The Physiological Basis for Faster Growth in the Sydney Rock Oyster, Saccostrea commercialis

11 Pages·1999·4.9 MB·English

by  B L Bayne

Reference: Bio/. Bull. 197: 377-387. (December 1999)
The Basis for Faster Growth in the
Physiological
Rock Saccostrea commercials
Sydney Oyster,
BRIAN L. BAYNE1'*, SUSANNE SVENSSON1'2 AND JOHN A. NELL3
1 Centrefor Research on Ecological Impacts ofCoastal Cities, Marine Ecology Laboratories All,
University ofSydney, NSW 2006, Australia; 2 Department ofZoophysiology, University ofGoteborg.
Medicinaregatan 18, 413 90 Goteborg, Sweden; and 3 NSW Fisheries, Port Stephens Research Centre,
NSW
Tavlors Beach, 2316, Australia
Abstract. Sydney rockoysters were sampled from a mass effectsofexogenous and endogenous factors. Ofthe various
selection experiment for growth (the "selected" category) endogenous factors involved, genotypic composition may
and from a control ("not selected") population and held in play a significant part.
the laboratory at three ration levels. We evaluated three Genetic properties may affect growth in various ways
models to explain fasterrates ofgrowth by selected oysters. (Koehn, 1991), including correlations between growth rate
Selection resulted in oysters feeding at up to twice the rate and genetic heterozygosity (Mitton and Grant, 1984; Zouros
and with greater metabolic efficiency than controls. A field el al., 1988; Britten, 1996). For marine bivalve molluscs in
experiment confirmed that selection leads to faster rates of particular, presumed interactions between genotype and
feedingacross awiderange offoodconcentrations. Selected growth are of particular interest in aquaculture (Newkirk,
oysters also grew more efficiently, at a smaller cost of 1980), and the selective breeding of oysters has often suc-
growth (Cj: mean values for Cg were 0.43 J J ' in ceeded in increasing average rates ofgrowth (see review by
selected individuals and 0.81 J J~' in the controls. In Sheridan, 1997).
contrast, oysters in both categories showed similar meta- Analysis ofthe bioenergetics ofgrowth is useful in stud-
bolic rates at maintenance, i.e., at a ration supporting zero ies seeking to link phenotypic variability in growth to ge-
growth. There was no evidence that differential energy netic causes. This approach involves the dissection of
allocation affected the balance between total metabolic re- growth into its component processes, as represented by the
quirements above and below zero net energy balance. By "balanced energy equation" of Winberg (1956; see review
esixzpeesriamnedntaignegs,witthhesnelsetcatneddaardnidzicnogntrtohleodyasttearsfoorfdsiizfef,erewnet hbayvWeiedseesrc,rib19e9d4h).owFogreneextaimcpallel,y bParesseedntlataintdudCionanlodviefrfe(r1e9n9c2e)s
found no effects of age on the differences due to selection. in the growth rate of the fish Menidia menidia were due to
Faster-growing oysters feed more rapidly; invest more en- differences in both food consumption rates and somatic
egcirregonywctyhp:,eratnhjdaonualrseelaoibwnlegeers-ttogeradol;wloiscnahgtoeiwnldeasisvihedinugeahlresgr.y npeetrugnriotwotfhtiesfsfuie- igdreonwttifhypehffyisciioelnocgyi.caIlnmethcihsanpirWesesmesnttosetxupdlya,inwoebsesertveoduvtarit-o
ability in growth in oysters. postulated three ways by
which an individual animal may increase its rate ofgrowth
Introduction above that of other individuals, when held in the same
The physiological processes that constitute growth are of environmental conditions. Though not fully independent
fundamental interest. A striking feature ofgrowth in nature nor mutually exclusive, these are sufficiently different in
isits variability amongst individuals, which is a result ofthe both underlying mechanisms and likely ecological conse-
quences to act as useful alternative models to explain vari-
ability in growth rate among individuals. We then evaluated
R*eTcoeivwehdo1m6 Dcoercreemsbpeornde1n9c98e;sahcocuelpdtedbe29adSderpetsseemdb.erE-1m9a9i9l.: blb@bio. these models by comparing oysters artificially selected for
usvd.edu.au faster growth with control, "not-selected," oysters.
