Figure 12.3. A continuum of larval migration by crab species results in horizontal zonation of species in the plankton between the upper estuary and the edge of the continental shelf: Rhithropanopeus harrisii larvae occur in the upper estuary, Dyspanopeus sayi larvae occur in the mid to lower estuary, Libinia spp. larvae occur in the lower estuary to the plume front, Ovalipes ocellatus and Uca spp. larvae occur from the upper, mid, or lower estuary (depending on the species) to the inner shelf, Callinectes sapidus larvae occur from the lower estuary to the outer shelf, and Cancer spp. larvae occur on the shelf. Larval migration by fishes also results in a predictable spatial gradient with larvae either largely remaining in the estuary, remaining on the shelf, or migrating from the shelf to the estuary: Anchoa mitchelli and Gobiosoma bosc larvae primarily occur in the estuary, Pomato-mus saltatrix larvae occur on the shelf and enter the estuary as postlarvae, and nonestuarine-dependent species (Citharichthys arctifrons, Hippoglossina oblonga, Etropus microstomus, Peprilus triacanthus, Pri-onotus spp.) largely remain on the shelf.

at the same time, and different stages of the same species move in opposite directions at the same time. Therefore, currents alone cannot explain the observed horizontal distributions of larvae, and behavior must be regulating migrations between adult and larval nursery areas. The continuum of larval transport patterns was maintained by the behavioral exploitation of circulation patterns. This was determined by coupling repeated extensive horizontal surveys with intensive vertical surveys relative to tides, currents, winds, and stage of development. This approach showed a clear relationship betweeninterspecific differences inhorizontal and vertical distributions of larvae. The vertical distributions were shown to be largely under behavioral control by using a comparative hypothesis testing approach that contrasted passive eggs with active larvae.

Upper estuary. Larvae remained within the upper Hudson River estuary (between the George Washington and Verrazano bridges) by exploiting typical two-layer flow. Residual flow was seaward at the surface, landward along the bottom and zero at 4-5 m depth at the shallowest site near the George Washington Bridge and 8-10 m depth at the deepest site near the Verrazano Bridge. At the shallow site, larvae remaining above 5 m would be swept downstream, those near 5 m would slosh back and forth over the tidal cycle without any net transport, and those at the bottom wouldbe carried upstream. Larvae could use this two-layer flow as a conveyor belt to move up and down the estuary provided that they effectively regulated their vertical position. The problem is that the water column is very dynamic over the tidal cycle. During spring tides when tidal currents were strongest, the water column at the shallow site was thoroughly mixed by flood and ebb currents every six hours and was partially mixed at the deep site. Can larvae regulate their vertical positions under such dynamic conditions?

The ability of larvae to regulate depth was tested by comparing the vertical distributions of non-motile eggs and swimming larvae (Fig. 12.4). Eggs were passively mixed throughout the water column by strong ebb and flood currents every sixhours as expected. By contrast, larvae were not mixed like eggs, and instead, apparently regulated depth. For example, all four larval stages of both species of crabs remained near the depth of no net transport and stayed near adult populations (Fig. 12.5). All four larval stages of R. harrisii were present upstream and were rare downstream, clearly indicating that they did not travel far after hatching. Larvae of the second species of mud crab, D. sayi, occurred in higher salinity waters downstream, thereby replacing R. harrisii larvae. Thus, similar vertical distributions of larvae of the two crab species resulted in larval retention within the estuary.

Larvae of R. harrisii have been shown to remain in the upper reaches of two other estuaries by remaining near the level of no net motion (Bousfield, 1955, Cronin and Forward, 1986). Laboratory studies suggest that both R. harrisii

Figure 12.4. Upper panel: expected vertical distributions of eggs and larvae over the tidal cycle in deep and shallow portions of the estuary, provided that turbulent mixing overwhelms larval behavior. Lower panel: expected distributions of larvae if they regulate depth effectively. Larvae may remain at one depth or change depth during development, and they may periodically change depth relative to tidal or light-dark cycles.

