Physical Oceanography at Dalhousie University

Click here to load reader

  • date post

  • Category


  • view

  • download


Embed Size (px)

Transcript of Physical Oceanography at Dalhousie University

Larval Flatfish Distributions and Drift on the Southern Grand Bank
Kenneth T. Frank and John W. Loder Department of Fisheries and Oceans, Bedford /nstltute of ()ceanography, P.O. Box 1006, Dartmouth, N.5. 82Y 4A2, Canada
James E. Carscadden Department of Fisher ies and Oceans, Northwest At lant ic Flsher ies Centre, P.O. Box 5667, St . John's, Nf ld. AIC 5Xl , Canada
Wil l iam C. Leggett Department of Biology, McCil l University, 1205 Ave. Dr. Penfield, Montreal, Que., H3A 1Bl , Canada
and Christopher T. Taggart Department of Fisher ies and Oceans, Northwest At lant ic Fisher ies Centre, P.O. Box 5667, St . lohn's, Nf Id. AIC 5Xl , Canada
Frank, K. T. , J . W. Loder, J . E. Carscadden, W. C. Legget t , and C. T. Taggart . 1992. Larval f la t f ish d is t r ibut ions and dr i f t on the southern Crand Bank. Can. J . F ish. Aquat . Sci . 49: 467-483.
lchthyoplankton and hydrographic surveys of the southern Crand Bank in September of 1986, 1987, and 19BB revealed substantial correspondence between the areal distributions of larvae of three flatfish species and tem- perature below the thermocline. Depth-averaged densities of American plaice (Hlppoglossoides p/atessoides) were negatively correlated with temperature whereas yellowtail f lounder (Limanda ferruginea\ and witch flounder (Clyptocephalus cynoglossus) densities were positively correlated with temperature. In spite of large interannual differences in abundance, the larval distributions showed similar structure from year to year. Using estimates of larval age inferred from length frequency distributions and literature values for growth rate, in conjunction with moored current measurements, estimates of spawning times and locations were obtained for each species. These estimates were compared with historical information on the distribution of prespawning fish for each species to examine the hypothesis of passive larval drift. The results indicate that in most, but not all cases, the larval distributions and currents are consistent with passive larval drift for particular growth rates and vertical distri- butions. However, the observations are not adequate to rule out alternative mechanisms involving behaviour.
Des relev6s de l ' ichtyoplancton et des paramdtres hydrographiques effectu6s dans le secteur sud des Crands Bancs en septembre 1986,1987 eI 19BB ont r6v6l6 une correspondance importante entre la r6partit ion spatiale des larves de trois espdces de poissons plats et la temp6rature de l 'eau sous la thermocline. Les densit6s, en fonction de la moyenne des profondeurs, des larves de plie canadienne (Hippoglossoldes p/atessoides) 6taient en corr6lation n6gative avec la temp6rature, tandis que les densit6s de larves de l imande ) queue jaune (l lrnanda ferruginea) et de plie grise (C/yptocephalus cynoglossus) 6taient en corr6lation positive avec cette variable. Mal916 d'importantes diff6rences inter-annuelles de l 'abondance, la r6partit ion des larves pr6sentaient une structure semblable d 'une ann6e A l 'autre. A l 'a ide d 'est imat ions de l 'Age des larves 6tabl ies ) par t i r des d is t r ibut ions de la fr6quence des longueurs, de taux de croissance tir6s d'ouvrages publi6s et de quantif ications courantom€- t r iques, les auteurs ont obtenu des est imat ions du moment et du l ieu de la f ra ie de chaque espdce. l ls ont ensui te compar6 ces estimations ) des donn6es historiques sur la r6partit ion de chaque espdce avant la fraie pour tenter d'6tayer l 'hypothdse d'une d6rive passive des larves. Les r6sultats r6vdlent que dans la plupart des cas, mais pas tous, la r6partit ion des larves et les courants appuient cette hypothdse dans le cas de certains taux de croissance et de r6partit ion verticale. Toutefois, les observations ne suffisent pas i 6carter d'autres m6canismes li6s au comportement.
Received December 21, 1990 Accepted October 10, 1991 (JA84s)
common approach in investigations aimed at under- standing recruitment to marine fish populations and its variability is a search for and quantification of the factors
R e e t r l e t 1 d c e c m h r c l 9 9 Q
Accept1 le 10 octobre 1991
that determine the distribution and survival of the early life stages. Hypotheses include those in which purely environmen- tal factors such as water drift (or lack thereof) dominate (e.g. Myers and Drinkwater 1988) and those requiring significant behavioural components such as vertical migration in relation to currents (e.g. Sinclair 1988). The possibility of retention, or an extended residence time in a particular area, has been sug- gested for many species and has been attributed both to purely
