+ All Categories
Home > Documents > The effect of selection treatments on Mytilus edulis , modifications of genetic and physiological...

The effect of selection treatments on Mytilus edulis , modifications of genetic and physiological...

Date post: 22-Jan-2023
Category:
Upload: uacm
View: 0 times
Download: 0 times
Share this document with a friend
12
Mar Biol (2008) 153:1141–1152 DOI 10.1007/s00227-007-0885-8 123 RESEARCH ARTICLE The eVect of selection treatments on Mytilus edulis, modiWcations of genetic and physiological characteristics Neil LeBlanc · Réjean Tremblay · JeV Davidson · Thomas Landry · Mary McNiven Received: 4 December 2006 / Accepted: 7 December 2007 / Published online: 10 January 2008 © Springer-Verlag 2007 Abstract This study examined the eVects of two selection treatments (elevated water temperature and air exposure) on the genetic and physiological characteristics of the juve- nile marine mussel, Mytilus edulis (<10 mm). Genetic eVects were measured on Wve allozymes and Wtness assessed using physiological tests to estimate energy bal- ance (scope for growth) as well as size, growth and sur- vival. The in vitro treatments resulted in 48% mortality from an air exposure of 11 h at 27°C and 76% mortality from a 6-h exposure to 33°C water. Survivors (n = 1,152) of each treatment along with controls (n = 2,304) were measured and randomly placed in compartmentalized cages. Mussels were deployed to three bays in Prince Edward Island, Canada and monitored over a 10-month period. Initially, both of the treatments had an eVect on mussel size and increased the heterozygosity of the surviving mussels. Physiological analyses after 3 months in the Weld showed that the two treated mussels showed lower meta- bolic rate that the control group. After 10 months in the Weld, the treated mussels were larger and had lower mortal- ity than the untreated control mussels. Unexplained envi- ronmental interaction in each of the bays had an eVect on allelic frequencies and heterozygosity. Overall, the results demonstrate that simple husbandry techniques can be used to increase the productivity of mussel seed and heterozy- gosity measures can be used to assess Wtness. However, more Weld data is needed to determine the consistency of the increased productivity and if the increased productivity justiWes the costs of a selective treatment. Furthermore, because the level of heterozygosity in juvenile mussel pop- ulations can vary considerably, both spatially and tempo- rally, it may be eVective as a warning of future natural mass mortality when overall heterozygosity levels are found to be low. Introduction A positive correlation between multi-locus heterozygosity (MLH) and various Wtness characteristics has been docu- mented for several species including Mytilus edulis (Mitton 1994). Characteristics of Mytilus species that have been examined include growth, survival and metabolic rate (Koehn and GaVney 1984; Bayne and Hawkins 1997; Myrand et al. 2002). Although considerable research has been conducted on the link between MLH and Wtness and many explicative hypotheses have been developed, there is still no single explanation of the relationship. The two func- tional theories used in explaining the relationship are over- dominance and associative overdominance. In the former, scored loci are directly involved in the Wtness relationship, Communicated by M. Kühl. N. LeBlanc · J. Davidson · M. McNiven Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, PE, Canada C1A 4P3 R. Tremblay Institut des sciences de la mer, Université du Québec à Rimouski, 310 allée des Ursulines, Rimouski, QC, Canada G5L 3A1 T. Landry Mollusc Section, Department of Fisheries and Oceans, Science Branch, Gulf Fisheries Centre, P.O. Box 5030, Moncton, NB, Canada E1C 9B6 N. LeBlanc (&) Joint Research and Development Division, Departments of Virology, The National Veterinary Institute (SVA) and The Swedish University of Agricultural Sciences (SLU), Ulls väg 2B, 751 89 Uppsala, Sweden e-mail: [email protected]
Transcript

Mar Biol (2008) 153:1141–1152

DOI 10.1007/s00227-007-0885-8

RESEARCH ARTICLE

The eVect of selection treatments on Mytilus edulis, modiWcations of genetic and physiological characteristics

Neil LeBlanc · Réjean Tremblay · JeV Davidson · Thomas Landry · Mary McNiven

Received: 4 December 2006 / Accepted: 7 December 2007 / Published online: 10 January 2008© Springer-Verlag 2007

Abstract This study examined the eVects of two selectiontreatments (elevated water temperature and air exposure)on the genetic and physiological characteristics of the juve-nile marine mussel, Mytilus edulis (<10 mm). GeneticeVects were measured on Wve allozymes and Wtnessassessed using physiological tests to estimate energy bal-ance (scope for growth) as well as size, growth and sur-vival. The in vitro treatments resulted in 48% mortalityfrom an air exposure of 11 h at 27°C and 76% mortalityfrom a 6-h exposure to 33°C water. Survivors (n = 1,152)of each treatment along with controls (n = 2,304) weremeasured and randomly placed in compartmentalizedcages. Mussels were deployed to three bays in PrinceEdward Island, Canada and monitored over a 10-monthperiod. Initially, both of the treatments had an eVect onmussel size and increased the heterozygosity of the surviving

mussels. Physiological analyses after 3 months in the Weldshowed that the two treated mussels showed lower meta-bolic rate that the control group. After 10 months in theWeld, the treated mussels were larger and had lower mortal-ity than the untreated control mussels. Unexplained envi-ronmental interaction in each of the bays had an eVect onallelic frequencies and heterozygosity. Overall, the resultsdemonstrate that simple husbandry techniques can be usedto increase the productivity of mussel seed and heterozy-gosity measures can be used to assess Wtness. However,more Weld data is needed to determine the consistency ofthe increased productivity and if the increased productivityjustiWes the costs of a selective treatment. Furthermore,because the level of heterozygosity in juvenile mussel pop-ulations can vary considerably, both spatially and tempo-rally, it may be eVective as a warning of future natural massmortality when overall heterozygosity levels are found tobe low.

Introduction

A positive correlation between multi-locus heterozygosity(MLH) and various Wtness characteristics has been docu-mented for several species including Mytilus edulis (Mitton1994). Characteristics of Mytilus species that have beenexamined include growth, survival and metabolic rate(Koehn and GaVney 1984; Bayne and Hawkins 1997;Myrand et al. 2002). Although considerable research hasbeen conducted on the link between MLH and Wtness andmany explicative hypotheses have been developed, there isstill no single explanation of the relationship. The two func-tional theories used in explaining the relationship are over-dominance and associative overdominance. In the former,scored loci are directly involved in the Wtness relationship,

Communicated by M. Kühl.

N. LeBlanc · J. Davidson · M. McNivenDepartment of Health Management, Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, PE, Canada C1A 4P3

R. TremblayInstitut des sciences de la mer, Université du Québec à Rimouski, 310 allée des Ursulines, Rimouski, QC, Canada G5L 3A1

T. LandryMollusc Section, Department of Fisheries and Oceans, Science Branch, Gulf Fisheries Centre, P.O. Box 5030, Moncton, NB, Canada E1C 9B6

N. LeBlanc (&)Joint Research and Development Division, Departments of Virology, The National Veterinary Institute (SVA) and The Swedish University of Agricultural Sciences (SLU), Ulls väg 2B, 751 89 Uppsala, Swedene-mail: [email protected]

123

1142 Mar Biol (2008) 153:1141–1152

while in the latter the scored loci act as neutral markersfor either balanced polymorphisms or deleterious alleles(Gosling 1992). Two explanations for associative overdomi-nance have been postulated. The Wrst theory of associativeoverdominance states that the association is caused by link-age disequilibrium, i.e. markers are physically linked tonearby Wtness genes. The second assumes that genetic vari-ability at the marker loci represent variability across thegenome (Tsitrone et al. 2001).

