Introduction
Seabirds are becoming increasingly threatened, and at a faster rate globally than all other groups of birds, as a result of problems related to the marine environment, particularly the interaction with fisheries (Croxall Reference Croxall2008, BirdLife International 2010). Seabird foraging distributions frequently overlap with commercial fisheries, often leading to negative effects such as incidental mortality of seabirds in fishing equipment and competition for common resources (Tasker et al. Reference Tasker, Camphuysen, Cooper, Garthe, Montevecchi and Blaver2000, Montevecchi Reference Montevecchi, Schreiber and Burger2002, Furness Reference Furness2003). In this context, mapping human activities and understanding their interaction with seabirds constitute important steps in the development of adequate conservation planning schemes. Growing concern worldwide has resulted in an increasing number of studies exploring the spatial relationship between fisheries and seabirds, confirming or highlighting the potential existence of negative effects (e.g. Phillips et al. Reference Phillips, Silk, Croxall and Afanasyev2006, Petersen et al. Reference Petersen, Phillips, Ryan and Underhill2008, Zador et al. Reference Zador, Parrish, Punt, Burke and Fitzgerald2008, Pichegru et al. Reference Pichegru, Ryan, Le Bohec, van der Lingen, Navarro, Petersen, Lewis, van der Westhuizen and Grémillet2009, Copello and Quintana Reference Copello and Quintana2009).
Magellanic Penguins Spheniscus magellanicus and Imperial Cormorants Phalacrocorax atriceps are two widely distributed seabirds in Patagonia, Argentina (Yorio et al. Reference Yorio, Frere, Gandini and Harris1998) and due to their numbers, biomass, and role as high-level predators, they are likely to be important components of these coastal ecosystems. Both species are also important local and regional economic resources. The Magellanic Penguin, internationally recognised as ‘Near Threatened’ (IUCN 2008), is one of the main ecotourism attractions. Some of its colonies receive over 100,000 visitors per season, generating important revenues at local and regional scales (Yorio et al. Reference Yorio, Frere, Gandini and Schiavini2001a). Imperial Cormorants are the main source of commercially harvested guano in central and southern Patagonia (Punta Reference Punta1996).
Over 25% of the total breeding population of both Magellanic penguins and Imperial Cormorants in Patagonia breed at Golfo San Jorge (Frere et al. Reference Frere, Quintana and Gandini2005, Schiavini et al. Reference Schiavini, Yorio, Gandini, Raya Rey and Boersma2005). This coastal sector, located in central Patagonia, is one of the priority areas for breeding seabirds in Argentina (Fundación Patagonia Natural 1996, Yorio Reference Yorio2000). The area also constitutes the main fishing grounds for one of the most important trawl fisheries in terms of economic return, targeting Argentine Red Shrimp Pleoticus muelleri, and an important fishing ground for the Argentine Hake Merluccius hubbsi (Bezzi et al. Reference Bezzi, Renzi, Hirsuta, Santos, Tringali, Ehrlich, Sánchez, García de la Rosa, Simonazzi, Villarino, Sánchez and Bezzi2004, Subsecretaría de Pesca y Acuicultura 2007). Recognition of the environmental value of this area has resulted in the recent decision by the Federal and Provincial Governments to designate this area as a marine protected area (Parque Marino Patagonia Austral), and thus information is urgently needed for the development of a management plan. Our goal was to evaluate the at-sea distribution of breeding Magellanic Penguins and Imperial Cormorants, explore the potential spatial overlap between these seabirds and commercial trawl fisheries, and discuss its implications for the effectiveness of the newly designated marine park to protect their populations.
