Internet & Web

9 pages
4 views

Feeding activity and opercular pressure transients in Atlantic salmon ( Salmo salar L.): application to feeding management in fish farming

Please download to get full document.

View again

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
Ecological and economical sustainability of marine aquaculture operations depend on proper feeding management. Feed wastage from overfeeding is a source of pollution, represents futile use of precious marine resources, and undermines the economic
Transcript
  FISH TELEMETRY Feeding activity and opercular pressure transientsin Atlantic salmon (  Salmo salar   L.): application to feedingmanagement in fish farming Jo Arve Alfredsen   Ba˚rd Holand   Torfinn Solvang-Garten   Ingebrigt Uglem   Springer Science+Business Media B.V. 2007 Abstract  Ecological and economical sustainabil-ity of marine aquaculture operations depend onproper feeding management. Feed wastage fromoverfeeding is a source of pollution, representsfutile use of precious marine resources, andundermines the economic viability of operations.Additionally, underfeeding reduces growth andfish welfare. Finding an optimal feeding regime interms of temporal and spatial distribution of thefeed ration require intimate knowledge of theindividual feeding behaviour of fish sustainingintensive culturing conditions. Fish telemetry hasproved to be a valuable tool for studying spatialbehaviour in sea cages, however there are cur-rently no practical methods available with respectto detection of actual feed intake in fish on theindividual level. The present study investigatespressure transients arising in the opercular cavityof farmed Atlantic salmon ( Salmo salar   L.) inconnection with feeding, and whether such mea-surements can serve as an indication of feedingestion. A technical solution to the sensingproblem based on a differential pressure trans-ducer is presented along with typical pressuresignal traces obtained during feeding in a hardwire tank experiment. Measurements showedconsiderable variation of sub-ambient pressuretransients (1.5 kPa ± 0.95) and their duration(519 ms ± 117), suggesting that the fish modulatesits strike intensity depending on the particularfeeding situation. Despite variations in scale,opercular pressure waveforms have distinct struc-tural features that repeat between feeding in-stants. From a signal processing point of viewwaveforms provide sufficient information withrespect to isolation and detection of feedingincidents, which is important with respect to apotential implementation of the sensing principlein a telemetry tag design. Issues regarding devel-opment and application of a telemetry systembased on this sensing principle are discussed. Keywords  Feeding behaviour    Feedingmanagement    Opercular pressure    Telemetry   Atlantic salmon Introduction Fish farming, using primarily sea cages, usuallyinvolves massive aggregations of fish in relatively Guest editors: P. R. Almeida, B. R. Quintella,M. J. Costa and A. MooreDevelopments in Fish TelemetryJ. A. Alfredsen ( & )    B. Holand    T. Solvang-GartenDepartment of Engineering Cybernetics, TheNorwegian University of Science and Technology,NO-7491 Trondheim, Norwaye-mail: jo.arve.alfredsen@itk.ntnu.noI. UglemNorwegian Institute for Nature Research, Tungasletta2, NO-7485 Trondheim, Norway  1 3 Hydrobiologia (2007) 582:199–207DOI 10.1007/s10750-006-0554-9  small volumes, where living conditions and fishbehaviour deviate significantly from the condi-tions experienced and the behaviour exhibited inthe natural environment. For example, outgrowthof Atlantic salmon ( Salmo salar   L.) is typicallyconducted in sea cages. Investigation of farm fishbehaviour is important because it reflects criticalinformation about fish welfare and represents thepremises for efficient farm management in termsof optimum feeding regimes and sustainableproduction. The behaviour of intensively culturedfish can not be extrapolated from knowledgeabout free-ranging fish, which suggests that effortshould be invested in adaptation and develop-ment of suitable methods for monitoring fishbehaviour under such conditions. As an effectivetool for studying fish behaviour on the individuallevel, fish telemetry has already found severaluseful applications within aquaculture research ingeneral (Juell & Westerberg, 1993; Baras &Lagarde `re, 1995; Begout Anras, Kadri, Juell &Hansen, 2000; McFarlane et al., 2002; Cooke,Thorstad & Hinch, 2004).Feeding management in fish farms encom-passes the control of ration size and type as wellas the temporal and spatial delivery of feed.Improper feeding management has serious impacton both productivity and environment, and theindustry is reported to have considerable poten-tial for improvement (Houlihan, Boujard &Jobling, 2001). Overfeeding weakens the feedconversion ratio (FCR = kg. feed used/kg. bio-mass gained), pollutes the environment throughmicrobial