Cross-shelf transport of pink shrimp larvae: interactions of tidal currents, larval vertical migrations and internal tides

of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report

Leadership & Management


Views: 16 | Pages: 18

Extension: PDF | Download: 0

Cross-shelf transport of pink shrimp larvae: interactions of tidal currents, larval vertical migrations and internal tides
  MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog SerVol. 345: 167–184, 2007 doi: 10.3354/meps06916 Published September 13 INTRODUCTION Most coastal species of fishes and invertebratesspawn offshore in shelf waters and early life stagesmigrate to coastal estuarine nursery grounds. Al-though little is known about the ecological coupling ofestuaries and the coastal ocean, progress has beenmade in understanding transport mechanisms of im-portant species that connect offshore and coastalgrounds (for review see Shanks 1995). The dominantmechanisms that may yield successful cross-shelftransport include wind-driven Ekman transport,upwelling fronts, moving frontal systems, countercur-rents generated by coastal eddies, coastal boundarylayers and non-linear internal tides and bores (e.g.Shanks 1995, 2006, Pineda 1999, Epifanio & Garvine © Inter-Research 2007 ·*Email: Cross-shelf transport of pink shrimp larvae:interactions of tidal currents, larval vertical migrations and internal tides Maria M. Criales 1, *, Joan A. Browder 2 , Christopher N. K. Mooers 1 ,Michael B. Robblee 3 , Hernando Cardenas 1 , Thomas L. Jackson 2 1 Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami,Florida 33149, USA 2 NOAA Fisheries, Southeast Fisheries Science Center, 75 Virginia Beach Drive, Miami, Florida 33149, USA 3 US Geological Survey, Center for Water and Restoration Studies, 3110 SW 9th Avenue, Ft Lauderdale, Florida 33315,USA ABSTRACT: Transport and behavior of pink shrimp Farfantepenaeu s duorarum larvae were investi-gated on the southwestern Florida (SWF) shelf of the Gulf of Mexico between the Dry Tortugasspawning grounds and Florida Bay nursery grounds. Stratified plankton samples and hydrographicdata were collected at 2 h intervals at 3 stations located on a cross-shelf transect. At the Marquesasstation, midway between Dry Tortugas and Florida Bay, internal tides were recognized by anom-alously cool water, a shallow thermocline with strong density gradients, strong current shear, and ahigh concentration of pink shrimp larvae at the shallow thermocline. Low Richardson numbersoccurred at the pycnocline depth, indicating vertical shear instability and possible turbulent transportfrom the lower to the upper layer where myses and postlarvae were concentrated. Analysis of verti-cally stratified plankton suggested that larvae perform vertical migrations and the specific behaviorchanges ontogenetically; protozoeae were found deeper than myses, and myses deeper than postlar-vae. Relative concentrations of protozoea in the upper, middle and bottom layers were consistent witha diel vertical migration, whereas that of postlarvae and myses were consistent with the semidiurnaltides in phase with the flood tide. Postlarvae, the shallowest dwellers that migrate with a semidiurnalperiodicity, experienced the largest net onshore flux and larval concentrations were highly correlatedwith the cross-shelf current. These results provide the first evidence of an onshore tidal transport (atype of selective tidal stream transport, STST), in decapod larvae migrating in continental shelfwaters offshore, ca. 100 km from the coast and at a depth of 20 m, while approaching the coastal nurs-ery grounds. Longer time series would be necessary to establish whether internal tides play any rolein the larval onshore transport of this species and determine if the STST is the dominant onshoretransport mechanism.KEY WORDS: Selective tidal transport · Pink shrimp larval behavior · Internal tides · Farfantepenaeus duorarum Resale or republication not permitted without written consent of the publisher   Mar Ecol Prog Ser 345: 167–184, 2007 