UBC Faculty Research and Publications

Disease resistance is related to inherent swimming performance in Atlantic salmon Castro, Vicente; Grisdale-Helland, Barbara; Jørgensen, Sven M; Helgerud, Jan; Claireaux, Guy; Farrell, Anthony P; Krasnov, Aleksei; Helland, Ståle J; Takle, Harald Jan 21, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12899_2012_Article_201.pdf [ 691.18kB ]
JSON: 52383-1.0220708.json
JSON-LD: 52383-1.0220708-ld.json
RDF/XML (Pretty): 52383-1.0220708-rdf.xml
RDF/JSON: 52383-1.0220708-rdf.json
Turtle: 52383-1.0220708-turtle.txt
N-Triples: 52383-1.0220708-rdf-ntriples.txt
Original Record: 52383-1.0220708-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessDisease resistance is related to inherentswimming performance in Atlantic salmonVicente Castro1,2,3, Barbara Grisdale-Helland4,5, Sven M Jørgensen1, Jan Helgerud6, Guy Claireaux7,Anthony P Farrell8, Aleksei Krasnov1, Ståle J Helland2,4,5 and Harald Takle1,3,5*AbstractBackground: Like humans, fish can be classified according to their athletic performance. Sustained exercise trainingof fish can improve growth and physical capacity, and recent results have documented improved disease resistancein exercised Atlantic salmon. In this study we investigated the effects of inherent swimming performance andexercise training on disease resistance in Atlantic salmon.Atlantic salmon were first classified as either poor or good according to their swimming performance in ascreening test and then exercise trained for 10 weeks using one of two constant-velocity or two interval-velocitytraining regimes for comparison against control trained fish (low speed continuously). Disease resistance wasassessed by a viral disease challenge test (infectious pancreatic necrosis) and gene expression analyses of the hostresponse in selected organs.Results: An inherently good swimming performance was associated with improved disease resistance, as goodswimmers showed significantly better survival compared to poor swimmers in the viral challenge test. Differencesin mortalities between poor and good swimmers were correlated with cardiac mRNA expression of virus responsivegenes reflecting the infection status. Although not significant, fish trained at constant-velocity showed a trendtowards higher survival than fish trained at either short or long intervals. Finally, only constant training at highintensity had a significant positive effect on fish growth compared to control trained fish.Conclusions: This is the first evidence suggesting that inherent swimming performance is associated with diseaseresistance in fish.BackgroundDiseases represent the main constraint for the success ofan expanding aquaculture industry. Atlantic salmon(Salmo salar) farmers can experience severe fish lossesdue to both infectious and non-infectious diseases, usu-ally during the seawater growth stage. Infectious pancre-atic necrosis (IPN), pancreas disease (PD), infectioussalmon anemia (ISA), as well as the sea lice parasite(Lepeophtheirus salmonis K) represent some of the mosthazardous diseases [1,2], but losses have also been asso-ciated with non-infectious diseases such as cardiac fail-ures [3,4]. Biosecurity countermeasures to control thedisease situation include vaccines and pharmaceuticals,as well as improvements of the genetic material, feedsand husbandry practices. The aim of current and futurecountermeasures is to strengthen the fish robustness,which is the capability to combine fast growth and nor-mal organ development with improved resistance toboth disease and physiological challenges.Sustained exercise training has been documented toconfer higher robustness to cultured fish, includingincreased somatic and cardiac growth, cardiac perform-ance, aerobic capacity of the muscle, oxygen carryingand extraction capacity and improved bone quality[5-10]. Khovanskiy et al. [11] found that exercised chumsalmon (Oncorhynchus keta) displayed lower mortalityassociated with an improved osmoregulatory capacityafter seawater transfer when compared to untrained fish.Going further, we have recently demonstrated directeffects of sustained exercise on disease resistance, show-ing that survival of Atlantic salmon challenged with in-fectious pancreatic necrosis virus (IPNV) was 13%* Correspondence: harald.takle@nofima.no1Nofima, Ås, Norway3AVS Chile S.A., Puerto Varas, ChileFull list of author information is available at the end of the article© 2013 Castro et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Castro et al. BMC Physiology 2013, 13:1http://www.biomedcentral.com/1472-6793/13/1higher for fish subjected to a moderate interval-trainingregime for six weeks prior to smoltification when com-pared with fish held at a low, constant swimming speed[12]. Thus, sustained exercise in fish can induce a similarrobustness effect as in humans, where a moderate aer-obic training is also known to decrease the risk of infec-tions and chronic life-style diseases [13,14].It has been observed that exercised salmonids [12,15]and non-salmonids, (Plecoglossus altivelis [16]; Chalcar-burnus chalcoides mento [17]; Morone saxatilis [18];Sparus aurata [19]; Danio rerio [20]) exhibit improvedgrowth due to improved feed efficiency, higher feed in-take or a mix of both. Several studies have reported onthe relationship between improved growth performanceand disease resistance in fish (see reviews by Merrifieldet al. [21,22]). For example, Gjedrem [23] suggested thata breeding program selecting for growth also induced apositive genetic response for disease resistance, althoughconflicting results exist [24]. Recently, an association be-tween exercise-induced growth and improved disease re-sistance was shown in Atlantic salmon [12]. Thepossibility of a linkage between these two factors, growthand disease resistance, is of obvious importance for thefish farming sector.Migratory fish such as salmonids have a great inherentcapacity for sustained aerobic swimming. Benefits fromexercise seem to be maximized at speeds close to theoptimal swimming speed (Uopt), where energy use ismore efficient and the cost of transport is minimized[25]. Exercising fish at speeds other than Uopt results inadditional energy usage for locomotion, even at lowspeeds due to behavioral changes (e.g. increased aggres-sion and spontaneous activity). Further, the highestspeeds may prove stressful and unsustainable comparedwith Uopt [20]. Because of their natural swimming be-havior and high aerobic capacity, salmonids are naturallyamenable to long-term continuous exercise training,provided sustainable water velocities are used. This is incontrast to terrestrial animals that more typically requireresting periods between bouts of exercise training. Inhumans, where most exercise training research has beenperformed, the intensity seems to be a fundamental fac-tor affecting the individual’s systemic immunity. Whileengaging in regular moderate exercise activity seems toenhance immune functions [13], high intensity