Vertical distribution of benthic invertebrate larvae during an upwelling event along a transect off the tropical Brazilian continental margin

Journal of Marine Systems 79 (2010) 124–133

Contents lists available at ScienceDirect

Journal of Marine Systems

j ourna l homepage: www.e lsev ie r.com/ locate / jmarsys

Vertical distribution of benthic invertebrate larvae during an upwelling event along atransect off the tropical Brazilian continental margin

Marcos Y. Yoshinaga a,⁎,1, Paulo Y.G. Sumida a, Ilson C.A. Silveira a, Áurea M. Ciotti b, Salvador A. Gaeta a,Luiz F.C.M. Pacheco a, Andréa G. Koettker a

a Instituto Oceanográfico da Universidade de São Paulo, São Paulo, CEP: 05508-120, Brazilb UNESP Campus do Litoral Paulista. Praça Infante Dom Henrique s/n° São Vicente, SP, CEP: 11330-900, Brazil

⁎ Corresponding author. Tel.: +55 11 30916543.E-mail address: [email protected] (M.Y. Yosh

1 Present Address: MARUM, University of Bremen, LeGermany.

0924-7963/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.jmarsys.2009.07.007

a b s t r a c t

Article history:Received 18 September 2008Received in revised form 21 July 2009Accepted 22 July 2009Available online 6 August 2009

Keywords:SE Brazilian coast (23°S and 42°W)Shelf dynamicsCoastal upwellingZooplanktonInvertebrate larvae

Abundance and composition of marine benthic communities have been relatively well studied in the SEBrazilian coast, but little is known on patterns controlling the distribution of their planktonic larval stages. Asurvey of larval abundance in the continental margin, using a Multi-Plankton Sampler, was conducted in across-shelf transect off Cabo Frio (23°S and 42°W) during a costal upwelling event. Hydrographic conditionswere monitored through discrete CDT casts. Chlorophyll-a in the top 100 m of the water column wasdetermined and changes in surface chlorophyll-a was estimated using SeaWiFS images. Based on the larvalabundances and the meso-scale hydrodynamics scenario, our results suggest two different processesaffecting larval distributions. High larval densities were found nearshore due to the upwelling eventassociated with high chlorophyll a and strong along shore current. On the continental slope, high larvalabundance was associated with a clockwise rotating meander, which may have entrapped larvae from aregion located further north (Cabo de São Tomé, 22°S and 41°W). In mid-shelf areas, our data suggests thatvertical migration may likely occur as a response to avoid offshore transport by upwelling plumes and/orcyclonic meanders. The hydrodynamic scenario observed in the study area has two distinct yet extremelyimportant consequences: larval retention on food-rich upwelling areas and the broadening of the tropicaldomain to southernmost subtropical areas.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

The vast majority of marine benthic invertebrate groups possess aplanktonic larval stage, which can spend from hours to months in thewater column before settling on a suitable habitat (Thorson, 1950).The distribution of adult populations is a direct consequence of larvaldispersal and survival (Shanks, 1995), and over the past 20 years,efforts have been made to investigate the interactions among physicaloceanography, larval distribution, dispersal and recruitment of larvae(Roughgarden et al., 1991; Poulin et al., 2002; Shanks and Eckert,2005; Shanks and Brink, 2005). Spatial and temporal patterns in larvaldistribution can be explained through a number of different processes,such as physical advection of larvae and the local production of larvalfood (Nakata et al., 2000; Botsford, 2001; Garland et al., 2002; Poulinet al., 2002).

The ability to control larval vertical position will affect the nettransport for a particular species (Shanks, 1995), as behavioralpatterns in conjunction to physical processes will determine whether

inaga).obener Straße, 28359 Bremen,

ll rights reserved.

larvae are exported, retained, or concentrated in specific locations(Clough et al., 1997; Cowen et al., 2000; Paris et al., 2002). Cross-shelftransport of larvae has been attributed to wind-driven drifting,internal tidal waves, meso- and large-scale circulation features, andupwelling and downwelling (e.g., Roughgarden et al., 1991; Shanks,1995; Nakata et al., 2000; Botsford, 2001; Poulin et al., 2002; Shanksand Eckert, 2005; Ma et al., 2006; dos Santos et al., 2008). The effectsof upwelling or downwelling on larval distribution can be accessedfrom the knowledge of vertical distribution of larvae (Shanks andBrink, 2005).

The upwelling system of Cabo Frio is an important site of primaryproductivity in the Brazilian coast (Gonzalez-Rodriguez et al., 1992;Gonzalez-Rodriguez, 1994), although it is generally considered weakcompared to eastern boundary coastal upwelling systems. Here, wepresent data on vertical distribution of benthic invertebrate larvaeduring a coastal upwelling event off the tropical Brazilian coast (CaboFrio, between 22°58'S, 42°03'W and 24°33'S, 41°23'W) and itsrelationship to local oceanographic conditions during repeatedcross-shelf transects. Our goal is to provide some insights on probablemechanisms of larval dispersion in the area. Similar research wasconducted in different oceanographic settings including the persistentupwelling off Chile (Poulin et al., 2002), the seasonal upwelling off thecoast of California (e.g., Roughgarden et al., 1991; Wing et al., 1995),

125M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

and the weak upwelling off the central-east coast of the US (e.g.,Shanks et al., 2000; Garland et al., 2002; Shanks et al., 2003; Shanksand Brink, 2005; Ma et al., 2006). Is worth to note that previousstudies (Shanks et al., 2000; Garland et al., 2002; Poulin et al., 2002;Shanks and Brink, 2005; Ma et al., 2006) concentrated efforts withinthe first 30 km of the coast, directly influenced by the upwelling front.Our study was expanded to a region out to ~100 km from the coast,and consequently, deeper zones in the water column were sampled.We observed simultaneous occurrence of coastal upwelling andupwelling filaments flowing southwards from the coast at Cabo Frio,which may represent additional constrain to the distribution ofbenthic population along the boundary between tropical andsubtropical domains in the SE Brazilian coast. Given the circulationscenario, we propose a physical/biological interaction that explainsthe association of benthic larvae in the water column and the specifichydrodynamic features of this region.

