Observations on the distribution of meroplankton during an upwelling event

The distribution of the larvae of benthic invertebrates wasinvestigated relative to hydrographic structures as a test ofthe hypothesis that larvae behave as if they are passive particles.Observations of larval and oceanographic distributions weremade off Duck, North Carolina, USA in August 1994. Conditionswere characterized by wind-driven coastal upwelling; flow wasgenerally offshore near the surface and onshore below the pycnocline.Within 5 km of the shore the pycnocline was bent upward by theupwelling and it intersected the surface along most of the transects.In zooplankton samples, 20 taxa of larvae were counted (10 bivalveveligers, nine gastropod veligers and one polychaete larvae).Using cluster analysis, larvae were separated into groups withsimilar patterns of distribution and similar affinities to waterproperties. The larvae in Cluster 3 did not display a consistentdistribution pattern beyond that they tended to be found inwarmer surface waters. An earlier paper described the distributionof larvae in the same location during a downwelling event [A.Shanks et al. (2002) J. Plankton Res., 24, 391–416]. Twoof the clusters identified during this previous study were quitesimilar in composition to Clusters 1 and 2 in this study. Inboth studies, Cluster 1 larvae were found below the pycnocline,but during the upwelling event they were transported shorewardwith the advection of the subpycnocline waters by the upwellingcirculation. Within 5 km of the shore, Cluster 1 larvae werefound at depths shallower than the base of the pycnocline andwere often found in patches of high larval concentration. Thepatches were located where the waters were upwelling. Cluster2 larvae were found within 5 km of the shore in both studiesand tended to be highly concentrated in convergences or divergences.Larvae in Cluster 1 generally appeared to be dispersing as passiveparticles, except within the zone of upwelling where they mayhave been swimming against the upwelling flow leading to higherlarval concentrations. Cluster 2 larvae appeared to be consistentlyconcentrated in areas of vertical currents, suggesting thatthey may be attempting to maintain a preferred depth in theface of the vertical flow which would lead to high larval concentrationand nearshore larval distributions despite extensive cross-shelfmovement of water. Despite their slow swimming speeds, the larvaein Clusters 1 and 2 were not swept offshore by the upwellingevent.     

Diet composition of larval Leiostomus xanthurus in and about the Mississippi River plume

The diet of the larvae of the sciaenid fish Leiostomus xanthuruscollected within the Mississippi River plume differs from thatof larvae collected in adjacent Gulf of Mexico shelf waters.Larvae collected in plume water had eaten twice as many foodorganisms as had larvae collected in shelf waters. Larvae collectedwithin the plume had eaten mostly small food organisms (tintinnids,copepod nauplii, pelecypod veligers and invertebrate eggs),whereas larvae collected in shelf waters had eaten mostly largerfood items (copepodite and adult copepods). Differences in theabundance of food organisms in and differences in the photicenvironments of these water masses may account for the differencesin diet.     

Veligers of an introduced bivalve,Limnoperna fortunei,are a new food resource that enhances growth of larval fish in the Paraná River (South America)

1. Larvae of ‘sábalo’, Prochilodus lineatus, whose adults represent over 60% of overall fish biomass in the Río de la Plata Catchment, have been observed to feed intensively on veligers of the exotic bivalve Limnoperna fortunei. 2. To assess the effects of this dietary shift on the growth of P. lineatus, 28‐day laboratory experiments were carried out feeding newly hatched P. lineatus larvae with three diets: zooplankton artificially enriched with L. fortunei veligers; natural zooplankton; and zooplankton artificially enriched with cladocerans and copepods. The average length, weight and gut contents of the fish larvae were assessed weekly and metabolic rates of fish larvae were measured. 3. Proportions of veligers in gut contents were always higher than those in the experimental diet: 100, 76 and 21% for veliger‐enriched, natural and low‐veliger diets, respectively. Larvae fed a veliger‐enriched diet grew to a significantly larger size than larvae fed the other two diets. In energetic balance comparisons using metabolic rates and prey energy content, all three diets were sufficient to support metabolism and growth. The greatest values of excess energy at the end of each week were in the veliger‐enriched experiments. 4. Feeding on veligers of L. fortunei significantly enhances the growth of P. lineatus larvae and supports the idea that this new and abundant resource is selectively preyed upon by P. lineatus during its larval stage. Higher growth rates may stem from the higher energy contents of veligers compared to crustaceans and/or from the lower energy costs of capturing slower prey.     

