THE REGULATIO NOF PARTICLE TRANSPORT WITHIN THE VENTRAL GROOVE OF THE MUSSEL GILL(MYT1LUS £DUUS)IN RESPONSETO
VARIATIONS IN ENVIRONMENTAL CONDmONS
by NicoleBertineRichoux
Athesis submitted to the SchoolofGraduateStudies
in partial fulfilmentoft he requirements for the degreeof
Master ofScience
Departmentof Biology MemorialUniversityofNewfoundland
December 1997
SI.John' s Newfoundland
.+.
ofAcqutsitionsandNationalCanadaUbr.uy BC'iographicServices 39SWeIingIon Streec etta-ONK1A 0N4ea.-
BibflOtheque nationale du Canada Acquismooset servicesbibliographiques
395.~Wellingtan oa-ONK1AONt
e--
The author bas granted a non•
exclusive licence allowing the National Library ofCanada to reproduce,loan. distribute or seU copies ofthis thesisinmicroform, paper or electronic fonnats.
The author retainsowners hip of the copyrigbtinthisthesis.Neitherthe thesisnor substantial extracts from it maybe printed or otherwise reproducedwithout the author's permission.
L'auteur a accorde une licencenon exclusive pennettant
a
laBibliotheque nationale du Canada de reproduire,preter,distribuerou vendre descopiesde cette thesesous la forme de microfiche/film,de reproduction sur papier ousurformat electronique,
L'auteur conserve la proprietedu droitd'auteurqui protegecette these. Ni 1atheseni des extraitssubstantiels de celle-ci ne doivent
eire
imprimes ou autrement reproduits sansson autorisation.0-612-34220-4
Canada
ABSTRACT
In order to understan dthoroughlythefeeding processes ofthe blue mussel, My/ilusedulis,inresponseto thevariableenvironmentalconditio nsitexperiencesin nature,itisimportantto examineindividually thedifferent compo nents that compriseits feeding system.The ciliated ventralfoodgrooverepresents oneoftheseprimary components.withinwhich the majorityoffood particlestrapped bythegill are transported to the labialpaJpsandgut.Theabilityof musselstoadjust food transportrateswithinthis groove couldserveasan important feedingregulatorymechanism in response to variations in theenvironment.
Usingvideo endoscopy,mucous strandvelocitiesin the ventralgrooveofmussels Mytilusedulisweredeterminedover variabletime periodsandin respo nseto short-term manipulationsofambientpanicle concentration. temperatureand panicletype.Mucous strandvelocitiesdecreasedwith increasingambient panicle concentratio n andpanicleload onthegill.andalso whenpanicletypewas switchedfrom algae to sediment.This evidencesupportsthe hypothesis thatM.edulispossesses compensation mechanisms
involvingpanicletransport atthelevel of theventralgrooveciliatodeal withshort-term
changesin theenvironment.Furt hermore,mucous strandvelocityin the ventralgroove increased whentheambient temperature ofmusselsacclimatedto4.SoCwasincrease d to lSoC.Thisrespon se is consistentwithstandard physiologicalrespon ses ofciliarysystems tochangesin temperature .Changesin velocity over timewere also observed.but there wasnoconsistentpatte rn.
ACKNOWLEDGEMENTS
twouldlike tothank:my supervisor,Dr.RayThompson,for his valuable input throughout the productionofthis thesis.Thanksalso to my committee members.Dr.Don Deibel and Dr.PatDabinett, for their suggestions.The assistance and advice of Elizabeth Hatfield, Alexander Bochdansky and Trevor Avery are greatly appreciated.as well as the supportandpatienceof my familyand friends.
This research was funded byfellowships from Memorial University of NewfoundlandandanNSERC research granttoDr.RJ. Thompson.
iii
TABLE OF CONTENTS
Abstract .
Acknowledgements. ListofTables Listof Figures_._
...ii ...iii ... ...vi ...vii
CHAPTERI-INTRODUCTION . I
1.1Backgrou nd.. ..1
1.2Description of feeding processes ofMytilus edulis_ .3
1.3Structure andfunction ofcilia.. . 10
1.4 Controlof cilia... . 12
[.5 Mucusstructure andrheology.. . 13
l.6 Objectives.. .. . . 14
CHAPTER 2 - METHODS 15
2.1 Collectionand maintenanceof mussels.. ...,_ IS
2.2 Video setup .. .. . 15
2.3 Particleconcentrationand temperature experiments (static system) 17
2.3.1Setup . 17
2.3.2Markers.. 17
2.3.3Videotaping . [8
2.3.4Clearance rates 23
2.4 Time effect experiments.. . 23
2.4.1Setup . 23
2.4.2Videotaping . 25
2.4.3Clearance rates... . 26
2.5 Particle qualityexperiments. . 28
2.S.1Setup.. . 28
2.5.2Sediment.. 30
2.6 Acute temperaturechange experiments.. . 30
2.7Gill filamentmeasurements.. . 30
2.8 Video analysisand variables.. . . 32
2.9 StatisticaJanalyses 33
CHAPTER 3 - RESULTS 3.1Gill filament measurements 3.2 Mucousstrand size...
...41 ...41
. . .41
3.3Statistical analyses . 45
3.3.1Final model.. . .. .45
3.3.2Particleload.... . 45
3.3.3Particleconcentration.. . .48
3.3.4 Clearancerate.. . 50
3.3.5Time effect experiments 50
3.3.6Particlequalityexperiments.. . 65
3.3.7Acute responsesto temperature changes 68
CHAPTER 4· DISCUSSION 71
4.1Gill filament measuremen ts 71
4.2Mucous strandsize 72
4.3Particle concentration,particleloadand clearance rate . 72
4.4 Temperature effects 77
4.5 Time effects 79
4.6 Particletype... . 80
4.7Summary.. . 82
References . .. 83
LISTOFTABLES
Table 2.1:Experimentalconditions to which mussels(MytiluseduJis)weresubjected duringparticle concentrationandacclimatedtempera tureexperimentsin astaticor flow-
through system. .. . _... . 20
Table2.2:Experimentalconditionsto which mussels(Mytilusedulis)were subjected during timeeffect experimentsinaflow-throughsystem.. 27 Table2.3:Experimentalconditions to which mussels(My /ilus edulis)were subjected duringparticlequality experimentsusing a flow-throughsystem _ _ 29 Table2.4:Experimental conditionstowhichmussels(Mytilus eduJis)were subjected during non-acclimation temperatureexperimentsusing a flow-through system. . J 1 Table2.5:Variablesconsideredinthe statisticalanalyses of pooledpaniclevelocitydata
inMytilusedulis. . . ,..38
Table3.1:Summary or filament width measurementsattheventralgroove ofMyIiius edulisusing bothimage analysis and astereoscope withcalibrated micrometer 42 Table3.2:P valuesassociated witheach significant explanatory variableinthefinal GLM
(Equations3.1and 3.2).... . .46
Table 3.3:Summary ofstatic system experimentshavingsignificantlinear regression slopesfortime
«
120minutes) vsparticlevelocityin theventral grooveofMytilusedulis.; .55
Table4.1.Someexamplesof particle velocitieson the frontalsurfaceandventralgroove
ofbivalvesin differentstudies. . 74
LIST OF FIGURES
Figure1.I.Flow chartshowingthedifferent compart mentsinthe suspension feedingand
digesti ve systems ofMytilusedulis _ 4
Figure1.2.Diagranunaticcross-sect ionof thegills ofMytilus edulis 5 Figure1.3Diagramm aticcross-sectionof one filament showing four differentciliarytypes
on the gillsofMytilus edulis. . 6
Figure1.4.Sideviewof one filament ofMytilusedulisshowing thethree tractstaken by
particles to arrive in the ventralfood groove . . 8
Figure1.5.Frontal view of the three (factsleadingfrom thefilamentstothe ventral
grooveinMytilusedutis. . 9
Figu re.2.1.Static systemsetup andvideotaping equipment usedin panicleconcentration
andtemperature experimentswithMy/i/usedulis 16
Figure2.2.Position of the opticalinsertio ntubein thepallial systemofMytilusedulis duringvideota pingof the ventralgroove... . 19 Figure 2.3.Flow-through systemsetup used intemperatu re.time andpaniclequality
experiments withMytilus edulis. . 24
Figure2.4.VentralgrooveofMytiluseduliscontainingathin mucous strand 34 Figure2.5.Ventral grooveofMytilus eduliscontaining a mediummucous str andwith
one micrcspherevisible . 35
Figu re2.6.VentralgrooveofMytilu seduliscontainingathickmucous strand and two
micro sphe res.. .. . . 36
Figure2.1.Ventral groove ofMytiluseduliswitha mucousstrandextendingoverthe
filamenttips .31
Figure3.1.Therelationshipbetweenpanicle velocityandmucous strandsizeinthe
ventralgroove ofMytilusedulis 43
Figu re3.2.The relationshipbetweenparticleloadandmuco us strandsizein the ventral
groove ofMy/ilusedulis 44
vii
Figure3.3.The relationshipbetweenparticlevelocityintheventral groove ofMytilus
edulisand particleload. . ..47
Figure 3.4.The relationship between particlevelocity in the ventralgrooveofMy/ilus
edulisandambient particleconcentration. . 49
Figure 3.5.The relationshipbetweenparticlevelocity in theventral grooveof
Myutus
edutisandclearance rate 5I
Figure 3.6.The relationship betweenclearancerateofMy/ilus edulisand particle
concentration.. . .. . 52
Figure3.7:A.The change in particlevelocity in theventral grooveof mussel#22over timeinstatic systemexperiments
B.Thechange in particlevelocity in theventral groove of mussel #24over
time instaticsystemexperiments... .. 53
Figure3.8:A.The change in particlevelocity in theventral groove of mussel #16over timeinstaticsystem experiments
B.Thechange in particle velocityin theventralgroove of mussel#18over
timeinstaticsystemexperiments.. . 54
Figure 3.9.The relationshipbetween particlevelocity in the ventralgrooveof mussel #9
andtime spent feeding inexperiment #36.. ...56
Figure3.10.Therelationshipbetweenparticle velocity in theventral groove ofmussel #9
and time spent feeding in experiment #37.. .. 57
Figure3.11.The relationshipbetweenparticle velocityon theventralgroove of mussel
#30 and time spent feedingin experiment #39 58
Figure 3.12.The relationshipbetween particle velocity onthe ventral grooveofmusse!