2. Study site

The upper 100 m of the water column of the Southeast BrazilianBight (SBB, 23°S to 28°S) is influenced by three water masses: theTropical Water (TW, TN20° and SN36.4) flowing as the surfaceexpression of the Brazil Current (BC); the cold and nutrient rich SouthAtlantic Central Water (SACW, Tb20 °C and Sb36.4) flowing belowTW and as the thermocline portion of the BC; and the Coastal Water(CW), a low salinity water resulting from fresh water input of small tomedium sized estuaries along the SBB (Campos et al., 1996, 2000;Silveira et al., 2000). At Cabo Frio, the SBB is characterized by arelatively narrow shelf (~50 kmwide) with an abrupt change (N–S toE–W) in coastline orientation (Campos et al., 2000; Rodrigues andLorenzzetti, 2001). Under N–NE winds, which are common duringsummer and blow parallel to the coast in Cabo Frio, surface watermoves offshore (via Ekman transport) resulting in the upwelling ofSACW (Castro and Miranda, 1998). Blooms of phytoplankton arecommonly observed as a consequence of SACW upwelling in coastalareas off Cabo Frio (Valentin, 1984; Gonzalez-Rodriguez et al., 1992).Matsuura (1996) studying the sardine spawning off the SE Braziliancoast identified the coastal upwelling of Cabo Frio as the key factorsupporting the regional fisheries productivity. Except for Cabo Frio,primary production in continental shelf and open waters off the SBBcan be considered under oligotrophic conditions, with a strongdepletion of nutrients in the euphotic zone associated with thewarm TW (Metzler et al., 1997). Alongshore variations are observed in

Table 1Information on sampling date, time of the MPS, local depth, integrated chl-a, total larvae andstudy.

Station # Distance offshore(km)

Date Local time Depth(m)

chl-a(mg m−2)

Larvae(ind. m−3)

Echi

1 191 2/7/01 7:55 PM 2474 13.5 22.1 4.02 153 2/8/01 10:18 AM 1960 14.0 6.2 0.43 116 2/8/01 2:47 PM 1191 12.0 2.4 0.44 77 2/8/01 8:48 PM 160 11.8 6.5 0.15 40 2/9/01 5:12 AM 120 22.2 18.4 0.06 3 2/9/01 10:19 AM 40 45.0 29.2 0.07 40 2/9/01 8:05 PM 123 19.4 20.8 0.18 58 2/9/01 11:28 PM 142 18.5 49.2 0.09 77 2/10/01 3:27 AM 160 12.9 4.8 0.010 107 2/10/01 7:06 AM 713 15.8 5.6 0.011 107 2/11/01 12:45 PM 708 19.2 7.2 0.112 78 2/11/01 4:48 PM 161 14.9 2.3 0.013 58 2/11/01 8:57 PM 145 18.4 9.8 0.114 40 2/11/01 11:42 PM 122 31.3 12.8 0.015 22 2/12/01 4:45 AM 116 19.6 66.9 0.316 5 2/12/01 8:57 AM 40 35.2 18.0 0.117 40 2/12/01 1:02 PM 123 15.6 24.4 0.018 22 2/13/01 3:38 PM 113 16.1 26.1 0.019 3 2/13/01 9:25 PM 36 37.2 5.4 0.0

the vicinity of Cabo Frio, with upwelling cells and plumes frequentlyfound southwards from Cabo Frio and Cabo de São Tomé, as well asnorthward (Lorenzzetti and Gaeta, 1996; Carbonel, 1998). Thismechanism was described by Calado et al. (2006) as coastal watersexcursions onto oceanic areas promoted by meanders of the BC.

3. Methods

Samples were collected during the multidisciplinary DEPROASExperiment (Ecosystem Dynamics of the Southwest Atlantic Con-tinental Shelf) in February 7th to 13th 2001 aboard the R/V Prof. W.Besnard (Oceanographic Institute, University of São Paulo). A total of19 stations were sampled along a single transect, oriented perpendic-ular to Cabo Frio, which extended up to 200 km off the coast. Thetransect was visited four times and samples were collected at 3, 5 or 6stations within a two-day average sampling per transect. Local depthsvaried from 40 to 2500 m, thus covering coastal, shelf and slopewaters (Table 1 and Fig. 1).

At each station, CTD casts and discrete water sampling werefollowed by deployments of a Multi-Plankton Sampler (MPS, 333 µmmesh size) set to sample five strata in the top 100 m of the watercolumn (at 20 m intervals). In nearshore areas (b100 m depth) wesampled only 2–3 strata using 10 m intervals. Flow meters were usedto compute the volume of water filtered in each net. Samples werepreserved in 4% buffered seawater formalin, sorted under a stereo-microscope and the larvae were enumerated to major taxa (echino-derms, gastropods, bivalves, cirripedians, brachyurans, stomatopodsand anomurans). Brachyuran larvae present in shelf and slope waterswere identified to the lowest possible taxonomic level.

