www.elsevier.nlrlocateraqua-online
Feeding behaviour of greenback flounder larvae,
ž
/
Rhombosolea tapirina Gunther with differing
¨
exposure histories to live prey
E.S. Cox
), P.M. Pankhurst
School of Aquaculture, UniÕersity of Tasmania, PO Box 1214, Launceston, Tasmania 7250, Australia
Accepted 6 September 1999
Abstract
This study examined the feeding performance of cultured larvae of the greenback flounder
Rhombosolea tapirina, using the live feed organisms, Brachionus plicatilis and Artemia sp., to
determine the primary sensory modality involved in feeding, and the effect of previous exposure to prey on subsequent prey selection. The proportion of larvae that fed on rotifers in the light
Ž400–700 nm daylight fluorescent tubes — light intensity of 5–6 mmol sy1 my2., increased
significantly from 66% to 96% from day 12 to day 27 post-hatching, respectively. In comparison, the proportion of larvae that fed on rotifers in total darkness, never exceeded 5% during the same period. This indicates that greenback flounder larvae are primarily dependent upon vision to feed.
Ž .
Prior feeding experience of greenback flounder larvae with either rotifers only R-treatment or a
Ž .
mixed diet of Artemia and rotifers A & R-treatment significantly affected subsequent prey selection when larvae were offered a mixed diet of Artemia and rotifers. It did not, however,
Ž .
affect the temporal onset of selection of the novel prey species Artemia by R-treatment larvae. Therefore, the differences in prey selection by larvae, with or without prior exposure to Artemia prey, were not due to the inability of larvae to handle and ingest a novel prey species, but they reflect positive selection for familiar prey. The latter indicates a learned component in the feeding behaviour of fish larvae. This has implications for the timing of the introduction of new live prey species during intensive culture of marine fish larvae. q2000 Elsevier Science B.V. All rights reserved.
Keywords: Feeding behaviour; Experience; Fish larvae; Learning; Flounder
)Corresponding author. Queensland Department of Primary Industries and Fisheries, Northern Fisheries Centre, PO Box 5396, Cairns, Queensland, 4870, Australia. Tel.:q61-07-4035-0158; fax:q61-07-4035-141; e-mail: [email protected]
0044-8486r00r$ - see front matterq2000 Elsevier Science B.V. All rights reserved.
Ž .
1. Introduction
Irrespective of the feeding mode adopted in adult life, most early life history stages of fish are selective predatory planktivores. This involves sensory detection of individual
Ž
prey by the predator, prior to initiation of the predator strike Arnold and Holford, 1990; .
Browman and O’Brien, 1992; Miller et al., 1993 . Most first-feeding fish larvae are
Ž .
dependent upon vision for prey detection Blaxter, 1986 , although non-visual senses Ž have also been implicated in prey detection by selective planktivorous fish larvae Batty
.
and Hoyt, 1995; Salgado and Hoyt, 1996 . This is in contrast to the non-selective particulate mode of feeding displayed by juveniles and adults of some fish species, in which gill rakers are used to filter prey from the environment without prior
discrimina-Ž .
tion of individual prey Janssen, 1980; Gibson and Ezzi, 1985; Batty et al., 1986 . Selective planktivory and non-selective particulate feeding are both active feeding behaviours and are not to be confused with passive prey engulfment, which may occur at low rates, especially in marine species that must drink seawater for osmotic control ŽTytler and Blaxter, 1988 ..
The success of prey capture by planktivorous fish larvae depends upon larval age, size, and motor and physiological competence, all of which increase during ontogeny ŽBlaxter, 1986 . Detection of a prey organism does not always result in a predator strike,. implying a prey selection process is involved. Studies that compare the prey spectra available in the environment with prey ingested by fish larvae, confirm that prey
Ž
characteristics besides size strongly affect patterns of prey selectivity Checkley, 1982;
. Ž .