Methods
Study area and species
The Golfo San Jorge (Fig. 1) extends from Cabo Dos Bahías (44°55′S, 65°32′W) to Cabo Tres Puntas (47°06′S, 65°52′W), and includes an area of more than 32,000 km2. Thirteen of the 17 Patagonian breeding seabirds nest on islands of the northern coastal sector of the gulf, including a significant proportion of the Patagonian population of some species (Yorio et al. Reference Yorio, Frere, Gandini and Harris1998). Magellanic Penguins breed at 19 colonies of between 30 and 96,000 nests totalling over 240,000 breeding pairs (Schiavini et al. Reference Schiavini, Yorio, Gandini, Raya Rey and Boersma2005) while Imperial Cormorants breed at 17 colonies of between 20 and 3,000 nests totalling approximately 13,300 breeding pairs (Frere et al. Reference Frere, Quintana and Gandini2005). These estimates of breeding pairs represent about one quarter of the Magellanic Penguin and Imperial Cormorant populations breeding in Patagonia, Argentina (950,000 and 55,000 breeding pairs, respectively). The Magellanic Penguin breeding population in coastal Patagonia is estimated to be over 70% of the global population (1,300,000 pairs; BirdLife International 2009); unfortunately there are no good estimates of the global population of Imperial Cormorant (Wetlands International 2006). Magellanic Penguins mainly feed on small schooling fish (see review in Schiavini et al. Reference Schiavini, Yorio, Gandini, Raya Rey and Boersma2005). Their diet at Golfo San Jorge is unknown, but available information from other locations in Chubut province indicates that they mainly feed on Engraulis anchoita, with the contribution of Merluccius hubbsi in some colonies (Gosztonyi Reference Gosztonyi1984, Frere et al. Reference Frere, Gandini and Lichtschein1996, Scolaro et al. Reference Scolaro, Wilson, Laurenti, Kierspel, Gallelli and Upton1999). Imperial Cormorants are a pursuit diving species that feed in relatively coastal waters mostly on fish, although they also include crustaceans, molluscs and polychaetes (see review in Frere et al. Reference Frere, Quintana and Gandini2005). At Golfo San Jorge, main prey recorded during the breeding season consisted of Engraulis anchoita, Merluccius hubbsi, Patagonotothen sp., Helcogrammoides cunninghami and P. muelleri (Punta et al. Reference Punta, Yorio and Herrera2003, Bulgarella et al. Reference Bulgarella, Cella Pizarro, Quintana, Sapoznikow, Gosztonyi and Kuba2008, Yorio et al. Reference Yorio, Copello, Kuba, Gosztonyi and Quintana2010 Magellanic Penguins and Imperial Cormorants have been recorded taking advantage of discards at different fisheries operating at Golfo San Jorge, but in relatively low frequencies and/or numbers (Yorio and Caille Reference Yorio and Caille1999, González-Zevallos and Yorio Reference González-Zevallos and Yorio2006, González-Zevallos unpubl.data), and thus it is unlikely that their populations benefit much from this supplementary food source.
Map of Golfo San Jorge, Argentina, indicating the location of Caleta Malaspina and the Parque Marino Patagonia Austral (see text).
The recently designated Parque Marino Patagonia Austral (Patagonia Austral Marine Park) (Law 26446, December 2008) includes an area of 750 km2, extending from the high tide mark one nautical mile (1.85 km) offshore, and is under the joint administration of the Government of the Province of Chubut and the National Parks Administration of Argentina (Figure 1). The goals of the Park are to protect representative samples of terrestrial, coastal and marine ecosystems, to ensure the continuity of natural processes and to promote sustainable activities compatible with the conservation of the Park. Several human activities, including commercial fisheries, use the northern sector of Golfo San Jorge. About 70 freezer trawl vessels targeting Argentine Red Shrimp and 20 ice trawlers targeting Argentine Hake operate mostly from September to May. Average annual catches for shrimp and hake fisheries during the study period were estimated at 6,000 and 23,000 tons, respectively (Secretaría de Pesca de la Provincia de Chubut, unpubl. data). In addition, a few small coastal trawlers also fish for Argentine Hake. Fishing grounds have remained the same during the study period (Secretaría de Pesca de la Provincia de Chubut unpubl. data).
