Modelling the distribution of two fish species in seamounts of the Azores

M. Machete^[46], T. Morato^{[46], [47]} and G. Menezes^[46]

1. INTRODUCTION

Seamounts are biologically distinctive habitats of the open ocean exhibiting a number of unique features (see Rogers 1994 for a review). They are characterized by the presence of substantial aggregations of deep-bodied fishes in the water column (Boehlert and Sasaki 1988, Koslow 1996, 1997, Koslow et al. 2000). These aggregations are supported in the otherwise food-poor deep sea by the enhanced flux of prey organisms past the seamounts and the interception and trapping of vertical migrators by the uplifted topography (Tseytlin 1985, Genin, Haury and Greenblatt 1988, Koslow 1997). The discovery of these fish aggregations coincided with declines in shallow-water traditionally exploited stocks (Watson and Pauly 2001) and led to seamounts being increasingly targeted by trawlers throughout the world’s oceans (e.g. the massive, but short-lived, fishery for pelagic armourhead (Pseudopentaceros wheeleri) in the North Pacific; the development of orange roughy and oreosomatid fisheries in the waters around New Zealand and southeastern Australia and subsequently in the North Atlantic (Rogers 1994, Clark 1999, Koslow et al. 2000) and elsewhere. Deep-sea seamount fish communities are highly susceptible to overfishing because they are long-lived, slow growing species with late maturity and low recruitment rates (Koslow 1997, Rico et al. 2001). Thus, when managing such fisheries, caution is required to reduce the risks of overexploitation (Clark 2001, Morato 2003).

2. THE AZOREAN FISHERY

The Azores (located in the region of 36-40°N, 24-32°W) are the most isolated Archipelago of the North Atlantic (Santos et al. 1995) with nine islands spread along a tectonic zone running WNW-ESE (Figure1). Fishing activity started with the colonization of these islands during the 15th century and fish constituted one of the main human subsistence resources (Menezes 1996). In the last two decades the situation has changed with artisanal exploitation having been successively replaced with commercial fishing (Santos et al. 1995). As a consequence, the abundance of several species, and thus the catch rates of the commercial fleet, have started declining over the last few years, while fish stocks have already displayed sign of intensive exploitation (Menezes 2003).

Since no trawl fishery operates around seamounts in the Azores, bottom longlining that targets demersal and deepwater species comprises the most important fishery for the local economy. In fact, even though this fishery does not exceed 5 000 tonnes a year, they still represent a considerable value (Silva and Krug 1994, Silva, Krug and Menezes 1994, Krug 1995, Menezes 1996). The blackspot seabream (Pagellus bogaraveo) has traditionally been the main target species of this fishery, but in recent years several other species, such as the alfonsino (Beryx splendens), have also become important. Most of these species are confined to seamounts, offshore banks and upper-slopes of the islands where bottom longlining occurs down to 1000m depth. Despite the large area (1 million km²) of the Azorean Exclusive Economic Zone (EEZ), the potential area for commercial bottom longlining occupies only approximately 3 percent of the zone.

Little is known about the number of seamounts, their characteristics or their associated fish populations despite the importance of seamounts for local fishing activities. The purpose of this paper is to (a) identify seamounts in the Azorean EEZ, (b) compile information about seamounts’ characteristics, i.e. location, minimum and maximum depth, area of the summit and the seamount, elevation above the seafloor and distance to the nearest seamounts, etc. and (c), estimate some indices of relative abundance using the CPUE of two fish species (alfonsino and blackspot seabream) at several seamounts. Because the size, degree of isolation (Menezes 2003) and slope (Clark, Bull and Tracey 2001) of the seamounts are, among other physical features, important ecological determinants of the abundance of exploited seamount fish populations, this work will include (d), a preliminary attempt of modelling alfonsino and blackspot seabream abundances in seamounts using the above-mentioned physical characteristics as predictors variables. These models can be shown to be useful for predicting the abundance of seamount-inhabiting fish species, especially in data-deficient situations.

3. METHODS

3.1 Seamounts’ physical characteristics

The seamounts around the Azores considered by this study were identified using bathymetric contour maps. Only those that satisfied the following criteria were considered:

i. having the peak shallower than 1 200 m depth, the limit above which most commercially important fish communities inhabit (Menezes 2003)

ii. having an elevation above the seafloor greater than 100 m (as described by Clark et al. 2001) and

iii. having a distance from adjacent seamounts greater than 2 nautical miles (nm) and ability to determine that the catch is from single seamount (Clark et al. 2001).

Bathymetric data used to estimate depth contour maps were taken from the "Global seafloor topography from satellite altimetry and ship depth soundings" database (Smith and Sandwell 1997, <http://topex.ucsd.edu/sandwell/sandwell.html>). A Kriging method was used to interpolate data and build bathymetric contour maps using Surfer 7.05 (Surface Mapping System Golden Software Inc.). Areas and distances were estimated using MapViewer 4.00 (Thematic Mapping System, Golden Software Inc.).

The characteristics chosen to describe seamounts were: (a) latitude and longitude of the centroid of the seamount, (b) minimum depth, (c) elevation (i.e. depth range between the peak and base of the seamount), (d) base area, (Areabase; for seamounts with a base deeper than 2000 m, the area of base was taken as the area of the 2000 m contour), (e) a slope index (seebelow) that represents the average steepness of the flanks of the seamount, and (f), distance to nearest seamount (nm).

3.2 Relative abundance indices

Relative abundance indices (using catch per unit effort) were estimated for alfonsino and blackspot seabream from data collected in 2002 by observers on board two 27 m commercial bottom longline vessels, the Cidade Celestial and Íris do Mar, from 34 fishing events on eightseamounts in the Azorean EEZ. CPUEs were estimated as the number of fish (CPUEn) and catch weight (in kg; CPUEw) of fish per 1 000 hooks.

3.3 Multiple regression models

Multiple linear-regression models were used to estimate abundance indices for alfonsino and blackspot seabream based on the following assumptions: (a) exploration rates are similar on all seamounts, (b) all longline fishing sets target both species on all seamounts sampled and(c), longline catch rates (CPUE) are an indication of relative abundance of each species on each seamount.

Multiple regression models were computed having the indices of fish abundance (CPUE_n and CPUE_w) as dependent variables and the physical seamounts characteristics as independent variables. The resulting equations that had better fits were used to predict the index of abundance for the two fish species on the seamounts for which there were no data.

At this stage, only three physical characteristics of seamounts (predictors) were taken into account for multiple regression models:

i. Area of the seamount shallower than 850m (Area_<850)

The area of the seamounts is known to limit the abundance of alfonsino (Vinnichenko 1997). Since blackspot seabream is highly dependent on benthic habitats (Morato et al. 2001), seamount area might also affect their abundance. Thus, it is expected that the larger the area of the seamount, the higher the values that the abundance index might display. Since these two fish species are known to occur within the top 850 m of the water column (Menezes 2003), for the purpose of this study the area of the seamount shallower than 850 m was considered.

ii. The slope of the seamount (Slope)

Slopes of seamounts are positively correlated with the biomass of some seamount aggregating fish (Clark et al. 2001). Thus, the abundance index of alfonsino is expected to increase with a corresponding increase in slope of the seamount. In the case of seabream, such relationship may not hold true. The average slope of the flanks of the seamounts were estimated as Arctangent [elevation/v(Areabase/p)] expressed in degrees.

iii. Distance to the nearest seamount (Dist.)

The degree of isolation of a seamount may affect fish abundance (Menezes 2003) and thus, the distance (nm) to the nearest seamount was estimated from bathymetric contour maps. The majority of seamount fishes form local populations (Vinnichenko 1998), that, in general, remain throughout their life cycle in the vicinity of the seamount (Clark et al. 2001, Vinnichenko 1998). Exchange of genetic material among populations probably occurs only during the early life-history stages through passive dispersion of eggs and larvae by currents. However, blackspot seabream display ontogenetic changes in habitat preference with juveniles inhabiting the waters of island shelf, whereas adults move to deeper waters and to offshore seamounts (Menezes 2003, Morato et al. 2001). In general, the abundance index is expected to increase with a corresponding decrease in distance from neighbouring seamounts.