aerobictraining results in acute and chronic states of impairedimmunity [26].On top of training effects, there seems to be anequally large inherent variation in exercise capacityamong fish and humans. For example, juvenile rain-bow trout (Oncorhynchus mykiss) can be classifiedaccording to their inherent swimming performanceas either poor or good swimmers. Interestingly, suchclassification was associated with several cardiac andmetabolic capacities after 9 months of common rear-ing [27].This study aimed to evaluate if inherent swimmingperformance in juvenile fish affect disease resistance. Ju-venile fish were identified according to their inherentswimming performance by pre-screening them in aswim challenge test. Fish classified as either poor orgood swimmers were then trained at four differentregimes to investigate if training differentially affectedthem. After smoltification, a controlled disease challengewith IPNV allowed us to assess differences in disease re-sistance among and within the two performance groupsand four training regimes. This was further examined byanalyzing expression levels of sensitive virus responsivegenes (VRGs) in head kidney and cardiac tissues. Inaddition, exercise-induced effects on robustness wereevaluated by growth performance and feed efficiency.ResultsDisease resistance is related to inherent swimmingperformanceMortality following the IPN challenge started 18 daysafter the introduction of virus-shedding fish and reacheda plateau around day 38 post-challenge. Inherent swim-ming performance showed a strong association with sur-vival after the IPN challenge test. Fish initiallycategorized as good swimmers had significantly bettersurvival than poor swimmers (86.1 and 77.6%, respect-ively; p = 0.02) when analyzed across all groups(Figure 1A). When survival was examined independentof the inherent swimming performance, differencesamong training regimes showed no significant difference(p = 0.21). The continuous-velocity training regimestended to improve survival compared to the controltrained group (87.1, 84.2 and 82.2% survival for respect-ively M, H and C), whereas interval-velocity trainingregimes tended to negatively impact survival (78.2 and75.3% survival for Sint and Lint, respectively; Figure 1B).When survival was examined in light of swimmingperformance, exercise training did not significantly affectdisease resistance of poor and good swimmers. Never-theless, exercise may have had a larger impact on diseaseresistance of poor swimmers since larger changes inmortality were observed in exercised groups of poorswimmers compared to good swimmers (Figure 2).Virus responsive gene expression reflects and supportsmortality dataTo verify mortality data from IPNV, infection level wasanalyzed in challenged fish at termination of the diseasetrial. Quantification of IPNV (by real-time qPCR inhead-kidney) in surviving fish showed low prevalence ofvirus-positive fish (33%) and low level of viral transcriptsin positive fish. Heart tissue was also tested and foundCastro et al. BMC Physiology 2013, 13:1 Page 2 of 12http://www.biomedcentral.com/1472-6793/13/1negative for all fish. Thus, sampled fish were in a latestage of infection with either low levels or no virusreplication.Gene expression analysis was performed to investigateamong-groups differences in host immune correlates ofdisease response. For initial screening of correlates, tran-scriptome analysis by microarray of poor and goodswimmers was assessed, since the strongest contrast inmortality was associated with swimming performance.This resulted in 21 genes with significantly higher tran-script abundance in poor compared to good swimmers(t-test, p < 0.05) (Figure 3). By function, all genes werepreviously identified as virus-responsive genes (VRGs); agroup of genes displaying a common activation/tran-scription to most of the known viruses infecting Atlanticsalmon. VRGs are sensitive antiviral markers reflectingthe infection status and the level of viral transcripts incells [28]. Thus, poor swimmers seemed to have higheractivation of antiviral immune genes compared to goodswimmers at the end of the infection trial.To further substantiate these results and to evaluateeffects of the different training regimes with sufficientbiological replication, expression of six VRGs was ana-lyzed in heart ventricle tissue of poor and good swim-mers from the C, M and Lint exercise-trained groupsusing qPCR. Results showed that induced levels of VRGsin poor swimmers were mainly explained by a strong ex-pression in Lint exercised fish (Figure 4A). Controltrained (C) and M trained poor swimmers had equal ex-pression level of VRGs. Within the good swimmers,VRG expression levels were higher in M and Linttrained compared to control trained fish.We further analyzed expression of eight VRGs in headkidney, where IPNV replication was observed. Expres-sion levels within exercised good swimmers showed asimilar trend as observed for heart tissue (Figure 4B).Figure 1 Cumulative survival of swimming performance andexercised groups during IPN challenge. A: The inherentswimming performance of the fish had a significant effect ondisease resistance, with the good swimmers showing a highersurvival (86.1%) than the poor swimmers (77.6%). B: Fish exercised atconstant speeds for 10 weeks showed a trend towards higherdisease resistance compared to those subjected to interval trainingregimes (Medium intensity (M) = 87.1%; High intensity (H) = 84.2%;Control (C) = 82.2%; Short interval (Sint) = 78.2% and Long interval(Lint) = 75.3%) (p = 0.21). *indicates significant difference, Mantel-Coxtest; p < 0.05.Figure 2 Interaction between inherent swimming performance and exercise training on disease resistance. Only those regimes includedin the gene expression analysis are displayed. These reflect the lower (Medium), middle (Control) and higher (Long interval) mortalities found inresponse to IPN infection. Data is shown within poor (left) and good (right) swimmers. Though not significant, differences in mortalities werelarger between poor and good swimmers for the Long interval, smaller for Control and minimum for the Medium intensity regime. Initial numberof fish on the challenge ranged from 100 to 112 per regime.Castro et al. BMC Physiology 2013, 13:1 Page 3 of 12http://www.biomedcentral.com/1472-6793/13/1Within poor swimmers, VRG expression levels were sig-nificantly lower in M compared to both control and Linttrained fish (p < 0.05). In contrast to heart tissue, VRGexpression levels in head-kidney of control trained goodswimmers were significantly lower compared to controltrained poor swimmers (p = 0.03). Thus, the increasedoverall mortality observed for poor swimmers wasreflected by stronger expression levels of antiviral im-mune genes in Lint trained (heart) and control trained(head-kidney) fish as compared to good swimmers.While no evidence for any positive effects of exercisetraining on mortality and infection status was observedfor good