Bivalve Macoma Balthica Research Articles

Genetic architecture in a marine hybrid zone: comparing outlier detection and genomic clines analysis in the bivalve Macoma balthica

The role of natural selection in speciation has received increasing attention and support in recent years. Different types of approaches have been developed that can detect genomic regions influenced by selection. Here, we address the question whether two highly different methods--F(ST) outlier analysis and admixture analysis--detect largely the same set of non-neutral genomic elements or, instead, complementary sets. We study genetic architecture in a natural secondary contact zone where extensive admixture occurs. The marine bivalves Macoma balthica rubra and M. b. balthica descend from two independent trans-Arctic invasions of the north Atlantic and hybridize extensively where they meet, for example in the Kattegat-Danish Straits-Baltic Sea region. The Kattegat-Danish Straits region forms a steep salinity cline and is the only entrance to the recently (ca. 8000 years ago) established brackish water basin the Baltic Sea. Salinity along the contact zone drops from 30‰ (Skagerrak, M.b.rubra) to 3‰ (Baltic, M.b.balthica). Both outlier analysis and genomic clines analysis suggest that large parts of the genome are influenced by non-neutral effects. Contrasting samples from well outside the hybrid zone, outlier analysis detects 16 of 84 amplified fragment length polymorphism markers as significant F(ST) outliers. Genomic clines analysis detects 31 of 84 markers as non-neutral inside the hybrid zone. Remarkably, only three markers are detected by both methods. We conclude that the two methods together identify a suite of markers that are under the influence of non-neutral effects.

Growth and production of the bivalve Macoma balthica (Linnaeus, 1758) (Cardiida: Tellinidae) in the southeastern part of the Baltic Sea

The growth and production of the Baltic clam Macoma balthica in the southeastern part of the Baltic Sea were studied. The shell length of M. balthica reached 23.5 mm, the maximum age was 13+ years. The linear growth was described by the von Bertalanffy equation for shallow-water area (depths 9–40 m): L τ = 23.99(1 − e −0.1293(τ − (−0.9578))), and for the deep-water area (41–81 m): L τ = 20.61(1 − e −0.1813(τ − (−0.5608))). The annual production was lower (25.35 ± 1.72 kJ/m2) in the shallow-water area than in the deep-water area (71.23 ± 4.48 kJ/m2), with values of P s /B ratio 0.44 and 0.38, respectively.

Deposit-feeders accumulate the cyanobacterial toxin nodularin

Blooms of toxic cyanobacteria may potentially affect food web productivity and even be a human health hazard. In the Baltic Sea, regularly occurring summer blooms of nitrogen-fixing cyanobacteria are often dominated by Nodularia spumigena, which produces the potent hepatotoxin nodularin. Evidence of sedimentation of these blooms indicates that benthic fauna can be exposed to nodularin. In a one month experiment, we simulated the settling of a summer bloom dominated by N. spumigena in sediment microcosms with three species of sediment-dwelling, deposit-feeding macrofauna, the amphipods Monoporeia affinis and Pontoporeia femorata and the bivalve Macoma balthica, and analyzed nodularin in the animals by HPLC–ESI–MS (high-performance liquid chromatography–electrospray ionization–mass spectrometry). We found nodularin in quantities of 50–120 ng g −1 DW. The results show that deposit-feeding macrofauna in the Baltic Sea may contribute to trophic transfer of nodularin.

Relative sea level changes during the Holocene in the Sisimiut area, south‐western Greenland

AbstractHolocene relative sea level (RSL) changes have been investigated by analysing and dating isolation sequences from five lakes near Sisimiut in south‐western Greenland. The transitions between marine and lacustrine sediments were determined from elemental analyses and analyses of macroscopic plant and animal remains. Radiocarbon dating was used to provide minimum ages for the transitions and to construct a RSL curve. Dating of a shell of the marine bivalve Macoma balthica indicates that deglaciation of the lowlands occurred in the early Holocene, at around 10 900 cal a BP. The RSL curve shows initial rapid regression from the marine limit at around 140 m, implying strong glacio‐isostatic rebound. We suggest that the margin of the Greenland Ice Sheet was located at the shelf break during the Last Glacial Maximum. Frequent remains of the ostracode Sarcypridopsis aculeata, which is a thermophilous brackish water species that is unknown from the extant fauna of Greenland, in one of the basins around 8500 cal a BP may mark the beginning of the Holocene thermal maximum in the region. Copyright © 2011 John Wiley & Sons, Ltd.