#33and timespent feeding in experiment#49. ..S9
Figure3.13.The relationshipbetweenparticle velocity on theventral groove ofmussel
#32and timespentfeeding in experiment #38. .... ..60
Figure3.14.The relationshipbetween particle velocity onthe ventral grooveofmussel
#33 and time spent feedingin experiment #50.... ...61
Figure3.l5.The relationshipbetweenparticle velocityon theventralgroove ofmussel
#30and timespent feeding in experiment #42 62
Figure 3.16.The relationship between particle velocity on theventral grooveofmussel
#24 and timespent feedinginexperiment#40 63
Figure 3.17.Therelationship between particlevelocity ontheventralgroove ofmussel #9
andtimespentfeedinginexperiment#41.. 64
Figure3.t8.The relationship betweenparticle velocityontheventralgroove ofmussel
#33 andtime spentfeeding inthe algae to inorganic sedimentexperiment#46. . 66 Figur e 3.19.The relationshipbetweenparticle velocityon the ventral groove ofmussel
#30andtime spentfeedingin the algae to inorganicsedimentexperiment #48 61 Figure3.20.The relationshipbetween particle velocity on theventral groove ofmusse!
#30andtimespent feedingintemperature experiment #43 69 Figure 3.21.Therelationship between particle velocityonthe ventral groove of mussel
#32and time spentfeedingin temperature experiment#44... . 70
ix
C~Rl-~ODUcnON
l.l Background
Bluemussels(Mytilusedulis)livein estuarineandcoastal areaswhere they experiencevariable temperaturesandconcentrationsofalgalcells,detritusandsilt.There are manypublished studiesconcerning the ratesatwhich suspensionfeedingbivalves process variousqualitiesand quantities of particles indifferentenvironmentalconditions (e.g.Bayneer al.,1976~Winter,1978;Navarro&Iglesias, L993).Mytilu sedulisis importantasan aquaculture species,anditisabundant inall major oceansinthe world whereit is knownto contributesignificantly tonutrientrecycling and resuspension (Asmus&Asmus, 1993;Dame,1993).For thesereasons,therehasbeen interestin determininghowenvironmental factorsaffect differentaspects of suspensionfeedingsuch as pumpingrate, clearancerate, selectionefficiency,ingestionrate,rates of feces and pseudofecesproduction. and absorption efficiency.Theadaptations of suspensionfeeders todifferent conditionshave proventobevery complexanddiverse. and most responses occurinconjunctionwith adjustments within the pallial system.
Since thepallialsystem hasprovento bedifficult to studyinvivodueto the presence of anopaqueshell,early experimentsonparticle transportvelocitiesor ciliary beatfrequencies,determ inedstroboscopically,weredone with excisedfragments ofgills, damagedspecimensor juveniles (e.g.Gray,1931;Hirasaka et al., 1957;Hoshi&
Hoshiyama,1963;Oral, 1967;Jergensen,1975;Stefano eta1.,1977;Catapaneet al., 1981;Jergensen&Ockelmann,1991).Recently,the technique of videoendoscopyhas been employed,thus allowingillvivoobservations ofthebivalvefeedingsystemand providingresearcherswith atoolwhichhasalready helped resolve some long-standing controversiesconcerningparticletransport onthe bivalve gill (e.g.Wardetal.,1993).
Endoscopyprovidesanexcellent opportunity toaddress furtherquestions concerning transportratesof particles by suspension feeders.
There are many factors which may potentially influence thevelocity ofparticleson bivalvegillsincludingthetemperature ofthesurroundingseawater,the bodysizeand feeding historyof theanimal,andthe density, volumeandtypeofeeUs towhich the animals are exposed.Wardetal.(199 1,1993 ) give mean particle velocities intheventral groove ofM. edulisattwo and three particle concentrations.respectively;Wardetal, (1994) report mean particlevelocitieson differentareasofthegills ofthe oyster, Crassostreavirgintca;and Beninger etaI.(1992)givevelocitiesinthe dorsal arch ofthe scaUop,Ptacopect enmagellanicus,at three differentparticle concentrations.Levintonet a1.(1996) report velocitieson the palp proboscidesand ctenidia of the deposit-feeder Yoldia limatulaand the etenidia ofMacomaspp..Others have measured the transport rates ofparticlesbythe frontal cilia alongthe ctenidial filaments of different species (e.g.
Ward etal.,1991;Tankersley,1996).These studies represent the bulk of datacollected for particlevelocities onthe gills ofsuspensionfeedersdeterminedinvivowith the endosco pe.Littleattempthasbeenmadeto determine whetherparticletransportrates varyaccordingtoenvironmentalconditions.or howanyresulting changesmay relate toor affectother pans ofthefeeding systemofM.edulis.
Ingeneral,there are twowaysof viewing the feedingprocessesofbivalves:1- as ciliary level responses, and2- assystem level responses.Theciliarylevelrepresents the basic unit ofresponsetochangesin theenvironmentandis determined by the physical properties ofthe cilia responsiblefor pumping waterand entrainingand transporting particles.Theseresponsesmay be direct physiologicalresponsesor indirectresponses arising from feedbackinformationoriginatinginothercompartmentsof the feeding system.If particle processingby the labial palps or thegut is arate-limitingstep,there maybe a feedback mechanism tothe gillswhichcauses a decrease orincrease in particle transport orcollection.Thisprocess would be anexampleofan indirectciliarylevel response.Alternatively,the decreaseinciliarybeat in responsetoa change in ambient temperature represe nts adirectciliary level response.
A systemlevelresponserepresentstheinnatephysiologicalprocesses of regulation of feedingbehaviourby suspensio n feeders inorder tomaximizenetenergy gain in responsetoenvironmentalchanges.The bivalvefeedingpathwaymay beviewed as a stepwise chainof components (Fig.1.1), and asystem levelresponseinvolvesthe adjustmentofoneor morefeedin gcompartmentsin ordertomaintainhomeostasisduring environmentalchange.For exam pie,bivalvesfeeding at highparticle concentrations could adopt one or more of the followingbehaviouralresponsesto regulateingestion:decrease clearance rate, increase productionof pseudofeces,or decrease time spent feeding (Foster.