Can. J. F ish. Aquat. Sci . , Vol .49, 1992
physical processes (e.g. American plaice (Hippoglossoides pLa- tessoides), Nevinsky and Serebryakov 1973; silver hake (Mer- luccius bilinearis) , Fahay 197 4) and vertical migration in rela- tion to depth-varying currents (e.g. Atlantrc herring (Clupea harengus harengus), Sinclair and Iles 1988).
Two physical factors that affect the degree to which eggs and larvae are retained within an area are its horizontal extent and the rate of horizontal water exchange. Together, these factors determine the water residence time (extent/exchange rate). The Grand Bank of Newfoundland, a continental shelf region used as a spawning ground by numerous species, has both a broad
plateau and weak subtidal currents (except in the Labrador Cur- rent around its perimeter) (e. g. Petrie and Anderson I 983; Loder et al. 1988). Therefore, retention of early l i fe stages might be expected to be a common feature on the Bank.
In this paper we report the distributions of the early life stages of three flatfish species on the southern Grand Bank, as observed in ichthyoplankton surveys conducted in September of 1986, 1987, and 1988. The species examined are American plaice, yellowtail flounder (Limanda ferruginea), and witch flounder (Glltptocephalus cynoglossus). There is limited infor- mation on the distribution of larvae and juveniles of these spe- cies in the area. However, extensive sampling of older juveniles (ages I and older), conducted since 1985, indicates that the southern Bank may be a juvenile nursery area for yellowtail
and plaice (Walsh l99l). We describe the distributions of lar- vae and 0-group juveniles in relation to the temperature field and use current measurements derived fiom moored instru- ments to estimate their movement and probable spawning sites. assuming that the dominant influence is water drif i. We use the results of these analyses to assess the hypothesis that passive drift from spawning areas can account for the observed larval distributions.
Materials and Methods
area (Fig. I ) was the Southeast Shoal centre 4,1o30'N. 50"W) of the Grand Bank, its
8 3 1 C 887 0
T A I L 43"
52 5 1 " Frc. l. Location map for the southern Grand Bank with the Table I for t imes and depths). l , 1986; I , 1987; O, 1988.
D 832
\ , , ,
Can. J. F ish. Aquat. Sci . , Vol .19, 1992
Teelr l . Summary of posit ions, depths (below surface), and periods with reliable data for current meter measurements on the southern Grand Bank in 1986-88. Mooring No. refers to the Bedford Institute "f O..^""gr"phy Ct
Water Data Mooring Latitude Longitude depth lnstrument return period
No. (N) (W) (m) depth (m) (calendar days)
l 986 714 44" t4.4',
1981 830 44"16.O' ,
I 988 887 44'15. l ' , 50"59.9 ' ,
shallowest area. The Shoal is or has been the site of major domestic and international f isheries for capelin, cod, and flatfishes. [t is a major spawning area for capelin and has been the site of studies on spawning (Carscadden et al. 1989) and early l i f 'e stages (Frank and Carscadden 1989) of capelin and on juvenile stages of f latf ish (Walsh l99l). The principal features of its spring-fall physical oceanographic regime are an approximate two-layer stratif ication (e.g. Loder l99l), weak subtidal currents, and a branch of the Labrador Current along the shelf break to the east (e.g. Petrie and Anderson 1983).
Physical Oceanography
The current and hydrographic fields in the Southeast Shoal region were monitored in 1986, 1987, and 1988 using moored sensors deployed from spring to fall (Ross et al. 1988) and hydrographic surveys. The returned data include time series from two Aanderaa current meters at each of one site in 1986, three sites in 1987, and one in 1988 (see Fig. I and Table l, and Ross et al. 1988 for details). Temperature and salinity dis- tributions were obtained during the l986-88 September larval surveys from a CTD mounted on the BIONESS net system (Sameoto et al. 1980) and from XBT casts. Temperature and salinity data for other times during spring-fall of 1986-88 were obtained fiom dedicated hydrographic (CTD and XBT) surveys and from the Marine Environmental Data Service (MEDS). The MEDS data were screened for suspect values which were removed, and all the hydrographic data were subsequently grouped into time intervals of 1-2 wk duration which provided quasi-synoptic coverage of the region.
Spawning has never been directly observed for any of the three species studied. Published historical accounts of adult dis- tributions, the approximate l imits of the distribution of pre- spawners, and the relationship between these prespawning dis- tributions and bottom temperatures were therefore used to a delimit the expected spawning times and locations.
