INTRODUCTION
Thresher sharks (Family Alopiidae) comprise three highly migratory pelagic species (Alopias vulpinus, A. pelagicus and A. superciliosus), all of which occur in the Pacific Ocean, including Peruvian waters (Smith et al., Reference Smith, Rasmussen, Ramon, Cailliet, Camhi, Pikitch and Babcock2008; Cornejo et al., Reference Cornejo, Velez-Zuazo, González-Pestana, Kouri and Mucientes2015). Two evolutionarily significant units of A. pelagicus have been identified in the Pacific Ocean, with eastern and western Pacific populations (Cardeñosa et al., Reference Cardeñosa, Hyde and Caballero2014). Members of this family exhibit low intrinsic reproductive rates, and are susceptible to over-exploitation (Smith et al., Reference Smith, Rasmussen, Ramon, Cailliet, Camhi, Pikitch and Babcock2008; Oldfield et al., Reference Oldfield, Outhwaite, Goodman and Sant2012; Dulvy et al., Reference Dulvy, Fowler, Musick, Cavanagh, Kyne, Harrison, Carlson, Davidson and Sonja2014). On the west coast of the USA, the population of A. vulpinus showed a marked reduction in size and decreases in landings after less than a decade of commercial exploitation during the late 1970s and early 1980s (Holts, Reference Holts1988; Hanan et al., Reference Hanan, Holts and Coan1993). In Peru, national landing records suggest that A. vulpinus is the fourth most captured shark species, where it is landed mainly in the north of the country (Gonzalez-Pestana et al., Reference Gonzalez-Pestana, Kouri and Velez-Zuazo2016). However, it is likely that species misidentification occurred between A. vulpinus and A. pelagicus (Velez-Zuazo et al., Reference Velez-Zuazo, Alfaro-Shigueto, Mangel, Papa and Agnarsson2015).
Peruvian shark fishery monitoring and management are at a relatively early stage, and no specific management plans have been implemented for the Alopiidae family. Ecosystem-based fisheries management is more holistic than single-species approaches and considers indirect effects on food webs. This approach requires an ecological understanding of exploited species and their communities. Yet, a lack of ecological data and data on how fishing alters ecosystems function are common limitations that have hampered the implementation of ecosystem-based approaches (Essington & Punt, Reference Essington and Punt2011). Peru has adopted an ecosystem-based approach for management of the Large Marine Ecosystem of the Humboldt Current but this does not include any specific measures related to sharks. Quantitative diet composition estimates provide a basis for understanding a species’ prey spectrum and its overall trophic ecology (Bizzarro et al., Reference Bizzarro, Carlisle, Smith and Cortés2017). Thus, determining the diet composition of thresher sharks is important for developing ecosystem models and management.
The diets of thresher sharks have been studied in the eastern Pacific Ocean (Ecuador, Mexico and California) and Australia (Preti et al., Reference Preti, Smith and Ramon2001, Reference Preti, Smith and Ramon2004; Polo-Silva, Reference Polo-Silva2004; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009, Reference Polo-Silva, Galván-Magaña, Newsome, Grijalba-Bendeck and Sanjuan-Muñoz2013; Rogers et al., Reference Rogers, Huveneers, Page, Hamer, Goldsworthy, Mitchell and Seuront2012; Galvan et al., Reference Galvan-Magaña, Polo-Silva, Berenice Hernández-Aguilar, Sandoval-Londoño, Ruth Ochoa-Díaz, Aguilar-Castro, Castañeda-Suárez, Cabrera Chavez-Costa, Baigorrí-Santacruz, Eden Torres-Rojas and Andrés Abitia-Cárdenas2013; Rosas-Luis et al., Reference Rosas-Luis, Loor-Andrade, Carrera-Fernández, Pincay-Espinoza, Vinces-Ortega and Chompoy-Salazar2015), and these studies have shown squids and teleosts to be the main prey. Squid have important roles in marine food webs, both as predators and as prey (Coll et al., Reference Coll, Navarro, Olson and Christensen2013). Because of their life history traits (e.g. fast growth) they can have large trophic impacts on food webs (Ehrhardt, Reference Ehrhardt1991; Coll et al., Reference Coll, Navarro, Olson and Christensen2013). They are an important component of the eastern Pacific pelagic ecosystem because of their abundance (Olson & Waters, Reference Olson and Watters2003). Thus, their predators might play an important role in squid population dynamics (Coll et al., Reference Coll, Navarro, Olson and Christensen2013). As a result, it is important to understand the predator–prey interactions between sharks and squids.