Several studies showed that heterozygote Wtness correla-tion (HFC) increased or become apparent when animals aresubmitted to stressful conditions, like elevated water tem-perature, anaerobic exposure, copper exposure, starvationand crowding (Gentili and Beaumont 1988; Scott andKoehn 1990; Beaumont and Toro 1996; Tremblay et al.1998; Myrand et al. 2002; Pluess and Stöcklin 2004). Thesestudies found that heterozygote individuals had an advan-tage in surviving under stressful conditions. MLH andmaintenance metabolism (basic metabolic rate needed toremain alive) in M. edulis have been found to be negativelycorrelated (Tremblay et al. 1998; Myrand et al. 2002), andincrease when stressful conditions were present (Gentiliand Beaumont 1988). These Wndings suggest that exposingmussels to levels of lethal stress should remove homozyg-otes or conversely heterozygotes should have an advantagein surviving.

In our experiment, we were interested in validatingwhether controlled lethal stressful conditions in M. eduliscould (1) increase the level of heterozygotes in mussel pop-ulations, (2) increase the physiological performance ofmussels and (3) increase the growth and survival in diVer-ent Weld conditions (diVerent bays).

Materials and methods

Treatments and Weld monitoring

Mussels used for the physiological and genetic measure-ments came from the experimental set-up presented inLeBlanc et al. (2005). BrieXy, approximately 30,000 indi-vidual spat were manually collected from suspended oystercages in St Peter’s Bay (46°23.231N, 62°41.599W) inPrince Edward Island (P.E.I., southern Gulf of St Law-rence, Canada) and transported to a laboratory facility.Experiments were conducted to expose mussels to the air orelevated water temperature. In keeping with other litera-ture, a minimum level of mortality of 50% was targeted(Wallis 1975; Cotter et al. 1982; Tremblay et al. 1998), forthe purpose of detecting any eVect of treatments. Approxi-mately 5,000 individuals were exposed to ambient air at anaverage temperature of 27.2°C § 0.5°C and average rela-tive humidity of 55.6 § 10% and stopped after 11 h when

mortality observed were 48% (LeBlanc et al. 2005). Forelevated water treatment, approximately 8,000 individualswere exposed to a constant water temperature of 32.6°Cand salinity of 27.6 ppt and stopped after 6 h when mortal-ity was 76% (LeBlanc et al. 2005).

Survivors from the air and elevated water temperatureexposures, as well as untreated control mussels, wereplaced in cages for individual monitoring in St Peter’s,New London and Tracadie Bays. All three bays are usedextensively for mussel aquaculture. Although the bays aresimilar in depth, temperature and salinity (Fisheries andAquaculture Division 2001; Crane 2003), three Weld siteswere used to detect any interaction between the treatmentsand the environment. All the cages were attached to mussellong lines in the bays at a water depth of 1.8–2.1 m(LeBlanc et al. 2005). After 10 months, mussels in each ofthe bays were measured (length; umbo to furthest point onthe posterior edge) and mortality was counted. All of thedata from mussels sampled for genetic analysis (n = 231)are included in the dataset from LeBlanc et al. (2005)(n = 394).

Allozymes analysis

The initial sample (time 0 after treatment) was comprisedof 80 mussels each randomly selected from survivors of theair exposure, elevated water temperature selection treat-ments and control specimens. Control mussels had beenkept in an upweller at ambient temperature until sampling.Finally, random samples of the air exposure, elevated watertemperature treatments and control (n = 30 per treatmentper bay) were taken after 10 months in Weld conditions ateach three bays. All samples were stored at ¡80°C untilanalysed. Genetic markers chosen were polymorphic allo-zymes implicated in metabolic functions and already dem-onstrated to be related to Wtness parameters in M. edulis(Myrand et al. 2002).

Allozyme analysis was carried out on horizontal acetateplate using the Hebert and Beaton (1989) method, asdescribed in Tremblay et al. (1998). The polymorphicenzymes studied were mannose phosphate isomerase(MPI*, EC 5.3.1.8), phosphoglucomutase (PGM*, EC2.7.5.1), octopine dehydrogenase (ODH*, EC 1.5.1.11), glu-cose phosphate isomerase (GPI*, EC 5.3.1.9) and leucineaminopeptidase (LAP*, EC 3.4.11). A standard of all knownalleles was prepared by mixing homogenates of individualsof diVerent genotypes. The standard was used on each gelfor comparison to aid in exact allele identiWcation.

Physiological measures

After exposition to stressful treatment, physiological testswere performed on 47 mussels from the control group, 28

123

Mar Biol (2008) 153:1141–1152 1143

from the air exposure group and 32 from the elevated tem-perature stress group to calculate scope for growth, anenergy balance equation used to calculate the energy avail-able for growth and reproduction (Widdows and Johnson1988). In order to calculate scope for growth, the physio-logical measures for oxygen consumption, absorptioneYciency and clearance rate were performed. Mussels weresent the same day (in insulated iceboxes with sea-ice) forthe physiological testing at the Centre Aquacole Marin deGrande-Rivière (Québec, Canada). Upon arrival, musselswere placed separately in individually numbered plasticmesh cages (vexar, mesh-size 9 mm) and kept in Xowthrough 1,000 l tanks. The individual cages allowed thetransfer of mussels to physiological chambers withoutsevering their byssus (Tremblay et al. 1998; Myrand et al.2002). Mussels were held in untreated and unWltered sea-water (similar to P.E.I. Weld conditions) for 1 month toallow for acclimation of laboratory conditions. After accli-mation (temperature = 5°C, salinity = 30 ppt), clearancerate, oxygen consumption and absorption eYciency weremeasured to estimate the scope for growth of each mussel.With excretion representing less than 5% of the total energybudget in mussels (Tremblay et al. 1998; Bayne et al. 1999;Honkoop et al. 2003), it was not considered in this study.Each day, a total of eight mussels were chosen at randomfrom each of the treatment groups and control for measure-ments until all specimens had been analysed.

Oxygen uptake was measured in a closed respirometrychamber (500 ml) Wlled with oxygen-saturated Wltered(0.45 �m) seawater, held in a temperature controlled waterbath at 5°C. Each mussel was acclimated to the respirome-try chamber for 1 h. Oxygen consumption was determinedby sealing the chamber and measuring the reduction in %O2

with a YSI (5331) polarographic analyser and electrode(Yellow Springs, OH, USA). Seawater was well-mixedwith a magnetic stirrer. The output signal was monitoredcontinuously on a chart recorder until a decrease of at least20% O2 was reached. Oxygen uptake was determined bymeasuring the decrease in oxygen in the seawater with anYSI (5331) polarographic analyser and electrode coupled toa chart recorder. Under normal feeding conditions, oxygenconsumption represents routine metabolism (Thompsonand Bayne 1972). After performing the physiological testsfor calculating scope for growth, the mussels were starvedfor 8 days (in Wltered seawater at 1 �m) at the same temper-ature and salinity conditions following which oxygenuptake was measured to quantify their standard meta-bolism, an estimate of maintenance or basal metabolism(Myrand et al. 2002).

Clearance rate is deWned as the volume of water clearedof suspended particles per unit time and biomass (Widdowsand Johnson 1988). Clearance rates were measured using astatic system (experimental chamber of 500 ml), in which

the rate of decrease in particle density (algal depletion) inthe metabolic chamber was monitored (Riisgård 1991;Navarro et al. 2000). These measurements were carried outusing a particle Beckman coulter-counter Z1, Wtted with a100-�m oriWce tube. The experimental medium was kepthomogeneous by gentle aeration and the clearance rate wasevaluated at 10-min intervals for 1 h using visual inspectionof valve position. Following Gilek et al. (1992), the greatestdiVerence between two consecutive measurements duringthis period was assumed to be the clearance rate.

Assimilation represents the product of ingested energyand absorption eYciency (Widdows and Johnson 1988).Absorption eYciency was measured by the Conover (1966)ratio that consists of {(F ¡ E)/[(1 ¡ E)F]}, where F = ash-free dry weight:dry weight ratio of food and E = ash-freedry weight:dry weight ratio of the faeces. Samples of thediet mixture were collected during the experiments, as wellas faeces in each container. Faeces from individual musselswere collected, making sure that no pseudofaeces (i.e.rejected suspended particles) were produced and mixed withfaeces (Honkoop et al. 2003). Samples were Wltered throughpre-ashed, pre-weighed 47-mm glass Wbre Wlters, rinsedwith isotonic ammonium formate (3.2%), dried at 80°C for48 h, cooled to room temperature in a desiccator, weighed,combusted at 450°C overnight, cooled to room temperaturein a desiccator and Wnally weighed to estimate the organicand inorganic fraction contained in the food and faeces.