Figure 12.4. Upper panel: expected vertical distributions of eggs and larvae over the tidal cycle in deep and shallow portions of the estuary, provided that turbulent mixing overwhelms larval behavior. Lower panel: expected distributions of larvae if they regulate depth effectively. Larvae may remain at one depth or change depth during development, and they may periodically change depth relative to tidal or light-dark cycles.

and D. sayi larvae may remain at their preferred depth by swimming upward until they reached light intensities and hydrostatic pressures sufficient to cause them to sink to a lower depth, and they may fine tune their vertical distributions over the tidal cycle by responding to changes in temperature and salinity.

Tidal vertical migrations also may facilitate retention of crab larvae in the upper estuary. They were first described for R. harrisii larvae in the laboratory, but they were not detected in the present study nor were they evident in a previous field study (Cronin and Forward, 1986). It is possible that the amplitude of vertical migration about the level of no net motion is too slight to be detected without finely sampling the vertical distributions of larvae over the tidal cycle. Another possibility is that larvae are overwhelmed by the strong tidal currents at these sites, because tidal vertical migrations by

D. sayi larvae have been detected in a quiet backwater of Long Island Sound where turbulent mixing may be much reduced (Hovel and Morgan, 1997). Crab postlarvae also undertake tidal vertical migrations (Christy and Morgan, 1998). Both crabs metamorphose to postlarvae after ten to fourteen days, sink lower in the water column, and undertake tidal vertical migrations back upstream to adult habitats (Fig. 12.5).

As for crab larvae, the vertical distributions of anchovy and goby larvae differed from eggs, suggesting that they regulated depth. However, they regulated depth differently than did crab larvae. These fish larvae appeared to undertake both tidal and diel vertical migrations at the downstream site. Tidal vertical migrations should lead to the upstream transport of larvae. This was the first time that tidal migrations have been observed for either species, although upstream transport has

Blue Crab Larvae
Figure 12.5. Larval migration by mud crabs (Rhithropanopeus harrisii, Dyspanopeus sayi) and blue crab (Callinectes sapidus) between the estuary and shelf, depicting changes in behaviors relative to physical transport processes.

been inferred from the horizontal distributions of eggs and larvae of both fishes before (Massman, Norcross, and Joseph, 1963; Schultz et al., 2000).

The vertical distributions of larvae of both fish species changed at the upstream site, and vertical migrations were no longer evident. However, larval distributions still differed from those of passively distributed eggs. Goby larvae occurred primarily in bottom waters throughout the tidal cycle, which would enable them to move slowlyupstream. These larvae school near the bottom before settling in shallow oyster beds (Breitburg, 1989),butitremains unclear whether the changes in vertical distributions of anchovy larvae represent a behavioral switch in response to changing environmental conditions or whether larvae were simply overcome by strong mixing from ebbing surface currents. Anchovy larvae occurred in the upper water column near slack tides and appeared to be mixed throughout the water column near mid-flood and mid-ebb tides. An extensive study of anchovy larvae that was conducted upstream from the present study showed that this pattern was maintained throughout the uppermost region of the estuary (Schultz et al., 2000). Thus, anchovy larvae may have moved upstream by undertaking tidal vertical migrations until they reached the shallow uppermost estuary, whereupon they may have been overcome, slowing upstream transport.

Tidal vertical migrations apparently were not essential for crab and fish larvae to be retained in the upper estuary, despite expediting upstream transport. Maximal ebb currents occurred at the surface, so that if larvae maintained their position at mid-depth, they would not be transported as far downstream as they would be transported upstream by a velocity jet that formed beneath the pycnocline duringflood tides. However, larvae were mixed throughout the water column by ebbing surface currents and may have spent some time near the surface in maximal ebb currents and some time near the bottom in minimal ebb currents. Therefore on average, larvae still would have been transported seaward more slowly during ebb tide than they would have been transported upstream by remaining near the velocity jet during flood tide.

In summary, larvae remained in the upper estuary by regulating depth. All larval stages were collected, the vertical distributions of larvae differed from nonmotile eggs over tidal cycles and were consistent with behaviors that would foster retention, and interspecific differences in vertical distributions of larvae were evident that would result in differential transport.