swimmers, results implied beneficial effects interms of reduced infection status (lower VRG expres-sion) from M trained poor swimmers.Results produced for 8 genes by microarray and qPCRwere in close concordance (Pearson r = 0.92; p = 0.001).Exercise training effects on growthAfter six weeks of exercise training, no significant differ-ences in thermal growth coefficient (TGC) were foundamong training regimes and control trained fish. At theend of the freshwater phase of the experiment (10 weeksof training plus 1 week detraining), TGC was signifi-cantly higher (p < 0.05) in the high intensity (H) trainingregime (1.61) compared with the control (C) trainedgroup (1.50) (Table 1). The other training regimes onlyshowed a trend towards higher TGC compared to thecontrol trained group (p < 0.1). At the end of exercisetraining, medium intensity (M) trained fish showed asignificantly higher condition factor (CF) than C. Feedintake showed significant differences among trainingregimes, with the higher values belonging to the H andshort interval (Sint) regimes at the end the first six train-ing weeks, while H and long interval (Lint) had the high-est feed intake in the second part of the trainingexperiment. After six weeks of training, the only signifi-cant differences for feed efficiency ratio (FER) amongtraining groups were H and Sint groups being lowerthan both C and M (Table 1).DiscussionIn this study we found that the inherent swimming per-formance of juvenile Atlantic salmon is associated withdifferences in survival to an infectious disease challengeafter seawater transfer, with good swimmers displaying asignificantly higher disease resistance than poor swim-mers. Exercise training had no significant effect on dis-ease resistance, though a trend towards improvedperformance was seen for fish being trained at constantcompared to interval regimes. Though not significant,results further argue for exercise training affecting poorswimmers more strongly than good swimmers. Mortal-ities were supported by mRNA expression levels of aFigure 3 Differentially expressed genes in poor versus good swimmers post IPN challenge. Microarray analyses resulted in 21 genesshowing higher transcript abundance in cardiac muscle of poor swimmers compared to good swimmers. By function, all of these genes havebeen previously identified as virus responsive genes (VRGs). Data for Poor and Good swim performance groups is log2-ER (expression ratio) ± SEM(n = 9).Castro et al. BMC Physiology 2013, 13:1 Page 4 of 12http://www.biomedcentral.com/1472-6793/13/1subset of VRGs reflecting the antiviral response status.Growth was promoted by exercise training, though onlysignificant for the highest intensity regime (H), whileswimming performance did not show an associationwith growth performance.The impact of inherent swimming performance ondisease resistanceThe inherent swimming performance of individual ju-venile salmon was positively associated with diseaseresistance. Fish classified as good swimmers showed8.5% better survival against IPN virus than poor swim-mers when challenged 3 months after the swimmingperformance classification. Claireaux et al. [27] demon-strated that good swimmers of a cohort of rainbowtrout, classified by a similar methodological approach asused in this work, retained this advantage nearly ninemonths later, despite a common rearing environmentand similar growth performance, and displayed a signifi-cantly better cardiac capacity and morphology comparedFigure 4 Expression of virus-responsive genes in tissues of exercised and swim performance groups post IPN challenge. Improvedsurvival after IPN challenge was associated with the expression of virus responsive genes (VRGs) as assessed by qPCR. Higher expression levels ofVRGs in cardiac tissue of poor compared to good swimmers from the Lint regime (A), and in head kidney of poor compared to good swimmersfrom the control group (B) reflected the overall higher mortality of poor swimmers in comparison to the good swimmers. Further, VRGsexpression in both tissues was in concordance with differences in survival between interval (Lint) and constant speed (Control and Medium)training regimes. Bars represent expression ratio ± SEM relative to a pooled control sample and normalized against two reference genes (18SrRNA and elongation factor 1α) with correction for PCR efficiency. For A and B, each bar is a composed index of 6 (n = 8 fish/swim-performance/regime) and 8 (n = 5 fish/swim-performance/regime) VRGs, respectively. Genes included: RSAD2 (radical S-adenosyl methionine domain containingprotein2) (A + B), IFIT5 (interferon-induced protein with tetratricopeptide repeats 5) (A + B), STAT1 (signal transducer and activator of transcription 1)(A + B), VHSV2 (viral haemorrhagic septicaemia virus-inducible protein) (A + B), BAF (barrier-to-autointegration factor) (A + B), GIG1 (interferon-inducible Gig1) (A + B), RIG-I (DEAD/H (Asp-Glu-Ala-Asp/His) box polypeptide) (B), MDA5 (interferon induced with helicase C domain 1) (B). abc:Denotes significant difference (p < 0.05; paired t-test) between training regimes. Other symbols (*§) denotes significant difference (p < 0.05; pairedt-test) between poor and good swimmers within each training regime.Castro et al. BMC Physiology 2013, 13:1 Page 5 of 12http://www.biomedcentral.com/1472-6793/13/1to poor swimmers. This and the present study collect-ively suggest that a simple screening test for swimmingperformance can efficiently distinguish between fish withlow and high robustness, with the latter possessing bet-ter cardiac capacity and disease resistance. It must benoted that none of these studies could find differencesin growth performance between poor and goodswimmers.In addition to the effects of the inherent swimmingperformance of fish on robustness, exercise trainingappeared to have a stronger, though not significant,modulatory effect on the poor swimmer’s disease resist-ance. While the M training regime showed a tendencyto improve the survival rate of the poor swimmers, theLint regime tended to decrease the survival of the poorswimmers. The possible interaction effect between in-herent poor swimming performance and training regimeon survival was supported by expression analysis ofVRGs in surviving fish from the different trainingregimes and performance groups at the end of the IPNchallenge. Results showed that improved survival ofgood swimmers was associated with lower expressionlevels of virus responsive genes, probably reflecting anoverall lower level of infection pressure, a more rapid orefficient viral clearance and/or a reduced antiviral statusto recover and regain homeostasis. Thus, the ability torapidly clear or reduce virus replication and antiviral im-munity at the end of a viral infection might be importantfor survival. In a previous study, we demonstrated thatthe improved survival induced by sustained training ofjuvenile Atlantic salmon, was related to a specific cardiactranscriptome