2001, Sponaugle et al. 2005). Another potential mech-anism for cross-shelf transport includes tides in associ-ation with a vertical movement behavior.Selective tidal stream transport (STST) is an impor-tant mechanism by which planktonic organisms usevertical differences of water velocity in a shear currentsystem to promote transport in an ‘appropriate’ direc-tion (for review see Forward & Tankersley 2001,Queiroga & Blanton 2005). Decapod larvae that spendsome portion of their life cycle in estuarine systems useSTST to promote transport, resulting in retention,export or reinvasion (Shanks 1995). The STST hasbeen implicated in transport of larvae and postlarvaeentering coastal nursery grounds, but not in cross-shelftransport on shelf waters. A STST has been identifiedin postlarvae of penaeid shrimps including the pinkshrimp Farfantepenaeu s duorarum (i.e. Rothlisberg etal. 1996, Criales et al. 2006), crab megalopae (for re-view see Forward & Tankersley 2001, Queiroga & Blan-ton 2005) and estuarine fish larvae (i.e. Rowe & Epi-fanio 1994). In order to immigrate to or remain withinthe estuary, planktonic larvae need to have some kindof behavioral adaptation to overcome the opposite fluxof the water associated with the tide. A growing bodyof laboratory and field observations suggests that ver-tical displacements are not a passive phenomenon, butthat larval behavior indeed influences this process (e.g.Forward et al. 2003). Environmental factors involved incontrolling the vertical migrations arelight, pressure and gravity (e.gSulkin1984), but salinity, temperature andturbulence also affect behavior (e.gWelch & Forward 2001). Most of thesefactors change cyclically, and larvaerespond with cyclical vertical move-ments.The pink shrimp is an important spe-cies in southern Florida; it supports themultimillion dollar Tortugas fishery onthe southwestern Florida (SWF) shelf inthe vicinity of Dry Tortugas and KeyWest and serves as an important preyspecies in Florida Bay, the principal in-shore nursery habitat (Browder et al.2002) (Fig. 1). As with most coastal– es-tuarine species, the spawning andnursery grounds of the pink shrimpinFlorida are spatially separated.Spawning areas are northeast of theDry Tortugas and juveniles occupynursery grounds in Florida Bay about150 km to the east-northeast (Roberts1986). Larval development is rapid,passing through 5 nauplii, 3 proto-zoeae, and 3 myses in about 15 d(Dobkin 1961), and 3 to 6 additional planktonic postlar-val stages in another 15 d (Ewald 1965). Two main path-ways have been hypothesized for pink shrimp larvaemigrating from Dry Tortugas to Florida Bay: (1) larvaedrift south-southeast downstream with the Florida Cur-rent front and enter Florida Bay through the tidal inletsof the middle Florida Keys (Munro et al. 1968, Criales etal. 2003); and (2) larvae move northeast across the SWFshelf and enter Florida Bay at its northwestern bound-ary (Jones et al. 1970, Criales et al. 2006). Transport via Florida Current– Florida Keys inlets was, until recently,the most widely recognized larval route because Ek-man transport generated by winds and coastal counter-current flow generated by cyclonic eddies may providefavorable onshore transport mechanisms along theFlorida Keys coast (Lee & Williams 1999, Yeung et al.2001). Nevertheless, recent results from a comparativestudy on these 2 migration pathways indicated that thegreatest postlarval influx (>70%) occurred at the north-western border of Florida Bay, suggesting a viableeast–northeast transport across the SWF shelf (Crialeset al. 2006). Florida Bay lies at the southern end of theFlorida peninsula and its western border connects di-rectly with the inner SWF shelf.Winds in southern Florida are seasonal; weak windsprimarily from the southeast dominate in summer,stronger winds from the northeast in the fall, and mod-erate to strong winds from the east to northeast in the 168Fig. 1. Study area showing the 3 sampling stations (DT = Dry Tortugas, MQ =Marquesas, ON = Onshore). The area enclosed by dashed line