Effects on growth, survival and bone development from start feeding lumpsucker (Cyclopterus lumpus) larvae with Artemia, copepods and fomulated feed

Transcription

1 Effects on growth, survivl nd one development from strt feeding lumpsucker (Cyclopterus lumpus) lrve with Artemi, copepods nd fomulted feed Julie Terese Hnssen Mrine Costl Development Sumission dte: June 2018 Supervisor: Elin Kjørsvik, IBI Co-supervisor: Kjell Inge Reitn, IBI Jn Ove Evjemo, IBI Norwegin University of Science nd Technology Deprtment of Biology

2

3 Acknowledgements Acknowledgements This Mster s thesis ws written t the Deprtment of Biology, t the Norwegin University of Science nd Technology (NTNU), Trondheim. The experimentl work ws performed t Centre of Fisheries nd Aquculture (Sel), NTNU. The writing of this thesis hs een under the guidnce of professor Elin Kjørsvik t the Deprtment of Biology NTNU nd co-supervisors Kjell Inge Reitn t Deprtment of Biology t NTNU nd Jn Ove Evjemo t SINTEF Ocen. I would like to thnk Elin Kjørsvik for excellent supervision, constructive feedck nd encourgement during oth experimentl work nd writing the thesis. Thnks to Kjell Inge Reitn nd Jn Ove Evjemo for vlule feedck. Thnks to Jn Ove Evjemo for lso helping with decpsultion of Artemi, Kjersti Rennn Dhl for guidnce with cultivtion of microlge, Frode Killingerg nd Arne Kjøsnes for technicl help during the strt feeding experiment nd Tor Brdl for good guidnce with the nlyticl work. I would like to thnk Jochim Lrsen Mrthinsen nd Jo vn Mil for the good coopertion nd vlule discussions during the experiment. Thnks to ll my gret clssmtes who mde the two yers in Mrine Costl Development joyful nd memorle experience. I pprecited ll the discussions, long study reks, lunches nd tps prties. Lst, I would like to thnk my fmily nd Jon-Anders for good support nd for lwys elieving in me. Trondheim, June 2018 Julie Terese Hnssen III

4 Astrct Astrct Slmon lice (Lepeoptherius solmonis) is mjor helth prolem in slmon (Slmo slr) production. Reduced lice sensitivity towrds severl of the chemicl tretments hve incresed the interest in clener fish used to iologicl remove slmon lice. The clener fish, lumpsucker (Cyclopterus lumpus) is reltively new quculture species nd there is little knowledge out the optiml rering conditions, nutrition nd the skeletl development. The im of this study ws to descrie the skeletl development in lumpsucker (C. lumpus) nd to evlute growth, survivl nd skeletl development in lrve fed with three different diets; Artemi, copepods (A. tons) nd formulted feed. One group ws fed enriched Artemi, second group fed copepods nd the third group ws fed comintion of A. tons (2-9 dph) nd formulted feed (from 7 dph). At 20 dph ll groups were wened to formulted feed nd fed it exclusively from dph. Strt feeding with Artemi improved lrvl growth nd survivl compred to lrve fed copepods nd formulted feed. The dry weight, wet weight nd stndrd length were significntly higher lredy from 6 dph nd throughout the experiment. Newly htched lrve hd one ossifiction in mouthprts, cleihtrum, vertere nd suction disc. Erlier onset of skeletl ossifiction relted to ge ws oserved in lrve fed Artemi. Significntly more vertere were fully ossified from dph nd more lrve hd ossified pterygiophores of first dorsl nd nl fins t 42 dph. In ddition, t 21 dph significntly more til fin rys hd ossified. However, the lrve fed copepods hd ossifiction in vertere, dorsl, nl nd til fins, urostyle nd hypurls t smller size. The smllest lrv with % fully ossified vertere ws 9.9 mm, over 0,8 mm shorter thn lrve fed Artemi nd formulted feed. The pterygiophores were ossified in lrve fed copepods lrger thn 8.3 mm, compred to 9,0 nd 8,8 mm for lrve fed Artemi nd formulted feed. The occurrence of severe skeletl nomlies (flt skull nd xis devitions) were 1 % t 50 dph. Fusion of two vertere were found in less thn 11 % of ll lrve t 50 dph. Twisted neurl rches, smll revition from the norml form ws oserved in significntly more lrve fed formulted feed. The occurrence ws highest in younger lrve nd considerly lower t 50 dph. IV

5 Smmendrg Smmendrg Lkselus (Lepeoptherius slmonis) er et stort helseprolem i produksjon v lks (Slmo slr). Redusert sensitivitet mot flere kjemiske ehndlinger hr økt interessen for ruk v rensefisk til å fjerne lkselus. Rensefiskrten rognkjeks (Cyclopterus lumpus) er en reltivt ny rt innenfor kvkultur og det er derfor mngel på kunnskp om optimle oppdrettsforhold, næringsstoffer og lrvens skjelettutvikling. Hensikten med dette studiet vr å eskrive skjelettutvikling hos rognkjeks (C. lumpus) og evluere vekst og overlevelse hos lrver fôret med tre ulike dietter; Artemi, hoppekreps (Acrti tons) og tørrfôr. En gruppe le fôret Artemi, den ndre gruppen le fôret hoppekreps og den tredje gruppen fikk en kominsjon v hoppekreps (2-9 dph) og tørrfôr fr 7 dph. Alle grupper le tilvendt til tørrfôr fr 20 dph og fôret det fr 22 dph. Målinger v vekst, overlevelse og einnlyser le rukt til å evluere effekten fr de ulike strtfôringsdiettene. Lrver frget med lizrin rød le rukt til å eskrive skjelettutviklingen og for å evluere forekomsten v skjelettvvik (nomlier). Strtfôring med Artemi resulterte i foredret vekst og overlevelse smmenlignet med lrver fôret med hoppekreps og tørrfôr. Tørrvekt, våtvekt og stndrd lengde vr signifiknt høyere llerede fr 6 dph og ut eksperimentet. Dette kn foreslå t DHA og EPA gitt i triglycerider i nriket Artemi sikrer tiltrekkelig mengde for norml vekst og overlevelse. De nylig klekte rognkjekslrvene hdde foreining i munndeler, gjelleue (cleihtrum), ryggvirvler og sugeskive. Skjelettutviklingen reltert til lder strtet tidligere i lrver fôret Artemi, mens lrvene fôret copepoder strtet foreiningen i ryggvirvler, rygg-, gtt- og hlefinner, urostyle og hypurler ved mindre størrelser. DHA tilført i fosfolipider i hoppekreps kn h ført til den tidligere einutviklingen. Forekomsten v lvorlige einvvik (nomlier) som fltt sklletk og kseforndringer (lordose, kyfose og skoliose) vr 1 % på 50 dph. Smmenvokste ryggvirvler le funnet i under 11 % v lle lrvene på 50 dph, men siden kun to ryggvirvler vr smmenvokst vil det trolig ikke påvirke lrvens lengde etrktelig. IV

6 Tle of contents Tle of contents Arevitions Introduction Use of clener fish in slmon quculture Optiml feed for lumpsucker lrve Skeletl development in mrine fish lrve Skeletl nomlies Aim of study Mterils nd methods Egg incution nd lrvl rering Feeding regime Live feed cultivtion Cultivtion of microlge Mintennce of copepods Cultivtion of Artemi Smpling Growth nd survivl Skeletl nlyses Ossifiction of vertere Ossifiction of fins Ossifiction of other prts Skeletl nomlies Sttisticl nlysis of dt Results Lrvl development Growth Dry- nd wet weight Dily weight increse Stndrd length Survivl Skeletl development Bone ossifiction in specific prts V

7 Tle of contents Suction disc Vertere Dorsl fins nd nl fin Til (cudl) fin nd til region Skeletl nomlies Anorml skull Anorml vertere nd xis devitions Twisted neurl rches Discussion The effects from different strt feeding diets on growth nd survivl Skeletl development nd the effects from different strt feeding diets Conclusion nd future perspectives References...56 Appendix Lrvl density Appendix Feeding mount formulted feed, Artemi nd A. tons Appendix Conwy Medium Appendix Dily weight increse Appendix Bone stining procedure Appendix Dry weight, wet weight nd stndrd length VI

8 Arevitions Arevitions Artemi group/tretment Copepod group/tretment Lrvl group fed Artemi 2-22 dph Lrvl group fed copepods (A. tons) 2-22 dph d Dys fter fertiliztion temperture (⁰C) DHA Dph DW DWI EPA Formulted feed group/tretment PUFA SE SGR SL WW Docoshexenoic cid (22:6n-3) Dys post htch Dry weight (mg) Dily weight increse Eicospentenoic cid (20:5n-3) Lrvl group fed copepod (A. tons) 2-9 dph nd formulted feed 7-50 dph. Polyunsturted ftty cid ( 20 cron toms) Stndrd error Specific growth rte Stndrd length (mm), mesured from the tip of the nose to the end of the vertere Wet weight (mg) 1