signature, suggesting lower levels of in-flammation and higher levels of immune effector mole-cules, antioxidant enzymes and xenobiotics clearancecapacity prior to an IPN challenge [12].The impact of training regimes on disease resistanceOverall, the three continuous training regimes (includingC) displayed a trend, though not significant, towardshigher survival compared to the interval trainingregimes. The continuous 0.65 body lengths (BL)s-1 M re-gime gave the best results, which is in agreement withour previous finding where Atlantic salmon pre-smoltstrained at a similar intensity for six weeks (0.8 BLs-1 for16 h and 1 BLs-1 for 8 h per day) showed 13% highersurvival following an IPN challenge test when comparedto control trained fish [12]. Such improvements are veryimportant in an industry context. In contrast to the im-provement in disease resistance from utilizing an inter-val training regime with mild speed changes as in thepreviously mentioned study, the ~3-fold daily changes inswimming speed applied for the Sint and Lint regimestended to reduce disease resistance against IPN com-pared to controls kept continuously at 0.32 BLs-1. SinceSint and Lint trained fish had theoretically swum thesame distance as the M trained fish, it could be arguedthat the relatively strong daily changes in water speedfor both interval regimes may be the cause of the appar-ent negative impact on disease resistance of these fish.We may speculate that the non-significant trend to-wards reduced disease resistance of the inherently poorswimmers trained with the Lint regime is due to a lowerTable 1 Growth parameters and dry matter intake of exercise trained Atlantic salmonC M H Sint LintBody weight (BW) (g) Start 40.9 ± 0.2 40.7 ± 0.2 40.6 ± 0.2 40.2 ± 0.4 41.0 ± 0.4W 6 70.5 ± 1.3 70.9 ± 0.7 72.4 ± 2.4 72.3 ± 0.6 71.4 ± 0.4W 11 95.4 ± 0.3 99.9 ± 1.1 99.4 ± 2.4 98.2 ± 2.2 100.4 ± 2.2Length (cm) Start 15.1 ± 0.03 15.0 ± 0.04 15.0 ± 0.01 15.0 ± 0.05 15.1 ± 0.03W 11 20.0 ± 0.04 20.2 ± 0.07 20.2 ± 0.16 20.2 ± 0.14 20.4 ± 0.14CF Start 1.18 ± 0.003 1.18 ± 0.002 1.18 ± 0.005 1.18 ± 0.002 1.18 ± 0.001W 11 1.18 ± 0.004b 1.20 ± 0.004a 1.19 ± 0.004ab 1.18 ± 0.002b 1.18 ± 0.008bTGC W 1-6 1.56 ± 0.05 1.59 ± 0.04 1.66 ± 0.09 1.68 ± 0.02 1.59 ± 0.01W 6-11 1.44 ± 0.08 1.64 ± 0.06 1.53 ± 0.05 1.47 ± 0.12 1.64 ± 0.10W 1-11 1.50 ± 0.01b 1.58 ± 0.03ab 1.61 ± 0.02a 1.59 ± 0.05ab 1.59 ± 0.03abRelative feed intake (% BW d-1) W 1-6 0.87 ± 0.02c 0.88 ± 0.02c 0.98 ± 0.03ab 0.99 ± 0.01a 0.92 ± 0.01bcW 6-11 0.65 ± 0.02c 0.69 ± 0.01bc 0.74 ± 0.01a 0.70 ± 0.02ab 0.74 ± 0.01aFER W 1-6 1.43 ± 0.04a 1.42 ± 0.02a 1.32 ± 0.00b 1.34 ± 0.00b 1.37 ± 0.02abW 6-11 1.66 ± 0.08 1.70 ± 0.03 1.52 ± 0.06 1.56 ± 0.11 1.58 ± 0.04C Control, M medium intensity, H high intensity, Sint short interval, Lint long interval, CF condition factor, TGC thermal growth coefficient, week (W) 6: End of firstsix weeks of training under a short day light photoperiod. W 11: End of 10 weeks of training and one week of detraining. Means in the same row with differentsuperscripts letters are significantly different based on one-way ANOVA (p < 0.05). Differences between the groups were assessed by the least-squares meansprocedure. Data are means ± SEM.Castro et al. BMC Physiology 2013, 13:1 Page 6 of 12http://www.biomedcentral.com/1472-6793/13/1acclimation capacity of these fish to the relatively strongchanges in swimming velocity compared to the goodswimmers. It is then plausible to suggest that poorswimmers suffered from higher stress levels when fol-lowing the Lint compared to the continuous speedregimes, which could potentially cause an impairment oftheir disease resistance capacities. Inherently good swim-mers, however, seem to have sufficient behavioral and/orphysiological plasticity as to avoid disease resistanceimpairment.The moderate intensity of the M training regime mayhave promoted an acclimative response in the poorswimmers, boosting their disease resistance to the levelof the good swimmers. Thus, if confirmed with newstudies, the overall profit of conducting M regime train-ing would be the achievement of a more homogenouspopulation when it comes to disease resistance. Thiswould imply an indisputable benefit for salmonproducers.The impact of training regimes on growth and feedutilizationBy definition, a robust fish must possess a good combin-ation of both high growth performance and disease re-sistance. Jobling et al. [29] and Davison [15] stated thathigher growth may be achieved for fish when training in-tensity lies between 0.75 and 1.5 BLs-1. Our results arein agreement with this; the H training regime (1.31 BLs-1)had significantly higher TGC than the control trainedgroup. Interestingly, the other three regimes (M, Sint andLint), which had an average water speed of 0.65 BLs-1,showed a tendency towards improved growth com-pared to control trained fish, suggesting the existenceof a correlation between growth and total work loadof the swimming-induced exercise. Higher growthgiven by the H training regime was mainly due toincreased feed intake associated with a lower feed effi-ciency and protein retention. This suggests that fishsubjected to that regime required more energy to satisfythe increased demand. Despite a lowered feed efficiency,increases in feed intake were sufficient to overcompensatethe needs of simultaneous swimming and growth. It couldbe argued that training at this intensity stimulated theregulation of neuroendocrine factors involved incontrolling feeding, resulting in an anabolic dominantstate. It is logical to think though, that growth will becompromised at higher water speeds than those testedhere, as has been found in salmonids [30,31] andother species, such as striped bass Morone saxatilis[32]. Indeed, routine gut blood flow, which is a basicrequirement for effective digestion, is reduced insalmon as they swim progressively faster and can stopwith abrupt stresses [33,34].Another effect that may contribute to exercise-induced growth is the possibility of salmon juvenileschanging from active to passive (ram) ventilation. Ramventilation is the capacity of some fish species to venti-late passively by opening their mouths when swimmingor facing high water currents, allowing water to passthrough the gills with enough pressure for gas exchangeto occur without the need for active branchial pumping[35,36]. The energy sparing effect of ram ventilation ran-ged from 8.4 to 13.3% in adult rainbow trout, whichshifted ventilation mode when swimming above 0.5-1BLs-1 [37]. Nevertheless, we cannot know for certain ifthis is also the case in this study.ConclusionsThis study provides the first evidence demonstratingthat inherent swimming performance in juvenile fishmay predetermine disease resistance later in life. Fishclassified as good swimmers showed a significantly lowermortality when challenged with IPN than fish classifiedas poor swimmers. Our results further suggest that theinherently poor swimmers are more sensitive to the in-tensity and