Magnhagen, 1985; Govoni et al., 1986; Meng and Orsi, 1991 . Jenkins 1987 suggested that some fish larvae display innate prey preferences at an early age, whereas others
Ž .
suggest that learning plays an important role in prey selection Werner et al., 1981 ; such that fish positively select for and are more effective at capturing prey that are
Ž .
familiar Checkley, 1982; Meyer, 1986; Coughlin, 1991; Wahl et al., 1995 .
In this study, feeding performance of larvae of the greenback flounder Rhombosolea
tapirina, was assessed in two light regimes, irradiance of 5–6mmol sy1 my2 and total darkness, to confirm that greenback flounder larvae are primarily visual feeders. Feeding response of two groups of larvae, one with prior exposure to rotifers only and one with
Ž
prior exposure to both Artemia and rotifers, was then examined in the light 5–6mmol
y1 y2.
s m to determine the effect of prior prey exposure on subsequent prey selection.
2. Materials and methods
2.1. Source of larÕae
Ž .
Two cohorts of greenback flounder larvae cohorts 1 and 2 , each from separate groups of broodstock, were reared during August 1996. Wild female flounder that were
Ž . Ž .
sexually mature ns15 were caught during June and August 1996 , from Waubs Bay,
capture, wild-caught females were treated with an intraperitoneal injection of LHRHa at
a dose of 100 mg kgy1 body weight. Forty eight hours after injection, and daily
thereafter, females were anaesthetised in a solution of 0.02% 2-phenoxyethanol, and checked for the presence of hydrated ovulated eggs. For each cohort of larvae, eggs from 2–3 ovulated females were stripped by the application of gentle pressure to the abdominal area; and, in a similar fashion, milt was expressed from 2–3 naturally spermiated males and it was collected in a syringe. Eggs were fertilised by adding 1 ml of milt per 100 ml of eggs, plus a small volume of seawater to activate the sperm. After 2–3 minutes, the eggs were transferred to a bucket of seawater from which fertilised eggs were skimmed from the surface and stocked into 200 l larval rearing tanks at a density of 50 ly1. Eggs and larvae were incubated in a recirculating seawater system in
which 25–50% of the tank volume was exchanged per day. Temperature was maintained at 12"18C, photoperiod at 13 h light: 11 h dark, and light intensity at the water surface
was 5–6mmol sy1 my2 during the photophase.
Live feed organisms, large strain rotifers, Brachionus plicatilis, andror Artemia
Ž .
nauplii and metanauplii INVE — Artemia Systems, Belgium , were enriched with
Nutripake and introduced into the tanks from day 4 post-hatching. Live feed regimes
varied according to experimental protocol and are described below.
2.2. Examination of feeding behaÕiour in light and dark
Greenback flounder larvae from cohort 1 were reared in a single 200 l culture tank.
Ž y1. Ž
Rotifers were introduced twice daily 5 ml from the time of first feeding day 4
. Ž y1.
post-hatching . Artemia nauplii were added once daily 1–2 ml in addition to the
rotifers from day 12 post-hatching. Feeding behaviour trials were conducted on days 12,
Ž .
15, 18, 21, 24 and 27 post-hatching in a stable temperature environment 128C"18C ,
y1 y2 Ž .
at two light intensity treatments: 0mmol s m absolute darkness and 5–6 mmol
y1 y2 Ž .
s m the light intensity at which larvae fed actively in the culture tank . Light was
Ž .
provided by broad-spectrum 400–700 nm Osram Daylight fluorescent tubes. The night before each experiment, 30 larvae were transferred from the 200 l larval culture tank into each of eleven 2.5-l black chambers, where larvae were maintained in static seawater culture. The chambers were then covered with blackout cloth and left undis-turbed overnight. There were 5 replicates per treatment and an additional chamber from which twenty larvae were sampled the next morning and examined under a dissecting microscope to confirm gut evacuation prior to commencement of the feeding trial. Large strain rotifers were then added sequentially at ten minute intervals to each of the
remaining ten chambers at a density of 2 mly1. The order of addition of rotifers was
2.3. Determination of the effect of preÕious exposure to prey
Fertilised greenback flounder eggs from cohort 2 were divided and stocked into two 200-l culture tanks, to form two treatments: treatment 1, in which larvae were
subse-Ž y1.
