Feeding habits and diet overlap of marine fish larvae from the peri-Antarctic Magellan region

The Magellan region is a unique peri-Antarctic ecosystem due to its geographical position. However, the knowledge about the distribution and feeding ecology of fish larvae is scarce. Since this area is characterized by low phytoplankton biomass, we hypothesize that marine fish larvae display different foraging tactics in order to reduce diet overlap. During austral spring 2009–2010, two oceanographic cruises were carried out along southern Patagonia (50–56°S). Larval fish distribution and feeding of the two most widely distributed species were studied, the smelt Bathylagichthys parini (Bathylagidae) and black southern cod Patagonotothen tessellata (Nototheniidae). Larvae of B. parini showed a lower increase in the mouth gape at size, primarily feeding during daytime (higher feeding incidence during the day) mostly on nonmotile prey (invertebrate and copepod eggs, appendicularian fecal pellets, diatoms). They showed no increase in feeding success (number, total volume of prey per gut and prey width) with increasing larval size, and the niche breadth was independent of larval size. Larvae of P. tessellata showed a large mouth gape at size, which may partially explain the predation on motile prey like large calanoid copepods (C. simillimus) and copepodites. They are nocturnal feeders (higher feeding incidence during night) and are exclusively carnivorous, feeding on larger prey as the larvae grow. Nonetheless, niche breadth was independent of larval size. Diet overlap was important only in individuals with smaller mouth gape (<890 μm) and diminished as larvae (and correspondingly their jaw) grow. In conclusion, in the peri-Antarctic Magellan region, fish larvae of two species display different foraging tactics, reducing their trophic overlap throughout their development.     

Diet breadth variability in larval blue whiting as a response to plankton size structure

Diet breadth (measured as the S.D. of the log of prey size per larvae; SLH) of blue whiting micromesistius poutassou larvae followed a quadratic equation with larval size. In small larvae, diet breadth in terms of size (SLH), the mean and the maximum of the log of prey size per larvae (MLH and XLH, respectively) increased with larval size as prey size selection shifted to larger prey. In contrast, large larvae tended to reduce diet breadth of prey sizes ingested, focusing on the larger prey that were abundant, instead of raising the upper limit of prey sizes because of the low abundance of larger prey. Except for larvae at the onset of first feeding, number of prey stayed constant or decreased in relation to larval size. Both patterns (in small and large larvae) maintained a constant rate of increase of gut carbon content with increase in larval size. Large larvae appear to maintain the increase in gut carbon content during ontogenetic development by reducing diet breadth (SLH) and increasing selection towards the larger prey that are abundant.     

Morphological development of larval cobia Rachycentron canadum and the influence of dietary taurine supplementation

The morphological development of larval cobia Rachycentron canadum from 3 days post hatch (dph) until weaning (27 dph) was examined using S.E.M. Two groups of fish were studied: a control group (CF), reared under standard feeding protocol, and a group in which prey items were enriched with supplemental taurine (4 g l(-1) day(-1) ; TF). TF fish grew faster (P < 0·001), attained greater size (mean ±s.e. 55·1 ± 1·5 v. 33·9 ± 1·0 mm total length) and had better survival (mean ±s.e. 29·3 ± 0·4 v. 7·1 ± 1·2 %) than CF fish. Canonical variance analysis confirmed findings with respect to differences in growth between the treatment groups with separation being explained by two cranial measurements. S.E.M. revealed that 3 dph larvae of R. canadum (in both groups) possess preopercular spines, superficial neuromasts on the head and body, taste buds in the mouth, an olfactory epithelium which takes the form of simple concave depressions, and primordial gill arches. Gill filaments start to form as early as 6 dph and lamellae buds are visible at 8 dph in both groups. In CF fish, the cephalic lateral line system continues its development at 12-14 dph with invagination of both supra- and infraorbital canals. At the same time, a thorn-like or acanthoid crest forms above the eye. At 14 dph, invaginations of the mandibular and preopercular canals are visible and around 22 dph enclosure of all cranial canals nears completion. In CF larvae, however, completely enclosed cranial canals were not observed within the course of the trial, i.e. 27 dph. In TF larvae, grooves of the cephalic lateral line system form 4 days earlier than observed in CF larvae of R. canadum (i.e. at 8 dph), with enclosure commencing at 16 dph, and completed by 27 dph. Along the flanks of 6 dph larvae of either treatment, four to five equally spaced neuromasts delineate the future position of the trunk lateral line. As myomeres are added to the growing larvae, new neuromasts appear such that at 16 dph a neuromast is associated with each myomere. By 27 dph, the trunk lateral line starts to invaginate in CF larvae, while it initiates closure in TF larvae. These findings elucidate important features of the larval development of R. canadum and show that dietary taurine supplementation benefits larval development, growth and survival in this species. Moreover, they suggest a conditional requirement for taurine in larval R. canadum.     

