- Email: [email protected]
Spatio-temporal variation in the reproduction timing of Atlantic Wolffish (Anarhichas lupus L) in Icelandic waters and its relationship with size
Fisheries Research 183 (2016) 404–409
Contents lists available at ScienceDirect
Fisheries Research journal homepage: www.elsevier.com/locate/fishres
Spatio-temporal variation in the reproduction timing of Atlantic Wolffish (Anarhichas lupus L) in Icelandic waters and its relationship with size Ásgeir Gunnarsson ∗ , Höskuldur Björnsson, Bjarki Elvarsson, Christophe Pampoulie Marine Research Institute, Skúlagata 4, P.O box 1390, 121 Reykjavík, Iceland
a r t i c l e
i n f o
Article history: Received 9 March 2016 Received in revised form 1 July 2016 Accepted 3 July 2016 Handled by Prof. George A. Rose Available online 18 July 2016 Keywords: Spawning time Length Spatio-temporal differences Iceland Temperature
a b s t r a c t Biological data were retrieved from 3694 female Atlantic wolffish Anarhichas lupus collected in Icelandic waters at four locations during the breeding season using a long–term study spanning from 2002 to 2013. The main objective was to investigate reproduction investment and timing. In the main spawning ground, little temporal differences were observed. In contrast, the peak of the spawning season was different among spawning grounds, suggesting spatial differences in the timing of reproduction. The size of females A. lupus was related to the spawning time with the larger fish spawning earlier than smaller ones. In addition, no significant pattern was found between temperature and spawning time. Except at its main spawning area where spawning begins in late September, spawning usually began in late August or beginning of September and was completed in early November. © 2016 Elsevier B.V. All rights reserved.
1. Introduction In temperate environments, seasonal differences in sea temperature and zooplankton abundance will typically affect the timing of reproduction of many fish. Spawning time, which is cyclical for most temperate fish species, is one of the main features affecting survival and growth of fish progeny, as favourable environments are likely to positively affect both attributes. Mortality is the highest for early life stages and has been related to temperature, food availability and predation (Houde, 1997). Temperature, which is one of the most important abiotic factors for the survival of the larvae (Houde, 1989), is also known to influence growth, mortality and juvenile fish size. It is also generally acknowledged that survival rate of larvae is higher when hatching synchronises to the zooplankton spring bloom and is asynchronous when predators are in highest abundance (Hjort, 1914; Cushing, 1969). Studies have also shown positive relationships between survival ratios of larvae or juvenile fish during their first year and year class strength for several fish species (Campana et al., 1989; Carscadden et al., 2013; Ojaveer et al., 2011; Sundby et al., 1989). The timing of the reproduction is therefore crucial for a successful recruitment
∗ Corresponding author. E-mail address: [email protected] (Á. Gunnarsson). http://dx.doi.org/10.1016/j.fishres.2016.07.002 0165-7836/© 2016 Elsevier B.V. All rights reserved.
and consequently for the stock productivity and sustainability. Reproductive timing has been related to temperature, photoperiod, food availability, demography and genetic diversity, inter-annual variation and differences between locations, and differences in spawning time (Bromage et al., 2001; Kjesbu, 1994; Lapolla and Buckley, 2005; Marteinsdottir et al., 2000; Morgan, 2001; Morgan et al., 2013; Ottera et al., 2012). The Atlantic wolffish Anarhichas lupus L. 1758 is widely distributed in the North Atlantic Ocean and is an important commercial species. Within its distribution, the spawning time of A. lupus varies considerably from September to October in Icelandic waters to late July-late September in the White Sea, September in Norwegian waters and autumn in Canadian waters (Jónsson, 1982; Templeman, 1986; Falk-Petersen et al., 1990; Pavlov and Novikov, 1993; Gunnarsson et al., 2006). In Iceland, spawning of A. lupus has been observed all around the country (Fig. 1), but a relatively small (2000 km2 ) area west of the country is considered to be the main spawning area. A. lupus exhibits a rather unusual reproductive strategy; the female coils around the eggs after spawning, and creates a demersal egg cluster that is later guarded by the male during the incubation period which is c. 800–1000 ◦ C-days (cumulative days temperature) (Keats et al., 1985; Ringø and Lorentsen, 1987; Pavlov and Moksness, 1995). In Icelandic waters, the fishery effort using bottom trawl has increased considerably at the main spawning ground during
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
different temperature regimes. The area between areas 2 and 3 was excluded from the analysis due to a lack of data (Fig. 1).
68° Area 3
Area 2 ●
●
67°
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
66°
Area 1
●
●
65°
●
●
●
●
●
●
●
1 10 20 30 40 50
●
●
●
●
● ●
●
●
●
●
●
●
●
64°
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
● ●
● ●
●
● ●
●
2.2. Sampling
● ●
● ●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
● ●
●
●
●
●
●
● ●
● ●
63° Area 4
62° 28°
26°
24°
405
22°
20°
18°
16°
14°
12°
10°
Fig. 1. Spawning () of Atlantic wolffish at Iceland, according to AGFS. The () represents stations where female A. lupus were found near spawning or recently spawned, the size of the circles denote number of such fish, (+) represents station where no such females were found.
spawning time from 1999 onwards, as the commercial value of A. lupus is highest during this period (S. Pétursson, pers. comm.). Before that, the A. lupus fishery was mainly comprised of longliners after the incubation period, outside the spawning areas. In the year 2002, in an attempt to improve the recruitment of A. lupus, part of the main spawning ground was closed during spawning and incubation time (area 1, Fig. 1). As the data and knowledge on spawning of A. lupus was scarce, sampling of data on spatialtemporal distribution of spawning in the area was increased. Based on this sampling and logbooks from fishing vessels, the size of the closed area was further extended in 2010 from 500 km2 to 1000 km2 and the closure period was extended from 1st of October – 1st of May to 15th of September – 1st of May. A. lupus abundance has declined drastically during recent years, especially in the north-west Atlantic Ocean where it was listed by the Canadian Species at Risk Act (SARA) as a species of ‘special concern’ (McCusker et al., 2008). In Icelandic waters, the recruitment was good from 1993 to 1999 but has decreased since then, and been at low level (40% of the highest observed values) from 2010 to 2014. The Marine Research Institute (MRI) of Iceland has given advice based on Maximum Sustainable Yield (MSY) since 2001 but fishing activities exceeded the advised catch for a number of years. However, since 2013, catches have been in accordance with the Total Allowable Catches (TAC) recommended by the MRI (MRI 2015). Concurrently, the fishable stock and the spawning stock have been stable and even increased slightly despite the relatively low recruitment. Nowadays, the trend in A. lupus spawning time is regularly monitored at the main spawning ground to estimate the benefit of the closure period. To observe possible influence of fisheries and environment on spawning time of A. lupus, we investigated potential relationships between the reproduction timing, size of the females, and temperature, at the main spawning ground as well as three other areas where this species is known to spawn. We also assessed spatio-temporal differences in spawning trends within these different spawning areas. 2. Materials and methods 2.1. Sampling areas Sampling areas were chosen to represent both the main spawning area of A. lupus (area 1 in Fig. 1) and other spawning areas with
Biological samples of female A. lupus were collected during 12 consecutive years (2002–2013) from four different areas in Icelandic waters (Fig. 1). All females caught were sampled during the autumn ground fish survey (AGFS) conducted by the MRI from late September to October, using a bottom trawl with 40 mm square mesh codend. In addition, samples were taken for stock assessment purposes from commercial landings from July to December to fully cover the spawning time of the species (Table 1 and Fig. 1). For the analysis of spatial distribution of spawning, only data from the AGFS were used except in 2011 and 2012. In the former year, the survey could not be completed correctly for technical reasons and in the latter, the data collected from areas 3 and 4 were biased. These data were consequently excluded from all analysis. During sampling, the total length (cm) of each female was measured and maturity stages determined visually in the field. The seabed temperature was measured with Scanmar thermometer which were located on the headline of the trawl and calibrated with Starmon mini temperature recorders before cruise.