spoilage of waste feed and spread of feed particles to populations of wild fish, andcauses waste of precious marine protein and lipidresources. On the other hand, underfeedingresults in reduced growth, elevated competitionand stress among the fish, and a weakened FCR(Ang & Petrell, 1997; Talbot, Corneillie, &Korsøen, 1999). Optimum feeding managementis a complicated decision process where theamount of information available for makingthe decisions may be limited. The ability of thefarmer to assess the size and amount of fish in thecages in order to determine an adequate pelletsize and ration (enough feed to support growthpotential) is obviously of great importance. How-ever, the decision regarding the temporal andspatial delivery of feed and the correspondingfeeding behaviour of fish during a particular meal,depending on the hunger level of the fish andenvironmental conditions such as light level andwater temperature, is what ultimately determinefeeding efficiency (Alver, Alfredsen, & Sigholt,2004).The problem of finding an optimum deliveryregime represents an active field of research andtechnological development. Most investigationsare based on the population level perspective of fish feeding, such as through controlled large scalefeeding experiments (Jobling et al., 2001). How-ever, observations made at the population levelwill primarily provide aggregated data on how thefish respond to feed, i.e. the relationship betweengross feed consumption and the average responseto a particular feature of the feeding strategy orenvironmental condition. These are averageresponses where signs of individual variationand interactions within the population are essen-tially hidden. In order to fully understand theunderlying mechanisms that govern feeding effi-ciency, observations should also be obtained atthe individual level (Jobling et al., 2001). Obser-vations of individuals or specific groups of fish canprovide valuable information on important mat-ters such as intrinsic feeding rates, feeding histo-ries, and consequences of social hierarchy. Thissuggests that the possibilities of fish telemetryshould be assessed with respect to observingindividual feeding behaviour and feed intake infish cages.Key to the development of a ‘‘feeding sensi-tive’’ telemetry system is the identification anddetection of a distinct signal that somehowreflects the act of feed ingestion (feeding signal).Once the detection principle is established itsfurther implementation in a conventional acoustictransmitter based system should be attainable.Ideally, the system should be able to reliablydetect every (and only those) feed pellets passinginto the tagged fish’ stomach. However, detectionof alternative signals giving indirect indications of feed ingestion should also be considered, partic-ularly if such feeding signals can lead to a robustsolution with respect to the sensing principle at anacceptable signal to noise ratio. In any case, thebalance between measurement accuracy, com- 200 Hydrobiologia (2007) 582:199–207  1 3  plexity, and invasiveness of the required technol-ogy need to be assessed.Fish use three different modes of prey capture:ram feeding, suction feeding, and manipulation(Norton & Brainerd, 1992). Suction feeding, inwhich the predator causes prey to be dragged intothe mouth through rapid buccal expansion, is theprincipal strike mode among many predatory fish(Alexander, 1967; Nemeth, 1997). In experimentsconducted with rainbow trout,  Oncorhynchusmykiss  (Van Leeuwen, 1984), it was demonstratedthat suction is an important component of theprey capture process. However, most fish exercisea combination of suction and ram feeding (Norton& Brainerd, 1992), as is probably the case forsalmon when presented with formulated feedpellets. As regards the question of identifyingpossible feeding signals, the production of pres-sure transients in the fish’s mouth during feedingmay turn out as a possible indicator. The buccaland opercular cavities are hydrodynamically con-nected through the gill archs (Van Leeuwen,1984) and the opercularcavityis readilyaccessiblefor pressure measurements by means of cannula-tion through the anterior wall of the cleithrum(Shelton, 1970; Norton & Brainerd, 1992). It isprobable that the invasiveness and complexity of this kind of instrumentation is feasible withrespect to implementation in a future telemetrysystem. However, the information present in thepressure signals must be analysed and reliablemethods of signal interpretation must be devel-oped. The present study investigates the charac-teristics of the pressure signal in the opercularcavity of Atlantic salmon during ingestion of typical feed pellets. The feasibility of this kind of instrumentation in a fish telemetry system andsome technical considerations with respect to itsimplementation are also discussed. Materials and methods AnimalsAtlantic salmon are the most important speciesemployed in Norwegian