The biology and ecology of thresher sharks in Peruvian waters is poorly understood, with no quantitative information of their diet currently available. This study sought to understand the feeding ecology of Alopias spp. in northern Peru by analysing stomach contents, including variability by body size, sex, location and season. This study is intended as a baseline for further research on the trophic ecology of thresher sharks in Peru.
MATERIALS AND METHODS
Sample collection
Samples were collected at five landing points in northern Peru (Zorritos, Mancora, Las Delicias, San Jose and Salaverry; Figure 1). Stomach contents were collected from small-scale driftnet fishing vessels between February and December 2015. Sharks were sexed and measured to total length (TL), with caudal fin depressed in line with the body to the cm below. Stomachs were extracted and preserved in a 10% formalin solution. Thresher sharks have similar morphological characteristics (Smith et al., Reference Smith, Rasmussen, Ramon, Cailliet, Camhi, Pikitch and Babcock2008); as a result, they can be misidentified. Therefore, a photographic record, when possible, was taken to verify the species. For data analyses, the samples were classified into two groups: individuals that could only be verified at the genus level (Alopias sp.) and individuals that could be accurately identified at a species level.

Fig. 1. Landing points where Alopias sp. samples were collected. The black line defines the division between zone 1 and zone 2, with the stomach content sample size (n).
Prey items were analysed in the laboratory and identified to the lowest possible taxon, counted and weighed. Identification guides were used to assist with identification of the teleosts and cephalopods. For cephalopods, their hard parts (i.e. beaks) were used for species identification (Iverson & Pinkas, Reference Iverson and Pinkas1971; Wolff, Reference Wolff1982; Lu & Ickeringill, Reference Lu and Ickeringill2002; Xavier & Cherel, Reference Xavier and Cherel2009), in addition, beaks were used to estimate total mass (TM) at ingestion, using regression equations (Lu & Ickeringill, Reference Lu and Ickeringill2002). For teleosts, otoliths and body remains were used for species identification (Chirichigno, Reference Chirichigno1998; Garcia-Godos, Reference Garcia-Godos2001).
Diet analysis
Diet was quantified using percentages by number (%N), weight (%W) and the frequency of occurrence (%O) (Hyslop, Reference Hyslop1980). Two diet indices were calculated: Index of Relative Importance (IRI) (1) and Prey-Specific Index of Relative Importance (PSIRI) (2). The IRI (1) was divided by the total IRI for all items to calculate the Index of Relative Importance on a per cent basis (%IRI; Cortes, Reference Cortes1997).

The Index of Relative Importance (IRI) was modified by the Prey-Specific Index of Relative Importance (PSIRI), which is additive with respect to taxonomic levels. This allows for more reliable comparisons between studies, as PSIRI values are not dependent upon taxonomic level or prey categories (Brown et al., Reference Brown, Bizzarro, Cailliet and Ebert2012). The %PSIRI was calculated using the equation of Brown et al. (Reference Brown, Bizzarro, Cailliet and Ebert2012) (2). We also present the %IRI results to facilitate comparison with other studies.

CUMULATIVE PREY CURVES
Cumulative prey curves were constructed to determine if an adequate number of stomachs had been collected to describe the diet (Jimenez-Valverde & Hortal, Reference Jimenez-Valverde and Hortal2003). The order in which stomachs were analysed was randomized 1000 times to eliminate bias. When a cumulative prey curve reaches an asymptote, the number of stomachs analysed is considered sufficient for describing the diet. According to Soberon & Llorente (Reference Soberon and Llorente1993) a slope value less than 0.1 indicates a good representation of the diet. A more rigorous procedure proposed by Bizzarro et al. (Reference Bizzarro, Robinson, Rinewalt and Ebert2007) establishes that to determine if a cumulative prey curve reaches an asymptote, the slope of the line generated from the endpoints should be compared to a line of zero slope (horizontal asymptote). The endpoints consist of the mean cumulative number of prey taxa generated for the final four stomach samples. Slopes are statistically compared using a Student's t-test and slopes that are not significant (P > 0.05), indicate that the curve reached an asymptote (Bizzarro et al., Reference Bizzarro, Robinson, Rinewalt and Ebert2007).
