Shallow water predation risk for a juvenile flatfish (winter flounder; Pseudopleuronectes americanus, Walbaum) in a northwest Atlantic estuary

Journal of Experimental Marine Biology and Ecology 304 (2004) 137 – 157 www.elsevier.com/locate/jembe

Shallow water predation risk for a juvenile flatfish (winter flounder; Pseudopleuronectes americanus, Walbaum) in a northwest Atlantic estuary John P. Manderson a,*, Jeffrey Pessutti a, John G. Hilbert a, Francis Juanes b a

James J. Howard Marine Sciences Laboratory, Behavioral Ecology Branch, Department of Commerce, NOAA/National Marine Fisheries Service, Highlands, NJ 07732, USA b Department of Natural Resources Conservation, University of Massachusetts, Amherst, MA 01003-9285, USA Received 8 April 2003; received in revised form 3 May 2003; accepted 15 December 2003

Abstract Many small fish, including several juvenile Atlantic flatfish, are most abundant in shallow areas presumable because these habitats enhance survivorship and/or growth. In this study, we investigated size-dependent depth distributions and the role of shallow habitats as predator refuges for age-0 winter flounder (Pseudopleuronectes americanus) in a northwest Atlantic estuarine nursery. Analysis of trawl surveys performed during the larval settlement period throughout the Navesink River and Sandy Hook Bay, New Jersey, showed that as fish increased in size, depth of occurrence gradually decreased, so that individuals >35 mm standard length (SL) were concentrated in habitats f1 m deep. Tethering in structurally simple and adjacent shallow and deep habitats showed that predation risk for flounder (30 – 50 mm SL) was low in shallow water ( < 1 m) and increased rapidly with depth. Summer flounder (Paralychthys dentatus), which were more abundant in trammel nets in deep habitats and included winter flounder in their diets, appeared to be important consumers of tethered fish. Our results indicate that following larval settlement, winter flounder emigrate from or suffer high mortality in deeper water to become concentrated in shallow habitats that can serve as predator refuges even when they lack complex physical structures. These results highlight the potential for functional habitat loss when natural and/or anthropogenic factors make shallow habitats unavailable to young fish. Published by Elsevier B.V. Keywords: Predation refuge; Nursery habitat quality; Tethering; Juvenile fish mortality

* Corresponding author. Tel.: +1-732-872-3057; fax: +1-732-872-3088. E-mail address: [email protected] (J.P. Manderson). 0022-0981/$ - see front matter. Published by Elsevier B.V. doi:10.1016/j.jembe.2003.12.004

138

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

1. Introduction Shallow coastal habitats throughout the world are thought to promote the survivorship of early life history stages of fish and invertebrates and thus serve as important nurseries (Blaber et al., 1995; Norcross et al., 1995; Adams and Ebersole, 2002; Nagelkerken et al., 2002). Structurally complex habitats including those formed by biogenic structures (e.g., submerged aquatic vegetation, reef building invertebrates) are common in shallow water and known to enhance survivorship (Stoner, 1982; Wilson et al., 1990; Tupper and Boutilier, 1997; Dahlgren and Eggleston, 2000; Holbrook and Schmitt, 2002; Peterson et al., 2002). Habitats lacking complex structures are generally not considered to be high in quality, but there is evidence that simple habitats in shallow water serve as predator refuges for a few round fish and decapod crustaceans (Ruiz et al., 1993; Dittel et al., 1995; Hines and Ruiz, 1995; Linehan et al., 2001; Clark et al., 2003). Simple shallow habitats may function as prey refuges because depth distributions of marine animals generally increase with body size and thus large predators are often less abundant in shallow water (Macpherson and Duarte, 1991; Paterson and Whitfield, 2000; Linehan et al., 2001). However, the refuge role of structurally simple shallow habitats may not be universal because some predators are not limited to deep water by body size or behavior (see Sheaves, 2001, for a discussion). Many flatfish, including winter flounder (Pseudopleuronectes americanus), use shallow habitats as nurseries (Gibson, 1997; Gibson et al., 1998, 2002; Norcross et al., 1999; Stoner et al., 2001; Amara and Paul, 2003). In estuaries, winter flounder settle in the spring in depositional habitats that can be > 5 m deep (Manderson et al., 2003). In many systems, larger juveniles are concentrated in shallow habitats 1 –2 m deep that are often structurally complex (e.g., vegetated beds, woody debris fields; Tyler, 1971; Wells et al., 1973; Black and Miller, 1991; Howell et al., 1999; Stoner et al., 2001; Phelan et al., unpublished manuscript; but see Armstrong, 1997) and thus, probably serve as predator refuges (Manderson et al., 2000). Juveniles also occur on shallow sandbars and beaches that lack complex physical structures (Tyler, 1971; Goldberg et al., 2002). While these shallow and simple habitats may serve as refuges, several important predators, including summer flounder (Paralichthys dentatus) and wading birds, forage in shallow water during some stages of the tide (Burger, 1983; Rountree and Able, 1992, 1997; Manderson et al., 2000). Thus, the efficacy of these habitats as refuges for juvenile flounder may be predator-specific and tidally dependent. We used field surveys and a tethering experiment to investigate the functional importance of shallow habitats for age-0 winter flounder on an estuarine nursery ground. We used beam trawl collections made near the end of the winter flounder settlement period to identify size-dependent changes in the depth distribution of young fish. We then examined depth-dependent variation in predation risk by testing the following null hypotheses in the field using tethering: 1. Predator encounter rates for flounder were the same in structurally simple shallow ( < 0.5 m) and adjacent deep ( > 1.3 m) habitats during the morning crepuscular period.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

