Reproductive biology of the cockfish , Callorhinchus callorynchus ( Chondrichthyes : Callorhinchidae ) , in coastal waters of the northern Argentinean Sea

The cockfish, Callorhinchus callorynchus, is a widely distributed holocephalan in the south-western Atlantic and an important resource for Argentinean fisheries. The reproductive characteristics of this species were studied in northern Argentinean coastal waters, where specimens arrive seasonally (winter-spring). Sexual dimorphism in maximum precaudal length (PCL) and total body mass (TM) was found, being females (PCL=630 mm; TM=3330 g; n=167) larger and heavier than males (PCL=482 mm; TM=1630 g; n=19). Maturity size for females was 466.22 mm PCL, representing 74% PCL of the largest female sampled. The mean number of ovarian follicles was 8.37 (± 4.84). The highest values of ovarian follicles diameter, gonadosomatic and oviducalsomatic indices were found in spring. A high proportion of mature females (44%) had atretic ovarian follicles, suggesting that they were in the resting stage of their reproductive cycle. The low number of mature males recorded and the absence of mature females with spermatophore masses or egg cases indicate that the study area would not be a reproductive zone. The seasonal presence of C. callorynchus in this region could be related to trophic movements. These results, obtained in an area with high fishing pressure on chondrichthyans, will be useful for implementing conservation and management measures.


Introduction
Members of the family Callorhynchidae (Subclass Holocephali) are oviparous marine chondrichthyans, widely distributed in the Southern Hemisphere (Didier, 2004).This monogeneric family (genus Callorhinchus) includes three living species (Nelson, 2006).The cockfish Callorhinchus callorynchus (Linnaeus, 1758) occurs from 23º S, in the south-western Atlantic, to 18º S in the Southeast Pacific (Cousseau, Perrotta, 2013).It is captured year-round as part of commercial bottom trawl fisheries in Argentina, Chile and Peru (Dagit et al., 2007;Consejo Federal Pesquero, 2009).Along the Argentinean Sea (south-western Atlantic Shelf between 35º S and 55º S), this species is abundant from coastal waters to 200 m depth and an important resource for recreational and small artisanal fisheries (Consejo Federal Pesquero, 2009;Cedrola et al., 2011;Cousseau, Perrotta, 2013).Argentinean landings of C. callorynchus represent 97% of the total south-western Atlantic catches of this species (Bernasconi et al., 2015a).In addition, there has been an increase in Argentinean landings from 1992 to 2011, with the highest one recorded in 2009 (2782 tons) (Sánchez et al., 2012;Navarro et al., 2014;Bernasconi et al., 2015a).
Chondrichthyan populations have low biological productivity and therefore they require careful management and monitoring (Walker, 2005a).Knowledge on life history characteristics (e.g.reproductive cycles, size at maturity, fecundity, nursery grounds and habitat) is necessary to develop stock assessment models and ecological risk assessment in order to implement effective sustainable management practices (Bonfil, 2005).Despite its fishery importance and the wide distribution of C. callorynchus off southern South America, its life-history characteristics were studied only in northern Patagonian waters off Argentina (41º-42º S 64º-65º W) (Di Giácomo, 1990Giácomo, , 1992;;Di Giácomo, Perier, 1994, 1996;Di Giácomo et al., 1994;Bernasconi et al., 2015aBernasconi et al., , 2015b) ) and in coastal waters off Coquimbo (30º15'S 70º30'W), Chile (Alarcón et al., 2011).In addition, the Argentinean port with the highest landings of chondrichthyans is located in the south-eastern Buenos Aires Province (Mar del Plata: 38º01'S 57º31'W), where the abundance of C. callorynchus was reduced by nearly 50% from 1994 to 1999 (Massa et al., 2004;Consejo Federal Pesquero, 2009).
The aim of this paper was to study the reproductive characteristics of C. callorynchus in northern Argentinean coastal waters (36°70'-38°10'S).Particular emphasis was put on estimating the size at first maturation and ovarian fecundity, assessing temporal variations in reproductive activity and determining whether the south-eastern coast of Buenos Aires is a reproductive area for C. callorynchus.This information will be essential to analyze the status of the population of C. callorynchus in the south-western Atlantic and implementing conservation and management measures.