representsthe Tortugas fishing grounds of pink shrimp Farfantepenaeus duorarum. The inset map indicates major currents in the Gulf of Mexico and off thecoast of Florida. YC = Yucatan Current, LC = Loop Current, FC = Florida Current, GS = Gulf Stream  Criales et al.:Cross-shelf larval transport winter and spring (Lee & Williams 1999). The oceancirculation around the Dry Tortugas is highly dynamicand is affected by the interaction of the Loop-FloridaCurrent system and warm core (anticyclonic) and coolcore (cyclonic) eddies, common phenomena associatedwith the Loop Current front (Fratantoni et al. 1998).Interactions with the Loop Current on the SWF shelfare sporadic and eddies may not play a major role indispersal of pink shrimp larvae because eddies usuallyform southwest of the Dry Tortugas (Lee et al. 1994)and the principal nursery ground is northeast of theDry Tortugas. Circulation on the SWF shelf is forcedmainly by tidal currents, winds and buoyancy fluxes(Weisberg et al. 1996). Tidal currents are strong in thecross-shelf direction; in contrast, subtidal flows areweak and mainly along-shore as a direct response towind events (Lee et al. 2002). Tides on the SWF shelfare mixed with contributions from both diurnal andsemidiurnal components (Smith 2000). Freshwater dis-charges from the Shark River affect a broad area of theSWF shelf (Lee et al. 2002). Vertical stratification of thewater column occurs in spring and summer as aresponse to variations in wind-mixing, heating andriver runoff (He & Weisberg 2002). Simulations of transport derived from a Lagrangian(horizontal) model coupled with larval vertical behav-iors identified STST as a viable transport mechanismfor pink shrimp larvae across the SWF shelf (Crialeset al. 2006). Planktonic stages, which migrate verti-cally to position themselves near the surface duringthe flood tide, could consistently travel between 100and 200 km in 30 d across the wide SWF shelf, andhypothetically 85% of the population could travel 150km. With this mechanism planktonic stages takeadvantage of the strong eastward surface flow thatdominates the SWF Shelf. A STST behavior previ-ously has been identified in pink shrimp postlarvaeentering estuarine nursery grounds, but not in earlierstages during their onshore migration across theshelf. Other potential mechanisms of cross-shelftransport on the SWF shelf, such as non-linear inter-nal waves and bores, Ekman transport or coastalboundary layers, have not been identified. Which ofthese mechanisms, if any, occur on the SWF shelf andtransport larvae to near-shore areas is unknown. Italso is unclear if similar mechanisms regulate trans-port throughout the journey from offshore spawningareas to estuaries or if estuarine ingress of coastalorganisms is more than a single stage process. Thisresearch was conducted to determine (1) potentialtransport mechanisms on the SWF shelf,(2) behaviorof early planktonic stages during their migrationacross the SWF shelf and possible environmentalcues associated with the behavior, and(3) implica-tions of behavioral patterns for transport. MATERIALS AND METHODSField sampling. Three stations on the SWF shelf,separated from each other by 60 km, were sampledalong a cross-shelf transect between the Dry Tortugasand Florida Bay, on board the RV ‘Gandy’ on July 2–5,2004 (Fig. 1). The Dry Tortugas (DT) station, the far-thest offshore, was selected for its proximity to theknown spawning grounds of pink shrimp (Jones et al.1970, Roberts 1986) about 30 km northeast of the DryTortugas at a depth of 30 m. The Marquesas (MQ) sta-tion was 30 km north of Marquesas at a depth of 20 m,and the onshore (ON) station was about 40 km south-west of Cape Sable on the western border of FloridaBay at a depth of 10 m. Sampling was conducted dur-ing the full moon in accordance with previous resultsthat showed the highest spawning activity of pinkshrimp females occurred between the full and lastquarter moon (Munro et al. 1968). A 300 kHz bottommounted Workhorse Acoustic Doppler current profiler(ADCP) was