design of the training regimes than goodswimmers. Finally, the results confirmed that sustainedexercise at high intensity stimulates growth performanceof Atlantic salmon, while exercise at lower intensitieshas less effect.The great variability in swimming performance withinpopulations of fish opens up novel possibilities forphenotype or marker-assisted selection in breeding pro-grams and as a discrimination tool to sort out poorjuvenile fish when it is still cost-effective.MethodsExperimental fishJuvenile Atlantic salmon belonging to the Salmobreedstrain were produced and reared at Nofima AS,Sunndalsøra, Norway. Freshwater stage procedures tookplace at the same research station, which is an approvedfacility under the Norwegian Animal Research Authority(NARA). Stunning and sampling of fish was done inagreement with the Norwegian regulations. As fish wereexposed to different sustainable water velocities that didnot induce any obvious stressful state, no specific NARAapproval was required according to Dr. G. Baeverfjord,member of the NARA board and local NARA officer atNofima AS.Swimming performance screening and trainingexperimental setupA total of 1355 fish were individually tagged (Passive Inte-grated Transponder (PIT), Glass tag Unique 2.12 × 12 mm,Jojo Automasjon AS, Sola, Norway) and measured (mass ±S.D. = 40.7 ± 0.2 g and fork lengths = 15.0 ± 0.3 cm) beforeCastro et al. BMC Physiology 2013, 13:1 Page 7 of 12http://www.biomedcentral.com/1472-6793/13/1being graded according to their swimming performance.Groups of approximately 100 fish were placed in a pre-conditioned 1.5 m diameter circular tank with an inner ringto reduce the swimming area to a 40 cm radius (Figure 5).The water inlet to the swimming area was tangentiallysituated on the side of the outer tank so that it generatedthe maximum water velocity. The inner ring was placedon four pieces of PVC (1 cm high) that allowed the waterto drain freely to the center of the tank, while preventingthe fish from leaving the swimming area. Maximum waterinflow generated water velocities of 42–20 cm s-1 nearestthe center, 73–58 cm s-1 in the middle of the stream and97–81 cm s-1 furthest from the center. A grid (paintedmetal meshing) was secured downstream of the water in-let to prevent fish from drifting backwards, and a flood-light placed above the grid encouraged fish to remainupstream. Water velocity and height (10–15 cm) werecontrolled by adjusting the water supply valve and theposition of the draining stand pipe. After being introducedinto the swimming flume, fish were left undisturbed for15 min at the lowest speed to acclimatize. Water speedwas then increased gradually every 1–2 min until half of afish group had been removed from the tank. Fish thatwere unable to swim against the increasing water currenttypically laid against the back-mesh grid. They wereremoved with a dip-net, identified (PIT tag reading) andplaced back in their rearing tank. During the trial, fishwere regularly and gently repositioned to ensure that theywould all be exposed to testing conditions and would notevade from the high speed outer portion of the swimmingring. Based on their swimming performance, fish were thenallocated to one of two groups. The first 50% that stoppedswimming were categorized as “poor swimmers”, and thelast 50% still swimming were the “good swimmers”. Bothpoor and good swimmers were randomly mixed among 16cylindro-conical experimental tanks (500 l, 82 cm in diam-eter, 77–86 fish tank-1) and left undisturbed for one weekbefore the start of the training regimes. The center of eachexperimental tank was fitted with a plastic pipe (31.5 cmdiameter), which reduced the area in the tank with lowestwater speed. A frequency-controlled pump (HanningElektro Werke, PS 18–300; Oerlinghausen, Germany)directed the water current and a wire mesh fence, attachedbetween the pipe and the edge of the tank, prevented thefish from drifting backwards. The water speeds werecalibrated by using the average speed measured at twelvepoints in the tank (four horizontal locations and threedepths at each location (Höntzsch HFA propeller,Waiblingen, Germany with HLOG software)). Five dif-ferent sustained exercise-training regimes were tested; thecontrol regime in quadruplicate tanks and the otherregimes in triplicate tanks. Three of the training regimeswere continuous velocity: the control (C; 5.7 cm s-1),medium intensity (M; 11.5 cm s-1) and high intensity(H; 23 cm s-1) regimes. At start of the 10-week trainingexperiment, these speeds were equivalent to 0.38, 0.77 and1.53 BLs-1 for C, M and H, respectively. As fish grewduring the trial, average relative water speeds werereduced to 0.32 (C), 0.65 (M) and 1.31 (H) BLs-1. Thetwo remaining training regimes used interval training,Figure 5 Swimming performance screening test. Groups of approx. 100 fish were placed in the screening tank each time and allowed toacclimate for 15 minutes at minimum water speed before start of the water speed increments. A plastic mesh (left side) prevented fish fromdrifting backwards. Inner ring was lifted one centimetre with the help of PVC pieces (not on diagram) to allow water drainage into the centre ofthe tank where the main outflow was located. Blue arrows indicate direction of water flow (clockwise).Castro et al. BMC Physiology 2013, 13:1 Page 8 of 12http://www.biomedcentral.com/1472-6793/13/1with daily increments in the relative water velocity from0.32 to 1.31 BLs-1 for either a single 8 h period (Longinterval; Lint) or four 2-h periods for a total high speedtraining period of 8 h (short interval; Sint). Theoretically,both interval groups swam the same distance as the Mgroup. The 10 weeks of training were followed by a one-week recovery at control speed prior transferring the fishto seawater. To stimulate smoltification during the ex-periment, fish were exposed to a short daylight regime(12–12 Light–dark) for the first six weeks, followed bycontinuous light for the remaining five weeks. Measure-ment of ATPase in gills (n = 10) sampled from eachgroup was conducted in a commercial laboratory,Havbruksinstituttet AS, Bergen, Norway, and confirmedthat all fish were sampled within the smolt-window(data not shown). Water temperature was measureddaily (10.5 ± 0.8°C) while oxygen saturation was mea-sured weekly and was maintained over 85% with oxygensupplementation. Dead fish were removed with dailyinspections and weighed.Growth and feed intakeDuring the training experiment, fish were fed to excessan extruded diet based on fish meal, ground wheat andfish oil (produced at Nofima AS, Fyllingsdalen, Bergen,Norway), using automatic feeders. The effluent water ofeach tank was led into a wire mesh box to enable sievingof waste feed. To minimize feed leaching, the effluentwater was directed to two different areas of the wire boxusing pinch valves on the water pipes, dependent onwhether feeding was occurring. The waste feed,expressed as dry matter (DM) content, was used to re-calculate daily feed intake in order to adjust ration levelevery second day (Helland et al., 1996). After the changein the photoperiod at the sixth week of training, feedingregime was increased from 12 (6.25 min h-1) to 24(3.12 min h-1) times per day.Fish were weighed in bulk after six weeks of trainingand a two-day fast. At the end of the trial, fish were indi-vidually re-weighed and re-measured.Viral challenge testFollowing the 1-week detraining period, approx. 