Deep-water shrimp fisheries in Latin America : a review

Commercial fisheries are expanding their activities into deeper water. The life history features of these deep-water resources make them more vulnerable to exploitation than most shallow-water resources. Moreover, the apparent lack of solid information about the ecology of most deep-water species represents a major limitation for the development and implementation of management strategies. This scenario has caused great concern regarding the sustainability of these resources and the possible environmental impacts on the deep-sea ecosystem. In Latin America, commercial fisheries are going deep as well, and considering the above-mentioned concerns, we felt the need to compile the available information about the deep-water shrimp resources and the current status of their fisheries in Latin America. Focusing on Mexico, Central America, Peru, Chile and Brazil, this review describes the exploited species, and, whenever available, the fishing fleet, fishery statistics, and management strategies. A total of 17 species (10 spp. of Penaeoidea; 7 spp. of Caridea) are of commercial interest in Latin America, but deep-water shrimps are currently fished only in Costa Rica, Colombia and Chile. An implemented management plan exists in Chile and Colombia, while Brazil approved fishery regulations for the aristeid fishery, which were never implemented. Considering the lack of information about the biology of the deep-water shrimps, which hinders the development of adequate management strategies, we see the urgent need to improve the communication and collaboration between the different stakeholders in Latin American. We suggest the establishment of a searchable and constantly updated database, which may serve as a valuable source of information for researcher and decision makers. Finally, we propose the development of regional research plans aimed towards supporting measurements for a sustainable use of deep-water shrimps in Latin America.


INTRODUCTION
Many marine fisheries, especially those in coastal waters, are fully exploited, threatened by overexploitation or already collapsed.According to Pauly (2009), demersal resources of all large shelves of the world have been depleted by the end of the 20th century.Declining catches in coastal zones and the concomitant increase in demand for seafood have forced the fishing industry to expand their activities into offshore and deep-water areas (Morato et al., 2006).As a consequence, the mean depth of bottom marine fish catches has increased from around 103 m in the early 1950s to 145 m in 2001, which represents a 42 m increase (Morato et al., 2006).Associated with this bathymetric shift in bottom fishery activities is an increase in the mean longevity of the catch, most dramatically since the early 1990s (Morato et al., 2006).
This trend is a serious concern, because deep-water species exhibit a series of life history traits, which make them especially vulnerable to overexploitation and extinction with a low capacity to recover from over-exploitation (e.g.Jennings et al., 1998;Cheung et al., 2005, Morato et al., 2006).Moreover, the apparent lack of solid information about the ecology of most deep-water species represents a major limitation for the development and implementation of management strategies for deep-water resources (Polidoro et al., 2008).Bensch et al. (2008) and Ramirez-Llodra et al. (2011) provided an overview of the current status of high seas bottom fisheries around the world, focusing principally on deep-water fisheries.The authors concluded that even basic information concerning total catch, by-catch species and life history of the deep-water target species is deficient and therefore not enough to generate conservation and management strategies.According to these authors, it is critical to assess these aspects if fisheries management intends to effectively ensure the long-term sustainability of these resources.
Deep-water shrimp fisheries have been developed in different parts of the world.In the northwest Atlantic, Canadian vessels began the exploitation of the northern shrimp Pandalus borealis between the late 1970s to early 1980s.Recently, many other countries reported catches of this species from high seas (Bensch et al., 2008).Different Heterocarpus species have been considered as large, relatively unexploited resources that may prove valuable to Pacific island nations and states; H. laevigatus and H. ensifer were identified as having the greatest commercial potencial.Ostazeski (1997) provided an overview of the deep-water shrimp fishery targeting H. laevigatus in northern Mariana Islands.He concluded that this fishery has yet to be fully exploited, but high costs involved in fishing in these deep-water areas may be more limiting than the sustainable yield.Well-established deep-water shrimp fisheries (Aristaeomorpha foliacea and, in particular, Aristeus antennatus; for review of the biology see Sardà et al., 2003;Politou et al., 2005) have been developed in the Mediterranean.Commercial fishing for these red shrimps started in the Ligurian Sea in the 1930s (Sardà et al., 2004).Stocks of the two aristeid shrimps collapsed in the 1970s, and fishing resumed following signs of stock recovery in 1985 (Sardà et al., 2004).A revision of the current status of the aristeid fishery in the Mediterranean is provided by Bensch et al. (2008).
Latin American fisheries are tending towards deeper water as well (Arana et al., 2009a).Several publications provide clear evidence for the increasing scientific and commercial interest in deep-water resources (e.g.Arana et al., 2002;Wehrtmann & Echeverría-Sáenz, 2007;Wehrtmann & Nielsen-Muñoz, 2009).More recently, Arana et al. (2009b) edited a compilation of information about deep-water fisheries and resources from both the Pacific and Atlantic coasts of Latin America.While commercial deep-water fisheries generally target bony and cartilaginous fishes (FAO, 2007), the situation in Latin America seems to be different: in this geographical zone, deep-water fishing activities focus largely on benthic and demersal invertebrates, and despite their relatively recent fishing history, 45% of these resources are already considered to be exploited or overexploited (Arana et al., 2009a).
Considering the above-mentioned situation, we felt the need to prepare an overview of the current status of the deep-water shrimp fisheries in Latin America.The present review compiles available information about these deep-water resources principally in Mexico, Central America, Peru, Chile and Brazil.In some of these countries/regions, no commercial fisheries have been established yet; in such cases, we gathered the available data obtained from scientific surveys carried out to assess the potential of the resources.The review provides a brief introduction to the deep-water shrimp species of commercial interest in Latin America, and presents, whenever available, information about the fishing fleet, fishery statistics, and management strategies.We finish our review with some recommendations for the future development of these fisheries and the associated research programs.

I.1. Introduction. National fisheries production in
Mexico has reached a plateau around an annual record of 1.5 million metric ton (CONAPESCA, 2008).There seems to be little possibilities of substantially increasing this production while considering only traditional fishery resources, which are mainly caught in shallow waters (<100 m depth) of the continental shelf.Thus, the exploration of potential deep-water fishery grounds and other new fishery resources appears to be an alternative to raise this production in the Exclusive Economic Zone (EEZ) of Mexico, which covers more than 3 million km 2 (Gracia, 2001).Nonetheless, only scattered information on the potential fishery resources of this enormous area is available.
Shallow water penaeid shrimps, which can be found both in coastal lagoons and in the continental shelf, constitute the most important fishing resource in Mexico.More than 15 species are reported to occur in the Gulf of Mexico (Vázquez-Bader & Gracia, 1994;Gracia & Hernández-Aguilera, 2005), but only six of them are commercially exploited (Farfantepenaeus aztecus, F. duorarum, F. brasiliensis, Litopenaeus setiferus, Sicyonia brevirostris and Xiphopenaeus kroyeri).Shallow water penaeid shrimps have been subjected to both artisanal inshore and industrial offshore fisheries and, as in many other parts of the world, these resources are overfished.High exploitation rates caused both growth and recruitmentoverfishing of some of the main shrimp species (F.duorarum, F. brasiliensis, L. setiferus) of the Gulf of Mexico (Gracia, 1996;Gracia & Vázquez-Bader, 1999).As a consequence of overfishing, the overall shrimp production of the Gulf of Mexico declined drastically after reaching its maximum yield during the 1970s-1980s (Gracia, 2004).Operations of the industrial offshore trawl fishery are limited to a depth of less than 100 m due to the bathymetric distribution of penaeid shrimps.Considering the intense trawling activities in the shelf bottoms of the Gulf of Mexico, it is unlikely to find unexploited shrimp fishing grounds of the traditionally exploited species.Therefore, the only possibility to increase shrimp catches is to find potential fishing grounds in deep-water areas in the Gulf of Mexico and in the EEZ of Pacific Mexico.
As a response to this situation, several initiatives were developed to obtain a better overview of the deep-water resources in Mexican waters.Regarding the Mexican Pacific, Hendrickx (2003Hendrickx ( , 2004) ) reported on the presence of four deep-water shrimp species (Benthesicymus tanneri Faxon, 1893; Acanthephyra brevicarinata Hanamura, 1984; Heterocarpus affinis Faxon, 1893; Pandalopsis ampla Bate, 1886) in the southeastern Gulf of California at depths between 600 and 2250 m.Hendrickx (2004) concluded that H. affinis could be considered as a target species.Starting in 2004, the project SAGARPA-CONACYT 040 aimed to carry out an exploratory fishery of deepwater shrimps in western Baja California (Flores et al., 2004).The results (Flores et al., 2004) showed the presence of Pandalus platyceros Brandt, 1851 between 126 and 252 m depth, and the species was considered as an important fishery resource yet not exploited in Mexico; the authors provided also a brief summary on the biology and fishery of this species.
Referring to the Atlantic coast of Mexico, Wicksten & Packard (2005) compiled records of 130 decapod species occurring on the continental slope (200-2500 m) and the abyssal plain (2500-3840 m) of the Gulf of Mexico.On the other hand, the Universidad Nacional Autónoma de Mexico is conducting a scientific research program, which began in 1999 and was then resumed in 2007 (Proyecto PAPIIT IN223109-3, Biodiversidad y Recursos Pesqueros Potenciales del Mar Profundo del Golfo de México).This project aims to study and assess potential fishing resources, mainly deep-water shrimps (between 300-1200 m depth) of the Gulf of Mexico.The ongoing research program has been conducted on board of the R/V JUSTO SIERRA.Five research cruises (1999,2007,2008,2009,2010) have been carried out as part of the ongoing project.Sampling stations were distributed along the entire Mexican Gulf of Mexico, from the Mexican border with the United States of America to the Mexican Caribbean Sea.