Prey selection and functional morphology through ontogeny of Amphiprion clarkii with a congeneric comparison

The relationship between the complexity of the feeding apparatus and prey selection through ontogeny was examined in Amphiprion clarkii larvae. Larvae were reared from 1 to 10 days post-hatch (dph) on a diet of rotifers, wild-caught plankton and newly hatched Artemia sp. nauplii. Results were compared with available data on the relationship between functional morphology and prey selection of Amphiprion frenatus to establish patterns of functional morphology and prey selection between the larvae of two species of coral-reef fishes. Larvae of both species exhibited an increase in selection of larger prey through ontogeny coincident with an increase in the complexity of the feeding apparatus. The first elements to ossify in larvae of both species were the pharyngeal teeth ( A. clarkii : 5 dph, near ± s.d. Standard length, L _S,4· 3 ± 0· 2 mm; A. frenatus : 5 dph, L _S5· 0 ± 0· 4 mm) which, in combination with the development of a more functional feeding apparatus, may have permitted larvae to better process new types and sizes of prey. Prey items, however, were selected differentially between the two fish species, which could not be fully explained by the functional state of the feeding apparatus. While prey selection is influenced by the functional state of the feeding apparatus, all aspects of larval fish biology (morphology, behaviour and physiology) should be considered.     

Early development of New Zealand hapuku Polyprion oxygeneios eggs and larvae

This study describes for the first time the normal development of New Zealand hapuku Polyprion oxygeneios embryos and larvae reared from fertilization to 11 days post-hatch (dph) at a constant temperature. Fertilized eggs were obtained from natural spawnings from communally reared captive wild broodstock. Eggs averaged 2 mm in diameter and had single or multiple oil globules. Embryos developed following the main fish embryological stages and required an average of 1859·50 degree hours post-fertilization (dhpf) to hatch. The newly hatched larvae (4·86 mm mean total length, L(T) ) were undifferentiated, with unpigmented eyes, a single and simple alimentary tube and a finfold that covered the entire body. Larvae relied on the energy from the yolk-sac reserves until 11 dph (7·33 mm mean L(T) ), when yolk-sac reabsorption was almost completed. Some of the major developmental stages from hatching to yolk-sac reabsorption were eye pigmentation (5 dph), upper jaw formation (7 dph), lower jaw formation (8 dph) and mouth opening (8-9 dph). By 9 dph, the digestive system consisted of pancreas, liver, primordial stomach, anterior and posterior gut; therefore, P. oxygeneios larvae would be capable of feeding on live prey. The developmental, morphological and histological data described constitutes essential baseline information on P. oxygeneios biology and normal development.     

Feeding of larval blue whiting and Atlantic mackerel: a comparison of foraging strategies

