Gulf of Mexico Science Volume 19 Article 3 Number 1 Number 1

2001 Abundance, Spatial Distribution, and Mortality of Young-of-the-Year Spotted Seatrout ( nebulosus) Along the Gulf Coast of Gary A. Nelson Florida and Wildlife Conservation Commission

Deborah Leffler Florida Fish and Wildlife Conservation Commission

DOI: 10.18785/goms.1901.03 Follow this and additional works at: https://aquila.usm.edu/goms

Recommended Citation Nelson, G. A. and D. Leffler. 2001. Abundance, Spatial Distribution, and Mortality of Young-of-the-Year Spotted Seatrout (Cynoscion nebulosus) Along the Gulf Coast of Florida. Science 19 (1). Retrieved from https://aquila.usm.edu/goms/vol19/iss1/3

This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf of Mexico Science by an authorized editor of The Aquila Digital Community. For more information, please contact [email protected]. Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye

Gulf of Mexico Sdeno•, 2001(1), pp. 30-42

Abundance, Spatial Distribution, and Mortality of Young-of-the-Year Spotted Seatrout ( Cynoscion nebulosus) Along the Gulf Coast of Florida

GARY A. NELSON AND DEBORAH LEFFLER

We used fixed-station and random-station sampling data from the period 1989- 97 to examine spatial and temporal patterns in the abundance and size structure of young-of-the-year (YOY) spotted seatrout, Cynoscion nebulosus, in three Florida estuaries. YOY seatrout first appeared at shallow-water (<1.5 m) seine sites in May-June in Choctawhatchee Bay () and in April-May in and Charlotte Harbor (both along the southwest Florida peninsula). Spotted seatrout were caught at deepwater (> 1.6 m) trawl stations within 1-3 mo of their initial appearance at shallow-water sites. Most spotted seatrout were caught in waters <3.7 m. Spring and summer peal's in YOY abundance, corresponding to strong influxes of newly spawned individuals, were observed only in southwest peninsula estuaries. Depending on the estuary, the occurrence of YOY spotted seatrout at shallow-water sites was associated with some combination of seagrasses, mangroves, salinity, depth, temperature, and mud. Estimates of total instanta­ neous mortality rates for YOY spotted seatrout in Tampa Bay were 0.027·d-1 for fixed sites and to 0.025·d-1 for randomly selected sites.

"'\}(lung-of-the-year (YO\') spotted seatrout, Cy- timate mortality rates of YOY spotted seatrout 1_ noscion nebulosus, play important ecologi­ along the gulf coast of Florida, USA. cal roles in estuarine and nearshore waters of Florida. They are prey for fish and birds (Carr and Adams, 1973; Johnson and Seaman, 1986) METHODS and prey upon a range of invertebrates and YOY spotted seatrout [$100 mm standard fish (Carr and Adams, 1973; McMichael and length (SL)] were studied in Choctawhatchee Peters, 1989), often to a degree that the abun­ Bay and (surface area: ca. dance of their prey may be affected (Johnson, 450 km2), a temperate estuary located in the 1982). western Florida Panhandle, and in Tampa Bay Despite the ecological importance of YOY (ca. 886 km2) and Charlotte Harbor (ca. 575 spotted seatrout, their population dynamics in km2), more subtropical estuaries located on western Florida estuaries have not been ade­ the gulf side of the Florida peninsula (Fig. 1). quately exan1ined. The published literature in­ All three systems are characterized by average cludes only information on estimates of age, depths of <5 m, salinities of 0-36 ppt, fresh­ growth, and spawning dates and descriptions water inflow from rivers, and expanses of bot­ of food habits and habitat associations (Spring­ tom vegetation, primarily seagrasses (Halodu.le er and Woodburn, 1960; Carr and Adams, wrightii and Thalassia testudinu.m.), in shallow ar­ 1973; McMichael and Peters, 1989). \•Vhether eas. seasonal changes in the abundance and size Spotted seatrout were sampled monthly structure of YOY spotted seatrout occur from 1989 to 1995 at shallow-water (<1.5 m) throughout entire estuaries or what factors in­ fixed seine and deepwater (> 1.6 m) fixed fluence their spatial distribution is unknown trawl sites. Monthly sampling began in 1993 in because past studies have had limited spatial Choctawhatchee Bay, in 1989 in Tampa Bay, coverages (sites were sampled in waters <1.5 and in 1991 in Charlotte Harbor. Fixed sites m) and short sampling durations ( <2 yr) were approximately evenly distributed (Springer and I•Voodburn, 1960; Carr and Ad­ throughout shallow- and deepwater areas. Fish ams, 1973; McMichael and Peters, 1989). In ad­ were collected in a 21.3-m X 1.8-m, 3.2-mm dition, natural mortality rates of YOY have not stretched-mesh seine or a 6.1-m, 38-mm been estimated. stretched-mesh otter trawl containing a 3.2-mm In this study, we used 4-9 yr of data to doc­ stretched-mesh codend liner. At beach sites, unlent seasonal changes in abundance and size seines were set acUacent to the shoreline and structure in shallow- and deepwater areas, to hauled onshore; at offshore sites, seines were identif}' factors that are associated with YOY set in open-water habitats away from the shore­ spotted seatrout spatial occurrences, and to es- line and retrieved offshore. Trawls were towed

© 2001 by the j\{arine Environmental Sciences Consortium of Alabama Published by The Aquila Digital Community, 2001 1 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3 NELSON AND LEFFLER-YOUNG-OF-THE-YEAR SEATROUT DYNAMICS 31

Tampa Bay

Charlotte Harbor Gulf of Mexico

..