2.3. Maturity Maturity stages were determined according to the maturity scale of Barsukov (1959), revised by Mazhirina (1988). Only two maturity stages were considered during this study, i.e. stage 3 which refers to fish that intend to spawn during the present spawning season and stage 4 which refers to fish that have recently completed spawning or were recovering after spawning (See Gunnarsson et al., 2006 for a full description).
2.4. Statistical analysis To investigate the effects of female size and spatio-temporal on the time of spawning a multivariate logistic regression model (Cox, 1958) of the form: paldy =
1
1 + exp −ˇald + ay
was estimated. Here paldy is the estimated proportion of fish that has spawned in area at length l, day of the year d, year y and ˇald = ˛a + ˇ1 d + ˇ2 l + ˇa3 d + ˇa4 l + ˇ5 ld + ˇa6 ld which represents the effect covariates and ay the year—area interaction used as a proxy for the spatio-temporal variability. Here the emphasis is on inferring the time of spawning, and its inter-annual variability, and therefore the year—area interaction were treated as a random effect. The parameters of the model were estimated using restricted maximum likelihood (as described in Bates et al., 2015). Variables were selected to the model by eliminating, using an exhaustive search, those variable combinations that contributed the greatest increase to the Bayesian Information Criterion score or BIC (Schwarz, 1978), resulting in a reduced model where the BIC score could not be improved. Models were considered comparable if the difference in the BIC score was less than 2, as suggested by Kass and Raftery (1995). 2 tests were used to compute variable significance in the reduced model.
406
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
Table 1 Mean of total length (±s.d., cm) for each year of female Atlantic wolffish sampled in area 1 and for each areas. n, depicts the number of female measured each year and area. Years
Areas
Year
L
2002 2003 2004 2005 2006 2007
67.4 64.0 67.7 68.5 66.3 69.5
± ± ± ± ± ±
7.63 9.66 9.11 8.47 8.93 10.25
n
Year
222 82 121 199 110 146
2008 2009 2010 2011 2012 2013
L 66.1 67.8 68.2 66.3 67.9 69.3
± ± ± ± ± ±
9.09 8.96 7.19 9.30 9.11 7.73
The results from the model estimation procedure were used to determine ST50 , i.e. the day of the year for which 50% of the fish had spawned, which was determined based on the following formula:
− ˛a + ˇ2 l + ˇa4 l + ay
3.1. Spawning distribution
L
n 67.6 61.3 61.0 67.4
± ± ± ±
8.86 12.02 16.01 11.58
2269 545 225 655
270
ˇ1 + ˇa3 + ˇ5 l + ˇa6 l
3. Results
1 2 3 4
280
To determine if estimated ST50 were different between years, areas or lengths at 95% confidence intervals (CIs) were constructed based on Normal approximation, where the variance was calculated using the delta method (as described in Oehlert, 1992), based on the estimated covariance from the model. Post-hoc differences between years or areas were subsequently determined based on the approach described by Tukey (1949) to account for multiple testing. The main spawning area (area 1) was the only investigated to assess the temporal variation of ST50 as data from other areas were considered insufficient to compare ST50 between years. Therefore the annual parameter estimates from the model were pooled (2002–2013) for each of the areas (2–4), and compared with pooled data from the main spawning ground (area 1). Linear regressions were used to examine relationship of seabed temperature and ST50 as temperature measurements were only available for AGFS samples. All statistical analyses were performed using R (version 3.2.2, see R Core Team, 2015 for further details).
Areas
254 220 218 137 460 100
a)
260 250 2004
2007
2010
2013
Year
b) 280
Day of the year
ST 50,ay =
n
270 260 250 60
70
80
90
3
4
Length (cm)
c) 280 270 260 250 1
2
Area
According to AGFS, spawning of A. lupus was widespread. However its main spawning ground was located West of Iceland in area 1 (Fig. 1). The number of females on maturity stages 3 and 4 was 242, 154, 85 and 90 in areas 1–4 respectively, which means 42%, 27%, 15% and 16% of the total number registered. These results indicate that area 1 was the most important area for spawning. 3.2. Model selection The model selection for spawning time five models were deemed equivalent with the lowest BIC value, 1) the full model, 2) the model omitting ˇa6 , 3) the model omitting ˇa6 and ˇ5 4) the model omitting ˇa4 , ˇ5 andˇa6 , and 4) the model omitting ˇa3 , ˇ5 andˇa6 . Although model 3 had the lowest BIC score (BIC ≈ 3 to other four models) model 4 was chosen as an interaction term between length and area, i.e. ˇa4 , was considered to be related to variations in sampling location between years within areas not actual differences in spawning time. 3.3. Spawning time In the main spawning ground (area 1), estimated ST50 did not differ significantly between years (p > 0.05) (Fig. 2a and Table 2). The range of ST50 was about seven days, the earliest was on day 271 in 2004 and the latest on day 278 in 2006. Based on ST50, a significant difference was only observed in spawning time between
Fig. 2. Estimated ST50 in area 1 by (a) length and (b) areas.
area 1 and 3 (p = 0.012, Fig. 2c and Table 2). Comparing spawning ogives between areas revealed significant interaction between days and areas (2 (3) = 60.299, p < 0.001, Fig. 3b) where shape of the spawning ogive differed significantly in areas 2 and 3 from 1 and 4 (2 (2) = 58.39, p < 0.001). Spawning was first observed on day of the year 260, 234, 239 and 246 in areas 1–4 respectively and the duration of the spawning time was 48, 57, 65 and 59 days respectively, or about two months for all areas except area 1 where spawning lasted one and a half months. For all areas the spawning was completed in late October or early November. 3.4. Fish length and spawning time In the main spawning ground (area 1), the interaction term between length of spawner and day of year was estimated to offer a non-significant improvement to the model (2 (1) = 0.195, p = 0.659). The proportion of females that had spawned varied significantly both as a function of length and day of year with larger fish spawning earlier (2 (1) = 58.71, p < 0.001 and 2 (1) = 2847.87, p < 0.001 respectively). Fig. 3a illustrates the estimated effect of length and day proposed by the model for area 1. In addition Table 2 and Fig. 2b shows the estimated ST50 from the model as a function of length.
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
407
Table 2 Estimate of the day of the year (±s.e.), by years and length (cm) in area 1 and for all areas. Years Year 2002 2003 2004 2005 2006 2007
Length ST50 276 274 271 273 278 274
± ± ± ± ± ±
1.7 1.7 1.8 1.7 1.8 1.7
Date
Year
1. Oct. 30. Sep. 27. Sep. 29. Sep. 3. Oct. 30. Sep.
2008 2009 2010 2011 2012 2013
ST50 272 272 275 275 274 278
± ± ± ± ± ±
1.7 1.7 1.7 1.7 1.7 1.8
Date
Length
28. Sep. 28. Sep. 1. Oct. 1. Oct. 29. Sep. 4. Oct.
60 70 80 90
Fig. 3. Estimated spawning ogives by day of the year and (a) length in area 1 and (b) areas, based on prediction for 70 cm fish from logistic regression.
3.5. Temperature The mean seabed temperature registered from AGFS was 7.38 ◦ C, 6.51 ◦ C, 5.33 ◦ C and 8.05 ◦ C for areas 1–4 respectively. The mean seabed temperature was highest in area 4 and lowest in area 3; the difference between these two areas was about 2.7 ◦ C. The variability of temperature in the main spawning ground (area 1) was relatively small. The average for the period 2002–2013 was 7.41 ◦ C and standard deviation 0.42 ◦ C. The lowest value was 6.77 ◦ C in 2002 and highest 8.15 ◦ C in 2008. This limited range makes it impossible to infer about the relationship between estimated ST50 and temperatures (p > 0.05). Indeed, the linear regression analysis did not reveal any significant relationship between the seabed temperature and estimated ST50 between years in the main spawning ground (area 1; p > 0.05) nor between seabed temperature and estimated ST50 when areas were compared (p > 0.05). 4. Discussion Despite the status of A. lupus as an endangered species and its important economic value in the North Atlantic, data on the reproduction of this species which are crucial for sustainable management are limited. The present study therefore aimed at