I.2. Presentation of the species.
Six main deep-water shrimp species of the families Aristeidae, Solenoceridae and Penaeidae were found in the 300-1200 m depth range (Gracia et al., 2010).Other crustaceans (geryonid crabs, deep-water lobsters and other shrimp species) and fishes of potential commercial interest were found, too.The following deep-water shrimps, however, stood out due to their abundance and size.
Aristaeomorpha foliacea (Risso, 1827) (giant red shrimp; local names: "camarón rojo gigante", "gamba española", and "camarão moruno" in Brazil) (Fig. 1).This aristeid shrimp is widely distributed and has been reported in the eastern Atlantic from the Bay of Biscay to western Sahara, Azores, Madeira, Canary Islands, the Mediterranean Sea, and off the coast of South Africa.In the western Atlantic, the species is known to occur from Massachusetts to Florida, in the Gulf of Mexico, and in the Caribbean Sea to Brazil.It has been also collected in the Indo-West Pacific from East Africa to Japan, New Zealand and Fiji (Pérez-Farfante & Kensley, 1997;Tavares, 2002).The results of the scientific surveys (Gracia et al., 2010) revealed the presence of A. foliacea in the entire Gulf of Mexico.This species sustains commercial fisheries in the Mediterranean (Figueiredo et al., 2001;Ragonese et al., 2001) and is currently exploited in Brazil (see below).The giant red shrimp has a bathymetric distribution from 250 to 1300 m depth, and is found to be associated with mud bottoms (Tavares, 2002).
Pleoticus robustus (Smith, 1885) (royal red shrimp; local name: "camarón rojo real") (Fig. 3).The distribution of this solenocerid shrimp is limited to the western Atlantic: from south of Martha's Vineyard (Massachusetts, USA), throughout the Gulf of Mexico and the Caribbean Sea to French Guiana (Holthuis, 1980;Pérez-Farfante & Kensley, 1997;Tavares, 2002).The royal red shrimp is the basis of a small but well-established fishery in the USA waters of the Gulf of Mexico off Alabama and Florida and Florida's Atlantic coast.The fishery is most active in the north of the Gulf of Mexico averaging 123380 kg per year since 1998 (Stiles et al., 2007).The species inhabits the upper continental slope between 180-730 m over blue/black mud, sand, muddy sand or white calcareous mud bottoms (Pérez-Farfante & Kensley, 1997;Tavares, 2002).During the scientific surveys in the Mexican Gulf of Mexico the species was collected between 213 and 764 m depth.A. Milne Edwards & Bouvier, 1909 (purple shrimp;local names: "alistado" or "camarón púrpura") (Fig. 4).The presence of this aristeid shrimp is restricted to the western Atlantic with a geographical distribution extending from Florida to Brazil, including the Gulf of Mexico and the Caribbean Sea (Holthuis, 1980;Pérez-Farfante & Kensley, 1997;Tavares, 2002).In Latin America, the species was seasonally exploited in French Guiana (Guéguen, 1998(Guéguen, , 2001)), and this species was also fished in Brazil, which is addressed later in this paper.Its bathymetric distribution ranges from 200 to 1144 m, although its density peaks are at 750 m depth (Pezzuto et al., 2006).In the Mexican Gulf of Mexico, the shrimp was found on mud bottoms above 300 m depth.

Aristeus antillensis
Penaeopsis serrata Bate, 1881 (pink speckled shrimp; local name: "camarón rosado moteado").The penaeoid shrimp has an amphi-Atlantic distribution and can be found on mud bottoms between 120 to 640 m (Pérez-Farfante & Kensley, 1997).This species has a potential commercial importance, but its compa- ratively small size may render its exploitation unappealing.Rathbun, 1901 (rose shrimp;local name: "camarón rosa").This species of the family Penaeidae is distributed in the western Atlantic, and has been reported from the Gulf of Mexico to Uruguay (Pérez-Farfante & Kensley, 1997).It can be found between 190-412 m, associated to mud bottoms (Pérez-Farfante & Kensley, 1997).This species is less important in terms of number and weight compared to the other deep-water shrimp species found in Mexican waters of the Gulf of Mexico.

Parapenaeus americanus
Many of these species have a wide worldwide distribution, but all of them are found along Latin America and some of them are already commercially exploited.Currently, deep-water shrimps are not subject to a commercial fishery in Mexico.Pleoticus robustus is the only deep-water shrimp that is supporting fishery activities in USA waters of the Gulf of Mexico.Sampling of benthic communities (1999Sampling of benthic communities ( -2010) ) was carried out day and night using a single semi-commercial shrimp trawl net (18 m mouth aperture, 4.5 cm stretched mesh body and 1.5 cm stretched mesh cod-end) after surveying the bottom with a sub-bottom profiler echosounder.When finding appropriate bottoms for trawling, exploratory trawling was done during 30 minutes at a speed of 2.5-3 knots.Distance travelled by the net was measured using the ship's high precision GPS; the distance covered in 30 minutes varied between 0.8 and 1.2 nautical miles, depending on sea conditions.I.4.Catch statistics.In the Gulf of Mexico, the most important deep-water shrimp species (in terms of biomass) were A. foliacea, A. edwardsiana and P. robustus, which together represented about 90% of the total weight catch (Fig. 5).Their mean sizes were also the largest ones registered along the scientific survey, varying between 35 and almost 60 mm CL.The remaining 10% of the deep-water shrimp biomass was constituted by A. antillensis, P. serrata and P. americanus, which were very abundant in some samples, but mean size of these species was comparatively small.The smallest species was P. americanus with a mean size of 22.5 mm CL (Table 1).