Water samples for chlorophyll-a (chl-a) analysis were collectedwith Niskin bottles coupled to the CDT-rosette system. Water wasfiltered on board using GF/F filters and chl-a concentrations (mg m2)were measured fluorimetrically (Turner Designs AU-10) in thelaboratory (Holm-Hansen et al., 1965) using 90% acetone and 24 hextraction. Surface chl-a was also estimated through satellite data.SeaWiFs radiometric data (Fig. 1) at Level 1A and nadir resolution of1.1 km, as well as daily meteorological data, were obtained from theNASAGSFC's Distributed Active Archive Center (DAAC). Chl-a (mgm3)was estimated by the Oc2 version 4 algorithm using SEADAS 4.4standard atmospheric correction and masks. Images were mapped toa cylindrical projection and colored coded to illustrate surfacechlorophyll in mg m−3 (Fig. 1).

individual major taxa densities for the first 100 m depth in each station sampled in this

nodermata Polychaeta Gastropoda Bivalvia Cirripedia Stomatopoda Brachyura

1.9 14.4 1.3 0.2 0.0 0.31.5 3.6 0.3 0.0 0.1 0.10.6 1.1 0.0 0.0 0.0 0.10.5 5.7 0.0 0.0 0.0 0.20.7 14.9 0.1 0.1 0.7 1.80.1 0.0 0.0 0.1 19.6 8.20.7 8.6 0.3 4.3 0.9 5.90.3 8.1 0.0 39.5 0.1 1.20.0 3.2 0.1 0.4 0.1 0.80.5 4.6 0.0 0.1 0.1 0.30.5 4.7 0.1 0.0 0.5 1.40.3 1.4 0.1 0.1 0.1 0.30.5 7.5 0.1 0.0 0.5 1.10.2 6.5 0.0 0.2 0.3 4.82.8 47.6 0.4 0.7 0.5 14.60.0 0.2 0.0 0.0 4.9 12.50.2 19.0 0.1 2.1 0.9 1.80.1 5.6 0.4 0.3 0.9 17.70.2 0.3 0.1 0.2 1.1 3.5

Fig. 1. SeaWiFS image from 02/13/2001 showing the study area transect and the depths sampled (open squares). 1 represents a plume of upwelled and chlorophyll rich water, 2 thecoastal upwelling event under investigation, 3 and 4 the excursion of coastal waters to offshore areas.

126 M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

A north-south velocity field (Fig. 2) was calculated from theobserved temperature and salinity data in the DEPROAS Experiment,using the sectional version of the Princeton Ocean Model (POM). Thismethod is detailed in Silveira et al. (2004) and was applied to theregion off Cabo Frio. The dynamical interaction between coastalcirculation and the meandering pattern of the BC in the DEPROASExperiment of 2001 was also described by Calado et al. (2006).

Fig. 2.North-South current velocities during a coastal upwelling event off Cabo Frio (seeSilveira et al., 2004). Negative velocities are southwestwards and contours intervals are0.1 m s−1.

Spearman's Rank correlation test was performed to identifypossible relationships between distance offshore, percentage of thewater column influenced by the SACW (Tb20 °C and Sb36.4, hereafter defined according to Silveira et al. (2000)), larvae densities andchl-a concentrations integrated for the upper 100 m.

4. Results

4.1. Physical oceanographic conditions, chl-a concentrations and larvaevertical distribution

Sampling was initiated on February 7th of 2001 at the offshore St. 1(Table 1). Low bottom water temperatures and high chl-a concentra-tions in nearshore areas of the first transect (Fig. 3), as well as thesatellite image from February 8th (Fig. 1), evidenced the presence of acoastal upwelling associated with a consistent bloom of phytoplank-ton. The north-south circulation pattern was computed only for thefirst transect since the time scales at which the velocity field evolvedwas slow compared to the duration of the DEPROAS Experiment (seedetails in Silveira et al., 2004). During the sampling of transect 1, therewere two distinct flow patterns (Fig. 2). The first was the southwardcoastal current flow associated with the upwelling regime centered atapproximately 20 km from the coast and occupying the first 20 m ofthe water column. We could also identify the vertical shear, withweaker bottom currents flowing in the opposite direction (Fig. 2). Off

Fig. 3. Larval spatial distributions in transect #1. Upper two panels showing temperature (°C) and chlorophyll-a concentration (mg m−3) profiles with dots representing total larvaldensities (ind. m−3). Shaded areas represent stations sampled at night. Note that horizontal axes showing stations number (above) and the distance from the coast (below). Lowerpanels showing the vertical distribution of the four most abundant taxonomic groups in each station.

127M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

the shelf, a robust clockwise rotating BC meander was observed inareas 100 to 200 km offshore (Fig. 2). Corroborating this feature wereboth satellite sea surface temperature (Calado et al., 2006) and oceancolor images showing an excursion of coastal waters from furthernorth of Cabo Frio into the oceanic realm (Fig. 1). The abundances oflarvae at St. 1 and St. 5 were higher than stations located at the outer-shelf or shelf-break areas of transect 1 (Table 1), and high larvaeconcentrationswere observed in subsurface strata 20–40m (Fig. 3). AtSt. 1, gastropods, followed by echinoderms, polychaetes and bivalvesdominated larval abundance (Table 1). In fact, the three lattertaxonomic groups were associated with offshore stations. Highdensities of those organisms at St. 1 (Fig. 3) were coincidently relatedto the external edge of the rotating meander of the BC (Figs. 1 and 2).Gastropods density was considerably higher than the other abundanttaxa (brachyurans, polychaetes and stomatopods) in continental shelfareas (Table 1).