Fish larvae employ different feeding strategies depending on area and season of spawning and hatching of larvae. Feeding and growth of larvae of blue whiting Micromesistius poutassou and mackerel Scomber scombrus from Porcupine Bank and the Celtic Shelf Break, west of Ireland, were compared based on prey concentrations in the environment and larval feeding behaviour. Both species were adapted to different environmental conditions. The mesopelagic blue whiting spawned in oceanic water that was well mixed. It was characterized by low production and low prey densities with minimum prey densities <1.0 organism 1⁻¹. Larvae of the Atlantic mackerel hatched later in the season in more productive water that was well stratified. Prey densities in the mackerel environment reached up to 1001⁻¹. Blue whiting larvae displayed a rather random distribution in the water column. Mackerel larvae <7 mm standard length ( L _s) were concentrated above the thermocline, while larvae >7 mm traversed the thermocline into deeper layers. Mackerel larvae >5mm L _s displayed marked cannibalism, exceeding 70%. Daily ration calculated on the basis of gut contents was rather low in both species: between 2.6 and 5.0% in blue whiting, but only 0.6 to 5.4% in mackerel. The results are discussed in relation to the respective environment both species encounter during their early larval life.     

Feeding habits of larval Mirogrex terraesanctae (Steinitz, 1952) in Lake Kinneret (Israel) I. Field study

Analysis of the food composition, distribution, lengths and dry weights of larval bleak, Mirogrex terraesanctae (Steinitz, 1952) in Lake Kinneret, Israel was conducted. Indices of electivity of ingested prey organisms were also determined. Larvae were surveyed throughout the 1989–90 reproductive seasons (Dec–Mar). Small larvae (<10.28 mm SL) preferentially select prey items smaller than 180 µ. Dominant zooplankters in the larval guts included: Bosmina longirostris, Ceriodaphnia reticulata and C. rigaudi, Asplanchna priodonita, Synchaeta oblonga, S. pectinata and juvenile cyclopoid copepods. Diaphanosoma brachiurum was rarely found (instars only) and no adult copepods were observed in larval guts. Some larvae specimens were found with up to 500 undigested (intact) accidentally ingested cells of Peridinium spp. per gut.     

不同寄主植物对三条橙灯蛾生长发育和繁殖的影响

在室内条件(温度(26±1)℃,相对湿度80%,光周期12L∶12D)下研究了天仙果(Ficus erecta Thunb.var.beecheyana(Hook.et Arn.)King)、大青(Clerodendrum cyrtophyllum Turcz)、华东油柿(Diospyros oleiera Cheng)、喜树(Camptotheca acuminate Decne)和杨梅(Myrica rubra(Lour.)Sieb.et Zucc)5种不同科的寄主植物对三条橙灯蛾Lemyra alikangensis(Strand)取食偏好、生长发育、繁殖及营养效应的影响。结果表明,三条橙灯蛾对不同寄主植物表现出明显的取食选择差异;不同寄主植物对三条橙灯蛾的幼虫发育历期、累计存活率、产卵量及营养效应等指标均有显著影响。三条橙灯蛾4龄幼虫偏好取食大青,对杨梅和喜树的选择率很低。幼虫历期以大青饲喂组最短为(39.20±1.79)d,杨梅饲喂组最长为(48.25±2.22)d;累计存活率以大青饲喂组最高(45.71%),杨梅饲喂组最低(28.57%);大青饲喂组产卵量最多为(553.75±61.69)粒/雌,杨梅饲喂组最少为(386.25±51.05)粒/雌。大青饲喂组三条橙灯蛾的相对生长率(0.30±0.03)mg·mg-1·d-1、食物利用率(13.46±1.84)%、近似消化率(62.04±6.29)%在各组中均最高,但食物转化率(21.65±0.95)%很低;杨梅饲喂组三条橙灯蛾的相对生长率(0.12±0.03)mg·mg-1·d-1、近似消化率(14.75±1.48)%很低,但是食物利用率(5.41±0.61)%较高,食物转化率(36.79±4.07)%最高。综合各指标认为:在5种供试植物中大青是三条橙灯蛾的最适寄主植物,杨梅则最不利于其生长发育和种群繁衍。     

Stable-isotope analyses reveal the importance of seagrass beds as feeding areas for juveniles of the speckled worm eel Myrophis punctatus (Teleostei: Ophichthidae) in Florida

