INTRODUCTION
It is widely acknowledged that many species of elasmobranchs are undergoing serious declines in abundance and are threatened with extinction (Musick et al., Reference Musick, Burgess, Cailliet, Camhi and Fordham2000; Stevens et al., Reference Stevens, Bonfil, Dulvy and Walker2000; Myers & Worm, Reference Myers and Worm2005). Although the most publicized threat comes from targeted shark fin fisheries, mixed fisheries where sharks are caught as incidental catch also pose a considerable threat (Musick et al., Reference Musick, Burgess, Cailliet, Camhi and Fordham2000). Dealing with this substantial level of shark by-catch in non-target fisheries may be the greatest challenge in managing global shark resources (Barker & Schluessel, Reference Barker and Schluessel2005). Overfishing and declining shelf sea fish stocks are driving the need to seek new fishing grounds and deep sea fish stocks have been targeted to help meet the demand for fish (Large et al., Reference Large, Hammer, Bergstad, Gordon and Lorance2003). With this comes the inevitable risk to deep water shark species. Defined as those species which are predominantly distributed, restricted to, or spend the majority of their life-cycle at depths below 200 m (following Kyne & Simpfendorfer, Reference Kyne and Simpfendorfer2010), many deep water sharks have slow growth rates, relatively large size and age at maturity and low fecundity making them particularly vulnerable to population depletion. The impacts of fishing activity in the deep sea are relatively poorly understood despite having good baseline data for many deep sea fishery ecosystems owing to their relatively young age and modern fisheries monitoring (Koslow et al., Reference Koslow, Boehlert, Gordon, Haedrich, Lorance and Parin2000; Bailey et al., Reference Bailey, Collins, Gordon, Zuur and Priede2009). An understanding of the biology and population size/structure of both the target and by-catch species is essential in order to effectively manage deep sea fisheries and to understand the impacts of fishing on the structure and function of deep sea ecosystems.
Deep water fisheries are well established in the Rockall Trough region which lies to the west of Ireland and the United Kingdom (Gordon, Reference Gordon2003). In 2003, a total allowable catch (TAC) system was introduced in an attempt to manage fisheries in the region more sustainably (Heymans et al., Reference Heymans, Howell, Ayers, Burrows, Gordon, Jones and Neat2011). CPUE estimates continued to decline for squaloid sharks including Centroselachus coelolepsis, Centrophorus squamosus, Dalatias licha and Deania calcea (Basson et al., Reference Basson, Gordon, Large, Lorance, Pope and Rackham2002; Heesen, Reference Heesen2003; Jones et al., Reference Jones, Beare, Dobby, Trinkler, Burns, Peach and Blasdale2005). Consequently, TACs for sharks were consistently reduced, until 2010 when the TAC for deep water sharks was reduced to zero, with no by-catch allowance. Landings of commercially targeted species of deep water sharks in this region have declined by 90% over the past decade which partly reflects TAC reduction, but also the decline of the stocks (ICES, 2010).