A preliminary distributional study of fish larvae near a ribbon coral reef in the Great Barrier Reef

Fish larvae from horizontal plankton tows along a single transect near outer ribbon reefs of the Great Barrier Reef in spring 1979 and summer 1980 had persistent distributional patterns. Larvae were identified to family and divided into young (preflexion) and old (postflexion) larvae, thus giving 28 taxa abundant enough for analysis. Non-uniform larval distributions were found for 81% of the 16 reef fish taxa with non-pelagic eggs, but for only 17% of the six reef fish taxa with pelagic eggs. Most differences in larval concentration were between the lagoonal and seaward sides of the reef. Only tripterygiid larvae had highest concentration just seaward of the reef, while larvae of 12 reef and three oceanic fish taxa occurred in highest concentrations on the lagoonal side of the reef. In five taxa of reef fishes, higher larval concentrations were found in the lagoonal backreef compared with the mid-lagoon habitat; but the reverse was not found in any taxon. Eleven taxa had indeterminate distributions, (i.e. no difference in concentration between stations). Mechanisms responsible for the distribution remain unknown, but we suggest that the view which considers fish larvae to be passively-drifting particles is unjustified without more information on larval behaviour.     

Distributional ecology of milkfish,Chanos chanos,larvae in the Great Barrier Reef and Coral Sea near Lizard Island,Australia

Synopsis Plankton hauls captured 682 milkfish larvae (2.1–12.3 mm) in the Great Barrier Reef Lagoon and Coral Sea during 1979–1986. Larvae were present from November to March, and absent in April, July and October. We analyzed concentration, abundance and size-frequency data and concluded that spawning took place in the Coral Sea or at the outer edge of the continental shelf, apparently following an adult spawning migration of at least 50 km. Larvae then moved inshore to at least our most inshore routine sampling site near midshelf. Some larvae may have remained for a period in the lee of reefs along the shelf edge. Larvae apparently occupied the upper few metres of the water column, thereby becoming susceptible to shoreward movement in the wind-driven surface layer. Movement from spawning sites to midshelf probably required active maintenance of vertical position by larvae which enabled passive movement with favourable currents, and perhaps horizontal swimming. By the time larvae reach midshelf, continued inshore movement by horizontal swimming alone is possible.     

Transport of molluscan larvae through a shallow estuary

Dispersal of invertebrate larvae is determined by larval swimmingbehavior, the length of planktonic development and the hydrodynamicregime. Larvae of estuarine invertebrates must refrain fromexport or invade an estuary after development in the ocean.This study investigates retention patterns of estuarine molluscsby measuring time series of larval abundance in relation tohydrodynamic processes. Previous investigations of larval dynamicshave generally focused on larger estuarine systems that areoften stratified and have relatively long hydraulic residencetimes. The estuary studied in this investigation supports densepopulations of infaunal clams yet has a water depth to tidalamplitude ratio near unity. To access processes affecting larvalretention, the circulation patterns of the estuary were measuredwith time series of salinity, temperature, pressure and horizontalvelocity. Transport rates of larvae between ocean and estuary,and within the estuary proper, were calculated from velocityand larval concentration time series. The daily residence timeof the estuary was determined for the summer spawning period.The results demonstrate that molluscan larvae were routinelytransported between the estuary and nearshore zone in tidalflows. Based on the magnitude of the horizontal current velocities,passive transport of larvae predominates during most of thetidal cycle in the estuary. Residence time calculations suggestthat the ability of larvae to remain in the estuary throughlarval development is unlikely, and there was no evidence ofselective retention of mature bivalve larvae in the estuary.Rather, larvae are exported rapidly from the estuary and undergodevelopment in the coastal ocean. Mesoscale physical processesin the coastal ocean probably control variation in the deliveryof larvae back to estuarine systems. Recruitment to this andsimilar estuaries must therefore be dependent on invasion.     