The feeding habits and habitats of the speckled worm eel Myrophis punctatus were studied on the mangrove edge of the Indian River Lagoon (IRL, Florida) using gut-content and stable-isotope analyses of carbon (δ(13) C) and nitrogen (δ(15) N). Four taxa were identified through analyses of gut contents, and the index of relative importance suggested that amphipods, microphytobenthos and annelids are the most important food sources in the fish's diet. To assess the feeding habits of the fish after their recruitment to the IRL, these food sources were collected from mangroves and nearby seagrass beds for isotope analyses. Stable isotopes constituted a powerful tool for discriminating fish prey items from mangroves (mean ± s.d.δ(13) C = -20·5 ± 0·6‰) and those from seagrass beds (mean ± s.d.δ(13) C = -16·9 ± 0·6‰), thus providing good evidence of food source origins. The 56 M. punctatus collected [10·0 < total length (L(T) ) < 16·2 cm] had average isotopic signatures of δ(13) C = -16·7 ± 0·2‰ and δ(15) N = 8·2 ± 0·1‰. A significant depletion in (13) C was observed for larger juveniles (15·0 < L(T) < 16·2 cm), suggesting that they found a portion of their food in mangroves. Estimation of the trophic level from stable isotopes (T(Liso)) was similar among different size groups of juvenile fish (T(Liso) = 3·2-3·5); therefore, M. punctatus was considered a secondary consumer, which is consistent with its zoobenthic diet. The concentration-dependent mixing Stable Isotope Analysis in R (SIAR) model revealed the importance of food sources from seagrass beds as carbon sources for all the fish collected, with a significant increase in mangrove prey contributions, such as annelids, in the diet of larger juveniles. This study highlights the importance of seagrass beds as feeding habitats for juveniles of M. punctatus after their recruitment to coastal waters.     

Diet of labrisomid blenny Auchenionchus variolosus (Valenciennes, 1836) (Labrisomidae) during its larval development off central Chile (2012–2013)

The diet, feeding success (prey number and total volume per gut, and maximum prey width) and trophic niche width of the labrisomid blenny Auchenionchus variolosus (Valenciennes, 1836) was studied during its larval development (3.93–17.26 mm standard length). Individuals were collected in October 2012 and October 2013 in nearshore waters (<500 m offshore) from Bahía Valparaíso, central Chile with Bongo nets. When compared to the same length range, larval A. variolosus collected during 2013 showed larger upper jaw‐at‐sizes than those from 2012. This coincided with a reduction in relative importance (%IRI) of the main prey item, copepod nauplii, from ~90 to ~73%. Feeding incidence (FI) was high throughout larval development, varying from 97.3 to 100%, being similar throughout the larval development. Prey items per gut (PIPG) ranged from 0 to 35 prey, showing no differences among years. Total volume per gut (TVPG) was positively correlated with larval length, and at given larval size, TVPG ingested by larval A. variolosus was larger during 2013 (0.0607 mm³) than during 2012 (0.0301 mm³). Prey width range was 47.47–700.94 μm and was positively correlated with standard length (SL). Niche breadth was independent of larval size and did not change during larval development in 2012 or 2013. The study helps to understand the trophic interactions occurring in nearshore waters off rocky reef environments from mid‐latitudes of the Southeast Pacific coasts.     

Effect of delayed first feeding on development and feeding ability of Paralabrax maculatofasciatus larvae

The impact of delayed feeding early in development on late feeding ability and development of spotted sand bass Paralabrax maculatofasciatus larvae was examined. Larvae were sampled from hatching until day 19 after delayed feeding for zero (control), 1, 2, or 3 days. Feeding incidence was evaluated as the percentage of larvae with food in the gut and feeding intensity was measured by direct counting of prey after dissection of the gut. Delayed feeding effects due to starvation were observed early in development. By day 5, significant differences ( P  < 0·05) were observed between controls and larvae submitted to degrees of delayed feeding, including total length ( L _T), eye diameter, and the size of the head, liver, gut, muscles and body. Differences were still apparent in L _T, and body, liver and muscle heights at the end of the studied period. Larvae under total starvation did not survive beyond day 5. Initial feeding incidence was 35, 60, 90 and 10% for larvae first fed on day 2, 3, 4 or 5, respectively. Mean ±  s . d . feeding intensity was 3·6 ± 0·8, 2·8 ± 1·3, 5·2 ± 0·3, and 10·2 ± 1·5 rotifers per larva depending on whether larvae were initially fed on day 2, 3, 4 or 5, respectively. The point of no return occurred between 4 and 5 days after hatching.     