In this study, the population biology of three deep-sea shark species found in the Rockall Trough region of the north-east Atlantic were examined. These species were selected as being most abundant at different depth strata on the continental slope: the black mouthed catshark Galeus melastomus Rafinesque, 1810, the longnose velvet dogfish Centroselachus crepidater (Bocage & Capello, 1864), and the white ghost catshark, Apristurus aphyodes Nakaya & Stehmann, Reference Nakaya and Stehmann1998. Owing to their deep water habitat, all of these species have been less well studied relative to many coastal elasmobranch species, especially A. aphyodes. In particular, very little is known about the age structure and growth patterns for the three shark species, both key parameters to understand the population biology of a species. The ageing of deep-sea sharks presents a unique problem because: (1) they have been shown to have much less calcification of the vertebrae than shallow and coastal species (Cailliet, Reference Cailliet, Pratt, Gruber and Taniuchi1990) rendering many shark ageing techniques (e.g. Cailliet et al., Reference Cailliet, Martin, Kusher, Wolf, Welden, Prince and Pulos1983; Cannizzaro et al., Reference Cannizzaro, Rizzo, Levi and Gancitano1995; Wintner, Reference Wintner2000) unsuitable; and (2) the vertebra of deep-sea sharks exhibit a deep cone structure which limits the penetration of stains and cleaning agents (Correia & Figueiredo, Reference Correia and Figueiredo1997). Gennari & Scacco (Reference Gennari and Scacco2007) presented a new technique to age the velvet belly dogfish, Etmopterus spinax, using acid to etch the softer translucent band (Correia & Figueiredo, Reference Correia and Figueiredo1997) as well as utilizing a stain (cobalt(II) nitrate) to gain further definition and visibility. This may provide a more effective technique for ageing deep-water shark species. It is important to note that within this study shark ‘ages’ remain unvalidated owing to current validation methods being unsuitable for deep sea species; thus all ‘ages’ are here on presented as ‘band counts’.
Therefore, the aims of this study were twofold: (1) to provide fundamental information on the population biology (size structure, length/mass relationships and maturity ogives) of G. melastomus, C. crepidater and A. aphyodes; and (2) to apply the band enhancement ageing method described by Gennari & Scacco (Reference Gennari and Scacco2007) to try to determine the age structure and to derive growth curves for each study species.
MATERIALS AND METHODS
Sample collection
This study was based upon samples collected during an annual Marine Scotland—Science deep water trawl survey (Cruise 1209S) of the Rockall Trough area during September 2009 (Neat et al., Reference Neat, Kynoch, Drewery and Burns2010; Campbell et al., Reference Campbell, Neat, Burns and Kunzlik2011). Samples were collected by bottom trawling from onboard the FRV ‘Scotia’ in a depth range of 500–1800 m on the continental slope between 54°N and 59°N (Figure 1). Trawl duration of approximately sixty minutes was maintained throughout the survey (not including time for descent and ascent of the trawl gear). The bottom trawl (BT184) was rigged with 16 in rock-hopper ground gear (Jackson Trawls Ltd, Peterhead, UK), 1700 kg doors of area 5.82 m2 (Morgere, St Malo, France), 100 m sweeps and floats rated to 2500 m. Warp to bottom-depth ratio ranged between approximately 2.5:1 and 2:1 decreasing gradually with depth. The cod-end was fitted with an internal liner with 20 mm mesh size which ensured retention of fish as small as a few centimetres in length.
Fig. 1. Trawl sites in and around the Rockall Trough, north-east Atlantic, for the Marine Scotland—Science deep water trawl survey, September 2009 (Cruise 1209S).
A total of 30 trawls were conducted lasting an average of 58 ± 12 minutes, ranging in depth from 523 to 1796 m. Length, mass, sex and maturity status were recorded for all sharks caught. Total length (TL, measured to the nearest centimetre) was taken as the anterior tip of the snout to the posterior tip of the caudal fin, with the caudal fin depressed along the anterior–posterior axis. Total mass (W) was measured to the nearest gram. Sex was determined by the presence (male) or absence of claspers. Maturity status was established by macroscopic examination of the gonad using the scale presented in Stehmann (Reference Stehmann2002): individuals at stage three or more (i.e. capable of successful sexual reproduction), for either sex, were considered mature.
For logistical reasons it was not possible to retain all sharks for ageing so a length-stratified sub-sample was kept for each species to represent the whole size range of individuals caught. For Galeus melastomus, a total of 106 (68 females, 38 males) individuals were caught and 64 (40 females, 24 males) were retained for ageing. For Centraselachus crepidater, a total of 173 (100 females, 73 males) individuals were caught with 89 (51 females, 38 males) retained for ageing. For Apristious aphyodes, a total of 33 (23 females, 10 males) individuals were caught, all of which were retained for ageing. Heads of specimens were removed by a single dorsal-ventral cut just anterior of the dorsal fin origin and frozen for later removal of vertebrae in the laboratory. Where target species were caught sampled sharks made up an average of 4.7% of the total trawl mass (range: 0.3–16.3%). Twenty per cent of hauls captured none of the target species.