Intraspecific egg size variation and sperm limitation in the broadcast spawning bivalve Macoma balthica

Broadcast spawners are exceptionally suited and simple models for studying parental investment in offspring, because direct post-spawning investment is nonexistent. However, a comprehensive understanding of the large variation that exists in their egg sizes is still lacking. One of the main hypotheses states that variation in fertilization conditions underlies some of the egg size variation, as larger eggs are larger targets for sperm. Here, we test the hypothesis that egg size may be locally tuned to expected ambient sperm concentrations during fertilization. In accordance with the hypothesis, we find that in the bivalve Macoma balthica (L.) adult density as a proxy for sperm concentration correlates strongly (correlation coefficient −0.87) with egg size in the field. Optimisation modeling confirms the negative relationship between optimal egg size and sperm concentration for M. balthica and this is independent of the fertilization model used. Discrepancies between models and observations remaining include larger egg sizes overall and a concave predicted relationship that is not obvious in the data. The results suggest that in M. balthica sperm limitation may play a role in fertilization success and in shaping egg size variation, and that locations with high population densities may make disproportionately large contributions to the next generation.

Indirect effects of non-lethal predation on bivalve activity and sediment reworking

Deposit-feeders are the dominant bioturbators of aquatic sediments, where they profoundly impact biogeochemical processes, but they are also vulnerable to both lethal and non-lethal predation by a large variety of predators. In this study, we performed a series of experiments to test the effects of predation avoidance on the feeding activity and sediment reworking intensity of the deposit-feeding bivalve Macoma balthica. Feeding activity at the sediment–water interface and sediment reworking intensity (vertical displacements of inert particle tracers) were monitored using image analysis techniques for treatments including and excluding the predatory shrimp, Crangon crangon. Detection of C. crangon by M. balthica resulted in an immediate retraction of the feeding siphon and a reduction in feeding activity. M. balthica also buried deeper into the sediment in the presence of C. crangon. This predator avoidance behaviour indirectly affected sediment reworking modes and rates, increasing the thickness of the bioturbated sediment layer as well as the non-local transport of sediment particles at depth. Conversely, feeding activity and sediment reworking processes remained unaffected when C. crangon was present, but isolated from the sediment, suggesting that predator perception in M. balthica is tactile (i.e. induced by direct encounter) rather than being chemosensory. Collectively, these results demonstrate that predatory avoidance behaviour by benthic infauna can significantly impact benthic bioturbation and the incorporation of organic matter into the benthic food web.

Brown shrimp ( Crangon crangon, L.) functional response to density of different sized juvenile bivalves Macoma balthica (L.)

Variability in infaunal bivalve abundance in the Wadden Sea is largely determined by recruitment variability. Post-settlement, but pre-recruitment bivalve mortality is high and related to the occurrence of their most abundant predator, the brown shrimp Crangon crangon. To investigate if the mortality patterns of newly settled bivalves can be explained by the foraging behavior of brown shrimp, we carried out experiments on shrimp functional response to three size classes of juveniles of the Baltic Tellin Macoma balthica. The functional response curves for all three prey sizes (0.62 mm, 0.73 mm, and 0.85 mm) were the hyperbolic Holling's type II. The attack rate was highest for the smallest prey size ( a = 0.31, medium and large prey a = 0.22); the handling time was longest for the largest prey size ( T h = 29 s, small and medium prey T h = 15 s). Thus, a large body size is advantageous for the bivalves over the whole density range. Knowledge of individual foraging behavior is needed to model predation mortality of bivalves. The consumption rates in the experiment were theoretically high enough to account for M. balthica mortality in the field.

Macrobenthos recruitment success in a tidal flat: Feeding trait dependent effects of disturbance history

Juvenile growth rate and production of the bivalves Macoma balthica and Cerastoderma edule and the polychaete Nereis diversicolor were investigated on two contrasting tidal flat sediments that represented different recovery stages following disturbance. Juvenile M. balthica and N. diversicolor grew faster and yielded a higher production in the early succession community where fresh microalgal food supply was higher and physical stress was lower in comparison with the mature community, while C. edule recruitment success did not differ between recovery stages. Growth rate responses were positively related to the consumer specific microalgal dietary requirement suggesting the roles of microalgal food availability and interference within the mature community on juvenile feeding efficiency. In addition to the enhanced growth, lower post-settlement dispersal in the more stable sediments of the early succession community also contributed substantially to the higher cohort production of M. balthica and N. diversicolor at early recovery stages. This study emphasises that the recruitment success of tidal flat macrobenthic juveniles is both context- and trait dependent (i.e. disturbance history and feeding trait). As recruitment is the foundation upon which all subsequent interactions within the community take place, our results illustrate the important role of biotic–physical interactions that affect food supply and sediment stability to tidal flat benthic community dynamics.