Individuals were caught using hooks and lines or bottom gillnets and examined immediately after landing or frozen for subsequent analysis in the laboratory.Sex, precaudal length (PCL, in mm), total body mass (TM, in g), eviscerated body mass (EM, in g) and gonad mass (GM, in g) were recorded for each individual.
For males, inner clasper length (CL, in mm) was measured from the insertion of the pelvic fin to the distal end (Compagno, 1984).For females, maximum oviducal gland width (OGW, in mm) and uterus width (UW, in mm) were recorded.In the case of mature females, liver mass (LM, in g), oviducal gland mass (OGM, in g), and the number and diameter of the largest vitellogenic ovarian follicles (OFD, in mm) were also recorded.
For both sexes, maturity stage was assessed by visual examination of reproductive organs following guidelines by Stehmann (2002), with modifications.Males were assessed as immature when they had testes undeveloped, genital ducts thread-like, frontal tenaculum without denticles and flexible claspers shorter than pelvic fin length.Maturing males had developing testes, genital ducts beginning to coil, frontal tenaculum with small denticles and flexible claspers longer than pelvic fin.Mature males had enlarged and round testes, highly coiled genital ducts, distinct seminal vesicle present (in green colour), frontal tenaculum with numerous sharp denticles and rigid claspers.Females were considered to be immature when they had ovarian follicles undifferentiated, oviducal glands undeveloped and uteri uniformly thin.Maturing females included those individuals with ovarian follicles translucent of different sizes, developing oviducal glands and uteri enlarging caudally.Mature females had vitellogenic ovarian follicles (>5 mm diameter with yellow coloration), crescent-shaped enlarged oviducal glands and wide uteri over their entire length.
Data analyses.To study the temporal size variation, length-frequency distributions were compared between periods using the two-sample Kolmogorov-Smirnov test (Crawley, 2005).Samples of C. callorynchus were grouped bimonthly for temporal data analysis due to the small number of individuals present at some sampling times.

e160137[3]
The PCL-TM and PCL-EM relationships were calculated in both sexes by the equation TM or EM=a PCL b , where a is the intercept and b is the slope which were estimated from logarithmically transformed data (Ricker, 1973;Froese, 2006).The value of b provides information on fish growth and it is used to determine whether weight increase is isometric (b=3) or allometric (negative allometry: b < 3, or positive allometry: b > 3).The null hypothesis of the isometric growth was tested with the t-test and 95% confidence limits of the parameter b were estimated (Zar, 2010;Ogle, 2016).
The onset of maturity was assessed from changes in reproductive organs of males (CL and GM) and females (OGW and UW) in relation to PCL.Also, the proportion of mature individuals in 20-mm PCL intervals was calculated.A logistic model was fitted to binomial maturity data (immaturematuring or mature) in order to estimate the length at which 50% of individuals (PCL 50 ) were sexually mature: P=1(1 + exp(α +(β PCL))) -1 , where P is the estimated proportion of mature individuals at a given length class, and α and β are model parameters (Roa et al., 1999;Crawley, 2007).
Symmetry and functional parity of the ovaries were analyzed.For this purpose, differences in the mean GM, number of ovarian follicles and OFD between right and left ovaries of mature individuals were tested with Wilcoxon paired-sampled test (Zar, 2010).Ovarian fecundity was estimated as the total number of vitellogenic follicles in ovaries and it was plotted against female PCL.Subsequently, Pearson's correlation was used to evaluate possible changes associated with size increment.Furthermore, the number of atretic follicles in ovaries was recorded, considering flaccid follicles with granulated yolk invaded by narrow blood vessels as atretic (Lucifora et al., 2004).
In addition, gonadosomatic (GI=100 × GM/EM), hepatosomatic (HI=100 × LM/EM) and oviducalsomatic (OGI=100 × OGM/EM) indices (Di Giácomo, Perier, 1994;Capapé, 1999) were calculated for mature females.To assess temporal trends in female reproductive condition, the variations in GI, HI, OGI, OFD and number of vitellogenic ovarian follicles between periods were compared using oneway ANOVA or Kruskal-Wallis test when the assumptions of normality or homogeneity of variances were violated.When significant differences were detected, post hoc comparisons with Tukey's honestly significant difference (HSD) or Dunn's test were performed.All statistical analyses were performed using the R statistical software, version 3.0.1 (R Development Core Team, 2013).