I.3 Scientific sampling method.
Catch per unit effort (CPUE) in kilograms per hour was used as an index of deep-water shrimp abundance and was arranged by 100 m depth strata.CPUE in the scientific survey varied from values near zero (0.02 kg/h) to a maximum CPUE of 16.4 kg h -1 .Catch rates showed high variability between different strata and geographical areas sampled.This agrees well with the typical patchy distribution of deep-water shrimps (Anderson & Lindner, 1971;D'Onghia et al., 1998;Belcari et al., 2003).Nonetheless, higher catches were consistently found in the 300-700 m depth interval (Fig. 6).The catch rates obtained in the Gulf of Mexico scientific survey are within the range of CPUE values reported for several deep-water shrimp fisheries in the world: for example, in Latin America, there are high mean annual CPUE values reported for A. edwardsiana fisheries in Brazil from 4.7 to 14 kg h -1 (Dallagnolo et al., 2009a).On the other hand, CPUE values reported for A. foliacea and A. antillensis from the Brazilian deep-water shrimp fisheries are considerably lower, varying between 0.76-6.3kg (this paper) and 0.005-2.4kg (Dallagnolo et al., 2009a), respectively.CPUE registered in deepwater shrimp fisheries for A. foliacea and A. antennnatus in the Mediterranean Sea and eastern Atlantic off Portugal have a range similar to the CPUE obtained in the Gulf of Mexico (0-12 kg h -1 ) (D'Onghia et al., 1998;Figueiredo et al., 2001;Carbonell & Azevedo, 2003;Can & Aktas, 2005).Also, Pezutto et al. (2006) have reported CPUE figures for A. edwardsiana in the southern region and the northern border of the ZEE Brazil that varied between 6.5 to 9.7 kg h -1 .
Deep-water shrimp catch rates recorded in the Gulf of Mexico seem to be lower than the mean annual CPUE of some of the deep-water fisheries conducted in the world.However, results of an exploratory scientific survey maybe not totally comparable to those of commercial fishing operations of a specialized fleet of commercial trawlers.Statistics of a developed fishery are more precise due to a larger fishing effort and an increased knowledge concerning profitable fishing grounds.Nevertheless, the CPUE recorded in the Gulf of Mexico serves as an index of the abundance of deep-water shrimp.Gracia et al. (2010) pointed out that about 40% of the hauls in the SW Gulf of Mexico resulted in high catch rates of above 5 kg h -1 .It is also important to highlight that deep-water shrimp CPUEs compare well with, and sometimes are even higher than current CPUEs obtained in the depleted coastal white and pink shrimp fisheries (1.1-10 kg) of the Mexican Gulf of Mexico (INAPESCA, 2009).These two coastal species play an important role in sustaining local shrimp fisheries in Mexico in spite of the apparent overexploitation of these resources.
Deep-water shrimp abundance indexes recorded for the Gulf of Mexico, as well as the extent of its potential distribution area between 300 and 1000 m depth (about 60,000 km 2 ; Fig. 7) reinforce the expectation that these species may represent a valuable fishing resource.In this large area there is a complex topography, including basins and submarine canyons, occasionally with high concentrations of deep-water shrimps (Fig. 7).It seems unlikely, however, that this potential fishery resource may reach   the high yields obtained in the coastal brown shrimp (F.aztecus) fishery (annual production of 7,000 metric ton with a CPUE per net between 20-65 kg h -1 (INAPESCA, 2010), whose stock is in good conditions and currently sustains the entire shrimp industry of the Gulf of Mexico (Gracia, 2004).However, the results obtained during the scientific research program seems to indicate that this potential fishing resource may render similar or even higher yields than those in the depleted coastal white (L.setiferus) and pink (F.duorarum) shrimp stocks; the annual production of these resources decreased to around 75 and 600 ton, respectively (Gracia, 2004;INAPESCA, 2005).It is difficult to estimate potential landing figures for these deep-water shrimps in the Gulf of Mexico with the currently available data, but it is reasonable to assume that the production will be low according to what usually occurs in established deepwater fisheries in other parts of the world (Jones et al., 1994;Politou et al., 2003;Pezutto et al., 2006).However, even if a relatively small-scale fishery can be established, it still could represent a valuable alternative fishing resource for the troubled Mexican shrimp industry of the Gulf of Mexico.
I.5.Conclusions.Developing this potential deepwater shrimp fishery requires an integrative research initiative that should consider the following aspects: 1) stock assessment of deep-water shrimps; 2) delimitation of fishing grounds, and 3) establishment of exploitation levels for these shrimps.Management strategies should include the use of the by-catch, which is composed mostly by other less abundant shrimps (mainly pandalid shrimps), geryonid crabs, deep-sea lobsters, and fishes that could be of commercial importance.Concurrently, it would be recommendable to develop processing and marketing strategies in order to obtain the maximum profit of these potential fishing resources.
The sustainable utilization of this potential deepwater shrimp resource will require reliable information on the biology and ecology of these deepwater shrimp species, especially on mortality and growth rates, life cycle history, and population renewal rates.Initiatives for an ecosystem-based management would be important to minimize possible impacts on the deep-sea ecosystem.This information is critical to define management strategies directed toward a sustainable use of these resources.Finally, we suggest that existing experiences (mainly in Latin America) concerning the responsible use of this deepwater shrimp resource should be consulted and adapted to the present situation in order to help in the development of management strategies that permit a maximum sustainable yield of this fragile fishing resource while minimizing the environmental impact of this fishery.

II. CENTRAL AMERICA
II.1.Introduction.Central American fisheries targeting crustaceans have focused on shallow water species of the Pacific coast.Information about the exploitation of deep-water decapods is scarce and poorly documented in national fishery statistics; for example, catches of Heterocarpus spp.were pooled until 1995 in Costa Rican statistics.Moreover, it seems advisable to interpret with caution the official statistics per country, because they are generally based upon the numbers provided by the fishing industry; there are no established procedures to validate the obtained data and to estimate unreported landings (FAO-OSPESCA, 2006).It might be possible that deep-water shrimps are fished in several countries of the region; however, official reports about these decapods are only available for Nicaragua and Costa Rica.The importance of deep-water shrimp fisheries for the region is difficult to assess; at least in Costa Rica the landings of the three commercially exploited deep-water shrimps (see below) comprised more than 50% of the annual shrimps landings between 1995 and 2005 (Wehrtmann & Nielsen-Muñoz, 2006;Álvarez & Ross-Salazar, 2010), but this scenario is most probably not representative for the entire Central American region.

II.2. Presentation of the species.
Three species are the target of commercial deep-water shrimp fisheries in Central America: the two pandalid species Heterocarpus vicarius and H. affinis, and the solenocerid shrimp Solenocera agassizii.
Heterocarpus vicarius Faxon, 1893 (northern nylon shrimp; local name: "camarón camello" or "camarón camellito") (Fig. 8).This pandalid shrimp is distributed between Gulf of California (Mexico) and Peru (Holthuis, 1980;Hendrickx et al., 1998).This species has been reported between 73 and 760 m, and was caught in traps at depths of up to 1454 m (Hendrickx 1995).According to Holthuis (1980) Wehrtmann ( 2011): ovigerous females measured between 29.4 and 45.4 mm CL and carried on average 15,008 newly-extruded eggs.During the incubation period, egg volume increased by 53.3%, while egg mortality during embryogenesis was estimated by 46.9%.Freshly-extruded eggs have an average volume of 0.045 mm 3 , those close to hatching 0.069 mm 3 , which corresponds to an egg volume increase during the incubation period of 53.3%.The average reproductive output was 0.178 and not statistically significant different with female size.Ovigerous females can be found year-round, but are more abundant between June-July and September-October (Wehrtmann & Nielsen-Muñoz, 2009).However, during their August-cruise in 1991, Hendrickx et al. (1998) collected only four ovigerous females of H. vicarius, and the size of these specimens ranged from 23.3 to 23.9 mm CL.

II.3. Fishing fleet. Deep-water shrimps are fished in
Central America with bottom trawling nets.The technical details regarding these shrimp trawlers may vary locally; however, the shrimpers typically use the "Florida type" boat with two lateral nets (Fig. 11).These vessels usually have a length of 16 to 24 m (Costa Rica: Álvarez & Ross-Salazar, 2010) and 18 to

II.4. Statistics.
Although deep-water shrimps might be mentioned somehow in national fishery reports, as far as we know, official data are only available from Nicaragua and Costa Rica.
Nicaragua.The commercial deep-water fishery is geographically concentrated in the Gulf of Fonseca (13°02'N, 87°48'W) and the Estero Real (12°57'N, 87°25'W), Pacific coast of Nicaragua.The fishery statistics of the Instituto Nicaragüense de la Pesca y Acuicultura (INPESCA) only mention deep-water shrimp as Heterocarpus spp., most probably referring to H. vicarius and H. affinis.According to these statistics (Fig. 12), highest landings were recorded in 1995 with 156,943 kg.However, since 2006 no more landings have been reported (INPESCA, 2008;OSPESCA, 2009).During 1993 and 2007, the number of operating deep-water shrimp trawlers varied between one and three (excluding those yearswhen no boats were fishing).The solenocerid shrimp Solenocera agassizii can be considered as the most important shrimp in the commercial deep-water shrimp fishery in Costa Rica (Wehrtmann & Nielsen-Muñoz, 2009).Annual landing of this species represented 27% (1995)(1996)(1997)(1998)(1999)(2000) and 29% (2001-2006) of the total shrimp landings in Costa Rica.By far, the highest annual landings (2.6 x 10 6 kg), were reported in 1986 (Fig. 14).The next but substantially lower peak (0.9 x 10 6 kg), was observed between 1994 -1995.Subsequently, catches dropped again and increased only after 2003, probably related to the fishing activities of The Rainbow Jewels S.A., a company, especially interested in this resource.The official statistics seem to indicate peak landings roughly every 10 years, however, always on lower levels.

II.5. Bycatch and discards.
As far as we know, Costa Rica is the only Central American country with information about the bycatch and discards associated with the deep-water shrimp fishery.In general, the entire bycatch is discarded, including squat lobsters (Pleuroncodes monodon), which are the target of a commercial fishery in other regions (Wehrtmann & Acuña, 2011).Several attempts to convince local fishermen and companies to utilize this resource in Costa Rica have not yet been successful.16).This finding may indicate significant changes in this fishery-based ecosystem, where trawling may have eliminated filter and detritus feeders, increasing scarcity of organisms feeding on phytoplankton and marine snow (see Pauly et al., 2009).It is speculated that such a condition may have favored the siphonophore outbreak observed along the Pacific coast of Costa Rica.The sudden massive appearance of "jellyfish" requires further attention.II.6.Conclusions.In Central America, deep-water shrimps are currently commercially exploited only in Nicaragua and Costa Rica.To the best of our knowledge, neither country has implemented a management plan for the deep-water resources.Due to the decreasing landings and increasing costs, more and more companies leave their trawlers in the harbor.This may favor the recovery of the stressed deepwater habitats, but will provoke also considerable social and economic consequences.The introduction of fishing restrictions and the implementation of a management plan, based on solid scientific data, seems to be indispensable to prevent the complete collapse of the shrimp resources (see Wehrtmann & Nielsen-Muñoz, 2009;Álvarez & Ross-Salazar, 2010).In this context, it is strongly recommended to improve collaboration between all relevant stakeholders (e.g.fishing sector, governmental sector, academic sector), thus facilitating information exchange regarding deep-water shrimp resources in order to support adequate decisions concerning fishing regulations at a national and regional level.