Smith,1975a).Since feeding responsesshould be observed and interpreted using a combination of theselevels,thepresent work will involve feedingsystemresponsesas well as ciliary responses of musselstovariationsin the environment.
The followingfoursectioos areincluded toprovidenecessarydescriptive informationregardingthefeeding systemofM.edulisand the dynamics ofmucociliary transport.The terms andconcepts developedhere will be usedinlaterchapters tohelp explain the results of thisstudy.
J.2 Descriptionof feeding processesofMytilusedulis
Mynlusedulishas a filibranch ctenidium,i.e.nonplicate(flat) and homorhabdic (filaments all of one type).Thectenidia are folded into large,W-shaped structures made up of inner andouter demibranchs,each composedof descendingand ascendinglamellae, suspended oneach side of the body from a ctenidial axis (Bayneet al., 1976;Gosling.1992;Fig.1.2).The ctenidia divide the mantlecavity intoinhalantand exhalantchambers.Thetwolamellae formingeach demibranch are joinedtogetherby interlamellarconnectivetissuejunctions,andadjacent filaments which comprisethe lamellaeare connected looselyby ciliated discs (Morton, 1979).Lateral cilia located alongthesidesof each filament(Fig.1.3) arearrangedincontinuous rows and createthe inhalantand the exhalant currents thatenter and leave viathe inhalantandexhalant
Figure 1.1.Flow chart showing the different compartments in the suspension feeding and digestive systems ofMytilus edulis.
dorsal
shell
+-... r-.r--- - ctenldial axis
exhalant chamber dorsal tract inhalant chamber descending lamella interlamellar junc tion
/-~'--
__ ascending lamella
l»~~---
ventral g roove mantle
ventral
Figure 1.2.Diagrammatic cross-sectionof thegills ofMytilus edulis.Arrows indicatethe directionof particletransporton thegillfilaments.
Redrawn fromBayne etal.,1976.
frontal c ilia laterofrontal cilia
lateral cilia
abfrontal cilia
Figure 1.3.Diagrammaticcross-sectionof one filamentshowing the four different types of ciliaon the gills ofMytilusedulis.
siphons, respectively.There is also evidence that suggests that the abfrontal ciliaactin concert with the lateral cilia and contribute to the water flow in mussels (Jones etal.•
1990).
Water enters the inhalant siphon, which spans the entire ventral surface ofthe mantle, and passes through the ostia in the ctenidial surface.Particles are captured on the filaments, with the aid oflatero-frontal cirri, and frontal cilia located on the crests ofthe filaments (Fig.1.3) move the food particles within a fine mucous layer or in direct contact with the frontal cilia ventrally to one of four ciliated food grooves (Ward et al.,1993;
Ward, 1996).Two dorsal ciliated tracts (Fig.1.2)may also serve as secondary transport paths for particles, but evidence shows that little material is transported here inMyliJus edutts.even at high panicle concentrations (Ward et al.,1993).Filtered water passes into the exhalant chamber and exits the mantle cavity via an exhalant siphon that is smaller than, and dorsal to,the inhalant siphon.
Particles arriving at the ventral groove can enter through one of three tracts (Foster-Smith, 1975b).Tract l represents the most direct route from the filament surface to the deepest part ofthe ventral groove, tract 2 allows particles entry midway between the deepest part of the groove and the crest ofeach filament,and tract 3 on the crest of each filament allows entry to the outermost part ofthe ventral groove (Figs.1.4&1.5).
Using a window technique to view the pallial system ofM. edulis(i.e.a hole was cut in the shell and a glass cover slip glued over it), Foster-Smith (1975b) observed that the smallest mucous strings enter the ventral groove from the frontal surface ofthe filaments via tract I, larger mucous strings via tract 2 and the largest via tractJ;the three tracts thus form a graded system of entry.He also observed that mussels could decrease entry of mucous strands into the groove by inclining the tilarnent tips forward and blocking off tracts I and 2.The purpose ofth.is graded entry with respect to mucous string size is unclear,sincethe majority of captured particles eventually arrive at aventral groove, and all mucous strings reach the labial palps for selection or rejection, regardless ofthe size of the mucous strand (Foster-Smith, I975b).The mucous strings are drawn out of the
J
t )
\
j
~
t I
J I
Figure 1.4.Side view ofone filament ofMytilusedulisshowing the three tracts taken by particlesto arrive in the ventralfood groove.Redrawn from Foster-Smith, 1915b.
\ ciliated
dis cs
/
Figure 1.5.Frontal viewof thethreetractsleadingfromthe filamentsto the ventralgrooveofMytilusedulis.Arrowsrepresent pathsof particles.Redrawn fromFoster-Smith, 1975b.
ventral groo vesby thelabial palps,which thendispe rse the parti cles forselectio nby the mechanical actionofthedorsal margin and ridged palp surface (Ward, 1996).
Studiesusing videoendoscopyhave confirmed thatparticlesare capturedand trans portedon the gills viamucociliary processesin the ventralgroovesandhyd rod ynamic processesin the dorsaltracts of thegillsofsome bivalves(Ward etel.,1993 ), but mucoci liary transportis predominanton thegillofMytilusedulis.Inthisspecies,the ventral food grooveistheprimaryroute by which food is carried within cohesive mucous strings to the labialpalps,althoughsome particlesdo travelwithinthe do rsaltractswhen animalsare exposed tohighparticleconcentrations.Pseud ofecesareremovedfromthe infrabranchi alchamberat thedorsal edgeoftheinhalant sipho nby ciliary actionofthe mantle and entrainmentwith the exhalant current (Morton,1983).Pseudofeces represent material filtered from suspensionbut rejectedbythelabialpalps before ingestion,and are produ ced by suspension feedersin ordertoremove excessparticles indense suspensions (Foster-Smith,1975a;Beningeretat.•1995) or toallo w forparticleselect ionby removing inorganic or other non-nu tritious particlesand leavingnutritious organic particlesbehind foringestion(Kierboe&Mahlenberg,1981;Ward et aI.,1997).
1.3Structur e and function of cilia
Sincethe ciliumis the basic functional unit ofparticle transport onthe bivalvegill, informatio n about the innervationandactivity ofcilia canberelatedtopart icle velocity dataobtaine dusingtheendoscope.Since cilia governthepumping rateandparticle transport capabilitiesofthegill,anyalteration10ciliaryactivity would have animpact on suspension feeding.Most ofthe research to datedeals with mammalian cilia.and the relati vely small amount ofwork on bivalves involves lateral cilia(e.g.Aiello,1960,1970;
Paparo,1972;Catapane atal., 1981;Murakami&Macheme r,1982; Jargensenetal., 1990;Jorgensen&Ockelmann, 1991;Prins et aI.,1991),although somefocus has beenon the frontal cilia(e.g.Nomura&Tomita. 1933;HirasakaetaJ.•1957;Hoshi&Hoshiyama, 1963).Little mentionis madeofventralgroovecilia.but presumably theyresemble
frontalcilia ratherthan lateral cilia intheir morphology and activity since ventraland frontalciliahavesimilar functions (i.e.transportofparticleswithinmucus).
All ciliamove water or mucusviaa recovery stroke,in which the cilium moves in a curvedpath bentclose tothe epithelialsurface.andaneffective or propulsivestroke, where the ciliumswings through an erect arc perpendicularto the epithelialsurface.Asa ciliummoves initsrecovery stroke, the entrained fluid aroundit comesinto contact with otherciliaandstimulates them to initiatemovementthat will minimizecontact between adjacent cilia.Itis inthismanner thata metachronalwaveispropagated,wherebythe beatingofmany cilia is organizedintocoordinated waves (Sleighetal.,1988).The main contribution ofmetachronism to thefunctionalefficiencyofciliarysystems isprobablyto maintaina regularflow (Sleigh&Aiello,1972),which ismore importantinwater•
propelling thanmucus-propelling systems.