Results
A total of 186 specimens of C. callorynchus (19 males and 167 females) were captured during June-July (17%), August-September (22%) and October-November (61%) in coastal waters of south-eastern Buenos Aires Province.Specimens of this species were not caught from December to May in the study area.The majority of males (84%) and females (74%) ranged from 420 to 460 mm PCL and 460 to 560 mm PCL, respectively (Fig. 2).Results from the Kolmogorov-Smirnov tests indicated no variation in female length-frequency distributions between June-July and August-September (D=0.12,n=70, p=0.96),June-July and October-November (D=0.21,n=128, p=0.23) and August-September and October-November (D=0.12,n=136, p=0.84).Females presented the maximum PCL observed (Tab. 1) and they were also heavier than males at sizes <482 mm PCL (Fig. 3).Males and females had isometric and allometric growth, respectively (Tab.1).Males were found in all months in which specimens of C. callorynchus were captured in the study area.The clasper increased remarkably from the 420 mm PCL onwards, when individuals began to mature (Tab.2; Fig. 4).The smallest mature male was 420 mm PCL (16.95 g GM; 77.82 mm CL) and the largest immature male was 425 mm PCL (1.65 g GM; 43.92 mm CL).All males >455 mm PCL (>29.17 g GM; >77.32 g CL) were mature (Tab.2; Fig. 4).Size at 50% maturity could not be estimated for C. callorynchus males due to the small sample size.
Tab. 2. Range and sample size for the reproductive variables recorded for male and female of Callorhinchus callorynchus by maturing stage.n, sample size; PCL, precaudal length (in mm); GM, gonad mass (in g); CL, inner clasper length (in mm); OGW, maximum oviducal gland width (in mm); UW, uterus width (in mm).Females were found in every month of capture of C. callorynchus.Mature females were predominant in each sampled period.Females with egg cases in their uteri or with spermatophore masses in the seminal receptacle were not found.Females showed a rapid increase in OGW and UW between 420 and 500 mm PCL (Figs. 5a,b).The smallest mature female measured 410 mm PCL (25.84 mm OGW; 18.12 mm UW) and the largest immature 490 mm PCL (16.68 mm OGW; 13.20 mm UW).All females >525 mm PCL (>24.62 mm OGW; 10.90 mm UW) were mature (Tab.2; Figs.5a,b).PCL 50 was estimated at 466.22 mm, which corresponded to 74% PCL of the largest female sampled (Fig. 5c).

Discussion
Callorhinchus callorynchus is present in coastal waters of south-eastern Buenos Aires Province (Argentinean Sea between 36°70'-38°10'S) from June to November (winterspring).Pellegrino, Cousseau (2005) reported the capture season for this species in the same area from September to March (spring-summer).Physical (temperature, salinity, depth and bottom sediment characteristics) and biotic factors (e.g.reproductive activity) may influence habitat use in chondrichthyans at all spatial scales (Simpfendorfer, Heupel, 2004).Adult migrations and habitat use of C. callorynchus in northern Patagonian waters (41-42ºS) were not affected by temperature and salinity (Di Giácomo, 1992).
The distribution pattern of individuals could be determined by prey distribution and availability (Di Giácomo, 1992).The low number of mature males found and the absence of mature females with spermatophore masses or egg cases in coastal waters of south-eastern Buenos Aires Province, suggest that the seasonal presence of this species during winter and spring would not be associated to mating and/or egg laying activity.This distribution pattern could be related to trophic movements as almost all specimens examined had gut contents (Melisa Chierichetti 2015, personal observation).For these reasons, further research is required to analyze habitat use of C. callorynchus in the study area in relation to environmental variables and biotic factors, such as food availability.e160137 [7] Sexual dimorphism in maximum size and body mass was found for C. callorynchus in the study area, with females (PCL=630 mm; TM=3330 g; EM=2820 g) larger and heavier than males (PCL=482 mm; TM=1630 g; EM=1330 g).Sexual size dimorphism is also a common characteristic observed in other holocephalans, with females larger than males (Freer, Griffiths, 1993;Francis, 1997;Nibam, 2011;Bell, 2012).The same pattern was observed for C. callorynchus in northern Patagonian waters (PCL males=620 mm; PCL females=760 mm) and in Chilean waters (PCL males=620 mm; PCL females=700 mm) (Di Giácomo, 1992;Alarcón et al., 2011;Bernasconi et al., 2015b).The differences in maximum size among studies for males (present study: n=19; northern Patagonian waters: n=147; Chilean waters: n=193) and females (present study: n=167; northern Patagonia waters: n =1571; Chilean waters: n=209) may result from differences in sample sizes.Moreover, the bias in maximum size observed among studies may be explained by sampling depth variations.Samples of specimens in Chilean waters were taken in coastal waters (<30 m depths) as in the present study, but in northern Patagonian waters samples were taken at depths from 20 to 190 m where individuals of C. callorynchus were segregated by size, with larger individuals at depths >50 m (Di Giácomo, 1992;Alarcón et al., 2011;Bernasconi et al., 2015b).For this reason, the frequency size distribution of C. callorynchus at depths >50 m should be analyzed in waters of south-eastern Buenos Aires Province.