4. RESULTS

4.1 Seamounts physical characteristics

Overall, 136 seamounts were identified from the bathymetric contour maps. The depths of their peaks ranged from close to the surface to approximately 1 200 m, while their base depths ranged from 550 to 2 000 m (Figure 2). Seamount mapping revealed a mean elevation of 460 m (SD = 351m), with mean peak of 813 m (SD = 298m) and mean depth at base of 1 273m (SD = 309 m). Most of the mapped seamounts had an elevation between 100 and 300m (Figure3). Thus, our study included 17 underwater mountains with heights above 1 000 m referred to as seamounts, 37 between 500 - 1 000 m referred to as knolls and 85 with elevations lower than 500 m referred to as hills. This classification is based on the US Board of Geographic Names (1981) in Rogers (1994).

The base area of Azorean seamounts ranged from 1.39 to 6 000 km² (mean = 337 m², SD = 720 km²). However, seamounts with a base area smaller than 200 km² were more numerous (Figure 4). The slope index of the seamounts ranged from 0.85 ^o to 9.82 ^o (mean = 3.68 ^o, SD = 1.94 ^o). Slopes ranging from 2 ^o to 5 ^o were more common (Figure 5).

With the exception of four isolated seamounts (Dist. = 35, 43, 58 and 61 nm) all 132 other seamounts have at least one seamount within 15 nm. The distance between seamounts ranged from 2 nm (predefined) to 61 nm (mean = 5.75 nm, SD = 7.33 nm), with most values occurring within 2 and 5 nm (Figure 6).

CPUE_n and CPUE_w for alfonsino ranged from 0.08 to 11.19 fish per 1 000 hooks and from 0.23 to 9.65 kg per 1 000 hooks respectively (Table 1). For blackspot seabream, CPUE ranged from 0.0 to 20.14 fish per 1000 hooks and from 0.0 to 16.18 kg per 1 000 hooks respectively (Table 1). Seamounts with higher CPUEs were Cruiser and A3 for alfonsino, and Cavala and A3 for blackspot seabream.

4.2 Multiple regression models

The preliminary multiple regression models developed using the variables of slope, Area<850 and Dist. are summarized in Table 2. The large standard errors associated with the estimated parameters denote weak relationships with the abundance indices; thus caution should be applied when interpreting the results:

i. Ln(CPUE_n)_alfonsino = - 5.947 + 0.00019 · Area_<850 + 1.106 · Slope + 0.490 · Dist
ii. Ln(CPUE_w)_alfonsino = - 4.909 + 0.00027 · Area_<850 + 0.966 · Slope + 0.404 · Dist
iii. Ln(CPUE_n)_seabream = - 1.827 + 0.00018 · Area_<850 + 1.207 · Slope - 0.142 · Dist
iv. Ln(CPUE_w)_seabream = - 2.620 + 0.00047 · Area_<850 + 1.134 · Slope + 0.076 · Dist

The regression analyses explained a high degree of the variability in the data (R₂>0.75) with the exception of the Ln(CPUE_w) for seabream (R² = 0.46) (Figure 7). The equations with better fits were for alfonsino [Ln(CPUE_w)] and for blackspot seabream [Ln(CPUE_n)]. These were used to predict the indices of abundance on the seamounts (Figures 8 and 9) for which we had no data.

5. DISCUSSION

This work was the first attempt to identify the seamounts in the Azorean EEZ and thus confidence in the results is still weak owing to the data deficient situations. The method used to identify seamounts, seafloor topography from satellite altimetry and ship depth soundings, lacked resolution and thus, prevented the identification of some seamount peaks or even the whole seamount. A better, but costly, solution would be to perform extensive in situ depth soundings that would provide good bathymetric data for the estimation of seamounts depths (peak and base), areas and slopes. In addition, because the estimates of catch-per-unit effort were based on data for only 34 fishing events and eight seamounts and were gathered by commercial fishing boats, they lack robustness and encompass a high degree of uncertainty. To surpass these limitations, it would be desirable to expand the ongoing research fishing cruises to a larger number of seamounts in the Azores. Another weakness of this work is that some assumptions may not hold true. This is mainly the assumption that the exploration rates are similar for all seamounts. Thus, it would be desirable to include an extra parameter in the multiple regression models (e.g. a dummy variable) to express the degree of exploitation of the seamount. Moreover, the variability in fish abundances could be attributed to more physical parameters, apart from those included in this study. When taking these concerns into account, the multiple regression models presented here should be treated with caution.

FIGURE 2. Base and peak depths of mapped Azorean seamount

TABLE 1
Data used for deriving multiple regression models to predict abundance indices of two fish species for seamounts lacking CPUE data

TABLE 2
Statistics of the preliminary multiple linear regression

S.e. in parenthesis.

The sea-bottom topography of the Azores region is complex mainly because of them increased tectonic and volcanic activities. This study showed that the Azores region is apparently dominated by ‘hills’. However, it is necessary to point out that our study did not consider all underwater features (e.g. base depths < 2 000 m, peaks < 1200m)and that some of the criteria used to identify seamounts established a priori may have biased the analysis.

Since the main topographic feature of the Azores are ‘hills’, and not ‘seamounts’ or ‘knolls’, it is likely that the study area may support a less conspicuous population of seamount aggregating fishes compared to that known for areas with similar topography in the world’s oceans (Vinnichenko 1999). Moreover, the mean peak depth (estimated at 813 m) was deeper than the maximum depth where the greatest abundance of the fish species studied is known to occur. This implies limited habitat availability for the two fish species in the region and, as a result, limitations in the area where the traditional bottom longline fleet could operate, as has been advocated by several authors, e.g. Menezes (2003). This is particularly the case for the Azorean fishing fleet, which, with the exception of few developing deepwater fisheries for Aphonopus carbo and Chaceon affinis, fishes to a maximum depth of 600 m. These fishing grounds represent less than one percent of the entire Azorean EEZ (Menezes 2003).

In the study area, most seamounts displayed gentle slopes ranging from 2o to 5 which implies the existence of habitat favourable to benthic-dwelling fish species. Indeed, the blackspot seabream tends to inhabit gentler sloping areas and exhibits a higher association with the benthic habitat, where it feeds on fish and benthic invertebrate such as ophiurids (Morato et al. 2001). In contrast, gentle slopes may not provide favourable conditions for alfonsino, since it tends to aggregate along intermediate slopes of seamounts and feeds in the water column on the flux of prey organisms that pass the seamounts (Morato-Gomes et al. 1998) in a similar way to orange roughy (Clark et al. 2001).

FIGURE 8. Plot of the estimated index of abundance (kilos of fish per 1 000 hooks) for Alfonsino in the seamounts in the Azores. Area of the circles proportional to the natural logarithm of the CPUE_w

FIGURE 9. Plot of the estimated index of abundance (number of fish per 1 000 hooks) for Blackspot seabream in the seamounts in the Azores. Area of the circles proportional to the natural logarithm of the CPUE_n

Most of the mapped seamounts in the Azores can hardly be considered isolated since the distance among them ranged from 3 to 5 nm. For those fish populations that tend to migrate among seamounts, it is probable that the study area may provide favourable conditions for their dispersal. This may be particularly true for some demersal fish species, such as the blackspot seabream, which migrate from coastal waters surrounding islands to offshore seamounts during different stages of their life cycle (Menezes 2003, Morato et al. 2001). On the other hand, the lack of isolated seamounts may not provide favourable conditions for alfonsinos. This may be particularly true because isolated seamounts are likely to have a greater enhancement of primary production caused by the particular local hydrographic conditions (Genin and Boehlert 1985, Dower, Freeland and Juniper 1992, Odate and Furuya 1998, Mouriño et al. 2001), which results in increased prey densities (Boehlert and Genin 1987). Such food resources attract fishes, such as the alfonsino (Morato-Gomes et al. 1998), that prey on macroplankton and nekton. Therefore, the degree of isolation should play an important role in the formation of different habitat types, which attract fish species possessing different attributes and life strategies. For instance, overexploitation of isolated fish communities that lack the ability to migrate among seamounts could result to local extirpation of fish stocks (Menezes 2003). Thus, the degree of isolation should be seriously considered when developing management plans for seamount fisheries. The lack of isolated seamounts along with the gentle slopes of Azorean seamounts and other factors may explain the low abundance of fish aggregation species found by Vinnichenko (1999) on these seabed features.