Erodibility of a mixed mudflat dominated by microphytobenthos and Cerastoderma edule, East Frisian Wadden Sea, Germany

Sediment erodibility and a range of physical and biological parameters were measured at an intertidal site in the German Wadden Sea area in June, September and November 2002 and February and April 2003 in order to examine the influence of macrozoobenthos and microphytobenthos on sediment erodibility and the temporal variation. The study site was a mixed mudflat situated in the mesotidal Baltrum–Langeoog tidal basin at the East Frisian barrier coast. The mud content at the site was about 35% and the filter-feeding cockle Cerastoderma edule was the dominating macrozoobenthic species (by biomass). The erodibility of the sediment showed strong temporal variation with high erosion thresholds in spring and late summer and significantly lower thresholds during the rest of the study period. The erosion thresholds were strongly dependent on the contents of chlorophyll a (chl a) and colloidal carbohydrates, both indicators of the content of microphytobenthos, in this environment primarily benthic diatoms. The content of microphytobenthos was high in September 2002 and April 2003, and regression analysis indicated that this was the only likely reason for the low erodibility found at these times. A biostabilisation index of about 4.5 was found for a situation with both abundant biofilms and cockles. A direct influence of Cerastoderma edule on erodibility was not observed, in contrast to other recent studies. The presence of C. edule at the site results in biodeposition of fine-grained material and the presence of C. edule will therefore probably increase the content of fine-grained sediments at the surface compared to an abiotic situation. Increasing the amount of fine-grained material in mixed sediments has previously been shown to reduce the erodibility of the sediments and C. edule will therefore in this way indirectly stabilize the bed. However, although C. edule may constitute the main part of the biomass at some intertidal sites, other and more vigorous bioturbators and deposit-feeding species (e.g., the bivalve Macoma balthica, the gastropod Hydrobia ulvae or the amphipod Corophium volutator) may completely hide its effect on sediment erodibility if these species are present in high numbers.

Settlement of Macoma balthica larvae in response to benthic diatom films

The role of multi-species benthic diatom films (BDF) in the settlement of late pediveliger larvae of the bivalve Macoma balthica was investigated in still-water bioassays and multiple choice flume experiments. Axenic diatom cultures that were isolated from a tidal mudflat inhabited by M. balthica were selected to develop BDF sediment treatments characterized by a different community structure, biomass, and amount of extracellular polymeric substances (EPS). Control sediments had no added diatoms. Although all larvae settled and initiated burrowing within the first minute after their addition in still water, regardless of treatment, only 48–52% had completely penetrated the high diatom biomass treatments after 5 min, while on average 80 and 69% of the larvae had settled and burrowed into the control sediments and BDF with a low diatom biomass (<3.5 μg Chl a g−1 dry sediment), respectively. The percentage of larvae settling and burrowing into the sediment was negatively correlated with the concentration of Chl a and EPS of the BDF. This suggests higher physical resistance to bivalve penetration by the BDF with higher diatom biomass and more associated sugar and protein compounds. The larval settlement rate in annular flume experiments at flow velocities of 5 and 15 cm s−1 was distinctly lower compared to the still-water assays. Only 4.6–5.8% of the larvae were recovered from BDF and control sediments after 3 h. Nonetheless, a clear settlement preference was observed for BDF in the flume experiments; i.e., larvae settled significantly more in BDF compared to control sediments irrespective of flow speed. Comparison with the settlement of polystyrene mimics and freeze-killed larvae led to the conclusion that active selection, active secondary dispersal and, at low flow velocities (5 cm s−1), passive adhesion to the sediment are important mechanisms determining the settlement of M. balthica larvae in estuarine biofilms.