quently exposed only to rotifers 5 ml ; and treatment 2, in which larvae were
Ž y1. Ž y1.
subsequently exposed to both rotifers 5 ml and Artemia 1–2 ml . Prey selection
Ž .
by larvae from treatment 1 prior exposure to rotifers onlysR-treatment larvae and
Ž .
treatment 2 prior exposure to rotifers and ArtemiasA & R-treatment larvae was then examined in feeding trials in which larvae were offered a mixed diet of both rotifers and
Artemia. In this fashion, Artemia nauplii were a novel prey for R-treatment larvae.
Feeding behaviour trials were conducted in 3-l chambers on days 11, 14, 17, 20, 23, 26 and 29 post-hatching at 12"18C and with light intensity at the test chamber water
surface of 5–6 mmol sy1 my2. There were five replicates per treatment, plus an
additional chamber to confirm gut evacuation of larvae. The night before each experi-ment 30 larvae from the appropriate 200-l culture tank were stocked into each of the 11 test chambers, which were covered and left overnight in the dark. The next morning, gut evacuation was confirmed in larvae from a single chamber and a mixed diet of rotifers
Ž y1. Ž y1.
1 ml and newly hatched Artemia nauplii 1 ml were added sequentially into the
test chambers, in a random order, at ten minute intervals, at a total density of 2 mly1. The larvae were left undisturbed to feed for 1 h before 20 larvae from each chamber were examined for presence of rotifers andror Artemia in the gastro-intestinal tract. All feeding responses were recorded using a presence or absence criterion.
Rotifers and Artemia used in the feeding trials were first washed through a 200mm
screen and collected on a 100 mm screen to produce a discrete size fraction of prey.
Ž .
Mean lorica width and length of rotifers ns50 and mean width and length of
Ž .
Artemia ns50 , were measured to determine absolute dimensions of prey.
Ten greenback flounder larvae from each of the R- and A & R-treatments, were randomly sampled from the culture tanks on each of the days tested. Larvae were anaesthetised in 0.02% 2-phenoxyethanol and examined under a dissecting microscope
Ž
fitted with an eyepiece graticule. Standard length SL — distance from the rostral tip of .
the larva to the posterior end of the notochord was measured to confirm that any differences in feeding responses of larvae in the two treatments were not due to differential growth rates of larvae.
2.4. Statistical analysis
Ž .
A two-way analysis of variance ANOVA and a Tukey–Kramer multiple compari-son of means test were used to analyse the effect of increasing age on feeding response
Ž
of larvae in the light and dark. Residual values replicate means subtracted from
. Ž
treatment means of arcsin6transformed data were normally distributed Shapiro–Wilk
.
test, Prob-Ws0.139 for data from the light treatments, but data from the dark
Ž .
treatments were not normally distributed Shapiro–Wilk test, Prob-Ws0.000 .
In the prior experience feeding trial, larvae were offered two prey species, which Ž
resulted in three possible feeding responses selection of Artemia only, rotifers only or .
both Artemia and rotifers . These data were analysed using a multiple analysis of
Ž .
variance MANOVA , in conjunction with a canonical distribution analysis, to test for the treatment effect on the three possible larval feeding responses with increasing larval age.
3. Results
3.1. Feeding behaÕiour in the light and dark
The ability of greenback flounder larvae to feed in the light differed markedly from the feeding ability of larvae maintained in the dark, with a consistently higher proportion
Ž .
of larvae feeding in the light on all days tested Fig. 1 . There was a significant Ž . difference in the proportion of fish feeding in the light on day 12 post-hatching 66% , compared with the proportion of larvae feeding in the light on day 27 post-hatching
Ž96%. Žtwo-way ANOVA, dfs5, ns20, Prob)Fs0.0002 , with a trend of increas-.