139

2. Predator encounter rates for flounder in adjacent shallow and deep habitats were similar during high and low tides. We also sampled larger fish and their diets in the habitats to identify potentially important predators.

2. Materials and methods 2.1. Changes in winter flounder depth distribution with body size Age-0 winter flounder were collected with replicate 1-m beam trawls (N = 2; 3 mm mesh liner) at 84 stations located throughout the Navesink River Sandy Hook Bay estuarine system (NSBES), New Jersey, during late May through early June, 1997 and 1998 (for sampling details, see Stehlik and Meise, 2000; Stoner et al., 2001). We selected May surveys because: (1) a broad size range of fish, from settlers [8 mm standard length (SL)] to individuals 50 mm SL were collected and; (2) water temperatures were cool throughout the estuary (13 – 21 jC) and unlikely to cause fish to avoid shallow habitats which can become exceptionally warm ( > 28 jC) during the summer (Casterlin and Reynolds, 1982; NEFSC, unpublished data). To determine whether depth distributions of flounder changed with body size, we tested for a significant interaction between depth of occurrence and body length in a Generalized Additive Model (GAM). Abundance (no. individuals 10 M 2 of tow) of flounder in 5 mm length classes from 10 to 50 mm SL was calculated for each station and survey and used as the dependent variable in the GAM. The independent variables were length class and station depth (0.5 –15 m) measured with an electronic depth finder and not corrected for tide. The GAM was constructed with a cubic spline smooth [4 degrees of freedom (df)] and a Poisson link function because abundance approximated a Poisson distribution (Hastie and Tibshirani, 1990). Changes in flounder depth distribution with size were also examined by plotting median and quartile depths for each 5 mm length class. Sand shrimp (Crangon septemspinosa) > 30 mm total length (TL) feed on newly settled flounder < 20 mm SL (Bertram and Leggett, 1994; Witting and Able, 1995; Taylor, 2003) and were also abundant in the May collections. As a result, we used the same graphical method to identify changes in the depth distribution of shrimp with body size. 2.2. Evaluation of shallow water predation risk 2.2.1. Collection and maintenance of experimental animals We collected winter flounder with a haul seine and summer flounder with hook and line in the NSBES. Animals were transported to the Howard Marine Sciences Laboratory, Highlands, New Jersey, and maintained in continuous flow-through tanks supplied with seawater from Sandy Hook Bay (temperature, 15– 22 jC; salinity, 19 – 23x). June and July photoperiods were simulated in the laboratories with computerized lighting systems. We provided all holding tanks with 2 –3 cm of washed fine sand