The second sampling of the transect was initiated on 9 February atSt. 6, located 3 km from the shore (Table 1),where SACWdominated thewater column (Fig. 4). As during in the first transect, section 2 showedhigh chl-a concentrations within 60 km offshore (Figs. 3 and 4), mostlypositioned close to the sea bottom (~70 m) where SACW was flowingtowards the coast with the persistence of the upwelling event. Thehighest larval concentration was observed at St. 6 (Table 1), withstomatopods and brachyurans accounting for 97% of the sample,coinciding with high concentrations of chl-a (Fig. 4). At St. 7, thedominant taxa were gastropods, cirripedians and brachyurans. Most of

the larvae at St. 7 were found in surface waters, and decreased inabundance with depth (Fig. 4). At St. 8, the most abundant larvae werecirripedians (80.3%), which were mostly found at 80–100 m (Fig. 4).Larval concentrations at St. 9 and St. 10 were lower than at the shelfstations (Table 1). At St. 9, gastropod larvae were most abundant,although cirripedians and brachyurans were numerically important aswell, with high numbers of individuals caught in the 80–100 m strata(Fig. 4). At St. 10, gastropods again were the most abundant (81.9%),however they were caught at greater depths than at stations St. 7 andSt. 9 (Fig. 4).

On 11 February, the section was visited a third time with samplingstarting on the slope and progressing towards the coast (Fig. 5). In orderto track the dynamics of the coastal upwelling event, the second andthird sections were limited to 150 km offshore, and consequentlystations located in areasN1000 m depth were not visited. Gastropodsfollowed by brachyurans were the most abundant larvae at stations St.11, 12 and 13 (stationsN60 km from shore). At St. 11 those organismswere positioned in subsurfacewaters (20–60mdepth), in deeper layersat St. 12 (mostly in the strata 60–80m, together with cirripedians), andin surface waters at St. 13 (Fig. 5). However, at St. 14, brachyuransoccupied thefirst strata,whereas gastropodsweremore abundant at the40–60 m layer (Fig. 5). While gastropods dominated the larvalabundance at St. 15 (71.2%), brachyurans and stomatopods representedalmost 99% of total larvae at St. 16 (Table 1 and Fig. 5).

On 12 February the fourth transect covered ~100 km and larvaewere only sampled at three stations within 40 km offshore (Fig. 6).

Fig. 4. Same as Fig. 3 for transect #2.

128 M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

The coastal upwelling event persisted throughout the DEPROASExperiment, and as observed in transect 3 (Figs. 5 and 6), highestchl-a concentrations were associated with the SACW, both near thecoast and in subsurface waters 40 km from the coast. At St. 17, therewas a dominance of gastropods followed by cirripedians and bothlarvae were most abundant at the deeper strata sampled (Fig. 6).Brachyurans and gastropods accounted for the highest densities at St.18 and the most abundant taxa were found in the first strata of thewater column (Fig. 6). At the last station sampled, brachyurans andstomatopods showed the highest densities and were associated withthe first 20 m of the water column, whereas gastropods andcirripedians were found down to 20–30 m strata (Fig. 6).

4.2. Larvae spatial distribution and correlation analysis

The spatial distribution of larvae show that the highest densitiesoccurred at St. 15 (66.9 ind. m−3), St. 8 (49.2 ind. m−3), St. 6 (29.2 ind.m−3), St. 18 (26.1 ind. m−3), St. 17 (24.4 ind. m−3) and St. 1 (22.1 ind.m−3) the most offshore station (Table 1). The dominant groups weregastropods (42%), followed by brachyurans (24%), cirripedians andstomatopods (13%) and polychaetes (3%). Gastropod larvae wereparticularly abundant at St. 15, where numbers reached 47.6 ind. m−3,representing 72% of the total. Larval abundance across shelf showed aU-shaped distribution with higher numbers at the coastal upwellingand in the gyre (Fig. 7).

Crustacean larvae were more abundant nearshore, and at St. 16they represented more than 90% of the larvae sampled, withbrachyurans density of 12.5 ind. m−3. This trend was observed for

all crustaceans including Brachyura, Stomatopoda, Cirripedia, Anom-ura and Palinuridea. Cirripedians reached a conspicuous peak of39.5 ind. m−3 at St. 8 and stomatopods were more abundantparticularly at St. 6 with 19.6 ind. m−3. At the most offshore St. 1, St.2 and St. 3, however, crustaceans accounted for less than 10% of thetotal. Polychaete larvae were abundant at St. 2 and 3, with 28 and 24%of the total, respectively, and reached maximum integrated density atSt. 15 (2.8 ind. m−3). Echinoderm abundance was high at the deeperstations, reaching a maximum of 19% (4.1 ind. m−3) of the total larvaeat St. 1. In shallower stations, echinoderms were not numericallyabundant, accounting for less than 5% of total larvae.

The analysis of correlation showed, as expected, significantnegative relationships between distance of the coast and severalbiotic variables (chl-a, total larval density and crustaceans) and % ofSACW (Table 2). Brachyura and Stomatopodawere the onlymajor taxaof larvae positively correlated to the % of SACW in the water column,whereas significant negative correlations were found between % ofSACW and bivalves, gastropods and polychaetes (Table 2). Integratedconcentrations of chl-a correlated positively with integrated larval,brachyuran and stomatopod densities.

4.3. Brachyuran larvae

A total of 351 brachyuran larvae were identified in 31 taxa. Twentyout of 31 taxa belonged to coastal and shelf species with their adultcounterpart distribution ranging from the intertidal to depths ofapproximately 160 m (Table 3). The remaining taxa included specieswith wider bathymetric distribution or individuals only identified at

Fig. 5. Same as Fig. 3 for transect #3.