Fig. 1. Percentage of greenback flounder larvae feeding on rotifers at two light intensities, 0mmol sy1 my2 Ždark treatment and 5–6. mmol sy1 my2 Žlight treatment , with increasing age of larvae. Dark cross-hatched.
Ž .
bars days 21 and 24 of age indicate feeding incidence in the dark on Artemia nauplii, which were
Ž .
inadvertently transferred into the test chambers along with the larvae. Values are means "SE of five
Ž .
ing feeding incidence from days 15–24 post-hatching. The proportion of larvae that fed in the dark on rotifers was consistently low, ranging from 2% on day 15 to a maximum of 5% on day 18 post-hatching, and it did not change significantly during ontogeny
Žtwo-way ANOVA, dfs5, ns20, Prob.)Fs0.883. Žnot withstanding non-normal
.
distribution of data . On days 21 and 24 post-hatching, 26% and 34% of larvae, respectively, had fed on Artemia nauplii when these prey items were inadvertently transferred into the test chambers along with the larvae. The proportion of larvae feeding
Ž .
Fig. 2. Percentage of greenback flounder larvae feeding on rotifers only no bar fill , rotifers and Artemia
Ž . Ž . Ž .
nauplii cross-hatched bars , and Artemia only black bars in: a larvae that had previous exposure to only
Ž . Ž .
rotifer prey R-treatment larvae ; and b larvae that had previous exposure to both rotifers and Artemia
on Artemia in the dark on days 21 and 24 post-hatching cannot, however, be compared with the proportion of larvae feeding on rotifers on these days, because larvae poten-tially had 16 h over-night to feed on Artemia, compared with 1 h to feed on rotifers during the rotifer feeding trial.
3.2. The effect of preÕious exposure to prey on subsequent prey selection
Greenback flounder larvae selected rotifers only, Artemia nauplii only or a mixture of rotifers and Artemia. Feeding responses of A & R-treatment larvae were markedly
Ž .
different from feeding responses of R-treatment larvae Fig. 2a,b . Prior feeding
Ž
experience had a significant effect on subsequent larval prey selection MANOVA: .
Pillai’s trace statistic, dfs39, Fs5.9279, Probs0.000 . Results from canonical
Ž .
discriminate analysis CDA confirmed a difference in prey selection of A & R-treatment
Ž .
larvae, compared to prey selection by R-treatment larvae Fig. 3 . Canonical variate 1 explained 69% of the variation in larval feeding response, whereas canonical variate 2 explained 26.9% of the variation in larval feeding response. The variation between the proportion of larvae that fed in the A & R-treatment and the R-treatment along canonical variate 1, was largely due to the proportion of larvae that ingested Artemia. The variation in feeding response between the A & R-treatment and the R-treatment larvae along canonical variate 2 was largely due to the proportion of larvae that ingested only rotifers. In addition, the variation between the two treatments along both canonical
Fig. 3. Canonical distribution analysis plot showing the variation in the proportion of R-treatment and A&R-treatment greenback flounder larvae feeding on either rotifers, Artemia nauplii or both rotifers and
Ž
Artemia. Circles indicate 95% confidence ellipses for A&R- and R-treatment larvae, respectively Biplot rays:
.
variate 1 and 2 was due, to a lesser degree, to the proportion of larvae that ingested both
Artemia and rotifers.
Ž .
A high proportion of R-treatment larvae between 50%–80% consistently fed on
Ž .
rotifers only Fig. 2a . No R-treatment larvae fed on Artemia prior to 14 days-of-age, at which time 7% of larvae ingested both Artemia and rotifers. Fewer than 50% of R-treatment larvae selected both Artemia and rotifers on any one day thereafter. On day 29 post-hatching, there was little difference in the proportion of R-treatment larvae selecting rotifers only and those selecting both prey species. On only two occasions, days 20 and 26 post-hatching, did a small percentage of R-treatment larvae select only
Ž .
Artemia 5 and 1%, respectively .