TROPHIC NICHE WIDTH
To evaluate the trophic niche width, the Levin standardized (Bi’) index was used based on %N values using the following formula (3):

Pij is the proportion of the prey j in the diet of the consumer i and n is the number of prey species. Values range from 0 to 1, where values closer to 0 indicate a diet dominated by few prey items (i.e. greater degree of specialization) and values closer to 1 indicate a lesser degree of specialization (Labropoulou & Eleftheriou, 1997).
TROPHIC POSITION
The relative trophic position (TP) was calculated for each individual based on the %PSIRI values of the prey species present in the stomach content. The equation of Christensen & Pauly (Reference Christensen and Pauly1992) was used (4).

DCij is the composition of the diet in which j is the proportion of prey items in the diet of the predator i and TLj is the trophic level of the prey items. Trophic positions for prey species were taken from two studies: Espinoza (Reference Espinoza2014) which presents values for northern Peru and, only when local values were not reported, Cortes (Reference Cortes1999) was used (Table suppl. 1).
DATA ANALYSIS
Diets of the two groups were analysed to determine if differences in the diet exist: at the genus level (Alopias spp.) and at the species level (individuals for which species identification had been confirmed through photographic records). Further analysis also depended on four factors: size classes, sex, location and season. Sharks were allocated into two size classes based on the estimated minimum size at maturity: sharks that measured greater than 282 cm TL for females and 267 cm TL for males (Cailliet et al., Reference Cailliet, Martin, Harvey, Kusher and Welden1983; Chen et al., Reference Chen, Liu and Chan1997; Liu et al., Reference Liu, Chen, Liao and Joung1999; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007). These values represent the minimum size at maturity for all three thresher shark species (Smith et al., Reference Smith, Rasmussen, Ramon, Cailliet, Camhi, Pikitch and Babcock2008). The division of the study area (zone 1: Zorritos, Mancora and Las Delicias; zone 2: San Jose and Salaverry) was justified based upon biogeographic characteristics of the Tropical East Pacific and Warm Temperate South-eastern Pacific marine provinces where the landing points are located (Spalding et al., Reference Spalding, Fox, Allen, Davidson, Ferdaña, Finlayson, Halpern, Jorge, Lombana, Lourie, Martin, Mcmanus, Molnar, Recchia and Robertson2007) (Figure 1). In addition, the bathymetry along the northern Peruvian margin changes: the continental shelf is narrower in zone 1 (width: 5.5–55.5 km, average: 26 km) in comparison to zone 2 (width: 40.7–129.6 km, average: 96.3 km) (Duperret et al., Reference Duperret, Bourgois, Lagabrielle and Suess1995). The division of the seasons was based upon the seasonality of chlorophyll-a concentration and primary production; for which the highest levels occurred during the austral summer and autumn (Pennington et al., Reference Pennington, Mahoney, Kuwahara, Kolber, Calienes and Chavez2006). Therefore, data were divided into two seasons: season 1 (June to November) and season 2 (December to May).
Non-metric dimensional scaling (nMDS) ordinations generated from a Bray–Curtis similarity matrix on numeric abundance of prey (%N) was used for two purposes: to determine whether a difference in diet exists between the two groups (at the genus level, Alopias spp., and at the species level), and to determine whether body size, sex, location or season exerted the greatest overall influence on the dietary composition of thresher sharks. An overall one-way analysis of similarities (ANOSIM) was used to test whether dietary compositions differed significantly by generating a R-statistic and a P-value. R-statistic values describe the extent of similarity (Clarke, Reference Clarke1993), with values near 1 indicating that the two groups are entirely separate, and values close to 0 indicating that there are no differences between the groups. This was also tested for two different purposes: to determine if there were differences in diet (1) between the two taxonomic groups (Alopias spp., and at the species level), and (2) between the factors (e.g. body size). Similarity percentages (SIMPER) were employed to determine the dietary categories that typified particular groups and/or contributed most to the similarities between groups (Clarke, Reference Clarke1993).