III. PERU
III.1.Introduction.Before the mid 1960s, it was assumed that the benthic and benthopelagic fauna off the Peruvian coast was poor, mostly probably due to the observed oxygen deficiency (Schweigger, 1943;Popovici, 1963), which supposedly prevented the presence of commercially interesting aggregations of groundfish.However, results published by Del Solar et al. (1965) revealed a rich biocenosis in deep-water environments off Peru, characterized by low oxygen contents with Peruvian hake (Merluccius gayi peruanus) as the characteristic and indicator species.Since the 1960s, a hake trawl fishery was developed, becoming one of the most important fisheries in Peru (Wosnitza-Mendo et al., 2007;Guevara-Carrasco & Lleonart, 2008).
An exploratory and experimental deep-water fishing using nontraditional gears and fishing methods was developed in 2001, revealing the presence of king crabs (Lithodidae) off Callao (12°S).These crabs were considered as a potential and alternative resource for a nontraditional trap fishery (Salazar et al., 2001) The Patagonian toothfish (Dissostichus eleginoides) is the target of another deep-water fishery in Peru carried out by a small fleet (5-6 boats) using longlines.This species is included in the Peruvian landing statistics since 1995 (Sancho, 2002).Fishing takes place along the Peruvian coast, from Salaverry (08°S) to Ilo (18°S), and at depths between 1000 and 1500 m.In recent years (2005 to 2010) the average yearly landing was of 130.3 ton (V.Aramayo, IMARPE, pers.comm.).
Since the middle 1960's, the Peruvian Marine Research Institute (IMARPE) has carried out research cruises along the Peruvian coast to investigate the archibenthic and bathyal fauna.The results of these cruises revealed the presence of more than 250 species and contributed significantly to our knowledge on the deep-water fauna (Del Solar & Álamo, 1970;Del Solar & Mistakidis, 1971;Vílchez et al., 1971;Del Solar & Flores, 1972;Kameya et al., 1997;JDSTA-IMARPE, 2004;Barriga et al., 2009).The following species of deep-water shrimps stand out due to their wide distribution and abundance in Peruvian waters: Haliporoides diomedeae, Benthesicymus tanneri, Nematocarcinus agassizii, Heterocarpus vicarius, H. hostilis, and Pasiphaea magna.Several studies were carried out to describe the biology and population dynamic of these decapods in order to generate enough information to promote sustainable fisheries targeting these potential resources (Méndez, 1981;Kameya et al., 1997;Barriga et al., 2009).
More recently (between 2007 and 2010), four research cruises were carried out by IMARPE, with the Spanish research vessel "Miguel Oliver", to characterize the principal faunal components of the bentho-demersal subsystem of the bathyal and archibenthic zones (between 200 to 1500 m depth) off the Peruvian coast from Puerto Pizarro (03°30'S) to Morro Sama (18°00'S).A bottom trawl (randomly stratified sampling) and the swept area method were applied to determine the distribution and relative abundance of the fauna (Alverson & Pereyra, 1969;Espino & Wosnitza-Mendo, 1984;Sparre & Venema, 1995).The total area was divided in eight zones (two zones per cruise) and three depth strata (200 to 500 m, 500 to 1000 m, and 1000 to 1500 m depth).A total of 272 hauls (30 min of effective trawling) were carried out with a mean speed of approximately three knots using a bottom net type LOFOTEN (for details see Barriga et al., 2009).The following sections synthesize the principal results concerning the distribution and abundance of deep-water shrimps in Peruvian waters.
The bathymetric distribution of H. diomedeae changes with latitude, with a distinct deepening from south to north.The presence of the species, at large depths, may be related to the physical characteristics of the Antarctic Intermediate Water (AAIW), where the temperature oscillates between 7.5°C and 3.5°C.It is assumed that a reverse relationship exists between the species abundance and the intensification of these water mass in the southeastern Pacific (Noziglia & Arana, 1976).
All investigations have indicated H. diomedeae as the most important deep-water shrimp in Peru due to several reasons: (1) its wide distribution and abun-   Vélez et al., 1992;Kameya et al., 2006).In Peru, H. diomedeae is widely distributed between 600 and more than 1500 m deep.Three areas of high abundance were detected: (1) north of 08°S (Malabrigo), where the highest biomass was observed in concentrations around 50 and more than 150 kg km -2 , (mean about 139 kg km -2 , especially in hauls deeper than 600 m); (2) the area between 08°S (Malabrigo) and 16°S (San Juan de Marcona); in this zone, between 08°S and 09°S, the species was present in low abundances (16 kg km -2 ) and almost absent in the remaining area; (3) south of 16°S, where H. diomedeae was found to be present in low abundances (mean biomass was 8.7 kg km -2 ), mainly between 700 y 1200 m depth (Fig. 19).
Benthesicymus tanneri Faxon, 1893 (red royal shrimp; local name: "gamba" in Chile, "langostino rojo de profundidad" or "gamba roja" in Peru).Méndez (1981) described morphological characteristics of this species from Peruvian material.B. tanneri is known to occur in Gulf of California, Panama, Colombia, Galapagos Islands, Ecuador and Peru.This species is soft and fragile, and rarely arrives at the surface intact.So, it would be a poor candidate for a fishery.Male and female shrimps collected on the continental slope off Peru ranged in size between 13.5-25.0mm and 11.0-30.7 mm CL, respectively (Méndez, 1981).The results of our recent studies (2007-2010) confirm a wide distribution of B. tanneri in Peruvian waters between 600 and 1500 m depth; mean biomass was estimated to be 9.2 kg km -2 , and the areas of highest concentrations were located north of 10°S (Huarmey) (Fig. 19).Nematocarcinus agassizii Faxon, 1893 (local name: "camarón rojo de profundidad").This deep-water shrimp belongs to the family Nematocarcinidae and it is adapted to walk on the soft bottom seafloor covering the continental slope (Méndez, 1981).Morphological characteristics of Peruvian specimens were described by Méndez (1981), and the size of males and females collected by IMARPE varied between 17.2-20.2mm CL and 17.4-26.0mm CL, respectively.
Scientific cruises carried out between 2007 to 2010 revealed the presence of N. agassizii in the northern zone (Fig. 19), with highest abundances off Tumbes (until 04°10'S) between 500 and 800 m depth: the mean biomass was around 12 kg km -2 , ranging from 0.5 to 45 kg km -2 .The southern boundary was observed to be located off Chimbote (11°S).A few individuals were observed also off Ilo (17°30'S), thus expanding the geographic distribution reported by Méndez (1981); she indicated the limit at 16°30'S based on observations published by Del Solar & Flores, 1972).
Heterocarpus vicarius Faxon, 1893 (northern nylon shrimp; local name in Peru: "camaron rojo de profundidad" or "camaron nylon) (Fig. 8).The geographic distribution of this pandalid shrimp extends from the Gulf of California to southern Peru (Méndez, 1981).In Peru, this benthic shrimp inhabits the continental slope between Banco de Mancora (03°24'S) and Mollendo (17°S); its vertical distribution ranges from 383 to 800 m (Méndez, 1981).Results of our scientific cruises carried out between 2007 and 2010 indicated highest biomass in areas north of Paita (05°00'S) at depths below 500 m, where the biomass was generally higher than 50 kg km -2 with a mean of 125 kg km -2 .South of Paita the species was less abundant (<0.5 kg km -2 ), and few individuals were collected off Ilo (17°37'S), probably indicating the southern border of its distribution in Peruvian waters (Fig. 19).
Heterocarpus hostilis Faxon, 1893 (local name: "camarón nailon").This pandalid shrimp is distributed from Gulf of Panama to Isla del Coco in Peru.Del Solar (1972) reported the species in Peru from 3°30'S to 10°5'S, occurring at depths between 695 and 800 m.Morphological characteristics of H. hostilis in Peru are described by Méndez (1981).The results of the scientific cruises carried out between 2007 and 2010 (Fig. 19) showed that the species is common and abundant (between 0.5 and 50 kg km -2 ) north of 11°00'S (Huacho), mainly concentrated in depths over 1000 m.
Pasiphaea magna Faxon, 1893 (local name in Peru: "camarón vidrio").Morphology, distribution and commercial importance in Peru were described by Méndez (1981).This shrimp is a pelagic species, which occurs as "accidental catch" in benthic trawls.The species is distributed along the Peruvian coast, inhabiting bathyal and archibenthic zones between 600 m and 1500 m.According to the results of the scientific cruises (2007)(2008)(2009)(2010), this shrimp is not very abundant (mean biomass was 1.3 kg km -2 ), with the exception of an area off Tumbes (north of 04°00'S), where biomass values reached up to 50 kg km -2 (Fig. 19).