According to Sleigh(1982) andSleigh et a1.(1988),mucus-propellingciliahave important morphologicaland functional distinctionsfrom water-propellingcilia.Mucus•
propelling ciliaare shorter,-5to7urnin length. and displayintermittentand irregular metachronism.asopposed to water-propelling cilia whicharetoto 20 urnlong and exhibit moreregular and continuousmetachronism.Although theseare bothimportant adaptationsthat increasethe efficiencyof transport by each type of cilium,itseemsthat thetwociliary typesrespond similarlytochanges in theirenvironment.Forexample, lateralcilia, which propel water, havebeenshown to increase inbeatfrequency in responseto increased temperature (Catapaneet al., 1981;Jergensen&Ockelmann. 1991;
Prins et al., 1991) as have mucus-propellingfrontalcilia (Gray,1929;Hirasaka et al., 1957;Hoshi&Hoshiyama, 1963).
water-propellingciliatransporthomogeneousfluid,whereasmucus-propelling cilia are coveredbya two-layerfluidconsistingof a low-viscositypericiliaryfluidclosest to the epithelialsurface.andan outerviscoelastic mucous layerwhich is penetrated by the cilia tips during the effectivestroke(Sanderson&Sleigh..1981).Althoughmost of the information concerningthe structureandfunctionof muws-propelling ciliais derived
II
frommammalian ciliatransporting mucus ata liquid/air interface,the same type of propulsionis alsofoundin aquaticanimalshaving noliquid/air interface above the mucus (Sleigh,1989).Beningeretal.(1997)recentlyconfirmed thatthistwo-layerfluid model for mucociliarytransportappliestothe gillsofMytilusedulis,As a general rule, ifthe beat frequencyof the ciliaisincreased,the tip velocity of the effectivestroke also increases.both resulting in a higher velocity of the mucus (Sleigh.1982).
Changesin particle transportmayreflect changes in theactivityof frontalcilia(i.e alterations inbeatfreque ncy,fonn or pattern of beat, or metachronal coordination among cilia) (Murakami,1989),thepatternof frontalciliarycurrents,or thequantity and viscosityof the traveling mucus(Tankersley,1996).Beat frequencyand pattern of beat are controlled bycellular activitiessuchas changesin membrane potential(Eckert,1972), and the cellular activitiesare in tum regulated bynervous and humoralcontrol (seesection 1.4).Metachron ismis a productof mechanicalinteractionamongciliaand istherefore independentof cellularmechanisms(Murakami,1989).
1.4 Controlof cilia
Paparo (1972) succeededin tracing branchial nerve fibres from the visceral gangliontothe ciliatedepitheliumof thegill filaments,and other researchers observed cessationinbeatingof ciliaafter severingthe branchial nerve,thussuggesting that ciliaare undernervous control(Gray,1929;Aiello,1960).Nervous controlof ciliary activity could thereforemediatesome observed effects onsuspension feeding due to changes in the ambient environment such as temperature.particle type and particle concentration.
Further researchhasdetermined that the lateral,frontal and latero-frontalciliated celts are innervated and regulatedby serotonergic and dopaminergicneuronsoriginating inthe central nervous system(Aiello, 1960,1970;Catapaneet aI., 1978).Increasedlevels of serotonin,which can be stimulatedby factorssuchas increased temperature, resultin an increase inthe beatingratesof lateralcilia (Catapane et al.,1981).In addition,ciliary activity variesat different locations on the gill at one time (Dral,1968),and numerous
adjacentfilaments areknownto be innervated. by the same nerve bundle(Aiello and Guideri,1965).This evidencesuggeststhat theMytilusgill has the capacityforlocal controlover thevariousciliary systems.Ciliarybeatingisultimatelydepe ndent on the availabilityof ATPandconsequentlyontherates of glycolysisand respiration,whichare intumstimulated by serotonin(Aiello,1970).Modul ated calciumconcentrationsserve as intrace llular messengers to regulate ciliaryactivity (Machemer&Sugino,1989).
1.5 Mucus structure and rheology
Mucus is anon-Newt onian fluid.and thus there are special implicationsfor its movementby cilia. For example,as the forcesappliedbyciliaincrease.thevisco sity of mucus decreases,there byfacilitatingits movement(Sleighetal.,1988).Atthe same time, mucusis viscoelasticand tendstoreturn to its originalshapeafter beingstretc hed,and according to Sleighetal.(1988), viscouseffects dominateat lowbeating freque ncies whereas elasticeffectsdominateathighfrequencies.Thesepropertieshave important but noteasilyunderstoodconsequences concerning the capture andtrans port ofpaniclesby mucus.The rheology of mucus.which involves the stress,strain and shearrate generated.
byciliary action,and theviscoelastic propertiesof mucus arecriticalfacto rs in mucociliary transport(Verdugo. 1982).Vogel (1994. p 20) avoideddealingwith any non-Newtonian fluids,and states that rheologyis...perhapsthe messiestandleastunderstood branch of fluidmechanics."
Mucus consistsof a concentrated mixture of glycoproteins(containing mucopolysaccharid es),proteoglycans,lipidsand othercompone ntsthat extendto form a highly cross-linkedviscoelasticgel (Blake&Sleigh.1974;Litt etaI.•1976).Mucus secretions are verycomplexbecausetheyshare both fluid-likeand solid-likeproperties that are constantlyaltered.The viscoelasticcohesionof astring of mucustendsto keep the entirestring movingtogether as a unit, unlikewater,evenifsections ofciliaare inactive(Sleigh.1989).
I]
Mucusinthe ventral groove ofMyti/us edulisis composed ofequal mixtures of acid-dominant and neutral-dominant mucopolysaccharides, resulting in an intermediate viscosity mucus compared with the highly viscous mucus, composed solely of acid•
dominant rnucopolysaccharides, produced by some other bivalves (Beninger et al.,1993).
All filtered material istransported to the labial palpsin this intermediate viscosity mucus within the ventral grooves ofM.edutis.
1.6 Objectives
The present study focuses on one compartment ofthe feedingsystem ofMytilus edutts-the ventral groove.This area ofthe gill is important because the majority offood particles captured by the gills are transported along one of four ventral grooves before being ingested. as food or rejected as pseudofeces.The purpose ofthis study is to determine if and how food particle velocitiesin the ventral groove ofMytitusedulis change in response to different environmental factors.Data from the literature are reviewed briefly to help explain changes in the rates offood transport in the ventral groove over time and in response to short-term manipulations ofambientparticle concentration, temperature (in both acute and acclimated situations) and particle type.
CHAPTER 2 - METHODS
2.1 Collection and maintenance of mussels
Blue mussels(Mytilus edulis)were collected at Bellevue,Trinity Bay.
Newfoundland.by SCUBA divers in August 1995 and March 1996 and were kept in flowing seawater at the Ocean Sciences Centre, Logy Bay.Since physiological rate functions ofbivalves are known to be affected by body size(e.g.Winter.1973;Widdows et al.,1979;Navarro&Winter, 1982), a relatively small sizerange of mussels was selected (- 6·8 em length).Mussels were batch fed two to four times per week with lsochrysissp.cultured in f72-Si medium (Guilliard, 1983) at
-Irc.
Each mussel was scrubbed clean of epiphytes and either labeled with plastic tags using Super Glue or placed on a labeled plastic tray.Shell length and width were recorded, and Velcro strips were glued onto the right or left valve of each mussel, depending on the experimental setup used,for subsequent experiments.During winter,mussels were transferred from the flow•through water tables to 50 L refrigerated tanks in order to keep the animals acclimated to the appropriate temperature.
2.2 Video setup
Experiments that required the use ofan endoscope were set up according to Ward etal.(1991). AnOlympus endoscopefittedwith a 1.7mmdiameter optical insertion tube (OIT)was attached to an optical zoom adapter (maximum magnification -150x, Scholiy Fiberoptic) and a charge-coupled-device colour (Sanyo) or black and white camera (Cohu 6500)(Fig. 2.l).AHi8 video recorder and a monitor were connected to the camera, and light was provided by a fibre optics cold light source.The endoscope andcamera were mounted on a micromanipulator (Fine Science Tools) and metal stand so that the position and angle ofthe endoscopecould be adjusted.