Eleven-day-old A & R-treatment larvae, like R-treatment larvae, selected only rotifers ŽFig. 2a,b . After day 11 until day 17 post-hatching, a high proportion of A & R-treatment.
Ž . Ž .
larvae 60–80% selected rotifers only Fig. 2b . The proportion of A & R-treatment larvae selecting only rotifers decreased thereafter, never exceeding 31% of larvae. On
Ž .
day 14 post-hatching, a small number of A & R-treatment larvae 1% selected only
Artemia. The proportion of larvae selecting only Artemia increased thereafter to 18% in
26-day-old larvae, but then decreased to 7% in 29-day-old larvae.
No larvae from the A & R- or R-treatments ingested Artemia prior to 14 days of age. Larvae started to ingest both rotifers and Artemia at the same age; however, a small number of A & R-treatment larvae also started to select Artemia only on the 14th day post-hatching. With the exception of 17-day-old larvae in the R-treatment group, a
Ž .
Fig. 4. Linear regression of change in standard lenght mm with increasing age of greenback flounder larvae
Ž 2 . Ž 2 .
for A&R-treatment larvae ys1.87q0.13 x, r s0.99 and R-treatment larvae ys1.97q0.13 x, r s0.99 .
Ž .
higher proportion of A & R-treatment larvae selected both Artemia and rotifers, when
Ž .
compared to the feeding response of R-treatment larvae Fig. 2a,b .
Ž .
Standard length mean"SE of greenback flounder larvae from cohort 2 increased
from 2.81 mm"0.06 in 7-day-old R- and A & R-treatment larvae, to a maximum of
5.53 mm"0.12 and 5.69 mm"0.10 in 29-day-old R- and A & R-treatment larvae,
Ž .
respectively Fig. 4 . Regression fits of change in SL with increasing age of larvae are
Ž 2 . Ž 2
described by the equations: ys1.97q0.13 x r s0.99 , and ys1.87q0.13 x r s
. y1
0.99 , respectively, representing an average daily growth increment of 0.13 mm day
Ž .
for both R-treatment and A & R-treatment larvae Fig. 4 .
3.3. Prey size
Ž .
Mean "SE, ns50 lorica length and width of rotifers was 278"3 mm and
Ž .
184"2 mm, respectively, whereas mean "SE, ns50 length and width of Artemia nauplii was 470"0.3mm and 186"20mm, respectively.
4. Discussion
For all stages of greenback flounder larvae examined, a consistently high percentage fed in the light, whereas a relatively low proportion fed in the dark. There was no significant change during ontogeny in the proportion of greenback flounder larvae feeding in the dark on rotifers. This, in conjunction with the high percentage of larvae feeding in the light, indicates that greenback flounder larvae are primarily dependent upon vision, to feed during the early pelagic larval phase. It should be restated that the apparently higher incidence of feeding by greenback flounder larvae in the dark on
Artemia, which were inadvertently introduced into the test chambers on days 21 and 24
Ž .
post-hatching 26% and 34%, respectively , cannot be compared with feeding perfor-mance on rotifers, because larvae potentially had all night to feed on the Artemia, compared with a 1 hour feeding duration once rotifers were introduced in the feeding trials. The general ontogenetic trend of increasing ability of greenback flounder larvae to capture prey in the light, confirms earlier reports that indicate that feeding ability of
Ž
larvae is initially poor, but increases with age of fish Houde and Schekter, 1980; Mills et al., 1986; Arnold and Holford, 1990; Coughlin, 1991; Browman and O’Brien, 1992;
.
Miller et al., 1992; Wahl et al., 1993 . This is consistent with growth related increases in mouth size and improvements in both motor and sensory capabilities, with the result that the duration of the feeding event for a given prey size decreases with increasing body
Ž
size of fish larva Arnold and Holford, 1990; Coughlin, 1991; Coughlin, 1994; Cook, .
1996 .