Fig. 6. Same as Fig. 3 for transect #4.

129M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

Fig. 7. Total density of larvae against distance from the shore. Upwelling and eddy zonesare approximate and may vary over time.

130 M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

higher larval taxon level, preventing an accurate determination oftheir bathymetric range. Zoea larvae of Parthenopidae sp. and Portu-nus cf. spinicarpus were the most frequent and abundant, occurringfrom nearshore stations up to 116 km offshore. Early stages of thesespecies were sampled only at coastal stations (inner and mid shelf) invarious depths, whereas later stages occurred in the outer shelf andslope at relatively deeper waters (Table 4). Pinnixa sp. and Eurypa-

Table 2Spearman's Rank correlation yields among variables used in this study.

Depth %SACW Chl-a Larvae Echinodermata Polyc

Depth 1 −0.658 −0.794 −0.611 0.336 0.464%SACW 1 0.393 0.110 −0.301 −0.5Chl-a 1 0.637 −0.262 −0.2Larvae 1 0.107 −0.0

We used integrated values of chl-a and larvae densities, station depths and percentage of SACand underlined coefficients with pb0.01.

Table 3Densities of Brachyura zoeae (ind. m−3) and spatial distributions of corresponding adult ta

Taxa/station # Shelf waters

19 15 7 13 4

Brachyura sp.1Brachyura sp.2 0.50 0.05Brachyura sp.3 0.03Homolidae sp. 0.03Calappidae sp. 0.02Calappa sp. 0.09 0.09Hepatus sp. 0.05Leucosiidae sp. 0.03 0.03Majidae sp. 0.06 0.04 0.07Anasimus latus 0.12 0.02Stenorhynchus seticornis 0.03Parthenopidae sp. 0.08 3.40 0.14 0.03Cancroidea sp.1Portunus cf. spinicarpus 0.91 0.38 0.12 0.21Callinectes sp.1Callinectes sp.2 0.03Arenaeus cribarius 0.11 0.01Goneplacidae sp. 0.02 0.21 0.20 0.14 0.04Xanthidae spp.1 0.03Xanthidae spp.2 0.56 0.14 0.09Eriphiidae sp. 0.09 0.02Eriphia gonagra 0.02Menippe nodifrons 0.03Eurypanopeus sp. 0.03Hexapanopeus sp.Pinnixa sp. 0.03Uca sp. 0.03Sesarmidae sp. 0.24 0.03Sesarma rectum 0.14Pachygrapsus sp.Pachygrapsus gracilis 0.05Not identified 2.44 13.04 0.26 0.09 0Total 3.43 14.60 5.39 1.10 0.13

nopeus sp. larvae were found mainly in deeper strata even at nightsamples. It is worth noting that larvae of estuarine species such as Ucasp., Sesarma rectum and Sesarmidae sp. were sampled very far fromcoast (St. 1).

5. Discussion

In this study, physical and biological time-series taken at a cross-shelf transect were used to infer the vertical and cross shelfdistribution patterns of benthic invertebrate larvae during a consistentcoastal upwelling event off Cabo Frio. This area is analogous to the eastcoast of the US, where several authors have investigated larvaltransport mechanisms during coastal upwelling and downwellingevents (e.g., Shanks et al., 2000; Garland et al., 2002; Shanks et al.,2003; Shanks and Brink, 2005; Ma et al., 2006). Those studiesdemonstrated that some types of larvae remained in inner-shelf areas(b30 km offshore) through depth-keeping mechanisms (Shanks andBrink, 2005) thus, avoiding passive cross-shelf transportation. Here,we revisited this question for a broader spatial scale (~100 kmoffshore).

During our sampling, the upper water column (100 m) wasvertically stratified (Figs. 3, 4, 5 and 6) with SACW moving onshore at

haeta Gastropoda Bivalvia Cirripedia Brachyura Stomatopoda

0.122 0.142 −0.510 −0.890 −0.79791 −0.631 −0.472 0.171 0.465 0.57820 −0.070 −0.171 0.429 0.844 0.73314 0.346 0.108 0.606 0.705 0.432

W in the 5 strata (0–100 m) sampled. In bold correlation coefficients with pb0.05, bold

xa.

Slope waters Adults distribution

11 3 2 1

0.02 Depend on speciesDepend on speciesDepend on speciesDepend on species

0.03 0.03 0.09 Coastal waters — 220 mCoastal waters — 220 mCoastal waters — 160 mCoastal waters — 160 mDepend on speciescoastal waters — 160 mCoastal waters — great depths

0.18 0.01 Coastal waters — 200 m0.01 Depend on species0.02 Coastal waters — 120 m0.01 Coastal waters — 90 m0.02 0.02 Coastal waters — 90 m0.04 Coastal waters — 70 m

Depend on species0.02 Depend on species

Depend on species0.02 0–5 m

0–5 m0–5 m

0.09 0–5 m0.07 0–25 m0.02 0–75 m0.30 Estuarine species0.28 Estuarine/coastal species

0.02 Estuarine species0.02 Coastal waters

Coastal waters0.42 0 0.03 0.141.39 0.16 0.09 0.27

Table 4Range distribution of brachyurans taxa in the water column and their zoeal developmental stage.