Fig. l. Map of Florida showing the locations of Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor.

at 1 knot for 10 min. Three seine hauls or trawl seine sites because visual assessment of bottom tows were made at each fixed site during day­ characteristics at trawl sites was hindered by light hours. Sampling occurred during the first depth. 2 wk of each ITlonth. Spotted seatrout were also sampled monthly Seasonal changes in YOY relative abundance and at randomly selected sites so more accurate es­ size structure.-To examine seasonal changes in timates ofYOY relative abundance could be de­ YOY relative abundance in the shallow- and termined and the spatial distribution of YOY deepwater areas, comparable monthly catch spotted seatrout in each bay could be de­ rates (mean number of individuals per 100m2) scribed. To coordinate sampling logistics, each were calculated from fixed-site and random­ bay was subdivided into four or five arbitrarily site sampling data by year. We separated YOY lettered zones. All zones encompassed about data used in all analyses from data on older equal areas. Within each zone, 1" latitude X 1" individuals by using maximum size limits se­ longitude microgrids, representing the sites to lected from monthly length-frequency plots. be sampled, were randomly selected within Monthly length frequencies were constructed randomly selected 1' latitude X 1' longitude by calculating the proportions of fish found in grids. At each site, one haul or tow was made each length class in each year, combining year­ with the use of the same gears and deployment ly distributions, and standardizing the cumu~ techniques as those used at fixed stations. lative proportions to 1. Length maximum size Monthly random sampling was conducted in limits used were in general agreement with Choctawhatchee Bay in 1996 and in Tampa those determined from analyses of the otoliths Bay and Charlotte Harbor in 1996 and 1997. of YOY spotted seatrout from Tampa Bay For all collections, total numbers of spotted (McMichael and Peters, 1989). In addition, seatrout were counted, standard lengths for up summary statistics (mean, minimum, and max­ to 100 randomly selected individuals per sam­ imum) of monthly length data were calculated ple were measured ( ± 1 mm), and all fish were and used to identify seasonal changes in size released. Salinity (ppt), temperature (C), dis­ structure at shallow-water seine sites. solved oxygen (ppm) levels, and depth (m) were also recorded at all sites. Dominant bot­ Depth distribution.-To determine whether YOY tom and shoreline vegetation and sediment were restricted to particular depth ranges dur­ (mud or sand) types were recorded only at ing the period surrounding peak abundance, https://aquila.usm.edu/goms/vol19/iss1/3 2 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 32 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

the cumulative frequency distributions of trawl Factors influencing YOY spatial occu.rrence.-We depths and depths at which YOY spotted sea­ examined variation in YOY seatrout occurrenc­ trout occurred were compared by using the es in shallow-water areas by using multiple lo­ Kolmogorov-Smirnov two-sample test (Siegel, gistic regression (Agresti, 1996) to model the 1956; Perry and Smith, 1 994). The cumulative presence/absence of spotted seatrout in ran­ frequency distributions for YOY spotted sea­ dom seine catches as a response surface to trout were constructed by weighting depth at year, deployment technique, sediment types, each random site by the number of YOY ­ bottom vegetation, depth, temperature, dis­ ted seatrout captured at that site. Trawl data solved oxygen, salinity, and shoreline vegeta­ from May-Sep. random sampling were com­ tion. We chose to examine spatial distribution bined over all years to examine overall trends on the basis of occurrences because initial rather than year-to-year variability. analyses with a delta-lognormal model (Lo et al., 1992; Stefansson, 1996) indicated that most i\llortalit;'.-Daily instantaneous total mortality of the total variation in YOY abundance that rates were estimated for spotted seatrout by could be accounted for by predictors occurred catch-curve analysis (Ricker, 1975). Natural in the presence/absence data. Therefore, the log-transformed values of mean number of in­ probability of the presence of spotted seatrout dividuals per 100 rn2 per age class were plotted was modeled as a binary response function of to select appropriate cutpoints for the de­ the following form: scending limb of the catch curve. To estimate mortality, the following equation was fitted to 11 log[P/(1 - P)] a + 2: f);X; + E, the abundance at age data by least-squares re­ i=l gression (PROC REG; SAS Institute, 1990): where P = Pr( Y = 1IX) is the response proba­ ln ( CPUE;) = ln ( CPUft0) - Z· A;, bilil:)', X; is the ith explanatory variable, n is the where ln( CPUE;) is the In-transformed mean number of explanatory variables, a is the in­ nmnber of individuals per 100 m 2 at age i, tercept, f); is the slope coefficient of explana­ E ln( CPUE0 ) is the intercept estimate, Z is the tory variable i, and is the error term (Agresti, estimate of total instantaneous mortality per 1996). day (slope), and A; is the age class in clays The selection of explanatory variables in the (Ricker, 1975). The age composition of YOY multiple logistic regression was accomplished spotted sea trout in seine catches was estimated by a stepwise logistic procedure (SAS Institute, from length and abundance data with the age­ 1997) with variables entered and removed on length predictive equation, the basis of a selection criterion of P ::; 0.05. Only data from July-Sep. 1996 for Choctaw­ 2 2 A = 12.472 + 1.836·L- 0.005·L , 1 = 0.88, hatchee Bay and June-Sep. of 1996 and 1997 derived from YOY otolith-aged spotted seatrout for Tampa Bay and Charlotte Harbor were in Tampa Bay (McMichael and Peters, 1989). used in the analyses because the spatial extent Mortality rates were estimated only for the of YOY seatrout distribution in each bay Tampa Bay population to avoid biases associ­ reached its maximum during these months. ated with potential interbay differences in The presence/absence of bottom vegetation growth rates of YOY spotted seatrout CWeste1~ (primarily seagrasses: H. wrightii, T. testudinum, heim and Ricker, 1978). '1\Te used data only and S)'ringodiu:m filiforme), mangrove (RAizopho­ from months during which the effects of im­ ra mangle, Avicennia genninans, and Lagunculm" migration in shallow-water areas appeared to ia racemosa) stands, rocks/ seawall structures, be minimal to avoid potential biases in the bare shoreline, overhanging shrubs/trees (e.g., rates of decline in abundance. Immigration Quercus spp.), reeds/marsh grasses ( e.g.,Juncus was assumed low when no decline in minimum roemerianus, Sjmrtina alternijlora), and mud size and/ or no substantial increases in catch were coded as dununy variables (i.e., 0 if ab­ rates were observed. Bias clue lo emigration sent and 1 if present). Deplh, temperature, sa­ was asstunecl low because there was no evi­ linil:)', and dissolved oxygen were transformed dence of migration fi·om the shallow-water ar­ with ln(x + 1). First-order interactions among eas, and the maximum size limits selected data the continuous and dummy variables were in­ on that could not avoid the seines. Abun­ cluded for selection in the initial model. Intel~ dance data were averaged over all years prior actions were considered important only if the to calculation of mean number of individuals associated main effects were also retained by at age to reduce the effects of interannual var­ the selection procedure, otherwise the inter­ iability. action was dropped and the analysis was re-

Published by The Aquila Digital Community, 2001 3 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3 NELSON AND LEFFLER-YOUNG-OF-THE-YEAR SEATROUT DYNAMICS 33