III. 3. Fisheries, actual status of resources and perspectives.
So far, no commercial deep-water shrimp fishery has been implemented in Peru.Therefore, the six potentially exploitable resources described above are virtually in pristine conditions, not subject to fishing pressure.Small amounts of these shrimps are accidentally caught as by-catch by industrial trawlers operating in northern Peru.Future strategies to begin any commercial exploitation of these species should consider the ecosystem approach, especially in northern Peru where the highest abundances and the greatest biological diversity of the bentho-demersal subsystem in Peruvian waters has been found.Since the mid-1900s, and with the arrival of larger and more powerful trawlers, offshore fishing operations started to be conducted along the Chilean coast.At the same time, the exploitation of the nylon shrimp (Heterocarpus reedi) began, achieving a record of 72,194 ton of landed shrimps in 1976.Nylon shrimp landings subsequently and progressively decreased, which promoted the search of new resources by researchers and companies, especially towards deeper waters.Besides the nylon shrimp, two other shrimp species with potential for a commercial exploitation were found in deep waters: the red royal shrimp (Haliporoides diomedeae) and the razor shrimp (Campylonotus semistriatus).These three species have been intensively studied, and the present review summarizes the obtained results.

IV.2.1. Distribution. The geographical distribution of H. reedi comprises the area off central-northern to central-southern Chile, between 23°48'S (south of Antofagasta) and 39°00'S (Puerto Saavedra).
Moreover, the presence of H. reedi has also been reported offshore Peru (Vílchez et al., 1971).In Chile, the spatial coverage between Antofagasta (23º30'S) and Coquimbo (30º30'S) is almost continuous.In contrast, the presence of the resource is scattered in the southern part of its distribution, forming small patches towards the north and south of the Río Maipo Canyon in the O'Higgins Region (33º30'S) and off Punta Nugurne in the Maule Region (35º55'S).Therefore, the fishing fleet has been operating mainly  The nylon shrimp lives in the continental slope at depths ranging between 155 and 800 m (Yáñez, 1973(Yáñez, , 1974;;Yáñez & Barbieri, 1974).Heterocarpus reedi occurs in an area where Subsurface Equatorial Water (ESSW) and Antarctic Intermediate Water (AAIW) mix between 200 and 500 m.According to Arana & Nakanishi (1971) and Arana et al. (1975), the depth changes of these water masses and the upwelling processes observed alongshore the Chilean coast promote the depth-related movement of this resource perpendicular to the coast.As suggested by the same authors, ovigerous females move towards shallower waters to release their larvae and later return to deeper waters.However, the nylon shrimp fisheries operate mainly at depths between the 150 and 500 m.

IV.2.2. Habitat.
This shrimp is generally found on mud bottoms, sedimentary rock, muddy sands and sandstone.According to Acuña et al. (2009), 70% of the observations correspond to catches over fine sand and 25.7% over medium sand grain.

IV.2.3. Life history parameters.
The species presents sexual dimorphism, which is externally not evident.
Females grow larger than males; maximum sizes correspond to 43 and 40 mm CL for females and males, respectively (Arana et al., 1976;Roa & Ernst, 1996).
According to Arana et al. (1976), the principal molting periods include March-May, July-September and November-January, a feature that may be associated to variations of the condition factor (Yany, 1974) and the chemical composition of the abdominal muscle.The molting increase varies between 9.7-10.6%and 7.2-8.9% in females and males, respectively (Arana et al., 1976).
Growth of H. reedi has been studied by Ziller (1993), who established the following growth parameters: L ∞ =37.67 mm CL; k=0.343 y t o =-0.117 in males and L ∞ =43.87; k=0.305 and t o =-0.17 in females.Roa & Ernst (1996) pointed out that the exploited fraction of the H. reedi population is comprised of five different annual classes, with clearly distinctive growth parameters between sexes, and values of L ∞ =40.68; k=0.199 and t o =-0.809 for males and L ∞ =48.34 mm CL; k=0.174 and t o =-0.51 for females.

IV.2.4. Fisheries gear.
Polyamide bottom trawling net with metallic gangways and 5 cm uniform meshes (2 inches), similar to those used in Chilean knife shrimp (Haliporoides diomedeae) fisheries are employed in the commercial fisheries of H. reedi.
Selectivity studies of trawling nets (considering the mean size at first maturity of 25 mm CL in females) revealed that mesh size should be at least 59 mm (2.5") to protect 50% of the females; such mesh size would slightly "overprotect" males as they show smaller sizes and would probably mature earlier (ECM, 2000a).

IV.2.5. Fishery statistics and management.
The nylon shrimp sustains the main shrimp fisheries in Chile, regularly exploited by industrial and artisanal fishery fleets.The commercial exploitation of this resource started in the 1950s as by-catch of the common hake fishery (Merluccius gayi gayi); subsequently it became a significant target resource as landing exceeded the 11,000 ton in 1968.Thereafter, a progressive decrease in the landing levels was observed, reaching 2,700 ton in 1980, to remain stable for a period of seven years around 3,000 ton.During the next years, a sustained increase in landing was observed until 1995-1996, when they once again exceeded 10,000 ton yr -1 (Fig. 22).
By late 1994, the nylon shrimp fisheries were subjected to a general access regime (i.e.free access).In 1995, it was declared as in a Full Exploitation State and regime.At the same time the fishing unit of the nylon shrimp (H.reedi) was defined as the fishing area covering the coast between the northern limit of Antofagasta (21º25'S) and the southern limit of Bio-Bio (38º28'S), from the eastern limit five miles offshore to the western limit corresponding to an imaginary line drawn at a 60-mile distance (Decree MINECOM Nº611 of 1995).This measure implicated the establishment of a global annual fishing quota, and the closure of the fisheries register for new applications for vessels and fishermen.Additionally, a biologically-based seasonal closure was established in the entire fishing area, from 01 July to 31 August of each calendar year.During the closed season it was forbidden to catch, market, transport, process and store the banned species and its derived products (SUBPESCA Decree Nº92 of 1998).
Subsequently in 2001, the fisheries were subjected to the Maximum Catch Limit per ship-owner, which involved the annual distribution of the global annual catch quota allocating 80% to the industrial sector (divided among ship-owners with valid fishing permits to execute extractive activities) and the remaining 20% to the artisanal sector.In addition to the latter measures, since 2003 the global fishing quota was divided per administrative region in order to avoid the concentration of fishing efforts and adapt the amount to be extracted according to the resource availability.Likewise, and aiming to achieve a yearround exploitation, the quota assigned to each region was subdivided into three periods: January-March, April-August and September-December.
A total of 26 industrial trawlers (Fig. 23) and five artisanal boats have been traditionally operating for this fishery, which executed nearly 2,500 hauls per year, with a mean duration of 2.5 h each.The yield has reached values close to 300 kg of nylon shrimp per trawling hour (t.h.) in the northern macro-zone, over 400 kg t.h -1 in the southern macro-zone with an average of nearly 350 kg t.h -1 (Zilleruelo & Párraga, 2009).
During the last ten years, a gradual change of the available biomass of the resource has been observed: from 10,000 ton in 1998 to a reported biomass oscillating between 28,000 and 30,000 ton in 2008 (Table 3).Consequently, Canales & Arana (2010) demonstrated that the catch per unit area (CPUA) increased from 532 kg km -2 in 1998 to more than 4,000 kg km -2 in 2006.This shows that the implementation of low catch quotas for this fishery has achieved a gradual recovery of the species, equivalent to an exploitation rate of 15% in the last year (Acuña et al., 2009).The implementation of global catch quotas below 5,000 ton would maintain the status quo situation for the exploitation of H. reedi, contributing to the maintenance and recuperation and thus avoiding overexploitation of the resource.
IV.2.7.Commercialization.Nylon shrimp catches are used for different marketing processes.The main process is freezing (peeled-tails, IQF), followed by fresh-chilled and dehydatation.Production excesses (carapace, appendix and exoskeletons) are used in flour production.Top-quality product volumes in the industry have remained between 700-850 ton yr -1 during the last years.The most significant markets for frozen products are Germany (44%), Japan (24%), Taiwan (15%), Denmark (8%) and UK (5%); the remaining 4% goes to different locations.Local markets offer fresh unpeeled shrimp tails.