9 Introduction 1. Introduction 1.1. Use of clener fish in slmon quculture Norwy is the world s iggest producer of Atlntic slmon (Slmo slr) nd lrge quntity is exported (Liu et l., 2011, FAO, 2017). Tody the min helth prolem in slmon quculture is infections from slmon lice. Slmon lice (Lepeoptherius slmonis) is n ectoprsite living of skin, lood nd mucus of slmonids. The risk of secondry infections nd osmotic stress increses when the ectoprsite rek the slmonid immune protection ginst the environment (Grimnes nd Jkosen, 1996). Chemicl, physicl nd iologicl tretment re used to remove slmon lice. Reduced sensitivity nd potentil resistnce towrds severl of the chemicl tretments used re oserved the recent yers (Lees et l., 2008, Aen et l., 2015). This hve incresed the interest in other possile tretment methods like iologicl tretment with clener fish. Clener fish is group of fish species tht et ded or infected skin nd ectoprsites of the ody surfce of other fishes. This symiosis is utilized in slmon frming to remove the lice of slmon. Bjordl (1991) studied clening ctivity in species of wrsse nd ws the first to report clening ehviour in four species of wrsse when exposed to lice infected slmon. The use of clener fish incresed from fish in 2006 to over 37 millions in 2016 (Fiskeridirektortet, 2017). In 2016 the industry used 43 % lumpsucker (Cyclopterus lumpus), 36 % wrsse species nd 21 % unspecified species to remove slmon lice (Fiskeridirektortet, 2017). In tempertures under 6 ⁰C wrsses enters dormncy with low ctivity nd feeding, nd perform d s lice eters (Kelly et l., 2014). Lumpsucker tolerte colder sewter nd is ctively feeding during the winter in tempertures down to 3-4 ⁰C (Imslnd et l., 2014, Imslnd, 2018). The production of lumpsucker incresed from million fishes from (Fiskeridirektortet, 2017). No stndrdized production procedure exists tody nd the informtion out nutritionl requirements, optiml diet nd rering procedure re miniml. Producers often experience high vriility in survivl nd individul lrvl size (Imslnd et l., 2014, Dhle et l., 2017) nd high mortlity in se cges is reported (Bornø et l., 2016). The lrvl growth rte, development nd survivl cn e ffected y rering conditions nd nutrition like reported in rering of other mrine fish lrve (Boeuf nd Le Bil, 1999, Plns nd Cunh, 1999, Downing nd Litvk, 2000, Chu et l., 2003). To succeed with producing 2

10 Introduction roust nd effective lice eter, knowledge out the requirements for norml lrvl development is importnt Optiml feed for lumpsucker lrve There is little informtion out the optiml strt feed for lumpsucker lrve. Producers hve oserved lrvl cceptnce of formulted feed from the strt of feeding, ut prolems with settlement of uneten feed in fish tnks nd high vriility in growth nd survivl is reported (Imslnd et l., 2014). In nture lumpsucker lrve live close to seweed nd the min food source is crustcens ssocited with the seweed (Ingólfsson et l., 2002). Ingólfsson et l. (2002) oserved tht lumpsucker lrve fed selectively nd ignored smller nd slow-moving prey orgnisms like rotifers. The min prey orgnisms they found in the stomch of lumpsucker lrv ws species from two copepod orders; Hrpticoid nd Clnoid. The strt of exogenous feeding is criticl phse for mrine fish lrve where optiml feed is importnt to ensure norml lrvl development nd survivl (Yúfer nd Dris, 2007). In strt feeding of mrine fish lrve live nd formulted feed re commonly used. Live feed, the nturl prey for mrine fish lrve is esy digestile nd reduces the wste from uneten feed. Their movement keeps them longer suspended in the wter column which give lrve more time to ctch them. The nutritionl content of the nturl feed is optiml for the fish lrve. Live feed is however more expensive compred to formulted feed nd knowledge out the cultivtion nd mintennce of the live feed species re required. Brine shrimp (Artemi) nd rotifers (Brchionus sp.) re commonly used live feeds in rering of mrine fish lrve. Stndrdized production procedures mke them esy nd ffordle s lrvl feeds. Rotifers re used s first feed for smller mrine fish lrve which cnnot et Artemi or dry feed from the strt of exogenous feeding. The level of the essentil polyunsturted ftty cids (PUFAs); EPA nd DHA in Artemi nd rotifers re lower thn the requirements of mrine fish lrve (Evjemo nd Olsen, 1997, Léger et l., 1987). Enrichment with mrine fish oils with sufficient content of lipids nd proteins re importnt to meet the nutritionl requirements. In Artemi nd rotifers the lipids re stored s energy in triglycerides (Coutteu nd Mourente, 1997). Artemi ctolize DHA to energy, nd when using Artemi s live feed second enrichment procedure is importnt to ensure sufficient mount of DHA (Evjemo et l., 1997). Storge of Artemi t 10 C efore use s feed reduce the ctolism of DHA (Sorgeloos et l., 2001). 3

11 Introduction DHA nd EPA re required for norml fish growth nd development (Srgent et l., 1997). DHA is found in cell memrnes, rin nd nerve tissue of mrine fish lrve (Srgent, 1995). Mny lrve re not cple of synthesizing sufficient mount of essentil ftty cids nd re dependent on supply from feed sources (Geurden et l., 1995). Erlier studies reported positive effect on growth, survivl, stress tolernce nd one ossifiction when DHA nd EPA were supplied in polr lipids (Coutteu et l., 1997, Gisert et l., 2005, Kjørsvik et l., 2009, Wold et l., 2009). Authors ssumed tht ftty cids provided in polr lipids (phospholipids) were esier digested nd ssimilted compred to ftty cids dded in neutrl lipids. Polr lipids (phospholipids) re importnt structurl nd functionl components of cell memrnes, rin nd neurl tissue (Tocher nd Hrvie, 1988) Copepods re the nturl prey for mny mrine fish lrve including lumpsucker lrve (Ingólfsson et l., 2002). Their nutritionl content re considered optiml for mrine fish lrve with sufficient mount of oth DHA nd EPA incorported in phospholipids (Evjemo et l., 2003, vn der Meeren et l., 2008, Evjemo nd Olsen, 1997). Strt feeding experiments with lln wrsse, cod nd turot strt fed with copepods (A. tons) hd etter growth, survivl nd stress tolernce compred to lrve fed Artemi nd rotifers (Kjørsvik et l., 2014, Øie et l., 2017, Hmre, 2006, Witt et l., 1984). Artemi nd rotifers hve lower content of phospholipids compred to copepods (Evjemo nd Olsen, 1997, Næss et l., 1995). Disdvntges with the use of copepods re tht they re difficult nd expensive to cultivte compred to Artemi nd rotifers. Formulted feed must fulfil severl structurl nd iochemicl requirements to stisfy the development of mrine fish lrve (Hmre et l., 2013). The disdvntges with formulted feed re the wste settling on the ottom of the tnks from uneten feed. This increse the orgnic lod nd provides more growth sustrte for cteri in the fish tnks (Dhle et l., 2017). Frequently clening is importnt to reduce the mount of orgnic mtter settling in fish tnks. A pilot study performed y NTNU, SINTEF nd lumpsucker producer (Nordlnd rensefisk) evluted if use of copepods (A. tons) s strt feed (2-14 dph) resulted in etter growth nd survivl thn formulted feed (Dhle et l., 2017). The lrve fed copepods otined higher growth nd survivl. The vrition in survivl etween lrve fed copepods were lower compred to those fed formulted feed. In ddition, the cteri composition in lrvl tnks fed 4

12 Introduction copepods ws less vrile. This my suggest tht strt feeding with formulted feed is less optiml compred to copepods Skeletl development in mrine fish lrve The fish skeleton supports ody posture, development nd locomotion, functions s ttchment point for muscles, protects orgns nd cells nd is reservoir for ions (Lll nd Lewis-McCre, 2007). Teleost ones re composed of osteolsts, osteocytes nd one lining cells, minerl phse nd n orgnic component of extrcellulr mtrix (Boglione et l., 2013). Osteolsts is the one forming cells. Osteocytes rek down one tissue, n importnt feture in one mintennce, repir nd remodelling. The one lining cells cover the one surfces nd cn differentite into osteogenic cells developing into osteoclsts (Prfitt, 2001). The minerl phse consists minly of clcium hydroxyptite slts emedded in the orgnic mtrix of collgen fiers type 1 nd 2 (Benjmin nd Rlphs, 1991). In generl, ones in dvnced teleost species lck the osteoclsts tht remodel ones nd perform clcium metolism. Insted clcium is tken up through the gills (Witten, 1997) nd the remodelling functions re performed y other cells. Most of the one remodelling is performed relted to growth (Witten nd Huysseune, 2009). Teleost one tissue continues to grow which mke the growth-relted mintennce possile throughout the fish life. However, if fish hve osteoclsts the differentition tkes plce in the hed kidney in fish (Witten nd Huysseune, 2009). These osteoclsts re mono nucler nd smller compred to multinucleted osteoclsts found in mmmls (Boglione et l., 2013). Bone formtion in fish is descried y three mechnisms; endochondrl, perichondrl nd intrmemrnous ossifiction (Boglione et l., 2013). Endochondrl ossifiction consists of severl steps replcing crtilginous templte with ones. Mny teleost species lck the endochondrl ossifiction mechnisms, nd the crtilge tissue usully found inside ones is replced y dipose tissue. Perichondrl one ossifiction re minly the ossifiction process found in fish. The perichondrium of crtilgous templte is trnsformed into periosteum which secrete one mtrix nd hve osteolst-similr functions (Boglione et l., 2013). During intrmemrnous ossifiction ones re formed y osteolsts differentited directly from mesenchyml cells without crtilgous templte (Frnz-Odendl, 2011). The mesenchyml cells differentite into osteolsts, lood vessels grow into the re with osteolsts nd one develops. The fish ones formed through intrmemrnous ossifiction 5