Vertebrae extraction and preparation
A series of incisions were made to defrosted heads to remove a section of vertebrae, typically 2–3 cm in length and consisting of 3–4 individual vertebra. Excess muscle tissue and the vertebral arches were removed manually. Vertebral sections of small specimens (i.e. G. melastomus and A. aphyodes) were stored in 95% ethanol within small vials until cleaning whilst the larger vertebral sections from C. crepidater were stored frozen at −20°C.
Cleaning of the vertebral column section largely followed the method outlined in Gennari & Scacco (Reference Gennari and Scacco2007) until the vertebrae separated from each other and no remaining tissue was visible. Each cleaning cycle consisted of immersion in a sodium hypochlorite solution (10–15% active chlorine) for 15 minutes followed by a 15 minute rinse in distilled water. The number of repetitions of the cleaning cycle varied depending on residual tissue volume and vertebrae size but a typical vertebral column section of G. melastomus or A. aphyodes required 4–5 repetitions. The larger vertebral sections of C. crepidater were found to be more resistant to cleaning and required longer immersion times (typically 6 h cf. 15 min for the smaller vertebrae). After cleaning, sections of the calcified wall surrounding the centra were removed to ensure that that the growth bands in the centra would be clearly visible after staining (Figure 2A). Clean vertebrae were stored in 70% ethanol solution until staining.
Fig. 2. (A) Sections of calcified wall surrounding vertebra centra which require manual removal. Calcification above and below black lines require removal; (B) example of band definition of vertebra of Galeus melastomus (male, TL= 57 cm) where N B = 5. Vertebra is viewed using a stereomicroscope under transmitted light. Scale bars = 1 mm.
Staining and reading of vertebrae
Staining of the vertebrae followed the method proposed by Gennari & Scacco (Reference Gennari and Scacco2007) with the following modifications: (1) vertebrae were immersed in the cobalt (II) nitrate solution for 1 to 5 min depending on the cone depth and observed degree of calcification; (2) the solution was agitated to allow penetration of the stain into the cavities formed by the deep coned nature of deep sea elasmobranch vertebra; and (3) after initial staining, vertebrae were rinsed thoroughly in distilled water to halt further staining as excessive staining prevented growth bands from being visible. Vertebrae were then immersed in an acid bath (1M hydrochloric acid and 70% ethanol in a 1:20 v:v) for approximately 30 s to etch the growth bands followed by a thorough rinsing in distilled water. After air-drying, vertebrae were viewed using a stereo-microscope under transmitted light and a digital image was taken using a Nikon Coolpix 4500 camera which was then used for reading of growth bands. Band pairs (Figure 2B), defined as a dark layer followed by a lighter layer (Cailliet et al., Reference Cailliet, Smith, Mollet and Goldman2006), were read independently by two readers. Each reader counted the number of band pairs (NB) for each vertebra once without knowledge of the length of the specimen from which the vertebra was taken or the other reader's count.