110 fishper training regime and training control, including bothperformance groups, were pooled and transferred to sea-water at VESO research station (Vikan, Norway) for anIPN challenge test. An additional nine fish per trainingregime and training control groups were similarly pooledand transferred to act as infection controls (unchal-lenged fish). The fish to be challenged as well as thoseacting as infection controls were acclimatized for oneweek in a separate 1.5 and 1 m3 tank (11 ± 0.2°C and0.5 l kg-1 min-1; water volumes were adjusted to achievesimilar densities). The IPN challenge test was performedby co-habitation and started when 20% of IPN-infectedchallenger fish were added to the experimental tank. Asimilar proportion of uninfected smolts were added tothe unchallenged tank, keeping similar densities in bothtanks. Challenger fish were previously marked by a finclip and injected with 5 ml of ~3 × 106 TCID50/ml ofIPNV (strain V-1244 cultured at the Norwegian Schoolof Veterinary Science). Throughout the challenge test,fish were observed and mortalities were recorded dailyfor 45 days, when the trial was terminated. A representa-tive selection of dead fish was confirmed positive forIPNV and was further bacteriologically examined on 2%NaCl blood agar plates as part of a standard procedureduring the challenge at VESO. All trials at VESO wereperformed according to Norwegian regulations for careand use of fish in research as stated in the Agricultureand Food Department regulation FOR 1996–01–15Nov. 23.Gene expression analyzesFish used for gene expression analyses were sampled atthe end of the IPN challenge test (day 45), when mortal-ity of all groups had leveled off. Challenged fish belong-ing to both swimming performance categories (poor andgood swimmers) were sampled for each of the regimesC, M and Lint, while nine unchallenged control trained(C) fish were used as hybridization controls for micro-array analyses. Cardiac ventricle and head kidney tissueswere immediately dissected from fish killed by a blow tothe head and stored in RNAlater (Ambion, Austin, TX,USA). Total RNA was extracted using TRIzol and puri-fied with PureLink RNA Mini Kit (Invitrogen, Carlsbad,CA, USA) following manufacturers guidelines. RNA con-centration was measured using NanoDrop 1000 spectro-photometer (Thermo Fischer Scientific, Waltham, MA,USA), and RNA integrity was assessed with an Agilent2100 Bioanalyzer (Agilent Technologies, Waldbronn,Germany). All samples had a RNA Integrity Number (RIN)above 9.Microarray analyses were performed with the salmonidoligonucleotide microarray (SIQ2.0, NCBI GEO platformGPL10679), consisting of 21 K features printed in dupli-cate on 4 × 44 K arrays from Agilent Technologies [38].Eighteen two-color microarray hybridizations were per-formed. Individual heart ventricle samples of challengedfish from poor and good swimming performers (n = 9each; 3 from each training regime) were competitivelyhybridized against a pooled sample consisting of equalamounts of RNA from the infection controls (n = 9) perarray. Unless specified otherwise, all reagents and equip-ment were from Agilent Technologies and used accord-ing to manufacturer’s protocol. Amplification andCastro et al. BMC Physiology 2013, 13:1 Page 9 of 12http://www.biomedcentral.com/1472-6793/13/1labeling of RNA (200 ng) were done using theLowInput-QuickAmp Labeling kit. Cy5 (test) and Cy3(control) labeled RNA was purified with RNeasy MiniKit (Qiagen, MD, USA), and the Gene ExpressionHybridization Kit was used for RNA fragmentation.Hybridizations were performed for 17 h at 65°C and ro-tation speed of 10 rpm in a hybridization oven. Arrayswere washed twice (Gene Expression wash buffers 1 &2) and immediately scanned with a GenePix 4100A(Molecular Devices, Sunnyvale, CA, USA) at 5 μm reso-lution and with manually adjusted laser power to ensurean overall intensity ratio close to unity between Cy3 andCy5 channels and with minimal saturation of features.The GenePix Pro 6.1 software was used for spot-gridalignment, feature extraction of fluorescence intensityvalues and spot quality assessment. Low quality spotswere filtered with aid of GenePix flags as well as by thecriterion (I-B)/(SI-SB) ≥0.6 where I and B are the meansignal and background intensities, respectively, and SIand SB are the respective standard deviations. Data wereexported into the STARS platform [38] for data trans-formation and Lowess normalization of log2-expressionratios (ER). Differentially expressed genes were selectedbased on spot signal threshold, log2-ER average > |0.6| inat least two individuals and significant difference be-tween groups (swimming performers) p < 0.05, n = 9Table 2 List of primers used for the qPCR reactionsGenes Short name Sequence 5' to 3' Accessionradical S-adenosyl methionine domain-containing protein 21 Rsad2 F-GTACCGCAGATGCACAACAC AF076620R-TTGACACTGCTTGGAGTTGCinterferon inducible protein Gig11 Gig1 F-GGCAACCTGAATCCAGAAGA DW569595R-GTCTGGACGCAGACTGATGAVHSV-inducible protein1 Vhsv2 F-GGTGAAGACCTGGACCTGAA BT072288R-TGACCCCTGTTGACCTTCTCinterferon-induced protein with tetratricopeptide repeats 51 Ifit5 F-CAGAGAGGTGCCAGGCTAAC BT046021R-TGCACATTGACTCTCCTTGGinterferon induced with helicase C domain 11 Mda5 F-CAGAGGTGGGGTTCAATGAT NM001195179R-AGCTCGCTCCACTTGTTGATDEAD/H (Asp-Glu-Ala-Asp/His) box polypeptide RIG-I1 Rig-I F-GACGGTCAGCAGGGTGTACT DY714827R-CCCGTGTCCTAACGAACAGTbarrier-to-autointegration factor1 Baf F-ACAGACCCCTCATCATCCTG BT049316R-CGGTGCTTTTGAGAAGTGGTsignal transducer and activator of transcription 1 isoform alpha1 Stat1a F-CGGTGGAGCCCTACACTAAG CB513054R-GGGATCCTGGGGTAGAGGTAinterferon-inducible protein Gig2-like1 Gig2 F-GATGTTTCATGGCTGCTCAA BT044026R-CTTTTCGGATGTCCCGACTAB-type natriuretic peptide2 BNP F-TCGACAAATCCGCAATAAGA CK883650R-TTGAGCCAATTCGGTCTAGCputative collagen alpha 12 put_Coll-a1 F-AACCCTGAACCCCTCAGTCT CA038317R-TGGTCCTACCGTCTGGTTTCleukocyte elastase inhibitor2 LEI F-TCTCAGATGGCAAAGGCTCT BT045959R-GTTGGCCAGTTTCAGGATGTelongation factor 1a3 EF1α F-CACCACGGGCCATCTGATCTACAA BT072490R-TCAGCAGCCTCCTTCTCGAACTTC18S rRNA3 18S F-GCCCTATCAACTTTCGATGGTAC AJ427629R-TTTGGATGTGGTAGCCGTTTCTCinfectious pancreatic necrosis virus_polyprotein4 IPNV F-CCGACCGAGAACAT AJ877117R-TGACAGCTTGACCCTGGTGAT1 Virus-responsive genes (Figure 4).2 Genes used for microarray validation.3 Reference genes used for normalization.4 IPNV gene used for relative virus quantification (described in Results section).Castro et al. BMC Physiology 2013, 13:1 Page 10 of 12http://www.biomedcentral.com/1472-6793/13/1(Student’s