Scope for growth was measured by subtracting theenergy respired and excreted from the energy absorbedfrom the food as follows (Widdows and Johnson 1988;Gilek et al. 1992): [P = A ¡ (R + U)] where P = energyincorporated into somatic growth and gamete produc-tion. A is the energy absorbed from the food, calculatedmultiplying clearance rate (l g¡1 h¡1) and AbsorptioneYciency. Absorption eYciency is calculated by theequation: [POM (mg¡1) £ 23 J mg¡1] ash-free dry masswhere POM = particulate organic matter. R is the energycatabolized. It is derived by the calculation of oxygenconsumption in ml O2 h¡1 and then the conversion in energyequivalents using the factor 1 ml O2 = 20.33 J (Widdowsand Johnson 1988).

Following measurements of scope for growth and stan-dard metabolism, a given mussel was divided in two sec-tions, which were weighed. From one section, a piece ofdigestive gland between 100 and 200 mg of wet mass wasfrozen at ¡80°C for allozymes analyses and 100 mg ofmantle tissue was preserved for species determination usingmolecular marker. The other section was dried at 65–70°Cfor 72 h. The total dry tissue mass of the complete musselwas estimated from dried section and the relative wet massof the two sections. Physiological measures were standard-ized for 1 g mussel using the allometric relationship asdescribed by Tremblay et al. (1998).

123

1144 Mar Biol (2008) 153:1141–1152

Species determination

All the mussels used in this experiment were identiWed asM. edulis or M. trossulus according to the PCR technique(Glu-5 marker) developed by Rawson et al. (1996) follow-ing DNA extraction with Qiagen DNeasy protocol kit foranimal tissues (Moreau et al. 2005).

Statistical analyses

Allelic and genotypic frequencies for the polymorphic lociwere obtained using GENETIX 4.05 (Belkhir et al. 1998).The Wxation index (Fis), which represents the Mendelianequilibrium and deviation from Hardy–Weinberg equilib-rium, was calculated using GENEPOP 3.3 (Raymond andRousset 1995). Index of genetic diVerentiation, Fst, wascalculated using methods of Weir and Cockerham (1984)and signiWcant diVerences between allelic frequencies weretested using the Fisher’s exact test; both of these tests areimplemented in GENEPOP 3.3. SigniWcance levels for sta-tistical tests were adjusted according to the sequential Bon-ferroni procedure (Rice 1989). Degree of heterozygosity(DH) was determined by scoring a mussel between 0 and 5based on the number of heterozygous loci. The eVect of thetreatments on DH was performed using Kruskal–Wallis andMann–Whitney tests because of the discrete nature of theresponse variable. Comparison of initial DH and DH after10 months was performed using Mann–Whitney tests. Theanalysis was performed using SPSS 13.0. The models usedfor analysing growth and mortality measures in the Welddata used treatment and bay as Wxed factors with cage as arandom factor to account for potential clustering withincages. Mixed ANOVA models were used for length datawith SAS 8.2 software. Random eVects binary logisticregression was used to analyse mortality with the softwareMLwiN 1.2. All statistical tests were interpreted using a 5%error level, and multiple pairwise comparisons were carriedout by the Bonferonni method. To establish relationshipbetween size, growth and survival with DH, Pearson corre-lations have been tested and ANOVA using general linearmodel have been developed to evaluate if initial length andDH could predict growth in the Weld. Physiological resultswere log + 1 transformed to normalize data and then eVectof treatment was estimated by ANOVAs followed bya posteriori multiple comparisons of Tukey HSD.

Results

Size and mortality after 10 months

Mussel size analysis was performed on the mussels sam-pled in this study. Size diVerences were observed between

treatments and between bays, as well as in a signiWcantinteraction (P < 0.05). The analysis determined that airexposed mussels were signiWcantly larger than the controlor elevated water temperature exposed mussels and thatlocation had an eVect on mussel size (Table 1). LeBlancet al. (2005) also looked at the length of the same musselpopulations after 10 months. The mussels measured for thecurrent study are in fact a subset of the one used in LeBlancet al. (2005). A comparison between the datasets foundonly one diVerence in the results. In the smaller dataset,mussels in New London Bay were found to be signiWcantlylarger than those in St Peter’s and Tracadie Bay. In thelarger dataset, mussels in New London Bay were onlyfound to be larger than those from Tracadie Bay (Table 1).Mussels from the control group had signiWcantly (P < 0.05)higher mortality (72%) during the 10-month period in theWeld than mussels from either treatment group. Musselsfrom the air-exposed group had a mortality of 57% whilethose exposed to elevated water temperature had 58% mor-tality (LeBlanc et al. 2005).

Genetic characteristics between treatments

The genetic characteristics of the mussels from the twotreatment groups along with the control are described inTable 2. Between each treatment, we observed signiWcantdiVerences in allelic frequencies for MPI*, ODH*, PGM*(Fisher’s exact test for each of these three loci X2 = 1,df = 10, P < 0.001). These diVerences were mainly relatedto diVerences between the initial stock, immediately afterstressful treatments (P < 0.05 for each treatment). Theinitial control mussels had signiWcant deviation from

Table 1 Adjusted treatment and bay mean mussel sizes after10 months, for dataset used by LeBlanc et al. (2005) and the subsetused in the genetic analysis performed in the current study

Size (mm with SE)

*a,b,c—categories indicated by the same letters are not statisticallysigniWcant at the 5% level9 Comparisons for treatment and bay refer to average eVects, due tosigniWcant interaction

Full data set* (n = 394)

Subset for genetic analysis (n = 231)

Treatment9

Control 31.69 (0.56)a 32.17 (0.69)a

Air exposure 35.06 (0.60)b 36.70 (0.67)b

Elevated water temperature 32.96 (0.63)a 33.83 (0.67)a

Bay9

Tracadie 30.91 (0.57)a 32.88 (0.74)a

New London 35.60 (0.76)b 36.86 (0.65)b

St Peter’s 33.20 (0.85)ab 32.96 (0.65)a

123

Mar Biol (2008) 153:1141–1152 1145

Table 2 Number of individuals used for analyses, allelic frequencies, observed (Ho) and expected (He) heterozygosities and Fis deWciency (+) orexcess (¡) from Hardy–Weinberg equilibrium at Wve loci for mussels initially from three treatment groups and from three locations after 10 months

Bold indicates signiWcant heterozygosity (P < 0.05)

Treatments: air exposure (A), control (C) and elevated water temperature (T). Locations: St Peter’s Bay (SPB), Tracadie Bay (TB) and NewLondon Bay (NLB)

Locus Populations

C-initial A-initial T-initial C-SPB A-SPB T-SPB C-TB A-TB T-TB C-NLB A-NLB T-NLB

GPI*

(n) 80 80 80 38 32 29 30 30 30 30 30 30

A 0.094 0.112 0.100 0.092 0.047 0.052 0.100 0.167 0.150 0.083 0.100 0.083

B 0.075 0.200 0.162 0.197 0.141 0.138 0.267 0.083 0.233 0.183 0.183 0.067

C 0.287 0.275 0.262 0.303 0.313 0.276 0.233 0.317 0.183 0.333 0.267 0.283

D 0.412 0.287 0.313 0.289 0.250 0.345 0.300 0.267 0.333 0.250 0.283 0.400

E 0.038 0.044 0.056 0.000 0.109 0.034 0.017 0.017 0.017 0.000 0.033 0.050

F 0.094 0.081 0.106 0.118 0.141 0.155 0.083 0.150 0.083 0.150 0.133 0.117

Ho 0.575 0.887 0.887 0.711 0.969 0.897 0.733 0.933 0.900 0.700 0.967 0.800

He 0.723 0.781 0.783 0.763 0.786 0.758 0.767 0.771 0.771 0.763 0.786 0.732

Fis 0.210 ¡0.131 ¡0.128 0.078 ¡0.217 ¡0.166 0.067 ¡0.194 ¡0.1541 0.100 ¡0.223 ¡0.076