13 Introduction re derml or memrne ones teeth, denticles, crnil derml ones, scles nd fin rys (Boglione et l., 2013). The fish skeletl tissue supporting the ody posture nd providing ttchment point for muscles is the notochord. It origintes from firous collgen sheth nd not from crtilge common in other vertertes. Secretion of firous collgen nd formtion of epithelil cells (notochordolsts) forms firous sheth which surround nd stiffen the notochord (Boglione et l., 2013). The verterl odies re estlished y minerliztion of the firous sheth (Witten nd Huysseune, 2009). The fin ones re composted of derml- nd n endoskeleton. The derml skeleton consists of fin rys nd the endoskeleton connects the fins to the rest of the fish skeleton (Witten nd Huysseune, 2007). In unpired fins, like dorsl, nl nd til fin the endoskeleton is connected directly to the xil skeleton, while pired fins is connected to girdles. The endoskeleton of nl nd dorsl fins consists of pterygiophores, nd in the cudl fin the urostyle is the endoskeleton supporting the fin. The pired pectorl fins re connected to shoulder girdle composed of oth endoskeletl nd derml elements. The lumpsucker skeleton is not well descried in literture. Voskooinikov nd Kudryvtsev (2014) descrie the one development in wild smpled lumpsucker eggs nd lrve. The shortest lrve with stndrd length (SL) 4.8 mm otined from eggs hd no one ossifiction. Ossifiction were first oserved in free living lrve longer thn 5.5 mm. At this size the mxill, premxill, dentl, certohyl, rnchiostegl rys, opercle, preopercle, cleihtrum, pelvis nd pelvic fin rys were ossified. Pelvic fins were modified into suction disc mking lrve cple of dhering to hrd sustrtes. In lrve lrger thn 6.5 mm teeth nd the fin rys of second dorsl, nl, nd cudl fin were ossified. The rows of teeth incresed from one to four rows during lrvl development. Ossifiction of nine vertere were oserved in 5.5 mm lrge lrve nd from 6.5 mm ll of them hd ossifiction, except from the two lst closest to the urostyle. The totl numer of vertere reported were 27 or 28. Urostyle nd hypurl ones supporting the cudl fin ossified in lrve longer thn 8.1 mm. The epurls lso providing cudl fin support ossified in lrve from 13 mm. 6

14 Introduction 1.4. Skeletl nomlies in mrine fish lrve Skeletl nomlies re common in rering of mrine fish lrve (Divnch et l., 1996). They re often found in fins, the spinl column nd the hed. Asence of the entire fin, fin rys or pterygiophores, decresed size, dupliction of fins nd other deformities re found in different fish species (Kocour et l., 2006, Georgkopoulou et l., 2010). Spinl nomlies include scoliosis (S-shped lterl side), lordosis (downwrd curvture) nd kyphosis (upwrd curvture) nd often comintion of these (Divnch et l., 1996). Axis devition ffects the fish ody shpe nd the growth re usully lower compred to fish without spinl nomlies. Other nomlies tht occur in the spinl column include fusion, disloction, compression nd shortening of the vertere (Boglione et l., 2001). Divnch et l. (1996) defined six different nomlies occurring in the hed region; Gill cover nomlies, pughedness, crossite, lower jw reduction, ventrlly projected hyornchil skeleton nd pike jw deformity. Environmentl, nutritionl nd genetic fctors cn interfere with norml skeletl development in fish. Wter temperture, ph nd wter currents (Hemmer et l., 1990, Divnch et l., 1997, Chtin, 1994, Adel et l., 2004, Georgkopoulou et l., 2010) nd the nutritionl content (Chu et l., 2003, Lll nd Lewis-McCre, 2007) re fctors oserved to cuse nomlies. Boglione et l. (2013) suggest tht severl fctors my cuse the sme skeletl nomly nd one fctor my cuse severl different nomlies. This mkes it difficult to identify the min reson ehind specific skeletl nomly. However, high proportion of skeletl nomlies during lrvl rering indicte tht something is less optiml. Villeneuve et l. (2005) relte the lipid content of feed to fish skeletl development in se ss. DHA nd EPA provided in phospholipids re reported to give etter skeletl development nd less nomlies in cod (Kjørsvik et l., 2009), lln wrsse (Øie et l., 2017), sess (Chu et l., 2003). Jump nd Clrke (1999, s sited in Chu et l., 2003) stte tht the highly unsturted ftty cids ffects specific nucler receptors involved in regultion of genes during skeletl development under ontogenesis in mmmls. DHA nd EPA might hve the sme impct during development of the fish skeleton. Protein with the optiml mino cid profile is importnt to support the fst one development in fish lrve (Chu et l., 2003). Deficiency of the mino cid tryptophn re reported to increse the occurrence of scoliosis in slmonids (Akiym et l., 1986). Moderte levels of protein hydrolystes in the diet ws reported to decrese the skeletl nomlies in gilthed serem, while high levels (over 12 % of dry mtter) resulted in more skeletl nomlies 7

15 Introduction (Gisert et l., 2012). Protein hydrolystes consists of peptide chins with mino cids or di- nd tri-peptides insted of whole proteins. (Chu nd Infnte, 1995) stte tht diet with optiml level of protein hydrolystes my ccelerte the mturtion process of the intestine. Vitmins nd minerls my lso ffect the skeletl development in fish. Lll nd Lewis-McCre (2007) compred severl studies of skeletl nomlies in mrine fish lrve nd found tht suoptiml levels of phosphorus, vitmin A nd K cn increse the occurrence of skeletl nomlies. Deficiency of phosphorous is reported to interfere with the one minerliztion in rinow trout (Fontgné et l., 2009), hddock (Roy, 2002) nd Atlntic hliut (Roy nd Lll, 2003). In hddock, uthors suggested tht phosphorus deficiency incresed the numer of osteoclsts performing one resorption. Phosphorus deficiency ws reported to increse numer of nomlies in the heml nd neurl spines in Atlntic hliut. Elevted levels of vitmin A (>188 ng mg -1 DW of the enriched live prey) in the enrichment diets of live feeds (Artemi nd rotifers) ws reported to result in more skeletl deformities in se rem (Fernández et l., 2008). Fernández et l. (2008) suggested tht vitmin A ccelerte the skeletl development nd high levels increse the numer of skeletl deformities. Roy (2002) oserved deformities in hddock fed diets with deficiency of vitmin K. Other vitmins nd minerls re lso suggested to e involved in verterl skeletl development. Vitmin E my e involved in the defence ginst free rdicls tht might dmge osteolsts nd stimulte osteoclsts (Lll nd Lewis-McCre, 2007). Dris et l. (2010) reported tht deficiency or elevted levels of vitmin D incresed the occurrence of skeletl deformities in se ss. They suggested reltion etween vitmin D levels nd the sorption of C 2+. Less C 2+ ws sored in se ss fed higher levels of vitmin D. 8

16 Introduction 1.5. Aim of study The im of this study ws to descrie the skeletl development in lumpsucker (Cyclopterus lumpus) nd to evlute growth, survivl, skeletl development nd nomlies in lrve fed with three different diets; Artemi, A. tons or formulted feed. These diets represent common strt feeding regimes of mrine fish lrve. Lumpsucker lrve were fed the three different diets the first 22 dys post htch (dph). Two groups were fed copepods (A. tons) nd enriched Artemi exclusively nd the third group ws fed comintion of A. tons (2-9 dph) nd formulted feed (7-22 dph). From 20 dph ll groups were wened to formulted feed nd fed formulted feed exclusively from 22 dph. The effects of the diets were mesured in growth (stndrd length, dry weight nd wet weight), survivl, skeletl development (onset of ossifiction in specific prts) nd skeletl nomlies. Bone nlysis were performed on lizrin stined lrve. The hypothesis ws tht lumpsucker lrve fed Artemi nd A. tons would hve etter growth, survivl, one development nd less nomlies compred to lrve fed formulted feed. The strt feeding experiment ws in collortion with two other mster students; Jochim Lrsen Mrthinsen studied the digestive system nd Jo vn Mil studied lrvl growth. 9

17 Mterils nd methods 2. Mterils nd methods The strt feeding experiment, growth- nd one nlysis of lumpsucker (Cyclopterus lumpus) ws performed t Norwegin University of Science nd Technology (NTNU) t Centre of Fisheries nd Aquculture (Sel). The experiment strted on My 2017 nd ws terminted in the end of June Growth nlysis were conducted prllel with the experiment nd one nlysis were performed in the utumn Egg incution nd lrvl rering Wild cught lumpsucker eggs were supplied from Morefish t Tjeldsergodden. The eggs (fertilized ) were incuted in 6,8 C for 198 d t Morefish. Before trnsported to Sel, they were disinfected with Buffodine. Eggs were incuted in drkness in two fmily incutors (Fishtech AS) with continuous wter flow, moderte ertion nd grdully incresing wter temperture from 8-10 ⁰C until htching t 307 d⁰. After htching (307 d⁰, 0 dph) the lrve were trnsferred with wter current to tues connecting the incutors to holding tnk. Lrve ws relesed into plnkton net plced in the holding tnk. The plnkton net prevented lrve from ttching to the tnk wlls. The 60 L white plstic holding tnk ws filled with 10 C sewter erted from the ottom. At 2 dph nine 1 L- mesuring jugs ws used to distriute lrve into nine 160 L lck cone-ottomed fish tnks. Lrvl density in ech fish tnk clculted when ending the experiment, rnged from lrve/l (tle A1.1, ppendix 1). Fish tnks were filled with 100 L treted sewter (34 ppt) of 10 C. The sewter tretment consisted of snd filtrtion, 1 micron filtrtion (mesh size 1 µm), heting, microil mturtion (Skjermo et l., 1997), degssing nd second 1 micron filtrtion. A ottom sieve with net size 700 µm ws covering the wter outlet in ll tnks. The wter exchnge rte ws incresed from 5-25 times dy -1 when necessry depending on the feeding mount (tle 2.1). Oxygen levels were mesured every third/forth dy using n opticl dissolved oxygen meter (ProODO, YSI, USA), nd kept over 80 % y djusting ertion dded to the tnk. The wter temperture mintined t 10 ⁰C ws mesured dily with Digitron digitl thermometer (2000T Type K, UK). Light regime consisted of 20 hours light using dylight fluorescent tues (Philips MASTER TL-D 90 Grphic, 18W/965) nd 4 hours drkness. 10