6. CONCLUSIONS

Deepwater fisheries in general, and seamounts fisheries in particular, are usually characterized by a boom and bust sequence (Koslow et al. 2000) with the targeted fish stocks showing signs of overexploitation within approximately ten years of the beginning of the fishery. This was the case for the orange roughy fishery off New Zealand, Australia, and the North Atlantic (Branch 2001), the pelagic armourhead fishery over seamounts in international waters off Hawaii (Sasaki 1986), and the blue ling (Molva dipterygia) fishery in the North Atlantic (Bergstad, Gordon and Large (s.d.)). Since fish stocks that aggregate around seamounts can be rapidly depleted, maintenance of seamount fisheries has depended on the discovery of unexploited seamounts.

There is a rising concern about threats to seamount ecosystems in the EEZ of coastal states and the high seas; and several countries, such as Canada, Australia and New Zealand, have begun to take action to protect such ‘fragile’ communities. In the Atlantic however, no action has been taken so far; yet, the ‘Oslo and Paris’ (OSPAR) Commission has listed seamounts as threatened habiats that require urgent conservation action. The developing ‘OSPAR Marine Protected Areas’ programme could provide some insight towards such an objective. In addition, seamounts dominated by hard substrata present in the EEZs of European Community country-members (e.g. Portugal) may also qualify as protection sites under the European Habitats Directive (1992, Natura 2000 code 1170 "reefs" in the Interpretation Manual of European Union Habitats EUR 15/2). Further action could include changes in fishing practices, such as switching from trawling (where this method of fishing is undertaken) to longlining, in order to minimize distruction of seabed habitats and associated fauna.

7. ACKNOWLEDGMENTS

The authors wish to thank Tony Pitcher (Fisheries Centre, UBC) and Vasiliki Karpouzi (Fisheries Centre, UBC) for their helpful comments and Ricardo S. Santos (Departamento Oceanografia e Pescas, UAç) for his support. T. Morato was financially supported through the scholarship BD 4473/2001 by the Portuguese National Science Foundation (FCT) while M. Machete’s participation in the Deep Sea 2003 Conference was supported through a travel grant by the Regional Government of the Azores (Fisheries Secretary).

8. LITERATURE CITED

Bergstad, O.A., J.D.M. Gordon & P. Large (s.d.). Is time running out for deepsea fish? ICES Webpage, <http://www.ices.dk/marineworld/deepseafish.asp>.

Boehlert, G.W. & A. Genin 1987. A review of the effects of seamounts on biological processes. pp. 319-334 in B.H. Keating, P. Fryer, R. Batiza & G.W. Boehlert (eds). Seamounts, Islands, and Atolls. Geophysical Monograph 43. Washington, D.C.

Boehlert, G.W. & T. Sasaki 1988. Pelagic biogeography of the armorhead, Pseudopentaceros wheeleri, and recruitment to isolated seamounts in the North Pacific Ocean. Fishery Bulletin 86: 453-466.

Branch, T.A. 2001. A review of orange roughy Hoplostethus atlanticus fisheries, estimation methods, biology and stock structure. South African Journal of Marine Science 23:181- 203.

Clark, M. 1999. Fisheries for orange roughy (Hoplostethus atlanticus) on seamounts in New Zealand. Oceanologica Acta 22: 593-602.

Clark, M. 2001. Are deepwater fisheries sustainable? - the example of orange roughy (Hoplostethus atlanticus) in New Zealand. Fisheries Research 51:123-135.

Clark, M.R., B. Bull & D.M. Tracey 2001. The estimation of catch levels for new orange roughy fisheries on seamounts: a meta-analysis of seamount data. New Zealand Fisheries Assessment Report 2001/75, 40 pp.

Dower, J., H. Freeland & K. Juniper 1992. A strong biological response to oceanic flow past Cobb Seamount. Deep-Sea Research 39: 1139-1145.

Genin, A. & G.W. Boehlert 1985. Dynamics of temperature and chlorophyll structures above a seamount: An oceanic experiment. Journal of Marine Research 43: 907-924.

Genin, A., L. Haury & P. Greenblatt 1988. Interactions of migrating zooplankton with shallow topography: Predation by rockfishes and intensification of patchiness. Deep-Sea Research 35: 151-175.

Koslow, J.A. 1996. Energetic and life-history patterns of deep-sea benthic, benthopelagic and semount-associated fish. Journal of Fish Biology 49(Supplement A): 54-74.

Koslow, J.A. 1997. Seamounts and the ecology of deep-sea fisheries. Americam Scientist 85: 168-176.

Koslow, J.A., G.W. Boehlert, J.D.M. Gordon, R.L. Haedrich, P. Lorance & N. Parin 2000. Continental slope and deep-sea fisheries: implications for a fragile ecosystem. ICES Journal of Marine Science 57: 548-557.

Krug, H.M. 1995. Biologia e avaliação do stock Açoreano de goraz, Pagellus bogaraveo. PhD Theses, Universidade dos Açores, 192 pp.

Menezes G.M. 1996. Interacções tecnológicas na pesca demersal dos Açores. Provas de acesso à categoria de assistente de investigação, Universidade dos Açores. 186pp.186 pp. + anexo.

Menezes, G.M. 2003. Demersal Fish Assemblages in the Atlantic Archipelagos of the Azores, Madeira, and Cape Verde. PhD Thesis, Universidade dos Açores.

Morato-Gomes, T., E. Sola, M.P. Grós, G. Menezes & M.R. Pinho 1998.........Trophic relationships and feeding habits of demersal fishes from the Azores: importance to multispecies assessment. ICESCM 1998/O:7.21pp.ICES CM 1998/7. 21pp.

Morato, T. 2003. Seamounts - hotspots of marine life. ICES Newsletter 40. <http://www.ices.dk/marineworld/seamounts.asp>..

Morato, T., E. Sola, M.P. Gros & G. Menezes. 2001. Feeding habits of two congener species of seabreams, Pagellus bogaraveo and Pagellus acarne, off the Azores (Northeastern Atlantic) during spring of 1996 and 1997. Bulletin of Marine Science 69: 1073-1087.

Mouriño, B., E. Fernández, P. Serret, D. Harbour, B. Sinha & R. Pingree 2001. Variability and seasonality of physical and biological fields at the Great Meteor Tablemount (subtropical NE Atlantic). Oceanologica Acta 24: 167-185.

Odate, T. & K. Furuya 1998. Well-developed subsurface Chlorophyll maximum near Komahashi No. 2 Seamount in the summer of 1991. Deep-Sea Research Part I 45: 1595- 1607.

Rico, V., J.M. Lorenzo, J.A., González, H.M. Krug, A. Mendonça, M.E. Gouveia & M. Afonso-Dias 2001. Age and growth of the alfonsino Beryx splendens (Low 1834) from the Macaronesian archipelagos. Fisheries Research 49: 233-240.

Rogers, A.D. 1994. The biology of seamounts. Advances in Marine Biology 30: 305-350.

Santos, R.S., S. Hawkins, L.R. Monteiro, M. Alves & E.J. Isidro 1995. Case studies and reviews. Marine research, resources and conservation in the Azores. Aquatic Conservation: Marine and Freshwater Ecosystems 5: 311-354.

Sasaki, T. 1986. Development and present status of Japanese trawl fisheries in the vicinity of seamounts. Pp 21-30 in R.N. Uchida, S. Hayasi & G. W. Boehlert (eds). The environment and researches of seamounts in the North Pacific. Proceedings of the Workshop on the Environment and Resources of Seamounts in the North Pacific, NOAA Technical Report NMFS 43.