377
378 B. L. BAYNE ET AL
First, in the increased acquisition model, an individual Table 1
may obtain more food perunit time by feeding more rapidly Mean (SD: n = 12} shell heights (cm) and whole weights {shellplus
than others, so increasing its metabolizable energy intake flesh, grams} ofoysters in thefourexperimental categories
(Present and Conover, 1992; Rist et al., 1997; De Moed el
til.. 1998). This may be evaluated by comparing a variety of Category Shell height Whole weight
traits of feeding behavior (BaWynee et al.. 1999) and relating Selected. Large 7.93 0.33a 52.95 1.05
the results to rates ofgrowth. tested the hypothesis that, Not selected. Large 7.89 0.47" 49.66 2.40C
under similar conditions of ration (quantity and quality), Selected. Small 6.37 0.32" 38.40 3.87d
oysters selected for faster growth will have faster rates of Not selected. Small 6.54 0.36b 36.96 2.84d
ingestion than do control (not-selected) oysters. Values sharing superscripted letters are not significantly different.
Second, in the modified allocation model, faster growth
may be the result of greater proportional allocation of en-
ergy to growth at the expense of other energy-demanding are referred to as "not-selected." These oysters were grown
We
processes, such as body maintenance (Wieser, 1989). under identical conditions to the selected oysters, within
evaluated one aspect ofthis model by estimating the meta- Nelson Bay, New South Wales, though they were from a
bolic rate of selected and not-selected oysters at mainte- natural larval settlement and not cultured as larvae within
nance, when growth is neither negative nor positive, to test the hatchery, as were the selected individuals. Within each
the hypothesis that selected oysters would show reduced category, we distinguished one group of "small" and one
maintenance rates. groupof"large" oysters (Table I). The selectedoysterswere
Finally, in the metabolic efficiency model, faster growth 23 monthsold. The ages ofthe not-selectedoysters werenot
may result from a higher growth efficiency (Present and known with certainty, but the "not-selected, small (NSS)"
Conover. 1992) from reduced metabolic costs of growth individuals are considered to be of similar age to the "se-
(Wieser, 1994). or from a combination of the two. This lected, large (SL)" individuals, though of smaller size, and
model was evaluated in two ways. Firstly, net somatic ofsimilar age and size to the "selected, small (SS)" oysters.
growth efficiency, defined as the proportion of metaboliz- The "not-selected, large (NSL)" oysters are thought to be
able energy intake allocated to growth, was determined. about 6 months older than the selected, large oysters.
Secondly, by measuring metabolic rates at different rates of Twelve oysters from each experimental category (Table
growth, we tested the hypothesis that reduced costs of I) were tagged for individual identification and held in the
growth correlate, amongst individuals, with increased water-table ofaresearch aquarium ofrecirculating seawater
growth rate. in Sydney. The aquarium contained 600 1 ofwater, ofwhich
We used the Sydney rock oyster, Saccostrea conunercia- 33% was replaced every 7 days. Water temperature was
lis (Iredale and Roughley). The New South Wales Fisheries controlled at 20 5C and salinity at 33 1.5%c.
Research Centre at Port Stephens. Australia, established a
mass selection program for these oysters in 1990 (Nell el The laboratory experiment was as follows:
al.. 1996). Four selected lines were established for faster 20 January to 8 Fehnuiry 1998 (20 days). No supple-
growth, and these have been bred in alternate years since. mentary food added; physiological measurements
After one generation of selection, Nell et at.. ( 1996) found made from 28 January to 5 February and labeled the
that oysters from two ofthe lines were heavier than oysters "field" condition.
from two control lines. After two generations, their weight 9February to4March (24 days). Food added to make up
was 18% greater than controls (Nell et a/.. 1998). Oysters the "middle ration" condition; measurements made
from the third generation ofselection (referred to in Nell et from 23 February to 2 March.
til.. 1998, as the "loose 2" selection line) were used in the 5 March to 18 March (14 days). Food added to comprise
present study and compared with control oysters. The ex- the "high ration" condition; measurements made 12 to
periment was designed to test hypotheses derived from the 18 March.
three models discussed above and to identify the physiolog- 19 March to 9April (22 days). No food added; the "low
ical characteristics that may explain enhanced growth in the ration" condition; measurements made from 6 to 10
selected lines.
April, after 16-20 days without food.