The fishes of northern and central Veracruz,Mexico

The hydrographic structure of Abra Harbour, a tidal embayment located at the seaward end of the highly polluted estuary of Bilbao, is influenced by the inflow of a polluted estuarine plume, the asymmetry of the harbour basin, and the tidal circulation pattern. Multivariate analysis of the spatial variability of the zooplankton between May 1981 and May 1982 showed that significant differences in zooplankton abundance and species composition occurred from the inner-eastern to the outer-western side, related to the horizontal structure of the system. The annual temperature cycle, however, was the major source of temporal variability, and the seasonal stratification in the water column was responsible for the predominance of vertical differences in zooplankton composition during the spring-summer period. Wind-induced turbulence and tides were other sources of variation. With increasing turbulence (rough sea), the spatial gradients in zooplankton composition were not as clear, and with decreasing tidal height the compositional differences in the horizontal dimension tended to be more evident. Spionid larvae accounted for strong local differences within the Harbour. They were usually segregated from other taxa, and mainly associated to the deeper waters characterized by a greater hydrological instability.     

Life cycle of Hydrobaenus kondoi Sæther (Chironomidae) at the middle reaches of the Kiso River, Japan

The life cycle and biological characteristics of the nuisance midge, Hydrobaenus kondoi Sæther were studied in 1989–90 at the middle reaches of the Kiso River, Japan. A study of seasonal abundance of both adults and larvae revealed that the Kiso River population was bivoltine. In laboratory cultures, the species required 840.7 degree-days to complete the life cycle. Larvae built a cocoon in the second instar for aestivation during the high temperatures of summer. The cocoons were found attached to sand grains in shallow water near shore.     

Temporal and spatial patterns in ichthyoplankton assemblages in bay and open coastal environments

The larval fish fauna occurring in temperate bay and shelf waters off Victoria, southern Australia, was found to be diverse, comprising taxa from 52 families. The most abundant groups collected were gobiids, tripterygiids, gobiesocids and clupeids. Fish egg concentrations were highest during spring and summer (September to February). Eggs of the Australian anchovy Engraulis australis occurred mainly during spring (September to November). Total larval fish concentrations were highest during summer (December to February), and were significantly higher at 1 km than 2 and 5 km from shore in offshore samples. Larval concentrations of a number of families, mainly reef‐associated taxa that attach their eggs to hard substrata, were also higher nearer to shore. These larvae are more developed upon hatching than those of pelagic spawners and more capable of avoiding passive drift. Multivariate analyses found that larval taxonomic composition did not vary significantly with distance from shore, but that seasonal and monthly groupings were evident, with different taxa dominating at different times of the year. Larvae of the families Gobiidae and Tripterygiidae occurred in all months, but were less abundant during winter. Spatial differences in the larval fish assemblage between offshore samples and samples taken in the bay were only apparent during summer. This was primarily due to a higher abundance of seagrass‐associated species, such as syngnathids and hemiramphids, utilizing specific habitats in the bay.     

Larvae of the invasive species <Emphasis Type="Italic">Limnoperna fortunei</Emphasis> (Bivalvia) in the diet of fish larvae in the Paraná River,Argentina

To assess the trophic impact of the planktonic larvae of the invasive Asian bivalve Limnoperna fortunei, introduced in South America around 1990, we investigated the gut contents of fish larvae collected at monthly intervals between October 2000 and March 2001 at three locations along a 600 km stretch of the Paraná River, and during November 2004 in two areas of marginal lagoons connected to the river near the city of San Nicolás, Argentina. Zooplankton was also collected in the lagoons in 2004. In total, 11,956 fish larvae were retrieved, of which 1,511 were used for detailed analyses. Of the 15 fish taxa collected, 11 had veligers of L. fortunei in their gut. Fish larvae with empty guts represented 60% (San Nicolás) to 72% (Paraná River) of the total number of fish. Proportions of feeding fish larvae with L. fortunei veligers in their guts varied between 20% (San Nicolás) and 56% (Paraná River); in 15% of the guts analyzed, L. fortunei was the only food item recorded. For those specimens that had consumed L. fortunei larvae and any other food item, L. fortunei was the most important item in 55% (Paraná River) to 71% (San Nicolás) of the animals in terms of biomass. No major temporal or spatial changes in the diet were observed along the Paraná River, but the relative biomass contribution of L. fortunei larvae differed strongly in fishes of different developmental stage. In protolarvae and mesolarvae, veligers accounted for 30–35% of the gut contents. In metalarvae, veligers accounted for only 3%, indicating enhanced food supply for the earliest fish life stages. Comparison of the relative proportions of the three main zooplankton types (L. fortunei veligers, cladocerans, and copepods) in the water and in larval fish guts indicates that L. fortunei is always selected positively over the other two prey types. While our results strongly suggest that the expansion of L. fortunei results in an enhanced food supply for local fish populations, they do not necessarily imply that the overall effect on the ecosystem in general, and on the fish fauna in particular is beneficial.     

