© 2005 International Council for the Exploration of the Sea
Comparison of performance of two size groups of farmed cod (Gadus morhua L.) juveniles following transfer to sea cages
North Atlantic Fisheries College Port Arthur, Scalloway, Shetland, ZE1 OUN, Scotland, United Kingdom
*Correspondence to L. McEvoy: tel: +44 1950 460023. e-mail: Lesley{at}johnsonseafarms.com.
In order to investigate the biologically and economically optimum size for sea transfer of cultured cod, replicated trials were undertaken using experimental cages (1.5 m x 1 m x 1 m) stocked with two size grades of cod juveniles previously "untested" for on-growing at sea in Shetland: "small grade" (SG), 9.3-g mean weight (±2.08 s.d.) and "large grade" (LG), 19.4-g mean weight (±4.48 s.d.). Survival was high in both grades tested, with no significant difference in overall mortality (7.3% (SG) and 8.4% (LG)). A significant difference was observed in the overall percentage growth per day (%SGR) (p = 0.02), with SG and LG exhibiting values of 1.3 and 1.08, respectively. Regression analysis of body weight gain over time between the two grades revealed a significantly higher proportional weight gain in SG (p = 0.01). Cost analysis revealed an initial saving of 8.5% by selecting SG juveniles. This saving was reduced to 2.9% when the cost of feeding these fish was taken into account. However, food wastage was higher in this study than would be expected in a commercial operation, suggesting that the actual saving for the on-grower could be greater than 2.9%. Selecting smaller sized juveniles for transfer will increase the growout period. However, this may be offset by the fact they may be transferred earlier, at a significantly lower price.
Keywords: Gadus morhua, growth, juvenile Atlantic cod, on-growing, survival
Received 13 June 2004; accepted 21 November 2005.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
Aquaculture in Scotland is dominated by salmon (Salmo salar L.) production. The farming of exclusively "marine" species, especially finfish, remains a small supplementary activity (Shields, 2001). However, changes in the economic climate, brought about by low salmon prices and the decline in wild cod (Gadus morhua L.) stocks, have caused farmers to begin to diversify into cod culture. Despite promising results from preliminary trials (Svåsand et al., 2004) and considerable interest being expressed by investors (Adoff et al., 2002), there remains a need to optimize certain protocols, especially at a regional level (Brown et al., 2003), both in larviculture and early on-growing.
Currently, a major impediment to the progress of cod aquaculture in Scotland and elsewhere is the lack of data regarding the optimum size of juveniles for transfer from hatcheries to on-growing sites at sea. Consequently, Scottish cod hatcheries have been forced to retain juveniles to a size of at least 30 g, loosely following the model of salmon smolt transfer. This has contributed to the reduction of production capacity and increased unit cost.
Cod hatcheries in the United Kingdom have been designed primarily to produce juveniles of 5 g or less for on-growing (P. Tarrant, pers. comm., NUFISH Ltd., Shetland). Therefore, their profitability depends on the implementation of this strategy. This strategy follows the model used for marine species such as sea bream (Sparus auratus L.) and sea bass (Dicentrarchus labrax L.). Existing production methods for these fish can be adapted for new species (Brown and Puvanendran, 2002), and it is widely believed that experience gained from salmon farming will be beneficial both in terms of management and technology in the on-growing of cod as well (Adoff et al., 2002; Engelsen et al., 2004).
Cod juveniles weighing less than 5 g have been stocked in ocean netpens off Newfoundland (King and Nardi, 2002), although the success of this transfer with respect to survival rate and growth was not documented. The applicability of a similar model to European cod farming is not known. In theory, it should permit an increase in juvenile production in existing hatcheries, lower the cost of juveniles, and increase the accessibility and profitability of cod farming. For the on-grower, this practice will only become justified when certain key biological (survival/growth) and economic (return/loss expected) concerns have been addressed.