deployed at the DT and MQ stationsatdepths of 30 and 20 m, respectively, to measurecurrents with vertical bin sizes of 1 m.At each station stratified plankton samples were col-lected at 3 fixed depths with a 1 m 2 Tucker trawl net(333 µm mesh). The Tucker trawl net was equippedwith 3 nets towed at a speed of 1.5 to 2.5 knots(0.8–1.3m s –1 ) and at a wire angle of 45 ° . Calibratedflow meters were mounted in the mouth of each net andmeasured volume of water, which ranged from 202 to835 m 3 (mean ±SD 445 ±112 m 3 ). Each net of theTucker trawl opened at the desired depth, and wastowed for approximately 10 min and then closed andretrieved. The depth intervals (d i ) at the DT stationwere every 10 m (d 1 = 5, d 2 = 15 and d 3 = 25 m); at theMQ station every 5 m (d 1 = 5, d 2 = 10 and d 3 = 15 m) andat ON station every 3 m (d 1 = 3, d 2 = 6 and d 3 = 9 m).Samples were collected at each station at 2 h intervals.At the DT and MQ stations samples were collected for24 h, and at the ON station for 14 h. Plankton samplingat the DT station began at 12:00 h on July 2 and contin-ued until 10:00 h on July 3, 2004, at MQ station from17:00 h on July 3 to 15:00 h on July 4, 2004 and at theON station from 19:00 h on July 4 to 07:00 h on July 5,2004. The day–night cycle in July in southern Florida is14 h light/10 h dark, so 7 samples were collected duringthe daylight and 5 in the dark at DT and MQ stationsand 2 samples during the daylight and 5 in the dark atON station. Water column temperature and conductiv-ity were measured with a Seabird SBE-25 CTD cali-brated within 2 mo before use. The CTD casts wereconducted immediately after each plankton tow fromnear the surface to within 2 m of the bottom. Salinityand density were computed from the conductivity,depth and temperature readings. 169  Mar Ecol Prog Ser 345: 167–184, 2007 Physical data analysis. The ADCP time series at5min intervals were smoothed with a low-pass filter.Hourly averaged wind speed and direction data wereobtained from the Dry Tortugas–C–MAN station anddata were smoothed with a 3 h and 24 h low-pass filter.Hourly tidal currents for the ON station were estimatedusing predicted tides software (Tides & Currents 1999,Nobeltec Nautical). Current and wind vectors wereresolved into cross-shelf, u = east (+) and west (–), andalong-shore v  = north (+) and south (–) components,with no rotation to maintain a proper orientation to thecoast. The x  -component represented flood and ebbcurrent with flood defined as positive values and ebbnegative.Wind vectors were converted to wind stress(theforce applied to the sea surface by the wind)calculated as: (1)where τ x,y correspond to the cross-shelf and alongshore wind stress (dynes cm –2 ), respectively,  ρ a is airdensity (1.3 × 10 –3 g cm –3 ), C  d is a drag coefficient (1.5 × 10 –3 ), is wind velocity vector and u is cross-shoreand  ν along-shore wind components.At the MQ station, vertical density gradients, currentshear and Richardson numbers (Ri ) were calculatedfrom the 24 h time series. Richardson numbers, whichare indicators of vertical turbulent mixing, are indica-tive of exchange between layers. For this calculation,density gradients and gravity act as stabilizing factors,and the current shear acts as the destabilizing factor(Baines 1986). The ratio between these 2 effects (gradi-ent) is called the Richardson number, which is adimensionless measure given by:(2)where g = gravity acceleration 9.8 m s –1 , ρ = density ofseawater (kg m –3 ), ∆ρ / ∆ z  = density gradient at 1mdepth interval, ∆ u ( v  )/ ∆ z  = current shear (cross-shoreand alongshore) at 1 m depth interval. If Ri falls below 1  ⁄  4 , the current shear is relatively strong, and stratifica-tion is relatively weak. When this happens the watercolumn is unstable; billows form and produce turbu-lent rotors, wherein the fluid from one layer gets mixedinto the other layer. Larval data analysis. Pink shrimp larvae weresorted, identified and classified into the 3 main plank-tonic stages: protozoeae, myses and postlarvae(Dobkin 1961). The