Orientation and navigation in coastal and estuarine zooplankton

The axiom that zooplankton species exhibit upwards migration behaviour at night is challenged by recent findings. Such behaviour is not universal, may vary during ontogeny, and is occasionally reversed. Moreover, in some estuarine and coastal zooplankton species vertical migration rhythms are of tidal, not diel, periodicity. There is evidence for several species that vertical migrations are endogenously controlled, occurring in constant conditions in the laboratory, suggesting that they have arisen under considerable selection pressure. They appear to play a significant role in orientation and navigation of coastal and estuarine zooplankton, predicated on the selective advantage of closure of life cycles. Vertical migrations between water masses moving in different directions at tidal, diel or longer timescales permit dispersal and recruitment, or retention, of planktonic larvae and adults in favourable ecological locations. Exogenous factors serving as cues for, or directly controlling, vertical migration rhythms include light, hydrostatic pressure, salinity, temperature, water movements and gravity responses, besides biological factors such as the timing of larval release, duration of larval life and predator/prey interactions. Behavioural characteristics should be quantified and factored into dispersal models which assume that zooplankton adults and larvae are advected as passive particles.     

Pelagic larvae of benthic invertebrates of the Vostok Bay,peter the great bay,Sea of Japan: Composition,phenology, and population dynamics

For the first time, year round and long-term seasonal research on the composition, phenology, and population dynamics of pelagic larvae of bottom invertebrates were conducted in the Vostok Bay of the Sea of Japan. Larvae of 98 taxa of various ranks were found. The larvae of bivalves and polychaetes were the most diverse. The larvae of bottom invertebrates occurred in plankton year-round. Their density was at its minimum in January (less than 100 ind./m³), and its maximum from June to November (more than 1000 ind./m³). The larvae of polychaetes dominated in the autumn and winter seasons, whereas bivalve and gastropod larvae dominated in the spring and summer seasons. It was shown that the proportion of pollution tolerant meroplankton larvae of the Polydora and Pseudopolydora genera was high in the Vostok Bay, and could indirectly indicate increasing anthropogenic pressure in the area.     

Larval biology of butterflyfishes (Pisces,Chaetodontidae): what do we really know?

Synopsis Relatively little is known of the pelagic portion of the life history of butterflyfishes. Eggs are small (<1 mm), pelagic and hatch in less than 30 hours. Most species pass through a so-called tholichthys larval interval characterized by elaborate, distinctive head spination:Coradion larvae have different head spination. While older chaetodontid larvae can be identified by adult characters, young (preflexion) larvae generally cannot now be identified below family. In tropical plankton studies chaetodontid larvae averaged <0.1% of larvae captured, and occurred in 13% of samples. This rarity is a major hindrance to further work, but is not unexpected in view of adult abundance. Larvae of a few taxa are most abundant in shelf waters, but larvae of many chaetodontid taxa seem to be most abundant in oceanic waters. In either case, waters near reefs have the fewest chaetodontid larvae. Offshore maxima of larvae appear to exist a few kilometers seaward of Great Barrier Reef ribbon reefs. Chaetodontid larvae may prefer the upper portion of the water column. Both size and age at settlement vary widely within the family and the large genusChaetodon, and the latter varies widely within species. Average size at settlement is less than 20 mm and age is less than 40 days. No correlation was found between size and age at settlement. Behaviour and feeding of chaetodontid larvae are essentially unstudied. Chaetodontid larvae seem to be least abundant in winter. The implications of these conclusions are discussed and some suggestions for further research are made. In all areas more work is needed.     