Estuarine plume. The first step in determining how larvae migrate between the Hudson River estuary and New York Bight was to map the location of the front that separates turbid, river water from clearer, shelfwater. The plume is a dynamic feature that moves back and forth with the tides, winds, and runoff, and therefore, it must be mapped continuously relative to larval distributions to determine how larvae are exchanged between es-tuarine and coastal waters. Surveying horizontal distributions across the plume front revealed an interspecific continuum of larval transport away from shore; larvae of some species occurred farther offshore than did others. At one end of the spectrum, R. harrisii larvae were absent, again indicating that these larvae were retained effectively in the upper estuary despite net seaward flow, and larvae of D. sayi remained almost exclusively (98 percent) in the mouth of the estuary. Larvae of spider and lady crabs mostly were released near the mouth of the estuary and moved farther offshore during development than did D. sayi larvae. The two larval stages of spider crab take only six to eight days to develop (Johns and Lang, 1977), enabling them to largely remain shoreward of the front while most postlarvae eventually were transported seaward of it. The lady crab has five larval stages, and late stage larvae of the lady crab were much more abundant seaward of the front than were late stage larvae of the spider crab. Few blue crab and fiddler crab larvae occurred in the estuary, and they became increasingly abundant farther offshore, indicating that they may have been transported into the study area from elsewhere.

The differential transport of larvae may have been facilitated by interspecific differences in the vertical distributions of larvae during development. Dyspanopeus sayi larvae were retained in the mouth of the estuary by remaining near the level of no net motion, as was evident in the upper estuary. Spider crab larvae were transported toward the plume front by occurring throughout the water column, and lady crab larvae were transported out to sea by primarily occurring in outwelling surface currents. Like lady crab larvae, blue crab larvae mostly occurred near the surface but apparently were carried in the opposite direction toward shore. Therefore, mean depth distributions alone may not entirely account for differential offshore-onshore transport.

Ontogenetic vertical migrations may have facilitated differential onshore-offshore transport. Spider and lady crab larvae both appeared to undertake classic ontogenetic vertical migrations, wherein early stage larvae occurred primarily in the neuston and late stage larvae or postlarvae occurred deeper in the water column. Therefore, early stage larvae of these species all may have been transportedseawardbyoutwellingsurface currents early in development, while seaward transport may have slowed or reversed late in development. In contrast, blue crab larvae appeared to undertake a reverse ontogenetic vertical migration (Fig. 12.5). By becoming increasingly more abundant in the neuston during development, blue crab larvae and especially postlarvae may have been transported into our study area from elsewhere.

The plume front was a persistent feature throughout the duration of sampling, but there was little evidence that larvae aggregated there, even though tremendous numbers of ctenophores and salps did. Gelatinous zooplanktons may have physically displaced larvae or they may have greatly reduced larval densities by predation. However, it may be even more likely that larvae did not aggregate at the front because most of them occurred beneath converging surface currents.

New York Bight. Of the four taxa that hatched in the estuary and were common in the estuarine plume, only lady and blue crab larvae were abundant on the open shelf. Fiddler crab larvae also were abundant on the shelf even though they were uncommon in the estuary and plume. First stage larvae of these three taxa were most abundant nearshore, and later larval stages were increasingly more abundant farther from shore during development. However, the distance that larvae traveled from shore differed among taxa. Fiddler crab larvae remained closest to shore and were most prevalent at stations 1 and 2 throughout development while rarely occurring beyond station 3 even late in development. Lady crab larvae also were most common at station 1 throughout the larval period but occurred in low concentrations as far offshore as station 4 beginning midway during development. Unlike larvae of these taxa, blue crab larvae were at least as common at stations 2 and 3 as they were at station 1, were common as far from shore as station 4 late in development, and even were collected at station 5 in low concentrations. Progressive seaward transport of larvae during development may have been due to Ekman transport by prevailing northeasterly winds, because larvae of all three taxa frequented surface waters, especially at night during diel vertical migrations.