Silva, H.M & H.M. Krug 1994. Estado de exploração do goraz nos Açores. Relatório da 13ª Semana das Pescas dos Açores: 125-132.

Silva, H.M., H.M. Krug & G.M. Menezes 1994. Bases para a regulamentação da pesca de demersais nos Açores. Arquivos do DOP, Série Estudos, No. 4/94: 41 pp.

Smith, W.H.F., & D.T. Sandwell 1997. Global seafloor topography from satellite altimetry and ship depth soundings. Science 277:1957-1962.: 1957-1962.957-1962.

Tseytlin, V.B. 1985. Energetics of fish populations inhabiting seamounts. Oceanology 25: 237-239.

Vinnichenko V.I. 1997. Russian investigations and deepwater fishery on the Corner Rising Seamount in Subarea 6. NAFO Scientific Council Studies 30: 41-49.

Vinnichenko, V.I. 1998. Alfonsino (Beryx splendens) biology and fishery on the seamounts in the open North Atlantic. ICESCM1998/O:13, 8pp. ICES CM 1998/13, 8 pp.

Vinnichenko, V.I. 1999. Russian investigations and fishery on seamounts in the Azores area. Semana das Pescas,

Watson, R. & D. Pauly 2001. Systematic distortions in the world fisheries catch trends. Nature 424: 534-536.

A life history approach to the assessment of deepwater fisheries in the Northeast Atlantic

M.W. Clarke
Marine Institute, GTP Parkmore
Galway, Ireland
<[email protected]>

1. INTRODUCTION

It has been generally stated that deepwater fishes cannot sustain high levels of exploitation because of their characteristic slow growth, longevity and low reproductive output. However deepwater fish species display a wide variety of life-history strategies, occupying diverse positions along the K-r continuum. Many teleosts display intermediate or conservative life-history characteristics, but the squalid sharks are more stringent K-strategists. Data were used in life-history analyses to assess the sustainability of these mixed-species deepwater fisheries. While there may be scope for compensatory changes in fecundity such scope is likely to be limited, especially for sharks.

The International Council for Exploration of the Sea (ICES) defines deepwater fisheries as those in waters deeper than 400-500 m. Such fisheries have developed rapidly in recent years in ICES Sub-divisions VI and VII (Figure1). This rapid expansion is due to the decline (or indeed collapse) of many traditional stocks. Some of these deepwater fisheries are long established, for example the Norwegian longline fishery for ling (Molva molva) and tusk (Brosme brosme) (Connolly, Kelly and Clarke 1999) while others are by now well established, for example the pelagic trawl fisheries for blue whiting (Micromesistius poutassou) and greater argentine (Argentina silus) (Gordon 2001). Others have developed in the last 10 years but are now quite advanced, such as the French mixed-species trawl fishery (Charuau, Du Pouy and Lorance 1995) and the Spanish deepwater longline fisheries for sharks, forkbeard (Phycis blennoides) and mora (Mora moro) (Pineiro et al. 2001). In most recent years further expansion of fishing to grounds such as Hatton Bank for Greenland halibut (Reinhardtius hippoglossoides), blue ling (Molva dypterygia) and sharks (Langedal and Hareide 2000, Pineiro et al. 2001) have taken place, and a new fishery for orange roughy west of Ireland has developed in most recent years.

Despite the established nature of many Northeast Atlantic deepwater fisheries, data for routine stock assessments are sparse and the lack of adequate, up-to-date information has prevented stock assessments being carried out in ICES since 2000 (Anon. 2002). Most deepwater fisheries in this area are mixed-species fisheries and this leads to problems in assessment and management. Though most deepwater species conform to the general K-selected life history mode, there is considerable variation within these mixed species assemblages in terms of vulnerability to overexploitation (Clarke et al. 2003). Dulvy et al. (2000) highlighted the dangers of exploiting a mixed-species assemblage; the local extirpation of the most vulnerable species may proceed unnoticed as happened in the case of common skate (Dipturus batis) in the Irish Sea. Various authors have used basic life history information as a tool in assessing the resilience of species to exploitation (Brander 1981, Jennings, Reynolds and Mills 1998). The ICES Advisory Committee on Fisheries Management took a simple approach and ranked deepwater species in the NE Atlantic according to a range of life history variables and used this to illustrate the differing risks associated with stock depletion in these species. A more complicated approach was taken by Smith, Au and Show (1998) who calculated "intrinsic rebound potentials" for 26 shark species, incorporating density dependence terms in their analyses. These authors point out that even simple life history data are difficult to collect, so the maximum benefit should be obtained from them.

Most deepwater fisheries developed recently. But already there is strong evidence from around the world that such fisheries may not be sustainable. It is unclear whether roundnose grenadier (Coryphaenoides rupestris) in the Northwest Atlantic will ever recover (Atkinson 1995) and there is evidence that many stocks of orange roughy (Hoplostethus atlanticus) in New Zealand have followed a similar fate (Clark 2001). It is clear that assembling data needed for conventional management will take a long time, in fact often longer than a deepwater fishery might be expected to last (Haedrich, Merrett and O’Dea. 2001). Management should ideally be based on population dynamics, including fisheries-dependent and fisheries-independent data, for example catch numbers at age and abundance and biomass indices collected on an annual basis. Yet for deepwater species this sort of information is mostly lacking. While great efforts have been taken to collect and refine time-series of catch and effort data, this process is only slowly allowing for assessments to be carried out. For deepwater stocks much of the data available relate to life history of target species. This paper aims to demonstrate that even basic life history information can, in itself, provide a framework for the advisory process for deepwater fisheries.

2. MATERIALS AND METHODS

Life history parameters and derived variables for the main deepwater species presented by Clarke et al. (2003) were augmented with data from the literature (Table 1). The basic parameters were maximum size (cm), maximum age, Brody growth coefficient (K), natural mortality (M), length at maturity (L50) and age at maturity (Age50). The ratios of size and age at maturity to maximum size and age were derived. These ratios provide a more meaningful indication of when in the fish’s growth, or life span, maturity actually is attained.

Table 1
Life history parameters of the deepwater and continental shelf dwelling species derived from Irish Marine Institute studies

* Length measurements for Coryphaenoides rupestris are pre-anus length, with total length in parenthesis. all other measurements are total length apart from Eutrigla gurnardus which are fork length

Longevity and length data for ling and blue ling were obtained from Bergstad and Hareide (1996). Size and age at maturity data for ling, blue ling and tusk were obtained from Magnusson et al. (1997), using median of male and female values. Longevity and length data for tusk were obtained from Magnusson et al. (1997) and an estimate of M from Anon. (1996). Maximum age of orange roughy and cardinalfish (Epigonus telescopus) were taken from Talman et al. (2002). For orange roughy, an estimate of M, L50 and maximum size were taken from Branch (2001), and references therein. For greater forkbeard, longevity and growth data were obtained from Casas and Pineiro (2001) and an estimate of M derived from data therein using the technique of Rikhter and Efanov (1976). Maturity data for this species were obtained from Kelly (1997). Life history data on blue whiting were obtained from routine sampling programmes underway in the Irish Marine Institute, while an estimate of M was obtained from Anon. (2003). The life history data above, were used to rank the main deepwater species in order of increasing conservatism in life history mode. The most conservative species was assigned the lowest rank for each life history trait.

The economic value of each species was calculated from records of average Irish prices in Ireland (Marine Institute 2003) and used to rank the species in order of price a tonne. Species that form aggregations that are targeted by the commercial fisheries were assigned ranks of 1, and those that are, in general, dispersed were assigned a rank of 2. Along with the economic value of the species this information can be used to highlight which species are more vulnerable in terms of attractiveness to fishing.