15
M
Figure.2.1. Staticsystem setup andvideotaping equipment used in particleconcentration andtemperature experiments withMylilusedults.(A) outertankcontaining coolantwater andconnectedwith temperature controlunit,(B) inner experimentalchambercontaining seawaterandalgalsuspension,(C) adjustableplasticstand,(D) mussel,(E) endoscope,(F) zoomadaptor,(0) camerahead,(H)micromanipulator,(I)support stand,(J) light source, (K)cameracontrols, (L)Hi 8 recorder,(M)monitor,(N) temperature control unit.
2.3 Particle concentrationand temperature experiments(staticsystem) 2.3./Setup
The majority of the panicle concentrationand temperatureexperimen tswere carriedoutin an aerated static system(Fig.2.1).For each preparation"one musselwas attachedwithVelcroto an adjustableplasticstand that rested inside a IOL container.A seawater/algaemixturewithinthe container was aeratedandmaintainedata constant tempera ture usinga cooling unit (NeslabInstru ments).Each animalwasallowedtograze down a predetermined concentrationoflsochrysissp.(rangingfrom6to 188 cellsIlrl)in the tOLcontainer for 30to 60minutes before video taping was initiated.Control experim ents without musselsdemonstrated thatthealgalcells didnotsinkinthe experimentalchamber.
2.3.2 Markers
To serveasmarkerparticlesduring an experiment,approximately200J.Llof fluorescent yellow plast icchips(2- 10urn diameter) suspendedin seawater wereadded to the experimentalcontainerafter theinsertionof theOITinto the pallial cavity of a mussel.
Fluoresce ntyellow particles showed thehighest contrast againstthe darkbrown background ofalgae-rich mucustravelingalong the gillfilaments.as comparedwithred.
orangeorpink fluorescen t particles.thus allowing foraccuratemucou s strandvelocity measurements (see section2.8).Panicle concentrationversustime plotslatershowedthat the small amountof markeraddedhad anegligible effect on thetotalpanicle concentration.Asanalternative toplasticchips.if a particularly clearand undistorted imagewas obtainedafter initialinsertion oftheOIT,small drops of20or 25 um diameter fluorescent yellowmlcrospberes(Fluoresbrite PlainMicro spheres.Polysciences) suspendedin seawater were added viaapipette directly to theinhalantsipho nto serve as calibration beadsduring subsequent imageanalysis offilaments (see section2.7) .After microsphereshadbeen in the ventral groovefor5·10 minutes. the lessexpensiveplastic chip suspension was pipettedinto the experimentaltank.
17
1.3.3 Videotaping
Inorde rto observethe same regionofthe gill duringallexperiments,the OIT was alwaysinsertedinto theposterior region ofthe shellgape (fig.2.2).This region proved tobe lesssensitive to disturbances from movementsofthe OIT than moreanterior regions of the mantIc.Videotaping was initiated once a suitableimage was obtained and the animalappeared tobefeedingnormally (i.e.wide shell gape with extendedmantle and siphons).No attemptwasmade to prop theshell valvesopen once the musselbegan feeding., and if theanimalwas disturbed enough to close downon the OIT,videotaping wasceaseduntilit relaxed again.Byallowing the musseltoclose when disturbed.the numberofdisturbance times couldbe recordedandtherebyprovideanindication of whether the mussel was feeding nonnallyduringeach experiment.Adjustmentofthe endoscopeusing the micromanipulatorwas minimized inorder to decreasethe possibility ofdisturb inga musselduringan experiment. Ifa musselwas disturbedtoo often(e.g.
greate rthan-10 timesinone hour),that session wasnotincludedinsubsequent analyses.
Repeatedexperimentsusing thesame groupof mussels werecompletedin order to determinethe variance of mucousstrand velocities within and between individuals.Forty•
five panicle concentrationexperimentswerecompleted usingt5differentanimals (Table 2.l ), although many morewere attempted.Failureofan experiment was generally due to intermittent feedingofamussel orits sensitivityto thepresence of theOlT.Ofthe 43 staticsystem experiments,16 were runat temperaturesof4to 6°C,theremainder at 14to 1SoC. These temperature slie within thenatural temperaturerangeofMytttusedulisin Newfound landwaters(0° • 200e )andareeasyto maintaininthelaboratory.To allow for ampleacclimationtime.all mussels weremaintained atthe appropria tetemperatu refor at least three weeksprior to each experiment,althoughtheytypicallyrequireonly 14days to physiologicallyandbiochemically adapttoa changein temperature (Bayneetal.,(976).
Experiment 45 was the only experimentin which algaewere added tothe sus pension while themussel was grazing;inall otherstaticsystem experimentsthe mussel was allowedto graze down thesuspensionwithoutreplacement,mainlybecause of the difficultiesinvolvedinmonitoring theparticleconcentrationand maintaininga suitable
tabial patp
anterior
cten idium surfa ce
inhalant currents
posterior adductor muscle
exhalant c urrent
posterior
OIT
Figure 2.2.Position of Ol'T in the pallialsystem ofMytilus edulisduring filming of the ventralgroove.Small arrows indicatethe movement of particleson the ctcnidial surface.
19
Table2.1:Experimentalconditionsto whichmussels(Mytilusetlulis)were subjected duringpanicleconcentrationand acclimatedtemperatureexperiments in a staticor flow-throughsystem(experiments 3Sand SIwerethe onlytwo perfonnedin a flow-throughsystem).Twocolumns are includedto differentiate concentrations duringthe entireexperimentwith concentrations duringventralgroovevideotaping (there wasnormallya20-45minutefeedingperiodbeforethestan of videotaping).Experiment 23 tookplaceovera longtimeperiodand was videotapedin threepans,thereforethreepanicle concentration ranges,clearanceratesand panicleloads are given,althoughtheyareallderivedfromthe same preparation.
Experiment 45 was theonlystatic systemexperimentin whichalgaewerereplacedwhen30%of the suspension had been removedbythe mussel.Three panicleconcentration ranges,clearance rates and panicleloadsarealsolistedforthis experiment.Exp-experiment#,Mus>animal#,T
=
temperature,N·numberofvelocitymeasurementstaken,Pert.conc.tcr)>panicle concentrationrangeover whichthe clearanceratewascalculated,Pan.cone.(videoj> particle concentration rangeduringvideotaping, Clear.rate>clearancerate,Loadepanicle load,Ssize·relative thickness of mucous strand(arbitraryunits,values1-4), Prevofed=previousfeedinghistoryof eachanimal(I·notpreviouslyfeeding;2:::
previouslyfeeding, see section 2.8, page33for explanation),Exp.lime-duration of experiment.
:::
E.p Dale Mus T N Part. Part. Clear. Load S Prev Elp.'C cone, eenc. rale pan min" size red lime
pan""I" pan""r' mlmin" _106 min
(er) (video)
I Sc:p2l'9S
"
I' '2 65-11 H-15 1".0 ].6 2 I 1202 OcU'95
, "
]] 26- 10 16-10 1K.9 1.0 2 I 110] 0.:112'95 22
"
20 ]0-27 29-27 14." 0.1 I I 100,
5678 Octl]Oct2J'95OcU4'IJSOct25'1J5Novl'9S'9S H221899" " " "
66]8H9 91·36117125-1'11-6-6] 63·361~-I972-1511-7 116.095,082.]82.1 U5.1U1.0 212I IIII 11201200590"
50 IU-109 119-10 1 29.5 ].2,
I 1059 Nov6'9S 22 H 25 170- I) 3 l00·I3~ 32.6 '.6
,
I 80IU l'ebT 96 2' I' 13 125-97 107·100 22.0 2.]
,
I 100II Feb lS'96 9 I' 29 136-73 105-73 62.0 5.6
,
I 90Table 2.1(cant.)
Exp Date Mus T N Part. Part. Clear. Load S Prev Exp
-c cone. cone. rate partmin-I size red time
partj.lrl part}lr1 mlmin"· .106 min.