The distance that larvae were transported offshore may have depended more on the location of larval release and larval development time than on the time that larvae spent in surface waters. Fiddler crab larvae were not transported as far offshore as were lady crab larvae, even though most of them occurred in the neuston rather than in near surface waters (1 to 10 m deep) and the two taxa have the same number of larval stages (5) and similar development times (Williams, 1984). Rather, lady crab larvae had a head start; they hatched near the mouth of the estuary, whereas fiddler crab larvae hatched throughout the lower estuary (Williams, 1984). Blue crab larvae were transported farthest offshore, because they are releasednear the mouths of estuaries, have three more larval stages, and were most abundant in the neuston of the three taxa (Fig. 12.5).

Spending time near the surface may be imperative for larvae to remain near the parental estuary. When these crabs develop, prevailing subsurface currents in New York Bight flowed to the southwest, and winds blew in the opposite direction to the northeast generating a countercurrent mid-shelf. Therefore, larvae in the neuston may be transported northward against the prevailing subsurface flow. Northward transport clearly occurred for the surface-dwelling larvae of blue and fiddler crabs. Few first stage larvae of either species were collected in the Hudson River estuary, the estu-arine plume or farther offshore, and later larval stages generally increased during development in New York Bight. Therefore, larvae likely were released elsewhere and transported into our study area. Larvae likely originated from southern populations, because they generally became increasingly more abundant all along the southern transect during development. The time required to travel northward from southern populations likely explains why late stage were more prevalent than were early stage larvae along the southern transect. By the time long-lived blue crab larvae metamorphosed to postlarvae, they again abounded along the northern transect, indicating that they traveled farther north than did fiddler crab larvae.

The proportion of lady crab larvae along the southern transect also increased during development, which again is consistent with northward transport. However, fewer of these slightly deeper dwelling larvae may have been transported northward than were larvae of the other taxa, because a lower proportion of them occurred along the southern than the northern transect. Furthermore, first stage larvae were abundant and larval density progressively declined during development, which is consistent with cumulative larval mortality without the extensive immigration shown by the other taxa. Thus, lady crab larvae largely appeared to be retained in New York Bight with some immigration from southern populations, and most blue and fiddler crab larvae in New York Bight likely originated to the south. Further evidence ofnorthward transport was provided by the appearance of expatriate larvae of southern species that do not occur in New York Bight as adults, such as the box crab Calappaflammea.

Now that multiple studies ofverticalmigrationby a single species are beginning to accumulate, it is becoming increasingly clear that the vertical migration behavior of larvae changes in space and time. Although lady crab larvae consistently appeared to undertake a classic ontogenetic vertical migration in the estuarine plume, in New York Bight and elsewhere in the Middle Atlantic Bight (Epifanio, 1988), blue crab larvae undertook a reverse ontogenetic vertical migration in the plume and a classic ontogenetic vertical migration farther offshore, elsewhere in the Middle Atlantic Bight (Epifanio, Valenti, and Pembroke, 1984; Epifanio, Little, and Rowe, 1988) and in the laboratory (Sulkin et al., 1980). Furthermore, fiddler crab larvae appeared to undertake a reverse vertical migration during our 48-hour study and in the laboratory (Anastasia, 1999), but they undertook classic ontogenetic vertical migrations elsewhere in the Middle Atlantic Bight (Epifanio et al., 1988). Moreover, a reverse tidal vertical migration was evident for lady crab larvae in New York Bight, but classic vertical migrations were apparent in the Hudson River plume. Finally, rock crab larvae appeared to undertake both classic andreverse ontogenetic and tidal vertical migrations, depending when and where larvae were sampled between the mouth of the Hudson River estuary to the edge of the shelf. This variation probably is not random noise but is under rhythmic behavioral control. This has been demonstrated by contrasting the timing of vertical swimming by blue crab postlarvae from onshore and offshore waters in the laboratory (Forward and Rittschof, 1994) and by fiddler crab larvae that were collected from different tidal regimes and observed under constant conditions in the laboratory (Anastasia, 1999). Plasticity in vertical migration behavior of larvae in response to changing oceanographic conditions may play an important role in facilitating cross-shelf migrations between adult habitats and larval nursery areas, and we need to increase our understanding of the underlying mechanisms generating plasticity in vertical migration.