Taxa/station # Shelf waters Slope waters

Inner Mid Outer

19 15 7 13 4 11 3 2 1

Callinectes sp.1 0–20 mIV

Hexapanopeus sp. 20–40 mII

Pachygrapsus sp. 0–20 mV

Parthenopidae sp. 0–20 m 0–60 m 0–20 m 0–20 m 40–80 m 80–100 mI I–II II IV IV–V IV–V

Portunus spinicarpus 0–30 m 0–20 m 0–40 m 20–40 m 20–60 mI I–II VI VI VI

Arenaeus cribarius 0–20 m 0–20 m 0–20 mVI VI VI

Callinectes sp.2 0–20 m 0–20 m 0–20 mII III V

Eurypanopeus sp. 60–80 m 20–40 mI I

Eriphiidae sp. 0–20 m 0–20 m 0–20 mIII III III

Pinnixa sp. 20–40 m 60–80 mV III–IV

Uca sp. 0–20 m 20–40 mI–III I–III

Sesarma rectum 0–40 m 0–20 mI–II II

Sesarmidae sp. 0–40 m 60–80 m 20–80 mI–IV III–IV II–III

Pachygrapsus gracilis 0–20 mI

Calappa sp. 0–20 m 0–20 mII II

Hepatus sp. 0–20 mIII–IV

Leucosiidae sp. 0–20 m 0–20 mfinal III–IV

Anasimus latus 60–100 m 0–20 mII II

Menippe nodifrons 0–20 mII

Eriphia gonagra 0–20 mIII

131M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

subsurface depths. The increased number of coastal species, such asbarnacle, brachyuran, stomatopod and other crustacean larvae, as wellas positive correlations between larvae abundance and chl-a (Tables 1and 2) suggest an association with the upwelling event. In the CaboFrio region, the vertical movement of the 18 °C isotherm was used toestimate upwelling speed; vertical current speed in nearshore areaswas estimated at the order of 0.1 mm s−1 (Valentin et al., 1987).However, the observed strong shore-parallel flow in surface waters,flowing southwards within the first 40 km with up to 20 cm s−1

(Fig. 2), is likely the most important factor determining the net larvaltransport.

The coastline irregularities of Cabo Frio and the spatial variation inwind stress can force the formation of a cool upwelled water plume,which intensifies in the offshore direction (Carbonel, 1998). Inaddition, several authors have described the connection of thisplume with the meandering pattern of the BC (Campos et al., 2000;Silveira et al., 2004; Calado et al., 2006). Coastal jets of cold andchlorophyll-rich waters could be seen not only off Cabo de São Tomé(22.5°S and 41.5°W, north of the study area, see also Calado et al.,2006), but also southwards of Cabo Frio (Fig. 1), and according toLorenzzetti and Gaeta (1996) they can be found up to 300 kmsouth from Cabo Frio. It is likely that coastal larvae entrapped in theseplumes are advected along the continental margin to offshore regions;otherwise some larvae must have the ability to avoid offshoretransport.

Elevated larval abundances at St. 1 were associated with the BCmeso-scale meander observed in Figs. 1 and 2, coming from the Caboof São Tomé shelf area. This scenario becomes particularly clear whenthe meander pattern (Fig. 2), thermal structure and larval concentra-tion (Fig. 3) are confronted. It seems that the higher abundances oflarvae were located at the offshore (and intense) border of themeander (St. 1). From the offshore edge of the clockwise meander(St. 1) to the center of the meander (St. 2 and 3, where the velocitiesare basically nil, Fig. 2), the number of larvae decreased considerably(Fig. 3, Table 1). Although gastropod was the most abundant group inoffshore waters, larval composition, with highest abundances ofechinoderm, polychaete, bivalve and sipunculan larvae, was relativelydistinct compared to coastal larvae (Table 1). In the case of gastropods,a significant number of larvaewere probably transported bymeandersdrifting cross-shelf to offshore areas (Fig. 3 and Table 1). Few larvae ofcoastal brachyuran species were identified in stations 1, 2 and 3(Table 3). Nevertheless, larvae of Uca sp., Sesarma rectum andSesarmidae sp.1 were sampled over 30 km offshore (Table 3), whichis unusual since estuarine species rarely occur in distances furtherthan 10 km from the coast (Shanks and Eckert, 2005; dos Santos et al.,2008). On the other hand, larvae of coastal species such as Hepatus sp.,Leucosiidae sp., Anasimus latus, Menippe nodifrons, Calappa sp. andEriphia gonagra were not sampled further than 58 km offshore inthe present work, despite being found at surface (Tables 3 and 4)where offshore transport tends to be more important. Our results

132 M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

suggest that larval vertical migration may prevent the loss of larvaefrom local populations through offshore surface advection tooligotrophic open waters off the Brazilian coast by strong cross-shelftransport during upwelling events off Cabo Frio region (Figs. 1 and 2).Moving to deeper layers, larvae may be carried back to the coast in thecolder SACW. However, it is clear that our results could be biased bythe single sample taken at the most offshore stations and the lack ofidentification to the species level, thus we are limited to suggest thatupwelling plumes as wells as meso-scale meanders and eddies couldpotentially play an important role for larvae dispersal in the tropicalBrazilian coast.