Statistical analyses
Statistical analyses were conducted using SPSS™ v.14 and Minitab™ 15. A sex-ratio was calculated for each species (female:male) and the Chi-squared test (Zar, Reference Zar1996) used to test for differences between this ratio and an expected ratio of 1:1. The non-parametric Mann–Whitney U-test was used to examine differences in length and mass distribution by sex owing to the non-normal distribution of the data. The length–mass relationships for each sex in each species were calculated using the following equation:
$$LnW=Lna+bLnL\quad \lpar \hbox{King}\comma \; 2007\rpar$$where W is the mass (g), L is the TL (cm), and a and b are constants. The b values for each sex were tested for isometric growth against a value of b = 3 (where 3 = isometric growth) using a t-test. Total length at 50% maturity (TL50) was calculated for each sex within each species using the logistic equation
$$Y{\rm=}\displaystyle{1 \over {\left[{1+e^{ - r\left({L - L_{50} } \right)} } \right]}}$$where Y is the percentage of mature individuals in each size class L (cm) and r is a constant (King, Reference King2007). The index of average percentage error (IAPE) was estimated for reader's growth band counts following Beamish & Fournier (Reference Beamish and Fournier1981). The IAPE is defined as:
$${\rm IAPE=}\displaystyle{{100} \over N}\sum\limits_{j=1}^N {\left({\displaystyle{1 \over R}\sum\limits_{i=1}^R {\displaystyle{{\left[{Xij - Xj} \right]} \over {Xj}}} } \right)}$$where N is the number of fish aged, R is the number of times fish are aged, Xij is the ith age determination for the jth fish, and Xj is the average estimated age of the jth fish. The non-parametric Wilcoxon test (Conover, Reference Conover1971) was used to test the null hypothesis that there were no differences between reader counts. In addition, the average coefficient of variation (CV) between counts was calculated using:
$$CVj=100\percnt \times \displaystyle{{\sqrt {\sum {_i^R=1\displaystyle{{\left({X_{ij} - X_j } \right)} \over {R - 1}}} } } \over n}$$where CV j is the age precision estimate for the jth fish (Campana, Reference Campana2001). Assuming that one band pair is formed each year, the von Bertalanffy growth model was used to estimate L ∞:
$$L_t = {\rm L}_\infty \left[{1 - \exp ^{ - k\left({ - t_0 } \right)} } \right]$$where L t = length at ‘age’ (i.e. number of band pairs) t; L ∞ = asymptotic length; k = the rate at which L ∞ is reached (‘years’−1) and t 0 = ‘age’ (‘years’) of the fish at theoretical zero length (King, Reference King2007).
RESULTS
Galeus melastomus
Galeus melastomus specimens were collected between 533 and 1049 m depth. More females than males (Chi square test, χ22 = 8.49, P < 0.05) were caught (1♀:0.56♂) with females ranging in total length from 32 to 69 cm and males ranging from 34 to 64 cm (Figure 3). Body mass ranged from 100 to 1370 g in females and 128 to 777 g in males. Galeus melastomus showed significant sexual size dimorphism with females attaining a greater mass and length than males (Mann–Whitney U-tests; total length, U = 723, N 1 = 68, N 2 = 38, P < 0.05; mass, U = 637, N 1 = 68, N 2 = 38, P < 0.05). Relationships between Ln length and Ln mass for males and females are presented in Figure 4. Both male and female G. melastomus exhibited isometric growth (t-test, ♂, t = -0.55, df = 66, P = 0.58; ♀, t = 0.21, df = 36, P = 0.83). All maturity stages were present in the G. melastomus catch with a maturity rate of 65% for females and 55% for males respectively. Figure 5 presents the maturity ogives for female and male G. melastomus with TL50 values calculated as 59.7 cm for females and 55.6 cm for males respectively. Both reader's agreed on band pair counts in G. melastomus vertebrae for a subsample of 32 individuals (50% of the total sample size). There was disagreement of band counts for the remaining 32 samples with the largest disagreement being two band pairs (3.13% of the disagreements). However, most disagreements were by only one band pair. The IAPE between readers was estimated as 2.75% with the average CV between reader's counts being 14.1%. There were no significant differences present between the two reader's total counts (Wilcoxon signed-ranks test. T = −3.24, N = 64, P < 0.05). Band pair counts ranged from 0 to a maximum of 7. Using the band pair counts as an estimate of age to fit a von Bertalanffy growth function to the size at ‘age’ data, males attained a lower asymptotic length than females with L∞ values of 60.8 cm and 69.3 cm for males and females, respectively. Values for k and t 0 are not presented as the lack of smaller individuals in the data set render these values meaningless.