Cod aquaculture tends to employ pre-existing salmon on-growing infrastructure and sites. Since most existing salmon on-growing sites in Scotland, and notably Shetland, are within areas of at least moderate wave/tidal exposure, on-growers have expressed serious concerns regarding the possibility of increased mortality in cod juveniles. Cod are described as having naturally low-speed swimming modes (Bjørnevik et al., 2003). Moderate levels of exercise stimulate growth in teleost fish; conversely, high speed exercise is thought to retard growth, although this is species dependent (Davison, 1997; Johnston, 1999). Prolonged exercise may be a factor affecting survival in post-transfer juveniles, and it is generally accepted that cod require lower energy sites than salmon.
The aim of this study was to investigate the biological performance and economic implications of stocking two smaller grades of juveniles into experimental cages in an effort to determine whether smaller grades of cod can be successfully transferred directly from the hatchery to sea sites or whether interim land-based nursery sites are required for cod up to 30 g.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Juvenile cod
A total of 3000 hatchery reared Atlantic cod juveniles was sourced from the Marine Hatchery of the North Atlantic Fisheries College (NAFC, Port Arthur, Scalloway, Shetland). All fish originated from NAFC's own locally sourced, captive broodstock. Fish were of the same age and from the same batch of juveniles. Two size grades of cod were chosen for the purposes of this trial, 1500 individuals per grade: a small grade (SG) with an average weight of 9.3 g (±2.08 s.d. (standard deviation)) and a large grade (LG) averaging 19.4 g (±4.48 s.d.). In all, 1500 SG cod were transferred on 15 October 2003 into five experimental cages, equivalent to 300 fish per replicate. This was repeated on 16 October for LG fish. Average initial stocking densities were 1.24 kg m–3 and 2.57 kg m–3 for SG and LG, respectively.
Experimental cages and study site
Ten small experimental cages were constructed for the purpose of this trial. Accessibility, the need for replication, and the need to witness feeding and behavioural characteristics were key considerations within the design objective. The frames, measuring 1.7-m length x 1.2-m depth x 1.7-m width, were constructed from stainless steel to provide rigidity for the internally suspended net. The net was made from 10-mm mesh and anti-fouled, with dimensions of 1.5-m length x 1-m depth x 1.5-m width, equivalent to a total net volume of 2.25 m3 (Figure 1). The nets were produced by Shetland Net Services (Blacksness Pier, Scalloway).
|
The site chosen at Johnson Seafarms Ltd (Vidlin, Shetland) is considered a moderate energy site with a depth range of 5–8 m. Mean near-surface current velocity measured 4.7 cm s–1. Secured in a random formation to two existing skeletal 10-m2 commercial cages, the experimental cages were suspended in an upright position in the water column approximately 1 m above the seabed. This strategy was employed to reduce possible contamination from the benthos and to minimize exposure to surface turbulence.
Experimental design
The trial was carried out between 15 October and 12 December 2003, a total of eight weeks. Individual cages were slowly hauled to the surface each morning by hand until the top netting breached the surface. This allowed control over feeding and access for daily monitoring (mortality, temperature, feeding, behaviour, visual signs of disease, predation, and cannibalism). Weather permitting, the cages were kept at the surface during the day and slowly lowered each evening.
For reasons discussed later, fish were fed to 4% body weight per day for the first week, and to satiation from this point onwards. Behavioural response to feed was used as a measure of satiation. The size of feed followed manufacturer's guidelines and was recalculated fortnightly in light of updated growth data. The following grades of EWOS Marine Diet were used (diet name is followed by cod size guide as provided by EWOS Ltd): 005 (5–15 g), 010 (10–25 g), and 020 (20–80 g).
Sampling for growth rate parameters was undertaken fortnightly, randomly selecting 20 fish per replicate by hand. Fish were culled by a percussive blow to the head. The samples were observed immediately in the laboratory for general health, and then body weight (to the nearest g) and length were measured (total length, mm), and condition factor calculated. At the end of the trial, all remaining fish were counted out of each replicate and transferred to other cages. This information was used in conjunction with daily records to check for discrepancies or unaccountable losses.
Calculations and statistics
Condition was calculated according to Fulton's condition factor, K = 100 WL–3, W = wet body weight (g) and L = total length (cm). The specific growth rate (SGR) was calculated as SGR = (ln Wf – ln Wi)/t, where Wf and Wi represent the final and initial weight values, respectively, and t is the number of days between sampling points. The daily percentage increase in juvenile weight was then calculated as %SGR = (eSGR – 1)100% (Kjørsvik et al., 2004).