Larval dispersal due to shelf hydrography depends greatly on theirvertical distribution and time spent in different water layers. In orderto remain close to suitable areas for settlement, coastal species mustavoid seaward advection by sinking towards SACW depths. dos Santoset al. (2008) reported a decapod diel vertical migration on thePortuguese coast, where larvae remained between 20–55 m duringday time, migrating towards the surface at night. A similar strategywas reported in central Chile, where competent gastropod larvaeavoid large-scale offshore transport by linking reverse diel verticalmigration and upwelling circulation, and thus restricting theirposition to a zone delimited by the upwelling plume (Poulin et al.,2002). However, no clear pattern on diel vertical migration wasobserved during our sampling period. Cirripedia and Brachyurasampled at night, for example, were positioned in high numbers atthe surface (0–40 m interval) at St. 7, whereas in St. 9 the majority oflarvae were found at 80–100 m strata (Fig. 4). On the other hand,Parthenopidae sp. and Portunus cf. spinicarpus, the most abundantand frequent Brachyuran taxa, were collected near surface duringnocturnal sampling and in greater depths during the day, corroborat-ing the most common diel vertical migration pattern among decapodslarvae (Queiroga and Blanton, 2005). By partitioning time betweenthe surface layer moving seaward and the bottom layer, where thecompensating counter-current develops, larvae can avoid seawarddispersal (Peterson, 1998).

From an evolutionary perspective, Shanks and Brink (2005)highlighted the potential of slow-swimming coastal larvae to adjusttheir vertical distribution and, thus, remain close to suitable settle-ment areas. Larvae sinking to subsurface layers would be transportedback to the inner shelf, if a continuous intrusion of SACW occurs. Thisbehavior may also maintain larval populations close to high foodsupply zones (i.e., over inner and mid-shelf chlorophyll-rich areas;Figs. 1, 3, 4, 5 and 6), particularly in the study area. Gastropoda,Cirripedia and Brachyura were the most abundant larvae in shelfwaters, and all of them showed a vertical displacement consistent tothe depth-keepingmechanism suggested by Shanks and Brink (2005).The stations where high number of larvae was observed down toN60 m in the water column were located in areasN40 km from thecoast (at Sts. 8, 9, 12 and 17, see Figs. 4, 5 and 6). Those areas are off thezone delimited by the shore-parallel coastal current (Fig. 2).Consistently, the highest larval densities occurred at surface watersin areasb40 km from the coast, which are characterized by high chl-aconcentrations (at Sts. 6, 15 and 18).

6. Conclusions

Despite limitations concerning larval identification and spatio-temporal scales used in the present study, our findings revealed thateven estuarine and coastal species experiencing excursions to mid-shelf areas (N50 km from the coast, St. 8 in Fig. 4) can avoid offshoretransport by sinking down to 80–100 m in the water column, whereSACW could potentially transport them back to coastal areas duringupwelling events. Alternatively, excursions of upwelling plumes couldserve as corridors for tropical larvae dispersion along subtropicalareas in the Brazilian continental margin. Future studies in the studyarea, employing technological advances on sampling, processing and

species identification, are needed to comprehensively address the roleof mesoscale features, such as upwelling drifting plumes and cyclonicmeanders, influencing benthic larval distribution in thewater column.

Acknowledgments

The authors wish to thank FAPESP (Grant 01/00165-2 for PYGS and04/06369-7 for MYY) and CNPq/PRONEX for the financial support toDEPROAS project. Special thanks to C. L. Lopes, M. L. Zani-Teixeira, M.Katsuragawa, T. E. Silva and M. Pompeu whose help was veryimportant to the present work. T. P. Costa is acknowledged for editingthe figures. We thank NASA, DAAC and the SeaWiFs Project for dataand software. A. Shanks and R. Rykaczewski provided improvingcomments to the manuscript.

References

Botsford, L.W., 2001. Physical influences on recruitment to California Currentinvertebrate populations on multiple scales. ICES J. Mar. Sci. 58, 1081–1091.

Calado, L., Gangopadhyay, A., Silveira, I.C.A., 2006. A parametric model for the BrazilCurrent meanders and eddies off southeastern Brazil. Geophys. Res. Lett. 33,L12602. doi:10.1029/2006GL026092.

Campos, E.J.D., Ikeda, Y., Castro, B.M., Gaeta, S.A., Lorenzzetti, J.A., Stevenson, M.R., 1996.Experiment studies circulation in the Western South Atlantic. EOS Trans. Am.Geophys. Union. 77, 253–259.

Campos, E.J.D., Velhote, D., Silveira, I.C.A., 2000. Shelf-break upwelling driven by BrazilCurrent cyclonic meanders. Geophys. Res. Lett. 27, 751–754.

Carbonel, C., 1998. Modelling of upwelling in the coastal area of Cabo Frio (Rio deJaneiro, Brazil). Braz. J. Oceanogr. 46, 1–17.

Castro, B.M., Miranda, L.B., 1998. Physical oceanography of the Western AtlanticContinental Shelf located between 4°N and 34°S. Sea 11, 209–251.

Clough, L.M., Ambrose, W.G., Ashjian, C.J., Piepenburg, D., Renaud, P.E., Smith, S.L., 1997.Meroplankton abundance in the Northeast Water Polynya: from oceanographicparameters and benthic abundance patterns. J. Mar. Syst. 10, 343–357.

Cowen, R.K., Lwiza, K.M., Sponaugle, S., Paris, C.B., Olson, D.B., 2000. Connectivity ofmarine populations: open or closed? Science 287, 857–859.

dos Santos, A., Santos, A.M.P., Conway, D.V.P., Bartilotti, C., Lourenço, P., Queiroga, H.,2008. Diel vertical migration of decapod larvae in the Portuguese coastal upwellingecosystem: implications for offshore transport. Mar. Ecol. Prog. Ser. 359, 171–183.

Garland, E., Zimmer, C.A., Lentz, S., 2002. Larval distribution in inner-shelf waters: theroles of wind-driven cross-shelf currents and diel vertical migrations. Limnol.Oceanogr. 47, 803–817.

Gonzalez-Rodriguez, E., 1994. Yearly variation in primary productivity of marinephytoplankton from Cabo Frio (RJ, Brazil) region. Hydrobiologia 294, 145–156.

