Cleaner fish have been used widely by the salmon farming industry in northern Europe as part of an integrated strategy for the control of sea lice (Bjordal 1990; Kvenseth 1996; Costello 2006). Sea lice continue to be one of the main health issues in farming salmon (Pike and Wadsworth 2000) and have a high economic cost to farming operations (Costello 2009). Wrasse are also stocked to combat the potential development of lice that are resistant to sea lice medicines (Lees et al. 2008). While wild-caught fish are currently used, estimated at over 10 million in Norway in 2012 (Sveier EWOS Norway, 2011, pers. comm.), attempts are being made to develop culture techniques for wrasse in finfish hatcheries (Treasurer 2011). There is concern regarding the welfare of wild-caught and farmed fish when held in net cages with Atlantic salmon, Salmo salar.
The issues that wrasse face in salmon cages are requirements for shelters for protection when at rest and from tides and currents; supplementary feeding when lice numbers are low; and care during farm operations such as grading, and moving salmon, and net cleaning.
Increasing emphasis is being placed on ensuring the welfare of fish and good practice in fish production throughout the production cycle as a growing number of consumers and retailers value ethically produced food (Scott and Ellis 2007). This has led to the European Animal Welfare Platform (EAWP, 2011) producing guidelines for the production of fishes such as Atlantic salmon, and has been driven by fish welfare assurance policies such as the RSPCA (2012) Freedom Foods, that have delineated commercial standards for Atlantic salmon aquaculture. These standards are expected to apply equally to other fish stocked with salmon in cages. Various measures have been proposed as indicators of fish welfare such as fish physiology, health, and behavior (Huntingford et al. 2006). Stress can be measured by cortsol directly or non invasively (Scott and Ellis 2007), but this can be time-consuming and not practical for routine monitoring. For practical measurements of farmed fish, a number of non intrusive techniques that can be applied readily on farms were suggested depending on species and circumstances (Huntingford et al. 2006). The fish condition factor can be measured, but this requires measurements of length and weight and comparison with standard values. However, fish condition alone does not indicate all causes of stress.
The objective of this study was to assess various measures of physical condition and welfare in five species of wrasse: goldsinny, Ctenolabrus rupestris; rock cook, Centrolabrus exoletus; corkwing, Crenilabrus melops; cuckoo, Labrus mixtus; and ballan wrasse, Labrus bergylta, to compare the relative condition of the five species when held in aquaculture tanks.
An additional operational farm study was also carried out to assess the fin condition of cage held wrasse through an 18–24 mo salmon production cycle.
Results
Fin Erosion and Splitting
The condition factors were 1.46 for rock cook, 1.43 for corkwing, and 1.42 for ballan wrasse (Fig. 1) and these did not differ significantly (F =0.69, P =0.50), although all were significantly higher (P <0.05) than 0.98 for cuckoo wrasse and 1.36 for goldsinny, Figure 1. The condition factor of five species of wrasse held in holding tanks for 3 mo. Although there were variations, these were due to differences in body shape and did not refiect difference in condition. Sample size=50 fish for all species with the exception of cuckoo wrasse=25. which refiects the slimmer body shape of these species. Eye and mouth damage were negligible in the five wrasse species, with only one corkwing with a physically damaged eye, which had a healed opaque epidermal area, and another corkwing with snout abrasion. There was no skin hemorrhaging in any fish. These rare conditions were too uncommon to analyse statistically.
Fin erosion in all wrasse species and on all fins was low with index scores of <0.56 (Fig. 2). The dorsal fin erosion was slight in goldsinny (0.55±0.61 SD) and cuckoo wrasse (0.56±0.69) but were significantly higher (q =3.26, P =0.001) than in ballan (0.30 ±0.46) which was also significantly higher than in rock cook (0.10±0.30) (q =2.41, P =0.02) and corkwing (0.13±0.30) (q =0.12, P = 0.02) (Fig. 2). A proliferation of epithelial cells was identified when the fins were examined in microscopy, but no active bacterial involvement was seen in these lesions.
Splitting of the dorsal, anal, and pectoral fins was also negligible (index circa 0.01) in all wrasse species. Fin splitting was primarily observed on the caudal fin, and this was significantly higher (ANOVA, P <0.05) than for splitting in all other fins in the five wrasse species. Splitting of the caudal fin was high (average index 2.8±0.57) in rock cook wrasse and also corkwing (2.17±0.59) and these values were significantly higher (q =13.38, P =0 and q =5.47, P =0) than in goldsinny (average index 1.55±0.25) and ballan wrasse (1.56±0.76) (q =13.04, P =0 and q =5.29, P =0), with the two latter species not significantly different (q =0.11, P =0.91). Fin splitting was lowest and significantly less (q =2.75, P =0.007) in cuckoo wrasse (1.13±0.81).
This baseline assessment of possible welfare indices in wrasse indicates that splitting of the caudal fin can be the most sensitive indicator of poor condition or damage. It also indicates that rock cook and corkwing wrasse are the species most likely to incur physical damage, although this was at a moderate level and not at a severity that compromised survival. No growth data are available for the acclimation period, and the mortality during this 12 wk period was 0.45% (SD 0.21) of stock/week with much of this due to capture and transport trauma.
FEI and FSI scores were assigned to ballan wrasse on stocking the fish on the farm, in winter, and just prior to final harvest (Fig. 3). The ballan wrasse examined at the end of the production cycle showed no damage to the body in any fish, and there were no signs of emaciation. There was no skin, jaw, or eye damage at any sampling point. Fin erosion was minimal on stocking, during the production cycle and on harvest (P >0.05). The level of fin splitting on stocking on the farm was mild (0.59±0.80 SD) and confined to the tail.