RESULTS
Of the 128 individuals examined, only 19 (14.8%) presented an empty stomach. Specimens of Alopias spp. not identified to species level (N = 90) measured 162–356 cm TL (mean ± SD: 291 ± 38.6), while individuals of Alopias pelagicus (N = 38) were 206–385 cm TL (mean ± SD: 291 ± 31). Length-frequencies varied by factor season, area, sex and maturity stage (Fig. Suppl. 1).
For Alopias spp., prey composition comprised 13 prey taxa: three teleosts and 10 cephalopods (including unidentified cephalopods; Table 1). According to the %PSIRI, the most important prey species were the Humboldt squid (Dosidicus gigas) (63.86%), unidentified cephalopods (10.69%), Patagonian squid (Doriteuthis gahi) (7.03%) and Peruvian hake (Merluccius gayi) (5.01%). The diet of A. pelagicus consisted of 10 prey taxa: one teleost and nine cephalopods (including unidentified cephalopods; Table 1). The most important prey species were D. gigas (65.3%PSIRI) and sharpear enope squid (Ancistrocheirus lesueuri) (9.28%). ANOSIM showed no significant differences between the diets of Alopias spp. and A. pelagicus (Fig. suppl. 2).
Table 1. Prey composition in stomach contents of thresher sharks: Alopias sp. (not identified at a species level, N: 90) and Alopias pelagicus (N: 38).

SD, standard deviation.
Percentage by number (%N), percentage by weight (%W), percentage by frequency of occurrence (%O), index of relative importance, expressed as a percentage (%IRI), and prey-specific index of relative importance (%PSIRI).
* Argonauta sp., Tremoctopus violase.
According to Soberon & Llorente (Reference Soberon and Llorente1993), the cumulative prey curves showed a trend toward an asymptote with a slope value less than 0.1 (0.03 for Alopias spp. and 0.06 for A. pelagicus) (Figure 2). However, according to Bizzarro et al. (Reference Bizzarro, Robinson, Rinewalt and Ebert2007), curves for both groups did not reach an asymptote since slopes were significantly different from a zero slope (P < 0.05).

Fig. 2. Randomized cumulative prey curve of Alopias sp. (A) and A. pelagicus (B) (grey points: observed data; black points: predicted data).
The trophic niche widths for Alopias spp. (Bi = 0.19) and A. pelagicus (Bi = 0.16) were narrow, with low standardized Levin index values indicating that the diet of Alopias shows a high degree of specialization and is dominated by a small number of prey species. The majority of stomachs contained between one (32%) and two (41%) prey species, with D. gigas the most common prey species (74% of stomachs that contained a single prey species were of D. gigas, and 76% of stomachs that contain two prey taxa also contained D. gigas) (Figure 3). Differences between factors were identified: individuals in the season 1, zone 1, size class II and females presented a greater degree of specialization (Table 2). The average trophic positions were high for both A. pelagicus (4.4 ± 0.13) and Alopias sp. (4.5 ± 0.14). No differences in average trophic positions between factors were identified.

Fig. 3. Number of prey species and number of Dosidicus gigas prey in stomach contents of thresher sharks (Alopias sp.).
Table 2. Levin's standardized index showing trophic niche width between factors in the diet of thresher sharks (Alopias sp.).

Season 1: June to November. Season 2: December to May. Zone 1: Zorritos, Mancora and Las Delicias. Zone 2: San Jose and Salaverry. Size class I: lower than the benchmark. Size class II: higher than the benchmark. Benchmark was 282 cm TL for females and 267 cm TL for males.
For Alopias spp., the average number of prey taxa per stomach content was 3 ± 1 (range: 1–6), and the number of prey items in stomachs ranged from 1–44. For A. pelagicus, the average number of prey species per stomach content was 2 ± 1 (range: 1–4), and the number of prey items per stomach ranged from 1–26. The greatest number of prey items was found in a female shark (Alopias spp.) that measured 292 cm TL which had 44 pairs of otoliths (equivalent to 44 M. gayi).