Areas ST50 280 276 271 267
± ± ± ±
1.8 1.7 2.3 4.5
Date
Area
5. Oct. 1. Oct. 27. Sep. 23. Sep.
1 2 3 4
ST50 274 268 263 265
Date ± ± ± ±
1.7 3.7 3.5 4.7
30. Sep. 23. Sep 18. Sep. 20. Sep.
investigating the impact of fish size and temperature on the spawning time of A. lupus, as well as spatio-temporal differences in duration of the reproduction and timing in different Icelandic spawning areas. The results presented here revealed that variability in spawning time is relatively small in the main spawning ground (area 1). In addition, larger fish spawn earlier than the smaller ones. For several fish species, cod included, it has been shown that the larger individual spawn earlier than the smaller (Carscadden et al., 1997; Lambert, 1987; Marteinsdóttir and Petursdottir, 1995; Wright and Gibb, 2005). Further, cod studies have demonstrated that fishing effort might lead to drastic changes in spawning time and duration by catching more larger than smaller fish in spawning aggregation (Rose et al., 2008; Scott et al., 2006; Zemeckis et al., 2014). Relationships could not be found between the seabed temperature and the reproduction time of A. lupus. The observed lack of differences in the peak of spawning (ST50 ) between years in the main spawning area of A. lupus (area 1) could be explained by the fact that temperature differences from year to year were too small to affect the reproduction pattern of the species or that temperature trajectories were similar between years. Indeed, if several studies have shown that higher temperature during the vitellogenesis or during the spawning season will delay the spawning of A. lupus (Tveiten and Johnsen, 1999; Tveiten et al., 2001), temperatures trajectories might also be important. Evidence showed that A. lupus reproducing at around 8 ◦ C will delay the peak of spawning of approximately 9 days compared to A. lupus reproducing at 4 ◦ C (Tveiten et al., 2001). On the other hand, differences of 2–5 ◦ C in temperatures did not affect the peak of spawning of A. lupus during two consecutive years (Pavlov and Moksness, 1996). One important feature when comparing the two aforementioned studies remains the direction of temperature changes which affect maturation and spawning time (Wang et al., 2010). In the former case (Tveiten et al., 2001), temperature changes were abrupt while in the latter (Pavlov and Moksness, 1996) temperature trajectories were similar between years and changes gradual. Comparing the peak of spawning (ST50 ) between spawning grounds revealed that ST50 was about two weeks later in warmer waters of area 1 than in colder waters of areas 2 and 3, which is a much more pronounced delay than the one observed by Tveiten et al. (2001) for a lower variability of temperature. Although constant light regimes have been shown to protract the spawning season of A. lupus to 9–10 months (Pavlov and Moksness, 1994), it did not seem to affect spawning time during our study as spawning times were different between locations despite a similar photoperiod. The most striking results were the observed stability of spawning in area 1 and its delay compared to spawning in other areas such as area 3, which might be explained by the availability of food following the hatching of larvae. The spawning season was about 2 months in all areas, except in area 1 where it was about 1.5 months, a result in accordance with previous observation from the White Sea (Pavlov and Novikov, 1993). Indeed, larvae that hatched from eggs spawned in the first half of September might have a lower
408
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
survival rate due to unfavourable environmental food conditions in area 1. Another possible explanation is that egg size greatly affects offspring fitness as well as larval survival since larger eggs usually results in larvae with higher energy resources (Brooks et al., 1997). By delaying spawning for two weeks, females from area 1 might produce higher quality eggs and ensure higher survival rate of their offspring. This could explain why area 1 is the largest spawning area for A. lupus, and exhibited the largest effective population size (Pampoulie et al., 2012). What favours the area 1 as a spawning ground for A. lupus is currently unknown but factors such as temperature, currents and substrate might be important. Recent research on the substrate quality in this area shows a relatively flat bottom characterised by glacial remains and iceberg scouring, localised at a water depth of 150–160 m. The thin sediment cover is mostly gravelly sand with dispersed finer material. Video records show layers of crust which are regarded as appropriate nesting holes for A. lupus. The origin of the crust is still under consideration (G. Helgadottir, pers. comm.). Recent studies on tagging at area 1 (Gunnarsson et al., unpublished) and genetic structure within Icelandic waters have nevertheless revealed contradicting results, i.e. a high degree of spawning ground fidelity and an absence of genetic structure (Pampoulie et al., 2012). Despite the absence of genetic structure at neutral markers, A. lupus seemed to exhibit life history traits variation within Icelandic waters (Gunnarsson et al., 2006) which might be important to consider for a proper management. Apart from the main spawning ground some uncertainty is related to the definition of the spawning areas used in this study. Although spawning females occurred in these areas, the sampling was coarser outside of the main spawning grounds, in particular in the southern area, and some practical considerations, such the number of available samples, were taken into account when defining the three other spawning areas. Thus a finer spatial structure, such as two spawning grounds within a study area, cannot be ruled out which may explain variation in the shape of the spawning ogive between areas 1 and 3, versus 2 and 4. During the present study, the relatively large sampling effort in the area 1 was implemented to find the optimal closure time, which could protect the fish during the incubation period without compromising the fishery when the price for the fish is highest on the market. Therefore, the correct “closure time” would typically start when spawning events are observed. Although the fish begin to aggregate in this area beginning of August, the “closure time” was thus based on the lack of observation of spawned individual before the 20th of September. The current start of “closure time” (day 260; 15th of September) seems to fit well with the attempt to protect males in their nest. The observed stability of the spawning time in this area during this study makes it possible to select one specific “closure time” that does not have to be re-evaluated every year, but the timing of the spawning event will need regular monitoring. A similar compromise was made with regard to the size of the closed area which now constitutes about 2/3 of the main spawning ground. Concurrently to these “protective” measures, TAC has decreased since 2006 and catches followed the advice from the MRI from 2013 onwards. Despite all these efforts to improve the recruitment of A. lupus, the recruitment index seems to remain stable (low level from 2011 onwards) and although the “closure time” of the main spawning ground seems suitable, further research is needed to potentially understand the factors at stake.
Acknowledgements We would like to thank Mary Frances and two anonymous reviewers for their constructive comments and efforts to improve the manuscript.