There was no significant difference (F =0.09, P =0.91) in the caudal FSI prior to stocking, during the winter and on harvest. From the assessment of wrasse welfare indices it is concluded that the condition of the ballan wrasse assessed by indices for fin erosion and splitting was maintained in cages during the production cycle. In the operational farm study the overall mortality for the period from May to December 2011 was 1121 wrasse of 13,040 fish stocked (8.6%) on all cages on the farm.
Discussion
Operational Welfare Indictors that have been used by welfare assessment schemes have been growth, stress, and fin damage (Turnbull et al. 2005). Huntingford et al. (2006) gave a list of indicators including changes in ventilation rate, behavior, reduced food intake, condition, poor growth, abnormalities, injuries including fin damage, disease, and reduced reproductive performance. A common measure of fish physical condition and welfare has been the use of indices of fish erosion and splitting (Moutou et al. 1998). Hoyle et al. (2007) suggested that the term “fin damage” was more descriptive of this condition. Fin erosion has been highlighted as a fish welfare indicator and has been defined as injury to live tissue including nerve endings (Ellis et al. 2008). It is one of several fish welfare indicators (Turnbull et al. 2005) and has the advantage that, as it is external, it is readily visible and easily distinguished and understood by farm and technical staff, and fin erosion was successfully demonstrated as an external indicator in salmonids (North et al. 2006; Adams et al. 2007; Noble et al. 2008).
These indices have been successfully applied in salmonids where fins can be damaged by aggressive interaction, eroded by high stocking densities, poor water quality, and general fish degradation due to poor culture conditions or inadequate feeding (MacLean et al. 2000), as well as rainbow trout, Onchorhynchus mykiss; amago, Onchorhynchus masou masou (Flood et al. 2010); and sea bass (Arechavala-Lopez et al. 2013). Fin damage measurements have ranged from a 2-point classification of erosion based on affected surface fin area and a qualitative 3-point scale of fin splitting (Moutou et al. 1998; MacLean et al. 2000) to a 6-point scale in rainbow trout (Hoyle et al. 2007; Ellis et al.
2008).
Fin erosion has been quantified by a range of methods (Noble et al. 2008). A fin index that is
equivalent to relative fin length was developed to quantify the extent of erosion (Kindschi et al. 1991) and this has been straightforward to apply as it avoids subjective assessment. This was improved following possible drawbacks caused by allometric effects of growth on fin length, effects of fin erosion on the caudal fin, and effects of stock and environment on fin shape (Pelis and McCormick 2003).
Predictive relationships were developed to improve assessment of severity of fin damage in rainbow trout. While these relationships can be measured to improve quantification, these techniques can be too complicated for use in production systems where a photographic qualitative scale may be more applicable (Hoyle et al. 2007). There are other forms of physical damage that are immediately recognizable such as scale loss caused by physical damage or aggression, damage to the snout or jaw, and also damage to the eyes. Condition factor has also been used successfully in both farmed and wild ballan wrasse and was found to decline by 2.5–3.6% in the 6 wk after being stocked in salmon cages (Skiftesvik et al. 2013).
The use of substrates/structures to reduce stress in wrasse and to increase survival, especially overwinter, can be beneficial to wrasse welfare. Survival of wrasse may also be increased by providing supplementary feeding when lice numbers are low. Further development is required of the use of wrasse hides and their effective use as a refuge from predators and to provide rest areas.
This study outlines baseline data on the welfare of five species of wrasse in aquaculture tanks and, for ballan wrasse, additional data on the welfare of these fish when stocked with Atlantic salmon in cages. The tank study indicates similar mortality of wrasse compared to the field trials of ballan wrasse stocked in salmon cages. Most of these mortalities on the farm occurred during two routine site activities. The first occurred during a sea lice bath treatment using Alphamax, a pyrethroid, when a total of 463 wrasse mortalities was removed, 3.6% of initial numbers stocked.
These fish showed over-infiated swimbladders and the rapid lifting of the nets did not allow adjustment of the swimbladder. Further, there may have been stress involved with the sea lice treatment due to the lifting of the nets, crowding of the fish, and the medicine used. It is unlikely that the medicine itself would have been toxic. The second mortality event occurred when the nets were lifted in December to transfer some salmon offsite. In this case, the lifting of the nets was responsible for the over-infiation of the swimbladder. Improvements to farm operations can be made to improve wrasse welfare, such as ensuring nets are raised slowly during sealice treatments and other farm operations and checking nets during net cleaning operations.
A modified fin damage index based on the scales of Hoyle et al. (2007) and Noble et al. (2008) was used in this study and was found to be a suitable tool for assessing fin damage in wrasse. The erosion and splitting of the fins were the most commonly identified signs of physical damage and these indices can be used by farmers to routinely assess the condition of wrasse on receipt and while stocked in salmon cages. In this study rock cook and corkwing were more prone to damage compared with the other wrasse species. The indices were low for fin erosion while fin splitting was more prevalent and was pronounced on the caudal fin. The caudal fin was also the most severely damaged fin in sea bass Dicentrarchus labrax with an index of over 1 in farmed fish (ArechavalaLopez et al. 2013). The occurrence of eye damage was low, recorded in only one rock cook, and none of the fish examined had mouth damage. There was evidence of filamentous bacterial involvement in the more advanced levels of fin erosion in goldsinny wrasse during the breeding season (T. Turnbull, Scottish Sea Farms, South Shian, Scotland, pers. comm.). It was considered that the bacterial infection was secondary to minor lesions arising from physical abrasion or fin nipping. However, these bacteria were of mixed origin and could not be identified in the current study.
May 2014