MPI*

A 0.138 0.225 0.069 0.211 0.359 0.138 0.233 0.200 0.150 0.150 0.200 0.400

B 0.819 0.731 0.894 0.789 0.641 0.810 0.767 0.783 0.833 0.833 0.783 0.517

C 0.044 0.044 0.038 0.000 0.000 0.052 0.000 0.017 0.017 0.017 0.017 0.083

Ho 0.150 0.463 0.188 0.263 0.469 0.310 0.400 0.367 0.200 0.233 0.167 0.433

He 0.309 0.413 0.195 0.332 0.460 0.322 0.358 0.346 0.283 0.283 0.346 0.566

Fis 0.519 ¡0.114 0.045 0.226 ¡0.002 0.053 ¡0.226 ¡0.042 0.308 0.191 0.531 0.250

ODH*

A 0.019 0.094 0.025 0.039 0.016 0.103 0.050 0.067 0.000 0.017 0.050 0.017

B 0.925 0.844 0.813 0.895 0.984 0.879 0.767 0.917 0.950 0.967 0.850 0.967

C 0.056 0.063 0.162 0.066 0.000 0.017 0.183 0.017 0.050 0.017 0.100 0.017

Ho 0.075 0.237 0.188 0.184 0.031 0.034 0.133 0.167 0.100 0.067 0.233 0.067

He 0.141 0.275 0.313 0.194 0.031 0.216 0.376 0.155 0.095 0.065 0.265 0.065

Fis 0.472 0.144 0.406 0.233 0.000 0.845 0.610 ¡0.058 ¡0.036 ¡0.009 0.136 ¡0.009

LAP*

A 0.075 0.144 0.131 0.171 0.156 0.034 0.167 0.133 0.183 0.033 0.150 0.083

B 0.369 0.438 0.331 0.408 0.344 0.414 0.417 0.350 0.333 0.333 0.383 0.250

C 0.556 0.412 0.525 0.421 0.484 0.552 0.400 0.500 0.483 0.617 0.450 0.650

D 0.000 0.006 0.013 0.000 0.016 0.000 0.017 0.017 0.000 0.017 0.017 0.017

Ho 0.463 0.637 0.525 0.632 0.719 0.517 0.600 0.600 0.500 0.500 0.600 0.500

He 0.549 0.618 0.597 0.627 0.623 0.523 0.638 0.609 0.622 0.507 0.628 0.508

Fis 0.164 ¡0.026 0.127 0.013 ¡0.139 0.029 0.104 0.032 0.212 0.031 0.061 0.032

PGM*

A 0.112 0.175 0.094 0.132 0.063 0.121 0.250 0.067 0.083 0.100 0.117 0.100

B 0.587 0.444 0.688 0.461 0.500 0.552 0.583 0.550 0.667 0.667 0.567 0.667

C 0.300 0.381 0.219 0.408 0.438 0.328 0.167 0.383 0.250 0.233 0.317 0.233

Ho 0.200 0.663 0.375 0.263 0.656 0.621 0.400 0.533 0.267 0.367 0.400 0.267

He 0.552 0.627 0.471 0.604 0.555 0.574 0.569 0.546 0.486 0.491 0.565 0.491

Fis 0.642 ¡0.050 0.209 0.582 ¡0.168 ¡0.064 0.302 0.040 0.465 0.269 0.307 0.470

All Ho 0.293 0.577 0.432 0.411 0.569 0.476 0.453 0.520 0.393 0.373 0.473 0.413

All He 0.458 0.546 0.475 0.511 0.499 0.487 0.551 0.494 0.459 0.429 0.527 0.480

123

1146 Mar Biol (2008) 153:1141–1152

Hardy–Weinberg equilibrium as measured with Fis forGPI*, MPI*, ODH* and PGM* loci and a large overalldeWcit in heterozygosity. The elevated water temperaturetreatment had signiWcant deWcits in heterozygosity atODH*, LAP* and PGM* loci. However these deWcits werenot as large as in the control mussels and overall observedheterozygosity (Ho) was only slightly lower than expectedby Hardy–Weinberg equilibrium. Immediately after the airexposure treatment we observed any deWcits in heterozy-gosity only at ODH* locus. After 10 months in the Weld twoloci, GPI* and LAP*, did not show signiWcant heterozygotedeWcits in control or treatment mussels in any of the threebays. A signiWcant deWcit at the ODH* and PGM* lociremained in control mussels placed in Tracadie Bay and asigniWcant deWcit at the PGM* locus of control musselspersisted in St Peter’s Bay. The elevated water temperaturetreated mussels in St Peter’s Bay continued to exhibit adeWcit at the ODH* locus, while in Tracadie and New Lon-don Bay a signiWcant deWcit remained at the PGM* locusfor the same group. The air treated mussels only showed asigniWcant deWcit at the MPI* and PGM* loci in New Lon-don Bay. The overall Ho scores reveal that in each bay theair treated mussels always had the highest observed hetero-zygosity.

Kruskal–Wallis analysis (df = 2, X2 = 63.7 P < 0.001)conWrm that the initial treatments had an eVect on DH ofmussels. Mann–Whitney tests, found that compared tomussels in the control group (DH = 1.46) both stress treat-ments had the eVect of signiWcantly raising the DH from theinitial control sample (elevated temperature, DH = 2.16,w = 4,285, P < 0.001 and air exposure, DH = 2.90,w = 5,298, P < 0.001). Also, the air exposure caused a moreimportant increase in the DH comparatively to the elevatedwater temperature exposure (w = 7,616, P < 0.01).

After 10 months of growth in diVerent bays, a Kruskal–Wallis analysis showing no signiWcant diVerences in theDH of mussels between bays (df = 2, X2 = 5.7 P = 0.06).Thus, data presented for each treatment are pooled from allthree bays (Table 3). In the elevated water temperature andthe air exposure treatments no diVerences in DH wasdetected in mussels from the 10-month period compara-tively to initial values (Table 3). However, mussels fromthe control sample had a signiWcant increase in their DHover the 10-month of growth (Table 3).

Relationship between size, growth and survival with heterozygosity

Pearson correlations for DH and initial lengths reveal a sig-niWcant correlation in the control sample (P-value = 0.01)with a value of 0.276 but not for the treated mussels(Table 4). No signiWcant correlations were observedbetween heterozygosity and length of mussels after10 months in the Weld. General linear models on lengthafter 10 months in the Weld with DH and initial length ascovariates, using bay and treatment as Wxed factors andcondo cages as a random variable, showed that DH (df = 1and 203, f = 0.96, P = 0.33) and initial length (df = 1 and203, f = 0.07, P = 0.79) were not signiWcant factors in pre-dicting length after 10 months. The diVerences in size at thebeginning of the experiment were completely compensatedfor by diVerential growth in the Weld. A second model wasrun to check for possible interaction between treatment andDH. There was no signiWcant interaction (df = 8 and 24.3,f = 1.43, P = 0.24).

Heterozygosity was altered by the two types of mortalitystress that mussels were subjected to in this experiment.Both initial treatments caused substantial levels of mortal-ity and resulted in higher levels of heterozygosity in thesurviving mussel groups (Tables 2, 3). In the Weld, the con-trol mussel group suVered signiWcantly higher mortalitythan either of the treated mussel groups, and was the onlyone of the three groups to see a signiWcant increase in heter-ozygosity (Tables 2, 3).