(er) (video)
12 Feb16'96 22 14 25 152-47 95-47 130.0 8.7 3 I 90
13 Feb20'96 24 14 18 75-48 61-48 50.0 2.7 2 I 90
14 Mar5'96 18 14 28 94-22 46-22 127.0 4.3 2 I 120
15 Mar6'96 4 14 II 127-93 126-111 38.0 4.5 3 I 90
~ 16 Mar14'96 4 14 20 81-27 51-30 124.0 5.0 2 I 90
17 Mar15'96 16 14 20 79-27 47-28 122.0 4.4 2 I 90
18 Mar19'96 16 14 12 232- 165 I70-I ·n 62.1 9.8 2 I 90
19 Mar20'96 24 14 17 253- 140 188-1H 64.8 11.0 3 I 90
20 Apr3'96 30 6 II 66-54 59-55 25.4 1.5 I I 80
21 Apr4'96 31 6 21 73-37 53-37 72.3 3.3 3 I 105
22 Jun.26'96 9 6 14 120-102 109-10 3 13.6 1.4 2 I 90
23 Jull l'96 9 6 28 136-85 1l1·90 -t·U 4.5 3 2 350
II 85-45 63-52 6·U 3.6
22 45-13 33-13 88.3 2.3
24 }u1l4'96 9 6 28 123·79 105-82 49.6 4.7 2 I 90
25 Jull5'% 16 6 II 124-77 86-66 81.8 6.4 3 2 80
26 Ju1l8'96 9 6 21 103-65 84-68 52.8 3.9 3 I 90
27 AugS'96 9 6 13 83..6] 73-63 37.7 2.6 2 2 80
28 Aug6'96 16 6 18 120-78 107-79 74.6 6.9 3 2 80
29 Aug6'96 9 6 19 22-13 19·14 84.5 1.4 2 2 80
30 Aug7'% 16 6 20 17-9 13-9 96.7 l.l 2 2 80
~o o ooo
z
.... 'lICI ....r--
~~~N~_~ ~~~N "" NNN ~ ~
~~ N N 2 N N~ O ~~~O ~~ ~ C __ON ~ OO O_~ ~O ~ O ~ _ O
~~§~~ g~~ ~ ~ ~~~~~ ~~
~~ ~~N'lICI N ~~ ~ ~ ~ ~ ~ ~ ~ N N _ N _ ...._ N ...._ .... - ....
.,."'.,. .,. .,. .,..,. .,. .,.
--- -- -
-
.... ....
.,....,....... .... .... .... ...
.,.imageonthe video monitoratthe sametime.Many of theanimals remained healthyand seemed toadapt to the presence ofthe OIT overlongperiods oftime.Someofthese mussels were used for many monthsin different experiment sinaddition tomussels broughtinat a later date.
1.3.4Clearance rates
Twentymlsamplesof seawaterwere taken at-15minuteintervalsbeginning when ananima)wasfirstplacedin theexperimental containerand continuinguntilthe experiment was terminated(the duration of most of theseexperimentsranged from1.5to 2 hours,dependingon the behaviourofthe mussel ).Anyfeces or pseudo fecesproduced bya musselduringanexperiment were removed witha pipette in orderto prevent resuspensionofbiodeposits.Samples taken fromthe experimentalchamberwere counted (sizerange 3to64 um diameter ) with aMuhisizer (Coulter Elect ro nics)fittedwith a 100)UJlorificetube.
Todetermine theclearance rate of eachanimal,the slope of the regression of natural logarithm ofpanicleconcentration versus time wasmultipliedbythevolumeof the experimentalcontainer. If theclearance ratechangedduringthedurationofan experime nt.the longest linearsectionoftheregressionline wasused.to calculatethe clearance ratefor thatanimal(oneclearancerate perexperimentwas necessaryfor clarity andforstatisticalpurpo ses).The particle concentrationrange overwhich eachclearance ratewas determi ned isrecorded in Table 2.1.
2.4 Time effectexperiments 2.4./Setup
Aflew-throughsystemwasusedtodeterminewhetherthere werelongterm time effect sonthevelocityof mucus inthe ventral groove(Fig.2.3).Withthissystem,the ambient part icleconcentrationcould be maintained for many hoursand theclearancerate of the musseland particle velocitieson the ventralgroove record ed.A 1.3Lexperimental
23
,
/-
Il l!' · U ..l
...
...now..04 ~ < 0:0 .
.;
~Y'\°o
<A)
Figure 2.3.Flow-through syste msetup used intemperature. time and part iclequality experimentswithMytiiusedulis.SeeFig.2.1forendoscope andvideo tapingequipment.
(A) head ertank,(8)fow restrictor,(C)algalculture,(0)peristaltic pump,(E) experimentalchamber,(F)musse l (G) outertank.containingcoolantand connec tedto temperaturecontrol unit.(H)siphoncup,(I)standpipe.
containerwas connected toa20L header tank continuouslyfilled with aerated seawater and to a separate10L carboy containing fluorescentmarkerpanicles in a dense culture of Isochrysis.The flow ofseawaterfrom the header tank to the experimental tank was controlledwith differentsized flow restrietors (plastic plugswithsmall holesdrilled through the center),and the aeratedalgalculturewas added to thesystem atspecific rates using a peristaltic pump (MinipulsJ.Mandel Scientific).The outflowfrom the experimental chamber consisted ofa siphon hose that was connected to a cup containing seawater,anda standpipethatmaintainedthe water levelin the experimental chamber.
Flow velocities through the system were calculatedusingthe timetaken to till a graduated cylinderwith seawater from the outflow tube.
After using the apparatus a few times. itwasnoticedthat the mussels adapted to the presenceof the marker panicles after beingexposed to them for2 toJhours.The animals were able to separate the marker panicles from the algae-rich mucous strings traveling within the ventral grooveand relocatethem to the outer edges ofthegroove wherethe particles were susceptibleto dislodgmentby feedingcurrents.thusmaking it difficultorimpossible to determine the muco usstring velocities.In order to counteract this effect, two carboys were used:one with marker particlesadded,and one without.For timeperiods in which the ventralgroove was not being videotaped during an experiment (see section2.4.2fordetails),the carboy whichcontained only algae was connected to the flew-throughsystem,whereasthecarboy which also containedmarkerpanicleswas connected tothe systemonlyduringvideo taping periods.This procedure preventedthe animals from adapti ng to the markersand creating this 'marker particleshunt'.The experimental chamber was immersed ina refrigeratedwater bath,andan aerator was placed next to the seawaterinlet to equilibrate<hlevels andresuspendthealgae before the waterreachedthemussel onthe other side of a baffle (Fig.2.3).
2..1.2 Videotaping
Each mussel was proppedat an angleinthe experimental chamber usinga small plastic stand to allow for insertionof the endoscope atan angleofapproximately 45°
25
(Velcro wasnotused duringflow-through experiments).Animals were positionedso that the inhalantsiphon facedtheseawaterinlet(i.e.facingthe opposite directionfrom animals in the staticsystemexperiments).Once aflowrate was achievedwhichprevented each musselfromremovingmorethan300!cl of the suspendedcells (flowraterangedfrom150 to 429ml min"),and thedesiredparticle concentrationwas obtained,a15to20minute adjustment periodfolloweduntilthemussel's shellgape permittedentryof the OfTand videotapingwasinitiated.
Ninetimeeffect experiments, each lasting from-3 to 48 hours,were successfully completed (Table2.2).During each experiment,the ventralgroove was videotaped for 15 to 20minutes,after which themussel was left to feed in theabsence of the OITfor-45 minutesto 2hoursbeforeOIT reinsertion. For someof thevery longexperiments, musselswereleft to feedundisturbedin the experimental setupfor 5to14 hour sbetween videotapings. In ordertodetermineifthere were any feeding history effects,the mussel in experiment40was allowed tofeedona predeterminedparticleconcentrationfor 23 hours priortovideotaping, whereas animalsinexperiments 41and 42 werekeptwithoutfood for2 dayspriorto videotaping.
2.4.3Clearance rates
Inordertodeterminethe clearancerate of each mussel,twentyml samplesof seawaterwere taken fromtheinflow,siphonandoutflow locationsduring each experiment.According toHildrethand Crisp(1976), the calculatio nofthe clearance rate for a bivalvein a flow-through system shouldtake into account the dilution effect of inflowingwaterasit enterstheexperimentaltank.To do this,onemust determine the particleconcentrationin the area immediatelysurrounding the mussel, which in this case is the siphonarea. and apply the equation:
CR=C,-C.'FR
C, Equatio n2.1
Table2.2:Experimentalconditions towhichmussels(My/illls edlllis)were subjectedduringtimeeffectexperiments in a flow- throughsystem.SeeTable2.1legendfor descriptionsofcolumn headings.Particleconcentrationand clearancerate valuesare meanswithstandarddeviationsin brackets.Themusselin experiment 40 hadbeenfeeding for 23hoursbeforefilming,and experiments40,41 and 42 are partitionedsinceclearancerates or particleconcentrations changedslightlyduringthe experiments.