18 Mterils nd methods Three 5x5 cm drk grey silicone mts (mcron king sheets, Cls Ohlson, Norwy) were plced in ech fish tnk to crete lrger surfce for lrve to ttch to (Imslnd et l., 2015). Skimmers were instlled t the wter surfce in tnks fed formulted feed. The lrvl tnks fed live feed were clened dily to remove feces, uneten food nd ded lrve. Since feeding with formulted feed incresed the sedimenttion of uneten food, those tnks were clened twice dily. A ottom sieve covering the wter outlet ws clened every second dy Feeding regime The strt feeding experiment consisted of three feeding regimes with ech three replictes (fish tnks), presented in tle 2.1. One group ws fed Artemi from 2-22 dph; first newly htched nd not enriched Artemi from 2-4 dph, then 50/50 comintion of newly htched unenriched nd enriched Artemi from 5-6 dph nd from 7-22 dph enriched Artemi. The two other groups were fed copepod, A. tons from 2-22 dph nd 2-8 dph. At 6-8 dph the formulted feed group ws wened to nd fed formulted feed (Gemm micro 150) until 22 dph. All groups were wened to formulted feed (Gemm micro 300) from dph nd fed exclusively for the rest of the experiment (22-50 dph). Artemi nd A. tons were fed to lrvl tnks from 100 L feed reservoirs using peristtic pumps. Feeding ws scheduled 6 times/dy during the light period (t 14.00, 18.00, 22.00, 03.00, nd o clock). The feeding mount ws incresed grdully from live prey dy -1 tnk -1 (Tle A2.1 in Appendix 2). The desired density of live feed in lrvl tnks per feeding ws live prey L -1 feeding -1. The formulted feed Gemm micro 150 nd 300 distriuted y Skretting AS ws fed with feeding utomts (Sterner disk feeder 905) plced over funnels with running wter flow. They were progrmmed to feed desired feeding mount using the progrm Normtic We Server, from Normtic AS. The feeding mount (Tle A2.1 in Appendix 2) ws incresed in reltion to estimted lrvl density nd growth following recommendtions from Skretting. 11

19 Feeding regime Form. feed Copepod Artemi Mterils nd methods Tle 2.1. Specifiction of the strt feeding experiment with dtes, dys post htch (dph), dy degrees (d ), wter exchnge rte (times tnk volume/dy), feeding regimes nd smpling dys for dry weight (DW), wet weight (WW), stndrd length (SL), mortlity, one nlysis nd ftty cid nlysis. Dte My 2017 June Dph d Wter exchnge 5 times/dy 10 times/dy 15 times/dy 25 times/dy Artemi (newly htched) Artemi (newly + enriched) Artemi (enriched with Multigin) Formulted feed (Gemm micro 300) Copepods (A. tons) Formulted feed (Gemm micro 300) Copepods (A. tons) Formulted feed (Gemm micro 150) Smpling DW WW SL Formulted feed (Gemm micro 300) x x x x x x x x x x x x x x x Bone nlysis x x x x x x x x x x x x x x 12

20 Mterils nd methods 2.3. Live feed cultivtion Artemi nd microlge were cultivted during the strt feeding experiment. Microlge ws used s feed for copepod (A. tons), which ws produced nd delivered from C-feed nd mintined until fed to the lumpsucker lrve Cultivtion of microlge Rhodomons ltic (Clone NIVA 5/9 Cryptophyce: Pyrenomondles) ws cultivted semicontinuously in two 200 L trnsprent polycronte plstic cylinders filled with sewter. In dvnce the sewter (34 ppt) ws snd filtered, heted to C, filtered through 1 µm mesh, chlorinted nd dechlorinted. The chlorintion ws performed y use of 0,25 ml chlorte (NOCl) per litre sewter for minimum 5 hours without ertion. Afterwrds the sewter ws dechlorinted with 0,03 g sodium thiosulfte penthydrte (N2S2O3) per litre sewter under hevy ertion for minimum 5 hours. The microlge were cultivted with constnt illumintion y 6 fluorescent tues (Aur t8 ultimte W) on three sides, nd ertion with 1-2 % CO2 dded from the ottom. Dily, % of the microlge culture ws hrvested nd fed to the copepods, see descriptions in mintennce of copepods. The remining culture ws diluted with treted sewter nd 1 ml/l sewter of Conwy medium ws dded (ppendix 3 modified from Wlne (1979)) Mintennce of copepods A. tons (clone DFH.AT1) of stges CI-CIII ws delivered from C-feed AS in plstic continers three times during the experiment. The CI-CIII stges re µm (Hgemnn, 2014). The density in ech continer ws clculted mnully nd the desired prey density for one dy feeding ws distriuted to holding tnks filled with 100 L of treted sewter (34 ppt nd 10 C). The sewter tretment ws the sme s for the wter in lrvl fish tnks. At low tempertures copepods hve reduced grzing nd development slows down (Mren Rnheim Ggnt, C-feed, pers. comm, 10. My 2017). To mintin the low grzing A. tons ws fed 6 L microlge (R. ltic) per 100 L holding tnk. The density of fed microlge ws not mesured. 50 % of the wter ws flushed out nd renewed with fresh sewter dily. The content of one holding tnk ws fed to lrvl tnks dily. The ottom outlet ws covered with 100 µm mesh sized filter. 13

21 Mterils nd methods Cultivtion of Artemi Artemi cysts (EG INVE Aquculture, Belgium) were hydrted in fresh wter ( g cysts, 4,9 L wter, ⁰C) with hevy ertion nd decpsulted ccording to (Sorgeloos et l., 1977). The decpsulted cysts were stored t 4 ⁰C for mximum 7 dys nd htched in 60 L cylindricl tnks filled with sewter (34 ppt). Aertion ws dded t the ottom nd wter temperture ws kept t ⁰C. After 24 hours, unhtched cysts were removed y turning of ertion for 5 minutes nd flushing out sediments through the ottom vlve for ½ second. The newly htched Artemi ws concentrted in sieve nd trnsferred to new 60 L tnk filled with sewter with the sme temperture nd hevy ertion. Enrichment with 10 g MultiGin medium (produced y Biomr) per 60 L sewter ws dded two times over 24 hours. After enrichment, Artemi ws wshed gently for minutes in n Artemi-wsher, concentrted in sieve nd trnsferred to feeding tnk filled with 55 L sewter (32 ppt nd 10 C). The holding tnks hd stgnnt wter flow nd hevy ertion dded t the ottom. Artemi were fed to lrve over period of 24 hours. The size of Artemi nuplii instr 2 (6-8 h) is µm (Léger et l., 1987) Smpling Smpling for dry weight (DW), wet weight (WW), stndrd length (SL) nd one nlysis ws sixteen times t regulr intervls during the experiment (tle 2.1). Smpling size ws incresed from lrv/tretment from 0-51 dph presented in tle 2.2. All lrve were smpled rndomly from lrvl tnks using plstic cups or with smll hnd net, nesthetized with tricine methne sulfonte (MS-222 Finquel, Agent Chemicl Lortories Inc., USA) nd rinsed in distilled wter efore further nlysis. Tle 2.2. Smpling size (lrv/tnk) for growth, one nd ftty cid nlysis. At 0 nd 2 dph smpling size re the totl numer tken efore lrve were distriuted to lrvl tnks. Smple size (lrv/tnk) Dys post htch Growth Bone nlysis 0, , 10, 13, 17, 21, 24, 29 34, ,

22 Mterils nd methods 2.5. Growth nd survivl Lrvl stndrd length (SL) ws mesured from the tip of the mouth to the end of the notochord (Figure 2.1). Lrve were photogrphed using stereomicroscope (Leic MZ 7.5, Germny) equipped with cmer (Axiocm Erc 5s, Crl Zeiss Microscopy GmH, Germny) nd softwre Zen 2.3 for windows. Stndrd length ws mesured from pictures using the Windows softwre ImrgeJ (version 1.8.0_112). After photogrphing, lrve were dried with wipes to remove excess wter, plced in preweighed tin cpsules nd wet weight (WW) ws mesured with n ultr-microlnce weight (Mettler Toledo UMX2). For mesurement of dry weight (DW) the tin cpsules with lrve were dried in 60 ⁰C for minimum 48 hours, efore weighed. The specific growth rte (SGR) nd percentge of dily weight increse (% DWI) ws clculted ccording to Ricker (1958) presented in eqution 2.1 nd 2.2, where DW1 nd DW2 re dry weight t time t1 nd t2. SGR = ln DW2 ln DW1 t 2 t 1 [2.1] % DWI = (e (SGR) 1) 100 [2.2] Lrvl survivl ws clculted sed on dily registrtion of ded lrve from 3-50 dph nd counting of remining lrve when ending the experiment. Numers were corrected for smpled lrve for growth nd one nlysis. Figure 2.1. Stndrd length mesurement on lumpsucker lrve photogrphed with stereo Lupe equipped with cmer. 15