IV.2.8. Conclusions.
Although a significant recovery of the nylon shrimp biomass has been observed, during the past 10 years the stock still has not consolidated at catch levels over 5,000 ton yr -1 .A better chance of recovery may be achieved with annual catches lower then the aforementioned, thus allowing an increase in the current biomass and achieve exploitation rates of the order of 10%.(Faxon, 1893) (Fig. 17)

IV.3.1. Distribution and life history charcteristics in Chile.
Haliporoides diomedeae is distributed through a continuous "band" in the central area of Chile, occasionally interrupted by topographic features (San Antonio submarine canyon, 33º31'S) or non suitable areas for trawling such as Punta Duao (34º48'S) and .The species can be found between 400-1000 m off central Chile (Arana et al., 2003b) and in the southernmost area between 500-900 m (Leiva et al., 1997).Red royal shrimp concentrations during exploratory fishing were found to be associated with the approximate depth location of the 5°C isotherm (Noziglia & Arana, 1976).The species lives semiburied in soft bottoms and associated with different decapods and fish species (Zilleruelo & Párranga, 2009).According to data obtained from the Chilean fisheries, H. diomedeae represents 52% of the weight catch in commercial hauls (Zilleruelo & Párraga, 2009).
The size of physical maturity of H. diomedeae in Chile is reached between 30.0 and 32.5 mm CL in males and 36.5 and 38.0 mm CL in females (Noziglia & Arana, 1976;Arana et al., 2003b).Males should reach maturity in the second year and females in the third year (Noziglia & Arana, 1976).Females grow faster (b=2.304)than males (b=1.994), which means that females are heavier than males for a given size (Arana & Cristi, 1971).
Both sexes show polymodal size frequency distributions, males with the most significant mode between 25 and 35 m CL and females between 30 and 50 mm CL (Arana et al., 2003b).The sexual proportion of the species reveals a clear female predominance in the catches (60%); small size-classes (<25 mm CL) are dominated (60-100%) by males, while in the larger size groups females comprise more than 80% (Gaete & Arana, 1986;Arana et al., 2003b).

IV.3.2. Fishing gear.
The nets used in the H. diomedeae fisheries are similar to those previously described for the nylon shrimp (Heterocarpus reedi) fisheries.However, the H. diomedeae fishery acquired more importance within the demersal crustacean fisheries off central Chile only since 1999 when the trawling fleet started to land volumes of certain commercial importance.Since 2001, the H. diomedeae fisheries represent an alternative in view of the restrictive measures implemented for other crustacean species commercially exploited off Chile.This allowed a gradual increase of the landings, reaching 705 ton in 2003, the highest annual value for H. diomedeae so far reported for Chile.Thereafter, catches have decreased, with landings at levels between 90 and 215 ton yr -1 (Fig. 24).The latitudinal distribution of the resource may be compared to that of H. reedi, although at a higher depth.In Chile, for administrative purposes, red royal shrimp fisheries cover the maritime area between 18º21'S and 43º44'S.However, commercial fisheries of the species are executed principally between 32ºS and 36ºS as accompanying fauna of the nylon shrimp fisheries; occasionally hauls are carried out to specifically catch H. diomedeae.Through the analysis of 6,043 hauls of the industrial and artisanal fleet, Párraga et al. (2010) were able to document that higher catch rates (37 and 62 kg per trawling hour) were obtained between 33º53'S and 34º41'S at depths exceeding 500 m; trawling was more successful between November and March (53 kg per trawling hour) than between April and October (42 kg per trawling hour).

Table 3. Main operational variables of the direct assessment cruises for nylon shrimp (Heterocarpus
IV.3.4.Current status of the resource.The red royal shrimp fisheries are managed under the "General Access Regime" (Ley General de Pesca y Acuicultura, Chile).Its commercial exploitation is restricted to catches of the trawling fleet dedicated to fish and other benthic crustaceans.

IV.3.5. Commercialization.
The species shows a fast post-catch deterioration caused by melanosis (blackening).The species undergoes different production processes to be commercialized as raw, cooked and frozen (peeled tails, IQF) products, and deveined or undeveined (stomach).Local markets offer fresh tails.
IV.3.6.Conclusions.Fishing performance for the red royal shrimp is lower than those reported for other crustacean species currently exploited in Chilean waters.According to the estimated vulnerable biomass levels (Arana et al., 2002(Arana et al., , 2003b)), the resource does not allow the development of large-scale extraction activities; therefore, it is more likely that fishing of H. diomedeae act as a complementary activity to the exploitation of other resources such as H. reedi, P. monodon and C. johni.Notwithstanding, the red royal shrimp fisheries represent a relevant economic activity and an important alternative if other crustacean fisheries will be closed once the established quotas are achieved or biological or extractive close seasons are put into force.
IV.4.2.Habitat.This shrimp is found off the central Chilean coast associated with the Antarctic Intermediate Water mass (AAIW), located between 400-1,000 m of depth.This water mass is characterized by low temperatures (4º to 7ºC), along with minimum salinities (>34.4),high oxygen content (>2.5 mL L -1 ) and relative low in nutrient contents (Arana et al., 1976).The razor shrimp inhabits mainly muddy sand bottoms (Arana et al., 1976).

IV.4.3. Life history parameters.
In general, our knowledge about life history features of representatives of the Campylonotidae is limited.They seem to have a short larval development, undergoing at least two zoeal stages and a decapodid stage (Thatje et al., 2001).Descriptions of the first larval stages and a key for their identification have been provided by Thatje et al. (2001).In Argentinean waters, C. semistriatus is the largest species among the three congeneric species (C.vagans, C. capensis, C. semistriatus) found in the southern Atlantic, with sizes between 17.0 and 35.5 mm CL (Torti & Boschi, 1973).In commercial hauls carried out off central Chile, the size of C. semistriatus varied between 11 and 37 mm CL; mean lengths in males was 19.8 mm and 25.9 mm CL in females, indicating a clear size difference between sexes (Arana et al., 2002(Arana et al., , 2003b)).Arana & Ahumada (2006) emphasized the possible existence of a relationship between the mean size of individuals and the depth of catch, meaning that size increases with depth, so that the largest mean sizes can be found between 700 and 800 m.
Females are more abundant than males (malefemale proportion of 1:2.3) in the catches (Arana & Ahumada, 2006).Small individuals are generally males, while females are predominant in specimens > 25 m CL; the sex ratio is 1:1 in individuals measuring around 21 mm CL (Arana & Ahumada, 2006).These size and sex patterns have led to the conclusion of a possible consecutive protandric hermaphroditism in the species (Torti & Boschi, 1973) from Argentinean waters, with intersex individuals between 22.3 and 25.5 mm CL.This trait, however, is yet to be confirmed for C. semistriatus inhabiting Chilean waters.
Females reach sexual maturity between 21.1 and 23.4 mm CL (Arana et al., 2003b).The parameters of the size-weight relationship indicate a negative allometric growth for this shrimp with b values between 2.57 and 2.66 for males and 2.59 and 2.67 for females (Arana & Ahumada, 2006).
IV.4.4.Fishing gear.The nets used in the C. semistriatus fisheries are similar to those already described for the nylon shrimp (H.reedi) and the red royal shrimp (H.diomedeae) fisheries.
IV.4.5.Fishery history.Since this shrimp lives below the depth range where the industrial fleet operates, the available information about this resource corresponds mainly to the results obtained from exploratory or stock assessment studies (Arana et al., 2002(Arana et al., , 2003b;;Arana & Ahumada, 2006).Currently, C. semistriatus is extracted as part of the by-catch of the nylon shrimp fishery operating in the deepest distribution stratum of H. reedi or, more commonly, associated with the H. diomedeae fishery, carried out in depth ranges similar to those established for the razor shrimp.
IV.4.6.Current status of the resource.The resource is only caught as bycatch of H. reedi and H. diomedeae fisheries with only a few ton per year.Currently, its exploitation is not regulated nor subordinated to any extraction regime.
IV.4.8.Conclusions.The offshore waters of northern and central Chile have been for several decades the area of an active crustacean fishery carried out by the commercial trawling fleet.These fisheries are mainly targeting the nylon shrimp (H.reedi), the yellow squat lobster (Cervimunida johni), and the red squat lobster (Pleuroncodes monodon), representing the raw material for high added-value products.Due to the reduction of the fishing quotas in the aforementioned species, despite its relatively low abundance and the great depth where the larger individuals occur, the razor shrimp (C.semistriatus) may become a resource to be extracted as a complementary species to the previously mentioned species (Arana & Ahumada, 2006).