Females of C. callorynchus in this study presented negative allometric growth (b < 3), which means that individuals become more elongated with increasing length.However, PCL-TM relationships for both sexes of this species showed negative allometry in northern Patagonian waters and positive allometry (b > 3) in Chilean waters (Di Giácomo, 1992;Alarcón et al., 2011).These differences result from the fact that within species the variance in weight-length relationships can be substantial, depending on the season, population, and annual or regional differences in environmental conditions (Froese, 2006).Besides, in northern Patagonian waters females were heavier than males at size >500 mm PCL, after reaching size at maturity (Di Giácomo, 1992).In contrast, in this study females were heavier than males at size <482 mm PCL, before reaching size at maturity.
The PCL 50 estimated for females in this study (466.22 mm; n=167) was similar to that recently obtained in northern Patagonian waters (474.80 mm; n=398) (Bernasconi et al., 2015b) but smaller than those previously reported in that zone (490 mm; n=123) (Di Giácomo, Perier, 1994) and in Chilean waters (500 mm; n=40) (Alarcón et al., 2011).These differences may be attributable to differences in the methods for assessing maturity and/or in sample size used in each study.Another possible explanation, taking into account that C. callorynchus has a widely distribution off southern South America, is that perhaps there are multiple discrete populations of this species with latitudinal variation in size at maturity, as was suggested for several chondrichthyan species around the world (Templeman, 1987;Parsons, 1993;Lucifora et al., 1999;Chiaramonte, Petovello, 2000;Mabragaña, Cousseau, 2004;Colonello et al., 2007;Ebert et al., 2008;Barnett et al., 2009).In turn, the PCL 50 estimated in the study area was lower than those values found in northern Patagonian waters, which may be associated with the fact that coastal waters of south-eastern Buenos Aires Province support the highest pressure on chondrichthyans in Argentina (Massa et al., 2004).In the present study, the PCL 50 occurred once C. callorhynchus female reached 74% of the maximum observed PCL, maturing at a smaller proportion than C. capensis females in South Africa (88% of the maximum size) (Nibam, 2011).The delayed size at maturity in holocephalans, a general pattern in chondrichthyans, indicates that they are particularly sensitive to fishing pressure (Walker, Hislop, 1998;Stevens et al., 2000).
The existence of paired functional ovaries is a common feature among oviparous chondrichthyan species (Hamlett, Koob, 1999).In C. callorynchus both ovaries are functional but differences were found in mass between right and left ovaries of mature females, which was not due to differences either in the number or size of follicles.In mature chondrichthyans, ovaries consist of small follicles, developing follicles of various sizes, preovulatory follicles undergoing atresia, and corpora lutea, which are all embedded together in a dense stroma of connective tissue (Walker, 2005b).The mass difference between right and left ovaries in C. callorynchus may be related to differences in the amount of stromal tissue between ovaries, as it was suggested for other oviparous chondrichthyans, such as the smallnose fanskate Sympterygia bonapartii (Mabragaña et al., 2002).