Some like it cold: populations of the tellinid bivalve Macoma balthica (L.) suffer in various ways from a warming climate

Because of its relatively low tolerance to elevated temperatures, Macoma balthica (L.) may be one of the first important marine species in temperate coastal areas to suffer from a warming climate. For the last few decades, the abundance of the species in the Dutch Wadden Sea has been se- riously declining. At lower latitudes, the southern edge of its range recently shifted several 100s of km to the north. To understand changes in abundance related to high temperatures, we studied short-term population responses to warmer than average seasons using data from a long-term (>35 yr) sampling program in the western Wadden Sea. The observed relationships included reduction of reproductive output and recruitment in years starting with mild winter-spring periods, reduction of adult survival in years with warm summers, reduction of annual migration in ~8 mo olds to more favourable areas in mild winters, reduction of growth rates in warmer than average growing seasons, and enhancement of seasonal weight loss in milder than average winters. We conclude that elevated temperatures nega- tively affect population dynamics of M. balthica in a number of ways, i.e. via recruitment or mortality as well as growth. These negative influences were observed already far below the upper lethal tem- perature of the species and in populations living ~1000 km poleward of the warm edge of the species' range. The outcome of M. balthica's energy balance appears to be decisive for their continued exis- tence by effects not only on individual weights but also on mortality as well as reproduction.

Inheritance of predominantly hidden shell colours in Macoma balthica (L.) (Bivalvia:Tellinidae)

Adults of the bivalve Macoma balthica live burrowed in mud or sand, which makes their shells invisible so that a vision-related selective advantage of colour morphs is hard to imagine. Nevertheless, this species harbours a conspicuous colour polymorphism of white, yellow, orange and red. Here we present data from laboratory-reared crosses that demonstrate a genetic basis for this colour polymorphism. The results suggest a simple model of inheritance of colour class: four alleles at a single locus displaying linear hierarchy of dominance, in the following order of increasing dominance: white < yellow < orange < red. Shell colour intensity is also shown to be a heritable trait in M. balthica. Furthermore, bicolouration may to a degree be a heritable trait, but the sample size is too small to draw any conclusions about possible inheritance mechanisms. Heritable colour variation can now, with the proposed model, be used as a genetic marker. Furthermore, the proposed model will provide the basis for further analysis of the evolutionary forces maintaining this colour polymorphism that is hidden from view for most of the life of this bivalve.

Behavioral effects of low dissolved oxygen on the bivalve Macoma balthica

Hypoxia, a dissolved oxygen concentration (DO) below 2 mg l – 1 , is a significant stressor in many estuarine ecosystems. Many sedentary organisms, unable to move to avoid hypoxic areas, have metabolic and behavioral adaptations to hypoxic stress. We tested the effects of hypoxia on the behavior and mortality of the clam Macoma balthica, using four levels of dissolved oxygen in flow-through tanks. We used five replicates of each of four treatments: (1) Hypoxic (DO mean ± SE = 1.1 ± 0.06 mg O 2 l – 1 ), (2) Moderately hypoxic (DO 2.6 ± 0.05 mg O 2 l – 1 ), (3) Nearly normoxic (DO 3.2 ± 0.04 mg O 2 l – 1 ), (4) Normoxic (DO = 4.9 ± 0.13 mg O 2 l – 1 ). We lowered the dissolved oxygen with a novel fluidized mud-bed, designed to mimic field conditions more closely than the common practice of solely bubbling nitrogen or other gasses. This method for lowering the DO concentrations for a laboratory setup was effective, producing 1.4 l min –1 of water with a DO of 0.8 mg O 2 l – 1 throughout the experiment. The setup greatly reduced the use of compressed nitrogen and could easily be scaled up to produce more low-DO water if necessary. The lethal concentration for 50% of the M. balthica population (LC 50) was 1.7 mg O 2 l – 1 for the 28-day experimental period. M. balthica decreased its burial depth under hypoxic and moderately hypoxic (~2.5 mg O 2 l – 1 ) conditions within 72 hours of the onset of hypoxia. By the sixth day of hypoxia the burial depth had been reduced by 26 mm in the hypoxic tanks and 10 mm in the moderately hypoxic tanks. Because reduced burial depth makes the clams more vulnerable to predators, these results indicate that the sub-lethal effects of hypoxia could change the rate of predation on M. balthica in the field.