Table 3 Mann–Whitney test comparing initial degree of heterozygosity and degree of heterozygosity (scored 0–5) after 10 months

Treatment n-Initial n-Field Initial mean (SE) Field mean (SE) w-Value P-value

Control 80 98 1.46 (0.11) 2.07 (0.10) 5,856 <0.001

Air exposure 80 92 2.90 (0.12) 2.59 (0.11) 7,389 0.07

Elevated water temp. 80 89 2.16 (0.12) 2.14 (0.10) 7,541 0.94

Table 4 Pearson correlation of degree of heterozygosity (5 scoredloci) and initial length and degree of heterozygosity and length after10 months

Treatment—sample Pearson correlation

P-value

Control-initial length 0.276 0.01

Air exposure-initial length 0.073 0.52

Elevated water temperature-initial length 0.069 0.54

Control-Weld length 0.140 0.23

Air exposure-Weld length 0.048 0.68

Elevated water temperature-Weld length 0.160 0.15

123

Mar Biol (2008) 153:1141–1152 1147

Physiology and relationship with degree of heterozygosity

ANOVA indicated that treatments have signiWcant eVect onstandard and routine metabolism rates of mussels (Table 5).Posteriori tests indicated that mussels from the controlgroup had a higher routine metabolic rate (VO2) than thetwo treatment groups, air exposure (t = 4.1, P < 0.001) andelevated water temperature (t = 2.8, P = 0.05). The airexposed mussels and the mussels exposed to elevated watertemperature were both found to have signiWcantly lowerbasal metabolic rates than the mussels in the control group.The t-values were 3.12 and 3.2, respectively with P-values < 0.01. The two treatment groups and the controlgroup had similar clearance rates, absorption eYcienciesand scope for growth measurements (Table 5) and no sig-niWcant diVerences were found between bays.

In the mussels used for physiological measurements, asigniWcant diVerence in heterozygosity, measured by theDH (0–5 in relation to loci), was found among the treat-ment groups (Table 5, Kruskal–Wallis analysis). Musselsfrom the group exposed to the air had a mean heterozygos-ity of 3.25, compared to 3.13 for mussels exposed to tem-perature stress and 1.41 for the control group (Table 5).Both treatment groups were signiWcantly diVerent from thecontrol group with t-values of 6.3 and 5.7, w-valuesof 1,534 and 1,763, respectively, and all P-values < 0.01.The relationship between the physiological measurementsaVected by the treatments was analysed using a model withtreatment as Wxed eVects and DH as a covariate. Routinemetabolism and scope for growth did not show any signiW-cance in the general linear model. DH was signiWcant inthe model run on standard metabolism (df = 1 and 103,f = 10.9, P < 0.01).

Species determination

All mussels characterized in this study were determined tobe 100% M. edulis.

Discussion

Genetic eVect of selective treatment

This paper presents the results of artiWcial genetic selectionon physiological characteristics and performance (growthand survival) of M. edulis in diVerent Weld conditions. Aprevious paper (LeBlanc et al. 2005) covered in detail thesurvival and size analysis from the initial stress treatmentsand Weld study. This previous study revealed that the airexposure treatment caused a mortality of 47.8% and theelevated water temperature exposure caused a mortality of76%. The treatments resulted in surviving populations thathad diVerent mean lengths. The mean length of the air-exposed mussels was 8.07 mm, the mean length of the ele-vated water temperature exposed mussels was 6.49 mm andthe mean length of control mussels was 6.81 mm. Over the10-month period treated mussels had signiWcantly lowermortality than untreated mussels; air exposure: 56.8%, ele-vated water temperature: 58.3%, control: 71.8%.

In the present study survival was utilized to measureWtness and evidence of the HFC is present from the allo-zyme analysis of both mussel groups that were exposed tostress treatments. The surviving mussels from both treat-ments had a signiWcantly higher DH compared to theuntreated control spat (Table 4). Along with lower DH, theuntreated control mussels had a signiWcant deWcit in overallheterozygosity. No signiWcant overall deWcit was observedfor the two treated groups, indicating that the stress treat-ments did selectively remove more homozygous mussels.This Wnding is consistent with other studies that found theHFC in mussels increases during times of stress (Gentiliand Beaumont 1988; Myrand et al. 2002). Other studieshave found the HFC is only present or accentuated understress. They cover a variety of species and stresses: Ranatemporaria (common frog) and food availability (Lesbarr-eres et al. 2005), Scabiosa columbaria (perennial plant) andcompetition from other species (Pluess and Stöcklin 2004),

Table 5 Physiological measurements and degree of heterozygosity on same individuals mussels after stressful treatments

Measurements included routine metabolism (ml O2 g¡1 h¡1), absorption eYciency (%), clearance rate (l g¡1 h¡1), scope for growth (J g¡1 h¡1) andstandard metabolism (ml O2 g¡1 h¡1). Standard errors are in bracketsa Results indicated values of ‘f’ from ANOVA analysis for physiological measures and value of ‘X2’ from Kruskal–Wallis analysis for degree ofheterozygosity

Measurement n Control n Air exposure n Elevated temperature

Resultsa df Treatment P-value

Routine 28 0.46 (0.04) 47 0.29 (0.04) 32 0.28 (0.02) 5.7 2 and 104 0.01

Absorption 26 58.7 (2.3) 43 64.3 (2.3) 30 64.0 (2.7) 2.5 2 and 97 0.09

Clearance 28 3.95 (0.31) 43 3.93 (0.35) 32 3.78 (0.34) 0.1 2 and 101 0.90

SFG 26 61.3 (7.4) 40 82.6 (9.6) 30 77.3 (8.5) 2.0 2 and 94 0.14

Standard 28 0.17 (0.01) 47 0.12 (0.01) 32 0.12 (0.01) 7.2 2 and 104 <0.01

Heterozygosity 28 1.41 47 3.25 32 3.13 27.2 2 <0.01

123

1148 Mar Biol (2008) 153:1141–1152

M. edulis (marine mussel) and crowding (Gentili and Beau-mont 1988). In our study, the strongest evidence of HFCcame from survivors of the stress treatments. The resultsfrom mussels sampled immediately after the stress treat-ments indicated that the air exposure treatment was themore eVective selection technique for seed quality.Although it had a signiWcantly lower mortality than the ele-vated water temperature treatment, the survivors of the airexposure were more heterozygous than others groups. Sur-vivors of the air exposure displayed an excess of heterozy-gosity compared to Hardy–Weinberg expectations for allloci and were signiWcantly larger than mussels from the ele-vated water temperature treatment or control. Furthermore,the elevated water temperature mussels did have deWcits atthree individual loci (ODH*, LAP* and PGM*). Increase ofheterozygous individuals after a stressful event inducingmortality have already been observed in several musselsexperiments (Gentili and Beaumont 1988; Koehn 1991;Skibinski and Roderick 1991; Myrand et al. 2002).

Attempts to distinguish which theory explains the HFChave been aided by the development of various geneticmarkers, however there is still no deWnitive explanation.Pogson and Zouros (1994) demonstrated the theory of adirect relationship of the scored allozymes and growthbased on the positive correlation between growth in giantscallops (Placopecten magellanicus) and the level of heter-ozygosity at seven scored allozymes and the lack of rela-tionship between growth and eight random cDNAfragments derived using restriction fragment length poly-morphisms. A study on Atlantic salmon using allozymesand microsatellites found similar results, where Wtness traits[length, weight and Xuctuating asymmetry (FA)] werefound to be related to allozyme heterozygosity but notmicrosatellite heterozygosity (Borrell et al. 2004). Otherstudies suggest that associative overdominance bestexplains HFCs or at least heterozygosity is not evidence ofdirect selection (Waldmann 2001; Bierne et al. 2003).Waldmann (2001) used breeding experiments on the plantScabiosa canescens to show that FA was correlated to thelevel of inbreeding, while a previous experiment he per-formed found no link between FA and allozyme heterozy-gosity. Bierne et al. (2003) found that in a Mytilus hybridzone, neutral markers can show high variation similar toallozymes, which casts doubt on selection theories. Othershave questioned the universality of the relationshipbetween heterozygosity and Wtness (Jorgensen 1992;Britten 1996) and suggest that when all the research isconsidered, evidence of a positive relationship betweenheterozygosity and Wtness is sparse and inconclusive. Morerecent studies have examined the role of metabolism andprotein turnover in explaining HFC (Bayne and Hawkins1997; Tremblay et al. 1998; Bayne et al. 1999; Myrandet al. 2002) but distinguishing between competing theories

explaining HFC is still diYcult (Bierne et al. 2000). Thecomplexity of the interaction between genetic factors andthe environment may be the reason why no conclusiveexplanation has been elucidated. Evidence is mounting tosuggest that more than one of the theories of HFC may bevalid and they can diVer spatially, temporally and overlap(Heath et al. 2002; Myrand et al. 2002; Rand et al. 2002;Véliz et al. 2004).