The mean ovarian fecundity for C. callorynchus off the south-eastern Buenos Aires Province (8.37 ± 4.84) was higher than that of C. capensis off South Africa (mean=5.30± 3.78) (Nibam, 2011).The total number of ovarian follicles was dependent on female body size in C. callorynchus, as was observed in other holocephalans (Barnett et al., 2009;Nibam, 2011;Bell, 2012) and oviparous elasmobranchs (Ebert, 2005;Quiroz et al., 2009;Perier et al., 2011;Scenna, Díaz de Astarloa, 2014), which suggest that reproductive potential increases with size.However, it is important to consider that the method of simply enumerating mature follicles to estimate fecundity, under the assumption that all oocytes will be ovulated and become viable egg cases, may be inadequate for oviparous species that reproduce continuously, as enlarged follicles are always present and ovulation may occur at any time (Barnett et al., 2009;Bell, 2012).At the same time, this assumption is not met for C. callorynchus because atretic ovarian follicles are commonly observed from the beginning of the female maturation.Atretic follicles form by degeneration of preovulatory follicles and resorption of yolk from the oocytes, atresia can occur at any follicle size (Walker, 2005b).Consequently, the ovarian fecundity of C. callorynchus may be lower than the one estimated in this study.
e160137 [8] Mature females of C. callorynchus in the south-eastern Buenos Aires Province were collected mainly in October-November (spring) coinciding with high values in GI, OGI, and OFD.In northern Patagonian waters, reproductive activity of C. callorynchus extends nearly throughout the year, mating and egg laying occur in spring and early summer when females with egg cases in their uteri were found at depths ranging from 20 to 40 m, followed by a short period (one month) of gonadal recovery (Di Giácomo, Perier, 1994).The onset and rate of egg laying is apparently temperature-dependent in oviparous chondrichthyans (Holden, 1975;Boisvert et al., 2015).Optimal temperature for egg productivity in C. milii from Australian waters is 16ºC and temperatures below 14ºC would be an environmental trigger to signal the end of the reproductive season, when females would leave the warm coastal waters for deeper waters (Boisvert et al., 2015).Water temperature off the south-eastern Buenos Aires Province during winter and spring are generally <14ºC (Guerrero, Piola, 1997;Lucas et al., 2005).Therefore, it may not be an optimal temperature for egg production in C. callorynchus.On the other hand, neonates were not caught and immature individuals were scarce throughout the sampling period, suggesting that the study area is not a nursery ground for C. callorynchus.However, further research with additional sampling using other fishing gear would be necessary to asses this latter hypothesis, as juveniles may not be catchable by the used fishing gear.Finally, a high proportion of mature females of C. callorynchus (44%) caught in the present work had atretic ovarian follicles.Bell (2012) defined resting females of holocephalans as those with corpora lutea or atretic follicles in their ovaries.Probably, most mature C. callorynchus female that seasonally migrate to coastal waters off southeastern Buenos Aires Province are in the resting stage of their reproductive cycle.
The relationship between parental stock and recruitment success must be of prime consideration in the development of an efficient strategy for the exploitation of chondrichthyan stocks (Hoenig, Gruber, 1990).Consequently, it would be essential to carry out sampling of C. callorynchus at depths >50 m to fully understand the reproductive characteristics of the species in northern Argentinean coastal waters (36°70'-038°10'S).Although the coastal waters off south-eastern Buenos Aires Province would not be a reproductive area for C. callorynchus, a high number of mature females (possibly in the resting stage) arrive seasonally to this zone, where they are subject to fishing mortality.For this reason, this species should be monitored from a fisheries management and species conservation perspective.

Fig. 1 .
Fig. 1.Map of the study area.Continuous lines are isobaths (in meters).The inset shows the location of the study area in Argentina.

Fig. 5 .
Fig. 5. Relationships between precaudal length and a. uterus width and b. oviducal gland width of immature (black circles), maturing (white rectangles), and mature (grey triangles) Callorhinchus callorynchus females.c.Proportion of mature females in 20-mm precaudal length intervals, showing the logistic ogive fitted (black line) to the observed values (white diamonds) and the 95% confidence intervals (dotted lines).Long-dashed lines represent the size at 50% maturity.

Fig. 7 .
Fig. 7. Bimonthly variation in a. hepatosomatic index, b. gonadosomatic index, c. oviducalsomatic index, d. diameter of ovarian follicles, and e. number of ovarian follicles in Callorhinchus callorynchus females.The box represents all observed values between the 25th and 75th quartiles, and contains the median (black circles) for each category.The box plot whiskers represent the minimum and maximum observations for each category in this study.Numbers in parentheses represent the sample size for each period and distributions with the same letter are not significantly different (p > 0.05).