number of larvae at each depth ( d) was standardized to larval concentration (N d  = larvae × 10 3 m –3 ) of seawater filtered. The standardized larvalconcentrations (N i  ) were used to compare catches oflarvae among depth profiles (every 2 h) at a single sta-tion. Larval abundance (larvae × 10 2 m –2 ) was used tocompare catches of larvae among stations:(3)where A is the larval abundance, i  is net 1, 2, 3 fishedduring a tow, N i is thenumber of larvae in net i, D  i  isthe depth interval sampled by net i, V  i  is the volume ofwater filtered by net i. Larval concentrations were log-transformed [ln( x  +1)] and tested for normality (Shapiro-Wilks) and ho-mogeneity of variance (Barlett’s test). A 3-way ANOVAfor each life stage was conducted on the dataset of nat-ural log-transformed concentrations. Factors analyzedwere depth (three levels = surface, middle and bottomlayer), light (two levels = light [daytime] and no light[nighttime]), and tidal phase (two levels = ebb andflood). Residuals from the ANOVAs were tested for sig-nificant autocorrelation using the ‘acf’ function of S-PLUS. Then the srcinal input data for the ANOVA wasfiltered to remove the autocorrelation (Emery & Thom-son 1997) and the ANOVA was repeated using the fil-tered data as input. Filtering was accomplished byauto-regression. The filtered dataset consisted of theresiduals of the auto-regression, which contained all in-formation in the srcinal data except the autocorrela-tion that was removed. Our ANOVA design was unbal-anced; thus, the type III sum of squares was computedfor determining significance (Underwood 1981).The weighted mean depth (WMD) of the larvae forprofile (i) was calculated at the MQ station to inspectthe variation of the vertical position of larvae in thewater column and identify the type of temporal pattern(diurnal or semidiurnal variation):(4)N d  is the concentration of larvae at specific depth; D  z  is the mean depth of sampling. A mean WMD wascalculated for each larval stage.To evaluate the physical factors that most influencedlarval concentrations at each profile we used a step-wise multiple regression model. The predictive vari-ables used in the model were wind stress ( τ  x, τ  y ), cross-shelf and along-shore currents, mean water temper-ature, salinity, density and turbulence as indicated byRichardson numbers (Ri).The instantaneous larval flux (LF id  ) (larvae m –2 s –1 )was calculated for each larval stage to understand howlarvae were transported. Onshore transport dependson the quantity of organisms present in a given waterparcel and the magnitude and direction of the flow towhich they are subjected (Yannicelli et al. 2006). TheLFwas defined as the product of the instantaneousWMDNN i dz d d  D  =× = ∑∑ 13 AV D  i i i i  = × = ∑ N 13 Rig =−× ( )  +  ( ) ρρ ν∆∆∆∆∆∆ z uzz  22 V   τ ρ ν (,) (,) xy  CVu = ad uru 170  Criales et al.:Cross-shelf larval transport larval concentration (N id  ) at each depth ( d) and profile (i) by the instantaneous east to west water velocity ( U  id, ms –1 ) at each depth ( d) and profile (i) :LF id = N id  x U  id  (5)The larval velocity relative to the water column (rel-ative larval velocity) provides a quantification tounderstand whether LF results from net gain or lossvelocity (Rowe & Epifanio 1994). For each profile thevertically integrated larval velocity (LV i  ) in m s –1 wascalculated by:(6)The east to west water velocity ( U  5-10-15 )was aver-aged (AV i  ) at each profile. The difference between thelarval velocity (LV i  ) and average water velocity (AV i  )resulted in the relative larval velocity (RLV i  ). If watervelocity at any given time differs among depths, larvaemay gain or lose velocity by concentrating at the bot-tom, middle or surface layers (Queiroga et al. 1997).This simple larval model takes into account the verticalmigration of larvae in a stratified water column. How-ever the model did not considerer the shear in the tidalvelocity, and the vertical distribution of larvae wasrepresented at only 3 points or