23 Mterils nd methods 2.6. Skeletl nlyses Skeletl nlyses were not performed on ll lrv smpled. Ten lrve/tnk t five smpling times (0, 10, 21, 34 nd 42 dph) nd 50 lrve/tnk t the end of the experiment (50 dph) were nlysed. Anesthetized nd rinsed lrve were fixted in 4 % prformldehyde (PFA) in phosphte uffered sline (ph 7.4, Apotekproduksjon AS; Norwy), nd stored t 4 C. Lrve were stined with lizrin red ccording to Kjørsvik et l. (2009). The one stining procedure (ppendix 5) contined five steps; (1) rehydrtion, (2) leching, (3) clering, (4) stining nd (5) preservtion. Rehydrtion in solutions of decresing mount of ethnol (95 %, 50 % nd 15 %) to distilled wter for hour in ech solution depending on lrvl size. A trnsprent plstic well try of different sizes (2x6 or 4x4 wells) depending on the numer of nd size of lrve, were used to seprte lrve smpled from different tnks when stining. After rehydrtion the lrve were stored in sodium orte uffer until the next dy if necessry. Then leched in 3% hydrogen peroxide in 1% potssium hydroxide (rtio 1:9) under strong light. The leching process ws monitored nd stopped when ll pigmenttion ws removed. Lrve under 10 mm were leched fter 1-2 hours, while lrve from mm used up to 6 hours. Clering with trypsin uffer to lmost trnsprent took 20 hours for the smllest lrve (>10 mm). The time ws prolonged to 2-4 dys nd uffer ws renewed dily for lrger lrve (10-20 mm). Clering ws difficult on the fttest lrve smpled t 50 dph nd stronger solution of the trypsin uffer ws used to shorten the time. The trypsin uffer did not remove the stomch content. Strving efore the lst smpling t 50 dph would proly reduced the stomch content. After clering the ones were stined with Alizrin for 1-2 dys depending on lrvl size. The stined lrve were preserved in 40 % glycerol in 1 % KOH for 2 dys nd photogrphed under stereomicroscope (Leic MZ75, Germny) equipped with cmer (Nikon Digitl Sight DS-5M L1, Jpn). Anlysis of lrve were performed using the stereomicroscope nd on pictures. ImgeJ (version 1.8.0_112) softwre for windows ws used to mesure stndrd length (Figure 2.2) nd counting numers of ossified vertere segments nd fin rys. 16

24 Mterils nd methods Photogrphed nd nlysed lrve were preserved in 70 % glycerol in 1 % KOH for 1-2 dys nd then 100 % glycerol in 4 C. Figure 2.2. Left: Stndrd length mesurement of one stined lumpsucker lrve. Right: Bone nomenclture in til region. VS = verter segment, HA = heml rch, NA = neurl rch, U = urostyle, HY = hypurls nd FR = fin rys (of til fin) Ossifiction of vertere Vertere segments were clssified s prtly ossified (P), compct (C), trnsprent (T) nd fully ossified (FO), depending on the intensity nd fill of colour in ech segment (Figure 2.3). Trnsprent segments hd no visile one ossifiction, prtly ossified segments hd sign of ossifiction ut were not filling the whole segment. In compct segments did the colour fill the whole segment ut without full intensity. Fully ossified vertere hd full intensity of colour, filling the whole segment. Neurl nd heml rches, urostyle nd hypurls were clssified s ossified if ny sign of ossifiction ws oserved (Figure 2.2). Figure 2.3. Clssifiction of vertere segments in lumpsucker lrve sed on sturtion of colour. FO = fully ossified, C = compct, P = prtly ossified nd T = trnsprent Ossifiction of fins Dorsl fins nd nl fin were scored 0-3 depending on the ossifiction (Figure 2.4). Score 0 equlled no ossifiction, 1 = prtly ossifiction of dorsl/nl fin rys, 2 = ossifiction of 17

25 Mterils nd methods dorsl/nl fin rys, ut not of pterygiophores, 3 = ossifiction of ll dorsl/nl fin rys nd pterygiophores. The til fin ws evluted y counting numer of ossified fin rys. Figure 2.4. Scoring of dorsl nd nl fin rys (0-3) sed on ossifiction. Scores for ll fins re indicted on left side of pictures. Score 0 = no ossifiction, 1 = prtly ossifiction of dorsl/nl fin rys, 2 = ossifiction of dorsl/nl fin rys, ut not of pterygiophores, 3 = ossifiction of ll dorsl/nl fin rys nd pterygiophores Ossifiction of other prts Suction disc, mouthprts (premxillre, mxillre, dentle nd the hyorid rch), cleithrum, urostyle nd hypurls where evluted s ossified when the first sign of ossifiction ws oserved. Ossifiction of urostyle nd hypurls relted to size ws visulized in tle, derived from (Sæle et l., 2004, Sørøy, 2012). Light grey re mrked SL of the smllest lrv when the first sign of ossifiction occurred in specific one to SL of the lrgest lrv without ossifiction. Drk grey re mrked which SL ll lrv hd ossifiction in tht one Skeletl nomlies An nomly is defined s devition from the norml shpe of ody prt or n orgn (Heierg, 2009). Anlysis of skeletl nomlies were performed on the pictures of one stined lrve. 18

26 Mterils nd methods Hed, mouthprts, vertere segments, heml nd neurl rches, urostyle, hypurls, dorsl fins, nl fin nd til fin were nlysed for skeletl nomlies. Numer of lrve with nomlies nd numer of nomlies per lrve were counted, nd the frequency of lrve with nomlies ws clculted. These definitions derived from Boglione et l. (2001) were used to identify nomlies oserved: Axis devition: o Lordosis: Inwrd curvture of the spine o Kyphosis: Outwrd curvture of the spine o Scoliosis: Sidewys S or C shped curvture of the spine Fused vertere: two or more verterl segments re fused nd might shre the sme neurl nd/or heml rch. Anorml neurl or heml rches: twisted rches (rches/spines form spirl t pex) Anorml hypurls Anorml fins: dorsl, nl or til fin rys or pterygiophores Flt skull: skull flt on dorsl side Anorml jw: shorter upper jw compred to the lower jw 2.7. Sttisticl nlysis of dt Dt were first tested for normlity using Shpiro-Wilk-test nd homogeneity of vrince with Levene s test. Norml distriuted dt were tested for significnt differences etween groups using one-wy ANOVA. The Welch test ws used to determine significnt differences, followed y post hoc tests; Student-Newmn-Keuls used for homogenous dt nd Dunnett T3 for non-homogenous dt. Not normlly distriuted dt were tested with non-prmetric test, nd to compre groups n independent smples Kruskl-Wllis test (2-sided, symptotic significnces) ws used. This test included pirwise comprison nd djustment of significnce vlues. Percentge dt were rcsine trnsformed efore sttisticl nlysis. A Persson liner correltion test were used to check for correltion etween dt. All sttisticl nlysis ws performed with IBM SPSS Sttistics v19.0 (SPSS Inc., Chicgo, USA) for Windows nd significnce level of p = 0.05 ws used for ll tests. Tles nd grphs were mde with Microsoft Office Excel

27 Results 3. Results 3.1. Lrvl development Lumpsucker lrve htched with ig yolk sc, open mouth nd pigmented eyes (Figure 3.1). Yellow-rown pigmenttion with drk dots were oserved in the hed nd the trunk, ut with vritions etween individul lrve. The til region ws trnsprent with fields of light eige pigmenttion. No fins were developed t 0 dph. Alredy from htching lrve hd suction disc on the underside of the ody which they used to ttch to surfces in the fish tnk. The ility to ttch to surfces ws even oserved in sedted, fixted nd ded lrve. From 6 dph fin rys of two dorsl fins, til fin, nl fin nd pectorl fins were developed. The first dorsl fin ws hlf covered with pigmented skin, while the covering of the second dorsl fin, nl fin nd til fin were trnsprent skin. The yolk sc ws no longer visile under the incresed pigmenttion in the hed. Approximtely ¼ of the trunk ws pigmented. At 10 dph the whole first dorsl fin ws covered with pigmented skin. Strt of flexion ws oserved in ll lrve from 10 dph. Light rown nd spotted pigmenttion ws visile from the til region to strt of the til fin. The hed ws drker pigmentted with light nd shiny horizontl stripe from ehind the eyes nd over the nose. A lrger suction disc with the sme pigmenttion s the hed ws developed. The mouth ws lso pigmented t 10 dph. At 21 dph the whole lrvl ody ws pigmented, except from fin rys of the second dorsl, til nd nl fin. All the fin rys hd grown in length. Lrger rown spots were oserved in the lrvl trunk with red colour in the lower hlf. The vertere were visile in the lrvl trunk up to 21 dph, ut t 29 dph it ws covered y drker pigmenttion. At 29 dph the fin rys were longer nd the horizontl stripe ws lighter nd shinier. The lrvl hed nd trunk hd grown in width. From 42 dph the lrve hd green-rown hed nd trunk, while the til prt hd developed some lighter spots. At the end of the experiment (50 dph) lrve hd wider trunk nd lighter pigmenttion with some drker spots. Their ody shpe egn to resemle n dult fish with wider ody shpe nd suction disc covering ¼ of the underside of the ody. 20

28 Results 0 dph 6 dph 0 dph 6 dph 10 dph 21 dph 10 dph 21 dph 29 dph 34 dph 29 dph 42 dph 50 dph Figure 3.1. Development from 0-50 dph of lumpsucker lrve fed Artemi. Lrvl ge nd stndrd length; 0 dph 6.0 mm, 6 dph 6.9 mm, 10 dph 7.2 mm, 21 dph 8.4 mm, 29 dph 9.8 mm, 34 dph 11.1 mm, 42 dph 12.2 mm nd 50 dph 13.6 mm. 21

29 Results The vrition in lrvl pigmenttion oserved in the copepod tretment t dph is illustrted in Figure 3.2. Lrve from the other groups hd less vritions in colour. The mjority of Artemi nd formulted feed fed lrve were drk grey-green. Lrv 1 in Figure 3.2 ws yellow with rown spots in the hed nd strt of the trunk nd smller lck spots scttered round the ody. Lrv 2 ws light grey with tint of lue nd scttered rown spots of different sizes. The third lrv ws red with smll drk spots. Lrv 4 hd drk pigmenttion with lighter rown spots nd pink re under the eyes. All lrve hd the light, shimmering stripe from ehind the eyes nd over the nose. The vrition etween individul lrv decresed nd t 50 dph less differences were oserved Figure 3.2. Exmples of different pigmenttion oserved in lrve from the copepod group, t dph. Age nd stndrd length (mm) of lrve: 1: 38 dph nd 9.2 mm, 2: 38 dph nd 9.8 mm, 3: 38 dph nd 9.8 mm, nd 4: 42 dph nd 9.8 mm. The lck r is 500 µm 22