Vulnerability of larval white perch,Morone americana,to fish predation

Synopsis The vulnerability of white perch, Morone americana, larvae to yearling bluegill, Lepomis macrochira, predators was examined in relation to larval size, nutritional condition and relative abundance of alternative prey. Short-term (15 min) predation trials were conducted in 381 tanks in the laboratory. Larval vulnerability was measured as the proportion of larvae killed and the proportion of successful attacks per predator in each 15 min trial. No significant differences in vulnerability were apparent among larvae regardless of feeding history at sizes < 6 mm SL. At larval lengths > 6 mm SL, size of larvae was the crucial determinant of their vulnerability. Percentage of larvae killed in 15 min was nearly 100% at sizes < 6 mm SL, decreased to 30% at a length of 12.0 mm SL and dropped to 18% at 14.0 mm SL. Larvae initially feeding at low food levels for 2–4 d exhibited decreased growth of 13–25% over the first 3 wks of life, and simulations based on laboratory results indicated that these growth deficits could result in 5- to 68-fold decreases in survivorship at 38 days after hatching (DAH). The relative abundance of alternative prey also had a pronounced effect on mortality of larvae. A 10-fold increase in alternative prey (Daphnia magna) abundance decreased bluegill predation rates on white perch larvae by 10–20%, while a 100-fold increase in Daphnia density decreased larval mortality by 75–90%.     

Quality of two aphid species (Rhopalosiphum padi and Sitobion avenae) as food for the generalist predator Tachyporus hypnorum (Col., Staphylinidae)

Abstract:  The quality of two aphid species ( Rhopalosiphum padi and Sitobion avenae ) as food for the staphylinid generalist predator Tachyporus hypnorum was investigated. Fruit flies ( Drosophila melanogaster ) were used as prey for comparison. T. hypnorum has been reported to prey on aphids, to consume large quantities and to show a high preference for aphids. This study showed that compared with fruit flies aphids are not high-quality prey for neither adult females nor larvae of T. hypnorum . Larvae were less able to utilize aphids than adult females. Reproduction was affected by both aphid species; a diet of S. avenae reduced fecundity, and a diet of R. padi reduced egg hatching success and prolonged the time spent in the egg stage compared with a fruit fly diet. For the T. hypnorum larvae pure diets of both S. avenae and R. padi resulted in high larval mortality. Overall ranking of the three prey types based on several fitness parameters differed between larvae and adults. The adult fitness parameters gave rather conflicting rankings of all three prey types though the two aphid species were of approximately the same overall value, whereas the larval fitness parameters gave identical rankings: D. melanogaster  >  S. avenae  >  R. padi . In the larval stage survival was greatly reduced by both aphid diets, while this was not the case for the adult females.     

Dietary niche breadth in a local community of passerine birds: an analysis using phylogenetic contrasts

The analysis of a local community of forest passerines (13 species) using phylogenetic contrasts shows a correlation between body size of bird species and mean prey size, minimum prey size, maximum prey size and the size range of dietary items. This suggests that larger birds drop small prey taxa from their prey list, because of the difficulty of capturing very small prey, for energetic reasons or because of microhabitat usage. We find some support for the third hypothesis. Dietary niche breadth calculated across prey taxa is not related to body size. Dietary niche breadth, however, is correlated with size-corrected measurements of the bill and locomotor apparatus. Long and slender bills increase the dietary niche breadth. Thus subtle differences constrain foraging and the techniques of extracting certain prey taxa form crevices. Dietary niche breadth and foraging diversity are positively correlated with population density: at least locally dietary generalists occur at higher breeding densities than specialists.     