Until quite recently, it was generally accepted that most marine fish larvae were
Ž .
obligate visual planktivores Blaxter, 1986 . However, there is increasing evidence that some larvae have the capacity to feed in the dark using non-visual senses to detect prey
ŽDabrowski, 1982; Townsend and Risebrow, 1982; Batty and Hoyt, 1995; Salgado and
.
Ž .
larvae of another flatfish, sole Solea solea , are active at night and feed very effectively
Ž .
in the dark Blaxter, 1969; Batty and Hoyt, 1995 using chemoreception and
mechanore-Ž . Ž .
ception to locate prey Batty and Hoyt, 1995 . Plaice larvae Pleuronectes platessa also feed in the dark, but only at significantly reduced levels compared to both the feeding
Ž
ability of sole larvae in the dark and their own feeding ability in the light Batty and .
Hoyt, 1995 . The non-visual sense organs of greenback flounder are poorly developed at
Ž .
first feeding Pankhurst and Butler, 1996 , but continued development during the larval period potentially provides for greater non-visual sensory input to prey detection as larvae grow. This is supported by the observation that adult greenback flounder readily
Ž .
feed in both light and dark conditions Chen et al., 1998 .
A field study which examined gut contents of greenback flounder larvae over a 24 Ž
hour period indicated that feeding occurred equally well during both night on a
. Ž
moonless night therefore precluding visually mediated prey detection and day Jenkins, .
1987 . The authors, however, advised caution in interpreting these data. The larvae, sampled in the single 24 hour period examined, had ingested a high proportion of bivalve veliger larvae, which are less readily digested than some other prey types, and it was possible that the apparent high rate of feeding in the dark may reflect retention of
Ž .
veliger shells in the gut Jenkins, 1987 . This did not, however, explain the occurrence of invertebrate eggs in the gut of larvae sampled at night. The incidence of feeding by greenback flounder in the dark in the present laboratory based study, albeit at a relatively low level compared to the proportion of fish feeding in the light, indicates that non-visual feeding cannot be discounted.
Previous exposure to a prey species significantly affected subsequent patterns of prey selection by greenback flounder larvae under light conditions. Differences in feeding observed in A & R-treatment larvae, compared with R-treatment larvae, were largely attributed to respective differences in selection of either Artemia nauplii only or rotifers only. Larvae in both treatments displayed similar growth profiles and this, in conjunc-tion with the screening of live prey prior to use in the feeding trials, minimises a
Ž .
confounding influence of either prey or larval and therefore gape size.
Differences in prey selection by R-and A & R-treatment greenback flounder larvae occurred despite the fact that larvae from both treatments started to ingest Artemia Žeither in combination with rotifers or alone for the first time at 14 days of age. This. indicated that the temporal onset of Artemia selection was not affected by prior prey exposure regimes, and that the differences in prey selection were not due to the inability of larvae in the R-treatment group to handle and ingest the novel prey species. It is
Ž .
likely, however, that capture and ingestion of Artemia the novel prey by R-treatment Ž
larvae involved increased handling times duration of visual orientation and prey
.
fixation prior to the strike compared with experienced A & R-treatment larvae, because in other studies the latency of the feeding response was longer in fish that had no prior
Ž .
experience of a prey species Meyer, 1986; Wahl et al., 1995 .
Ž .
largemouth bass Micropterus salmoides Colgan et al., 1986 , juvenile bluegill sunfish
Ž .
Lepomis macrochirus Werner et al., 1981 , cichlid fry Cichlasoma managuense ŽMeyer, 1986 and walleye Perca fla. Õescens Wahl et al., 1995 , which improved afterŽ .
Ž .
repeated periods of exposure to a novel prey. In addition, Salgado and Hoyt 1996
Ž .
reported that prey selection by fathead minnow larvae Pimephales promelas reflected ‘sensory learning’, because the feeding response of larvae changed in a predictable way according to the suite of sensory cues available to the larvae during a pre-conditioning
Ž
period. For example fish reared on live food in the dark under which conditions . chemosensory and mechanosensory cues are potentially available for prey detection , fed
Ž .
less on dead food in both the light when vision and chemosensory cues were available
Ž .
and dark when only chemosensory cues were available .