These data were used to derive further biological variables for these species. The ratios of size and age at maturity to maximum size and age were derived. These provide a more meaningful indication of when in the fishes’growth, or life span, maturity actually is attained. Estimates of natural mortality in this study (Table 1) were obtained using a method that assumes that this is the rate required to reduce a recruited population to 1 percent of its initial value (Annala and Sullivan 1996). In the present case, maximum age was taken to be the greatest observed age in samples. Estimates of fecundity and age at maturity were used to derive the potential rate of population increase the surrogate r’ - (Jennings, Reynolds and Mills 1998) as follows

r’ = ln (fecundity at length at 50% maturity)/age at 50% maturity

Fecundity at size at maturity was used for greater argentine and roundnose grenadier, but for the sharks, mean observed ovarian fecundity was used because there was no evidence of increased fecundity with size (Clarke 2000, Girard and Du Buit 1999). Age at maturity was not estimated directly for the sharks but predicted from the von Bertalanffy growth function for birdbeak dogfish (Deania calceus) and from mean length at age in the case of leafscale gulper shark (Centrophorus squamosus).

A Beverton and Holt (1957) yield per recruit analysis was done for two hypothetical species, one with a K-strategist life history, and the other with a more r-selected mode. This model assumes that fish growth is expressed by the von Bertalanffy growth function and that mortality is exponential (Ricker 1975).

3. RESULTS

Length and age data of the deepwater species are presented in Table 1 along with those of the shelf species. Maximum age attained (longevity) by these deepwater species varies. The shortest-lived species was forkbeard attaining ages of 9 years. The species that was estimated to reach the greatest age was leafscale gulper shark, attaining an age of 70 years. Roundnose grenadier was another long-lived species (60 years). Species with intermediate longevity were, in decreasing order, blue-mouth redfish (Helicolenus dactylopterus) (43 years), greater argentine (36 years), birdbeak dogfish (35years) and black scabbardfish (Aphanopus carbo) (32 years). The maximum ages reported were as determined from the studies outlined above. Apart from the work carried out for roundnose grenadier (Gordon, Merrett and Haedrich et al. 1995) and grey gurnard (Eutrigla gurnardus) (Connolly 1986) where marginal increment analysis was employed, these studies did not include any validation of the age estimates that were obtained.

Greatest age at 50 percent maturity was recorded for blue-mouth redfish (15.5 years) though maturity was attained by this species at a smaller size than the other species except for forkbeard. Forkbeard reached maturity at smaller size and age than any of the other species. Greater argentine (4 years) also matures early, while roundnose grenadier matured later (10 years). The deepwater species were longer lived than the shelf species and thus the estimates of natural mortality for the deepwater species were lower. The Brody growth coefficients (K) of the deepwater species indicate that they grow more slowly, reaching asymptotic size at a lower rate than the shelf species. Species displaying fastest growth, in terms of the Brody growth coefficient (K) from the von Bertalanffy growth model was forkbeard, followed by greater argentine. While roundnose grenadier displayed slow growth, the slowest growing of all species examined was blue-mouth redfish, displaying slower growth than the shark birdbeak dogfish. Though blue whiting reaches over 50 percent of maximum length before maturity, it matures at a relatively young age and is fast growing compared to the others (K = 0.19) (Table 2).

A more useful biological parameter than length or age at which 50% of the population reaches maturity is one that provides an indication of the age for the species when it reaches sexual maturity. Table 1 presents length and age at 50% maturity as ratios of maximum length and age in each case. Maturity was reached at largest proportion of maximum size in the case of the leafscale gulper shark (83 and 88% for males and females respectively). The other sharks also mature at high proportions of their maximum length. Roundnose grenadier mature at around 50% of maximum length but in terms of age at only about 18% of maximum. Blue-mouth redfish also attained maturity at an advanced size, though at an earlier percentage of maximum age than roundnose grenadier. The shelf dwelling species all reached sexual maturity at relatively small size and early age; in all cases first maturity was reached at less than 2 years. In contrast, of the deepwater species only roundnose grenadier matured at less than 50% of maximum size. Contrasting patterns of maturity with respect to age are also apparent. The shelf dwellers all matured in the first 12% of their life spans. Apart from roundnose grenadier the deepwater species reached maturity between 20 and 70% of their life spans. Fecundity estimates were only available for two of the teleosts. Greater Argentine females in the range 26.5-45 cm total length had fecundities in the approximate range 4 478 to 16 284 ova. Roundnose grenadier in the range 63-95 cm TL had fecundities in the approximate range 11 000 to 55 000 ova.

Table 2 presents the above data, combined with additional data from various literature sources, to present an overall picture of the varying life histories in the deepwater fisheries. Orange roughy was the most valuable species (2 603 €/t). Most of the remaining true deepwater species commanded lower market prices, between about 1 100 € and 1 500 €/t. The pelagic deepwater species had much lower market prices. Combined with the behavioural ranking, the economic data give some indication of vulnerability. For example, a high-value aggregating species such as orange roughy offers considerable incentive to fisheries. Combined with its conservative life history mode, these characteristics render it particularly vulnerable to exploitation.

The potential rate of population increase for four deepwater species and four shelf dwelling species is presented in Table 3, in order of increasing rate. These values show that the deepwater species all have slower rates of population increase than the shelf-dwelling species. The lowest rates of all are those of the sharks.

The Beverton and Holt yield per recruit simulations show some important differences between fisheries based on K- and r-strategists. K-strategist-based fisheries produce maximum yield at lower rates of fishing mortality than those based on r-strategists (Figure 2). r-strategists may be harvested at higher rates of fishing mortality than those based on K-strategists. The parameters used as inputs to this analysis are given in Table 4.

Table 2
Life history and economic parameters for deepwater species taken in fisheries west of Ireland and Britain

Table 3
Surrogate potential population replacement rate for four deepwater species compared with 4 continental shelf-dwelling species, calculated as fecundity at L₅₀/Age₅₀

Table 4
Parameters used to fit Beverton and Holt (1957) yield per recruit model for hypothetical K and r-strategist species

4. DISCUSSION

The percentage of maximum length at which maturity is reached was greatest in the case of the sharks (78-88%), which agrees with mean values calculated for elasmobranchs by Frisk, Miller and Fogarty (2001). The deepwater teleosts matured at lower percentages of maximum length with the values for roundnose grenadier and forkbeard less than 55% and in the range for the shelf-dwelling species. The shelf-dwelling species appear to reach sexual maturity while somatic growth proceeds. The ratio of age-at-maturity to maximum age represents the portion of time and growth that takes place before the adults invest in reproduction. Again, the sharks had the highest values, indicating that they live most of their lives before they mature. These contrasts between shelf and slope species agree well with published studies; Gordon et al. (1995) notes that slope dwelling fish only mature when somatic growth has slowed or ceased, indicating that on the deepwater slopes, energy is available for growth or reproduction, but not both (Merrett and Haedrich 1997). Estimates of the Brody growth coefficient (K) for the deepwater species predict moderate to low rates of growth to asymptotic size.

The ratios of age at maturity to maximum age for sharks are similar to those reported for shelf elasmobranchs by Frisk, Miller and Fogarty (2001). These authors tentatively suggest that compensatory responses to exploitation may explain earlier maturation in the sharks relative to other - unexploited - vertebrate classes. However, the high likelihood that these species have gestation periods of more than one, and perhaps more than two, years (Girard 2000) coupled with the possibility that they have prolonged periods of rest between reproductive events (Clark and King 1989, Clarke, Connolly and Bracken 2001) might indicate that the scope for compensatory change is limited. There have been reports of density dependent changes in fecundity in the shelf-dwelling squalid shark spinydogfish(spiny dogfish (Squalus acanthias) (Gauld 1979). However Portuguese dogfish (Centroscymnus coelolepis) and birdbeak dogfish do not develop subsequent batches of oocytes during gestation. This seems to support the view that the scope for compensatory changes in deepwater sharks is more limited than might be the case for their shelf-dwelling relatives.