Materials and Methods
Rations
Material and general procedures The food was unicelled algae Isochrysis galbana (strain
Oysters were provided by the NSW Fisheries Laboratory T-ISO) and Chaetoceros gruclUs, supplied as algal pastes
at Port Stephens. The selected oysters were as described by Reed Mariculture Inc.. California. Individual pastes were
above. Controls were from a commercial oyster farm and combined in the proportion 3 parts T-ISO to 7 parts C.
PHYSIOLOGY AND GROWTH OF ROCK OYSTERS 379
gracilis, and the cells were suspended in seawater in a
feedingreservoiratthedesiredconcentration (Table II). The
cells were dosed to the oysters by peristaltic pump, from
0930 to 1530 daily (middle ration) and 0930 to 1630 (high
ration). Cell concentrations in the trays were monitored
frequently with a particle size analyzer (Coulter Counter
model Zl). Samples ofcellsfromthe feeding reservoirwere
weighed after drying overnight at 80C, and then measured
for nitrogen content (by Leco CHN analyser).
measurements
Physiological
All oysters were measured for clearance rate, absorption
efficiency, rate of oxygen consumption, and rate of ammo-
nia-nitrogen excretion, at each ration level. Following pre-
liminary studies, care was taken not to use the same indi-
viduals in any two measurements without at least 24 h of
recovery from the stress of handling.
Clearance rate (CR). Clearance rate is a measure of the
volume of water cleared of algal cells per hour. When
pseudofeces (that is, material cleared from suspension but
not ingested) are not produced (as in this experiment), the
rate of ingestion of food is calculated as CR X [food
concentration].
Oysters were placed individually in 1-1 beakers in water
from the feeding trays and left undisturbed in a water bath
for 30 min. A beaker without an oyster was used as the
blank control. The water was gently aerated using Pasteur
pipettes coupled to a compressed air supply. About 600 ml
of the water was then siphoned off and replaced with
seawatercontaining algal cells at aconcentration equivalent
to the experimental ration. A 10-ml sample was taken after
10 min and then at 10-min intervals for a further 40 min.
Cell concentrations were measured with a particle size
analyzer (Coulter Counter model Zl).
Fordata analysis, acheck was first made for linearity (In
cell concentration plotted over time), then clearance rate as
Table II
Maximal concentrations ofcells (T-ISO + C. gracilis) andtotal
paniculate matter(TPM) in theoystertraysforeach ofthe
experimental rations
380 B. L. BAYNE ET AL.
vidually in 1 1 offiltered seawater and left undisturbed for 3 is based on a separate sample ofoysters (Svensson, unpub-
h. Beakers without oysters served as blank controls. Con- lished). Theexponent forgrowth was determined from rates
centrations of ammonia were measured using the phenol- of growth calculated (see above) for the high ration condi-
hypochlorite method of Solorzano (1969); a full set of tion. The values were as follows:
standards was analyzed for each experimental run. Rates of = =
excretion, as milligrams of ammonia-nitrogen per hour, Clearance rate: ft 0.64=1 0.1 13( 32)=
were calculated as: Oxygen consumption ra=te: J3 0.536 0.10=7(>i 25)
Excretion rate: /3 0.772 0.156(/( 50)
VNH4 - N = (Conccvpll - Conca,nlrill)-Vol//. Growth rate: j3 = 1.96 0.58(/i = 45).
winheerxpeeCriomnecnetxaptllaannddcCoonntrcoclonbleroa|kaerres,amremsopnecitaivceolnyc,eVnotrlatisiotnhse fleAshswmeeiagshutreodf tohveerexalplerciamteegnotrailesoWy(snte=rs4a8t),thteheemndeaonfdtrhye
volume of water in the beaker, and / is the incubation time experiment was 0.920 0.243 g. stand was set at 1.0 g.
(3 h).