References Barsukov, V.V., 1959. Wolffish Family (Anarhichadidae). Indian National Documentation, New Delhi. Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48. Bromage, N., Porter, M., Randall, C., 2001. The environmental regulation of maturation in farmed finfish with special reference to the role of photoperiod and melatonin. Aquaculture 197, 63–98, http://dx.doi.org/10.1016/S00448486(01)00583-X. Brooks, S., Tyler, C.R., Sumpter, J.P., 1997. Egg quality in fish: what makes a good egg? Rev. Fish Biol. Fish. 7, 387–416. Campana, S.E., Frank, K.T., Hurley, P.C.F., Koeller, P.A., Page, F.H., Smith, P.C., 1989. Survival and abundance of young atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) as indicators of year-class strength. Can. J. Fish. Aquat. Sci. 46, 171–182. Carscadden, J., Nakashima, B.S., Frank, K.T., 1997. Effects of fish length and temperature on the timing of peak spawning in capelin (Mallotus villosus). Can. J. Fish. Aquat. Sci. 54, 781–787, http://dx.doi.org/10.1139/cjfas-54-4-781. Carscadden, J.E., Gjøsæter, H., Vilhjálmsson, H., 2013. Recruitment in the Barents Sea Icelandic, and eastern Newfoundland/Labrador capelin (Mallotus villosus) stocks. Prog. Oceanogr. 114, 84–96, http://dx.doi.org/10.1016/j.pocean.2013. 05.006. Cox, D.R., 1958. The regression analysis of binary sequences. J. R. Stat. Soc. B 20, 215–242. Cushing, D., 1969. The regularity of the spawning season of some fishes. Extr. Du J. du Cons. Int. pour L’exploration la Mer. 33, 81–92. Falk-Petersen, I.B., Haug, T., Moksness, E., 1990. Oservations on the occurrence, sise, and feeding of pelagic larvae of the common wolffish (Anarhichas lupus) in western Finnmark, northern Norway. J. Du Cons. 46, 148–154. Gunnarsson, Á., Hjörleifsson, E., Thórarinsson, K., Marteinsdóttir, G., 2006. Growth, maturity and fecundity of wolffish Anarhichas lupus L. in Icelandic waters. J. Fish Biol. 68, 1158–1176, http://dx.doi.org/10.1111/j.1095-8649.2006.00990. Hjort, J., 1914. Fluctuations in the great fisheries of Northern Europe viewed in the light of biological research. Rapp. Procès-Verbaux des Réunions du Cons. Int. pour l’Exploration la Mer. 20, 1–228. Houde, E.D., 1989. Comparative growth, mortality, and energetics of marine fish larvae: temperature and implied latitudinal effects. Fish. Bull. 87, 471–495. Houde, E., 1997. Patterns and consequences of selective processes in teleost early life histories. In: Chambers, C., Trippel, E.A. (Eds.), Early Life History and Recruitment in Fish Populations. Chapman and Hall, Padstow, pp. 173–196. Jónsson, G., 1982. Contribution to the biology of catfish (Anarhichas lupus) at Iceland. J. Mar. Res. Inst. Reykjavik 4, 3–26. Kass, R., Raftery, A., 1995. Bayes factors. J. Am. Stat. Assoc. 90, 773–795. Keats, D.W., South, G.R., Steele, D.H., 1985. Reproduction and egg guarding by Atlantic wolffish (Anarhichas lupus: Anarhichidae) and ocean pout (Marcozoarces americanus: Zoarcidae) in Newfoundland waters. Can. J. Zool. 63, 2565–2568. Kjesbu, O.S., 1994. Time of start of spawning in Atlantic cod (Gadus morhua) females in relation to vitellogenic oocyte diameter, temperature, fish length and condition. J. Fish Biol. 45, 719–735. Lambert, T., 1987. Duration and intensity of spawning in herring Clupea harengus as related to the age structure of the mature population. Mar. Ecol. Prog. Ser. 39, 209–220, http://dx.doi.org/10.3354/meps039209. Lapolla, A., Buckley, L., 2005. Hatch date distributions of young-of-year haddock Melanogrammus aeglefinus in the Gulf of Maine/Georges Bank region: implications for recruitment. Mar. Ecol. Prog. Ser. 290, 239–249, http://dx.doi. org/10.3354/meps290239. G. Marteinsdóttir, G. Petursdottir, 1995. Spatial and temporal variation in reproduction of icelandic cod at selvogsbanki and nearby costal areas ICES C. 1995/G15. Marteinsdottir, G., Gunnarsson, B., Suthers, I.M., 2000. Spatial variation in hatch date distributions and origin of pelagic juvenile cod in Icelandic waters. ICES J. Mar. Sci. 57, 1182–1195. Mazhirina, G.P., 1988. Sex cycle of the barents sea catfishes: the biology of the fish in the seas of the European North. Trud. PINRO, 151–162. McCusker, M.R., Paterson, I.G., Bentzen, P., 2008. Microsatellite markers discriminate three species of North Atlantic wolffishes (Anarhichas spp.). J. Fish Biol. 72, 375–385, http://dx.doi.org/10.1111/j.1095-8649.2007.01701.x. Morgan, M.J., Wright, P.J., Rideout, R.M., 2013. Effect of age and temperature on spawning time in two gadoid species. Fish. Res. 138, 42–51, http://dx.doi.org/ 10.1016/j.fishres.2012.02.019. Morgan, M.J., 2001. Time and location of spawning of american plaice in NAFO divisions 3LNO. J. Northwest Atl. Fish. Sci. 29, 41–49. Oehlert, G.W., 1992. A note on the delta method. Am. Stat. 46, 27–29. Ojaveer, E., Arula, T., Lankov, A., Shpilev, H., 2011. Impact of environmental deviations on the larval and year-class abundances in the spring spawning herring (Clupea harengus membras L.) of the Gulf of Riga (Baltic Sea) in 1947–2004. Fish. Res. 107, 159–168, http://dx.doi.org/10.1016/j.fishres.2010. 11.001. Ottera, H., Agnalt, T.A., Kjesbu, O.S., Dahle, G., Jørstad, K., 2012. Is spawning time of marine fish imprinted in the genes? A two-generation experiments on local Atlantic cod (Gadus morhua L) populations from different geographical regions. ICES J. Mar. Sci. 69, 1722–1728. Pampoulie, C., Skirnisdottir, S., Danielsdottir, A.K., Gunnarsson, A., 2012. Genetic structure of the Atlantic wolffish (Anarhichas lupus L.) at Icelandic fishing
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409 grounds: another evidence of panmixia in Iceland? ICES J. Mar. Sci. 69, 508–515, http://dx.doi.org/10.1093/icesjms/fss017. Pavlov, D.A., Moksness, E., 1994. Production and quality of eggs obtained from wolffish (Anarhichas lupus L.) reared in captivity. Aquaculture 122, 295–312. Pavlov, D.A., Moksness, E., 1995. Development of wolffish eggs at different temperature regimes. Aquac. Int. 3, 315–335. Pavlov, D.A., Moksness, E., 1996. Repeat sexual maturation of wolffish (Arnarhichas lupus L.) broodstock. Aquaculture 139, 249–263. Pavlov, D.A., Novikov, G.G., 1993. Life history and peculiarites of common wolffish (Anarhichas lupus) in the White Sea. ICES J. Mar. Sci. 50, 271–277. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Found. for Statis Comput., Vienna, Austria https://www.R-project.org/. Ringø, E., Lorentsen, H., 1987. Brood protection of wolffish (Anarhichas lupus L.) eggs. Aquaculture 65, 239–241. Rose, G.A., Bradbury, I.R., DeYoung, B., Fudge, S.B., Lawson, G.I., Mello, L.G.S., Robichaud, D., Sherwood, G., Snelgrove, P.V.R., Windle, M.J.S., 2008. Rebuilding Atlantic Cod: Lessons from a Spawning Ground in Coastal Newfoundland, Resiliency of Gadid Stocks to Fishing and Climate Change. In: Kruse, G.H., Drinkwater, K., Ianelli, J.N., Link, J.S., Stram, D.L., Wespestad, V., Woodby, D. (Eds.). Alaska Sea Grant College Program, Fairbanks, http://dx.doi.org/10.4027/ rgsfcc.2008.11 (AK-SG-08-01). Schwarz, G., 1978. Estimating the dimension of a model. Ann. Stat. 6, 461–464. Scott, B.E., Marteinsdottir, G., Begg, G.A., Wright, P.J., Kjesbu, O.S., 2006. Effects of population size/age structure, condition and temporal dynamics of spawning on reproductive output in Atlantic cod (Gadus morhua). Ecol. Model. 191, 383–415, http://dx.doi.org/10.1016/j.ecolmodel.2005.05.015.
409
Sundby, S., Bjorke, H., Soldal, A., Olsen, S., 1989. Mortality rates during the early life stages and year-class strength of northeast Arctic cod (Gadus morhua L.). J. du Cons. Int. pour L’Exploration la Mer 191, 351–358. Templeman, W., 1986. Some biological aspects of atlantic wolffish (Anarhichas lupus) in northwest atlantic. J. Northwest Atl. Fish. Sci. 7, 57–65. Tukey, John, 1949. Comparing individual means in the analysis of variance. Biometrics 5 (2), 99–114. Tveiten, H., Johnsen, H.K., 1999. Temperature experienced during vitellogenesis Influences ovarian maturation and the timing of ovulation in common wolffish. J. Fish Biol. 55, 809–819. Tveiten, H., Solevåg, S.E., Johnsen, H.K., 2001. Holding temperature during the breeding season influences final maturation and egg quality in common wolffish. J. Fish Biol. 58, 374–385. Wang, N., Teletchea, F., Kestemont, P., Milla, S., Fontaine, P., 2010. Photothermal control of the reproductive cycle in temperate fishes. Rev. Aquac. 2, 209–222, http://dx.doi.org/10.1111/j.1753-5131.2010.01037.x. Wright, P.J., Gibb, F.M., 2005. Selection for birth date in North Sea haddock and its relation to maternal age. J. Anim. Ecol. 74, 303–312, http://dx.doi.org/10.1111/ j.1365-2656.2004.00924.x. Zemeckis, D.R., Dean, M.J., Cadrin, S.X., 2014. Spawning dynamics and associated management implications for atlantic cod. North Am. J. Fish. Manag. 34, 424–442, http://dx.doi.org/10.1080/02755947.2014.882456.