Experimental predation studies of malacophagous larvae of Sepedon fuscipennis (Diptera: Sciomyzidae) and aquatic snails

Experimental studies were conducted under both laboratory and field conditions to determine the effects of prey density, three levels of prey aggregation, water depth, and predator density on the number of snails killed per larva of Sepedon fuscipennis. Of these factors, predation rates were most influenced by prey density and water depth. The number of small (2–4.5 mm) Lymnaea palustris killed per larva of S. fuscipennis increased at a decreasing rate as prey density increased under shallow water conditions. Larvae killed a mean of 14 snails at a prey density of 200/m², while an average of 24 snails were killed per larva of S. fuscipennis at a prey density of 4000/m². This functional response to prey density was largely confined to third-instar larvae, and as water depth was increased the response was not apparent.A field study in which larval densities of S. fuscipennis were manipulated showed that the population density of smaller individuals of L. palustris (< 4.5 mm) was reduced when predator density was increased. Populations of Physa integra, Gyraulus parvus, and larger L. palustris were not significantly reduced by the malacophagous larvae at the levels tested.     

Parasitic copepods affect morphospace and diet of larvae of a temperate reef fish

The effects of ectoparasites on larvae of the clingfish Gobiesox marmoratus were evaluated at the dietary and morphometric levels. The larvae and ectoparasites were collected by nearshore plankton samplings during October, November and December 2013 off El Quisco Bay, central Chile. The standardized abundance of total larvae and those ectoparasitized larvae (PL) was positively related and high parasite prevalence was found throughout the sampling period (up to 38%). Geometric morphometrics analyses indicate main changes in the shape through early ontogeny and subtle but significant variations between PL and non‐parasitized larvae (NPL). Prey composition varied between PL and NPL; small size (<6 mm standard length, L_S) parasitized larval G. marmoratus ate mostly gastropod larvae, whereas small non‐parasitized specimens ate mainly cirripede nauplii. All larger (>8 mm L_S), pre‐settlement stages parasitized by Trifur and, or Caligus copepods had content in their gut, suggesting that ectoparasites did not diminish prey capture in host with larger size. Morphometric and dietary changes occurring during larval development were decoupled, both for PL and NPL. The maintenance of a slender, more hydrodynamic body through pelagic development and the ingestion of less‐mobile prey in PL suggests non‐lethal effects of ectoparasitism on rocky‐reef fish larvae.     

Vertical distribution and feeding of larval blue whiting in turbulent waters above Porcupine Bank

In April 1995 a patch of blue whiting larvae Micromesistius poutassou was found at low illumination levels below 20 m depth near Porcupine Bank, west of Ireland, together with high densities of copepod nauplii and reduced turbulence rates, suggesting that larval blue whiting vertical distribution was determined by prey concentration, illumination and turbulence. Most (83·8%) larvae (2·0–7·5 mm L _s) had food in their guts. Feeding incidence and feeding intensities increased with increasing larval length. Only larvae >5·5 mm reduced numerical in favour of weight-based feeding intensity, indicating a shift in dietary composition. Maxima of the diel rhythms of feeding incidence and intensities occurred at 1800 and 2100 hours and minima at dawn (0600 hours). Proportionately, more nauplii were eaten by day but more copepod eggs and tintinnids at night. The distinct diel pattern in larval blue whiting feeding suggests that any analysis of factors mediating feeding must take into account diel feeding cycles. Larval feeding was significantly affected by wind speed. The larvae ate more and larger items at low than at higher turbulence levels. The data suggest that the maximum level of turbulence was not beneficial for larval blue whiting, but more moderate wind speeds could have had an enhancing effect on larval feeding success.     