Uptake of diatoms in Baltic Sea macrozoobenthos during short-term exposure to severe and moderate hypoxia

The effect of severe and moderate hypoxia on food uptake of benthic macrofauna was studied in the laboratory. The hypothesis was that low oxygen concentrations negatively affect feed- ing at oxygen levels that have little effect on the studied animals' survival. The bivalve Macoma balthica, the priapulid Halicryptus spinulosus, the amphipods Monoporeia affinis (subadult & juve- nile) and Pontoporeia femorata (subadult) were offered the 14 C-labelled diatom Skeletonema costa- tum in 0.8 to 10.6 mg O2 l -1 . Feeding was measured as radioactivity uptake. Subadult amphipods were studied one species at a time (single) or together (mixed). Feeding changed in all amphipods at the lowest oxygen concentrations, but no effect was found for M. balthica and H. spinulosus. At the lowest concentration (0.8 mg O2 l -1 ) feeding by subadult M. affinis (single) was only 17% of the full oxygen saturation (10.6 mg O2 l -1 ), and, at 1.6 mg O2 l -1 , 14% of the feeding at 8.9 mg O2 l -1 . Juvenile M. affinis consumed more labelled algae at 3 mg O2 l -1 than at higher oxygen concentrations. M. balthica feeding was not affected. Little radioactivity uptake was registered for H. spinulosus at any oxygen concentration, showing that H. spinulosus is not a surface deposit feeder. The amphipods were the most sensitive to week-long oxygen deficiency. Survival decreased significantly in the lowest oxygen concentrations (0.8; 1.6 mg O2 l -1 ). Of the subadult M. affinis 15 and 65%, respectively, survived, compared with 0 and 58% for P. femorata. Juvenile M. affinis mortality was high in all oxygen concentrations, whereas most M. balthica and H. spinulosus survived.

Bioturbation changes the patterns of benthic emergence in zooplankton

We investigated the effects of two functionally different deposit feeders, the amphipod Monoporeia affinis and the bivalve Macoma balthica, on the benthic emergence of copepods (Acartia spp., Eurytemora affinis, Temora longicornis), cladocerans (Bosmina longispina maritima, Daphnia spp.), and rotifers (Synchaeta spp., Keratella spp., Notholca spp.) in the northern Baltic Sea. We performed 3.5‐week laboratory experiments in June‐July and in September‐October using surface sediment containing zooplankton eggs and various combinations of amphipod and bivalve densities. Both zoobenthic species increased egg mortality rates in Synchaeta spp. and in calanoids. Each amphipod consumed a maximum of one Synchaeta spp. egg and 31 calanoid eggs per day, while the bivalves consumed six Synchaeta and 143 calanoid eggs per day. Monthly consumption throughout the community matches or even exceeds the field densities of eggs. In summer, M. affinis decreased the daily emergence rate of Synchaeta spp. by 31‐53% and of M. balthica by 57%. The effect on calanoids was less severe: the emergence rate of Acartia spp. and E. affinis was significantly suppressed only by M. balthica (the reduction was 48% and 28% in Acartia spp. and in E. affinis, respectively). T. longicornis was the only species that benefited from macrofauna: together the two zoobenthic species resulted in a 2.7‐fold increase in the emergence of this calanoid. Bioturbation significantly influences the benthic emergence of zooplankton, with effects varying among species, and may thus affect the dynamics of pelagic populations.

Biocenoses of Anadyr Estuary (Bering Sea) and conditions of their existence

Some bottom biocenoses in the river mouth area of Anadyr Estuary are studied. Seasonal biocenosis of filamentous algae Acrosiphonia spp. occurs in the intertidal zone, biocenosis of the bivalve Macoma balthica is common at a depth of 4–8 m. Species diversity increases with depth. The distribution of biocenoses and population structure depend on the hydrological regime and substrate type in the surveyed area.

Habitat related growth and reproductive investment in estuarine waters, illustrated for the tellinid bivalve Macoma balthica (L.) in the western Dutch Wadden Sea

In estuarine areas, bivalve species can be found in a variety of environments, where they experience large differences in environmental conditions. In the present paper, the importance of different habitats (intertidal, subtidal and adjacent coastal waters) for the persistence of the population was evaluated for the bivalve Macoma balthica (L.) in the western Dutch Wadden Sea estuary. Intra-specific variation in growth and reproductive output were followed during the year and related to local abiotic conditions. Significant differences in growth and reproductive investment were found between locations. Young individuals were mostly found in the intertidal area, where growth in terms of somatic mass was good. These areas were not favourable for adult individuals, since growth in shell length was low and many individuals did not reproduce. In the subtidal, where the highest densities were found, somatic and gonadal mass indices were low. Coastal areas had the lowest densities and showed high growth in terms of shell length and body mass. The habitat with the highest reproductive effort per individual was not the most important habitat in terms of reproductive output due to differences in density and in size of the habitat type. For M. balthica, the subtidal habitat contributed the most to the reproductive output of the western Dutch Wadden Sea population although the highest reproductive output per individual was in the coastal area.