30 Results 3.2. Growth Dry- nd wet weight Men lrvl dry nd wet weight were significntly higher for lrve fed Artemi from 6-51 dph (Figure 3.3). The difference etween lrve fed Artemi nd the other groups fed copepod nd formulted feed incresed t 2-22 dph. When one of the groups fed copepods were wened to formulted feed (7-9 dph), the growth decresed nd dry- nd wet weight were significntly lower thn the other groups. The lrvl group still fed copepods incresed their weight until wening to formulted feed strted t 20 dph. Wening to formulted feed resulted in slowdown of in weight increse in the two groups fed live feeds. Lrve wened from Artemi hd lower weight increse from dph compred to erlier. The period with slow growth ws lsted to 29 dph for the lrve fed copepods. The rest of the formulted feed period hd ll groups stedy increse in dry nd wet weight. At the end of the experiment (51 dph) the lrve fed Artemi (22.7 mg) were significntly higher nd twice the dry weight of lrve fed copepods nd formulted feed (11.2 nd 11.9 mg). The wet weights t 50 dph were 108, 63 nd 64 mg for lrvl groups fed Artemi, copepods nd formulted feed. The Person correltion test showed positive liner correltion etween wet weight nd dry weight in ll groups (Figure 3.4), where the dry weight increse ws times the wet weight. 23

31 Men wet weight (log mg) Men dry weight (log mg) Results 1,4 Live feed period Formulted feed period 1,2 1,0 0,8 0,6 0,4 0,2 0,0 c c Artemi Copepod Formulted feed Dys post htch 2,2 2,0 1,8 1,6 1,4 1,2 1,0 0,8 0,6 0,4 Live feed period Formulted feed period c c Artemi Copepod Formulted feed Dys post htch Figure 3.3. Men lrvl dry weight (log mg tretment -1 ) from 0-51 dph nd men lrvl wet weight (log mg tretment -1 ) from 0-50 dph. Smpling size; n = 16 tretment -1 t 0 nd 2 dph, n = 15 tretment - 1 t 6, 10, 13, 17, 21, 24, 29, 34 nd 38 dph, n = 30 tretment -1 t 42, 45 nd 50 dph, nd n = 48 tretment -1 t 51 dph (only DW). First dotted line t 9 dph mrks the trnsfer from feeding copepod to formulted feed for the formulted feed group. Second dotted line t 22 dph, mrks the end of live feed period for Artemi sp. nd copepod tretments. All tretments hd formulted feed diets from 22 dph. Stndrd error (±) re represented y error rs. Letters indicte significnt differences etween tretments. 24

32 Dry weight (mg lrv -1 ) Results Artemi Copepod Formulted feed Wet weight (mg lrv -1 ) Figure 3.4. Correltion of lrvl dry weight (mg lrv -1 ) nd wet weight (mg lrv -1 ). Blck dotted line is the DW nd WW correltion for Artemi fed lrve, lue for copepod fed lrve nd green for formulted feed fed lrve. Person point-iseril liner correltion coefficients, R p 2 = 0,995 (Artemi), 0,995 (Copepod) nd 0,996 (formulted feed) nd the correltion ws significnt for ll groups. 25

33 Men dily weight increse (%) Results Dily weight increse The dily weight increse (% DWI) sed on dry weight ws negtive efore exogenous feeding strted t 2 dph (Figure 3.5). From 2-21 dph, the DWI ws significntly higher for lrve fed Artemi compred to the groups fed copepods nd formulted feed. Lrve fed Artemi incresed their dry weight 5.5 % per dy, while the other lrve groups grew 1.1 nd 0.6 %. After wening to formulted feed for ll groups the DWI ws % nd no significnt difference etween the groups ws found. The group wened to formulted feed t 9 dph hd the highest DWI for this period. The trend for the whole feeding period (2-50 dph) ws higher DWI for the Artemi lrvl group (6,3 %) thn for the copepod group (4.9 %) nd the formulted feed group (5.0 %). However, no significnt difference etween the groups ws found (p=0.053) Artemi Copepod Formulted feed No feeding Dys post htch Figure 3.5. Men dily weight increse (%, tretment -1 ) for time intervls representing the different feeding periods; 0-2 dph: efore feeding strted, 2-22 dph: live feed period, ut 7-22 dph: one group fed formulted feed, dph: formulted feed period for ll groups nd 2-51 dph: feeding period. Smpling size; n = 16 t 0 nd 2 dph, n = 15 tretment -1 t 6, 10, 13, 17, 21, 24, 29, 34 nd 38 dph, n = 30 tretment -1 t 42, 45 nd 50 dph, nd n = 48 tretment -1 t 51 dph (only DW). Stndrd error (±) re represented y error rs. Letters indicte significnt differences etween tretments. 26

34 Results Stndrd length Before feeding strted, the newly htched lumpsucker lrve hd men stndrd length of 6.4 mm (Figure 3.6). Alredy from 6 dph the lrve fed Artemi hd significntly higher stndrd length compred to the other two groups. The difference etween the Artemi fed group nd the lrve fed copepods nd formulted feed continued throughout the experiment. Both the groups wened from copepods to formulted feed (9 nd 22 dph) hd following period with slower increse in stndrd length. During this period the stndrd length ws slower for the group wened to formulted feed compred to oth the other two groups. At the end of the experiment the Artemi group hd men stndrd length of 13.7 mm, significntly longer thn the copepod nd formulted feed groups; 11.2 nd 11.8 mm. Significnt difference etween the copepod nd the formulted feed group were lso found t 50 dph. The Person correltion test showed significnt nd positive liner correltion etween lrvl dry weight nd stndrd length in ll groups (Figure 3.7). The mximum size otined for lrve from Artemi group hd stndrd length 16.1 mm nd dry weight 25.2 mg nd for the formulted feed group 13 mm nd 13.3 mg. The heviest lrve from copepod group ws 16.1 mg nd 12.6 mm long, while the longest ws 12.9 mm long nd weighted 12.5 mg. 27

35 Men stndrd lenght (mm) Results 14,0 13,0 Artemi Copepod Formulted feed No feeding Live feed period Formulted feed period 12,0 c 11,0 10,0 9,0 8,0 7,0 c c c 6, Dys post htch Figure 3.6. Men stndrd length (mm tretment -1 ) t 2, 6, 10, 13, 17, 21, 24, 29, 34, 38, 42, 45 nd 50 dph. Smpling size; n = 16 tretment -1 t 2 dph, n = 15 tretment -1 t 6, 10, 13, 17, 21, 24, 29, 34 nd 38 dph, nd n = 30 tretment -1 t 42, 45 nd 50 dph. Letters indicte significnt differences etween tretments nd stndrd error re represented with ± error rs. The first dotted line indicte the end of feeding copepods to formulted feed group, nd the second dotted line re the end of live feed period forartemi nd copepod groups. 28

36 Dry weight (mg lrv -1 ) Dry weight (mg lrv -1 ) Dry weight (mg lrv -1 ) Results Artemi Copepod Stndrd length (mm lrv -1 ) Stndrd length (mm lrv -1 ) 30 Formulted feed Stndrd length (mm lrv -1 ) 3 Figure 3.7. Correltion etween dry weight (mg lrv -1 ) nd stndrd length (mm lrv -1 ). 1: lrve fed Artemi, 2: lrve fed copepods nd 3: lrve fed formulted feed. Blue lines re the liner correltion lines. Person point-iseril liner correltion coefficient, r p 2 =0,965 (Artemi sp.), 0,935 (copepod) nd 0,974 (formulted feed) nd the correltion ws significnt for ll groups. 29

37 Results 3.3. Survivl For ll groups the men survivl ws over 79 % during the experiment (Tle 3.1). Lrve from the Artemi group hd significntly higher survivl thn lrve fed formulted feed dph. The lowest survivl ws oserved for the copepod fed lrve, ut the high vrition etween the three tnks in this tretment gve no significnt differences from the other tretments. At the lst smpling dy, the survivl of the lrve fed Artemi ws significntly higher thn oth the other groups. The vrition in survivl etween lrvl tnks ws from 34 dph higher in the copepod nd formulted feed lrvl groups. There ws 0 % vrition in survivl etween the three tnks fed Artemi. After wening to formulted feed, sudden increse in mortlity ws experienced from dph in copepod nd formulted feed fed lrvl groups. At the sme time hevy growth of cteril threds ppered in those tnks. The cteri threds were first oserved in copepod fed tnk (tnk 10). From dph the mortlity ws % higher thn in the two other tnks from the copepod tretment. Severl lrve got tngled into cteri treds nd suffocted nd n incresed numer of lrve were found ded on the ottom. Wter exchnge ws incresed more t this point to try to reduce the cteri. Different mount of cteri threds ws oserved in lrvl tnks fed copepod nd formulted feed the following dys. Tle 3.1. Men survivl (%) ± stndrd error (SE) per tretment on 21, 34, 42 nd 51 dph from dily mesurements t 3-51 dph. Smpling size n=3 tnks tretment -1. Significnt differences re indicted y different letters. Dph Men survivl (%) ± SE Artemi Copepod Formulted feed ± ± 0 99 ± ± 0 83 ± 7 89 ± ± 0 80 ± 7 87 ± ± 0 79 ± 6 83 ± 3 30