Esp Dale Mus T N Particle Cleara nce Load S Pre\' Esp.
·e
cone. rate pan mrl.Iz_
red timepart f.ll'l mlmin,l ·10' hours
(std dev] (.tddev) to>
...
36 Scp-l·96 9 \5 4l 162(6.22) 6'.0(11.\) \1.2 3 2 5.831 Scp lO'96
•
\4 IS '6.3(3.85) 58.5(I9 A) 5.6 2 2 5.838 Jan6'97 J2 5 13 119(1 1.8) 67.7(18.S) 8\ 2 \ 8
3. Jao7'97 30 4 21 11.5(1.83) 76.7(34.5) 0.88 2 \ 8
40 ScpI9"96 24 13 5' -12.6(·UK ) I22.7{20.l-l) 5.2 2 2 48
,.0
J-I.II.(l.66) IU O O.8) 5.2H Octl'Yll
,
13 73 lU(3.21) 110(28.0) 3.9 2 1 2843 27.5(3.21) 55.3(16. 1) I.l
42 No\'6'96 30 8.5 20 122(6A') 51.1(l 2A) 6.2 2 I
\,
55 3U('.85) U
"
FebI2·" 33 4 2' U7(0.3I) 5'.' (36.1) 0.27 2 I 3.650 Feb17·97 II 3.5 l4 27.4( 1.63) 54.1(16.0) 1.5 2 I 2.8
whereCR=clearance rate(mJmin"),
c.-=
inflowconcentration(particlesur '),C.=outflow concentration(particles~rl),
c. =
concentrationinthe siphonarea (particles~rl), and FR= flow rate (mJmin").Intheflow-throughexperiments carriedoutinthisstudy, thepanicle concentration at eachmussel'ssiphonwasvery similarto theparticle concentration measuredintheoutflowingseawater,andsincetheoutflowconcentrations tendedto be lessvariablethansiphon concentrations,thefollowing modified equation was used foraUclearance ratecalculations:CR=C,-C.'FR C.
2.5 Particle quality experiments
Equation2.2
2.5./ Setup
Two particlequality experiments weredone usingtheflew-throughsystem describedinsection2.4(Table2.3).Animalswere fed algalsuspensionsfor 3hours and then fed aninorganicsedimentsuspensionfor theremainder of eachexperiment.In experiment 46,thesedimentsuspensionwasintroducedintothe experimental chamber with a peristalticpump afterdisconnectingthe supply ofalgae.butmuch ofthe sediment settled within thetubing before reachingthemussel.evenwithhighpumpingrates.To avoid sedimentsettling in the tubing in thesecond experiment,sedimentwas suspendedin seawaterin large temperaturecontrolled tanksand wastransferreddirectlyto the experimental headertankwhen a particlequalitychangewasrequired.Thissetupresulted inashorter routebetweenthesedimentsuspension tankand the experimentalchamber and alsopermittedvigorousstirring ofthesediment withinthe headertank. which intum helpedtokeep the sediment suspendedforlonger periods.By runningthis systemwithout a musselpresent,asettlingrate of6.S%of the total sedimentparticles traveling tothe experimentalchamberwasdetermined(additionalstirringwithinthe experimentalchamber
Table2.3:Experimentalconditionsto whichmussels(Mytilus ed,,/is)weresubjected duringparticlequality experiments usinga flow-through system.SeeTable2.1legendfordescriptionsof columnheadings.Panicleconcentration andclearancerate values are means withstandard deviationsfollowingin brackets.Experiment 48is partitionedsinceclearancerateschanged slightly.
Exp Date Mus T N Particle Clearance Load S Prey Esp.
N 'C cone. rete pan mln'' size rod lime
'"
pa(SId dev]n~r' rnl(SId uev)min" '106 bcurs"
feb]'97 JJ U 2. 7.11(0.8.5) 6·U 4(J0.74) 0.44,.
MllrllJ'97 30 2.5 223. 8.42(0.74) 1S1(50.8)113(27.3) 1.3tu nl
was not possible sincethe resulting vibrations disturbedthe feeding mussel andmade videotaping impossible).Clearance rate measurements were thereforecorrected for paniclesettlement.
2.5.2 Sediment
Sediment was taken fromthe upper 10 cm of a core sample from250 m depth in Conception Bay, Newfoundland,and driedat60DC.Thesediment was thencombusted in a muffle furnace at 420DCfor 6 hours, cooledand groundwith a mortar and pestlebefore beingresuspendedin seawater.The size frequency distributionof thefinalsuspensionwas similar tothat oflsochrysis(i.e.modenear 3J!m diameter).
2.6 Acute temperature change experiments
Two temperatureexperimentswere doneusingthe flow-through system inwhich the animalsused wereacclimatedto 4 - SDCand then exposedto 14 - ISDCduringfeeding (Table2.4).After 2 hours offeeding at thelowertemperature (ambientseawa ter temperature).seawate rintheexperimental headertank was replacedwith heatedwater.It took-20 minutes forthe temperaturechangetobe completedin theexperimental chamber, and althoughthe mussels'mantle gape decreasedduringthechange,both musselscontinued to feed normally.Particleconcentrat ion measurements necessaryfor clearance rate calculationswererecordedcontinuously duringthe entireexperiment.
2.7 Gill filament measurements
Image analysis software (Mocha,.JandelScientific)was used totake filament width measurementsdirectlyfromthe videotapesof ventralgroovesmadeduringexperimentsin whichfluorescentmicrosphereswere used.Bycapturingavideoframewith a non•
distortedimage of a microsphereadjacent to theventralgroove filaments,the microsphere could be used as a referencepointwithwhichto measure the filament widths.Since the image analysismethodof measuring filaments isa noveltechnique.these on-screen
Table2.4:Experimentalconditionsto whichmussels(Mylilusedulis)weresubjected during accutetemperatureexperiments using aflow-throug hsystem.SeeTable2.1legend fordescriptionsofcolumnheadings.Particle concentrationandclearance ratevaluesaremeanswithstandarddeviations followinginbrackets.Both experimentsarepartitionedsinceclearancerates changedwhenthe ambient temperature was altered.
Exp Date Mus T N Particle Clearance Load S Prev Eap.
-c
cone, rate panmin" size fed timepart~rl mlmin" _106 hours
(sIddev] (sIddev]
.3 Jan2S'97 30
•
15 ..97 29.8(3.8.) H112(.1.3).0(2H) 3.31.0••
Jan2"97 32 5 23 11.8(0.77) 1.7.2(30.• 5) 1.7I. 38 1J).7(.1.••)
measurementsoffilaments fromlivingmusselswere later compared with post-mo rtem measurements offilamentsusing a dissect ing microscopeandcalibratedocular micrometer.
2.8 Videoanalysisand variables
Freez e-frameanalysesofthe videorecordings wereusedto determine the velocity of each mucous strand travelingwithin theventral groove ofademibranch.When a distinctfluorescentmarkerpanicle that wasfinnlylodged withinamucous strand(i.e.
showing consistent movementwithintheventral grooveand not affected bymantle currents) was observed, thevelocityof the particlewas calculatedfrom the number of framesrequired for it to travelacross thetips of a known numberof filaments.Since the videospeedwas 30frames per second(NTSC standard),and filamentsize waspreviously determined using image analysis.thevelocityofthepanicleinwas easily calculated.
Particlevelocity measurements were takenfrom eachvideotape at-2 minuteintervals.
when possible,andthe corresponding'time spent feeding'was calculatedusingvideo time and real time comparisonswithpauses takeninto consideration.