Since ANOSIM showed no significant differences between the diets of Alopias spp. and A. pelagicus, all samples were grouped (N = 128) for the subsequent analyses. The ANOSIM results showed significant differences in diet between the body size classes (R-statistic = 0.21, P < 0.01) and locations (R-statistic = 0.35, P < 0.01). The nMDS plot and the overall R-value indicated that there were small, but statistically significant differences in diet according to location and body size (Fig. suppl. 2). Individuals from Zone 1 fed primarily on D. gigas (83.60%PSIRI), while individuals from Zone 2 fed on a combination of D. gigas (44.01%) and M. gayi (23.59%) (Table 3). Both body sizes fed primarily on D. gigas; yet size class II presented higher PSIRI values (79.17%) than size class I (59.84%) (Table 3).
Table 3. Prey-specific index of relative importance (%PSIRI) of thresher sharks (Alopias sp.) according to location (zone 1: Zorritos, Mancora and Las Delicias; zone 2: San Jose and Salaverry) and body size (size class I: lower than the benchmark, size class II: higher than the benchmark; benchmark was 282 cm TL for females and 267 cm TL for males).

*Argonauta sp., Tremoctopus violase.
DISCUSSION
This study has shown that thresher sharks are top predators in the waters off northern Peru, with a diet composed mainly of Dosidicus gigas. These results complement the findings of other diet studies of Alopias spp. in the Eastern Pacific (Ecuador, California and Mexico) that found that cephalopods (e.g. D. gigas) and, to a lesser extent, teleosts (e.g. M. gayi) were important prey (Preti et al., Reference Preti, Smith and Ramon2001, Reference Preti, Smith and Ramon2004, Reference Preti, Kohin, Dewar and Ramon2008; Polo-Silva, Reference Polo-Silva2004; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009). In contrast, Rosas-Luis et al. (Reference Rosas-Luis, Loor-Andrade, Carrera-Fernández, Pincay-Espinoza, Vinces-Ortega and Chompoy-Salazar2015) found that A. superciliosus in Ecuadorian waters fed mainly on teleosts with squids of secondary importance. The present study indicated a higher trophic position (4.4–4.5) in comparison with thresher sharks (i.e. A. pelagicus, A. superciliosus) sampled off Ecuador (3.7–3.9; Polo-Silva, Reference Polo-Silva2004; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009), but a lower value than for thresher sharks (i.e. A. pelagicus (4.7), A. superciliosus (5.2)) sampled off Mexico (Li et al., Reference Li, Zhang and Dai2016). Previous studies of thresher sharks have also found only small differences in the diet according to body size, sex or season (Polo-Silva, Reference Polo-Silva2004; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009). Yet, it is important to consider that the assumptions in this study may be different from other studies since data were analysed both to the species level and to the genus level.
According to Soberon & Llorente (Reference Soberon and Llorente1993), the slopes of the cumulative prey curves indicated that overall sample sizes were sufficient to describe the diets of Alopias sp. and A. pelagicus. Yet, according to Bizzarro et al. (Reference Bizzarro, Robinson, Rinewalt and Ebert2007), sample sizes should be increased to improve the quantitative description of the diet. In addition, previous studies of the diets of Alopias have found a greater number of prey (20–27 taxa; Preti et al., Reference Preti, Smith and Ramon2001, Reference Preti, Smith and Ramon2004; Polo-Silva, Reference Polo-Silva2004; Polo-Silva & Grijalba-Bendeck, Reference Polo-Silva and Grijalba-Bendeck2007; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009). Therefore, the results of this study should be interpreted with caution. Future studies of thresher sharks in northern Peru should increase sample size toward improving their diet description accuracy. Since some species of cephalopods could not be identified to the species level, this represents a knowledge gap that also remains to be resolved.
In the South-eastern Pacific, the diet of A. pelagicus has only been studied in Ecuador (Polo-Silva, Reference Polo-Silva2004; Polo-Silva et al., Reference Polo-Silva, Rendón and Galván-Magaña2009). The authors found that this shark species fed primary on three species: D. gigas (75%IRI), purpleback flying squid (Sthenoteuthis oualaniensis) (12%IRI) and Panama lanternfish (Benthosema panamense) (9%IRI). Thus, the diet of A. pelagicus in Ecuador is composed of oceanic species. In the present study, D. gigas was also the main prey species of A. pelagicus (79%IRI) but S. oualaniensis and B. panamense were absent from the diet of A. pelagicus in northern Peru. In the case of S. oualaniensis, this may be explained by this species’ preference for warmer waters (e.g. Ecuador) than D. gigas (Nigmatullin et al., Reference Nigmatullin, Nesis and Arkhipkin2001).