The dangers of not validating age have been illustrated by Beamish and McFarlane (1983). The errors in ageing the Pacific ocean perch (Sebastes alutus) led to a management strategy that was less conservative than was prudent, given the great longevity and low natural mortality of that species. In the absence of tagging data, natural mortality is often estimated using the techniques of Rikhter and Efanov (1976), or Hoenig (1983) or is based on maximum age attained by a stock or species (Annala and Sullivan 1996). The problems with using these approaches may be illustrated with reference to black scabbardfish (Aphanopus carbo). Morales and Carvalho (1996), using whole otoliths, found ages of up to eight years while the results reported in this study from sectioned otoliths were up to 32 years. The resultant differences in estimates of M from the method used in this study, 0.57 and 0.14 respectively, give quite different results about the maximum yield of this species. Again, the need for validation of age is underlined.

Available data on reproduction in teleosts further strengthens the contrasts between shelf and slope. Roundnose grenadier and greater argentine produce small numbers of large eggs (Kelly, Connolly and Bracken 1996, Ronan, Bracken and Malloy 1993) that are characteristic of species inhabiting low-energy environments (Ekau 1991). There are difficulties in translating these fecundity data into annual egg production estimates. There have been several studies of reproduction in roundnose grenadier, but the results have been contradictory. Their spawning is prolonged throughout the year (Allain 2001, Bergstad 1990, Gordon and Hunter 1994, Kelly et al. 1996, Magnusson and Magnusson 1995), however Kelly et al. (1996) found this species to be a determinate spawner, though Allain (2001) considers fecundity to be indeterminate and that the number of egg batches produced each year is unknown. Spawning in greater argentine proceeds throughout the year (Magnusson 1988, Ronan et al. 1993) though there may be seasonal peaks in spawning intensity (Anon. 1995b). There is no published information on the nature of spawning in this species. Differing spawning strategies complicate comparisons between shelf and slope teleosts. The grey gurnard has an asynchronous strategy, spawning repeatedly throughout the breeding season (Connolly 1986). The Atlantic mackerel (Scomber scombrus) also has a protracted spawning period (Anon. 1999). Thus, these species spread their reproductive effort over time to counteract environmental variability by adopting a "bet hedging" strategy (Lambert and Ware 1984). This approach accommodates environmentally induced poor recruitment by increasing the temporal scale of reproductive output relative to that of the environmental fluctuation (Merrett and Haedrich 1997).

The deepwater sharks have much lower fecundities than the slope-dwelling teleosts, but share similar values with shelf-based relatives such as spiny dogfish (Squalus acanthias) (Holden and Meadows 1964). These species produce a small number of well-developed young, with a better chance of survival. This tends to support the idea that stock-recruitment relationships are more defined in these viviparous elasmobranchs than may be the case for many teleosts.

This study illustrates the differences in growth and reproduction between shelf and slope species. These differences might be illustrated using the concept of the K-r continuum. K-strategists tend to inhabit environments where there is little fluctuation, they achieve success by attaining large size, deferred reproduction and producing smaller numbers of more developed offspring (Begon, Harper and Townsend 1996). However there has been much criticism of the K-r concept. Stearns (1992) examines the evolution of life history traits and suggests that earlier authors had tended to consider that so-called K-strategists evolved under density-dependent conditions, while r-strategists evolved in density-independent conditions, a theory that he considers incorrect. Boyce (1984) states that this theory should only be applied to density-dependent models, a point which Stearns (1992) also makes. The deficiencies in this concept should be noted. The use of the model in the current study is by way of an simple generalization of the dichotomy in life-history strategies in these species, and does not consider the selection pressures on individual organisms that produced these traits.

The intrinsic rate of natural increase ® is the rate at which a population increases in size per individual in unit time. It is calculated as the mean number of offspring produced by an individual in its lifetime divided by the average time between the birth of an individual and the birth of the first offspring of that individual- cohort generation time (Begon et al. 1996). Given the uncertainties in reproductive biology of deepwater teleosts and elasmobranchs, it was not possible to calculate r. However, following the method of Jennings, Reynolds and Mills (1998), a surrogate value - the potential rate of population increase (r’) - was derived from the available data for four deepwater species and four shelf species. In this approach, fecundity at age at maturity provides an index of reproductive output and age at maturity an index of cohort generation time (Jennings, Greenstreet and Reynolds 1999). Ranking the species according to r’ suggests that the sharks are least resilient to fishing, followed by the slope teleosts. The shelf dwellers display markedly higher rates of potential population increase. There are no published estimates of the intrinsic rate of population increase ® for the shelf species in this study. Jennings, Reynolds and Mills (1998) state that r could not be calculated for such species because available data were from stocks that had already been exploited, and reduced life-spans would bias the estimates. These authors suggest that r’ is a useful surrogate, and produced estimates of this parameter for a range of shelf species. The parameter r’ incorporates the fecundity at L50, a surrogate for the mean annual egg production and Age50, a substitute value for cohort generation time. Due to uncertainties in the estimates of annual egg-production (see above), r was not calculated for the deepwater species, and the surrogate estimate (r’) was used instead.

Hoenig and Gruber (1990) suggested the possibility of ranking species according to their resilience to exploitation, based on life history characteristics. Smith, Au and Show (1998) calculated "intrinsic rebound potentials" for 26 shark species, incorporating density dependence terms in their analyses. Brander (1981) ranked skate species according to the total mortality the populations could withstand without collapsing. This approach was also taken by Walker and Hislop (1998) for North Sea skates. Smith et al. (1998) note the difficulties in obtaining all the necessary data, therefore it seems prudent to maximise the usefulness of such information for assessment purposes.

These estimates of potential population increase suggest that these deepwater species are less resilient to fishing pressure and that they will respond more slowly to decreased exploitation than those on the continental shelf. The deepwater sharks share their low rates of increase with shelf-dwelling sharks (Smith, Au and Show 1998, Walker and Hislop 1998) but the possibility that the deepwater sharks have long gestation periods of two or more years (Girard 2000) and the likelihood that they have prolonged resting periods between reproductive events (Clark and King 1989, Clarke, Connolly and Bracken 2001) may indicate that these deepwater elasmobranchs are more vulnerable than their shelf-based relatives.

However, life history models are limited in their application for stock assessment purposes., provide only a static view of the population that does not consider possible density-dependent factors. More importantly, this method does not allow any direct measure of how to set management targets. But, these methods can be a useful aid in situations of poor fisheries data. Combined with data on trends in abundance, they can highlight the potential of particular species to recover, if fishing effort is reduced. They can be used to classify species in a fishery along a continuum of relative resilience to exploitation. Another application is to provide informed decisions about prior distributions of the rate of increase for surplus production models.

Most deepwater fisheries on the continental slopes of the ICES area are multi-species in character, with the possible exception of the pelagic trawl fishery for greater argentine (though data from this fishery are lacking). Figure 3 illustrates the interactions between the main gear types in terms of the main species in the catch.

Roundnose grenadier is taken in the multi-species trawl fisheries with a range of species including greater argentine, deepwater sharks, black scabbardfish, blue ling and other species (Charuau, Du Pouy and Lorance 1995). Longline fisheries on the upper slopes target ling and tusk (Anon. 2000) while in deep waters another longline fishery targets mora, forkbeard and the sharks (Pineiro, Casas and Banon 2001). The diversity of species in longline catches is less than from trawl, but sharks tend to dominate the discards from longline harvests (Connolly and Kelly 1996). Orange roughy is taken in the mixed trawl fishery along with roundnose grenadier, and also in a directed fishery using specialised trawl gear, along with the cardinal fish.

Management of these fisheries should consider the vulnerability of each of the species. The ICES Working Group on the Biology and Assessment of Deep-sea Fisheries Resources has ranked the main deepwater species in order of their vulnerability, based on various life history parameters. In relation to these multi-species deepwater fisheries the question arises, how can a range of species be managed when they have a range of differing life history traits, though generally conforming to the K-strategist mode?