Learned feeding behaviour has implications for the intensive culture of marine fish larvae. Because of mouth size constraints and the energetic gains in larvae ingesting
Ž
larger prey as body size increases Polo et al., 1992; Bremigan and Stein, 1994; .
Paszkowski and Tonn, 1994 , intensive culture protocols for rearing larvae of marine teleosts usually involve a sequential transfer to live prey species of increasing size during ontogeny. Previous studies have indicated that learning to successfully feed on a
Ž
novel prey criteria including feeding capture success, latency of feeding duration and
. Ž
handling effort variously occurred after 1–5 days exposure to the novel prey Ware, .
1971; Werner et al., 1981; Meyer, 1986; Wahl et al., 1995 . This indicates that substantial temporal overlap is required during the transition to sequentially larger live feeds, or weaning diets, if feeding efficiency of larvae in culture is to be optimised.
Acknowledgements
This study was supported by a University of Tasmania Office for Research Grant to P.M. Pankhurst. Thanks are extended to Ned Pankhurst, Carolyn Barnett, Andrea Hobby and Polly Hilder for assistance with capture and hormone treatment of wild broodstock, and to Carolyn Barnett for assistance with statistical analysis.
References
Ž .
Arnold, G.P., Holford, B.P., 1990. The reactive perceptive field of larva plaice Pleuronectes platessa : a three dimensional analysis of visual feeding. Rapp. P.V. Comm. Int. Explor. Mer. 191, 474.
Batty, R.S., Hoyt, R.D., 1995. The role of sense organs in the feeding behaviour of juvenile sole and plaice. J. Fish Biol. 47, 931–939.
Ž .
Batty, R.S., Blaxter, J.H.S., Libby, D.A., 1986. Herring Clupea harengus filter-feeding in the dark. Mar. Biol. 91, 371–375.
Blaxter, J.H.S., 1969. Visual thresholds and spectral sensitivity of flatfish larvae. J. Exp. Biol. 51, 221–230. Blaxter, J.H.S., 1986. Development of sense organs and behaviour of teleost larvae with special reference to
feeding and predator avoidance. Trans. Am. Fish. Soc. 115, 98–114.
Bremigan, M.T., Stein, R.A., 1994. Gape-dependent larval foraging and zooplankton size: implications for fish recruitment across systems. Can. J. Fish. Aquat. Sci. 51, 913–922.
Ž
Browman, H.I., O’Brien, W.J., 1992. Foraging and prey search behaviour of golden shiner Notemiganus
.
Ž .
Checkley, D.M. Jr., 1982. Selective feeding by Atlantic herring Clupea harengus larvae on zooplankton in natural assemblages. Mar. Ecol. Prog. Ser. 9, 245–253.
Ž
Chen, W., Purser, J., Blyth, P., 1998. Circadian feeding rhythms in greenback flounder Rhombosolea
.
tapirina : role of light–dark cycles. Australian Marine Sciences Association 1998 Conference, Abstract, Adelaide, p. 30.
Colgan, P.W., Brown, J.A., Orsatti, S.D., 1986. Role of diet and experience in the development of feeding behaviour in largemouth bass, Micropterus salmoides. J. Fish Biol. 28, 161–170.
Cook, A., 1996. Ontogeny of feeding morphology and kinematics in juvenile fishes: a case study of the cottid fish Clinocottus analis. J. Exp. Biol. 199, 1961–1971.
Ž .
Coughlin, D.J., 1991. Ontogeny of feeding behaviour on first feeding Atlantic salmon Salmo salar . Can. J. Fish. Aquat. Sci. 48, 1896–1904.
Ž .
Coughlin, D.J., 1994. Suction prey capture by clown-fish larvae Amphiprion perider-aion . Copeia 1, 242–246.
Ž .
Dabrowski, K.R., 1982. The influence of light intensity on feeding of fish larvae and fry II. Rutilus rutilus L.