Oxygen:nitrogen ratio. The ratio, in molarequivalents, of Field measurements
oxygen consumed to nitrogen excreted serves as an index of Rates of ingestion were measured in the field on two
catabolic substrate (Bayne and Newell, 1983). and was occasions, as a test of the hypothesis that results of the
calculated to evaluate whether the oysters in the different laboratory experiment on feedingrates, and in the contextof
growth and sizecategories were utilizing different biochem- a model ofenergy acquisition, would be repeated under the
ical substrates, a difference which might then explain other more natural conditions of food availability. Selected and
observed metabolic differences. not-selected oysters of similar size (64.0 3.35 g and
Growth. The 12 oysters in each experimental category 62.7 2.86 g whole weight, respectively) were held over-
were weighed (shell plus flesh) at the beginning (Table 1) night in wide-mesh bags at the mouth of the Karuah River
and end of the experiment. Growth was calculated by sub- as it enters the Port Stephens estuary, on 9-10 September
traction and related to the period spent at high ration (14 and 3-5 November. 1998. This is an area used for cultivat-
days) to convert to a daily rate. At the end of the experi- ing oysters. The selected oysters were from the same mass
ment, the oysters were shucked and dry flesh weights de- selection as those used in the laboratory. During the field
termined after drying overnight at 80C. measurements, watertemperatures were 19.1 1.9C(over
Due to uncoupling in the growth of shell and tissue in both months) and salinities were 28.2 1.9%o (September)
bivalves (Hilbish, 1986; Lewis and Cerrato, 1997) conver- and 32.5 8.3%c (November). Total paniculate matter in
sions of total weight to weight of tissue, using a constant suspension was 8.0 2.3 and 29.3 5.8 mg 1 in
conversion factor, must be made with caution. For this September and November, respectively.
study, we derived such conversion factors for each individ- The oysters were placed individually in specially de-
ual at the end of the experiment. Given the relatively short signed trays (36 X 16 X 8 cm. with a sill at one end to
duration of the experiment, we considered it appropriate to reduce turbulent flow) at flow rates of450 15 ml min" '
use these factors to estimate equivalent dry flesh weight at of water pumped directly from the river. After 1 h all
the start, and so toestimate growth also as milligrams ofdry biodeposits were removed from the trays and the oysters left
tissue per day. undisturbed for a further 30 or 60 min. Feces and pseudo-
feces were then collected quantitatively, together with sam-
Converting rates to a standard body size ples ofsuspended particulate matter, and filtered onto ashed
and weighed GF/C filters. The filters were dried overnight at
This conversion was based on allometric relationships 80C, weighed, ashed at 450C for4 h, and weighed again.
between dry flesh weight and the measured physiological The results were used to calculate rates of filtration and
rates, following Bayne and Newell (1983): ingestion by the "biodeposition" method as described by
Iglesias et al. (1998) and Bayne er al. (1999).
where V/stanJ and VVsland are the standardized rate and dry Statistical anal\sis
W
flesh weight, respectively; Vmeas and meas are the rate and The results ofthe laboratory experiment were analyzed in
dry flesh weight as measured; and |3 is the allometric expo- three stages, using SYSTAT 6.0 (Wilkinson, 1996).
nent in the equation describing physiological rate as a
function of body size. 1. For each ration condition the physiological measure-
Estimates of/3 were derived forclearance and respiration ments, standardized to an animal size of 1 g dry flesh
ANOVA
rates across all experimental categories from the "field weight, were analyzed as a two-way with
condition" measurements. The exponent for excretion rate "selection" (i.e.. selected or not-selected) and "size"
PHYSIOLOGY AND GROWTH OF ROCK OYSTERS 381
(large or small) as the main effects. In all cases the middle~ration, 0.593 0.024: high ration, 1.071 0.068
"selection X size" interaction was not significant. mg 1 '.
2. Given the "repeated measures" nature ofthe experi- Clearance and ingestion rates. These were measured for
mental design, a different approach was taken to the middle and high ration levels only (Table III). Differ-
analyze the data across rations. Three groups of four ences due to size alone, following standardization to 1 g dry
individuals were first selected at random from each flesh weight, were significant only for ingestion rates at the
category (selection and size). These were then allo- middle ration. This result indicates that the effects ofage on
cated, again at random, to one of the three ration feeding rates were not greatly significant overall. The ef-
levels. A three-way ANOVA was then done, with fects of selection were greater, particularly at high ration,
"selection," "size," and "ration" as the main effects. where the selected, large oysters had significantly faster
Where both "selection" and "ration" showed signif- rates ofclearance and ingestion than the not-selected, large
icant effect, a regression analysis was performed, and not-selected, small oysters.