depths. RESULTSHydrographic features Dry Tortugas station (DT)—Vertical stratificationTime-depth series of temperature at the DT stationlocated at a depth of 30 m indicated strong stratifica-tion. A sharp and relatively deep thermocline was lo-cated between 15 and 22 m (Fig. 2A). Above and belowthe thermocline the variations of temperature wereminimal, ca. 29.5°C in the upper layer and ca. 25°C inthe lower layer. Salinity and density had a similar pat-tern with high gradients between 15 and 22 m(Fig.2B,C). Time series of current components ( u and  ν ) did not indicate changes in direction or speedthroughout the water column (Fig. 3). Cross-shelf cur-rents were higher in magnitude (from +40 to –60 cms –1 ) than along-shore currents (+30 to –40 cm s – ).Cross-shelf currents in the upper layer (3 to 8 m) wereonly slightly stronger than near-bottom (20 to 25 m)currents. Cross-shore and along-shore current data in-dicated the presence of mixed tides. In the cross-shelf,semidiurnal periodicity was evident with changes ofcurrent direction about every 6 h, and in the along-shore, currents showed a diurnal periodicity with anorthward flow centered in the middle of the night.Marquesas station (MQ)—Internal tidesTime-depth series of temperature at the MQ stationlocated at a depth of 20 m revealed a shallow thermo-cline located in the upper layer (Fig. 2D). The thermo-cline, halocline and pycnocline were located in theupper 8 m. Temperature and salinity data indicatedthat the depth of the thermocline and pycnoclinemoved up and down with the warmest isotherm(27.8°C) and freshest isohaline (36.24) intersecting atthe sea surface at periodic intervals (Fig. 2E,F). Theseoscillations occurred about every 12 h in synchroniza-tion with the frequency of semidiurnal tides. Watertemperature was stratified vertically; in the upper layerthe mean temperature was 27.5°C, in the middle-layer25.6°C and in the lower layer 25.3°C. The mean salin-ity in the upper layer was 36.27, in the middle-layer36.36, and in the lower layer 36.35. Time series of cur-rents indicated a similar speed and direction in bothcomponents (Fig. 4). The cross-shelf component had asemidiurnal periodicity, with a strong eastward flowoccurring in the middle of the night and another in themiddle of the day. A northward flow in the along-shorecurrent occurred approximately every 12 h.Time-depth series of density, current shear, andRichardson numbers (Ri ) were constructed at 1 m depthintervals for the MQ station (Figs. 5 & 6). Time series ofdensity indicated a marked stratification in the upper10m of depth with periodical oscillations (Fig. 5C).Current shear showed high values between 4 and 8 m inboth along-shore and cross-shelf components(Fig.5A,B). Average shear in the along-shore componentat 5 and 6m depth was slightly larger than in the cross-shelf component (Fig.6B). Ri numbers throughout theentire water column were low (0.1 to 1.8) indicating in-stability in the water column that enhances turbulentvertical mixing. The highest Ri numbers occurred at thepycnocline depth and the lowest occurred sporadicallyabove and below the pycnocline, in the wind-mixedlayer and the bottom tide-mixed layer respectively(Figs.5D & 6C). Interestingly, low Ri numbers at the shal-lowest depth occurred during daylight hours, whichcoincided with the largest concentration of pink shrimppostlarvae in the upper layer (see Fig. 10).Time series of average temperature and sea surfacetemperature at the DT (first 24 h) and MQ stations (last24 h) revealed that the temperature was about 2°Clower at the MQ station (Fig.7B). Similar resultsappeared at the temperature contours across the tran-sect where the 27°C isotherm, which was located atabout 20 m at the DT station, was near the surface atthe MQ station (Fig. 8). Prevailing atmospheric condi-tions before and during the sampling period were rela-tively calm with weak east-southeasterly winds(Fig.7C,D). The average cross-shore and along-shoreLVNN i id id id  U  =× ∑∑ 171
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!