aquaculture. Two spec-imens (A: standard length 52 cm, weight 1540 g;B: 50 cm, 1388 g) were selected and prepared forthe investigations of this study by installing apressure sensing device in their opercular cavity.In addition, a number of similar fish wereemployed for the purpose of experimental con-trol. The fish were acquired from a tank rearedpopulation of the Salmobreed-strain at Akvaf-orsk, Sunndalsøra, and transferred to BrattøraResearch Centre, Trondheim, where they werekept in rectangular 1,000 l flow-through tanks. Inorder to facilitate acclimation the fish were fedformulated feed pellets of the same size andquality (Skretting Nutra, 4 mm) as they were fedprior to transport. Feed delivery was adminis-tered by use of automatic feed dispensers threetimes per day (2.5% of body weight). Consistentfeed intake at the level prior to transport wasconsidered as evidence of acclimation, whichoccurred after four weeks. During experimentsinstrumented fish were accompanied by unalteredfish as control and to facilitate feeding behaviour.Temperatures during acclimation and the exper-iments were in the range 8.0–8.2  C. Animals werehoused and maintained according to nationalregulations and protocols applying to animal useand care.SurgeryPrior to surgery fish were anaesthetized byimmersion in a solution of 2-phenoxyethanol(2-PE, 0.65 ml/l). Following anaesthesia the fishwere transferred to specialized cradle wheresurgery could be performed with the fish’ gillspartially submerged in pure sea water. Waterwas poured over the fish repeatedly to keep theirskin moist during surgery, and the surgicalprocedure was timed to last about six minutes.After surgery the fish were revived in theirrespective holding tanks. The fish returned to astable position and normal respiration rateapproximately three hours after revival, andfeeding resumed after 24 h. The experimentemployed a ‘‘hard wired’’ experimental setupwhere the fish were physically connected to thedata recording equipment through a signal cable(diameter 2.8 mm, 3 wire + shield). Low frictionfour terminal slip rings (Poly-Scientific, AC6373)were introduced between the tagged fish and therecording equipment in order to avoid excessive Hydrobiologia (2007) 582:199–207 201  1 3  twisting of the cable as the fish moved around.After an initial period of habituation the cabledid not seem to affect the behaviour of the fishnoticeably.Opercular pressure was measured by insertinga medical grade plastic cannula (Venflon, BDMedical Systems, 1.1 mm outer diameter,0.8 mm inner diameter) into the opercular cavitywith the distal tip of the cannula protrudingapproximately 5 mm past the anterior wall of thecleithrum. This technique is similar to thetechnique described e.g. by Shelton (1970). Thecannula was positioned by inserting a hypoder-mic needle (2.4 mm outer diameter, 1.7 mminner diameter) through the tissue beneath thecleithrum. The needle carried the plastic cannulainto place and was subsequently removed. Theexternal portion of the cannula was fixed to theskin by four sutures (2/0 silk) 15 mm dorsally of the pectoral fin.The end of the cannula was flanged to 11 cmof flexible Tygon R3603 tubing (2.4 mm outerdiameter, 0.8 mm inner diameter). The remoteend of the tubing was attached to the positivepressure port of a miniature differential pressuresensor (Honeywell 26PC01 SMT, pressure ran-ge ± 6.9 kPa, frequency response 1 kHz). Thenegative port of the sensor was left open makingambient water pressure reference for the pres-sure measurements. Interior parts of cannula,tubing and pressure port were filled completelyby degassed fluid to assure adequate incompress-ible coupling to the pressure sensitive membraneof the sensor. Prior to surgery, the pressuresensor had been connected to an IC instrumen-tation amplifier (Texas Instruments, INA321) foramplification of the pressure signal, and bothcomponents were mounted on a purpose builtprinted circuit board and subsequently mouldedinto a solid epoxy cylinder of length 32 mm anddiameter 13 mm (denoted as the tag). The tagwas then attached to the fish by passing two finesurgical steel wires guided by two hypodermicneedles through the epaxial musculature justbelow the dorsal fin. The steel wires were tightlywrapped around the tag prior to surgery causingfirm attachment of the tag to the fish body. Anoutline of the arrangement can be seen in Fig. 1.ObservationsAfter the surgical procedure the pressure signal of the opercular cavity was readily accessiblethrough the flexible cable extending from thetag. The cable was connected to dedicated com-puter hardware for AD conversion of thepressure signal (National Instruments, DAQCard6062E, sampling rate 100 Hz, resolution 12 bit).Purpose built computer software using the Lab-VIEW programming system (National Instru-ments) was developed permitting real timevisualization of pressure waveforms and datarecording. Pressure data time series were re-corded