Because panicle velocity measurements weretaken atsmaller intervalsthanthe particl econcentrationmeasurements,it wasnecessary to calculate the panicle concentrationvalues at all timesusing plotsofInpanicleconcentrationvs time.The following equation wasused todetermine the panicle concentratio nsthatcorres pondedto eachvideomeas urement inthe static systemexperime nts:
Equation2.3
whereCt=panicle concentratio n(panicles
""r
l)at time t (minutes),C,«panicle concentration(panicles""r
l)attime O.and &"'" slope of the Inconcentrationvs timeplot.This equationwasnot required for theflow-thro ughexperimentssince the particle concentratio nwas kept constant.The part icle load (paniclesmin"),or total amountof foodmaterialmoving on the gillperunit time.correspondingto each velocity
measurementwasdeterminedbymultiplying theclearancerate(m1min" )byparticle concentra tion (particles llr')·I OOO.
Since thesize ofamucous strandwithin the ventralgroov e changedbetween.and sometimes within, experiments,the relative size ofeach strand wasalso recorded.Four categorieswere usedto indicatestrand thickness:1=thin, 2=medium,3=thick,and 4=
extended (Figs.2.4 •2.7).Category4describessituations wheretheoutside edges of the mucous strand wereextendingbeyondtheventral groove andobscuringthefilaments.It was notpossible to measure strandwidth directly using image anaJysis becausesections of the mucousstrands were oftenhidden bythecontraaionofthe widefilamenttips at the ventralgroove.Sincefood particlesareentrainedwithinmucuswhiletravelingin the ventralgroove.the terms'mucusstrandorstringvelocity', 'particlevelocity' and 'transport ratewithin the ventralgroove'are synonymous and are usedinterchangeably throughoutthis thesis.
Someofthe musselswereobservedtobefeeding intheirholdingtanksbeforean experiment,therefo re the previousfeedingconditionof each animal was recordedas either:1- musselhadnot been feedingbeforethestan ofthe experiment,or 2-mussel had been feedingprior totheexperiment.Also,since experimentswere perfonned over a periodof twenty consecutivemonths. itwasnecessary to test whethertherewasa seasonalormonth effecton paniclevelocity,therefore each experimentwas assignedto the appropriate month (I.20) andthevariable'month'included insubsequentanalyses.
Table2.5liststhe response variables thatwere includedin thestatisticalanalyses discussedin section2.9.
2.9Statisticalanalyses
Sincenearlyallof theexperimentsinthis studywere carried out using the same basicprotocolduringthefirst 120minutes(i.e.individual musselsfedwithIsocbrysissp.
atvariable panicle concentrati onsand temperatures),itwas possible to poolthe acute temperature.acclimatedtemperature,particle quality,time effectand panicle
33
Figure2.4.Ventral groove ofMytilus eduliscontaining a thinmucous strand.
Figure 2.5. Ventralgroove ofMyrilus eduliscontaining amedium mucous strand withone markerparticle visible.
35
Figure 2.6.Ventral groove ofMytilus eduliscontaining a thick mucous strand and twomicrospheres.
Figure2.7. Ventral groove ofMytiluseduliswithamucous strand extending over the filament tips.
37
Table2.5:Variables considered inthestatisticalanalysesofpooledparticlevelocity data inMytilusedulis.
Variable Abbreviation
lag for particle velocity lag particle concentration cone
clearance rate cr
particleload load
time time
temperature temp
mussel1# mus
previousfeedinghistory prev
month month
Typeofvariab le continuous continuous continuous continuous continuous categorical categorical categorical continuous
concentration experiments in order to determine an overall particle velocity model (see Tables 2.1- 2.4 for details of individual experiments). Restricting this pooled data set to time spent feeding less than 120 minutes kept time effects to a minimum;long-term time effects are addressed separately. Itwas assumed that using a static or flow-through system in different experiments would not affect the basic feeding physiology of the mussels
Data from all experiments, with the exception of experiment 40,were included in the pooled statistical analyses.Experiment 40 was a long-term experiment in which a mussel was allowed to graze on an algal suspension for 23 hours before videotaping. Since the standard procedure during the first 120 minutes of feeding was modified, this experiment was eliminated from the final model. The remaining data were divided into one of two temperature groups: one group consisting of32 experiments conducted in a temperature range of8.Sto 15°C(mean 14°C),and the second group consisting of28 experiments conducted in 3 to 6 °C seawater (mean 4.5°C).
A general linear model (GLM) was used to test the effects of the numerous categorical and continuous explanatory variables (Table 2.5) and their interactions on the mucus velocities in the ventral groove.Because all the data collected were derived from time series, significant autocorrelation of the residuals (detected by the Durbin-Watson test or the Arima procedure in the SAS statistical package) resulted when applyingany model to the pooled data.Since association ofresiduals violates one of the major assumptions of the GLM, various techniques were employed to remove the autocorrelation from the analyses. The most successful procedure for this particular data set involved creating a new variable consisting of lagged values of the responsevariable, YI_!, which resulted in a model of the type:
Equation 2.4
39
(Myers,1990).By includingthis new variable in the GLM,the associationof the residuals was taken intoaccount andtherefore corrected for in the analyses.
Separate analyses of individual experimentsusingregression analysis,ANOVA and correlationanalysiswere also completed in additionto the overall particlevelocitymodel.
Many ofthese models also incorporate a'lag'term to eliminateautocorrelation of the errorterms.
CHAPTER 3 - RESULTS
3.I Gill filament measurements
Filament width measurements at theventral margins oflivemussels determined usingimage analysisweresmaller thanthose obtainedfrom freshly dissected animalsusing a stereoscope.evenafteranunderestimationfactorwas appliedtothe imageanalysis measurements [fable 3.1).This factorof-7%represents the differencebetween measu rements ofthe diameters ofcalibrationparticles using image analysissoftware and theactual size ofthe calibration particlesprovidedbyPolysciences (25 urn).The underestimation factorwas appliedtothe grand mean ofthe image analysisfilamentwidth measurements togive acorrected mean of8J.8um, whichwas9.8% lowerthan the grand mean for stereoscopemeasurements(92.9um).
3.2Mucousstrand size
Particle velocity decreased with increasingmucous strandsize(Fig.3.J),and as strandsizeincreased so didthepanicle loadon the gill(Fig.3.2).Since strandsize was correlatedwithpanicleload.itwas decided that onlythe quantitativeand non-subjective variable 'load'wouldbe used inthe finalstatisticalmodel (section3.3).
41
Table3.1:Summaryoffilamentwidthmeasurements atthe ventralgrooveofMytitus edulisusingbothimageanalysisandastereoscopewithcalibratedmicrometer.Mussel= animal#.N"numberoffilamentmeasurementspermussel.
Image analysis:
Mussd N Mea. Standard
1"",1 devlatien
18 45 68.93 4.16
16 52 81.68 7.16
9 221 82.48 14.24
22 60 80.28 6.27
24 31 78.31 4.7
Grand 78.34 )1m •7%- 5.48 mean:
Corrected: 5.48+78.34=83.&)1m
Stereoscope:
Mussel N Mean Standard
(J,Un) deviation
3 8 101.8 0
5 16 82.9 10.4
12 IS 93.3 11.4
24 5 106 4.18
22 5 105 5.0
32 5 80 3.6
30 5 83 4.5
31 5 93 4.5
9 5 105 5.0
16 5 86 4.2
33 5 86 4.2
Grand 92.9,.m
700
600 1=thin
2=medium
500 3=thick
4=extended
~ i
400~E 300
" g
~ 200~
100
N=202 N=713 N=293
strand
N= 117
Figure3.1. The relationshipbetweenparticlevelocityand mucous strandsizeinthe ventralgrooveofMytiflls edulis.The mean andstandard deviation for particlevelocity in eachstrand sizecategoryare plotted,andsamplenumbersare included Strand sizeand velocity are significantlycorrelated
(r'-
0.44,n- 1325, P<0.0001)43
1.4e+7 1.20+7 1.0e+7 r~ 8.0.-.6
...
~ 6.0.-.6 'e!
~
4.De-.62.00-s
C.De+O N=202
f
N= 713
strand
N=117
Figure3.2.The relationshipbetweenpan icleloadand mucous strand sizeinthe ventral grooveofMyli/us eduits.The meanand standarddeviation forparticleloads in each strand sizecategory are plotted,and sample numbers areincluded.Strandsizeand panicle load are significantly correlated (r2=0.53,n=1325,P<O.OOOI)