Long-term and large-scale variability in productivity of the tellinid bivalve Macoma balthica on Wadden Sea tidal flats

A series of 27 successive annual estimates of production of soft parts of the bivalve Macoma balthica (L.) populating the tidal flats of Balgzand (western Wadden Sea, The Netherlands) was obtained using multi-station (15 sites) and long-term (27 yr: 1978 to 2004, inclusive) monitoring data on numerical densities and ash-free dry mass (AFDM). Annual production values were esti- mated from seasonal growth increments, which were positive in spring and early summer and (in adults) negative in the remainder of the year, when up to 70% of individual weight gain of the preceding growing season was lost. Annual estimates based on twice-annual sampling were checked and refined on the basis of a detailed 2 yr study. Weight loss proportions increased with age of ani- mals and with water temperature in winter. Net annual somatic production of the entire population varied from 1 to 9 (27 yr mean: 2.9 ± 0.3) g AFDM m -2 yr -1 . Lifetime production per individual averaged 0.007 g AFDM for recruits of ~3 mo old and 0.032 g AFDM for ~10 mo olds. Numerical den- sities of ~10 mo old recruits predicted, within narrow limits, subsequent cohort production by way of a positive linear relationship. However, production of 3 mo old recruits was on average higher at low than at high densities, owing to better juvenile survival at low densities. Local (between-station) variation in mean annual production was primarily related to local variation in recruitment. However, production per recruit was significantly higher at low than at high intertidal levels, in accordance with higher growth and lower mortality rates at lower levels. Strong temporal and local variation was also found in annual production/biomass (P/ - B) ratio values, which reflected the variable age compo- sition of the population: P/ - B ratio values were augmented whenever mean age was low. Consistently higher mean P/ - B ratio values (~0.6 vs. ~0.4 yr -1 ) were found at high near-coast compared with low off-shore tidal flats, in accordance with higher mortality rates on high flats. Environmental factors affecting variation in annual production included between-year variability in recruitment (which was related to winter temperature and shrimp predation pressure) and growth (which was related to diatom abundance).

Macoma balthica in Spain, a few decades back in climate history

The marine bivalve Macoma balthica has disappeared from the Spanish part of the Bay of Biscay during the past four decades. Investigating the possible role of climate change in forcing this species up north, we translocated it down south, back into two Spanish estuaries, and followed the thermal acclimatization of the metabolic rate during spring, summer and autumn. Our results reveal that in natural populations the respiratory response to temperature becomes down-regulated during summer and autumn. The respiration rate in the southward translocated populations became down-regulated to a safe level at very high temperatures (31 °C), but remained high at average temperatures instead. These translocated populations showed a gradual reduction of the condition-index, down to the level of starvation at the end of summer. Combined with an increased metabolic rate this indicates that the translocated specimens suffered from elevated maintenance rates. We conclude that short-term but frequent exposure to > 30 °C in the Spanish estuaries, induces elevated maintenance rates in M. balthica, and ultimately starvation. M. balthica indeed disappeared from the northern Spanish coast due to increasing summer maxima during the last decades. We prospect that the southern distribution limit of M. balthica will shift further north, and that the clam will eventually disappear from the entire Bay of Biscay with future increases in summer temperatures.

Effects of temporary food limitation on development and mortality of Macoma balthica larvae

Long-term observations (1973 to 2001) of populations of the intertidal bivalve Macoma balthica, in the western Wadden Sea, have suggested that larval mortality is strongly density- dependent. In addition, density-independent factors may affect larval mortality. One possible source for such an effect is food limitation. In laboratory experiments, M. balthica larvae were reared under food limiting conditions, both quantitatively (high or low food level) and temporally (starvation during the first, second or third week). The results indicated that larvae offered high food levels grew signifi- cantly faster (6.9 µm d -1 ), and metamorphosed earlier (16.5 d) and at greater length (264 µm) than larvae subjected to low food level (4.4 µm d -1 , 19.3 d and 244 µm, respectively). For both food levels, starvation in the first week resulted in a late metamorphosis at a large size, starvation in the second week resulted in an early metamorphosis at a small size, and starvation in the third week yielded intermediate results. Larval mortality was always lower in the low food condition, but the timing of starvation had no impact on mortality. This suggests that larvae, in vivo, do not die directly from food limitation alone. The results are discussed with reference to models of metamorphosis.