Fig. 3. Length–frequency distribution of female (left) and male (right) Galeus melastomus, Centroselachus crepidater and Apristurus aphyodes from the Rockall Trough, 2009.
Fig. 4. Natural logarithm length/weight relationships of Galeus melastomus, Centroselachus crepidater and Apristurus aphyodes from the Rockall Trough, 2009, with females (left) and males (right).
Fig. 5. Maturity ogives for Galeus melastomus (♀, top left, N = 68, L50 = 59.7 cm; ♂, top right, N = 38, L50 = 55.6 cm), Centroselachus crepidater (♀, middle left, N = 100, L50 = 75.4 cm; ♂, middle right, N = 73, L50 = 57.2 cm) and Apristurus aphyodes (♀, bottom left, N = 23, L50 = 56.9 cm; ♂, bottom right, N = 10, L50 = 49.0 cm) from the Rockall Trough, 2009.
Centroselachus crepidater
Centroselachus crepidater specimens were collected between 837 and 1512 m depth although very few individuals were caught around their mid-depth range (1050–1085 m). Significantly more females than males (Chi square test, χ22 = 4.21, P < 0.05) were caught (1♀:0.73♂). Centroselachus crepidater showed sexual dimorphism in length and mass with females being significantly longer and heavier than males (Mann–Whitney U-tests; total length, U = 11432, N 1 = 100, N 2 = 73, P < 0.05; mass, U = 11,328, N 1 = 100, N 2 = 73, P < 0.05). Few small individuals of either sex were caught (Figure 3). Females ranged in total length from 27 to 87 cm and males ranged in total length from 29 to 76 cm. Body mass ranged from 99 to 4726 g in females and 94 to 3024 g in males. The relationships between Ln length and Ln mass for male and female C. crepidater are shown in Figure 4. Both male and female C. crepidater exhibited isometric growth (t-test, ♂, t = 1.96, df = 98, P = 0.35; ♀, t = 0.94, d.f. = 70, P = 0.35). Nearly all maturity stages were found to be represented in this species over the duration of this survey with the exception of female stage five (embryo differentiation stage), of which no examples were found. Amongst females, 66.3% individuals were found to be mature and amongst males 71.2% individuals were found to be mature. Figure 5 presents the maturity ogives for C. crepidater males and females with TL50 values calculated as 75.4 cm for females and 57.2 cm for males respectively. Preliminary investigations showed that the preparation and staining method did not reliably distinguish between the growth bands in this species. A sub-sample of thirty individuals was selected at random and after being prepared and stained only 13.3% of samples (N = 4) showed any discernible growth bands and the maximum number of bands counted was two. Since this number of band counts was not comparable to similar sized fish in other ageing studies for this species (e.g. Irvine et al., Reference Irvine, Stevens and Laurensen2006), no age determinations were conducted for C. crepidater in this study.
Apristurus aphyodes
Apristurus aphyodes individuals were collected from a depth range of 1519–1569 m with significantly more females caught than males (Chi square test, χ22 = 5.12, P < 0.05) (1♀:0.43♂). Female A. aphyodes ranged in total length from 34 to 55 cm, whilst males ranged in total length from 36 to 52 cm. Un-eviscerated mass ranged from 136 to 609 g in females and 161 to 506 g in males. There was no difference in length and mass frequency between males and females for A. aphyodes (Mann–Whitney U-tests; total length, U = 428, N 1 = 23, N 2 = 10, P = 0.15; mass, U = 435, N 1 = 23, N 2 = 10, P = 0.08) (Figure 3). The relationships between Ln length and Ln mass for male and female A. aphyodes are shown in Figure 4. Both male and female A. aphyodes exhibited isometric growth (t-test, ♂, t = −0.26, df = 21, P = 0.80; ♀, t = −0.57, df = 8, P = 0.58). Few mature individuals of either sex were caught with only 13.0% of female and 20.0% of the males found to be mature. Figure 5 indicates that the estimated lengths for 50% maturity (TL50) were 56.9 cm and 49.0 cm for female and male A. aphyodes respectively. Like Centroselachus crepidater, the method used for developing vertebral growth bands in this study was unsuccessful and no growth bands were visible in any A. aphyodes individuals.