COMPOSITION,ABUNDANCE AND SEASONALITY OF LARVAL FISH IN RICHARDS BAY HARBOUR,KWAZULU-NATAL,SOUTH AFRICA

Summary

The present study examined the composition, abundance and seasonality of fish larvae occurring in Richards Bay Harbour, over an 18-month period from June 1991 to November 1992. Ichthyoplankton samples were collected every six weeks at top, mid and bottom depths in the dredged channel 100 to 200 m from the harbour entrance. A total of 7 163 fish larvae, representing 105 taxa and 53 families, was collected throughout the study period. The dominant families were the Engraulidae and Gobiidae representing 50% and 37% of the total catch, respectively. The most abundant larvae were those of the thorny anchovy, Stolephorus holodon, and an unidentified goby, Gobiid 12 (32% and 30% of the total catch, respectively). Estuarine-associated taxa dominated the total catch in terms of density (61%), with 11 out of the 28 taxa being totally dependent on estuaries at some stage in their life cycle. Turbidity and temperature were significant variables accounting for 27% of the variation in larval densities of estuarine-dependent species. The main peak in larval density occurred in January 1992 (summer) with a mean larval density of 160 larvae per 100 m³. Larvae of estuarine-associated taxa were predominantly at the flexion and postflexion developmental stages. Since larval densities of certain estuarine-associated species were significantly greater in bottom samples and on flood tides, it is suggested that selective tidal stream transport is a recruitment mechanism used by these species. The present study concluded that it is important to maintain and preserve the estuarine component of Richards Bay Harbour so that the system can function as an estuarine nursery for many of the fish species depending on it at certain stages in their life cycle.     

Roles of larval supply and behavior in determining settlement of barnacles in a temperate mangrove forest

Recruitment is often a major influence on the spatial distribution of populations of benthic marine invertebrates, but the contributions of different components of recruitment are not well known, with the added complication that the relative importance of various life-history processes may be scale-dependent. Previously, we have shown that over a large scale across a mangrove (Avicennia marina) forest in southeastern Australia, settlement of the barnacle Elminius covertus explained its patterns of recruitment, which in turn explained the distribution of adults on mangrove pneumatophores. Post-settlement mortality had little influence on this pattern. In contrast, small-scale vertical distributions of adult barnacles along individual pneumatophores were determined by the pattern of recruitment, which differed from the pattern of settlement, so post-settlement mortality determined the vertical patterns of adults.

In this study, we tested whether larval supply and/or settlement behavior influence the observed settlement patterns of E. covertus across a forest (from seaward to landward zones). We also tested whether larval supply could explain the vertical settlement patterns along the pneumatophores. A pumping system was used to collect cypris larvae from seaward, mid and landward zones of a mangrove forest and an adjacent, unvegetated shore and from three heights above the sediment surface. We also used transplantation of wooden stakes bearing microbial films and barnacle recruits between horizontal zones of the forest to determine whether settlement was influenced by these films or recruits.

Both cyprid supply and cyprid behavior were important factors in determining the patterns of settlement of E. covertus across the forest. Cyprid supply was a result of three-fold differences in immersion times of different (landward, mid and seaward) zones across the forest and a decrease in density of cyprids in the water column from the seaward zone of the forest to the landward sections. In the absence of mangroves immediately adjacent to the forest, there was no temporally consistent difference in cyprid density across the shore and even the differences in immersion time did not produce consistent differences in cyprid supply across the shore. Wooden substrata that had been immersed at seaward sections of the forest attracted consistently more settlers than substrata immersed initially at other sections of the forest and settlement could be induced beyond the normal distribution of adults of E. covertus by stakes transplanted from the seaward zone.

The vertical settlement pattern could not be explained by the supply of cyprids, suggesting that larval behavior must determine the vertical settlement pattern.     

Rhythmic aspects of estuarine migration of hatchery-reared Atlantic salmon (Salmo salar) smolts

Seasonal, diel and tidal rhythmic activity of hatchery-reared Atlantic salmon ( Salmo salar ) smolts migrating through a large estuary was studied by ultrasonic tracking of 46 individuals during two seasons. Prior to 10 May each year most smolts were inactive and remained near shore in shallow water. After 10 May nearly all smolts moved away from the release point into swift water and made rapid seaward progress. Initiation of migration each year occurred when river and hatchery water temperatures rose above 9°C. Migration in the estuary was largely passive drift, and as a result there were distinct tidal rhythms of ground ('swimming') speed and net seaward travel. There were no diel rhythms in ground speed or net seaward travel; smolts drifted seaward on the tides during both day and night. Smolts may be slightly deeper during day than night.     