All statistical analyses were carried out using GraphPad Prism® (Version 3.0). Testing for coherence with the Gaussian or "Normal" distribution was carried out when applicable on data sets using the Kolmogorov–Smirnov (KS) test. Percentage data for survival and %SGR values required arcsine transformation to stabilize the variance (Fowler et al., 1998). An unpaired (two-tailed) t-test was used to compare the means of survival and the overall mean %SGR between SG and LG. Linear regression was used to compare the slopes of proportional weight gain (%) over the trial period between SG and LG, since actual weights are not comparative.
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Survival
No significant difference in survival was recorded (t = 0.19, p = 0.85), with overall mortality at 7.3% and 8.4% for SG and LG, respectively. All losses, with the exception of two individuals, were accounted for. This would suggest that juvenile cod grown in this manner, at these size grades, are not prone to cannibalism.
Values, drawn from diagnosis by visual appearance only, indicated no significant difference in the frequency of vibriosis as the causative agent of mortality between the size grades (Table 1). A sharp decline in Vibrio-related mortalities was witnessed after week 2 of the trial. Sustained, low levels of predation were prevalent after week 1, with both grades showing similar mortality as a result of attempted predation by birds (Table 1).
|
Growth
Analysis of variance between replicates of SG and LG revealed no significant differences. The end of trial mean weight of SG (19.2 g ± 5.0 s.d.) is close to that of the initial mean weight of LG (19.4 g ± 4.49 s.d.) (Figure 2). Further analysis reveals a very similar weight frequency distribution between the two sets of data (Figure 3). The daily percentage increases in juvenile weight (%SGR) values of both SG and LG were highly variable (Table 2). The daily percentage increases in juvenile weight values were low at the first sample point (14 days) with values of 0.93 (SG) and 1.07 (LG). A large increment in %SGR is witnessed in SG (SGR = 1.29) after the fourth week (28 days); this is far less pronounced in LG (SGR = 1.15). SG fish showed higher overall %SGR values than LG (t = 2.88, p = 0.02). As would be expected with higher %SGR values, SG juveniles showed higher proportional weight increments (F = 7.23, p = 0.01) after week 2.
|
|
|
Initial condition values (K) of 1.0 (SG) and 1.07 (LG), increased over the trial period with end of trial values of 1.06 and 1.10, respectively.
Cost analysis
The current commercial selling price of cod juveniles is approximately one pound (Sterling, £) for a fish of 1 g, plus one penny for every subsequent gramme (Tarrant, pers. comm.). Therefore, the unit costs of SG and LG fish are equal to £1.08 and £1.18, respectively. Basic cost analysis shows that buying juveniles at 9 g is 8.5% less expensive than at 19 g. At a trial level (1500 individuals), this revealed an initial saving of £150.00 by selecting SG over LG fish (Table 3). Deducting the cost of feed required to bring the smaller 9-g fish up to a weight of 19 g reduced this initial saving from 8.5% to 2.9%; refer to the costing below:
|
|
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The smaller (SG) cod performed as well as the larger LG fish in terms of survival and growth. This would suggest that on-growers may expect no growth or survival penalties in selecting cod <10 g for growout. By selecting smaller fish, the growout period will be extended, the duration of which will depend on a number of factors including temperature (Björnsson and Steinarsson, 2002), the initial size of cod juveniles, time of transfer with respect to daylight length, and feeding regime.
The "snapshot" of growth data taken from this trial does not allow us to extrapolate whether the SG cod would have maintained or lost their size deficit with respect to LG over an entire growout period; furthermore, smaller fish generally show higher growth rates, as witnessed in this study, making this calculation problematic. It is believed that this size deficit could be reduced by the optimization of transfer time (on-growers may be able to coordinate sea transfer of cod with optimal temperatures experienced locally), and feeding and grading protocols.