The simulations based on Beverton and Holt’s (1957) yield per recruit model (Figure 2) show some important differences between species with what might be termed K-strategist life histories and those with r strategies. Fisheries based on K- strategists (such as the deepwater species in this study) achieve maximum yields at lower rates of fishing mortality (F) than those based on r-strategists. Thus r-strategists (analogous to the shelf-dwelling species) can sustain higher fishing mortalities. The resilience of these species can be gauged by r’ or preferably the intrinsic rate of population increase r. But depending on the rate of fishing mortality some species may decline, while others may sustain that level of exploitation. The lack of species-specific abundance indices may explain why, for example, well-known species such as the common skate almost disappeared from the Irish Sea (Dulvy et al. 2000) while other skates were more resilient to fishing. This highlights the dangers of exploiting multi-species assemblages without taking into account the differing life-histories of the species involved. Framing management objectives may involve choosing the most vulnerable species and setting reference points for fishing mortality based on guidelines based on the precautionary approach. According to the results of this study and the ICES report on deepwater fisheries (Anon. 2001) the most vulnerable species in the exploited deepwater assemblage in the waters west of Ireland and Scotland are the deepwater squalid sharks.

Haedrich, Merrett and O’Dea (2001) stated that management plans for deepwater fisheries cannot follow those developed for traditional shelf stocks. Indeed this statement could apply to traditional stocks also. The usual approach of providing single species advice for each species has been seen to be flawed and is now being changed. ICES is now aiming at providing advice on fisheries, rather than single stocks. The current advice for Irish Sea demersal fisheries, for example, is that fisheries should only proceed when there are no catches of cod or whiting (ICES 2003). This advice is based on analytical catch-at-age assessments that are not possible with deepwater species. However, basic life-history data can be used to produce fleet-based assessments of risk and support fisheries-based advice.

To frame fisheries-based advice for deepwater species, the first step should be to determine the species relationships in each fishery and for each depth range. This analysis can be used to evaluate the current and advised TACs, or effort levels. In other words, a TAC of 1 000t for roundnose grenadier would entail certain amount of sharks, blue ling, black scabbard and orange roughy, for example. The life history analyses provide guidance on which species are most at risk, and can be used by managers to decide on which species are priorities for management decisions. Then a series of scenarios can be run, based on catch or effort levels for the species that have been deemed as priorities. Such an approach along with incorporation of technical interactions, could underpin fishery-based management advice for mixed-species deepwater fisheries.

5. LITERATURE CITED

Allain, V. 2001. Reproductive strategies of three deepwater benthopelagic fishes from the northeast Atlantic Ocean. Fis. Res., 51: 165-176.

Annala, J.H. & K.J. Sullivan 1996. Report from the fishery assessment plenary, April-May 1996: stock assessments and yield estimates. NIWA: Wellington. Unpublished Report.

Anon. 1995a. Estimation of fish biomass in the Irish Sea by means of the annual egg production method. Belfast: Queen’s University of Belfast. Unpublished Report to European Commission.

Anon. 1995b. Report on the Study Group on the Biology and Assessment of Deep-sea Fisheries Resources. ICES CM 1995/Assessment: 4.

Anon. 1996. Report on the Study Group on the Biology and Assessment of Deep-sea fisheries resources. ICES CM 2000 Assess:8.

Anon. 1999. Report of the Working Group on Mackerel and Horse Mackerel Egg Surveys. ICES CM 1999/G:5

Anon. 2000. Report on the Study Group on the Biology and Assessment of Deep-sea fisheries resources. ICES CM 2000 ACFM:8.

Anon. 2001. Report of the Working Group on Biology and Assessment of deep-sea Fisheries Resources. ICES CM 2001/ACFM:23.

Anon. 2002. Report of the Working Group on Biology and Assessment of deep-sea Fisheries Resources. ICES CM 2002/ACFM: 16

Atkinson, D.B. 1995. The biology and fishery of roundnose grenadier (Coryphaenoides rupestris Gunnerus, 1765) in the north west Atlantic. In Proceedings of NATO Advanced Workshop on deepwater fisheries of the north Atlantic Slope. (Ed. A.L. Hopper). Amsterdam: Kluwer. pp. 51-111.

Beamish, R.J. & G.A. Mc Farlane 1983. The forgotten requirement for age validation in fisheries biology. Trans. Am. Fis. Soc., 112 (6): 735-743.

Begon, M., J.L. Harper, & C.R. Townsend 1996. Ecology. London: Blackwell Scientific, 1068 pp.

Bergstad, O.A. 1990. Distribution, population structure, growth and reproduction of the roundnose grenadier in the deep waters of the Skagerrak. Mar. Biol., 107: 25-39.

Bergstad, O.A. and N.-R. Hareide 1996. Ling, blue ling and tusk of the north-east Atlantic. Bergen: Report to Institute of Marine Research.

Beverton, R.J.H. & S.J. Holt 1957. On the dynamics of exploited fish populations. Ministry of Agriculture, Fisheries and Food Fisheries Investigations, Series II, no. 19. 533 pp.

Boyce, M.S. 1984. Restitution of r- and K- selection as a model of density-dependent natural selection. Ann. Rev. Ecol. Syst. 15: 427-448.

Branch, T. 2001. A review of orange roughy Hoplostethus atlanticus fisheries, estimation methods, biology and stock structure. South African Journal of Marine Science., 23: 181- 203.

Brander, K. 1981. Disappearance of Common skate Raja batis from the Irish Sea. Nature, 290: 48-49.

Casas, J. M. and C. Piñeiro, 2000. Growth and age estimation of greater forkbeard (Phycis blennoides Brünnich, 1768) in the north and westh of the Iberian Peninsula (ICES Division VIIIc and IXa). Fisheries Research 47 (19-25). 7 p.

Charuau, A., H. Du Pouy, & P. Lorance 1995. French exploitation of the deepwater fisheries of the north Atlantic. In Proceedings of NATO Advanced Workshop on deepwater fisheries of the north Atlantic Slope. (Ed. A.L. Hopper). Amsterdam: Kluwer. pp. 337-356.

Clark, M. 2001. Are deepwater fisheries sustainable? - the example of orange roughy (Hoplostethus atlanticus) in New Zealand. Fis. Res., 51: 123-135.

Clark, M. R. & K.J. King 1989. Deepwater fish resources off North Island New Zealand: results of a trawl survey, May 1985-June 1986. Wellington: Ministry of Agriculture and Fisheries. New Zealand Fisheries Technical Report, No. 11.

Clarke, M. 2000. Aspects of the biology of three exploited deepwater sharks Centrophorus squamosus, Centroscymnus coelolepis and Deania calceus (Elasmobranchii, Squalidae) from the continental slopes of the Rockall Trough and Porcupine Bank. Dublin: National University of Ireland. Unpublished Ph.D. Thesis, 310 pp.

Clarke, M.W., P.L. Connolly, & J.J. Bracken 2001. Aspects of reproduction of the deepwater sharks Centroscymnus coelolepis and Centrophorus squamosus from west of Ireland and Scotland. J. Mar. Biol. Ass. U.K.. 81: 1019-1029.

Clarke, M.W., P.L. Connolly & J.J. Bracken 2002. Age estimation of the exploited deepwater shark Centrophorus squamosus from the continental slopes of the Rockall Trough and Porcupine Bank. J. Fish Biol., 60: 501-514.

Clarke, M.W., P.L. Connolly & J.J. Bracken 2002. Catch, discarding, age estimation, growth and maturity of the squalid shark Deania calceus west and north of Ireland. Fis. Res. 56: 139-153.

Clarke, M.W., C.J. Kelly, P.L. Connolly and J.J. Bracken. 2003. A life history approach to the assessment and management of deepwater fisheries in the northeast Atlantic. Journal of Northwest Atlantic Fisheries Science. 31: 401-411.

Connolly, L. 1986. Aspects of the biology of the gurnard (Pisces; Triglidae) from the Irish Sea. Dublin: National University of Ireland. Unpublished Ph.D. Thesis, 394 pp.

Connolly, L. & C.J. Kelly 1996. Catch and discards from experimental trawl and longline fishing in the Rockall Trough. J. Fish Biol., 49 (Supplement A): 132-144.

Connolly, L., C.K. Kelly & M.W. Clarke 1999. Longline survey of the eastern slopes of the Rockall Trough. Dublin: Marine Institute Fisheries Leaflet, No. 180.