Gonzalez-Rodriguez, E., Valentin, J.L., André, D.L., Jacob, S.A., 1992. Upwelling anddownwelling at Cabo Frio (Brazil). J. Plankton Res. 14 (2), 289–306.

Holm-Hansen, O., Lorenzen, C.J., Holems, R.W., Strickland, J.D.H., 1965. Fluorometricdetermination of chlorophyll. J. Cons. Perm. Int. Explor. Mer 30 (1), 3–15.

Lorenzzetti, J.A., Gaeta, S.A., 1996. The Cape Frio upwelling effect over the South BrazilBight northern sector shelf waters: a study using AVHRR images. Int. Arch.Photogramm. Remote Sens. 31, 448–453.

Ma, H., Grassle, J.P., Chant, R.J., 2006. Vertical distribution of bivalve larvae along a cross-shelf transect during summer upwelling and downwelling. Mar. Biol. 149,1123–1138.

Matsuura, Y., 1996. A probable cause of recruitment failure of the Brazilian sardineSardinella aurita population during the 1974/75 spawning season. S. Afr. J. Mar.Sci. 17, 29–35.

Metzler, P.M., Gilbert, P.M., Gaeta, S.A., Lublan, J.M., 1997. New and regenerateproduction in South Atlantic off Brazil. Deep-Sea Res. I 44, 363–384.

Nakata, H., Kimura, S., Okazaki, Y., Kasai, A., 2000. Implications of meso-scale eddiescaused by frontal disturbances of the Kuroshio Current for anchovy recruitment.ICES J. Mar. Sci. 57, 143–152.

Paris, C.B., Cowen, R.K., Lwiza, M.M., Wang, D., Olson, D.B., 2002. Multivariate objectiveanalysis of the coastal circulation of Barbados, West Indies: implication for larvaltransport. Deep-Sea Res. I 49, 1363–1386.

Peterson, W., 1998. Life cycle strategies of copepods in coastal upwelling zones. J. Mar.Syst. 15, 313–326.

Poulin, E., Palma, A., Leiva, G., Narvaez, D., Pacheco, R., Navarrete, S., Castilla, J., 2002.Avoiding offshore transport of competent larvae during upwelling events: the caseof the gastropod Concholepas concholepas in Central Chile. Limnol. Oceanogr. 47,1248–1255.

Queiroga, H., Blanton, J., 2005. Interactions between behavior physical forcing in thecontrol of horizontal transport of decapod crustacean larvae. Adv. Mar. Biol. 47,107–213.

Rodrigues, R.R., Lorenzzetti, J.A., 2001. A numerical study of the effects of bottomtopography and coastline geometry on the Southeast Brazilian coastal upwelling.Cont. Shelf Res. 21, 371–394.

Roughgarden, J., Pennington, J.T., Stoner, D., Alexander, S., Miller, K., 1991. Collisions ofupwelling fronts with the intertidal zone: the cause of recruitment pulses inbarnacle populations of central California. Acta Oecol. 12 (1), 35–51.

133M.Y. Yoshinaga et al. / Journal of Marine Systems 79 (2010) 124–133

Shanks, A.L., 1995. Mechanisms of cross-shelf dispersal of larval invertebrates and fish.In: Mc Edward, L.R. (Ed.), Ecology of Marine Invertebrate Larvae. CRC Press, Florida,pp. 332–367.

Shanks, A.L., Brink, L., 2005. Upwelling, downwelling, and cross-shelf transport ofbivalve larvae: test of a hypothesis. Mar. Ecol. Prog. Ser. 302, 1–12.

Shanks, A.L., Eckert, G., 2005. Population persistence of California Current fishes andbenthic crustaceans: a marine drift paradox. Ecol. Monogr. 75, 505–524.

Shanks, A.L., Largier, J., Brubaker, J., 2003. Observations on the distribution ofmeroplankton during an upwelling events. J. Plankton Res. 25, 645–667.

Shanks, A.L., Largier, J., Brink, L., Brubaker, J., Hoof, R., 2000. Demonstration of theoffshore transport of larval invertebrates by the shoreward movement of anupwelling front. Limnol. Oceanogr. 45, 230–236.

Silveira, I.C.A., Calado, L., Castro, B.M., Cirano, M.J., Lima, A.M., Mascarenhas, A.S., 2004. Onthe baroclinic structure of the Brazil Current-IntermediateWestern Boundary CurrentSystem at 22°–23°S. Geophys. Res. Lett. 31, L14308. doi:10.1029/2004GL020036.

Silveira, I.C.A., Schmidt, A.C.K., Campos, E.J.D., Godoi, S.S., Ikeda, Y., 2000. A Corrente doBrasil ao Largo da Costa Leste Brasileira. Rev. Bras. Oceanogr. 48, 171–183.

Thorson, G., 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol.Rev. 25, 1–45.

Valentin, J.L., 1984. Spatial structure of the zooplankton community in the Cabo Frioregion (Brazil) influenced by coastal upwelling. Hydrobiologia 113, 183–199.

Valentin, J., André, D.L., Jacob, S.A., 1987. Hydrobiology in the Cano Frio (Brazil)upwelling: two-dimensional structure and variability during a wind cycle. Cont.Shelf Res. 7, 77–88.

Wing, S.R., Botsford, L.W., Ralston, S.V., Largier, J.L., Organ, L.E.M., 1995. Spatial structureof relaxation events and crab settlement in the northern California upwellingsystem. Mar. Ecol. Prog. Ser. 128, 199–211.