In this trial, cod transferred to sea at 9 g were not prone to higher levels of mortality than fish transferred at 19 g, when kept under identical conditions. On-growers have expressed concern over the susceptibility of juveniles of this size to Vibrio infection since at this stage they have not been fully vaccinated and are exposed to new environmental stressors. Initial mortality in both size groups as a result of vibriosis was high but declined rapidly after the first week of the trial, probably indicating that the fish had recovered from transfer stress. This trial generated only low numbers of mortalities that could be attributed to vibriosis. However, additional studies are needed to investigate the risk of disease during early transfer further.
Predation by birds was a direct result of not deploying anti-predator nets throughout this study. One would expect the values of predation to be much lower in a commercial operation employing anti-predator nets. Although not strictly representative of commercial conditions, small experimental fish cages, as employed by Qin et al. (2001), Vaccaro et al. (2004), and the present study, are useful to estimate performance criteria such as survival and growth of a species. However, within the cages deployed in this study, it is likely that the cod were subjected to increased levels of stress from daily hauling to the surface, exaggerated cage movement during bad weather, and higher interaction with predators than would be encountered in a commercial scale cage. Therefore, the levels of mortality should be viewed as overestimates compared with those expected in a commercial operation.
Concern over post-transfer cannibalism in smaller cod was unsubstantiated in this study. Significantly, however, during the first week of the trial, rationing of feed was inappropriate as a result of feed loss caused by wave energy and the small size of cages used. Optimal feeding regimes in seapens for juveniles of a size used in this study are not known. The strategy of over-feeding immediately after transfer to sea cages is common in salmon smolts to aid recovery and boost growth (Sales and Cumming, pers. comm., North Atlantic Fisheries College, Shetland). In the case of this trial, feeding to satiation was implemented to reduce potential cannibalism and to provide baseline growth and survival data. Size variation rarely remains small in a population of fish over time, as witnessed at the end of trial size distribution of LG fish (Figure 3). Cannibalism in cod may be severe in cohorts exhibiting large size variation, especially in inadequately fed groups (Folkvord, 1991). Ensuring adequate food distribution within a sea cage environment of "nursery phase" cod may be difficult, and cannibalism resulting from sub-optimal feeding strategies may ensue (Svåsand et al., 2004). There is an evident lack of research pertaining to optimal feeding strategies in post-transfer cod and, for the on-grower, this is of great concern. Optimal feeding regimes are widely viewed, therefore, as a necessary focus of research (Adoff et al., 2002). Further improvements in feeding regimes should help to reduce intra-population size variation, reduce cannibalism, and decrease grading frequency.
Cod in this trial were comparatively inactive feeders, unlike salmonid species. This likely reflects the different feeding behaviour of cod compared with salmonids and their lower levels of activity, rather than unsuitable culture conditions. From the results of this study and commercial observation, we propose that nursery phase cod may benefit from slow, controlled introductions of feed over lengthy periods with a bottom mesh, or a solid bottom net in parts that would permit grazing.
Condition values increased over the trial period. This improvement in the nutritional status or well-being of the fish (Grant et al., 1998; Wijekoon et al., 2003) would suggest that moderate exposure to wave energy is not detrimental to the survival or growth of cod of these sizes when fed to satiation. Under the environmental conditions experienced in this study, there appears to be no disadvantage in transferring small cod directly to sea, rather than employing interim, land-based nursery sites.
The economic appraisal of this trial revealed promising results. This study has shown that a significant initial saving may be realized by acquiring cod of a smaller size: in this instance a saving of 8.5% was realized by selecting SG over LG fish. After the cost of feeding was taken into account, this saving was reduced to 2.9%. This probably represents an underestimate, since feed wastage was particularly high, and no allowance was made for transportation savings. Therefore, this saving would probably be greater when applied to a commercial operation. For instance, twice the number of SG juveniles may be stocked for a given volume during transit, representing a significant saving on transportation alone.
As mentioned earlier, the growout period will be extended by stocking smaller fish. Furthermore, the on-grower will have to contend with additional costs related to labour, cage modifications, and vaccination costs. Irrespective of this, a significant initial saving may negate this extended on-growing period. In order to further validate this, collection and analysis of data from a full production cycle are required.
|
Acknowledgements |
|---|
We thank Johnson Seafarms Ltd for the provision of site and their keen participation throughout this trial. The research presented here was sponsored by the Colleges and Businesses in Partnership (CBP) scheme, a pilot initiative funded by the Highlands and Islands Enterprise and the Scottish Executive; and Shetland Fisheries Centre Ltd.