Contents lists available at ScienceDirect
Fisheries Research journal homepage: www.elsevier.com/locate/fishres
Spatio-temporal variation in the reproduction timing of Atlantic Wolffish (Anarhichas lupus L) in Icelandic waters and its relationship with size Ásgeir Gunnarsson ∗ , Höskuldur Björnsson, Bjarki Elvarsson, Christophe Pampoulie Marine Research Institute, Skúlagata 4, P.O box 1390, 121 Reykjavík, Iceland
a r t i c l e
i n f o
Article history: Received 9 March 2016 Received in revised form 1 July 2016 Accepted 3 July 2016 Handled by Prof. George A. Rose Available online 18 July 2016 Keywords: Spawning time Length Spatio-temporal differences Iceland Temperature
a b s t r a c t Biological data were retrieved from 3694 female Atlantic wolffish Anarhichas lupus collected in Icelandic waters at four locations during the breeding season using a long–term study spanning from 2002 to 2013. The main objective was to investigate reproduction investment and timing. In the main spawning ground, little temporal differences were observed. In contrast, the peak of the spawning season was different among spawning grounds, suggesting spatial differences in the timing of reproduction. The size of females A. lupus was related to the spawning time with the larger fish spawning earlier than smaller ones. In addition, no significant pattern was found between temperature and spawning time. Except at its main spawning area where spawning begins in late September, spawning usually began in late August or beginning of September and was completed in early November. © 2016 Elsevier B.V. All rights reserved.
1. Introduction In temperate environments, seasonal differences in sea temperature and zooplankton abundance will typically affect the timing of reproduction of many fish. Spawning time, which is cyclical for most temperate fish species, is one of the main features affecting survival and growth of fish progeny, as favourable environments are likely to positively affect both attributes. Mortality is the highest for early life stages and has been related to temperature, food availability and predation (Houde, 1997). Temperature, which is one of the most important abiotic factors for the survival of the larvae (Houde, 1989), is also known to influence growth, mortality and juvenile fish size. It is also generally acknowledged that survival rate of larvae is higher when hatching synchronises to the zooplankton spring bloom and is asynchronous when predators are in highest abundance (Hjort, 1914; Cushing, 1969). Studies have also shown positive relationships between survival ratios of larvae or juvenile fish during their first year and year class strength for several fish species (Campana et al., 1989; Carscadden et al., 2013; Ojaveer et al., 2011; Sundby et al., 1989). The timing of the reproduction is therefore crucial for a successful recruitment
∗ Corresponding author. E-mail address: [email protected] (Á. Gunnarsson). http://dx.doi.org/10.1016/j.fishres.2016.07.002 0165-7836/© 2016 Elsevier B.V. All rights reserved.
and consequently for the stock productivity and sustainability. Reproductive timing has been related to temperature, photoperiod, food availability, demography and genetic diversity, inter-annual variation and differences between locations, and differences in spawning time (Bromage et al., 2001; Kjesbu, 1994; Lapolla and Buckley, 2005; Marteinsdottir et al., 2000; Morgan, 2001; Morgan et al., 2013; Ottera et al., 2012). The Atlantic wolffish Anarhichas lupus L. 1758 is widely distributed in the North Atlantic Ocean and is an important commercial species. Within its distribution, the spawning time of A. lupus varies considerably from September to October in Icelandic waters to late July-late September in the White Sea, September in Norwegian waters and autumn in Canadian waters (Jónsson, 1982; Templeman, 1986; Falk-Petersen et al., 1990; Pavlov and Novikov, 1993; Gunnarsson et al., 2006). In Iceland, spawning of A. lupus has been observed all around the country (Fig. 1), but a relatively small (2000 km2 ) area west of the country is considered to be the main spawning area. A. lupus exhibits a rather unusual reproductive strategy; the female coils around the eggs after spawning, and creates a demersal egg cluster that is later guarded by the male during the incubation period which is c. 800–1000 ◦ C-days (cumulative days temperature) (Keats et al., 1985; Ringø and Lorentsen, 1987; Pavlov and Moksness, 1995). In Icelandic waters, the fishery effort using bottom trawl has increased considerably at the main spawning ground during
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
different temperature regimes. The area between areas 2 and 3 was excluded from the analysis due to a lack of data (Fig. 1).
68° Area 3
Area 2 ●
●
67°
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
66°
Area 1
●
●
65°
●
●
●
●
●
●
●
1 10 20 30 40 50
●
●
●
●
● ●
●
●
●
●
●
●
●
64°
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
● ●
● ●
●
● ●
●
2.2. Sampling
● ●
● ●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
● ●
●
●
●
●
●
● ●
● ●
63° Area 4
62° 28°
26°
24°
405
22°
20°
18°
16°
14°
12°
10°
Fig. 1. Spawning () of Atlantic wolffish at Iceland, according to AGFS. The () represents stations where female A. lupus were found near spawning or recently spawned, the size of the circles denote number of such fish, (+) represents station where no such females were found.
spawning time from 1999 onwards, as the commercial value of A. lupus is highest during this period (S. Pétursson, pers. comm.). Before that, the A. lupus fishery was mainly comprised of longliners after the incubation period, outside the spawning areas. In the year 2002, in an attempt to improve the recruitment of A. lupus, part of the main spawning ground was closed during spawning and incubation time (area 1, Fig. 1). As the data and knowledge on spawning of A. lupus was scarce, sampling of data on spatialtemporal distribution of spawning in the area was increased. Based on this sampling and logbooks from fishing vessels, the size of the closed area was further extended in 2010 from 500 km2 to 1000 km2 and the closure period was extended from 1st of October – 1st of May to 15th of September – 1st of May. A. lupus abundance has declined drastically during recent years, especially in the north-west Atlantic Ocean where it was listed by the Canadian Species at Risk Act (SARA) as a species of ‘special concern’ (McCusker et al., 2008). In Icelandic waters, the recruitment was good from 1993 to 1999 but has decreased since then, and been at low level (40% of the highest observed values) from 2010 to 2014. The Marine Research Institute (MRI) of Iceland has given advice based on Maximum Sustainable Yield (MSY) since 2001 but fishing activities exceeded the advised catch for a number of years. However, since 2013, catches have been in accordance with the Total Allowable Catches (TAC) recommended by the MRI (MRI 2015). Concurrently, the fishable stock and the spawning stock have been stable and even increased slightly despite the relatively low recruitment. Nowadays, the trend in A. lupus spawning time is regularly monitored at the main spawning ground to estimate the benefit of the closure period. To observe possible influence of fisheries and environment on spawning time of A. lupus, we investigated potential relationships between the reproduction timing, size of the females, and temperature, at the main spawning ground as well as three other areas where this species is known to spawn. We also assessed spatio-temporal differences in spawning trends within these different spawning areas. 2. Materials and methods 2.1. Sampling areas Sampling areas were chosen to represent both the main spawning area of A. lupus (area 1 in Fig. 1) and other spawning areas with
Biological samples of female A. lupus were collected during 12 consecutive years (2002–2013) from four different areas in Icelandic waters (Fig. 1). All females caught were sampled during the autumn ground fish survey (AGFS) conducted by the MRI from late September to October, using a bottom trawl with 40 mm square mesh codend. In addition, samples were taken for stock assessment purposes from commercial landings from July to December to fully cover the spawning time of the species (Table 1 and Fig. 1). For the analysis of spatial distribution of spawning, only data from the AGFS were used except in 2011 and 2012. In the former year, the survey could not be completed correctly for technical reasons and in the latter, the data collected from areas 3 and 4 were biased. These data were consequently excluded from all analysis. During sampling, the total length (cm) of each female was measured and maturity stages determined visually in the field. The seabed temperature was measured with Scanmar thermometer which were located on the headline of the trawl and calibrated with Starmon mini temperature recorders before cruise.
2.3. Maturity Maturity stages were determined according to the maturity scale of Barsukov (1959), revised by Mazhirina (1988). Only two maturity stages were considered during this study, i.e. stage 3 which refers to fish that intend to spawn during the present spawning season and stage 4 which refers to fish that have recently completed spawning or were recovering after spawning (See Gunnarsson et al., 2006 for a full description).