J M S J M S J M S J M S J M S J M S J M S J M S J M S

35 {:)' 28 ~ 21 ci. 14 ~ 7 I-

J M S J M S J M S J M S J M S J M S J M S J M S J M S Charlotte Harbor 35 0 28 ~ 21 ci.. 14 E 7 ~ 0.25 2l c 0.20 .{g 0.15 § 0.10 ff 0.05l 0.00 ~ JMSJMSJMSJMSJMSJMSJMSJMSJMS 1989 1990 1991 1992 1993 1994 1995 1996 1997

Fig. 2. Monthly relative indices of young-of-the-year abundance (no. of fish per 100 m 2) and mean monthly water temperature at fixed (FS) and random (RS) seine and trawl sites frorn 1989 to 1997 in Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor. Separate scales are shown for seine and trawl abundances.

peated. Plots of residual deviance and Pearson were used to determine relative ilTtportance of chi-square statistics, which provide measures of selected variables. Model regression coeffi­ how well the model predicts each sample ob­ cients were estimated by maximum likelihood servation, were examined to help identifY ob­ (SAS Institute, 1997). servations poorly fit by each model. If the ab­

solute value of deviance and Pearson residual RESULTS statistics for an observation were >2.0 and >3.0, respectively, those observations were re­ Seasonal changes in YOY relative abundance and moved from the data set and the stepwise anal­ size structu.re.-YOY spotted seatrout appeared yses were repeated (Beauchamp et al., 1992). first as postlarvae (9-17 mm) at shallow-water The final model's goodness-of-fit was evaluated seine sites during May or June in Choctaw­ by the deviance, Pearson residual chi-square hatchee Bay when average surface tenlpera­ statistics, the Hosmer-Lemeshow goodness-of­ tures were >27 C and during April or May in fit test (Hosmer and Lemeshow, 1989), and Tampa Bay and Charlotte Harbor when aver­ concordance measures (Beauchamp et al., age temperatures were >23 C (Fig. 2). Spotted 1992). Standardized regression coefficents seatrout were first collected at trawl sites in all https://aquila.usm.edu/goms/vol19/iss1/3 4 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 34 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

Choctawhatchee Bay Tampa Bay Charlotte Harbor 1.2 January 0.4 Januaz 0.4 n =2 !nr=18 0.8 n8 = 0 0.3 8 nr= 0 0.2 0.2 0.4 0.1 0.0 0.0 0.0 February 0.4 0.4 0.8 n8 = 0 0.4 nr=O 0.2 0.2 0.0 0.0 0.0 March March 0.8 0.8 0.8 n8 = 0 n8 = 2 0.4 nr= 0 0.4 0.4 nr=O 0.0 0.0 0.0 April April 0.8 0.4 0.2 n8 =1 n8 =67 0.4 nr= 0 0.2 0.1 nr= 1 0.0 0.0 0.0 May May 0.8 0.4 0.4 n = 311 n8 =1,176 8 0.4 0.2 nr=2 .. 0.2 nr= 1 c 0.0 0.0 0.0 0 June 0.3 June '€ 0.4 n =45 ns= 2,771 0.4 0 8 0.2 a. 0.2 nr= 11 nr=25 0.2 0.1 f\ e 0.0 0.0 0.0 0... July 0.2 0.2 0.2 n8 = 304 0.1 nr=53 0.1 0.1 .. :·: 0.0 0.0 0.0 August 0.3 August 0.3 0.2 n =1,886 n8 = 296 0.2 8 0.2 0.1 nr= 152 0.1 nr= 35 0.1 0.0 0.0 0.0 September 0.3 September 0.4 0.2 0.2 n8 = 266 n8 = 2,526 0.2 nr= 17 0.1 0.1 :. nr= ~.0 .. .. 0.0 0.0 0.0 0.3 October October 0.4 n = 2,044 0.2 0.2 ;n8 = 195 8 0.1 0.1 nr= 8 0.2 nr= 12 0.0 0.0 0.0 November 0.3 ;November 0.4 n =36 ~ !. 0.4 ;n =178 8 0.2 8 ;; nr=1 0.2 nr= 33:\ 0.1 idt 0.2 0.0 ::. 0.0 0.0 .. December 0.3 ~ n = 82 December 0.8 8 0.4 n =55 n8 = 3 0.2 8 0.4 nr= 1 0.1 / n:.~7 0.2 nr= 1 0.0 0.0 0.0 0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250 Standard Length (mm) Standard Length (mm) Standard Length (mm)

Fig. 3. Monthly length-frequency distributions of spotted seatrout captured at fixed and random seine (-) and trawl (· · ·) sites in Choctawhatchee Bay, Tampa Bay, and Charlotte Harbor from 1989 to 1997. n is the number of spotted seat:rout measured.

bays 1-2 mo after their initial appearance at in catch rates occurred in most years during seine sites, although their abundances in trawls Aug.-Oct. at seine sites in Tampa Bay and were very low (Fig. 2). Charlotte Harbor (Fig. 2). Low numbers of Catch rates of spotted seatrout at seine sites YOY spotted seatrout were generally captured peaked during July-Sep. in Choctawhatchee after Nov. when water temperature dropped Bay and during May-July in Tampa Bay and rapidly (Fig. 2). Charlotte Harbor (Fig. 2). During midsummer In all bays, the size range of YOY seatrout (July-Aug.), catch rates of YOY spotted sea­ captured at trawl sites was generally smaller trout declined slightly in most years when wa­ than the size range of seatrout captured at ter temperatures were maximmn in all bays ex­ seine sites during May-Oct. (Fig. 3). Few fish cept Choctawhatchee Bay (Fig. 2). In all bays, <30 mm SL were captured after Oct. in any of catch rates at trawl sites generally peaked in our collections, indicating settlement of post­ the same month or 1-2 mo after the peak at larvae had ended by Nov. (Fig. 3). Spotted sea­ fixed and random seine sites. A second peak trout (<75 mm SL) were caught at shallow-wa-

Published by The Aquila Digital Community, 2001 5 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3 NELSON AND LEFFLER-YOUNG-OF-THE-YEAR SEATROUT DYNAMICS 35