t-test). Recording of microarray experimentmetadata was in compliance with the Minimum Infor-mation About a Microarray Experiment (MIAME)guidelines (Brazma et al., 2001). Microarray results weresubmitted to GEO (GSE38603).Expression of single genes (VRGs) was assessed byquantitative real-time RT-PCR (qPCR) on heart ventricleand head kidney samples from swim performers of thethree groups C, M and Lint at day 45 post-challenge. Ex-pression levels were calculated relative to expressionlevels of the same nine infection control samples usedfor microarrays. A total of six and eight VRGs were ana-lyzed for heart (n = 16 per regime; 8 good and 8 poorswimmers) and head kidney (n = 10 per regime; 5 goodand 5 poor swimmers), respectively. The same samplesfrom both tissues were further scanned for viral pres-ence by using a set of specific IPNV primers (Table 2).Confirmation of array results by qPCR was based oneight genes (4 up- and 4 down-regulated) in the same in-dividual samples as used for microarrays.All qPCR primers were designed using the ePrimer3from the EMBOSS online package [39], except for theIPNV primers [40] and synthesized by Invitrogen(Table 2). Synthesis of cDNA was performed on 0.5 μgof DNAse treated (DNA-free; Ambion) RNA samplesusing TaqMan@ Reverse Transcription reagents (AppliedBiosystems, Foster City, CA, USA) and primed with anequal mix of oligo dT and random hexamers. PCR reac-tions were prepared manually and run in duplicates in96-well optical plates on a LightCyclerW 480 (RocheDiagnostics, Mainheim, Germany) using 2X SYBR GreenMaster mix (Roche), 5 μl of cDNA samples and a primerconcentration of 0.42 μΜ each in a final volume of12 μl. For all genes, cDNA was previously diluted 1:10(1:1000 for 18S). qPCR thermal cycling was as follows:5 min pre-incubation at 95°C, followed by 45 amplifica-tion cycles consisting of 95°C for 10 s, 60°C for 15 s and72°C for 15 s, followed by a melting curve protocol (95°Cfor 5 s, continuous increase from 65°C to 97°C) to assessspecificity of the amplicon. Fluorescence was measured atthe end of every extension step and throughout the mel-ting curve step. Cycle threshold (CT) values were calcu-lated using the second derivative method. Duplicatereactions differing more than 0.5 CT values were dis-carded, and values were averaged for relative quantifica-tion. PCR efficiency was assessed by six 10-fold serialdilutions of pooled sample templates for each primer pair.Relative expression ratios were calculated by thePfaffl method [41] with normalization against tworeference genes (18S and Elongation factor 1α). Anindex value of VRG expression was calculated foreach group (training regime or swimming perfor-mance) by averaging the relative expression ratio ofthe single genes.Calculation and statisticsRelative feed intake: 100 × (dry feed intake/mean bodymass (BM)/days fed).TGC: 1000 × [(BM10.33 – BM00.33)/Pday-degrees],where BM1 and BM0 are final and initial body masses,respectively.FER: (Wet fish gain + dead fish mass)/dry feed intake.CF: 100 × BM (g) × fork length (cm)-3.For growth and CF analyses, the individual fish datawere analyzed by analysis of variance in a hierarchicalmodel including the fixed effect of training regime andthe random effect of tank within regime. The mean datafor each tank were tested by variance analysis (meanscompared using the least-squares means procedure)(SAS software, version 9.1, SAS Institute, Inc., Cary, NC,USA). Percentage data were transformed (arcsine squareroot) before being subjected to analysis. Differences be-tween training regimes were considered significant atthe p < 0.05 level, and are presented as mean ± SEM.Differences in survival during the IPN challenge testwere evaluated using the Mantel-Cox test in GraphPadPrism (version 5.01, GraphPad Software, Inc., San Diego,CA, USA). For the microarray analyses, expression dif-ferences between the groups where assessed by Student’st-test; p < 0.05, and data are presented as log2ER ± SEM.Difference in expression levels for the indexed values ofpooled VRGs, was assessed by paired Student’s t-test,p < 0.05; between target and control groups in qPCR.Correlation between microarray and qPCR results forselected genes was assessed by Pearsons’ r.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsVC carried out the molecular studies, participated in samplings, datainterpretation and drafted the manuscript. BG performed the growth andnutrition studies analyses. SMJ interpreted the microarray data and draftedthe corresponding section. JH participated in the design of the study andrevised the manuscript. GC designed the screening test facilities, performedthe classification of fish according to swimming capacities, and revised themanuscript. APF participated in designing the study and critically revised themanuscript. AK analyzed the microarray data and revised the manuscript. SJHdesigned the fish training facilities and performed and participated ingrowth analyzes. HT obtained the funding, conceived and designed thestudy, coordinated and participated in samplings, data interpretation anddrafting of the manuscript. All authors have read and approved the finalmanuscript.AcknowledgementsWe would like to thank the technical staff at Nofima for feed production,caring of the fish and laboratory analyses. This study was funded by TheFishery and Aquaculture Industry Research Fund (FHF) and The ResearchCouncil of Norway (Grant number: 190067).Author details1Nofima, Ås, Norway. 2Institute of Animal Sciences, Norwegian University ofLife Sciences (UMB), Ås, Norway. 3AVS Chile S.A., Puerto Varas, Chile.4Aquaculture Protein Centre, CoE, Ås, Norway. 5Nofima, Ås, Norway.6Norwegian University of Science and Technology, Faculty of Medicine,Trondheim, Norway. 7Université de Bretagne Occidentale, LEMAR, Unité dePhysiologie Fonctionnelle des Organismes Marins, Ifremer, Plouzané, France.Castro et al. BMC Physiology 2013, 13:1 Page 11 of 12http://www.biomedcentral.com/1472-6793/13/18Faculty of Land and Food Systems, & Department of Zoology, University ofBritish Columbia, Vancouver, BC, Canada.Received: 5 March 2012 Accepted: 16 January 2013Published: 21 January 2013References1. Johnson SC, Treasurer JW, Bravo S, Nagasawa K, Kabata Z: A review of theimpact of parasitic copepods on marine aquaculture. Zool Stud 2004,43:229–243.2. Robertsen B: Can we get the upper hand on viral diseases in aquacultureof Atlantic salmon? Aquac Res 2011, 42:125–131.3. Poppe TT, Taksdal T: Ventricular hypoplasia in farmed Atlantic salmonSalmo salar. Dis Aquat Organ 2000, 42:35–40.4. Brocklebank J, Raverty S: Sudden mortality caused by cardiac deformitiesfollowing seining of preharvest farmed Atlantic salmon (Salmo salar) andby cardiomyopathy of postintraperitoneally vaccinated Atlantic salmonparr in British Columbia. Can Vet J 2002, 43:129–130.5. Hochachka PW: The effects of physical training on oxygen debt andglycogen reserves in trout. Can J Zool 1961, 39:767–776.6. Davie