Larvae and Juveniles (0-Group)
The sampling methods used are detailed in Frank and Carscadden (1989). Briefly, the horizontal and vertical distri-
Can. J. F ish. Aquat. St ' i . . Vol .49. 1992
butions of larvae of the three flatfish species were assessed tiom surveys conducted in September of igSO. 1987, and 1988. In 1986, a rectangular grid of 39 stations spaced at l0 nautical mile intervals along six transects was sampled. Eight additional stations were sampled along 44'15' centered on the shelf break. In 1987 the grid was expanded to the north, west, and south yielding a total of 7 I stations. In 1988. anorher grid of 7 I sta- tions was sampled with slightly different spatial coverage. We used a small-scale version (0.25 m'z) of the BIONESS, a mul- tiple opening and closing net. f itted with seven 333-pm-mesh nets. Seven discrete depths were sampled at each deployment and sample depths were spaced at 5-m intervals starting at 5 m (below surface). Each depth stratum was sampled for l0 min during which approximately 200 mr of water was fi l tered. Sen- sors positioned on the net frame provided real-time on-deck readouts of pitch, roll, and flow rate, as well as temperature, salinity, and depth (CTD). During the middle of each BIO- NESS tow, a horizontal tow of l0 min duration was made at I m depth with a 333-pm, 0.75-m diameter net.
All samples were preserved in 4Vc formalin buff'ered with sodium borate. All fish larvae and macrozooplankton were sorted from the samples and counted. Total length measure- ments were made using an ocular micrometer fbr larvae <15 mm and a mi l l imetre scale ru ler for larger larvae (> l 5 mmt and j uven i l es .
Variations in the depth distributions of larvae and juveniles of the three species were evaluated by calculating the verlical centre of mass of their distributions, 2.,,,:
( l ) z , n : 4 , r , 2 , ) , j - I , m
yhgre p, is the proportion of larvae in the jth depth interval, z, is its average depth, and nr is the number of intervals. The centroids of the horizontal distribution of verticallv averased larval density were computed from
) Not^ / ' ) \ v - L \ L t ^ e n , -
) N .
5 J A : l ' r t
whereN. is the larval density,,r, and,\,* are the horizontal coor- dinates of the ftth station. and n is the number of stations.
Infened Spawning Areas
For an egg or larva released at horizontal position X, at t ime 1, . i ts posi t ion at some later t ime l , is
t,, (3) X, : X, +
) ,, U,,(X,, t7 dt
where U.(X,, r) is its Lagrangian velocity (relative to the sea- floor) at t ime t. This velocity includes both the motion of the egg/larva relative to the water and the space- and time-depend- ent water velocity. To evaluate the influence of physical advec- tion on larval distributions, we assumed that the individual lar- vae move with the ambient water and approximated this velocity by the velocily u,,,(x^, l) observed at a current meter position r,,. Taking r, to be the midtime of the September survey, the mean release time can be approximated by
(4) t; : tr - (1, - l,)lg
l l 44 1 4 41 I L
tzt 262 t2t-262
where /, is the larval length from the September survey, /, the length at release, and g its average growth rate. Thus, the release position can be estimated from
f r r
u,,,(x,,,, t) dt
with X,, /,, and /, specified from the September survey data, a.(r) fiom the m<iored current measurements, and /, and g from literature values. The use ofcurrent observations from a sinsle position assumes that there is limited horizontal structure in ihe subtidal currents which dominate long term drift on the South- east Shoal. Cross-spectral analysis of the 1987 moored obser- vations indicates that the spatial scale of the subtidal currents over the Shoal is of the order 50- 100 km, although around the Shoal's perimeter, there is substantial structure aJsocrated with the Labrador Current and the steep bathymetry. Hence, the accuracy of X, estimated from (5) will generally decrease for earlier release times (older larvae) if positions are considerable distances from the mooring site.
In the analysis presented here, the length at release was taken as the hatching length and the growth rate chosen to be appro- priate to post-hatching. Strictly speaking, the infened release times and positions are therefore appropriate to hatching. How- ever, because the egg phase is short relative to the larval drift time, they also approximate the spawning times and positions.
Results Physical Oceanography
The temperature distributions observed along 44'15'N during the September survey in each year (Fig. 2) indicate a