with ration as the independent variable, to compare When the original ingestion rates (i.e.. before standard-
oysters from the different categories. izing them to 1 g dry flesh weight) were converted to
3. Finally, dataforeach individual oysteroverthe three percentages of body weight (as dry mass in milligrams),
ration levels were analyzed by linear regression, and significant differences among categories were evident
comparisons between categories were made on the (Fig. 1: high ration). Selected oysters at both middle and
basis ofthe average "within category" values for the high ration levels had faster relative ingestion rates than
slope and intercept in the fitted equations. These the not-selected oysters, with no significant differences
regressions were for three data points only, per indi- between large and small oysters. For example, at the
vidual: only those for which the level of significance middle ration. SL oysters ingested 1.14 0.47 %bw
in the analysis was P < 0.10 were used for compar- d"1 compared with 0.80 0.22 %bw d"1 for NSL
,
isons. individuals; the values for SS and NSS oysters were
The results of the measurements in the field were ana- 0.88 0.44 and 0.68 0.33 %bw d 1, respectively
lyzed by two-sample / test with pooled variance. (see Fig. 1 for the high ration data).
Absorption efficiency. There were no significant differ-
Results ences due either to selection or to size. Values were all
between 0.66 and 0.78 (mean 0.71 0.06), with slightly
Laboratory experiment lower values at high than at middle ration (P < 0.05).
Ration levels. The cell concentrations in the feeding trays Rates ofoxygen consumption. Differences among cate-
<
(Table II) were converted to equivalent dry mass using the gories (Table IVA) were significant (P 0.05) only forsize
constant 0.013 mg per 106 cells, and to nitrogen content effects at the middle ration, where the VO2 was higher for
using 5.6% N by weight. "Peak rations" are the levels the NSL oysters than for the NSS oysters.
recorded between 1100 and either 1500 (low and middle Differences between selected, large, and not-selected,
ration) or 1600 (high ration), and are the concentrations large, and between selected, small, and not-selected, small,
applied during the physiological measurements (except ex- oysters were not significant at any of the ration levels.
cretion rates, which were measured in filtered seawater). The effects of ration on rates of oxygen consumption,
Peak rations were as follows: low ration, 0.074 0.018; however, were highly significant for all categories, with
Table III
Clearance rates (CR: I h~'> andingestion rates (IR: mg h '}foroysters in each offourexperimental cuti'xurii'x. at twn rution*
CR IR
Category
382 B. L. BAYNE ET AL
rates converted as 26.5 J mg ' (Widdows and Hawkins,
TO 1989). The energy ingested per unit ofenergy respired was
then calculated.
cu
Q. At the middle ration, differences in this efficiency mea-
sure between categories were not significant; the overall
TO3 mean value was 0.83 0.38 joule ingested per joule re-
spired, which indicates aration level below the maintenance
requirement. At the high ration, average energy ingested per
LU unit respiration was higher (1.53 0.55 J J"'), with
<
significant differences due to selection (P 0.01). Selected
oysters (large and small: 1.86 0.78 J J"1) were more
efficient in this respect than the not-selectedoysters ( 1.20
0.55 J J"1).
CO
LOU This conclusion wasconfirmed by the analysis ofdata for
? individuals. Foreach oyster, aregression analysis was made
SL NSL SS NSS of respiratory energy loss. R (J d '), as a function of
CATEGORY ingested ration, IR (J d"1). Figure 2 shows the means and
standard deviations of the fitted slopes, grouped by cate-
Figure 1. Ratesofingestion (not standardized fordifferences in body gory. Categories SL and NSL (P < 0.02), and SS and NSS
weight), as percent of dry body weight per day, by oysters in the tour (P < 0.001) were significantly different. Differences be-
Seseexllpeeeccrttieemdde,,nstlmaaalrlglce.;atNegSoLr,iesNoatt-sheilgehcrtaetdi,onl;armgeea;nSsS. SSelDe.ctTehde,csamtaelglo;riNeSsSa.reNSoLt.- etvweere,nwseirzeecanottegosriigensifi(cSaLnt.vs.OSnSaavnedraNgeS,Lsvesl.ecNtSeSd).oyshtoewr-s
respired 0.24 J foreveryjoule ingested, across ration levels,
respiration rate increasing as ingested ration increased. Ox- compared with 0.45 J by the not-selected oysters.