DISCUSSION
This study has provided new information on the population biology of three deep sea shark species, Galeus melastomus, Centroselachus crepidater and Apristurus aphyodes in the Rockall Trough region. The results of this study are presented together with the available published data for the three species in Table 1.
Table 1. Summary of ecology and population biology of the deep-sea sharks Galeus melastomus, Centroselachus crepidater and Apristurus aphyodes. M and F, male and female; TL, total length; a and b are the coefficients from the length/weight power curve (W = aLb); TL50 is the length at 50% maturity.
1, Borges et al., Reference Borges, Olim and Erzini2003; 2, Capapé et al., Reference Capapé, Reynaud, Vergne and Quignard2008; 3, Carrasón et al., Reference Carrasón, Stefanescu and Cartes1992; 4, Clarke et al., Reference Clarke, Connoly and Bracken2001; 5, Correia & Figueiredo Reference Correia and Figueiredo1997; 6, Cortés Reference Cortés2000; 7, Cox & Francis Reference Cox and Francis1997; 8, Costa et al., Reference Costa, Erzini and Borges2005; 9, Daley et al., Reference Daley, Stevens and Graham2002; 10, Dulvy & Reynolds Reference Dulvy and Reynolds1997; 11, Blackwell Reference Blackwell2010; 12, IGFA 2001; 13, Irvine et al., Reference Irvine, Stevens and Laurensen2006; 14, Jones et al., Reference Jones, Tselepides, Bagley, Collins and Priede2003; 15, Last & Stevens Reference Last and Stevens2009; 16, Mendes et al., Reference Mendes, Fonseca and Campos2004; 17, Merella et al., Reference Merella, Quetglas, Alemany and Carbonell1997; 18, Nakaya & Stehmann Reference Nakaya and Stehmann1998. *, sex unspecified. ♦, no size range. ◊, un-validated.
In the present study which surveyed during one calendar month only, smaller individuals of all three target species were in low abundance in the catches, as has been reported for other studies on deep-sea sharks in the Rockall Trough region (Girard & Du Buit, Reference Girard and Du Buit1999; Clarke et al., Reference Clarke, Connoly and Bracken2001; Clarke et al., Reference Clarke, Connoly and Bracken2002) which also surveyed for short calendar durations. However, it is unlikely that this is an artefact of the fishing gear used in our study as crustaceans and teleost fish with maximum total lengths of less than 30 mm were retained in the same trawls. It is possible that ontogenetic or seasonal movements between habitats, as seen for other deep-water shark species, may explain the absence of smaller individuals in the Rockall Trough region. For example Deania calcea, another deep-water squaloid shark, is highly migratory (Clark & King, Reference Clark and King1989) with smaller individuals of D. calcea found off the coast of Portugal but absent in the Rockall Trough region where larger size-classes dominate (Clarke et al., 2002). It is possible that a similar migration pattern could exist for the morphologically similar C. crepidater. Small individuals of G. melastomus were also infrequently caught in our sampling. The available data on their diet indicate that juvenile G. melastomus feed mainly upon small invertebrates such as crustaceans switching to teleost fish with increasing size (Olaso et al., Reference Olaso, Velasco, Sánchez, Serrano, Rodríguez-Cabello and Cendero2005; Fanelli et al., Reference Fanelli, Rey, Torres and de Sola2009). It would be expected, in contrast to this study, that smaller individuals preying upon benthic crustaceans would be more prone to capture by bottom trawling than their larger conspecifics which feed upon pelagic teleosts such as blue whiting, Micromesistius poutassou, horse mackerel, Trachurus trachurus and Atlantic saury, Scomberesox saurus saurus (Olaso et al., Reference Olaso, Velasco, Sánchez, Serrano, Rodríguez-Cabello and Cendero2005); unless they occupy shallower water early in ontogeny moving into deeper water as they increase in size. This is supported by data in Rinelli et al. (Reference Rinelli, Bottari, Florio, Romeo, Giordano and Greco2005) who found an increase in mean TL with increasing depth for this species. Such an ontogenetic dietary shift has been reported for the spurdog Squalus acanthias (Jones & Geen, Reference Jones and Geen1977; Ketchen, Reference Ketchen1986). Clearly further studies on the trophic ecology and movement patterns of deep sea sharks are needed.