Dulvy, N.K., J.D. Metcalfe, J. Glanville, M.G. Pawson & J.D. Reynolds 2000. Fishery stability, local extinctions and shifts in community structure in skates. Cons. Biol. 14: 283-293.

EC Fair 1999. Developing deepwater fisheries. Oban: Scottish Association for Marine Sciences. Report to European Commission.

Ekau, W. 1991. Reproduction in high Antarctic fishes (Notothenioidei). Meerforschung, 33: 159-167.

FAO. 1995. Precautionary approach to fisheries. Rome: FAO. Fisheries Technical Paper, 350/1.52 pp.

Frisk, M.G., T.J. Miller & M.J. Fogarty 2001. Estimation and analysis of biological parameters in elasmobranch fishes: a comparative life history study. Can. J.Fis. Aq. Sci.., 58: 969-981.

Gauld, J. 1979. Reproduction and fecundity of the Scottish-Norwegian stock of spurdog, Squalus acanthias (L.). ICES CM 1979/H:54.

Girard, M. 2000. Distribution et reproduction de deux espèces de requins de grands fonds, les <<sikis>>, Centrophorus squamosus et Centroscymnus coelolepis exploités dans l’Atlantique Nord-Est. Rennes: ENSAR. Unpublished Ph.D. Thesis.

Girard, M., & M. H. Dubuit 1999. Reproductive biology of two deepwater sharks from the British Isles, Centroscymnus coelolepis and Centrophorus squamosus (Chondrichthyes: Squalidae). J. Mar. Biol. Ass. U.K. 79: 923-931.

Gordon, J.D.M. 1999. Management considerations of deepwater shark fisheries. In Shotton, R. (Ed) Case studies of the management of elasmobranch fisheries. FAO Fish. Tech. Paper No. 378/2: 774-819. FAO, Rome.

Gordon, J.D.M. 2001. Deepwater fish and fisheries: introduction. Fis. Res., 51: 105-111.

Gordon, J.D.M. & J.E. Hunter 1994. Study of deep water fish stocks to the west of Scotland. Scottish Association for Marine Science, Oban, Unpublished Report.

Gordon, J.D.M., N.R. Merrett & R.L. Haedrich 1995. Environmental and biological aspects of slope dwelling fishes of the North Atlantic slope. In Proceedings of NATO Advanced Workshop on deepwater fisheries of the North Atlantic Slope. (Ed. A.L. Hopper). Amsterdam: Kluwer. pp. 1-27.

Gordon, J.D.M., S.C. Swan, C.J. Kelly & N.-R. Hareide 1995b. Age determination of juvenile roundnose grenadier Coryphaenoides rupestris, a deep water macrourid fish: a preliminary report. ICES CM 1995/G:6, 6pp.

Haedrich, R.L., N.R. Merrett & N.R. O’Dea 2001. Can ecological knowledge catch up with deepwater fishing? A North Atlantic perspective. Fis. Res., 51: 113-122.

Hoenig, J.M. 1983. Empirical use of longevity data to estimate mortality rates. Fis. Bull. (US), 82: 898-903.

Hoenig, J.M. & S.H. Gruber 1990. Life-history patterns in the elasmobranchs: implications for fisheries management. NOAA Technical Report NMFS, no. 90, 1-16.

Holden, M.J. & P.S. Meadows 1964. The fecundity of spurdog (Squalus acanthias L.) and its use for age determination. J.Mar. Biol. Ass. U.K. 42: 179-197.

ICES 2003. Report of the Advisory Committee for Fisheries Management 2003. Copenhagen: International Council for the Exploration of the Sea.

Jennings, S., J.D. Reynolds & S.C. Mills 1998. Life history correlates of responses to fisheries exploitation. Proc. R. Soc.Lon., Series B, Biological Series. 265: 333-339.

Jennings, S., S.R.P. Greenstreet & J.D. Reynolds 1999. Structural change in an exploited fish community: a consequence of differential fishing effects on species with contrasting life histories. J. An. Ecol., 68: 617-627.

Kelly, C.J. 1997. Aspects of the biology of three deep water fish species from the continental slope of the Rockall Trough. Dublin: National University of Ireland. Unpublished Ph.D. Thesis.

Kelly, C.J., L. Connolly & J.J. Bracken 1996. Maturity, oocyte dynamics and fecundity of the roundnose grenadier from the Rockall Trough. J. Fish Biol., 49 (Supplement A): 5-17.

Kelly, C.J., P.L. Connolly & J.J. Bracken 1997. Age estimation, growth, maturity and distribution of the roundnose grenadier from the Rockall Trough. J. Fish Biol., 50: 1- 17.

Kelly, C.J., L. Connolly & J.J. Bracken 1999. Age estimation, growth, maturity and distribution of the bluemouth rockfish Helicolenus d. dactylopterus (Delaroche 1809) from the Rockall Trough. ICES J. Mar. Sci., 56: 61-74.

Kelly, C.J., P.L. Connolly & M.W. Clarke 1998. The deepwater fisheries of the Rockall Trough: some insights gleaned from Irish survey data. ICES CM 1998/O:40. 22pp.

Lambert, T.C. & D.M. Ware 1984 Reproductive strategies of demersal and pelagic spawning fish. Can. J. Fish.Aq. Sci., 41: 1565-1569.

Langedal, G. & N.-R. Hareide 2000 Rapport fra Forsoksfiske med line pa Hatton Bank. Bergen: Directorate of Fisheries. Unpublished Report.

Magnusson, J.V. 1988. On silver smelt (Argentina silus Asc.) in Icelandic waters. ICES C.M.1998 G:39

Magnusson, J.V., O.A. Bergstad, N.-R. Hareide, J.Magnusson and J. Reinert 1997. Ling, blue ling and tusk of the northeast Atlantic. Copenhagen: Nordic Council of Ministers.

Magnusson, J.V. & J. Magnusson 1995. The distribution, relative abundance and biology of the deep-sea fishes of the Icelandic slope and Reykjanes Ridge. In: Hopper, A.G. (Ed.) Deepwater Fisheries of the North Atlantic Oceanic Slope. Kluwer, Amsterdam, pp. 161-199.

Marine Institute 2003. The Stockbook, Galway: Marine Institute, 389 pp. Merrett, N.R. & R.L. Haedrich 1997. Deep-sea Demersal Fish and Fisheries. London: Chapman Hall. 283 pp.

Morales Nin, B. & D. Sena-Carvalho 1996. Age and growth of the black scabbardfish (Aphanopus carbo) off Madeira. Fis. Res. 25: 239-251.

Pineiro, C.G., M. Casas & R. Banon 2001. The deepwater fisheries exploited by Spanish fleets in the Northeast Atlantic: a review of the current status. Fis. Res. 51: 311-320.

Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Ottawa: Department of Fisheries and Oceans, Bulletin of the Fisheries Research Board of Canada. No. 191, 382 pp.

Rikhter, V.A. & V.N. Efanov 1976. On one of the approaches to estimation of natural mortality of fish populations. ICNAF Research Document, 76/VI/8, 12 pp.

Ronan, M.P., J.J. Bracken & J.P. Molloy 1993. Determination of biological characteristics of the greater argentine (Argentina silus) west of Ireland and Scotland. Dublin: University College Dublin. Report to the Commission of the European Communities, no. MA 2.605.

Smith, A.E., D.W. Au & C. Show 1998. Intrinsic rebound potentials of 26 species of Pacific sharks. Mar. Freshw. Res., 49: 663-678.

Stearns, S.C. 1992. The evolution of life histories. Oxford: Oxford University Press. 243 pp.

Talman, S., K. Krusic-Golub, S. Robertson, and C. Green, 2002. Age Estimation of deepwater fish species from the eastern north Atlantic. Queenscliff: Draft Report to An Bord Iascaigh Mhara (The Irish Sea Fisheries Board). 77 pp

Walker, P.A. & J.R.G. Hislop 1998. Sensitive skates or resilient rays? Spatial and temporal shifts in ray species composition in the central and north-western North Sea between 1930 and the present day. ICES J. Mar. Sci., 55: 392-402.