PS, Wells RMG, Tetens V: Effects of sustained swimming onrainbow-trout muscle structure, blood-oxygen transport, and lactate-dehydrogenase isozymes - evidence for increased aerobic capacity ofwhite muscle. J Exp Zool 1986, 237:159–171.7. Farrell AP, Johansen JA, Steffensen JF, Moyes CD, West TG, Suarez RK:Effects of exercise training and coronary ablation on swimmingperformance, heart size, and cardiac enzymes in Rainbow-Trout,Oncorhynchus-Mykiss. Can J Zool 1990, 68:1174–1179.8. Gallaugher PE, Thorarensen H, Kiessling A, Farrell AP: Effects of highintensity exercise training on cardiovascular function, oxygen uptake,internal oxygen transport and osmotic balance in chinook salmon(Oncorhynchus tshawytscha) during critical speed swimming. J Exp Biol2001, 204:2861–2872.9. Anttila K, Jaevilehto M, Manttari S: The swimming performance of browntrout and whitefish: the effects of exercise on Ca2+ handling andoxidative capacity of swimming muscles. J Comp Physiol B 2008,178:465–475.10. Totland GK, Fjelldal PG, Kryvi H, Lokka G, Wargelius A, Sagstad A, Hansen T,Grotmol S: Sustained swimming increases the mineral content andosteocyte density of salmon vertebral bone. J Anat 2011, 219:490–501.11. Khovanskiy IY, Natochin YV, Shakhmatova YI: Effect of physical exercise onosmoregulatory capability in hatchery-reared juvenile chum salmon,Oncorhynchus keta. J Ichthyol 1993, 33:36–43.12. Castro V, Grisdale-Helland B, Helland SJ, Kristensen T, Jorgensen SM,Helgerud J, Claireaux G, Farrell AP, Krasnov A, Takle H: Aerobic trainingstimulates growth and promotes disease resistance in Atlantic salmon(Salmo salar). Comp Biochem Phys A 2011, 160:278–290.13. Gleeson M: Immune function in sport and exercise. J Appl Physiol 2007,103:693–699.14. Mathur N, Pedersen BK: Exercise as a mean to control low-grade systemicinflammation. Mediat Inflamm 2008, doi:10.1155/2008/109502.15. Davison W: The effects of exercise training on teleost fish, a review ofrecent literature. Comp Biochem Physiol A 1997, 117:67–75.16. Nakagawa H, Nishino H, Nematipour GR, Ohya S, Shimizu T, Horikawa Y,Yamamoto S: Effects of water velocities on lipid reserves in Ayu.Nippon Suisan Gakk 1991, 57:1737–1741.17. Hinterleitner S, Huber M, Lackner R, Wieser W: Systemic and enzymaticresponses to endurance training in 2 Cyprinid species with different life-styles (Teleostei, Cyprinidae). Can J Fish Aquat Sci 1992, 49:110–115.18. Young PS, Cech JJ: Improved growth, swimming performance, andmuscular development in exercise-conditioned young-of-the-yearstriped bass (Morone saxatilis). Can J Fish Aquat Sci 1993, 50:703–707.19. Ibarz A, Felip O, Fernandez-Borras J, Martin-Perez M, Blasco J, Torrella JR:Sustained swimming improves muscle growth and cellularity in giltheadsea bream. J Comp Physiol B 2011, 181:209–217.20. Palstra AP, Tudorache C, Rovira M, Brittijn SA, Burgerhout E, van den ThillartGEEJ, Spaink HP, Planas JV: Establishing zebrafish as a novel exercisemodel: swimming economy, swimming-enhanced growth and musclegrowth marker gene expression. PLoS One 2010, 5. doi:10.1371/journal.pone.0014483.21. Merrifield DL, Dimitroglou A, Foey A, Davies SJ, Baker RTM, Bogwald J,Castex M, Ringo E: The current status and future focus of probiotic andprebiotic applications for salmonids. Aquaculture 2010, 302:1–18.22. Sweetman JW, Torrecillas S, Dimitroglou A, Rider S, Davies SJ, Izquierdo MS:Enhancing the natural defences and barrier protection of aquaculturespecies. Aquac Res 2010, 41:345–355.23. Gjedrem T: Genetic improvement of cold-water fish species. Aquac Res2000, 31:25–33.24. Overturf K, LaPatra S, Towner R, Campbell N, Narum S: Relationshipsbetween growth and disease resistance in rainbow trout, Oncorhynchusmykiss (Walbaum). J Fish Dis 2010, 33:321–329.25. Tucker VA: Energetic cost of locomotion in animals. Comp Biochem Physiol1970, 34:841–846.26. Nieman DC: Exercise effects on systemic immunity. Immunol Cell Biol2000, 78:496–501.27. Claireaux G, McKenzie DJ, Genge AG, Chatelier A, Aubin J, Farrell AP:Linking swimming performance, cardiac pumping ability and cardiacanatomy in rainbow trout. J Exp Biol 2005, 208:1775–1784.28. Krasnov A, Timmerhaus G, Schiotz BL, Torgersen J, Alanasyev S, Iliev D,Jorgensen J, Takle H, Jorgensen SM: Genomic survey of early responses toviruses in Atlantic salmon, Salmo salar L. Mol Immunol 2011, 49:163–174.29. Jobling M, Baardvik BM, Christiansen JS, Jorgensen EH: The effects ofprolonged exercise training on growth performance and productionparameters in fish. Aquacult Int 1993, 1:95–111.30. Greer Walker M, Emerson L: Sustained swimming speeds and myotomalmuscle function in trout, Salmo gairdneri. J Fish Biol 1978, 13:475–481.31. East P, Magnan P: The effect of locomotor-activity on the growth ofbrook charr, Salvelinus fontinalis Mitchill. Can J Zool 1987, 65:843–846.32. Young PS, Cech JJ: Optimum exercise conditioning velocity for growth,muscular development, and swimming performance in young-of-the-year striped bass (Morone saxatilis). Can J Fish Aquat Sci 1994,51:1519–1527.33. Thorarensen H, Gallaugher PE, Kiessling AK, Farrell AP: Intestinal blood-flowin swimming chinook salmon Oncorhynchus tshawytscha and the effectsof hematocrit on blood-flow distribution. J Exp Biol 1993, 179:115–129.34. Farrell AP, Thorarensen H, Axelsson M, Crocker CE, Gamperl AK, Cech JJ: Gutblood flow in fish during exercise and severe hypercapnia. CompBiochem Physiol A 2001, 128:551–563.35. Muir BS, Kendall JI: Structural modifications in gills of tunas and someother oceanic fishes. Copeia 1968, 2:388–398.36. Farrell AP, Steffensen JF: An analysis of the energetic cost of the branchialand cardiac pumps during sustained swimming in trout. Fish PhysiolBiochem 1987, 4:73–79.37. Steffensen JF: The transition between branchial pumping and ramventilation in fishes - energetic consequences and dependence onwater oxygen-tension. J Exp Biol 1985, 114:141–150.38. Krasnov A, Timmerhaus G, Afanasyev S, Jørgensen SM: Development andassessment of oligonucleotide microarrays for Atlantic salmon (Salmosalar L.). Comp Biochem Phys D 2011, 6:31–38.39. Rice P, Longden I, Bleasby A: EMBOSS: The European molecular biologyopen software suite. Trends Genet 2000, 16:276–277.40. Skjesol A, Skjæveland I, Elnæs M, Timmerhaus G, Fredriksen BN, JørgensenSM, Krasnov A, Jørgensen JB: IPNV with high and low virulence: hostimmune responses and viral mutations during infection. Virol J 2011,8:396.41. Pfaffl MW, Horgan GW, Dempfle L: Relative expression software tool(REST©) for group-wise comparison and statistical analysis of relativeexpression results in real-time PCR. Nucleic Acids Res 2002, 30:e36.doi:10.1093/nar/30.9.e36.doi:10.1186/1472-6793-13-1Cite this article as: Castro et al.: Disease resistance is related to inherentswimming performance in Atlantic salmon. BMC Physiology 2013 13:1.Castro et al. BMC Physiology 2013, 13:1 Page 12 of 12http://www.biomedcentral.com/1472-6793/13/1


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items