continuously during feeding episodes.Measurements were adjusted for the differencein depth between the tip of the cannula and thereference pressure port in the tag on the fish’back.Feeding behaviour was observed by employ-ing multiple underwater cameras (Type WA-TEC 107LH) in the fish tank. The camerapermitted a close view of the feeding fishwithout causing any disturbance. Video se-quences of the feeding fish were recorded toDVDs for subsequent analysis and correlationwith pressure recordings. Synchronization of video recordings and time series of opercularpressure was secured by adding accurate timeinformation to both videos and data files.Analysis was performed by identifying positiveoccurrences of feed ingestion on the videorecordings and inspecting the associated pres-sure signals looking for features of the wave-forms that distinguish feeding events. Fig. 1  Outline of the instrumentation of the fish used inthe experiment202 Hydrobiologia (2007) 582:199–207  1 3  Results The results presented here are based on mea-surements and observations of fish A. Fish B wastreated in a similar manner, but served mainly asa backup in case of technical failure and as anindicator with respect to detection of potentialsources of variability arising from the tagattachment procedure. However, variability be-tween two individuals appeared to be substan-tially lower than the intra-individual variabilityexperienced.Feeding related opercular pressure data onsalmonids are sparse. In order to get an impres-sion of the validity of the pressure measurements,it was decided to observe the fish’s respiratorywaveform first. Respiration in salmonids, and fishin general, is relatively well documented bypressure waveforms of the buccal and opercularcavities, EMG recordings from the musclesinvolved, and visual observations of moving skullstructures (Alexander, 1967; Shelton, 1970; Os-wald, 1978). Figure 2 shows three representativeperiods of the respiratory cycle as measured bythe tag. The most conspicuous property of thewaveform are the regular subambient drops inpressure which are known to correspond to thephase of the respiratory cycle where the opercularsuction pump is active (sucking water over thegills). The waveform also contains a relatively flatregion where the buccal pressure pump is active(pressing water over the gills). The flat region ispreceded by a ‘‘knee’’ and followed by a minorpeak which indicates the transitional phasebetween the two distinct pumping actions. Somenoise is present in the signal, but its magnitude istoo small to obscure the main features of thewaveform. The noise is apparently quantizationeffects due to the small signal amplitude of therespiration signal. Respiration is actually belowthe pressure range for which the tag electronics isdesigned to work optimally.The fish commenced feeding within 24 h of surgery. 4 mm feed pellets were delivered from anautomatic feed dispenser and the fish’s responsewas recorded on video while the correspondingpressure signal was recorded by the computersoftware. Positive feeding events were noted forsubsequent analyses. Generally, the opercularpressure during feed capture and ingestionshowed high variability with respect to the mag-nitude of the subambient pressure transient( P   = [0.8,4.6] kPa,   P   = 1.5 kPa,s.d. = ±0.95 kPa, N   = 16). The transients were, however, of consid-erably greater magnitude than corresponding su-bambient transients of the respiration signal. Also,the shape of the waveform that followed after theinitial transient endured substantial fluctuations.Such fluctuations are probably caused by irregularand rapidly changing water flows in the fish mouthas a result of manipulation and ingestion of thecaptured feed pellet. The duration of the subambi-ent transient was also investigated and it was foundthat this parameter endured less variation thanwhat was observed with the magnitude of thepressure transient ( t   = [360, 829] ms,   t   = 519 ms,s.d. = ±117 ms,  N   = 16). Compared to other spe-cies that have been investigated (non-salmonids)thesubambientopercularpressureofsalmonseemsto be of longer duration.Figure 3 shows a selection of four pressuresignatures of varying magnitude recorded duringthe experiment. The waveforms show commonfeatures with respect to their shape when differ-ences in scale are disregarded. The transientphase seems to be preceded by a short period of elevated pressure, presumably caused by a pre-paratory phase of buccal compression leading thestrike. Similar observations have been done withother species, e.g.  Micropterus salmoides  and Cichla ocellaris  of the Centrarchidae family 0 500 1000 1500 2000 2500 3000−250−200−150−100−50050100 Time (ms)    P  r  e  s  s  u  r  e   (   P  a   ) Fig. 2  Pressure measurements in the opercular cavityduring three cycles of respirationHydrobiologia (2007) 582:199–207 203  1 3
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x