Significantly higher numbers of females were caught for all target species. Our finding of a sex-ratio in favour of females for G. melastomus is in contrast with that reported by Rey et al. (Reference Rey, de Sola and Massutí2005) for G. melastomus in the Alboran Sea where males significantly outnumbered females. However, a sex-ratio in C. crepidater in favour of females was also reported by Clarke et al. (Reference Clarke, Connoly and Bracken2001) for the Rockall Trough in the same sampling region. In addition, Clarke et al. (Reference Clarke, Connoly and Bracken2001) also reported a sexually dimorphic length–frequency distribution for C. crepidater with females reaching a significantly larger size than males (Table 1). Interestingly, we caught few C. crepidater females containing pups and none collected at the embryo differentiation stage (Stage F5; Stehmann, Reference Stehmann2002) during this survey. Neat et al. (Reference Neat, Burns and Drewery2008) report that all C. crepidater caught in trawls conducted in the Anton Dohrn seamount region, located to the north of the Rockall Trough, during 2006 and 2007, were exclusively female and in the latter stages of pregnancy suggesting that this seamount may be a pupping and potential nursery area for C. crepidater.
Although data on the length–mass relationships for the three shark species under study are limited, the values for a and b obtained in the present study were found to be comparable to those reported in previous studies (Table 1). In particular for G. melastomus where a and b values from multiple studies (Merella et al., Reference Merella, Quetglas, Alemany and Carbonell1997; Borges et al., Reference Borges, Olim and Erzini2003; Mendes et al., Reference Mendes, Fonseca and Campos2004) were available, the results from this study lie close to the regression line obtained when plotting log a vs log b (Froese, Reference Froese2006) providing confidence in our results.
The TL50 estimates reported for G. melastomus and C. crepidater are comparable to values reported in earlier studies (Table 1) and in agreement for a lower TL50 for males than females. Costa et al. (Reference Costa, Erzini and Borges2005) estimated TL50 values of 49.4 cm for males (N = 31) and 69.7 cm for females (N = 35) for G. melastomus. Costa et al. (Reference Costa, Erzini and Borges2005) used clasper length to calculate TL50 for male G. melastomus in contrast with the internal examination used in this study. Estimates of TL50 for C. crepidater reported by Clarke et al. (Reference Clarke, Connoly and Bracken2001) were 51.9 cm for males and 68.1 cm for females, respectively. The TL50 values estimated by Clarke et al. (Reference Clarke, Connoly and Bracken2001) benefitted from larger sample sizes and multiple sampling methods compared to the present study. The estimates of TL50 for A. aphyodes presented here are the first conducted for this species. Nakaya & Stehmann (Reference Nakaya and Stehmann1998) have suggested that A. aphyodes of both sexes are fully mature between 47 and 50 cm TL and the estimates of TL50 for both sexes presented in this study are within/close to these suggested maturity ogives. However, our data would suggest that the upper boundary suggested by Nakaya & Stehmann (Reference Nakaya and Stehmann1998) is too low for females and is likely to be an artefact of the small sample size from which it was derived.