Foraging areas and spatial overlap
The study of foraging seabirds was conducted on birds breeding at Caleta Malaspina (45°10′S, 66°30′W) (Fig. 1), located at the southern sector of the Parque Marino Patagonia Austral, during the 2006 and 2007 breeding seasons. Breeding numbers of Magellanic Penguins and Imperial Cormorants at the study colonies were estimated at 2,700 and 270 pairs, respectively (Schiavini et al. Reference Schiavini, Yorio, Gandini, Raya Rey and Boersma2005, Frere et al. Reference Frere, Quintana and Gandini2005). GPS loggers (GPSlog, Earth and Ocean Technologies, Kiel, Germany) were deployed on 45 adult breeders of both sexes (seven female and eight male Magellanic Penguins; 14 female and 16 male Imperial Cormorants) during late November, corresponding to the early chick stage of both species (first 10–15 days after hatching). These loggers recorded time, latitude and longitude once per second, and the horizontal accuracy is considered to be better than 5 m for 90% of fixes (GPSlog Manual). The units were streamlined to minimise drag and had maximum dimensions of 95 × 48 × 24 mm (L × W × H) and weighed 65 g, representing less than 1.5% and 2.9% of Magellanic Penguin and Imperial Cormorant mean body mass, respectively (Yorio et al. Reference Yorio, García Borboroglu, Potti and Moreno2001b, Svagelj and Quintana Reference Svagelj and Quintana2007). Devices were attached to the feathers of the lower back using waterproof tape (Wilson et al. Reference Wilson, Pütz, Peters, Culik, Scolaro, Charrasin and Roupert-Coudert1997) to minimise hydrodynamic drag (Bannasch et al. Reference Bannasch, Wilson and Culik1994). Every effort was taken to minimize the stress caused to the birds during manipulation. The procedure was completed in less than five minutes and birds quickly returned to their nest. GPS loggers were retrieved when birds came back from the sea and returned to the nest. All birds equipped with devices showed no signs of discomfort due to the attachment of the device and continued breeding normally during the study period. Moreover, previous studies on the two species suggest no alteration of foraging behaviour by birds carrying similar devices (see Wilson and Quintana Reference Wilson and Quintana2004, Wilson et al. Reference Wilson, Kreye, Lucke and Urquardt2004, Quintana et al. Reference Quintana, Wilson and Yorio2007, Gómez Laich et al. Reference Gómez Laich, Wilson, Quintana and Shepard2008, Shepard et al. Reference Shepard, Wilson, Quintana, Gómez Laich and Forman2009). Localisations of birds at their nests were identified to ground-truth the data and were based on nest position obtained by a commercial GPS unit (GPS III plus, Garmin International). All penguins and half of the cormorants performed just a single foraging trip during the instrumented period. All recorded trips were used to develop maps (see below) while only one trip per bird was used for statistics to avoid pseudoreplication. Means are reported ± 1 SD.
Spatial analyses were made using ArcView 3.2 (Environmental Systems Research Institute, 1992–1999). Foraging at-sea distributions of seabirds were estimated by fixed kernel method (Worton Reference Worton1989) and ad hoc calculation of smoothing parameters as implemented in the Animal Movement v2.04 beta extension (Hooge and Eichenlaub Reference Hooge and Eichenlaub2000). The locations of observed hauls made by hake (n = 2,081) and shrimp (n = 12,378) trawlers between November 15 and March 15 of the 2001-2007 fishing seasons in the fishing area under Chubut jurisdiction south of 44°55’S were obtained from the database of the On-board Observer Program of the Chubut Province (Secretaría de Pesca de la Provincia de Chubut unpubl. data). The number of sampled vessels varied among years, representing approximately 14% and 17% of the hake and shrimp fishing fleet sizes, respectively. Utilisation areas for the two fishing fleets were estimated using these trawls by fixed kernel method (see above). GPS devices provided the time to the nearest second for each fix, and the amount of time spent by birds at fishing areas was estimated using the accumulated time for all fixes recorded within areas classified as high intensity use (95% contours). Sample sizes differ as not all foraging trips were completely recorded or headed to open waters where fishing takes place.
Seabird incidental mortality at trawl vessels
Information on the number of Magellanic Penguins and Imperial Cormorants killed in nets was gathered on board hake and shrimp trawlers for a total of 179 and 3,367 hauls respectively from 15 November 2004 to 15 March 2005. Information was obtained from the database of the On-board Observer Program of the Chubut Province. Bird mortality was expressed as average rates per haul. Distance (in km) to the nearest shore of each haul with a recorded incidental capture was obtained using ArcView 3.2. Bootstrapping was used to assess confidence intervals around the mean mortality rates (Efron and Tibshirani Reference Efron and Tibshirani1993). For each set of data, 3,000 random re-samplings (with replacement) of the original data set were produced, mean mortality rate was calculated, and the results were then sorted to obtain the 95% confidence interval (CI) of the mean (R Development Core Team 2004).