DiVerences in allelic frequencies at the MPI* and PGM*loci and possibly the ODH* locus between the two treat-ment groups immediately following exposure to elevatedwater temperature or emersion also suggest the possibilityof direct selection in response to diVerent environmentalstresses. A strong body of evidence has developed showingdirect selection related to salinity on the LAP* locus inM. edulis (Koehn and Hilbish 1987), however this was notfound for the closely related species, M. galloprovincialis(Gardner and Kathiravetpillai 1997). Several studies onmussels and other species provide evidence of direct selec-tion at various allozyme loci (Hummel et al. 1995; Planesand Romans 2004; Riginos and Cunningham 2005) butresults can be inconsistent (Zhukovskaya and Kodolova2003) and strong evidence of direct selection is rare andgroundbreaking (Bierne et al. 2003). Often the environmentplays an important role. In some environments, allozymesappear to act as neutral markers (i.e. the isozymes have nodiVerentiated eVects, thus their distribution results fromrandom recombination), while in other environments thesame markers show evidence of direct selection (or linkedlocus). This phenomenon has been noted in at two leastdiVerent studies of the acorn barnacle, Semibalanus balano-ides (Rand et al. 2002; Véliz et al. 2004), with selectiondiVering among locations. Shifts in allozyme behaviouroccurred on micro and macro geographic scales. In Randet al. (2002), the GPI* locus was found to be neutral inMaine and selective in Rhode Island. Furthermore, selec-tion at the MPI* locus was opposite in Maine and RhodeIsland with respect to high and low tide habitats. Becauseof the complex environmental interaction that can occur inthe Weld, repeated lab experiments involving selection treat-ments, should be considered to detect selection at speciWcloci from quantiWable stress.

Relationship with physiological characteristics

Mussels decrease their metabolism in response to air expo-sure (Shick and Widdows 1981), sometimes retarding orcompletely suppressing heart rate to reduce metabolicrequirements (Coleman and Trueman 1971) and in turnincrease their survival rate (Famme et al. 1981). Myrandet al. (2002) hypothesized that homozygotes may not beable to reduce their basal metabolism as well as heterozyg-otes during air exposure, a logical possibility given the

123

Mar Biol (2008) 153:1141–1152 1149

higher maintenance metabolic rates that were associatedwith homozygous mussels in that study. Similar resultshave been found in another experiment on M. edulis(Tremblay et al. 1998), as well as in our study. Understressful conditions the maintenance metabolism of musselsincreases greatly (Koehn and Bayne 1989). The reason forthis is hypothesized to be an increase in protein turnover(HoVman and Somero 1995), which is required to make thenecessary adjustment in maintaining the internal environ-ment under aerobic stress (HoVman and Parsons 1991).Therefore the negative correlation of metabolism to hetero-zygosity may also explain the selective mortality of thehomozygous spat in the elevated water temperature treat-ment in this study. The general increase in heterozygosityfollowing the two stress treatments, even at loci whereselection may have occurred, provides evidence supportingthe theory of heterozygote superiority under stressful con-ditions.

A theory for the heterozygosity-Wtness relationship isthat lower protein turnover of heterozygotes (Hawkins et al.1986) gives them more energy for somatic growth and ametabolic advantage to dealing with stressful conditions(Myrand et al. 2002). Metabolic tests on respiration in thethree groups were consistent with this theory. Mussels fromthe three groups were examined using scope for growthmeasurements in the laboratory under optimal environmen-tal conditions. Mussels from the control had similar energyabsorbed from the food (clearance rate £ absorptioneYciency) compared to mussels from the stress experi-ments, but more energy was catabolized to maintain theorganism. The control mussels used more energy than mus-sels from the stress experiments but did not compensate byabsorbing more food, thus these mussels showed a trendtowards lower scope for growth values (energy incorpo-rated into somatic growth and gamete production), althoughnot statistically signiWcant at the 5% level (Table 5). Underthe stressful condition of starvation, maintenance metabo-lism was lower in the treated mussels and was negativelycorrelated to DH. This Wts the hypothesis that heterozygosity-Wtness correlations are more readily observed when animalsare more stressed.

Relationship with growth

In samples taken at the beginning of the experiment onlythe control showed a signiWcant correlation of MLH andsize. In the survivors of the elevated water temperaturetreatment and air exposure the relationship may have beenconfounded by the presence of physiological factors relatedto the type of stress experienced by the mussels. In the ele-vated water temperature treatment, smaller mussels fromthis experiment were found to be more resistant to thestress, while larger mussels tolerated air exposure better

(LeBlanc et al. 2005). Bayne (1984) found that althoughsmall mussels (M. edulis) had a higher maintenance energyrequirement and lost more weight during starvation, smallindividuals were at an energetic advantage in responding toelevated temperature. The energy level was determinedusing scope for growth measurements (the energy availablefor growth and reproduction). Scope for growth is simplyan energy equation that calculates the amount of energy anorganism possesses which can be allocated to cell produc-tion (reproductive or somatic). Bayne (1984) found that astemperature was increased, smaller mussels (<0.6 g dryweight) maintained positive scope for growth while largermussels did not. This countervailing physiological factormay explain why any possible size-MLH relationship is notapparent and why deWcits at individual loci remained afterthe elevated water temperature treatment. It is less apparentwhy the correlation disappeared after air exposure, as thistreatment selected out smaller mussels. No other report oflarger juvenile mussels surviving air exposure was reportedfor M. edulis, however Sukhotin et al. (2003) found that inadult M. edulis survival in air for large mussels wasreduced compared to small and medium sized adults. Thelack of similarity in size and age and conditions of exposuremake it diYcult to draw any conclusions by comparing theseemingly opposite results of these two studies. One couldspeculate that when mass is very low, as in our study, desic-cation is the primary factor causing death but when musselsget larger and older other factors such as metabolism andtissue composition are more important for resistance to airexposure. How the physiological response to anoxia mayinterrupt the size-MLH correlation is unknown. The MLH-size relationship may stem from a metabolic advantageunrelated to other factors involved in the survival from des-iccation. For instance, this may result from Xuid retentionand the strength of the adductor muscle, which overshad-owed MLH-size correlation. Other results from anoxicexperiments on bivalves reveal that no distinct pattern ofsurvival with size has emerged. In the freshwater zebramussel, Dreissena polymorpha larger mussels (as deter-mined by shell length) have been found to surviveanoxic stress longer than shorter mussels (Matthews andMcMahon 1995; Paukstis et al. 1999). In both of thosestudies, the mussels used were larger than those in thepresent study, with shell length ranging from 12 to 34 mm inMatthews and McMahon (1995) and larger specimens weregreater than 16 mm in Paukstis et al. (1999). Matthews andMcMahon (1995) also studied the freshwater Asian clam,Corbicula Xuminea (10–30 mm shell lengths) and in thatspecies larger clams showed signiWcantly lower survival toanoxic stress.

In this experiment the DH, as measured, was not a goodpredictor for size. Pearson correlations and one-way ANO-VAs revealed that neither of the two stress treatments nor

123

1150 Mar Biol (2008) 153:1141–1152

the control showed any signiWcant relationship between DHand length. The lack of MLH-length relationship have beenalready observed in mussels (Gentili and Beaumont 1988;Beaumont 1991) and may result from the experimentaldesign in these experiments and the use of only Wve loci.Mussels were tracked individually throughout the course ofthe Weld trials resulting in the need to house them in sepa-rate compartments, in other words at very low density.Another experiment that reared M. edulis in individualcompartments in the Weld found no correlation betweengrowth and multi- or single-locus heterozygosity (GaVney,1986). Further, an experiment by Gentili and Beaumont(1988) found no correlation between DH and growth forM. edulis reared at low densities but did Wnd a signiWcantcorrelation when mussels were kept at high densities. TheMLH-Wtness relationship is usually weak (Pecon Slatteryet al. 1993; Bayne and Hawkins 1997) and even thoughmortality in the Weld was high, perhaps speciWc stressfulconditions are needed to accentuate or reveal the MLH-growth relationship.