This study applied the ageing technique of Gennari & Scacco (Reference Gennari and Scacco2007) to three species of deep-water shark, G. melastomus, C. crepidater and A. aphyodes, that differ in their mean depth profiles (see Table 1). The method developed by Gennari & Scacco (Reference Gennari and Scacco2007) was shown to be successful when applied to Etmopterus spinax owing to the low disagreement in band counts between readers (14.1% ±1 band). The results of this study indicate limited success with staining of growth bands only possible in G. melastomus. However, the high level of agreement in band counts between readers for G. melastomus indicates that the band pair data reported for this species in the present study can be treated with confidence. In addition, an IAPE of less than 5% is considered good in fish ageing studies and a CV greater than 10% is not uncommon in ageing sharks via vertebral analysis (Campana, Reference Campana2001). The low level of disagreement between readers is most likely to be attributable to the use of high definition digital images as recommended by Gennari & Scacco (Reference Gennari and Scacco2007). In the present study, disagreement in band counts was mostly by ±1 band, as has been reported previously by Correia & Figueredo (1997), with disagreements being more common in larger individuals. This is most likely to be attributable to: (1) the reduced distance between distal bands in older individuals (as a result of slowed rate of growth) making them harder to distinguish (Cortés, Reference Cortés2000); and (2) vertebral margin re-folding with increasing age found in many deep-sea species that exhibit deep coned vertebrae (Gennari & Scacco, Reference Gennari and Scacco2007). The maximum number of band pairs observed in the present study (seven) is similar to Correia & Figueredo (1997) who observed a maximum of eight. The L∞ values estimated in the present study for G. melastomus using the number of band pairs as a proxy for age correspond to the maximum observed size for the species (Rey et al., 2005; Rinelli et al., Reference Rinelli, Bottari, Florio, Romeo, Giordano and Greco2005). However, given the low numbers of smaller (younger) sharks in the present study, we have chosen not to present the k or t 0 values derived from the von Bertalanffy growth curves as they are likely to erroneous.
Clear resolution of the band pairs in the ventral centra of Centroselachus crepidater and Apristurus aphyodes using the staining technique of Gennari & Scacco (Reference Gennari and Scacco2007) was not successful. It is possible that the homoeothermic cold water masses of the deeper habitat in which these species live reduces the effects of seasonality and this, coupled with sporadic food availability, reduces deposition of growth materials in periodic intervals. However, C. crepidater have been successfully aged using dorsal fin spines where maximum growth band count was 54 years (Irvine et al., Reference Irvine, Stevens and Laurensen2006) assuming annual deposition. This suggests that either vertebrae centra are unsuitable for aging in deeper species as the lack of calcification makes growth bands indistinguishable from the rest of the centra or that the chemical composition of vertebral centra of deeper species makes the use of cobalt (II) nitrate as a stain unsuitable for distinguishing growth bands. Gennari & Scacco (Reference Gennari and Scacco2007) used Etmopterus spinax sampled from a depth range of 300–900 m in the Tyrrhenian Sea to develop their staining technique. Whilst E. spinax can be classed as a deep-water species (Kyne & Simpfendorfer, Reference Kyne and Simpfendorfer2010), these sample depths cover only the upper depth range of C. crepidater and may affect the direct transferability of this method to deeper dwelling species. Thus, conventional use of dorsal fin spines is thus likely to remain the established method for ageing C. crepidater and further study is required to find a suitable method for ageing A. aphyodes.
ACKNOWLEDGEMENTS
We are grateful to the skipper and crew of FRV ‘Scotia’. We thank William Reid and Ana Maria Santos for their assistance in sample and data collection. Thanks also go to Martyn Kurr and Andy Marriott for their time in reading growth bands from the digital images and also Andy Marriott for his advice on the statistical analysis part of this study. We thank the two anonymous referees for their helpful comments on the manuscript. This study formed part of the Master's degree thesis which was submitted by Daniel Moore to Bangor University.
FINANCIAL SUPPORT
Marine Scotland—Science deep-water trawl survey (Cruise 1209S) was funded by the Scottish Government (Grant number MF0763).