ygen consumption rates were converted to energy equiva- Excretion rates. At all ration levels, selected oysters
lents as 20.1 J ml O: ' (Gnaiger, 1983), and ingestion excreted more ammoniathan the not-selectedoysters (Table
Table IV
Metabolic measurementsforoysters infourexperimentalcategories at low, middle, amihigh ration levels
Ration
Category
PHYSIOLOGY AND GROWTH OF ROCK OYSTERS 383
u.o
T3
384 B. L. BAYNE ET AL.
Table V
Rates ofgrowth andgrowth efficiency tit hi^li ration, the estimatedcosts ofgrowth andmaintenance metabolic rate (Rmjln[),/oroysters in each of
fourexperimentalcategories
Category
PHYSIOLOGY AND GROWTH OF ROCK OYSTERS 385
different energy substrates at the time of the experiment. finding lends general support to the "energy acquisition"
Such differences might indicate different physiological model anddemonstratesthat the inferredgenetic component
states in the two growth categories, which would render of variability in feeding behavior supports faster growth,
more detailed metabolic comparisons complex. There was, both under natural circumstances in the field and in the
however, no significant difference in the O:N ratios (P > laboratory.
0.05), inspiteofdifferences in rates ofexcretion, suggesting Feeding rates were not only faster in selected oysters,
that selection for growth did not shift the normal seasonal they were also more metabolically efficient. Feeding is not
pattern of metabolism significantly in these oysters. itselfenergetically expensive in bivalve molluscs, although
Until we know more about nitrogen metabolism in this the total costs offeeding and digestion plus absorption may
species, we cannot fully interpret the observed differences account for up to 20% oftotal metabolic rate (Hawkins and
in excretion rates. However, the selected oysters both fed Bayne, 1992). Feeding rates may increase significantly
more quickly and excreted nitrogen at a higherrate than the without seriously compromising net energy yield. The rock
control individuals. This is not unexpected, but more infor- oysters selected for growth achieved a greater gain of en-
mation on the relationship between ingested and excreted ergy per unit ofenergy lost in metabolism than did control
nitrogen is needed before these observations can be set in oysters. The actual mechanisms of feeding that are respon-
context with selection for growth. sible for these differences remain unknown. We presume,
We proposed three models to explain the observed dif- however, that increasing rate of ingestion as a mechanism
ferences in rates ofgrowth: faster growing individuals may for increasing gross energy yield will be limited eventually
feed more rapidly; they may reduce their maintenance en- by decreased gut passage time, which ultimately limits
ergy requirement; or they may grow more efficiently than maximum absorption efficiency (Bayne et al., 1989).
slower growing individuals. The results of the laboratory Our second model, which we call the energy allocation
experiment supported the first and third, but not the second, model, was not supported by the results ofthe experiment.
of these models. Further, by experimenting with different The estimated rate of metabolism at maintenance varied
sizes (and therefore ages) of oysters, but correcting the between 250 and 360joules per day, standardized for oys-
wmeeasduermeomnesnttrastaecdcotrhdatinaggteo owbassernvoetdasizsei/grnaitfeicraenltatfiaocntsohripsi,n wteirtshosfel1egctdiroyn.flAenshawveeirgahgte, aofnd30it8inJcreda~s'ediswieqtuhiavgaleebnutttnoota
explaining most of the observed differences between indi-
maintenance requirement formetabolizable energy intake of
viduals. The exceptions were estimated maintenance meta-
bolic rate, which increased with age. and growth efficiency, 1.4% ofbody weight per day. This is similar to published
values forotherbivalves ofsimilar size (reviewedby Bayne
which declined with age.
and Newell, 1983) and accords with ourconclusion that the
Oysters from the selected line had faster clearance rates
(volume of water cleared of food cells per hour) than middle ration level in the laboratory was insufficient to meet
control, not-selected oysters. This was reflected in a 44% therequirements formaintenance oftheseoysters. Increased
increase in ingestion rates at the greatest ration. Because of maintenance costs with age have commonly been reported
for other bivalves (review by Griffiths and Griffiths, 1987).
a similarity across experimental categories in the efficiency
with which these cells were absorbed in the gut, an identical Studies with blue mussels, Mytilus edulis, by Hawkins et
difference in metabolizable energy intake was observed. In ul. (1986) and Bayne and Hawkins (1997), and with rain-
experiments with hybrid and inbred lines of Pacific oysters bow trout, Oncorhvnchus mykiss, by McCarthy et al.