We are just beginning to understand how larvae recruit back onshore, but they may have been transported shoreward by internal waves and directed onshore swimming (Fig. 12.5). Tidally generated internal waves propagate across the shelf transporting postlarvae that accumulate in surface convergences above the wave with them (Shanks, 1995). Onshore swimming may facilitate onshore migrations, because these postlarvae are able swimmers; blue crab larvae can swim about 10 cm/s (LuckenbachandOrth, 1992) andbluefishlarvae swim evenfaster at 2-3 bodylengths/s (Hunter, 1981). Recent evidence suggests that postlarvae may use chemical, auditory, and other cues to swim toward shore and locate estuaries (Kingsford et al., 2002). Crab and bluefish postlarvae remained at the surface after they crossed the estuarine plume front, but once inside estuaries, these chemical cues may initiate a change in behavior (Forward and Rittschof, 1994). All crab postlarvae investigated thus far undertake selective tidal migrations (Fig. 12.5) that transport them up estuary to adult habitats (Christy and Morgan, 1998).

Behavior ensures a steady supply of recruits, even though postlarvae of some species begin their journey from as far away as the edge of the continental shelf. Many investigators point to occasional large recruitment events as evidence that physical forcing rather than behavior controls larval supply. For example, downwelling winds episodically force blue crab postlarvae from nearshore coastal waters into the mouths of estuaries resulting in large recruitment events (Epifanio and Garvine, 2001). However in the absence of these wind events, they recruit regularly to estuaries. Winds are light throughout the peak recruitment season of blue

Table 12.1. Rank orders of Stage I larvae captured in plankton tows and postlarvae captured in the plankton or on settlement collectors in Flax Pond, New York

Flax Pond is a shallow embayment on the south shore of Long Island Sound. Larval production matched larval recruitment, regardless of whether larvae typically develop on the continental shelf or entirely within estuaries, except for Ovalipes ocellatus that did not recruit to the salt marsh. Larvae of two other crabs were present but were not included here because postlarvae were not reliably sampled (pinnotherids) or larvae were rare and the dispersal pattern is not well known (Hemigrapsus sanguineus).

Species Larval habitat Larval production Larval recruitment

Uca spp. Shelf 1 1

Dyspanopeus sayi Estuary 2 2

Ovalipes ocellatus Shelf 3 6

Sesarma reticulatum Estuary 4 3

Libinia spp. Shelf 5 4

Panopeus herbstii Estuary 6 5

Adapted from Hovel and Morgan, 1997.

crabs along the north central Gulf Coast, and yet blue crabs recruited to Mobile Bay, Alabama, every two weeks during minimum amplitude tides throughout the six-month study period for two consecutive years (Morgan et al., 1996). Therefore, blue crab larvae may cross the continental shelf and enter estuaries by exploiting reliable physical transport mechanisms.

Cross-shelf mechanisms may be as reliable as much better understood estuarine mechanisms. This surprising conclusion was reached by measuringlarvalproductionandrecruitmentofes-tuarine crabs to determine whether species that developed on the shelf recruited to a salt marsh in Long Island Sound as reliably as did those that developed entirely within estuaries (Hovel and Morgan, 1997). Indeed they did; postlarvae recruited back proportionally to the numbers oflar-vae that were released regardless of whether larvae were retained or exported (Table 12.1). The one exception was for lady crab that may not recruit to salt marshes. The same result was obtained in both other estuaries (North Inlet, South Carolina; Mobile Bay, Alabama) where this comparison was made (Christy and Morgan, 1998; Morgan unpublished data). Thus, recruitment appears to be reliable whether larvae leave the estuary or not along the Atlantic and Gulf coasts, thereby calling into question the widely held belief that larvae that migrate far away on the continental shelf and that develop for a long time in the plankton recruit less reliably than do species that do not migrate far and spend little time in the plankton. This also suggests that recruitment probably may not be decoupled from local production, as is commonly thought.

0 0

Post a comment