Two factors may explain why the correlation betweenheterozygosity and length was not apparent after 10 monthsin the Weld. First, diVerential mortality increased the hetero-zygosity in the controls and HFCs are often more apparentwhen heterozygosity is low (Zouros and Foltz 1987 as citedin Pecon Slattery et al. 1993; Myrand et al. 2002). This fac-tor may also explain the lack of MLH-size correlations inthe treated mussels both initially and in the Weld. Second,smaller mussels grew faster than larger mussels in the Weldexperiment. The biology behind this equalization process isunknown but it is possible that the low-density experimen-tal structure provided better opportunity to smaller musselscompared to a high-density arrangement where competitionfor space may favour larger individuals.

Conclusions

Although some studies have questioned the MLH-Wtnesslink (Jorgensen 1992; Britten 1996), many studies havefound evidence to support its existence in Mytilus (Mitton1994; Bayne and Hawkins 1997; Tremblay et al. 1998;Myrand et al. 2002). These studies vary in the strength ofthe correlation found and the reasons for this are numerous.Due to the complex nature of relating factors from genetics,physiology and ecology, it will likely take many more stud-ies to elucidate clearly the cause(s) of any MLH-Wtnessrelationships. As for the current discussion as to whetheroverdominance or associative overdominance is the basicpremise for explaining the link, it appears that the two maynot be mutually exclusive. The deWcit at the ODH* locusfound in the initial elevated water temperature treatmentand control but not in the air exposed mussels supports the

theory of overdominance; as does the diVerence in allelicfrequencies at three loci following the two diVerent stressregimes. The apparently more random nature of the shiftsin allelic frequencies at the GPI* and LAP* loci is consis-tent with the quantitative associative overdominancehypothesis. If in fact both processes are involved, this mayexplain why results show a large degree of variation fromstudy to study. It is also worth noting that even with thevast array of genetic analysis techniques now available toscientists, in most experiments genetic analyses only pro-vide a snapshot of information for a population or organ-ism, so caution must be exercised when employing andinterpreting diVerent techniques. DeWoody and DeWoody(2005) warn that using allozymes, microsatellites and SNPsto estimate genetic variability for the whole genome isinappropriate and Dufresne et al. (2002) found that sup-posed neutral markers, i.e. microsatellites, may not alwaysbehave in a neutral way.

Acknowledgements The authors are grateful to Garth Arsenault forhis technical assistance as well as Stephen Stewart (Stewart Musselfarms), Bob Fortune (United Mussel Farms) and Russell Dockendorf(P.E.I. Mussel King) for their participation in the Weld study. This workwas supported by the Prince Edward Island Aquaculture Alliance withfunds from the National Research Council Industrial Research Assis-tance Program, the Aquaculture and Fisheries Research Initiative, thePrince Edward Island Department of Agriculture, Fisheries, Aqua-culture and Forestry and the Department of Fisheries and Oceans.

References

Bayne BL (1984) Responses to environmental stress: tolerance, resis-tance and adaptation. Proceedings of the 8th European marinebiology symposium, University of Oslo, Norway, 14–20 August1983, pp 331–349

Bayne BL, Hawkins AJS (1997) Protein metabolism, the costs ofgrowth, and genomic heterozygosity: experiments with the mus-sel Mytilus galloprovincialis Lmk. Physiol Zool 70:391–402

Bayne BL, Hedgecock D McGoldrick RR (1999) Feeding behaviourand metabolic eYciency contribute to growth heterosis in paciWcoysters [Crassostrea gigas (Thunberg)]. J Exp Mar Biol Ecol233:115–130

Beaumont AR (1991) Genetic studies of laboratory reared mussels,Mytilus edulis: heterozygote deWciencies, heterozygosity andgrowth. Biol J Linn Soc 44:273–285

Beaumont AR, Toro JE (1996) Allozyme genetics of Mytilus edulissubjected to copper and nutritive stress. J Mar Biol Assoc UK76:1061–1071

Belkhir K, Borsa P, Goudet J, Chikhi L and Bonhomme F (1998)GENETIX, logiciel sous WindowsTM pour la génétique des pop-ulations. Laboratoire Génome et Populations, CNRS UPR 9060,Université de Montpellier II, Montpellier, France

Bierne N, Tsitrone A, David P (2000) An inbreeding model of associa-tive overdominance during a population bottleneck. Genetics155:1981–1990

Bierne N, Daguin C, Bonhomme S, David P, Borsa P (2003) Directselection on allozymes is not required to explain heterogeneityamong marker loci across a Mytilus hybrid zone. Mol Ecol12:2505–2510

123

Mar Biol (2008) 153:1141–1152 1151

Borrell YJ, Pineda H, McCarthy I, Vazquez E, Sanchez JA, Lizana GB(2004) Correlations between Wtness and heterozygosity at allo-zyme and microsatellite loci in the Atlantic salmon, Salmo salarL. Heredity 92:585–593

Britten HB (1996) Meta-analyses of the association between multilocusheterozygosity and Wtness. Evolution 50(6):2158–2164

Coleman N, Trueman ER (1971) The eVect of aerial exposure on theactivity of the mussels Mytilus edulis L. and Modiolus modiolus(L.). J Exp Biol Ecol 7:295–304

Conover RJ (1966) Assimilation of organic matter by zooplankton.Limnol Oceanogr 11:338–354

Cotter AJR, Phillips DJH, Ahsanullah M (1982) The signiWcanceof temperature, salinity and zinc as lethal factors for the musselMytilus edulis in a polluted estuary. Mar Biol 68:135–141

Crane C (2003) Unpublished salinity data (1998–2003) from theAugust survey. Prince Edward Island Department of Environmentand Energy, Water Management Division, Charlottetown

DeWoody YD, DeWoody JA (2005) On the estimation of genome-wide heterozygosity using molecular markers. J Hered 96:85–88

Dufresne F, Bourget E, Bernatchez L (2002) DiVerential patterns ofspatial divergence in microsatellite and allozyme alleles: furtherevidence for locus-speciWc selection in the acorn barnacle, Semi-balanus balanoides? Mol Ecol 11:113–123

Famme P, Knudsen J, Hansen ES (1981) The eVect of oxygen on theaerobic–anaerobic metabolism of the marine bivalve, Mytilusedulis L. Mar Biol Lett 2:345–351

Fisheries and Aquaculture Division (2001) P.E.I. mussel monitoringprogram 2001 report. Fisheries and Aquaculture Technical ReportSeries. Department of Fisheries, Aquaculture and Environment,Prince Edward Island

GaVney PM (1986) Physiological genetics of growth in marine bival-ves. Ph.D. dissertation, State University New York, Stony Brook

Gardner JPA, Kathiravetpillai G (1997) Biochemical genetic variationat a leucine aminopeptidase (LAP*) locus in blue (Mytilus gallo-provincialis) and greenshell (Perna canaliculus) mussel popula-tions along a salinity gradient. Mar Biol 128:619–625

Gentili MR, Beaumont AR (1988) Environmental stress, heterozygos-ity, and growth rate in Mytilus edulis L. J Exp Mar Biol Ecol120:145–153

Gilek M, Tedengren M, Kautsky N (1992) Physiological performanceand general histology of the blue mussel, Mytilus edulis, from theBaltic and North Seas. Neth J Sea Res 30:11–21

Gosling EM (1992) Genetics of Mytilus. In: Gosling E (ed) The musselMytilus: ecology, physiology, genetics and culture. Develop-ments in aquaculture and Wsheries science, vol 25. Elsevier,Amsterdam, pp 353–356

Hawkins AJS, Bayne BL, Day AJ (1986) Protein turnover, physiolog-ical energetics and heterozygosity in the blue mussel, Mytilusedulis: the basis of variable age-speciWc growth. Proc R Soc Lond B229:161–176