pronounced vertical stratification with a sharp thermocline near 20 m, consistent with historical observations (Loder 1991). Some interannual variability is apparent, including cooler temperatures and a less pronounced thermocline in 1986, and wanner near-surface waters and a slightly deeper thermocline in 1988. The vertical sections also show elevated near-bottom temperatures in the vicinity of the Shoal in each year, again consistent with historical observations (Loder 1991), and subzero temperatures along the Bank's eastern edge consistent with the presence of the cold intermediate layer associated with the Labrador Current (Petrie et al. 1988). Areal distributions of bottom temperature (Fig. 3) during early spring, early summer, and late summer (from the MEDS and hydrographic survey data) confirm the persistence of these features during spring and summer of each year, although with some variability in spatial structure and temperature from year to year (also see Carscadden et a l . 1989).
The current measurements derived from the moored instruments revealed that, although the subtidal currents are important to long-term drift on the southern Grand Bank (Loder et al. 1988), they are generally weak. The mean currents have magnituCes of up to 0.04 m/s, while the standard deviations of the low-frequency current components are typically 0.02- 0.05 m/s (Ross et al. 1988). The expected particle drifts resulting from the measured currents are illustrated in Fig. 4 which presents progressive vector plots for the centraf and eastem sites (Fig. l) in 1987. Typical of the central and western sites in the other years, the central-site drift is generally westward and significantly stronger at the upper meter. In
20 E. 5 a o o o
O 5 0 1 O O 1 5 O 2 O O k m
FIc. 2. West-east vertical sections of temperature along 44"15'N from the CTD mounted on the BIO- NESS during the September larval surveys of (a) 1986, (b) 1987, and (c) 1988.
Can. J. Fish. Aquat. Sci., Vol. 49, 1992
2 7 J U X E . 5 J U L Y t$a
2 6 J U X E . 5 J U L Y 1 3E7
I O J U X E . 3 J U L Y 1 S
5 2 W9 W
4 2 N
" r ' l - \ .e) l'VA
5 2 W
2 I A P A . . E I A Y 2 I A U G . . ' S E P I
4 ? N
5 2 W $ w
Ftc. 3. Bottom temperature distributions on the southern Grand Bank during early spring, early summer, and late summer in 1986, 1987, and 1988. Data are from dedicated hydrographic surveys and MEDS and have been sorted into time intervals with broad spatial coverage.
i 7 - 2 t s E P T .
Can. J. F ish. Aquat. Sci . , Vol .49, 1992 4 l l
?- 500 B
F o
K I LOME-TRES -100.0 -50.0 0.0
KILOMETRES 50.o t@.0 1n00 e000-m.0 -150_0
9? -rm.o
-330.0 -2d), -r5o.0 -t*o -ill.roJ.h"#.0 rml) 150.0 2d).0 25{ro -?5().0 -m.0 -t".. -tffirofr"rho# soo rm.o r5o.o e{oo
Ftc. 4. Progressive vector plots f iom current meters at the (a, c) central (mooring 830) and (b, d ) eastern (mooring t l32) si tcs in 1987. Crgsses denote 5-d intervals, labelled every 20 or 40 d.
contrast, the eastern-site drift is generally southward. These drifts are consistent with barotropic numerical model predictions (Greenberg and Petrie 1988; Hukuda et al. 1989) of a strong southward Labrador Current along the Bank's eastern edge and a weak west to northwestward drift across the Bank. The mean currents (Fig. 5) confirm the interannual persistence and large space scales of the characteristic drifts, but a lso indicate quant i ta t ive var iat ions.
The moored measurements at the central site also revealed a large decrease in temperature over much of the water column in August 1986. This was associated with both vertical mixing and the displacement of Labrador Current water onto the Shoal during the passage of a tropical storm (Loder and Ross 1988; Frank and Carscadden 1989) and contributed to the relatively cool 1986 September temperatures (Fig. 2).
Commercial abundances of American plaice, yellowtail flounder, and witch flounder are common on the Grand Bank. As a result, their seasonal distributions and those of prespawn- ers are documented. Of the three, yellowtail inhabits the shal-
lowest and warmest portions of the Bank. In fact. in this area, yellowtail is at the northern l imit of its commercial range. As a result, temperature appears to be an important regulator of its distribution. In the early 1960's yellowtail abundance on the Bank increased coincidenr with a decrease in haddock abun- dance and an upward trend in bottom temperature (Pitt 1970). During the 1950's and 1960's. the greatest concentrations of yellowtail occurred at depths of 57-64 m, coincident with bot- tom temperatures of 3.1-4.8"C (Pitt 1970). Research trawl sur- veys conducted during the 1970's and 1980's revealed the larg- est concentration of yellowtail to be at depths less than 100 m where water temperatures exceeded I'C lWells et al. 1988). Fitzpatrick and Miller ( 1979) used research trawl data to delimit the distribution (Fig. 6) of ripe fish berween 1952 and 1968. During this interval, peak spawning occurred during the latter half of June in…