IV.5. General conclusions.
The pandalid H. reedi is the main shrimp species exploited in Chilean waters, and has supported a relevant fishery due to its available biomass.This scenario has promoted the development of significant fishing activities, which involve an important industrial fleet exploiting this resource.Moreover, the nylon shrimp fishery maintains several companies dedicated to the processing and commercialization of these products for the international markets, which, in turn, generated numerous employment opportunities.
The historical fluctuations observed in H. reedi landings require especial attention when implementing management measures; the resource needs to be monitored periodically in order to analyze its condition, to assess accordingly annual global catch quotas and to promote administrative measures to facilitate a sustainable exploitation of this shrimp.
In the case of the red royal shrimp (H.diomedeae) and the razor shrimp (C.semistriatus), assessments (Arana et al., 2003b;Arana & Ahumada, 2006) have shown considerably lower biomasses compared to those reported for the nylon shrimp (H.reedi); moreover, the distribution of these species seems to be patchy and confined to few specific areas.Consequently, the exploitation of these (by-catch) species must be carefully observed to avoid overfishing and to permit a sustainable use of these deep-water shrimp resources.
Essentially artisanal in its beginning, the activity rapidly diversified from mid 1960's to early 1970's, as a response to government efforts to develop industrial fleets and facilities in Brazil.While well succeeded in adding a new scale to the national shrimp fisheries, building up an excessive fishing capacity coupled with management weakness lead to an overexploitation of most Brazilian shrimp stocks in a few years (Valentini & Pezzuto, 2006).
This scenario remained quite unchanged up to late 1990's, when local industrial trawler fleets expanded their operations to deeper areas of the Exclusive Economic Zone (EEZ), in search of new and profitable resources which could compensate economic losses faced in the traditional coastal fishing grounds (Perez et al., 2001).Concurrently, a new governmental policy trying to develop deep-water fisheries in Brazil has started, inaugurating a new chapter in its fishery history.This policy was based upon implementing a program of specialized foreign vessels chartered by national companies, which lasted for nearly 11 years.Both processes, mediated by learning and technological adaptations of the local fleets, have resulted in the establishment of new foreign and domestic fisheries, including bottom gillnetting for the monkfish (Lophius gastrophisus), trawling for Argentine hake (Merluccius hubbsi), Brazilian codling (Urophycis mystacea) and Argentine squid (Illex argentinus) and potting for geryonid crabs (Chaceon spp.) (see review in Perez et al., 2009a).
After some individuals had been caught incidentally in the year 2000, a deep-water shrimp fishery also emerged in 2003, when two foreign vessels started to target three valuable species pertaining to the family Aristeidae (see Pezzuto et al., 2006): the scarlet shrimp (locally called "carabinero"), Aristaeopsis edwardsiana (Johnson, 1867) (Fig. 2); the giant red shrimp (locally called "moruno"), Aristaeomorpha foliacea (Risso, 1827) (Fig. 1), and the purplehead gamba prawn (locally called "alistado" shrimp) Aristeus antillensis A. Milne Edwards & Bouvier, 1909 (Fig. 4).A brief summary about the distribution and ecology of these three species can be found in the present review, in the chapter about Mexico.This episode inaugurated a new Aristeidae fishery in Latin America.However, "carabinero" and "alistado" shrimps have long been seasonally exploited in French Guiana by local penaeid trawlers (Guéguen, 1998), while the "moruno" shrimp is known to sustain several well-established fisheries in Mediterranean waters (Sardà, 2000).
V.2.Fishing fleets.The fishery developed along most of the Brazilian EEZ; however, the fishing activities were concentrated in 11 slope-fishing grounds situated between Espírito Santo (18 o S) and Rio Grande do Sul (32 o S) states (Fig. 26).Catches were obtained from a very narrow bathymetric stratum between 700 and 750 m depth (Dallagnollo et al., 2009a); this situation imposed severe technological constraints to the domestic trawlers accustomed to operate in areas shallower than 500 m at most.
Consequently, deep-water shrimp fishery in Brazil was dominated by eight foreign stern trawlers, chartered by national companies, and an additional nationalized vessel with similar characteristics.Most of them between 30 and 40 m length, powered by approximately 1,000 HP engines, both equipped with processing facilities and freezing chambers (Fig. 27).These trawlers were capable to perform fishing trips lasting over one month.A variety of bottom trawl nets were employed by the fleet, which differed mostly in head rope length (45 to 95 m) and mesh size (40 to 80 mm).Trawling velocity ranged between 2 and 3 knots, and towing lasted on average 4 h.Each trawl swept a linear extension of approximately 21 km (maximum 73 km) (Dallagnolo, 2008;Dallagnolo et al., 2009a).4).While generally playing a secondary role in the fishery (20.6% of the total), Aristaeomorpha foliacea dominated in the fishing grounds N1 and C4 (Fig. 26), surpassing the production of A. edwardsiana (Dallagnolo et al., 2009a).Landings of this species increased continuously up to 2006, when a maximum of 51.7 ton was attained.Aristeus antillensis accounted only for 3% of the deep-water landings in Brazil.Its production increased from 0.3 ton in 2002 to 15.8 ton in 2005, but declined in the subsequent years (Table 1).Profitable catches of this species were obtained only in two small fishing grounds situated at nearly 19.5ºS (Pezzuto & Dias, 2009).
As compared to the other species, A. edwardsiana produced the highest mean annual catch rates (4.7 to 14 kg h -1 ).However, they revealed a declining pattern in time, reducing between 26% and 61% depending on the fishing area.On the other hand, and in spite of being generally lower (0.76 to 6.3 kg h -1 ), mean catch rates of A. foliacea increased substantially at rates varying between 327% and 835%.Based on biological data published by Pezzuto & Dias (2007), Dallagnolo et al. (2009a) hypothesized that the inversion of the mean catch rates in A. edwardsiana and A. foliacea may be related to their different life strategies: the former species is larger and mature later in life, while the latter is smaller and presumably faster-growing.This difference may confer an advantage to A. foliacea in expanding its distribution and abundance after the massive removal of A. edwardsiana from the fishing grounds.
Aristaeus antillensis produced the lowest catch rates among the Brazilian deep-water shrimps (0.005 to 2.4 kg h -1 ), which decreased by 65% between 200465% between and 200765% between (Dallagnollo et al., 2009a)).Such changes in commercial catch rates were confirmed by biomass assessments (Dallagnolo, 2008).As a conse-   (Dallagnolo et al., 2009b).However, this plan was never implemented due to unclear reasons (see details of the Brazilian fishery institutional framework in Perez et al., 2009a).The plan included: limits to the fleet size (two vessels); fishing area (18 o 20'S to 28 o 30'S, 500-1000 m depth); a global total allowable catch (TAC) for the three species together (60 ton year -1 ): gear restrictions (stern trawl method, minimum cod-end mesh size 60 mm stretched): by-catch limits (Chaceon spp.15% and L. gastrophysus 5% of the total catch); exclusion areas (a quadrant between Rio de Janeiro and São Paulo States, seamounts, coral bottoms and areas "N" of 21 o S when catches of A. antillensis attain 4.4 ton); rotation of harvestable areas ("N" and "S" from the main exclusion quadrant); and control measures as logbooks, satellite tracking and observers in all trips.
V.5.Conclusions.The commercial exploitation of aristeid shrimp resources in Brazil developed as a typical "gold rush" fishery.The production peaked only four years after the start of directed foreign fleet operations and stopped four years later.In spite of the governmental decision to implement a mandatory and detailed observer program, coupled with scientific efforts, aiming at obtaining and processing timely information about the species, management measures were not enforced throughout the period.Overexploitation of the most abundant and valuable species (A.edwardsiana) was perceived in a few years (Dallagnolo, 2008).As a consequence, and even considering the concurrent increase in catch rates and yields of A. foliacea, this fishery turned economically unprofitable and was abandoned.Contrary to the domestic gillnetting for monkfish and trawling for Argentine hake, Brazilian codling and Argentine squid fisheries in depths <500 m, which were made possible by minor modifications of existing vessels, technological and operational limitations impeded the development of a national fishery for aristeid shrimps in deeper areas of the Brazilian slope.While such restrictions would have contributed to alleviate the fishing pressure over the deep-water shrimp stocks, during the last years, governmental resistance to implement the respective management plan has placed the resource under potential risk.