Heath DD, Bryden CA, Shrimpton JM, Iwama GK, Kelly J, Heath JW(2002) Relationships between heterozygosity, allelic distance(d2), and reproductive traits in chinook salmon, Oncorhynchustshawytscha. Can J Fish Aquat Sci 59:77–84

Hebert DPN, Beaton MJ (1989) Methodologies for allozyme analysiscellulose acetate electrophoresis. Helena Laboratories, Texas, p 32

HoVman GE, Parsons PA (1991) Evolutionary genetics and environ-mental stress. Oxford University Press, New York

HoVman GE, Somero GN (1995) Evidence for protein damage at envi-ronmental temperatures: seasonal changes in levels of ubiquitinconjugates and HSP 70 in the intertidal mussel Mytilus trossulus.J Exp Biol 198:1509–1518

Honkoop PJC, Bayne BL, Underwood AJ, Svensson S (2003) Appro-priate experimental design for transplanting mussels (Mytilus sp.)in analyses of environmental stress: an example in Sydney Har-bour (Australia). J Exp Mar Biol Ecol 297:253–268

Hummel H, Bogaards R, Amiard-Triquet C, Bachelet G, Desprez M,Marchand J, Rybarczyk H, Sylvand B, de Wit Y, de Wolf L(1995) Uniform variation in genetic traits of a marine bivalverelated to starvation, pollution and geographic clines. J Exp MarBiol Ecol 191:133–150

Jorgensen CB (1992) Heterozygosity and energetics of growth in sus-pension-feeding bivalves: a re-examination. Ophelia 36:171–186

Koehn RK (1991) The cost of enzyme synthesis in the genetics ofenergy balance and physiological performance. Biol J Linn Soc44:231–247

Koehn RK, Bayne BL (1989) Towards a physiological and geneticalunderstanding of the energetics of the stress response. Biol J LinnSoc 37:157–171

Koehn RK, GaVney PM (1984) Genetic heterozygosity and growthrate in Mytilus edulis. Mar Biol 82:1–7

Koehn RK, Hilbish TJ (1987) The adaptative importance of geneticvariation. Am Sci 75:134–141

LeBlanc N, Landry T, Stryhn H, Tremblay R, McNiven M, DavidsonJ (2005) The eVect of high air and water temperature on juvenileMytilus edulis in Prince Edward Island, Canada. Aquaculture243:185–194

Lesbarreres D, Primmer C, Laurila A, Juha M (2005) Environmentaland population dependency of genetic variability-Wtness correla-tions in Rana temporaria. Mol Ecol 14:311–323

Matthews MA, McMahon RF (1995) Survival of zebra mussels (Dre-issena polymorpha) and Asian clams (Corbicula Xuminea) underextreme hypoxia. US Army Corps of Engineers. WaterwaysExperiment Station. Technical report EL-95-3

Mitton JB (1994) Molecular approaches to population biology. AnnuRev Ecol Syst 25:45–69

Moreau V, Tremblay R, Bourget E (2005) Distribution of Mytilusedulis and M-trossulus on the Gaspe coast in relation to spatialscale. J ShellWsh Res 24:545–551

Myrand B, Tremblay R, Sevigny JM (2002) Selection against bluemussels (Mytilus edulis L.) homozygotes under various stressfulconditions. J Hered 93:238–248

Navarro JM, Leiva GE, Martinez G, Aguilera C (2000) InteractiveeVects of diet and temperature on the scope for growth of thescallop Argopecten purpuratus during reproductive conditioning.J Exp Mar Biol Ecol 247:67–83

Paukstis GL, Tucker JK, Bronikowski AM, Janzen FJ (1999) Survivor-ship of aerially-exposed zebra mussels (Dreissena polymorpha)under laboratory conditions. J Freshwater Ecol 14:511–517

Pecon Slattery J, Lutz RA, Vrijenjoek RC (1993) Repeatability ofcorrelations between heterozygosity, growth, and survival in anatural population of the hard clam Mercenaria mercenaria.J Exp Mar Biol Ecol 165:209–224

Planes S, Romans P (2004) Evidence of genetic selection for growth innew recruits of a marine Wsh. Mol Ecol 13:2049–2060

Pluess AR, Stöcklin J (2004) Genetic diversity and Wtness in Scabiosacolumbaria in the Swiss Jura in relation to population size. Con-serv Genet 5:1–12 (uncorrected proof)

Pogson GH, Zouros E (1994) Allozyme and RFLP heterozygosities ascorrelates of growth rate in the scallop Placopecten magellanicus:a test of the associative overdominance hypothesis. Genetics137:221–231

Rand DM, Spaeth PS, Sackton TB, Schmidt PS (2002) Ecologicalgenetics of MPI* and GPI* polymorphisms in the acorn barnacleand the spatial scale of neutral and non-neutral variation. IntegrComp Biol 42:825–836

Rawson PD, Joyner KL, Meetze K, Hilbish TJ (1996) Evidence for intra-genic recombination within a novel marker that distinguishes mus-sels in the Mytilus edulis species complex. Heredity 77:599–607

Raymond M, Rousset F (1995) Genepop, version 1.2. Populationgenetics software for exact test and ecuminicisms. J Hered86:248–249

123

1152 Mar Biol (2008) 153:1141–1152

Rice WR (1989) Analyzing tables of statistical tests. Evolution43:223–225

Riginos C, Cunningham CW (2005) Local adaptation and speciessegregation in two mussel (Mytilus edulis £ Mytilus trossulus)hybrid zones. Mol Ecol 14:381–400

Riisgård HU (1991) Filtration rate and growth in the blue mussel,Mytilus edulis Linneaus, 1758: dependence on algal concentration.J ShellWsh Res 10:29–35

Scott TM, Koehn RK (1990) The eVects of environmental stress onthe relationship of heterozygosity to growth rate in the coot clamMulinia lateralis (Say). J Exp Mar Biol Ecol 135:109–116

Shick JM, Widdows J (1981) Direct and indirect calorimetric measure-ment of metabolic rate in bivalve molluscs during aerial exposure.Am Zool 21:983–996

Skibinski DOF, Roderick EE (1991) Evidence of selective mortality infavour of Mytilus galloprovincialis Lmk phenotype in Britishmussel populations. Biol J Linn Soc 42:351–366

Sukhotin AA, Lajus DL, Lesin PA (2003) InXuence of age and sizeon puMPI*ng activity and stress resistance in the marine bivalveMytilus edulis L. J Exp Mar Biol Ecol 284:129–144

Thompson RJ, Bayne BL (1972) Active metabolism associated withfeeding in the mussel Mytilus edulis L. J Exp Mar Biol Ecol9:111–124

Tremblay R, Myrand B, Sevigny JM, Guderley H, Blier P (1998) Bio-energetic and genetic parameters in relation to susceptibility of

blue mussels, Mytilus edulis (L) to summer mortality. J Exp MarBiol Ecol 221:27–58

Tsitrone A, Rousset F, David P (2001) Heterosis, marker mutationalprocesses and population inbreeding history. Genetics 159:1845–1859

Véliz D, Bourget E, Bernatchez (2004) Regional variation in the spa-tial scale of selection at MPI** and GPI** in the acorn barnacleSemibalanus balanoides (Crustacea). J Evol Biol 17:953–966

Waldmann P (2001) The eVect of inbreeding on Xuctuating asymmetryin Scabiosa canescens (Dipsacaceae). Evol Ecol 15:117–127

Wallis RL (1975) Thermal tolerance of Mytilus edulis of EasternAustralia. Mar Biol 30:183–191

Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysisof population structure. Evolution 38:1358–1370

Widdows J, Johnson D (1988) Physiological energetics of Mytilusedulis: scope for growth. Mar Ecol Prog Ser 46:113–121

Zhukovskaya EA, Kodolova OP (2003) EVects of genotypes at theunspeciWed esterase (Est) and leucine aminopeptidase (LAP*)loci on the temporal dynamics of the morphological variation inthe Black Sea mussel Mytilus galloprovincialis Lam. Biol Bull30:262–270

Zouros E, Foltz DW (1987) The use of allelic isozyme variation for thestudy of heterosis. Isozyme 13:1–59

123


Recommended