Topographically generated fronts, very nearshore oceanography and the distribution and settlement of mussel larvae and barnacle cyprids

Flow patterns adjacent to shore may prevent or aid shorewardmigration of benthic invertebrate larvae. We hypothesized thata front at the mouth of Sunset Bay, Oregon, prevents shorewarddispersal of larvae, significantly altering settlement of mussellarvae and barnacle cyprids. Settlement was measured at threesets of moorings (three moorings per site) distributed acrossthe front at Sunset Bay. From 6 July to 4 September 2000, sampleswere collected roughly every other day. Concurrently, we madevertical zooplankton tows adjacent to each mooring site andcollected physical oceanographic data. During upwelling-favorablewinds, the front was always present at the bay mouth, separatingsignificantly cooler, saltier and denser offshore water fromthat within the bay. During downwelling winds, the front brokedown and we found no significant difference in the surface physicaloceanographic parameters across the bay mouth. During upwelling,the concentration of mussel larvae was higher seaward of thefront than landward, but there was no significant differencein concentration during downwelling, suggesting that the frontmay act as a barrier to the shoreward dispersal of mussels.Mussel settlement was too low and sporadic to allow statisticalanalysis. There was no difference in cyprid concentrations acrossthe bay mouth whether the front was present or not. Cyprid settlementwas, however, nearly an order of magnitude lower at mooringsseaward of the front than at those landward. A significant cross-correlationwas found between settlement at the offshore mooring and tidalrange (r = 0.464, lag = 0 days) and between settlement at themid and inner moorings and downwelling winds (r = 0.532 midbay, r = 0.532 inner bay, lag = 0 days). Seaward of the front,settlement varied with tidal range, while landward of the front,most settlement occurred as brief pulses during downwellingwinds, periods when the front was not present. We found largedifferences in the distribution of cyprids, and mussel larvaeand cyprid settlement relative to the front; larval distributionsand settlement varied with upwelling versus downwelling windsand was due to differences in the very nearshore (i.e. within100–1000 m of shore) coastal oceanography.     

The location of parasites within their hosts: the behavioural component in the larval migration of Nippostrongylus bras1l1ensis in the tissues of the rat

The role of larval behaviour in successful completion of tissue migration is briefly discussed and it is related to the passive carriage of larvae along the ‘pipes and tubes’ of the host. Larvae of N. brasiliensis were injected into selected portions of the circulatory system and following periods of 5–60 min they were recovered from the blood, liver and lungs. Larvae were also immobilised in 0·4% piperazine, a dosage which permitted recovery in about 60 min. The dispersion of treated larvae was compared with that untreated controls. It was found that larvae were carried very rapidly in the blood stream and that they became lodged in the first capillary bed that they entered. They could not pass through capillary beds without movements (and/or secretions). A decreased number of adults developed after larvae were introduced via a series of routes which required the larvae to pass through an increasing number of ‘hurdles’ to migration.     

Dispersal of Deep-Sea Larvae from the Intra-American Seas: Simulations of Trajectories using Ocean Models

Using data on ocean circulation with a Lagrangian larval transport model, we modeled the potential dispersal distances for seven species of bathyal invertebrates whose durations of larval life have been estimated from laboratory rearing, MOCNESS plankton sampling, spawning times, and recruitment. Species associated with methane seeps in the Gulf of Mexico and/or Barbados included the bivalve "Bathymodiolus" childressi, the gastropod Bathynerita naticoidea, the siboglinid polychaete tube worm Lamellibrachia luymesi, and the asteroid Sclerasterias tanneri. Non-seep species included the echinoids Cidaris blakei and Stylocidaris lineata from sedimented slopes in the Bahamas and the wood-dwelling sipunculan Phascolosoma turnerae, found in Barbados, the Bahamas, and the Gulf of Mexico. Durations of the planktonic larval stages ranged from 3 weeks in lecithotrophic tubeworms to more than 2 years in planktotrophic starfish. Planktotrophic sipunculan larvae from the northern Gulf of Mexico were capable of reaching the mid-Atlantic off Newfoundland, a distance of more than 3000?km, during a 7- to 14-month drifting period, but the proportion retained in the Gulf of Mexico varied significantly among years. Larvae drifting in the upper water column often had longer median dispersal distances than larvae drifting for the same amount of time below the permanent thermocline, although the shapes of the distance-frequency curves varied with depth only in the species with the longest larval trajectories. Even species drifting for >2 years did not cross the ocean in the North Atlantic Drift.