38 Results 3.4. Skeletl development Figure 3.8 present the ge-relted skeletl development of the lrvl group fed with Artemi showed the with highest growth. Newly htched lumpsucker lrve hd ossifiction in mouthprts (mxill, premxill nd dentle), suction disc, opercle, cleihtrum, the first vertere with the ssocited neurl nd heml rches. At 10 dph, the fin rys of ll fins (dorsl, pectorl, nl nd til fins) were ossified. Lrvl skull/crnium nd rnchiostegl rys were lso ossified. Mouthprts hd more ossifiction nd teeth were oserved in the dentl ones. The end of the spine, the urostyle one hd end upwrds nd flexion hd strted. First sign of ossifiction in neurl nd heml rches were seen in newly htched lrve (0 dph). The heml nd neurl rches of compct, prtly nd fully ossified vertere were ossified. Lrve hd 28 neurl rches, the sme mount s vertere. The numer of heml rches ws 17 of norml length. Close to the stomch lrve hd 3-5 shorter heml rches, which not covered the stomch region (Figure 3.8). From dph mny of the heml nd neurl rches were not stright nd hd one/two ends (lrve 2-5, Figure 3.8). This were seen in lrve up to 11 mm, while lrve over 12 mm hd stright rches without the ends. The urostyle nd oth hypurl ones were ossified from 21 dph in lrve with SL 7-8 mm. More of the fin rys were ossified in ll fins. From dph, the fin rys hve grown longer. First ossifiction of pterygiophores of dorsl nd nl fins were oserved t 34 dph. Four rdils of the pectorl fins were visile under the rnchiostegl rys. The pterygiophores of dorsl nd nl fins grew longer nd in numer from dph. At 50 dph the neurl nd heml rches were stright nd met the end of the pterygiophores, mening tht ones covered the entire trunk. Hypurl supporting the til fin were fully ossified in lrve over 12 mm. High vrition ws found etween numer of fully ossified vertere in individul lrve (Figure 3.10). The smllest lrve with ll 28 vertere fully ossified ws 11.2 mm, while the iggest lrve without ll vertere ossified ws 16 mm. 31

39 Results 0 dph 10 dph 21 dph 34 dph dph dph 50 dph Figure 3.8. Visulistion of one development from 0-50 dph of lumpsucker lrve fed Artemi sp. Lrvl stndrd length: 0 dph 5.5 mm, 10 dph 6.4 mm, 21 dph 7.0 mm, 34 dph 8.9 mm, 42 dph 9.4 mm, 50 dph 11.5 mm. Blck rs equl 500 µm. 32

40 Results 3.5. Bone ossifiction in specific prts There ws difference etween the lrvl tretments when compring onset of ossifiction in vertere, til, nl nd dorsl fins, urostyle nd hypurls. The different feeding effect on skeletl development were found relted to oth ge nd size Suction disc Newly htched lrve hd ossifiction in pelvis, spine ry nd soft rys of the suction disc (Lrve 1 in Figure 3.9). Most prt of the development of the suction disc hppened the first 10 dys post htching. The pelvis ws more ossified nd the spine nd soft rys were longer nd wider. From dph the pelvis ossified more, the rys elongted, nd the size of the suction disc incresed. The spine nd soft rys were lrger nd more curved. After 21 dph were the chnges in ossifiction smll, nd comprison with the lrve from 50 dph indicted tht the suction disc ws complete lredy from 21 dph. 0 dph 10 dph 21 dph 50 dph Soft rys Pelvis Spine ry Figure 3.9. Visuliztion of one ossifiction on the ventrl side of lumpsucker lrve with the suction disc. Lrvl stndrd length t 0 dph 5.7 mm, 10 dph 6.5 mm, 21 dph 8.9 mm nd 50 dph 14.2 mm. Nomenclture defined y Voskooinikov nd Kudryvtsev (2014) re presented on lrve from 50 dph. Blck rs equl 500 µm. 33

41 Results Vertere Newly htched lrve hd n verge of 5.1 trnsprent nd 20.7 prtly ossified vertere (Tle 3.2). At 10 dph the lrve fed Artemi hd significntly higher numer of fully ossified vertere nd the difference from the other two groups continued throughout the experiment. All lrvl groups hd men totl numer of 28 verters from 10 dph, nd over 60 % of the vertere were compct ossified. The first fully ossified verter ws oserved in lrve fed Artemi t 10 dph, while the other groups hd none. One lrv from the formulted feed group hd only 27 vertere t 21 dph. Significntly less compct ossified vertere were found in lrve fed formulted feed from dph. The increse in fully ossified vertere ws low until 42 dph for ll lrvl groups. At 42 dph over 56 % of vertere of the Artemi lrve were fully ossified. The other two groups hd less thn 14 % fully ossified vertere t 42 dph. The increse in numer of fully ossified vertere ws high for the copepod nd formulted feed group from dph nd t 50 dph over 73 % of vertere were fully ossified. Artemi fed lrve hd t the sme time over 96 % of the vertere fully ossified. Lrve from the copepod group hd higher mount of fully ossified vertere t smller size compred to the other tretments (Figure 3.10). The smllest lrve with % of the vertere fully ossified ws 9.9 mm compred to 11.4 nd 10.7 mm for Artemi nd formulted feed fed lrvl groups. Over 90 % fully ossified vertere were oserved in 1,4 mm shorter lrve thn from the other groups. The difference in length etween the groups ws smller in lrve with ll 28 vertere fully ossified. A higher mount of the lrve fed Artemi hd over 90 % fully ossified vertere thn the other two groups. However, those lrve hd stndrd lengths over 14 mm, lrger lrve thn found in the other groups. 34

42 Results Tle 3.2. Men numer of vertere (lrv -1 ) clssified depending on the degree of ossifiction with men stndrd length (mm) ± stndrd error (SE) per tretment. Trnsprent = no visile one ossifiction, prtly = ossifiction strting t the rches root ut not filling the whole vertere, compct = ossifiction in the whole vertere ut not fully sturted colour, nd fully ossified = fully sturted colour filling the whole vertere. The totl numer of vertere is the sum of trnsprent, prtly, compct nd fully ossified vertere. Smpling size; n = tretment -1 t 0, 10, 21, 34, 42 nd 50 dph. Significnt differences etween tretments re indicted y different letters. Dph Tretment N Men SL (mm) ± SE Numer of vertere (lrv -1 ) ± SE Trnsprent Prtly Compct Fully ossified Totl 0 No feeding ± 0,1 5.1 ± ± ,8 ± 0,5 10 Artemi ± 0,1 1.9 ± ± ± 0,5 0,8 ± 0,1 28,0 ± 0,1 Copepod ± 0,1 1.7 ± ± ± 0,1 0 28,0 ± 0,0 Form. feed ± 0,0 1.6 ± ± ± 0,0 0 28,0 ± 0,0 21 Artemi ± 0, ± ± 0,3 3,8 ± 0,2 28,0 ± 0,0 Copepod ± 0, ± ± 0,7 1,9 ± 0,5 28,0 ± 0,0 Form. feed ± 0,1 c 1.0 ± ± 0.5 c 15.1 ± 0,8 1,5 ± 0,2 27,9 ± 0,1 34 Artemi ± 0, ± ± 0,2 4,7 ± 0,4 28,0 ± 0,0 Copepod ± 0,1 0.1 ± ± ± 1,0 2,4 ± 0,4 28,0 ± 0,0 Form. feed ± 0,1 0.3 ± ± ± 0,8 c 2,0 ± 0,2 28,0 ± 0,0 42 Artemi ± 0, ± 1,7 15,8 ± 1,7 28,0 ± 0,0 Copepod ± 0, ± ± ± ± 0.0 Form. feed ± ± ± ± ± Artemi ± ± ± ± 0.0 Copepod ± ± ± ± ± 0.0 Form. feed ± 0.1 c ± ± 0.5 c 20.7 ± 0.6 c 27.9 ±

43 Numer of fully ossifed vertere (lrv -1 ) Numer of fully ossified vertere (lrv -1 ) Results Artemi Artemi 100 % Copepod Form. feed % % Copepod Form. feed 100 % % 80 % Stndrd length (mm) Stndrd length (mm) Figure Numer of fully ossified vertere (lrv -1 ) plotted with lrvl stndrd length (mm). Totl numer of vertere per lrv ws 28. Lrve ws smpled t 0, 10, 21, 34, 42 nd 50 dph, nd smpling size, n = 10 tretment -1 t 0 dph, n = 21 tretment -1 t 10 dph, n = 30 tretment -1 dy -1 t 21, 34 nd 42, nd n = 150 tretment -1 t 50 dph. The grey lines re mrked with the percentge of vertere fully ossified e.g. 80 % = 80 % of the ll vertere in lrv is fully ossified. 36

44 Results Dorsl fins nd nl fin No ossifiction ws oserved in the nl nd dorsl fins in lrve from 0 dph. From 10 dph ll lrve hd ossifiction in the fin rys nd ws clssified with score 1 (Tle 3.4). Score 2 ws otined from 21 dph, nd significntly more lrve from the Artemi tretment hd score 2 in the second dorsl fin thn from the formulted feed tretment. The highest difference etween the tretments ws seen in the nl fin, where 97 % Artemi, 50 % copepod nd 17 % formulted feed fed lrve hd score 2. However no significnt difference ws found due to vrition etween lrvl tnks nd only three replictes (fish tnks) per tretment. Ossifiction in the pterygiophores (score 3) ws from 34 dph, nd t 42 dph in 100 % of the lrve fed Artemi. The numer of lrve fed formulted feed with score 3 in first dorsl nd nl fin were significntly lower thn lrve fed Artemi At the lst smpling dy (50 dph) over 99 % of the lrve from ll tretment hd ossifiction in pterygiophores. The development of nl nd oth dorsl fin relted to stndrd length ws similr. Lrve with stndrd length 5-6 mm hd score 0 nd score 1 were oserved in mm lrve except from one lrve of 8.8 mm (Figure 3.11). From mm lrve hd score 2, score 3 ws reched from 8.3 mm nd in ll lrve over 11.3 mm. Lrve from the copepod tretment reched the fin ossifiction score 3 in smller lrve compred to the other tretments, ut the difference etween the lrvl tretments ws not significnt. The size intervl with fin development from score 2 to 3 ws lso smller for the copepod fed lrve, which mens they hd fster fin development. From mm the fins developed to score 3, while fins in lrve fed Artemi nd formulted feed were nd mm. 37