~120 E Choctawhatchee Bay FS RS .s 100 ~ 80 . c .. ·. J Q) 60 . _J "E 40 C

~ 120 E Tampa Bay FS RS .s 100 .s:: 0, 80 c Q) 60 .. _J ...:!' . . "E 40 C

~ 120 E Charlotte Harbor FS RS .s 100 .s:: 0, 80 c J. Q) .t :J. 60 . ·~. ;~ :1 n }),..., ;-.... ·(·: .... _J .::. ... ~ MJ~. . "E 40 . . . C

ter sites through Dec. in Choctawhatchee Bay emigrated from the shallow-water area or were and through Feb. in Tampa Bay and Charlotte avoiding the sampling gears (Fig. 4). In Choc­ Harbor (Fig. 3). tawhatchee Bay, a similar pattern was indicat­ Plots of the monthly length statistics showed ed, but lengths did not decline after a summer that seasonal changes in the size structure of decline in catch rates. Mean length of spotted YOY spotted seatrout occurred annually in the seatrout in Tampa Bay continued to increase shallow-water areas. Mean length of YOY spot­ through fall (Oct.-Nov.) in most years, but it ted sea trout increased after the months of first generally declined through late fall (Nov.­ capture at fixed and random. seine sites (Fig. Dec.) in Charlotte Harbor (Fig. 4). L1). In Tampa Bay and Charlotte Harbor, mean length declined in Aug.-Sep. in most years, af­ Dej1th distribution.-About 90% of YOY spotted ter the summer catch rate decline (July-Aug.). seat:rout caught in trawls during May-Nov. Monthly minimum and, occasionally, maxi­ were captured in waters <3.7 m, <3.5 m, and mum lengths also declined during Aug.-Sep., <3.0 m in Choctawhatchee Bay, Tampa Bay, indicating both that smaller seatrout ( <25 mm and Charlotte Harbor, respectively. Few fish SL) were recruiting to shallow-water areas on ( <1% of total catches) were captured in waters or after the summer decline in abundance and >4 m. The Kolmogorov-Smirnov test showed that large seatrout (>90 mm SL) had either that the cumulative frequency distribution for https://aquila.usm.edu/goms/vol19/iss1/3 6 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 36 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

spotted seatrout depth was significantly differ­ in the initial selection phase remained in the ent from that of trawl depth in Tampa Bay and logistic analyses. The nonsignificance of most Charlotte Harbor; there was no significant dif­ goodness-of-fit statistics and the moderate to ference in Choctmvhatchee Bay, but occur­ high concordance measures indicated that the rences of spotted seatrout in trawl catches final models fit the spotted seatrout data ade­ there were few (Choctawhatchee Bay: D(KS quately after data from one random site in test statistic) = 0.31, nfish u·awls = 10, ntrawls = Choctawhatchee Bay, 11 in Tampa Bay, and 12 376, n.s.; Tampa Bay: D = 0.49, nfish trawls = 35, in Charlotte Harbor were removed from the ntrawls = 573, P ::; 0.001; Charlotte Harbor: D original data sets (Table 2). = 0.40, nfish u·awls = 17, ntrawls = 429, p::; 0.05). Mortality.-Estimates of daily instantaneous to­ Factors associated with YOY spatial occurnmce.­ tal mortality (Z) in Tampa Bay were calculated YOY spotted seatrout occurred at 28.7%- from the decline in relative abundance at age 52.3% of the seine sites sampled during the at shallow-water fixed and random seine sites; summer periods in 1996 and 1997 (Table 1). we used June data to avoid potential bias as­ In all bays, seatrout were captured most fre­ sociated with the influx of the late summer co­ quently at offshore seine sites and at sites with hort in July. Estimates were made for fixed-site bottom vegetation (Table 1). In Choctawhatch­ and random-site data separately with data from ee Bay, seatrout occurred most frequently at age classes 42-98 clays and 50-90 clays, respec­ sites with seawall/rock (34.7%), reed/marsh tively (Fig. 5). Estimates of Z were 0.027·cl-1 grass (35.1 %), or bare (29.0%) shorelines. In (fixed) and 0.025·cl-1 (random) (Table 3). Tampa Bay and Charlotte Harbor, YOY sea­

trout were captured most frequently at sites DISCUSSION with mangrove shorelines (57.4% and 59.5%, respectively) and less frequently at sites with Seasonal changes in YOY abundance and size struc­ seawall/rock (49.3% and 50.0%), reed/marsh ture.-Aclult spotted seatrout are believed to grass (44.8% and 40.0%), or bare (35.1% and spawn during the night in deep channels and 27.8%) shorelines (Table 1). In Choctawhatch­ depressions near grass flats in high-salinity ar­ ee Bay and Tampa Bay, YOY also occurred fre­ eas of estuaries when temperatures are >21 C quently at sites with mud sediments (Table 1). (Tabb, 1966; Helser et al., 1993). Pelagic sea­ The environmental parameters recorded at trout larvae are presumably dispersed via es­ the capture sites were relatively similar in all tuarine currents to nearshore areas, where three bay systems. Mean dissolved oxygen they settle near the bottom (Peebles and Tol­ ranged 5.8-6.5 ppm, and mean temperature, ley, 1988). The appearance ofpostlarvae (9-17 29.9-30.4 C, and mean depth was 0.7 min all mm) and small YOY spotted seatrout ( <30 bays (Table 1). Mean salinity was the only var­ mm) in shallow- and deepwater sites shows that iable that differed greatly between bays; it was settlement occurs to both areas. The absence lower (16.4 ppt) in Choctawhatchee Bay than of spotted seatrout at deepwater sites until 1- in Tampa Bay (27.2 ppt) and Charlotte Harbor 2 mo after their first appearance at seine sites (26.2 ppt) (Table 1). probably reflects the low levels of spawning ac­ The probability of capturing a YOY spotted tivity during the early season (e.g., Brown-Pe­ seatrout was related to some habitat attributes tersen et al., 1988) and the low probability of common to all three bays and to some that capturing individuals when abundance is low were bay specific. In all bays, the presence of with an inefficient sampling gear like a trawl. bottom vegetation was the most important var­ The occurrence of YOY sea trout mostly in wa­ iable positively associated with YOY occurrenc­ ters <3.7 m is probably clue to their depen­ es, as indicated by the standardized regression dence on seagrasses for cover (seagrasses are coefficents (Table 2). The probability of cap­ generally restricted to waters <3.0 m in Choc­ turing seatrout was positively related to sam­ tawhatchee Bay, Tampa Bay, and Charlotte pling depth and negatively related to salinity Harbor; M. 0. Ruark, Florida Marine Research in Tarnpa Bay and Charlotte Harbor, and in nstitute, St. Petersburg, FL, pers. comm.) and Choctawhatchee Bay and Tampa Bay, it was to the distribution of their prey, which are also positively related to dissolved oxygen levels restricted to shallow-water areas (McMichael (Table 2). The probability of capturing a YOY and Peters, 1989). seatrout was also positively related to the pres­ We believe that the differences between es­ ence of mud in Choctawhatchee Bay and to tuaries in the timing of the initial appearance the presence of mangroves in Tampa Bay (Ta­ of YOY seatrout and in the seasonal patterns ble 2). No first-order interactions considered of the abundance and sizes ofYOY are partially

