño 1+2 | -0.113 | 0 | 0.059 |
| Niño 3 | -0.125 | 11 | <0.050 |
| Niño 3.4 | -0.135 | 9 | <0.050 |
| ONI | -0.109 | 9 | 0.065 |
| PDO | -0.156 | 2 | <0.050 |
Figure 3 Smoothed time series of the incidental catch of spearfish and Pacific Decadal Oscillation (PDO) time series, with corresponding trend lines.
Incidental catch per type of set
Of the total 292.95 fish per set caught during our study period, 70.98% (207.95 fish per set) were caught when the fleet fished on floating objects, 19.17% (56.18 fish per set) were caught on dolphin sets, and 9.8% (28.81 fish per set) were caught when the fleet targeted free-swimming tuna schools.
The most important mode for incidental catches in sets made on floating objects was found south of the equator, between 5ºS and 10ºS. On the other hand, the most notorious mode for dolphin sets was found north of the equator, between 5ºN and 10ºN. No notorious latitudinal mode was found for incidental catches in sets made on free-swimming tuna schools (Fig. 4a).
Figure 4 Spatial distribution of total incidental catch of spearfish per unit effort, per type of set (on dolphin, unassociated, and on floating objects) for latitude (a) and longitude (b) during the study period.
Incidental catches of spearfish in sets made on floating objects was higher west of 100ºW; the most important mode for dolphin sets was found between 85ºW and 90ºW longitude, although there was a similar mode between 120ºW and 125ºW longitude. Most incidental catches in sets on free-swimming tuna schools occurred to east of 100ºW (Fig. 4b).
Discussion
Only ~1% of total purse seine sets resulted in the total incidental catch of spearfish, mostly in the open ocean, highlighting the rareness and oceanic behavior of this species. Shimose et al. (2010) suggested that spearfish occur mainly in areas far from the equator. However, our evidence shows that most spearfish were caught in an area near the equator, from 10ºS to 10ºN. As Shimose et al. (2010) noted, their results may be true for a specific time of the year (late summer, early fall). In our paper we present a bigger spatiotemporal picture, which helps better understand the spatial distribution of spearfish. In addition to the possible effects of seasonal migration, the fact that the data reported by Shimose et al. (2010) were collected during longline operations could indicate a source of bias in the spearfish catch estimate because longlines can reach up to 250 m depth (Bigelow et al. 2006), and spearfish are believed to inhabit shallower waters than other billfish (Nakamura 1985); this bias is expected to be lower when using data from purse seine sets, which use surface fishing gear.
The equatorial zone is a highly productive area because of the upwelling events that are generated by the trade winds (Martínez-Rincón et al. 2009); this upwelling process is manifested by a cold water tongue that extends from an area near the Central American coast to the International Date Line (Wyrtki 1981, p. 1206). In the Northern Hemisphere, there is a zone where temperature changes very little and the strong thermoclines occurring outside this cold water tongue, at around 5ºN-10ºN, favor the aggregation of some marine mammals and fish (Fiedler et al. 1992, Martínez-Rincón et al. 2009). The area around the cold water tongue (5ºN-10ºN) is an important feeding zone for large pelagic fish such as sharks, tunas, and billfish that feed on crustaceans, fish, and small squid, as a result of the upwelling events that occur in the convergence zone between the South and North Equatorial Currents (Mann and Lazier 1996, Galván-Magaña 1999, Bocanegra-Castillo 2007).
According to Shimose et al. (2010), spearfish feed mainly on fish from the Molidae family, and “[tend] to avoid the equatorial zone, where preferable prey may not be available.” Measurement and count data for members of the Molidae family is scarce (Matsuura 2015). Some Molidae, such as Mola mola, occur seasonally in Pacific Ocean waters off Southern California (USA), suggesting migratory activity (Cartamil and Lowe 2004), presumably in response to a preferred temperature range (Lee 1986). In spite of the possible migratory activity of spearfish prey, spearfish might opportunistically feed on available prey, since Nakamura (1985) noted that spearfish stomach contents vary spatially and seasonally. Seasonal sampling of stomach contents for this species may provide better insights on the factors that determine the spatial distribution of prey.