The feeding ecology of Morone americana larvae in the Chesapeake Bay estuarine turbidity maximum: the influence of physical conditions and prey concentrations

Feeding ecology of white perch Morone americana larvae, a major component of the ichthyoplankton community in the estuarine turbidity maximum (ETM) region of upper Chesapeake Bay, was evaluated in years of contrasting physical conditions. In 1998, high river flow led to an ETM with high turbidity and pronounced salinity stratification. In 1999, under low river flow conditions, turbidities in the ETM region were lower and the salt front was located 15 km up‐stream of its location in 1998. Copepodites and adults of Eurytemora affinis were the predominant prey in guts of all length classes of larvae (3·2–9·8 mm standard length, L _S) that were examined. Repeated‐measures, multiple regression analyses were conducted to evaluate the influence of several factors on feeding success by determining if physical conditions (temperature, current velocity, salinity, turbidity and light) and prey concentrations explained variability in mean gut fullness (gut contents mass per body mass) of small (<5 mm L _S) and large (>5 mm L _S) white perch larvae. Eurytemora affinis concentrations were significant in the statistical models for both small and large larvae. High concentrations of E. affinis , which enhanced encounter rates of white perch larvae with prey, may have been the most important factor determining feeding success in the ETM region. Larval feeding incidence (percentage of guts with food) was higher in 1998, suggesting that annual variability in river flow influenced larval feeding success by controlling physical and biological conditions in the ETM region.     

Small-scale patterns in distribution and feeding of Atlantic mackerel (Scomber scombrus L.) larvae in the Celtic Sea with special regard to intra-cohort cannibalism

Short-term variability in vertical distribution and feeding of Atlantic mackerel (Scomber scombrus L.) larvae was investigated while tracking a larval patch over a 48-h period. The patch was repeatedly sampled and a total of 12,462 mackerel larvae were caught within the upper 100 m of the water column. Physical parameters were monitored at the same time. Larval length distribution showed a mode in the 3.0 mm standard length (SL) class (mean abundance of 3.0 mm larvae =75.34 per 100 m³, s=34.37). Highest densities occurred at 20–40 m depth. Larvae <5.0 mm SL were highly aggregated above the thermocline, while larvae ≥5.0 mm SL were more dispersed and tended to migrate below the thermocline. Gut contents of 1,177 mackerel larvae (2.9–9.7 mm SL) were analyzed. Feeding incidence, mean number (numerical intensity) and mean dry weight (weight-based intensity) of prey items per larval gut were significantly dependent on larval size. However, while weight-based feeding intensities continued to increase with larval length, numerical intensity peaked at 4–4.9 mm SL, indicating a shift in the larval diet. While first-feeding larvae relied most heavily on copepod nauplii and eggs, larvae ≥5.0 mm SL initiated piscivorous feeding. All identifiable fish larvae were Atlantic mackerel. Thus, the piscivory was cannibalism. Larval feeding incidence and numerical feeding intensities peaked during daytime and were reduced at night. Daily ration estimates for first-feeding mackerel larvae <4.0 mm SL were extremely low = 1.43% body dry weight, but increased dramatically at 5.0 mm SL, i.e., at the onset of cannibalism, reaching >50% body dry weight in larva ≥8.0 mm SL. Received in revised form: 31 October 2000 Electronic Publication     

Consumption and gut evacuation rate of laboratory‐reared spotted seatrout (Sciaenidae) larvae and juveniles

The temperature and mass dependence of maximum consumption rate was measured for larval and early juvenile spotted seatrout Cynoscion nebulosus . Maximum consumption ( C _MAX) estimates were obtained from feeding and gut evacuation experiments on larvae (3·8–19 mm standard length, L _S) at three temperatures (24, 28 and 32° C), and maximum consumption experiments on juveniles at three temperatures (20, 26 and 32° C). Feeding levels were determined for larvae fed live prey ( Brachionus plicatilis and Artemia salina ) ad libitum . The midgut and total evacuation times were estimated for fish feeding continuously and discontinuously using alternate meals of tagged and untagged live prey. Temperature and fish size had significant effects on gut evacuation and consumption. The gut evacuation time increased with increasing fish size, and decreased with increasing temperatures. Mass‐specific midgut contents increased for small larvae <0·156 mg dry mass ( M _D)( c . 4 mm L _S), and decreased for larger larvae and juveniles. Maximum consumption was modelled by fitting a polynomial function to a reduced dataset of individuals feeding at high levels. The C _MAX model predicted an initial increase in specific feeding rate from 70 to 155% M _D day⁻¹ for small larvae, before declining for larger larvae and juveniles.