Results
Spatial foraging patterns of penguins and cormorants
In both study years, Magellanic Penguin foraging areas were located south of the colony, mostly following the coastline in 2007 (Figure 2). Imperial Cormorants travelled east and south of their colony, and although they foraged in similar areas during both study seasons, they showed a higher use of the internal waters of Caleta Malaspina during 2007 (Figure 3). Foraging range (maximum distance from the colony) varied between individuals and years, with a maximum of 120.1 and 68.2 km for Magellanic Penguins (Table 1, Figures 2a and 2b) and Imperial Cormorants (Table 1, Figure 3a and 3b), respectively. The duration of Magellanic Penguin foraging trips (defined as the time elapsed between departure and return from the nest/colony) varied between one and two days, while those of Imperial Cormorants were of only a few hours long (Table 1).
At sea distribution (expressed as locations per unit area) of foraging Magellanic Penguins from Caleta Malaspina, Golfo San Jorge, Argentina, during the chick stage of (a) 2006 (n = 5 trips) and (b) 2007 (n = 10 trips).
At sea distribution (expressed as locations per unit area) of foraging Imperial Cormorants breeding at Caleta Malaspina, Golfo San Jorge, Argentina, during the chick stage of (a) 2006 (n = 22 trips) and (b) 2007 (n = 25 trips).
Foraging patterns of Magellanic Penguins and Imperial Cormorants breeding at Caleta Malaspina, Golfo San Jorge, Argentina, during the 2006 and 2007 breeding seasons (mean ± SD with range in parentheses).
(1)n = 3; (2)n = 5
Seabird spatial overlap with fisheries and incidental mortality
In both years, foraging Magellanic Penguins and Imperial Cormorants showed a clear overlap with operating vessels of both trawl fisheries (Figure 4 and 5, respectively). Birds remained a significant proportion of their time at sea in waters of high intensity use (95% kernel contours) by both trawl fisheries (penguins: 17.9–66.2%; cormorants: 46.0–89.0%) (Table 2). During foraging excursions, birds reached the areas of high intensity use (95% kernel contours) soon after leaving their colony (penguins: < 1.30 h; cormorants: < 0.24 h) (Table 2), and distances travelled to fishing grounds were extremely short for both species (< 4.43 km) (Table 2).
Tracks of foraging Magellanic Penguins instrumented at Caleta Malaspina during the chick stage of 2006 (a and c) and 2007 (b and d), and spatial distribution of hauls made by hake trawlers (a and b) and shrimp trawlers (c and d) between November 15 and March 15 of the 2001–2007 fishing seasons at Golfo San Jorge, Argentina. Grey areas indicate the areas of high intensity use by hake and shrimp trawl fisheries (95, 75 and 50% kernel contours).
Tracks of foraging Imperial Cormorants instrumented at Caleta Malaspina during the chick stage of 2006 (a and c) and 2007 (b and d) and spatial distribution of hauls made by hake trawlers (a and b) and shrimp trawlers (c and d) between November 15 and March 15 of the 2001–2007 fishing seasons at Golfo San Jorge, Argentina. Grey areas indicate the areas of high intensity use by hake and shrimp trawl fisheries (95, 75 and 50% kernel contours).
Spatial distribution of Magellanic Penguin and Imperial Cormorant incidental mortality in hake (a and b) and shrimp (c and d) trawl fisheries operating at Golfo San Jorge between 2004 and 2005. Grey circles indicate number of individuals incidentally captured per trawl; black circles indicate location of colonies. Source: On-board Observer Program of Chubut Province.
Average time spent in fishing areas relative to the duration of the trip (%), average travel time (h) and distance to fishing grounds (km) for foraging Magellanic Penguins and Imperial Cormorants breeding at Caleta Malaspina, Golfo San Jorge, Argentina. Fishing areas are those represented by the 95% probability kernel contour (see Figures 4 and 5).