Polovina et al. (2009) found a decrease in the spearfish catch reported by the Hawaiian longline fishery from 1996 to 2006. Population decrease is assumed by the IUCN (Collette et al. 2011). Our data suggest that, at least in the eastern Pacific Ocean, the spearfish population is on the rise at a rate of ~4% per year, as evidenced by the positive slope in the linear regression analysis.
The incidental catch of spearfish was significantly correlated with SST extracted from the Niño 3 and 3.4 regions and PDO. However, the significance of correlation tests applied to data sets with a high number of observations should be addressed with care because the value of the coefficient at which the correlation becomes statistically significant is inversely related to the number of observations, so statistically significant correlations may still depict a weak correlation between the 2 variables. In contrast with the higher and more significant correlations found with SST extracted from the Niño 3 and 3.4 regions, correlation between ICPUE and SST from the Niño 1+2 region (the closest one to the coastline) was lower and nonsignificant, which weakly suggests that spearfish are affected mostly by SST variations that are due to processes that occur in the open ocean.
We found negative correlations between ICPUE and SST from the Niño regions, which suggests that spearfish prefer cool waters. However, higher correlations between these variables were found with 9-11 month lags; that is, higher incidental catches occurred ~10 months after the SST minimum occurred in the Niño regions. The SST minimum in the Niño 3.4 region occurred within the first 2 months of the year, so incidental catch is expected to be higher during winter, when SST starts to decrease towards the end of the year. A plausible explanation is that spearfish arrive in equatorial waters to spawn, since this species is believed to spawn in waters with relatively low temperatures (~25ºC, Nakamura 1985). Another plausible explanation is that spearfish feed on prey that are more abundant during that time of the year. However, we can only speculate on the mechanisms that are responsible for the time lag found between ICPUE and SST because literature concerning the basic biology of spearfish and/or its main prey is notably scarce, and these hypotheses would need to be further investigated.
The PDO is a temporal pattern that affects Pacific Ocean climate in time scales of decades, and it can be viewed as a “long-lived” El Niño climate phenomenon (Mantua and Hare 2002), since the cold and warm phases of both El Niño and the PDO show similar behavior (Mantua et al. 1997). The correlation between ICPUE and PDO was the highest and most direct (2 month lag) of all environmental correlations; moreover, ICPUE and PDO were the only 2 variables that showed a significant trend over time.
With this evidence, we suggest that spearfish inhabit zones close to the equator (e.g., outside the defined Niño zones), where upwelled water is cooler than equatorial waters. Also, spearfish appear to be more affected by environmental processes that occur at larger temporal scales than El Niño, such as the PDO. Since we used a standardized relative abundance index (ICPUE) that is not affected by possible temporal changes in fishing effort, we also suggest that the negative trend in PDO was an important factor causing the positive trend in the incidental catch of spearfish because, given the time scales at which the PDO affects climate in the Pacific Ocean, these cooler, preferred waters are expected to remain a longer time in the zone near the equator.
Most spearfish catches were associated with sets on floating objects. The affinity that certain fish species have for floating objects is not fully understood. It is believed that some fish use floating objects to trace high-productivity water masses, since the origin of these objects is usually a high-productivity zone, such as a river mouth (Hall 1992). Some authors suggest that floating objects tend to aggregate around certain oceanographic structures, such as fronts and gyres, where billfish prey also congregate (Sakagawa 1989, cited in Sosa-Nishizaki 1998). Solana-Sansores (2001) showed that the equatorial current system can retain floating objects from different origins in a latitudinal band from 15ºS to 10ºN and noted that that purse seine sets on floating objects are becoming more common. This could explain the high incidental catch of spearfish on this type of set, which occurs mainly west of 100ºW, depicting again this species’ preference for oceanic waters (Shimose et al. 2010).
In summary, only ~1% of purse seine sets resulted in total incidental catch of spearfish. Most spearfish were caught near the equator and in the open ocean on sets associated mainly with floating objects. We found a positive trend in the relative abundance of spearfish, a contrasting result with what has been reported for this species in oceanic waters near Hawaii. Spearfish appear to be more affected by interdecadal climatic events, such as the PDO, than by events of interannual frequency, such as El Niño. Interdecadal events are likely playing an important role in the positive trend in spearfish catches.