2.4. Statistical analysis To investigate the effects of female size and spatio-temporal on the time of spawning a multivariate logistic regression model (Cox, 1958) of the form: paldy =
1
1 + exp −ˇald + ay
was estimated. Here paldy is the estimated proportion of fish that has spawned in area at length l, day of the year d, year y and ˇald = ˛a + ˇ1 d + ˇ2 l + ˇa3 d + ˇa4 l + ˇ5 ld + ˇa6 ld which represents the effect covariates and ay the year—area interaction used as a proxy for the spatio-temporal variability. Here the emphasis is on inferring the time of spawning, and its inter-annual variability, and therefore the year—area interaction were treated as a random effect. The parameters of the model were estimated using restricted maximum likelihood (as described in Bates et al., 2015). Variables were selected to the model by eliminating, using an exhaustive search, those variable combinations that contributed the greatest increase to the Bayesian Information Criterion score or BIC (Schwarz, 1978), resulting in a reduced model where the BIC score could not be improved. Models were considered comparable if the difference in the BIC score was less than 2, as suggested by Kass and Raftery (1995). 2 tests were used to compute variable significance in the reduced model.
406
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
Table 1 Mean of total length (±s.d., cm) for each year of female Atlantic wolffish sampled in area 1 and for each areas. n, depicts the number of female measured each year and area. Years
Areas
Year
L
2002 2003 2004 2005 2006 2007
67.4 64.0 67.7 68.5 66.3 69.5
± ± ± ± ± ±
7.63 9.66 9.11 8.47 8.93 10.25
n
Year
222 82 121 199 110 146
2008 2009 2010 2011 2012 2013
L 66.1 67.8 68.2 66.3 67.9 69.3
± ± ± ± ± ±
9.09 8.96 7.19 9.30 9.11 7.73
The results from the model estimation procedure were used to determine ST50 , i.e. the day of the year for which 50% of the fish had spawned, which was determined based on the following formula:
− ˛a + ˇ2 l + ˇa4 l + ay
3.1. Spawning distribution
L
n 67.6 61.3 61.0 67.4
± ± ± ±
8.86 12.02 16.01 11.58
2269 545 225 655
270
ˇ1 + ˇa3 + ˇ5 l + ˇa6 l
3. Results
1 2 3 4
280
To determine if estimated ST50 were different between years, areas or lengths at 95% confidence intervals (CIs) were constructed based on Normal approximation, where the variance was calculated using the delta method (as described in Oehlert, 1992), based on the estimated covariance from the model. Post-hoc differences between years or areas were subsequently determined based on the approach described by Tukey (1949) to account for multiple testing. The main spawning area (area 1) was the only investigated to assess the temporal variation of ST50 as data from other areas were considered insufficient to compare ST50 between years. Therefore the annual parameter estimates from the model were pooled (2002–2013) for each of the areas (2–4), and compared with pooled data from the main spawning ground (area 1). Linear regressions were used to examine relationship of seabed temperature and ST50 as temperature measurements were only available for AGFS samples. All statistical analyses were performed using R (version 3.2.2, see R Core Team, 2015 for further details).
Areas
254 220 218 137 460 100
a)
260 250 2004
2007
2010
2013
Year
b) 280
Day of the year
ST 50,ay =
n
270 260 250 60
70
80
90
3
4
Length (cm)
c) 280 270 260 250 1
2
Area
According to AGFS, spawning of A. lupus was widespread. However its main spawning ground was located West of Iceland in area 1 (Fig. 1). The number of females on maturity stages 3 and 4 was 242, 154, 85 and 90 in areas 1–4 respectively, which means 42%, 27%, 15% and 16% of the total number registered. These results indicate that area 1 was the most important area for spawning. 3.2. Model selection The model selection for spawning time five models were deemed equivalent with the lowest BIC value, 1) the full model, 2) the model omitting ˇa6 , 3) the model omitting ˇa6 and ˇ5 4) the model omitting ˇa4 , ˇ5 andˇa6 , and 4) the model omitting ˇa3 , ˇ5 andˇa6 . Although model 3 had the lowest BIC score (BIC ≈ 3 to other four models) model 4 was chosen as an interaction term between length and area, i.e. ˇa4 , was considered to be related to variations in sampling location between years within areas not actual differences in spawning time. 3.3. Spawning time In the main spawning ground (area 1), estimated ST50 did not differ significantly between years (p > 0.05) (Fig. 2a and Table 2). The range of ST50 was about seven days, the earliest was on day 271 in 2004 and the latest on day 278 in 2006. Based on ST50, a significant difference was only observed in spawning time between
Fig. 2. Estimated ST50 in area 1 by (a) length and (b) areas.
area 1 and 3 (p = 0.012, Fig. 2c and Table 2). Comparing spawning ogives between areas revealed significant interaction between days and areas (2 (3) = 60.299, p < 0.001, Fig. 3b) where shape of the spawning ogive differed significantly in areas 2 and 3 from 1 and 4 (2 (2) = 58.39, p < 0.001). Spawning was first observed on day of the year 260, 234, 239 and 246 in areas 1–4 respectively and the duration of the spawning time was 48, 57, 65 and 59 days respectively, or about two months for all areas except area 1 where spawning lasted one and a half months. For all areas the spawning was completed in late October or early November. 3.4. Fish length and spawning time In the main spawning ground (area 1), the interaction term between length of spawner and day of year was estimated to offer a non-significant improvement to the model (2 (1) = 0.195, p = 0.659). The proportion of females that had spawned varied significantly both as a function of length and day of year with larger fish spawning earlier (2 (1) = 58.71, p < 0.001 and 2 (1) = 2847.87, p < 0.001 respectively). Fig. 3a illustrates the estimated effect of length and day proposed by the model for area 1. In addition Table 2 and Fig. 2b shows the estimated ST50 from the model as a function of length.
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
407
Table 2 Estimate of the day of the year (±s.e.), by years and length (cm) in area 1 and for all areas. Years Year 2002 2003 2004 2005 2006 2007
Length ST50 276 274 271 273 278 274
± ± ± ± ± ±
1.7 1.7 1.8 1.7 1.8 1.7
Date
Year
1. Oct. 30. Sep. 27. Sep. 29. Sep. 3. Oct. 30. Sep.
2008 2009 2010 2011 2012 2013
ST50 272 272 275 275 274 278
± ± ± ± ± ±
1.7 1.7 1.7 1.7 1.7 1.8
Date
Length
28. Sep. 28. Sep. 1. Oct. 1. Oct. 29. Sep. 4. Oct.
60 70 80 90
Fig. 3. Estimated spawning ogives by day of the year and (a) length in area 1 and (b) areas, based on prediction for 70 cm fish from logistic regression.
3.5. Temperature The mean seabed temperature registered from AGFS was 7.38 ◦ C, 6.51 ◦ C, 5.33 ◦ C and 8.05 ◦ C for areas 1–4 respectively. The mean seabed temperature was highest in area 4 and lowest in area 3; the difference between these two areas was about 2.7 ◦ C. The variability of temperature in the main spawning ground (area 1) was relatively small. The average for the period 2002–2013 was 7.41 ◦ C and standard deviation 0.42 ◦ C. The lowest value was 6.77 ◦ C in 2002 and highest 8.15 ◦ C in 2008. This limited range makes it impossible to infer about the relationship between estimated ST50 and temperatures (p > 0.05). Indeed, the linear regression analysis did not reveal any significant relationship between the seabed temperature and estimated ST50 between years in the main spawning ground (area 1; p > 0.05) nor between seabed temperature and estimated ST50 when areas were compared (p > 0.05). 4. Discussion Despite the status of A. lupus as an endangered species and its important economic value in the North Atlantic, data on the reproduction of this species which are crucial for sustainable management are limited. The present study therefore aimed at