Published by The Aquila Digital Community, 2001 7 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3

z t'1 TABLE 1. (A) Total number of seine sites (n) sampled, number of sites at which spotted seatrout occurred (Trout), and percentage of seatrout occurrences (%) t""' characterized by year, deployment technique, bottom vegetation, shoreline type, and sediment types, and (B) summary statistics for environmental parameters at CFJ 0 sites where spotted seatrout occurred duringJuly-Sep. in Choctawhatchee Bay andJune-Sep. in Tampa Bay and Charlotte Harbor. z Choctawhatchee Bay Tampa Bay Charlotte Harbor ~ Variable Category n Trout % n Trout % n Trout % d t""' A t'1 >rj Year 1996 108 31 28.7 170 89 52.3 124 48.4 >rj 60 t""' 1997 - 170 85 50.0 124 59 47.6 t'1 Deployment Offshore 54 20 37.0 226 121 53.5 128 68 53.1 Beach 54 11 20.4 114 53 46.5 120 51 42.5 r Bottom vegetation Unvegetated 72 9 12.5 134 39 29.1 52 12 23.1 d c Vegetated 36 22 61.1 206 135 65.5 196 107 54.6 z Shore Type Mangrove 0 0 0.0 169 97 57.4 208 103 59.5 0 Seawall 23 8 34.7 73 36 49.3 6 3 50.0 0 Reeds 37 13 35.1 29 13 44.8 5 2 40.0 71 >-l Shrub/trees 13 0 0.0 5 1 20.0 3 1 33.3 ::c Bare shore 31 9 29.0 37 13 35.1 18 5 27.8 t'1 Other 0 0 0.0 5 2 40.0 6 5 83.3 Unknown 4 1 25.0 22 12 54.5 2 0 0.0 ~ Sediment Mud 19 11 57.9 124 71 57.2 77 37 48.0 ~ CFJ Sand 89 20 22.5 214 103 48.1 171 82 48.0 Mean SE Range Mean SE Range Mean SE Range ~ B 10 0 Dissolved oxygen (ppm) 6.3 0.16 3.0-8.9 5.8 0.11 0.5-11.9 6.5 0.15 1.0-11.8 c Salinity (ppt) 16.4 0.65 0.0-27.7 27.2 0.28 0.3-37.0 26.2 0.49 0.8-39.0 >-l Temperature (C) 29.9 0.14 26.7-34.5 29.9 0.08 26.3-34.6 30.4 0.10 26.1-34.6 d Depth (m) 0.7 0.03 0.2-1.9 0.7 0.01 0.3-1.4 0.7 0.01 0.3-1.2

~H [') CFJ

(.)() https://aquila.usm.edu/goms/vol19/iss1/3 -0 8 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 38 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

TABLE 2. Results of the stepwise logistic regression of the presence of young-of-the-year spotted seatrout on year, ten>perature, salinity, depth, dissolved oxygen, seagrasses, shore vegetation, deployment technique, and sediment type for Choctawatchee Bay, Tampa Bay, and Charlotte Harbor.

Analysis of maximum likelihood estimatesa Parameter Parameter coefficient (SE) Waldx2 Standardized coefficients Choctawhatchee Bay Intercept -10.647 (3.9715) 7.19** Bottom vegetation 3.123 (0.6400) 23.81*** 0.817 Mud 2.529 (0.7444) 11.53*** 0.535 Dissolved oxygen 3.867 (1.9010) 4.37* 0.369 Goodness-of-fit statistics Chi-square 79.9 (P > 0.90) Deviance 81.2 p > 0.90) H-L test 9.8 (P > 0.19) n 107 Concordance 0.881 Tampa Bay Intercept 6.355 (2.5226) 6.34* Mangrove 0.744 (0.2584) 8.29** 0.205 Bottom vegetation 1.960 (0.2791) 49.31 *** 0.529 Salinity -3.468 (0.7991) 18.83*** -0.492 Depth 3.882 (0.9250) 17.61 *** 0.315 Dissolved m:ygen 0.847 (0.3566) 5.64* 0.168

Goodness-of~fit statistics Chi-square 319.3 (P > 0.60) Deviance 369.9 (P > 0.06) H-L test 14.2 (P > 0.08) n 329 Concordance 0.786 Charlotte Harbor Intercept 2.500 (1.7340) 2.08 n.s. Bottom vegetation 2.578 (0.4952) 27.11*** 0.561 Salinity -2.149 (0.5751) 13.97*** -0.349 Depth 4.348 (1.1826) 13.52*** 0.313 Goodness-of-fit statistics Chi-square 235.6 (P > 0.40) Deviance 282.7 (P < 0.02) H-L test 5.3 (P > 0.70) n 236 Concordance 0.741

a n.s. = not significant, *P :5 0.05, **P:::; 0.01, and ***P :5 0.001.

related to the estuary-specific differences in estuaries and twice (early spring and early sum­ the reproductive activities of adult spotted sea­ mer) in the southern peninsula estuaries (IGi­ trout. Reproduction begins about 1 n1o earlier ma and Tabb, 1959; McMichael and Peters, (lviarch-April) in southern Florida estuaries 1989; DeVries eta!., 1997). Thus, the first peak than in northern Florida Panhandle estuaries in abundance-observed during July-Sep. in (April-May) (DeVries et al., 1997; M. D. Mur­ Choctawhatchee Bay and May-July in Tampa phy, Florida Marine Research Institute, St. Pe­ Bay and Charlotte Harbor-is probably made tersburg, FL, pers. comm.), which would ac­ up of those individuals spawned during spring, count for YOY appearing 1 mo earlier in Tam­ and the second peak in abundance-observed pa Bay and Charlotte Harbor than in Choctaw­ during Aug.-Oct. in Tampa Bay and Charlotte hatchee Bay. Spawning activity is also known to Harbor-is probably made up of those individ­ peak only once (early summer) in panhandle uals spawned during summer. The decline in

Published by The Aquila Digital Community, 2001 9 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3 NELSON AND LEFFLER-YOUNG-OF-THE-YEAR SEATROUT DYNAMICS 39

2 Fixed Stations 1989-1995 1

0 -1 (\J--- E -2 ••• 0 0 • • ..- -3 • • ...... ••• Q) • 0... -4 ··- ..c (/) '+= Random Sampling c -1 1996-1997 ct:S Q) ..._~ -2 • c • -3

-4 .....• • ·- .