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Adoff G., Skjennum F.C., Engelsen R. (2002) Experience and prospects of Norwegian cod farming. Bulletin of the Aquaculture Association of Canada 102-1:8–11.
Bjørnevik M., Karlsen Ø., Johnston I.A., Kiessling A. (2003) Effect of sustained exercise on white muscle structure and flesh quality in farmed cod (Gadus morhua L.). Aquaculture Research 34:55–64.[CrossRef][Web of Science]
Björnsson B. and Steinarsson A. (2002) The food-unlimited growth rate of Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 59:494–502.
Brown J.A., Minkoff G., Puvanendran V. (2003) Larviculture of Atlantic cod (Gadus morhua): progress, protocols and problems. Aquaculture 227:357–372.[CrossRef][Web of Science]
Brown J.A. and Puvanendran V. (2002) Development of Atlantic cod larviculture in Newfoundland. Bulletin of the Aquaculture Association of Canada 102-1:5–7.
Davison W. (1997) The effects of exercise training on teleost fish, a review of recent literature. Comparative Biochemistry and Physiology 117A:67–75.
Engelsen R., Asche F., Skjennum F., Adoff G. (2004) New species in aquaculture: some basic economic aspects. In Moksness E., Kjørsvik E., Olsen Y. (Eds.). Culture of Cold-Water Marine Fish(Blackwell Science, Oxford) pp. 487–515 528 pp.
Folkvord A. (1991) Growth, survival and cannibalism of cod juveniles (Gadus morhua) – effects of feed type, starvation and fish size. Aquaculture 97:41–59.[CrossRef][Web of Science]
Fowler J., Cohen L., Jarvis P. (1998) Practical Statistics for Field Biology 2nd edn. (John Wiley & Sons, Chichester) 259 pp.
Grant S.M., Brown J.A., Boyce D.L. (1998) Enlarged fatty livers of small juvenile cod: a comparison of laboratory-cultured and wild juveniles. Journal of Fish Biology 52:1105–1114.[CrossRef][Web of Science]
Johnston I.A. (1999) Muscle development and growth: potential implications for flesh quality in fish. Aquaculture 177:99–115.[CrossRef][Web of Science]
King N.J. and Nardi G.C. (2002) Development of Atlantic cod larviculture in Newfoundland. Bulletin of the Aquaculture Association of Canada 102-1:12–17.
Kjørsvik E., Pittman K., Pavlov D. (2004) New species in aquaculture: some basic economic aspects. In Moksness E., Kjørsvik E., Olsen Y. (Eds.). Culture of Cold-Water Marine Fish(Blackwell Science, Oxford) pp. 204–278 528 pp.
Qin J.G., Ingerson T., Geddes M.C., Kumar M., Clarke S. (2001) Size grading did not enhance growth, survival and production of marron (Cherax tenuimanus) in experimental cages. Aquaculture 195:239–251.[CrossRef][Web of Science]
Shields R.J. (2001) Larviculture of marine finfish in Europe. Aquaculture 200:55–88.[CrossRef][Web of Science]
Svåsand T., Otterå H.M., Taranger G.L. (2004) The status and perspectives for the species. In Moksness E., Kjørsvik E., Olsen Y. (Eds.). Culture of Cold-Water Marine Fish(Blackwell Science, Oxford) pp. 433–474 528 pp.
Vaccaro A.M., Buffa G., Mirto S., Sara G., Mazzola A. (2004) Comparison of growth performance and biometric relationships in two reciprocal sturgeon hybrids reared in net cages (Sicily, Mediterranean). Aquaculture Research 35:552–558.[CrossRef][Web of Science]
Wijekoon M.P.A., Puvanendran V., Brown J.A. (2003) Interpopulation difference in growth of juvenile Atlantic cod (Gadus morhua) – implication for broodstock development. Bulletin of the Aquaculture Association of Canada 21–23 Special Publication 6.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||