In Peru, thresher sharks present similar diets to other shark species. In northern Peru, the smooth hammerhead shark (Sphyrna zygaena) has a similar diet to thresher sharks, feeding mainly on two species (i.e. Doryteuthis gahi and D. gigas) (Gonzalez-Pestana et al., Reference Gonzalez-Pestana, Acuña-Perales, Coasaca-Cespedes, Cordova-Zavaleta, Alfaro-Shigueto, Mangel and Espinoza2017). In Ecuador, blue sharks (Prionace glauca) also feed mainly on cephalopods (i.e. Ancistrocheirus lesueurii, Histioteuthis dofleini and D. gigas) (Loor-Andrade et al., Reference Loor-Andrade, Pincay-Espinoza and Rosas-Luis2017). Another ongoing study of the trophic ecology of pelagic elasmobranchs in northern Peru, using stable isotopes, found that Alopias spp., P. glauca and S. zygaena had strong trophic overlaps (Alfaro-Cordova pers. comm.). This suggests that these three commercial shark species are sharing resources, especially S. zygaena and Alopias spp (Alfaro et al., Reference Alfaro-Cordova, Del Solar, Gonzalez-Pestana, Acuña-Perales, Coasaca, Cordova-Zavaleta, Alfaro-Shigueto and Mangel2018). Such trophic interactions should be taken into account in any future development of trophic models and ecosystem-based management.
In the eastern Pacific, D. gigas is one of the most abundant and largest squids, and Peru reports the highest worldwide fishery landings (Nigmatullin et al., Reference Nigmatullin, Nesis and Arkhipkin2001). In Peruvian waters, only two species had previously been reported preying on D. gigas: S. zygaena and sperm whales (Physeter macrocephalus) (Clarke et al., Reference Clarke, MacLeod and Paliza1976; Gonzalez-Pestana et al., Reference Gonzalez-Pestana, Acuña-Perales, Coasaca-Cespedes, Cordova-Zavaleta, Alfaro-Shigueto, Mangel and Espinoza2017). In this study, results indicate that thresher sharks are also important predators, which could potentially have an impact on the overall population of D. gigas in northern Peru; however, more advanced studies are needed to verify this. The potential reduction in predation pressure on squids resulting from fisheries exploitation of their predators (Ward & Myers, Reference Ward and Myers2005; Smith et al., Reference Smith, Fulton, Hobday, Smith and Shoulder2007) might cause an increase in cephalopod biomass (Piatkowski et al., Reference Piatkowski, Pierce and Moraisda Cunha2001; Watters et al., Reference Watters, Olson, Field and Essington2008). As with the interplay of the shark species described above, these commercial species of thresher sharks and D. gigas, comprising both predator and prey, will require an ecosystem-based approach for both their populations to be managed effectively.
This study presents the first stomach content analyses for thresher sharks in Peru and the southernmost study of its diet in the eastern Pacific. As such, it can serve as a baseline to promote and guide additional, more focused and advanced studies of thresher shark trophic ecology. Future studies should pay close attention to accurately identifying this family to the species level. In addition, since this study consisted mostly of adult-sized animals, future diet studies of thresher shark diets should also include a wider range of body sizes to more adequately assess the life history of thresher shark trophic ecology.
SUPPLEMENTARY MATERIAL
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315418000504.
ACKNOWLEDGEMENTS
We would like to thank Sonia Valle and Aldo Indacochea for supporting this project and facilitating the use of the laboratory of Marine Biology at Universidad Cientifica del Sur. Also, thanks to José Carlos Xavier, Astrid Jiménez, Sergio Pingo, Silvia Kohatsu, Angie Sánchez and Akemi Arévalo for assistance, collecting, analysing and identifying samples. Bernabé Moreno for facilitating information.
FINANCIAL SUPPORT
This work was supported by Fondo para la Innovación, la Ciencia y la Tecnología (PIBA-369-2014), The DEFRA Darwin Initiative, UK, and The United States Embasssy in Peru.