-5 20 40 60 80 100 120 140 160 Estimated Age (days)

Fig. 5. Catch curves (least-square regression lines) fit to the relative abundance at age for YOY spotted seatrout at fixed stations and random sites.

mean and minimum length in late summer­ es) was the most important habitat variable early fall after the midsummer abundance de­ commonly associated with YOY spotted sea­ cline in most years in Tampa Bay and Charlotte trout occurrences in the three Florida estuar­ Harbor also confirms that the observed second ies studied. Studies on YOY spotted sea trout by peak in abundance is probably due to the set­ Chester and Thayer (1990), Rutherford et al. tling of the second cohort of YOY in the shal­ (1989), and McMichael and Peters (1989) also low areas (Fig. 2). found that YOY seatrout distribution was asso­ ciated with seagrasses. Because seagrasses pro­ Factors associated with YOY sjJatial occurrences.­ vide fish with both a place to hide from pred­ The presence of bottom vegetation (seagrass- ators and a place to forage for diverse prey,

TABL!c 3. Summary of regression statistics from the catch curve analyses of young-of-the-year spotted sea­ trout in Tampa Bay at fixed sites during 1989-95 and random sites during 1996-97. Data from only age classes 42-98 days at fixed sites and 50-90 days at random sites, respectively, were used in the analyses. All intercepts and slopes were tested for significance from zero.

In(CPUE;,) SE[In(CPUf;,)] z SE(L:) }

***P:::; 0.001; n.s. = not significant. https://aquila.usm.edu/goms/vol19/iss1/3 10 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 40 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

seagrass habitats are believed to be very im­ groves. In Tampa Bay, mangrove distribution portant to survival of YOY seatrout and other has been severely fragmented by urbanization estuarine fishes (Stoner, 1983; Chester and and shoreline development (Lewis et al., Thayer, 1990; Ruiz eta!., 1993; Rooker eta!., 1985), but in Charlotte Harbor, where the 1998). shoreline remains relatively undeveloped, The significant associations between depth, mangrove distribution is nearly continuous oxygen, and mud and the spatial occurrences (Hamm.et, 1990; Blewett, Florida Marine Re­ of YOY spotted seatrout were difficult to assess search Institute, Charlotte Harbor Field Labo­ without 1neaningful, first-order interactions in ratory, Port Charlotte, FL, pers. coml'n.). The the stepwise logistic analysis. However, because positive relationship between mangrove pres­ YOY abundance was strongly associated with ence and YOY seatrout occurrences found in seagrasses, the significance can be proposed in Tampa Bay, but not in Charlotte Harbor, may relation to the distribution of seagrasses. YOY then represent the effect of the plant's frag­ spotted seatrout occurrences in shallow-water mented shoreline distribution and its localized areas may be positively associated with water influence on the detritus-based food web that depth in Tampa Bay and Charlotte Harbor be­ supports prey of YOY seatrout. cause biomass of seagrasses increases with dis­ tance from shore and depth in estuaries (Eleu­ 1\llortalit,y.-Our estimates of daily mortality of terius, 1987). The positive association between YOY spotted seatrout in the shallow water of dissolved oxygen and seatrout presence may Tampa Bay were relatively low. These estimates reflect differences between seagrass areas (fixed station = 0.027; random station = (high oxygen production; Odum, 1957) and 0.025) were lower than those made for YOY unvegetated and/ or eutrophic sites (low oxy­ spotted seatrout in (0.035) (Ruth­ gen production) of the moderately and heavily erford et a!., 1989) and were comparable to developed shorelines of Choctawhatchee Bay but higher than those made for YOY of estua­ and Tampa Bay, respectively (Irby, 1974; Avery, rine-dependent species such as pinfish (Lago­ 1997). Because mud is commonly associated don rhomboides) in the same Florida estuaries with the patchily distributed seagrass beds in (0.021-0.023) (Nelson, 1998); gulf menhaden shallow-water areas of Choctawhatchee Bay (Brevoortia jJatronus) in Fourleague Bay, LA (pers. obs.), higher occurrences ofYOYat mud (0.017-0.021) (Deegan, 1990); and two sciaen­ sites may be expected there. ids; spot (Leiostomus xanthurus) in York River, Many fish species tend to be distributed VA (0.017) (Weinstein, 1983), and Atlantic along environmental gradients in estuaries, es­ croaker (Niicropogonias undulatus) in Rose Bay, pecially salinity gradients (Moser and Gerry, NC (0.023) (Currin eta!., 1984). 1989; Cyrus and Blaber, 1992; vVhitfield, 1999). Applying the age-length relationship for For YOY spotted seatrout, our results strongly spotted seatrout developed from. fish captured suggest that their spatial occurrences are neg­ during the early 1980s to length data collected atively correlated with salinity in bays with a from seatrout captured during 1989-97 intro­ wide salinity range. A similar pattern in the duced a risk of biasing the age structure of the abundance of large, post-spawning seatrout length distributions if changes in growth or was found by Helser eta!. (1993) in four Lou­ survival rates had occurred between the two isiana estuaries. The observed pattern may be time periods (Westerheim and Ricker, 1978). the result of physiological preferences for sa­ However, this potential bias may not be great linities that minimize metabolic costs and op­ because the length data collected in this study timize growth and survival (Wohlschlag and (1989-97) and the original otolith-length data ·vvakeman, 1978) or of behavioral responses to (collected in the early 1980s) used by Mc­ avoid stenohaline predators (Dahlberg, 1972; Michael and Peters (1989) were both com­ Odum., 1988). It is difficult to conclude which bined over several years, which would have mechanisms are operating in Tampa Bay and dampened interannual differences. Charlotte Harbor because data on growth and In summary, YOY seatrout first appeared in survival over the range of environmental gra­ shallow-water areas during May-June in Choc­ dients and predator distributions are currently tawhatchee Bay and during April-May in Tam­ lacking. pa Bay and Charlotte Harbor. In all bays, they Mangroves play an important role as a major were caught at deepwater sites within 1-3 mo source of detritus in south Florida estuaries after their shallow-water appearance. Spring (Lewis et a!., 1985). Many fish and inverte­ and sununer peaks in YOY abundance, corre­ brates eaten by YOY spotted seatrout rely on sponding to strong influxes of newly spawned the detritus-based food web supported by man- individuals, were observed only in Tampa Bay