(Crassostrea gigeis), Bayne et al. (1999) recorded faster (1994), have demonstrated how reduced rates of protein
clearance rates by hybrids in three out offour comparisons, turnover contribute to reduced metabolic costs and higher
consistent with observed differences in growth rate. Genet- rates of growth. These processes appear to be genotype
ically based differences in feeding behavior are an impor- dependent, and they support the concept of differential
tant component of differences in rates of growth among energy allocation (Wieser, 1989) as a means of increasing
individual oysters. growth. In our experiments, differences in maintenance
A similar difference between selected and not-selected metabolic rate between fast- and slow-growing oysters were
oysters was observed in the field experiment, in which a not statistically significant. This result merits further re-
different technique for measuring feeding behavior was search, however. For example, the data (Table V) show a
used (the biodeposition method in the field, in contrast to tendency towards highermaintenance metabolic rates in the
directcell counts in the laboratory), and the concentration of not-selected oysters, particularly among the smaller size
food was significantly greater (8.00 2.28 and 29.3 5.8 categories, but with high variance. In similar experiments
in the field, compared with a maximum of 1.07 0.07 mg with Pacific oysters, Crassostrea gigas (Bayne, in press),
1~' in the laboratory). Under these conditions the differ- we have observed significant differences in flmaim among
ences between the two experimental categories of oyster individuals, which correlated with differences in growth
were actually more marked than in the laboratory. This rate. The energy allocation model remains a possibility in
386 B. L. BAYNE ET AL
MEI
Metabolic select
energy loss
MEI
control
*Snaint
G
v control 'select
[-ve] zero [+ve] Growth
Figure4. A qualitative illustration ofthe main findingsofthis study. Metabolic energy loss isplotted asa
function of growth for oysters selected for fast growth, and for control oysters. Below the maintenance
requirement, where growth is negative, there is no difference in metabolic expenditure due to selection; the
maintenance metabolicdemand(ftma,nl) isthesameforbothexperimentalcategories. However,selectedoysters
achieve a higher metabolizable energy intake (MEI) than the controls and express a lowercostofgrowth. The
net result is an increased growth rate and a higher growth efficiency (Gxka vs. Gcomro|). Differences due to
selection have been exaggerated for illustration purposes.
the general case, therefore, although not supported directly selection for growth in the rock oyster resulted in individ-
by these data on Saccostreu. uals that had a greater intake of metabolizable energy by
Our third model concerned growth and metabolic effi- virtue of faster (and more metabolically efficient) feeding,
ciency. This was evaluated by estimating both the costs of and were able to use this intake more efficiently forgrowth.
growth and net growth efficiency in selected and not-se- Selected and control oysters did not differ in their energetic
lected oysters. The results supported the hypothesis that costs at maintenance. The field experiment confirmed that
selectedoysters would show a lowercostofgrowth (0.43 J selected oysters fed more rapidly than the controls. The
J ') than control oysters (0.81 J J~'). Both values are high challenge now is to analyze in more detail the feeding
compared with published values (Wieser, 1994; average for behavior and the metabolic processes that contribute to the
"ectothermic metazoans" of0.30J J~'), possibly reflecting costs ofgrowth and to link these processes more directly to
a relatively poor-quality diet and slow overall rates of observed individual differences in genotype.
growth. Nevertheless, the differences due to selection, and
the lack of significant differences due to age, are evident.
Clearly, selection for growth in this species, as in others Acknowledgments
(Bayne and Hawkins, 1997), involves selection for reduced
costs of growth. We are grateful to Shannon Long, who helped with many
Selected and not-selected oysters also differed in growth aspects ofthis work, and to Shannon and Alison Phillips for
efficiency measured as the proportion of metabolizable en- their work in the field. Graham Housefield provided invalu-
ergy intake utilized in tissue growth. This efficiency was able support throughout the project. The research was sup-
low in all cases forexample, 0.28 0.09 and 0.21 0.08 ported by a grant from the Australian Research Council to
for selected and not-selected oysters, respectively. The re- the Special Research Centre on Ecological Impacts of
sults do, however, support the hypothesis that selected oys- Coastal Cities. Susanne Svensson was supported by the
ters utilize a higher proportion of absorbed ration for Foundation for Strategic Environmental Research, Sweden
growth, and do so at a reduced cost ofgrowth relative to the (Sucozoma Project. DNR 95005). The manuscript benefited
controls. significantly from the comments ofGee Chapman and Tony
In summary (Fig. 4), our experiments indicate that mass Underwood.