(1)n = 12; (2)n = 9
A total of 142 Magellanic Penguins and 54 Imperial Cormorants were killed in nets at both trawl fisheries operating in the Golfo San Jorge during the 2004/2005 fishing season. Mortality rates varied between 0.02 and 0.34 individuals per haul, depending on species and fishery (Table 3). The record shows that incidental captures occurred close to shore in both study years (Magellanic Penguins: 6.3 ± 4.6 km, range = 0.5–18.0, n = 142 birds; Imperial Cormorants: 6.7 ± 4.9 km, range = 1.1–20.9, n = 54 birds, Figure 6). During the evaluated period, shrimp vessels operated at a mean distance from the coast of 5.0 ± 4.6 km (range = 0.1–92.5, n = 3,367 hauls) while hake vessels operated at an estimated distance of 11.5 ± 5.3 km (range = 0.2–37.4; n = 179 hauls).
Mortality rate of Magellanic Penguins and Imperial Cormorants at the hake and shrimp trawl fisheries operating in Golfo San Jorge, Argentina, during the 2004/2005 fishing season.
Discussion
Results show that during the chick stage, Magellanic Penguins and Imperial Cormorants forage mostly in coastal waters inside the gulf, travelling up to 120 and 70 km from the colony, respectively. Magellanic Penguins and Imperial Cormorants were recorded feeding in similar coastal sectors, although the latter showed larger foraging ranges. Distances travelled by Magellanic Penguins in our study were similar to those recorded during the chick stage at other locations in Patagonia (Wilson et al. Reference Wilson, Scolaro, Grémillet, Kierspel, Laurenti, Upton, Gallelli, Quintana, Frere, Müller, thor Straten and Zimmer2005, Boersma et al. Reference Boersma, Rebstock, Stokes and Majluf2007, Boersma and Rebstock Reference Boersma and Rebstock2009). It should be noted, however, that Magellanic Penguin foraging patterns obtained in this study should be interpreted with caution given that they were derived from a relatively low sample size. Imperial Cormorants feeding areas were also located mostly close to shore. Results of their foraging patterns agree with previous evaluations in the same study area, where Imperial Cormorants instrumented with radio-transmitters travelled further than 8 km from the colony (Sapoznikow and Quintana Reference Sapoznikow and Quintana2003) and individuals instrumented with time-depth recorders foraged up to depths of 70 m (at least 10 km from the colony) (Quintana et al. Reference Quintana, Wilson and Yorio2007). Although the information obtained in the present study is restricted to the early chick stage, it provides valuable insights into potential conflicts between foraging seabirds and fisheries. However, further studies are needed to assess spatial needs of foraging birds at other stages of the breeding cycle, as studies on Magellanic Penguins and other seabirds have shown that spatial requirements are larger during the incubation stage (e.g. Boersma et al. Reference Boersma, Stokes and Strange2002, Huin Reference Huin2002, Boersma and Rebstock Reference Boersma and Rebstock2009).
Both Magellanic Penguins and Imperial Cormorants foraged within the waters of the gulf, overlapping with hake and shrimp fisheries during their foraging trips. Birds breeding at Caleta Malaspina arrived at fishing grounds shortly after leaving their colony and, depending on the season, spent a significant part of their time at sea within the waters used by both fishing fleets. The observed foraging ranges and time spent by both species at hake and shrimp fishing grounds suggest a high probability of spatial conflict. Potential spatial conflicts with fishing operations may not be restricted to seabirds breeding at Caleta Malaspina. Magellanic Penguins and Imperial Cormorants nest in a total of 19 and 17 colonies, respectively, along the 100 km coastal sector of northern Golfo San Jorge (see Study Area). All these colonies are located within less than 20 km of the areas of high intensity use by hake and shrimp trawl vessels, with more than 70% of them at even less than 5 km (Figure 7). Even recognising the potential variability in foraging range among colonies, a simple extrapolation of the maximum foraging ranges obtained for penguins and cormorants from Caleta Malaspina to neighbouring colonies in northern Golfo San Jorge indicates that at all these locations feeding seabirds and trawl vessels can show a significant spatial overlap (Figure 7).