Published by The Aquila Digital Community, 2001 11 Gulf of Mexico Science, Vol. 19 [2001], No. 1, Art. 3 NELSON AND LEFFLER-YOUNG-OF-THE-YEAR SEATROUT DYNAMICS 41

and Charlotte Harbor. YOY spotted seatrout CYRUS, D.P., AND S.J. M. BLAilER. 1992. Turbidity and were generally restricted to depths <3.7 m in salinity in a tropical northern Australian estuary all bays. The occurrence of YOY spotted sea­ and their influence on fish distribution. Estuarine trout in shallow-water areas was associated with Coastal Shelf Sci. 35:545-563. DAHLBERG, M. D. 1972. An ecological study of Geor­ the presence of seagrasses and mangroves; with gia coastal fishes. Fish. Bull. 70:323-353. salinity, dissolved oxygen, and depth; and with DEEGAN, L.A. 1990. Effects of estuarine environmen­ the presence of rnud sediments. Estimates of tal conditions on population dynamics of young­ total instantaneous mortality for YOY spotted of-the-year gulf menhaden. Mar. Ecol. Prog. Ser. seatrout in Tampa Bay were comparable to 68:195-205. published values for other sciaenid species. DEVIUES, D. A., C. D., BEDEE, C. L. PALMER, Al'ID S. A. BORTONE. 1997. Age, growth, maturity, and size cornposition of spotted seatrout, Cynoscion nebulo­ AcKNOvi'LEDGMENTS ws, in the Panhandle Region of Florida. Final re­ Funding for this study was provided in part port for Florida Department of Environmental Protection, Marine Resources Grants MR019 and by the State of Florida Recreational Fishing Li­ MR024. cense and in part by the Department of the ELEUTERIUS, L. N. 1987. Seagrass ecology along the Interior, U.S. Fish and Wildlife Service, Federal coasts of Alabama, Louisiana, and Mississippi. In: Aid for Sportfish Restoration, Project Number Proceedings of the symposium on sub-tropical­ F-43 to the Florida Fish and Wildlife Conser­ tropical seagrass of the southeastern United vation Commission. We thank the staff of the States. M.]. Durako, R. C. Phillips, and R. R. Lew­ Fisheries-Independent Monitoring Program at is, II (eels.), Fla. Mar. Res. Pub!. 42:11-25. the Florida Marine Research Institute for their HAMMETT, K. l'vi. 1990. Land use, water use, stream­ dedication to sampling. H. ]. Simpson provid­ flow characteristics, and water quality characteris­ ed encouragement throughout the study. Com­ tics of the Charlotte Harbor inflow area, Florida. U.S. Geol. Surv. Water-Supply Paper 2359-A. ments by Tim MacDonald, Daryl Pierce, Mike HELSER, T. E., R. E. CoNDREY, AND J. P. GEAGHAl'l. Murphy, Bob McMichael, Judy Leiby, and Jim 1993. Spotted seatrout distribution in four coastal Quinn improved the quality of the manuscript. Louisiana estuaries. Trans. Am. Fish. Soc. 122:99- 111. LITERATURE CITED HOSMER, D. W.,JR., AND S. LEMESHOW. 1989. Applied logistic regression. John Wiley and Sons, Inc., New AGRESTI, A. 1996. An introduction to categorical data York. analysis. John ~Wiley & Sons, Inc., New York, NY. IRBY, E. v\T., JR. 1974. A fishing survey of Choctaw­ AvERY, IN. 1997. Distribution and abundance of ma­ hatchee Bay and adjacent Gulf of Mexico waters. croalgae and seagrass in Hillsborough Bay, Flori­ Fla. Mar. Res. Pub!. 2. da, fi·om 1986 to 1995, pp. 151-166. In: Proceed­ JOHNSON, D. R. 1982. Effects of rainfall, recruitment, ings of the Tampa Bay area scientific information and operation of the Cedar Bayou Electric Gen­ symposium 3. S. F. Treat (ed.). Tampa Bay Region­ erating Station (Baytown, Texas) on the dynamics al Planning Council, Tampa, FL. of fish and macrocrustacean communities in the BEAUCI-IAMP,J.J., S. W. CHIUSTENSEN, AND E. P. SMITH. brackish water intake and discharge area. Ph.D. 1992. Selection of factors afiecting the presence cliss., Texas A&!VI Univ., College Station, TX of brook trout ( Salvelinus fontinalis) in Adirondack ---, AND W. SEA!\IAN, JR. 1986. Species profiles: lakes: a case study. Can.]. Fish. Aquat. Sci. 49:597- life history and environmental requirements of 608. coastal fishes and invertebrates (south Floricla)­ BROWN-PETERSEN, N., P. THOMAS, Al'ID C. R. ARNOLD. spotted seatrout. U.S. Fish Wild!. Serv. Bioi. Rep. 1988. Reproductive biology of the spotted sea­ 82(11.43); U.S. Army Corps of Engineers, TR EL- trout, Cynoscion nebulosus, in South Texas. Fish. 82-4. Bull. 86:373-387. KLil1IA, E. F., AND D. C. TABB. 1959. A contribution CAIU{, W. E. S., AND C. A. ADAl\IS. 1973. Food habits to the biology of the spotted , Cynoscion of juvenile marine fishes occupying seagrass beds nebulosus (Cuvier), from northwest Florida, with a in the estuarine zone near Crystal River, Florida. description of the fishery. Fla. Board Conserv. Trans. Am. Fish. Soc. 102:511-540. Mar. Res. Lab. Tech. Ser. 30. CriESTER, A.J., AND G. W. THAYER. 1990. Distribution LEWIS, R. R. III, R. G. GILMORE, JR., D. W. CREWZ, AND of spotted sea trout ( Cynoscion nebulosus) and gray W. E. 0DUM. 1985. Mangrove habitat and fishery snapper (Lutjanus griseus) juveniles in seagrass resources of Florida, pp. 281-336. In: Florida habitats of western Florida Bay. Bull. Mar. Sci. 46: aquatic habitat and fishery resources. W. Seaman, 345-357. Jr. (eel.). Florida Chapter of the American Fishel~ CURRIN, B. M., J. P. REED, AND J. M. MILLER. 1984. ies Society, Eustis, FL. Growth, production, food consurnption, and mor­ Lo, N. C., L. D. JACOBSON, AND J. L. SQUIRE. 1992. talil:)' of juvenile spot and croaker: a comparison Indices of relative abundance from fish spotter of tidal and nontidal nursery areas. Estuaries data based on delta-lognormal models. Can. J. 7(4A):451-459. Fish. Aquat. Sci. 49:2515-2526. https://aquila.usm.edu/goms/vol19/iss1/3 12 DOI: 10.18785/goms.1901.03 Nelson and Leffler: Abundance, Spatial Distribution, and Mortality of Young-of-the-Ye 42 GULF OF MEXICO SCIENCE, 2001, VOL. 19(1)