Ž .
and Perca fluÕiatilis L. . Zool. Jb. Physiol. 86, 353–360.
Gibson, R.N., Ezzi, I.A., 1985. Effect of particle concentration on filter- and particulate-feeding in the herring Clupea harengus. Mar. Biol. 88, 109–116.
Govoni, J.J., Ortner, P.B., Al-Yamani, F., Hill, L.C., 1986. Selective feeding of spot, Leiostomus xanthurus, and Atlantic croaker, Micropogonias undulatus, larvae in the northern Gulf of Mexico. Mar. Ecol. Prog. Ser. 28, 175–183.
Houde, E.D., Schekter, R.C., 1980. Feeding by marine fish larvae: developmental and functional responses. Env. Biol. Fish 5, 315–334.
Ž . Ž .
Janssen, J., 1980. Alewives Alosa pseudoharengus and ciscoes Coregonus artedii as selective and
Ž .
non-selective planktivores. In: Kerfoot, W.C. Ed. , Evolution and Ecology of Zooplankton Communities. The University Press of New England, Hanover, pp. 580–586.
Jenkins, G.P., 1987. Comparative diets, prey selection, and predatory impact of co-occurring larvae of two flounder species. J. Exp. Mar. Biol. Ecol. 110, 147–170.
Magnhagen, C., 1985. Random prey capture or active choice? An experimental study on prey size selection in three marine fish species. Oikos 45, 206–216.
Meng, L., Orsi, J.J., 1991. Selective predation by larval striped bass on native and introduced copepods. Trans. Am. Fish. Soc. 120, 187–192.
Meyer, A., 1986. Changes in behaviour with increasing experience with a novel prey in fry of the central
Ž .
American cichlid, Cichlasoma managuense Teleostei: Cichlidae . Behaviour 98, 145–167.
Miller, T.J., Crowder, L.B., Rice, J.A., 1993. Ontogenetic changes in behavioural and histological measures of visual acuity in three species of fish. Environ. Biol. Fish. 37, 1–8.
Miller, T.J., Crowder, L.B., Rice, J.A., Binkowski, F.P., 1992. Body size and the ontogeny of the functional response in fishes. Can. J. Fish. Aquat. Sci. 49, 805–812.
Mills, E.L., Confer, J.L., Kretchmer, D.W., 1986. Zooplankton selection by young yellow perch: the influence of light, prey density and predator size. Trans. Am. Fish. Soc. 115, 716–725.
Pankhurst, P.M., Butler, P., 1996. Development of the sensory organs in the greenback flounder, Rhombosolea tapirina. Mar. Fresh. Behav. Physiol. 28, 55–73.
Paszkowski, C.A., Tonn, W.M., 1994. Effects of prey size, abundance, and population structure on piscivory by yellow perch. Trans. Am. Fish. Soc. 23, 855–865.
Ž .
Polo, A., Yufera, M., Pascual, E., 1992. Feeding and growth of gilthead seabream Sparus aurata L. larvae in relation to the size of the rotifer strain used as food. Aquaculture 103, 45–54.
Salgado, S.D., Hoyt, R.D., 1996. Early behaviour formation in fathead minnow larvae, Pimephales promelas: implications for sensory function. Mar. Fresh. Behav. Physiol. 28, 91–106.
Townsend, C.R., Risebrow, A.J., 1982. The influence of light level on the functional response of a zooplanktivorous fish. Oecologia 53, 293–295.
Tytler, P., Blaxter, J.H.S., 1988. The effects of external salinity on the drinking rates of the larvae of herring, plaice and cod. J. Exp. Biol. 138, 1–15.
Wahl, C.M., Einfalt, L.M., Hooe, M.L., 1995. Effect of experience with piscivory on foraging behavior and growth of walleyes. Trans. Am. Fish. Soc. 124, 756–763.
Ž .
Ware, D.M., 1971. Predation by rainbow trout Salmo gairdneri : the effect of experience. J. Fish. Res. Bd. Can. 28, 1847–1852.