Estimated overlap between hake and shrimp trawl fisheries and foraging Magellanic Penguins and Imperial Cormorants breeding in colonies in the north of Golfo San Jorge, Argentina. The hatched area shows the potential foraging area of Magellanic Penguins (a and c) and Imperial Cormorants (b and d) during the early chick stage, estimated using the maximum foraging distance obtained for the two species at the Caleta Malaspina colonies (see text). Grey areas indicate the areas of high intensity use by hake (a and b) and shrimp (c and d) trawl fisheries (95% Kernel contours); black circles indicate location of colonies.
Hake and shrimp trawl fisheries produce significant amounts of fishery discards, which are thrown overboard (Cordo Reference Cordo2005, González-Zevallos and Yorio Reference González-Zevallos and Yorio2006). These attract large number of seabirds, including Magellanic Penguins and Imperial Cormorants (González-Zevallos and Yorio Reference González-Zevallos and Yorio2006). Both species take advantage of discards using the pursuit diving method, obtaining discards directly from, or falling off, the net during haul-back, thus increasing their chances of becoming entangled (González-Zevallos and Yorio Reference González-Zevallos and Yorio2006). This attraction to fishing vessels to make use of waste is very likely an important factor influencing the mortality from drowning in nets observed in this and previous studies in Golf San Jorge (Gandini et al. Reference Gandini, Frere, Pettovello and Cedrola1999, González-Zevallos and Yorio Reference González-Zevallos and Yorio2006). Unfortunately, the effects of the estimated incidental mortality on the population dynamics of penguins and cormorants breeding at Golfo San Jorge is still unknown, but mortality of adult individuals should not be underestimated given the life-history traits of seabirds (Furness and Monaghan Reference Furness and Monaghan1987, Saether and Bakke Reference Sæther and Bakke2000). Similarly, information is insufficient to evaluate the effects of trawl fisheries on penguin and cormorant trophic resources through extraction and habitat disturbance.
Spatial foraging patterns of Magellanic Penguins and Imperial Cormorants suggest that fisheries operations could potentially affect their populations breeding within the gulf and, in addition, trigger the occurrence of negative interactions with wildlife based tourism and guano harvesting. Although Magellanic Penguin and Imperial Cormorant populations are among the targets of the recently designated marine park, the defined boundaries do not appear to be adequate for their effective protection. The obtained information shows that both Magellanic Penguins and Imperial Cormorants forage mostly beyond the limits of the park, where they overlap with trawl fisheries and may be killed in fishing gear. Marine protected areas are in general ineffective for the protection of highly mobile species such as seabirds (Boersma and Parrish Reference Boersma and Parrish1999, Hyrenbach et al. Reference Hyrenbach, Forney and Dayton2000, Yorio Reference Yorio2000), but some degree of protection of seabird breeding populations might be achieved by including relatively small extensions of adjacent sea to colony locations because these birds are central place foragers. However, given the scale at which penguins and cormorants forage it is unlikely that the needed protection will be afforded by the marine park, and thus the spatial design of the park needs to be re-evaluated and/or off-boundary actions and specific governance arrangements will have to be developed to address the complexities of the marine system under consideration (Yorio Reference Yorio2009). It is expected that the information presented in this study will help raise awareness among managers and stakeholders of the potential conflicts between fisheries and breeding seabirds and contribute to both the development of the management plan for the recently designated Marine Park and a conservation strategy for the Golfo San Jorge’s marine wildlife and environments.
Acknowledgements
Research was funded by grants from the Wildlife Conservation Society and Proyecto ARG/PNUD ARG 02/018 Conservación de la Diversidad Biológica y Prevención de la Contaminación Marina en Patagonia. We thank R. Wilson, E. Shepard, A. Gómez-Laich, and R. Vera for field assistance and J. Lancellotti for help in data analysis. We also thank Centro Nacional Patagónico (CONICET) for institutional support, Soriano S.A. and the On-board Observer Program of the Chubut Province for logistical support and access to data, and the captains and crews of the hake shrimp trawlers for their help and advice.