McMICHAEL, R H., AND K. M. PETERS. 1989. Early life SAS INSTITUTE. 1990. Proc REG procedures. SAS/ history of spotted sea trout, Cynoscion nebulosus (Pi­ Stat user's vol. 2. Release 6.04 ed. 3, SAS Institute sces: ), in Tampa Bay, Florida. Estuaries Inc., Cary, NC. 12:98-100. ---. 1997. Proc LOGISTIC procedures. SAS/ MosER, M. L., AND L. R. GERRY. 1989. Differential STAT Software: changes and enhancements effects of salinity changes on two estuarine fishes, through Release 6.12. SAS Institute Inc., Cary, NC. Leiostomus xanthu.nts and l'v!icropogonias undulatus. SIEGEL, S. 1956. Nonparametric statistics for the be­ Estuaries 12:35-41. havioral sciences. McGraw-Hill Book Company NELSON, G. A. 1998. Abundance, growth, and mo1~ Inc., New York. tality of young-of-the-year pin fish, dwmboi­ SriUNGER, V. G., AND K. D. WooDBUiu'\1. 1960. An eco­ logical study of the fishes of the Tampa Bay area. des, in three estuaries along the gulf coast of Flm~ ida. U.S. Fish. Bull. 96:315-328. Fla. Board Conserv. Mar. Res. Lab. Prof. Pap. Ser. ODUM, H. T. 1957. Primary production measure­ 1:1-104. ments in eleven Florida springs and marine turtle­ STEFANSSON, G. 1996. Analysis of groundfish survey abundance data: cornbining the GLM and delta grass community. Limnol. Oceanogr. 2:85-97. approaches. ICES]. Mar. Sci. 53:577-588. 0DUM, W. E. 1988. Comparative ecology of tidal STONER, A. W. 1983. Disu-ibution of fishes in seagrass fi-eshwater and salt marshes. Annu. Rev. Ecol. Syst. meadows: role of macrophyte biomass and species 19:147-156. composition. Fish. Bull. 81:837-846. PEEBLES, E. B., AND S. G. TOLLEY. 1988. Distribution, TABB, D. C. 1966. The estuary as a habitat for spotted gJ-owth and mortality oflarval spotted sea trout, Cy­ seatrout, Cynoscion nebulosus. Am. Fish. Soc. Spec. noscion nebulosus: a comparison between two adja­ Pub!. 3:58-67. cent estuarine areas of southwest Florida. Bull. WEINSTEIN, M. P. 1983. Population dynamics of an Mar. Sci. 42:397-410. estuarine-dependent fish, the spot (Leiostomus xan­ PERRY, R 1., AND S.J. SMITH. 1994. Identifying habitat thunts), along a tidal-creek-seagrass meadow coen­ associations of marine fishes using survey data: an ocline. Can. J. Fish. Aquat. Sci. 40:1633-1638. application to the northwest Atlantic. Can.]. Fish. WESTERHEIM, S. j., AND W. E. RICKER. 1978. Bias in Aquat. Sci. 51:589-602. using an age-length key to estimate age frequency RicKER, vV. E. 1975. Computation and interpretation distributions. J. Fish. Res. Board Can. 35:184--189. of biological statistics of fish populations. Bull. WHITFIELD, A. K. 1999. Ichthyofaunal assemblages in Fish. Res. Board Can. 191:1-382. estuaries: a South African case study. Rev. Fish RooKER,]. R, G.]. HoLT, AND S. A. HoLT. 1998. Bioi. Fish. 9:151-186. Vulnerability of newly settled red drum (Sciaenops WOHLSCHLAG, D. E., AND j. M. WAKEMAN. 1978. Salin­ ocellatus) to predatory fish: is early-life survival en­ ity stresses, metabolic responses and distribution harlCed by seagrass meadows? Mar. Bioi. 131:145- of the coastal spotted seau-out, Cynoscion nebulas u.s. 151. Conu-ib. Mar. Sci. 21:171-185. Rurz, G. M., A. H. HINES, AND M. H. PosEY. 1993. FLORIDA FISH AND WILDLIFE CONSERVATION Shallow water as a refuge habitat for fish and crus­ taceans in non-vegetated estuaries: an exmnple COMMISSION, FLORIDA MARINE RESEARCH IN­ from Chesapeake Bay. Mar. Ecol. Prog. Ser. 99: STITUTE, 100 EIGHTH AVENUE SE, ST. PETERS­ 1-6. BURG, FLORIDA 33701-5095. PRESENT ADDRESS RUTHERFORD, E. S., T. W. SCHMIDT, AND J. T. TIU·!ANT. ( GAN): MASSACHUSETTS DIVISION OF MARINE 1989. Early life history of spotted sea trout ( Cynos­ FISHERIES, ANNISQUAl'v1 RivER MARINE FISHER­ cion nebulosus) and gray snapper (Lutjanw griseus) IES STATION, 30 EMERSON AVENUE, GLOUCES­ in Florida Bay, National Park, Florida. TER, l'viAsSACHUSETTS 01930. Date accepted: Bull. l\hr. Sci. 44:49-64. December 10, 2000.

Published by The Aquila Digital Community, 2001 13