VI. OTHER COUNTRIES
Deep-water shrimps are commercially exploited along the Pacific coast of Colombia.The term "deep-water" in Colombian fisheries refers to depths below 72 m (Puentes et al., 2007) or below 36 m (INVEMAR, 2009).According to INVEMAR (2009), commercially exploited deep-water shrimps include Solenocera agassizii as well as Farfantepenaeus brevirostris and F. californiensis (in the present review, we do not consider these two penaeid shrimps as deep-water species).Puentes et al. (2007) mentioned additionally Heterocarpus spp.(according to the authors H. vicarius and/or H. hostilis); however, these pandalid shrimps are not the target species and are obtained as by-catch associated to the fishery of the abovementioned three shrimp species.Puentes et al. (1994) and Rueda et al. (2004) provided a description of the shrimp trawler fleet operating in Pacific Colombia.In 1993, the deep-water shrimp trawler fleet consisted of 23 boats, while in 2003 17 boats were used exclusively to extract deep-water shrimps, and an additional seven trawlers were allowed to fish in both shallowand deep-water grounds.The trawlers used for the deep-water shrimp fisheries have a length varying between 15.0 and 24.4 m, and most of them measure between 20-22 m (Rueda et al., 2004).Landings of the three above-mentioned deep-water shrimps varied between roughly 575 ton (1993) and 1240 ton (2000 and 2006); the most recent data ( 2008) indicate landings of 753 ton, comprised mainly (55.2%) of S. agassizii (INVEMAR, 2009).In the specific case of S. agassizii, reported landings peaked in 2003 with approximately 688 ton (Fig. 28); since 2006, catches tend to continuously decrease, reaching 118 ton in 2009 (Barreto et al., 2010).As mentioned by Barreto & Bordo (2008), 69% of the fished population of S. agassizii consists of immature individuals; these authors characterize this deep-water resource as fully exploited.The study of Puentes et al. (2007) revealed the presence of 54 fish species belonging to 35 families, 11 crustacean species within 11 families, and one mollusk species, as by-catch of the deep-water shrimp fishery in Colombia; all crustaceans, including species of potential commercial importance (Heterocarpus spp.and Pleuroncodes monodon) are considered as discards.Several authors (Pedraza et al., 2002;Rubio et al., 2005) indicated that a trap fishery for H. hostilis exists, operating between 600 and 900 m depth.As far as we know, no official statistics about this trap fishery are available, and this fishery has been abandoned (V.Puentes, Ministerio de Ambiente, Vivienda y Desarrollo Territorial Colombia, pers.comm.)The incorporation of turtle excluder devices (TED), a fishing ban between January and February, and global catch quota ((for 2008: 930 ton for the three above-mentioned deep-water shrimp species (Barreto & Borda, 2008)) are the management strategies implemented by the Colombian authorities (Ruedo et al., 2004;Barreto & Borda, 2008;Barreto et al., 2010;Campo-Soto, 2010).species occurred predominantly at a depth between 600 and 700 m (Guéguen, 1998).Statistics concerning the annual catches of A. edwardsiana are available only until 1993 (see Guéguen, 1998): reported landings started in 1988 with 93 ton, decreased continuously to 34 ton in 1990, and reached its highest value in 1991 with 258 ton; thereafter, landings decreased again to 140 ton in 1993.According to Guéguen (1998), the CPUE of this fishery fluctuated between 0.341 ton h -1 (1988) and 0.175 ton h -1 (1990).It remains unclear whether this deep-water shrimp fishery in French Guiana is still active or has been abandoned.

VII. CONCLUSIONS AND RECOMMENDATIONS
Commercial deep-water fisheries are currently carried out in Chile, Colombia, and Costa Rica.Our compilation revealed the presence of a total of 17 deep-water shrimp species with different levels of potential for a commercial exploitation from Latin American offshore waters, including ten penaeoid and seven caridean species (Table 5).Currently, four species (Haliporoides diomedeae, Solenocera agassizii, Heterocarpus affinis, H. reedi and H. vicarius) are the target of a commercial fishery, and three other species (Aristaeomorpha foliacea, Aristaeopsis edwardsiana, and Aristeus antillensis) have been commercially exploited in Brazil, but these fisheries are now closed.Aristaeopsis edwardsiana have been reported also to support a seasonal fishery in French Guiana since 1988, but there are no data available about the current status of this fishery (Guéguen, 1998).While S. agassizii and H. vicarius are fished in Costa Rica and Colombia, the three aristeid species have been reported from both Brazil and Mexico, but they were exploited exclusively in Brazil.
To the best of our knowledge, an implemented management plan for the exploitation of deep-water shrimps only exists in Chile.However, in Colombia fishing closure (January and February) and annual catch quota were put into force to protect the deepwater shrimp resource (Barreto & Borda, 2008;Barreto et al., 2010;Campo Soto, 2010).In Brazil, detailed fishery regulations for the aristeid fishery were approved but never implemented.As a consequence, the main target-species (A.Edwardsiana) became rapidly overexploited, and the fisheries have been abandoned.In the remaining countries studied in the present review, there are no special management strategies directed towards the sustainability of these valuable deep-water resources.
Table 5.General information and current fishery status of the main deep-water shrimp species around Latin America."Geographic distribution" indicates the known geographic distribution of the species in Latin America, and "Country" refers to the country where a commercial fishery exists or not; the term "unknown" means that it is unclear if this fishery is still active.References concerning bathymetric and geographic distribution of the species are mentioned in the text; otherwise they are indicated at the end of the table.Tabla 5. Información general y situación de las principales especies de camarones alrededor de América Latina.La "Distribución geográfica" indica el área conocida en que se distribuye la especie en América Latina, y "País" se refiere al país donde se desarrolla o no la pesquería comercial; el término "desconocido" significa que no está claro si la pesquería es todavía activa.La referencia respecto a la batimetría y distribución geográfica de las especies está mencionada en el texto; en caso contrario, está señalada al final de la tabla.

Figure 5 .
Figure 5. Species compositions (percentages) by weight of deepwater shrimp catches based on data collected during cruises carried out between 1999 and 2010 in the Gulf of Mexico.Figura 5. Composición porcentual en peso por especie de camarones de aguas profundas, obtenida de datos recolectados en cruceros realizados entre 1999 y 2010 en el golfo de México.

Figure 7 .
Figure 7. Potential deepwater shrimp habitat and bottom topography of the Mexican Gulf of Mexico.Figura 7. Hábitat potencial de camarones de aguas profundas y topografía del fondo marino en el sector mexicano del golfo de México.

Figure 16 .
Figure 16.Mass occurrence of siphonophores in a deep-water shrimp haul carried out off Pacific Costa Rica, March 2011 (Photo: Tayler McLellan Clarke).Figura 16.Masa de sifonóforos obtenida en un lance de pesca dirigida a la captura de camarones de aguas profundas frente a la costa Pacífica de Costa Rica, marzo 2011 (Foto: Tayler McLellan Clarke).
Thus, in 2003 a trap fishery for king crab Paralomis longipes begun, reaching a maximum yield of 9.4 ton in June 2008; during 2009 and 2010 landings of this king crab fishery declined considerably (S.Aguilar, IMARPE, pers.comm.).

Figure 19 .
Figure 19.Distribution and relative biomass (kg km -2 ) of six deep-water shrimps off the Peruvian coast.Data come from four research cruises carried out between 2007 and 2010 with the R/V "Miguel Oliver".Figura 19.Distribución y abundancia relativa (kg km -2 ) de seis camarones de aguas profundas frente a la costa de Perú.Datos obtenidos en cuatro cruceros de investigación realizados entre 2007 y 2010 con el R/V "Miguel Oliver".
. After World War II, Chile experienced a remarkable development of its fisheries, which reached record landings exceeding 8 million ton by 1994.Although most captures involved pelagic fish species, benthic crustaceans inhabiting coastal areas or the continental shelf, such as crabs (Cancer spp., Homalaspis plana), yellow squat lobster (Cervimunida johni) and red squat lobster (Pleuroncodes monodon) were also extracted.

V. 3 .
Statistics.From 2002 to 2009, production of deep-water shrimps in Brazil summed a total of 719 ton (fresh weight).Aristaeopsis edwardsiana constituted the dominant species in the catches, accounting for 75.4% of the landings between 2002-2009.Catches of this species increased steadily from 13.0 ton in 2002 to 182.6 ton in 2005, declining to only 13.7 ton in 2009 (Table

Table 4 .
Landing statistics of the deep-water shrimps in Brazil, between 2002 and 2009.All values in kg.Tabla 4. Desembarque de camarones de aguas profundas en Brasil, entre 2002 y 2009.Todos los valores están expresados en kg.
Source of information: Grupo de Estudos Pesqueiros (Universidade do Vale do Itajaí)