Areas ST50 280 276 271 267
± ± ± ±
1.8 1.7 2.3 4.5
Date
Area
5. Oct. 1. Oct. 27. Sep. 23. Sep.
1 2 3 4
ST50 274 268 263 265
Date ± ± ± ±
1.7 3.7 3.5 4.7
30. Sep. 23. Sep 18. Sep. 20. Sep.
investigating the impact of fish size and temperature on the spawning time of A. lupus, as well as spatio-temporal differences in duration of the reproduction and timing in different Icelandic spawning areas. The results presented here revealed that variability in spawning time is relatively small in the main spawning ground (area 1). In addition, larger fish spawn earlier than the smaller ones. For several fish species, cod included, it has been shown that the larger individual spawn earlier than the smaller (Carscadden et al., 1997; Lambert, 1987; Marteinsdóttir and Petursdottir, 1995; Wright and Gibb, 2005). Further, cod studies have demonstrated that fishing effort might lead to drastic changes in spawning time and duration by catching more larger than smaller fish in spawning aggregation (Rose et al., 2008; Scott et al., 2006; Zemeckis et al., 2014). Relationships could not be found between the seabed temperature and the reproduction time of A. lupus. The observed lack of differences in the peak of spawning (ST50 ) between years in the main spawning area of A. lupus (area 1) could be explained by the fact that temperature differences from year to year were too small to affect the reproduction pattern of the species or that temperature trajectories were similar between years. Indeed, if several studies have shown that higher temperature during the vitellogenesis or during the spawning season will delay the spawning of A. lupus (Tveiten and Johnsen, 1999; Tveiten et al., 2001), temperatures trajectories might also be important. Evidence showed that A. lupus reproducing at around 8 ◦ C will delay the peak of spawning of approximately 9 days compared to A. lupus reproducing at 4 ◦ C (Tveiten et al., 2001). On the other hand, differences of 2–5 ◦ C in temperatures did not affect the peak of spawning of A. lupus during two consecutive years (Pavlov and Moksness, 1996). One important feature when comparing the two aforementioned studies remains the direction of temperature changes which affect maturation and spawning time (Wang et al., 2010). In the former case (Tveiten et al., 2001), temperature changes were abrupt while in the latter (Pavlov and Moksness, 1996) temperature trajectories were similar between years and changes gradual. Comparing the peak of spawning (ST50 ) between spawning grounds revealed that ST50 was about two weeks later in warmer waters of area 1 than in colder waters of areas 2 and 3, which is a much more pronounced delay than the one observed by Tveiten et al. (2001) for a lower variability of temperature. Although constant light regimes have been shown to protract the spawning season of A. lupus to 9–10 months (Pavlov and Moksness, 1994), it did not seem to affect spawning time during our study as spawning times were different between locations despite a similar photoperiod. The most striking results were the observed stability of spawning in area 1 and its delay compared to spawning in other areas such as area 3, which might be explained by the availability of food following the hatching of larvae. The spawning season was about 2 months in all areas, except in area 1 where it was about 1.5 months, a result in accordance with previous observation from the White Sea (Pavlov and Novikov, 1993). Indeed, larvae that hatched from eggs spawned in the first half of September might have a lower
408
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409
survival rate due to unfavourable environmental food conditions in area 1. Another possible explanation is that egg size greatly affects offspring fitness as well as larval survival since larger eggs usually results in larvae with higher energy resources (Brooks et al., 1997). By delaying spawning for two weeks, females from area 1 might produce higher quality eggs and ensure higher survival rate of their offspring. This could explain why area 1 is the largest spawning area for A. lupus, and exhibited the largest effective population size (Pampoulie et al., 2012). What favours the area 1 as a spawning ground for A. lupus is currently unknown but factors such as temperature, currents and substrate might be important. Recent research on the substrate quality in this area shows a relatively flat bottom characterised by glacial remains and iceberg scouring, localised at a water depth of 150–160 m. The thin sediment cover is mostly gravelly sand with dispersed finer material. Video records show layers of crust which are regarded as appropriate nesting holes for A. lupus. The origin of the crust is still under consideration (G. Helgadottir, pers. comm.). Recent studies on tagging at area 1 (Gunnarsson et al., unpublished) and genetic structure within Icelandic waters have nevertheless revealed contradicting results, i.e. a high degree of spawning ground fidelity and an absence of genetic structure (Pampoulie et al., 2012). Despite the absence of genetic structure at neutral markers, A. lupus seemed to exhibit life history traits variation within Icelandic waters (Gunnarsson et al., 2006) which might be important to consider for a proper management. Apart from the main spawning ground some uncertainty is related to the definition of the spawning areas used in this study. Although spawning females occurred in these areas, the sampling was coarser outside of the main spawning grounds, in particular in the southern area, and some practical considerations, such the number of available samples, were taken into account when defining the three other spawning areas. Thus a finer spatial structure, such as two spawning grounds within a study area, cannot be ruled out which may explain variation in the shape of the spawning ogive between areas 1 and 3, versus 2 and 4. During the present study, the relatively large sampling effort in the area 1 was implemented to find the optimal closure time, which could protect the fish during the incubation period without compromising the fishery when the price for the fish is highest on the market. Therefore, the correct “closure time” would typically start when spawning events are observed. Although the fish begin to aggregate in this area beginning of August, the “closure time” was thus based on the lack of observation of spawned individual before the 20th of September. The current start of “closure time” (day 260; 15th of September) seems to fit well with the attempt to protect males in their nest. The observed stability of the spawning time in this area during this study makes it possible to select one specific “closure time” that does not have to be re-evaluated every year, but the timing of the spawning event will need regular monitoring. A similar compromise was made with regard to the size of the closed area which now constitutes about 2/3 of the main spawning ground. Concurrently to these “protective” measures, TAC has decreased since 2006 and catches followed the advice from the MRI from 2013 onwards. Despite all these efforts to improve the recruitment of A. lupus, the recruitment index seems to remain stable (low level from 2011 onwards) and although the “closure time” of the main spawning ground seems suitable, further research is needed to potentially understand the factors at stake.
Acknowledgements We would like to thank Mary Frances and two anonymous reviewers for their constructive comments and efforts to improve the manuscript.