45 Results Tle 3.4. Men percentge of lrv with first dorsl, second dorsl nd nl fin ry scores nd lrvl men stndrd length (mm) ± stndrd error (SE) for the tretments. Smpling dys; 10, 21, 34, 42 nd 50 dph, nd smpling size, n = tretment -1. Fin ry score 0 = no ossifiction, 1 = ossifiction of less thn hlf of the length of dorsl/nl fin rys, 2 = ossifiction in over hlf the length of dorsl/nl fin rys, ut not pterygiophores nd 3 = ossifiction of ll dorsl/nl fin rys nd pterygiophores. - = 0 % lrve with tht prticulr score. Significnt difference etween the tretments is indicted y different letters. Dph Tretment Men SL Dorsl fin 1 score Dorsl fin 2 score Anl fin score n (mm) ± SE Artemi 6.7 ± Copepods 6.3 ± Form. feed 6.2 ± Artemi 8.2 ± Copepods 6.9 ± Form. feed 6.4 ± 0.1 c Artemi 9.8 ± Copepods 8.4 ± Form. feed 8.2 ± Artemi 11.9 ± Copepods 9.6 ± Form. feed 10.1 ± Artemi 13.6 ± Copepods 11.4 ± Form. feed 11.6 ± 0.1 c

46 Anl fin score Dorsl fin 1 score Dorsl fin 2 score Results Artemi Copepod Formulted feed 0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 Stndrd length (mm) 1 Artemi Copepod Formulted feed 0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 Stndrd length(mm) 3 Figure Fin ry score of first dorsl, second dorsl nd nl fin in individul lrv with stndrd length (mm). Lrve smpled 2 1 Artemi Copepod Formulted feed 0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 Stndrd length (mm) t 0, 10, 21, 34, 42 nd 50 dph, nd smpling size, n = 10 tretment -1 t 0 dph, n = 21 tretment -1 t 10 dph, n = 30 tretment -1 dy -1 t 21, 34 nd 42, nd n = 150 tretment -1 t 50 dph. Fin ry score 0 = no ossifiction, 1 = ossifiction of less thn hlf of the length of dorsl/nl fin rys, 2 = ossifiction in over hlf the length of dorsl/nl fin rys, ut not pterygiophores nd 3 = ossifiction of ll dorsl/nl fin rys nd pterygiophores. Scoring for different tretments re lifted to see differences etter. 39

47 Averge numer of til fin rys (lrv -1 ) Results Til (cudl) fin nd til region Newly htched lrve hd no ossifiction in the til fin rys, urostyle or hypurl ones. The verge numer of til fin rys were 9-10 t 10 dph nd incresed with ge to rys (Figure 3.12). Lrve fed Artemi hd significntly more ossified fin rys thn lrve from the other tretments from dph. At 21 dph the verge numer of fin rys of Artemi fed lrve ws 15 ossified fin rys, while lrve fed copepods nd formulted feed hd 14 nd 12. The til fin of Artemi fed lrve ws fully developed from 21 dph, since the numer of til fin rys not continued to increse. For the other tretments ws the fully developed til fin oserved t 34 dph. The mximum numer of fin rys t 50 dph ws 18 for the lrve fed Artemi, while mx 17 rys ws oserved in lrve from the other two tretments. The size- relted ossifiction of til fin rys showed shift where lrve under mm hd under 15 ossified rys nd igger lrve hd over 15. From lrvl size mm most of the ossifiction of til fin rys hppened. The numer of til fin rys in individul lrve vried from in lrve over 7 mm. Ossifiction in over 15 til fin rys ws oserved in 0.3 mm smller lrve fed copepods compred to the other tretments (Figure 2.13). 16,0 15,0 Artemi Copepod Form. feed c c 14,0 13,0 c 12,0 11,0 10,0 9,0 8, Dys post htch Figure Men numer of ossified til fin rys ± stndrd error (SE) lrv -1 tretment -1. Lrve smpled t 0, 10, 21, 34, 42 nd 50 dph, nd smpling size, n = 10 tretment -1 t 0 dph, n = 21 tretment - 1 t 10 dph, n = 30 tretment -1 dy -1 t 21, 34 nd 42, nd n = 150 tretment -1 t 50 dph. Error rs represent ± SE nd significnt differences etween tretments re indicted y different letters. 40

48 Men numer of til fin rys (lrv -1 ) Men numer of til fin rys (llrv -1 ) Results Artemi Copepod Form. feed Artemi Copepod 10 Form. feed Stndrd length (mm) Stndrd lenght (mm) Figure Men numer of ossified til fin rys relted to size, stndrd length (mm) lrv -1 tretment -1. Top-left: ll tretments, top-right: Artemi sp., lower-left: copepod nd lower-right: formulted feed. Smpling size, n= lrv tretment -1 on 0, 10, 21, 34, 42 nd 50 dph. Lrv from 0 dph hd no ossified til fin rys. 41

49 Results Ossifiction of the urostyle ws first oserved in ll lrve from ll tretment over 9.0 mm (Tle 3.6). The first sign of ossifiction in 6.5 mm long lrv fed copepod, while the first lrve fed Artemi nd formulted feed were 7.5 nd 7.0 mm. Hypurls strted to ossify in ll lrve over 7.5 mm fed Artemi nd in 6.5 mm lrve fed copepod nd formulted feed. Tle 3.6. Bone ossifiction of ones in til region (urostyle, hypurls nd til fin rys) nd presence of flexion relted to size, stndrd length (mm). Light grey represents the shortest lrv oserved with sign of ossifiction, nd drk grey is the size where ll lrve hd ossifiction in tht specific prt. All lrve over 9.5 mm hd ossifiction in urostyle, hypurls, til fin rys nd presence of flexion. Urostyle Lower hypurl Upper hypurl Tretment Artemi Copepod Form. feed Artemi Copepod Form. feed Artemi Copepod Form. feed Stndrd length (mm) ,

50 Results 3.6. Skeletl nomlies The skeletl nomlies found in lumpsucker lrve were twisted neurl rches, flt hed, norml vertere nd xis devitions (Tle 3.7 nd Figure 3.14), while norml jw ws not oserved in ny lrv. Other nomlies oserved were norml fin rys or pterygiophores presented in lrv d nd e in Figure None of the nomlies were oserved in lrve t 10 dph. From dph there ws only significnt difference etween the tretments in occurrence of twisted neurl rches. Tle 3.7. Men percentge of lrv with nomlies (%) ± stndrd error (SE); flt hed, norml vertere segments, xis devitions (lordosis, kyphosis nd scoliosis) nd other deformities (fins). Smpling size, n=30 lrve tretment -1 t 21, 34 nd 42 dph, nd n=150 lrve tretment -1 t 50 dph. Dph Tretment Men percentge of lrve with nomlies (%) ± SE Anorml skull Anorml vertere Axis devitions Other nomlies 21 Artemi 10 ± ± 3 Copepod 13 ± ± 5 Form. feed 11 ± ± 8 c Twisted neurl rches 34 Artemi - 3 ± ± 11 Copepod 7 ± 3 3 ± ± 10 Form. feed 13 ± 7 3 ± ± Artemi - 7 ± ± 5 Copepod 3 ± ± 3 Form. feed 3 ± ± 3 90 ± 5 50 Artemi - 6 ± 0 1 ± 0 1 ± 1 11 ± 3 Copepod 1 ± 1 11 ± 2 1 ± 0 1 ± 1 17 ± 5 Form. feed - 7 ± 1 1 ± 0 1 ± 1 32 ± 4 43

51 Results Anorml skull Lrve oserved with flt hed hd fltter dorsl side of the skull compred to the verge skull shpe (lrv c in Figure 3.14). The highest numer of lrve with flt hed ws oserved t dph. At dph decresed the occurrence of flt hed nd under 3 % were found to hve the nomlity the lst two smpling dys. No significnt difference etween the tretments were found due to high vritions etween tnks Anorml vertere nd xis devitions Two types of norml vertere were oserved; (1) two vertere fused without seprtion etween (lrv, Figure 3.14) nd (2) vertere with different position compred to the rest of the spine (lrv, Figure 3.14). Fused vertere were only oserved in the two lst vertere of the cudl prt of the spine. No significnt difference ws found etween the tretments. The occurrence ws highest on 50 dph when 6, 11 nd 7 % of the lrve fed Artemi, copepod nd formulted feed hd norml verters. Axis devitions; lordosis, kyphosis nd scoliosis were only found in 1 % of lrve t 50 dph from ll tretments (Tle 3.7). This equlled in totl five lrve ffected from ll tretments. One lrv fed formulted feed hd oth lordosis nd kyphosis (lrv f in Figure 3.14). Lordosis lone ws found in one lrv fed Artemi nd one lrv fed formulted feed. Kyphosis ws oserved lone in one lrv fed copepod nd one other lrv from the sme tretment hd scoliosis. No significnt difference ws found etween the tretments regrding xis devitions Twisted neurl rches Twisted neurl rched were oserved in lrve from dph (Tle 3.7 nd lrv g in Figure 3.14). The sme spirl t pex were lso found in the heml rches. Significntly less percentge of lrve from the Artemi group hd twisted neurl rches t dph. The lst smpling dy, the occurrence of twisted neurl rches decresed for ll tretments. Lrve fed formulted feed hd significntly more twisted neurl rches thn the other tretments t 50 dph. The numer of twisted rches per individul lrve lso decresed t 50 dph, where the verge numer per lrve ws under 1 for ll tretments (Figure 3.14). The sme ws seen in the Artemi nd copepod lrvl groups t 42 dph, where the men numer of twisted neurl rched per lrve ws under 2. 44

52 Results c d e f g Figure Exmples of oserved skeletl nomlies in lrve from dph. ) Fused two lst vertere, ) norml position of the second lst vertere, c) flt skull on dorsl side, d) pterygiophore of nl fin orientted wrong wy, e) norml position of pterygiophore of second dorsl fin, f) lordosis nd kyphosis nd g) twisted neurl rches. 45