References Barsukov, V.V., 1959. Wolffish Family (Anarhichadidae). Indian National Documentation, New Delhi. Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48. Bromage, N., Porter, M., Randall, C., 2001. The environmental regulation of maturation in farmed finfish with special reference to the role of photoperiod and melatonin. Aquaculture 197, 63–98, http://dx.doi.org/10.1016/S00448486(01)00583-X. Brooks, S., Tyler, C.R., Sumpter, J.P., 1997. Egg quality in fish: what makes a good egg? Rev. Fish Biol. Fish. 7, 387–416. Campana, S.E., Frank, K.T., Hurley, P.C.F., Koeller, P.A., Page, F.H., Smith, P.C., 1989. Survival and abundance of young atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) as indicators of year-class strength. Can. J. Fish. Aquat. Sci. 46, 171–182. Carscadden, J., Nakashima, B.S., Frank, K.T., 1997. Effects of fish length and temperature on the timing of peak spawning in capelin (Mallotus villosus). Can. J. Fish. Aquat. Sci. 54, 781–787, http://dx.doi.org/10.1139/cjfas-54-4-781. Carscadden, J.E., Gjøsæter, H., Vilhjálmsson, H., 2013. Recruitment in the Barents Sea Icelandic, and eastern Newfoundland/Labrador capelin (Mallotus villosus) stocks. Prog. Oceanogr. 114, 84–96, http://dx.doi.org/10.1016/j.pocean.2013. 05.006. Cox, D.R., 1958. The regression analysis of binary sequences. J. R. Stat. Soc. B 20, 215–242. Cushing, D., 1969. The regularity of the spawning season of some fishes. Extr. Du J. du Cons. Int. pour L’exploration la Mer. 33, 81–92. Falk-Petersen, I.B., Haug, T., Moksness, E., 1990. Oservations on the occurrence, sise, and feeding of pelagic larvae of the common wolffish (Anarhichas lupus) in western Finnmark, northern Norway. J. Du Cons. 46, 148–154. Gunnarsson, Á., Hjörleifsson, E., Thórarinsson, K., Marteinsdóttir, G., 2006. Growth, maturity and fecundity of wolffish Anarhichas lupus L. in Icelandic waters. J. Fish Biol. 68, 1158–1176, http://dx.doi.org/10.1111/j.1095-8649.2006.00990. Hjort, J., 1914. Fluctuations in the great fisheries of Northern Europe viewed in the light of biological research. Rapp. Procès-Verbaux des Réunions du Cons. Int. pour l’Exploration la Mer. 20, 1–228. Houde, E.D., 1989. Comparative growth, mortality, and energetics of marine fish larvae: temperature and implied latitudinal effects. Fish. Bull. 87, 471–495. Houde, E., 1997. Patterns and consequences of selective processes in teleost early life histories. In: Chambers, C., Trippel, E.A. (Eds.), Early Life History and Recruitment in Fish Populations. Chapman and Hall, Padstow, pp. 173–196. Jónsson, G., 1982. Contribution to the biology of catfish (Anarhichas lupus) at Iceland. J. Mar. Res. Inst. Reykjavik 4, 3–26. Kass, R., Raftery, A., 1995. Bayes factors. J. Am. Stat. Assoc. 90, 773–795. Keats, D.W., South, G.R., Steele, D.H., 1985. Reproduction and egg guarding by Atlantic wolffish (Anarhichas lupus: Anarhichidae) and ocean pout (Marcozoarces americanus: Zoarcidae) in Newfoundland waters. Can. J. Zool. 63, 2565–2568. Kjesbu, O.S., 1994. Time of start of spawning in Atlantic cod (Gadus morhua) females in relation to vitellogenic oocyte diameter, temperature, fish length and condition. J. Fish Biol. 45, 719–735. Lambert, T., 1987. Duration and intensity of spawning in herring Clupea harengus as related to the age structure of the mature population. Mar. Ecol. Prog. Ser. 39, 209–220, http://dx.doi.org/10.3354/meps039209. Lapolla, A., Buckley, L., 2005. Hatch date distributions of young-of-year haddock Melanogrammus aeglefinus in the Gulf of Maine/Georges Bank region: implications for recruitment. Mar. Ecol. Prog. Ser. 290, 239–249, http://dx.doi. org/10.3354/meps290239. G. Marteinsdóttir, G. Petursdottir, 1995. Spatial and temporal variation in reproduction of icelandic cod at selvogsbanki and nearby costal areas ICES C. 1995/G15. Marteinsdottir, G., Gunnarsson, B., Suthers, I.M., 2000. Spatial variation in hatch date distributions and origin of pelagic juvenile cod in Icelandic waters. ICES J. Mar. Sci. 57, 1182–1195. Mazhirina, G.P., 1988. Sex cycle of the barents sea catfishes: the biology of the fish in the seas of the European North. Trud. PINRO, 151–162. McCusker, M.R., Paterson, I.G., Bentzen, P., 2008. Microsatellite markers discriminate three species of North Atlantic wolffishes (Anarhichas spp.). J. Fish Biol. 72, 375–385, http://dx.doi.org/10.1111/j.1095-8649.2007.01701.x. Morgan, M.J., Wright, P.J., Rideout, R.M., 2013. Effect of age and temperature on spawning time in two gadoid species. Fish. Res. 138, 42–51, http://dx.doi.org/ 10.1016/j.fishres.2012.02.019. Morgan, M.J., 2001. Time and location of spawning of american plaice in NAFO divisions 3LNO. J. Northwest Atl. Fish. Sci. 29, 41–49. Oehlert, G.W., 1992. A note on the delta method. Am. Stat. 46, 27–29. Ojaveer, E., Arula, T., Lankov, A., Shpilev, H., 2011. Impact of environmental deviations on the larval and year-class abundances in the spring spawning herring (Clupea harengus membras L.) of the Gulf of Riga (Baltic Sea) in 1947–2004. Fish. Res. 107, 159–168, http://dx.doi.org/10.1016/j.fishres.2010. 11.001. Ottera, H., Agnalt, T.A., Kjesbu, O.S., Dahle, G., Jørstad, K., 2012. Is spawning time of marine fish imprinted in the genes? A two-generation experiments on local Atlantic cod (Gadus morhua L) populations from different geographical regions. ICES J. Mar. Sci. 69, 1722–1728. Pampoulie, C., Skirnisdottir, S., Danielsdottir, A.K., Gunnarsson, A., 2012. Genetic structure of the Atlantic wolffish (Anarhichas lupus L.) at Icelandic fishing
Á. Gunnarsson et al. / Fisheries Research 183 (2016) 404–409 grounds: another evidence of panmixia in Iceland? ICES J. Mar. Sci. 69, 508–515, http://dx.doi.org/10.1093/icesjms/fss017. Pavlov, D.A., Moksness, E., 1994. Production and quality of eggs obtained from wolffish (Anarhichas lupus L.) reared in captivity. Aquaculture 122, 295–312. Pavlov, D.A., Moksness, E., 1995. Development of wolffish eggs at different temperature regimes. Aquac. Int. 3, 315–335. Pavlov, D.A., Moksness, E., 1996. Repeat sexual maturation of wolffish (Arnarhichas lupus L.) broodstock. Aquaculture 139, 249–263. Pavlov, D.A., Novikov, G.G., 1993. Life history and peculiarites of common wolffish (Anarhichas lupus) in the White Sea. ICES J. Mar. Sci. 50, 271–277. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Found. for Statis Comput., Vienna, Austria https://www.R-project.org/. Ringø, E., Lorentsen, H., 1987. Brood protection of wolffish (Anarhichas lupus L.) eggs. Aquaculture 65, 239–241. Rose, G.A., Bradbury, I.R., DeYoung, B., Fudge, S.B., Lawson, G.I., Mello, L.G.S., Robichaud, D., Sherwood, G., Snelgrove, P.V.R., Windle, M.J.S., 2008. Rebuilding Atlantic Cod: Lessons from a Spawning Ground in Coastal Newfoundland, Resiliency of Gadid Stocks to Fishing and Climate Change. In: Kruse, G.H., Drinkwater, K., Ianelli, J.N., Link, J.S., Stram, D.L., Wespestad, V., Woodby, D. (Eds.). Alaska Sea Grant College Program, Fairbanks, http://dx.doi.org/10.4027/ rgsfcc.2008.11 (AK-SG-08-01). Schwarz, G., 1978. Estimating the dimension of a model. Ann. Stat. 6, 461–464. Scott, B.E., Marteinsdottir, G., Begg, G.A., Wright, P.J., Kjesbu, O.S., 2006. Effects of population size/age structure, condition and temporal dynamics of spawning on reproductive output in Atlantic cod (Gadus morhua). Ecol. Model. 191, 383–415, http://dx.doi.org/10.1016/j.ecolmodel.2005.05.015.
409
Sundby, S., Bjorke, H., Soldal, A., Olsen, S., 1989. Mortality rates during the early life stages and year-class strength of northeast Arctic cod (Gadus morhua L.). J. du Cons. Int. pour L’Exploration la Mer 191, 351–358. Templeman, W., 1986. Some biological aspects of atlantic wolffish (Anarhichas lupus) in northwest atlantic. J. Northwest Atl. Fish. Sci. 7, 57–65. Tukey, John, 1949. Comparing individual means in the analysis of variance. Biometrics 5 (2), 99–114. Tveiten, H., Johnsen, H.K., 1999. Temperature experienced during vitellogenesis Influences ovarian maturation and the timing of ovulation in common wolffish. J. Fish Biol. 55, 809–819. Tveiten, H., Solevåg, S.E., Johnsen, H.K., 2001. Holding temperature during the breeding season influences final maturation and egg quality in common wolffish. J. Fish Biol. 58, 374–385. Wang, N., Teletchea, F., Kestemont, P., Milla, S., Fontaine, P., 2010. Photothermal control of the reproductive cycle in temperate fishes. Rev. Aquac. 2, 209–222, http://dx.doi.org/10.1111/j.1753-5131.2010.01037.x. Wright, P.J., Gibb, F.M., 2005. Selection for birth date in North Sea haddock and its relation to maternal age. J. Anim. Ecol. 74, 303–312, http://dx.doi.org/10.1111/ j.1365-2656.2004.00924.x. Zemeckis, D.R., Dean, M.J., Cadrin, S.X., 2014. Spawning dynamics and associated management implications for atlantic cod. North Am. J. Fish. Manag. 34, 424–442, http://dx.doi.org/10.1080/02755947.2014.882456.