140

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

(mean grain diameter [GD] = 0.14 mm, S.D. = 0.02). Winter flounder were fed chopped clam ad libitum while summer flounder were maintained on a diet of live killifish (Fundulus heteroclitus). 2.2.2. Tethering We tethered winter flounder using 2.24 kg test monofilament line that was threaded through the mouth, out the dorsal operculum and then tied in a loose loop. This method minimized tissue damage and the release of body fluids which can unnaturally increase the vulnerability of tethered prey to predators (Curran and Able, 1998). Tethers were 50 cm long in laboratory experiments and 75 cm long in field experiments. 2.2.3. Laboratory experiments examining tethering artifacts We determined whether flounder mortality was elevated by handling stress associated with tethering by monitoring the survivorship of 10 tethered and 10 untethered fish (30 – 50 mm SL) for 72 h. The fish were held separately in 2 m long  0.6 m wide  0.3 m deep tanks. We examined the effects of tethering on flounder behavior by videotaping individual tethered (N = 10) and untethered (N = 10) fish (30 – 50 mm SL) in glass sided tanks (0.6 m long  0.15 m wide  0.40 m deep) provided with fine sand substrata (GD = 0.14 mm). The fish were introduced to the tanks at 0830 h EST and filmed for 5 min h 1 until sunset (f1900 EST). We measured activity as the proportion of time fish spent moving in the water column or along the substratum. We measured the effect of tethering on the vulnerability of winter flounder to predation by exposing tethered and untethered fish (30– 50 mm SL) to individual summer flounder (310 – 380 mm TL). We selected summer flounder because they are important piscine predators of post-settlement stage winter flounder > 20 mm SL (Manderson et al., 2000). Tethered and untethered prey were offered independently to the predators to evaluate effects on mortality rather than prey selection. A satiated predator was introduced into each of 10, 2.5 m diameter  0.5 m deep tanks provided with 2– 3 cm of fine sand (GD = 0.14 mm). After a 24-h starvation period, the predators were isolated within the tanks in opaque PVC cylinders (0.7 m diameter  0.6 m deep). Five tethered or untethered winter flounder were randomly assigned to the tanks and introduced to areas outside the cylinders. Tethered fish were attached to 20 g washers that were buried in the sand. Predators were released from the cylinders after 1 h and allowed to feed for 24 h. The predators were then removed from the tanks which were drained to recover surviving prey. Five replicates of the tethered and untethered treatments were performed. 2.2.4. Field experiment examining shallow water predation risk 2.2.4.1. Study sites. Tethering and sampling for predators were performed in the lower Navesink River where age-0 winter flounder have been historically abundant during July and August (Fig. 1; Stoner et al., 2001; Goldberg et al., 2002; NEFSC unpublished data). We selected three study sites with medium sand substrata (GD = 0.82 mm, 0.32– 1.53 mm)

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

141

Fig. 1. Location of the Navesink River Sandy Hook Bay estuarine system in the New York Bight apex and aerial photograph of the three study sites in the lower Navesink River where tethering and sampling for predators and prey were performed.

and no complex physical structures. The tidal range in the study area averaged 103 cm (spring tide = 150 cm, neap tide = 70 cm). Each site included a sand bar or beach less than 30 m from a relatively deep habitat (Fig. 1). We maintained depths of V 50 cm in the shallow habitats while the deep habitats averaged 256 cm (S.D. = 126; Table 1). During sampling, we measured depths with a

142

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

Table 1 Mean depths and depth differences (range) in centimeters in the shallow and deep habitats during high and low tide where tethering and trammel netting were performed Site

A B C

Low tide

High tide

Deep

Shallow

Difference

Deep

Shallow

Difference

133 175 356

37 45 50

96 (73 – 118.4) 159 (124 – 194) 305 (257 – 352)

225 209 466

45 50 36

180 (150 – 210) 180 (100 – 260) 430 (406 – 453)

See Fig. 1 for site locations.

meter stick or sounding lead and temperatures (23 – 27 jC) and salinities (20 – 27 x) with an electronic probe. 2.2.4.2. Tethering in the field. We used chronological devices similar to those described by Ha (1996) to measure predator encounters for tethered winter flounder (Fig. 2). Each device consisted of a digital stopwatch with start/stop and reset switches wired to magnetic reed switches. The devices were encased in epoxy resin with steel plates positioned just above the start/stop reed switches. Tethered fish were attached with small swivels to magnets that were placed on the steel plates to start the watches. The watches continued to run until predators attacked tethered fish and pulled the magnets away from

Fig. 2. Chronological device used to measure time to predator encounter for tethered winter flounder.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

143

the steel plates and reed switches. The force necessary to remove the magnets from the plates was 95 g. Before the start of experiments, flounder were tethered in the laboratory and placed in 1-l jars with mesh covers in flow through tanks for a 14-h recovery period. The fish were transported to the field in the jars in a cooler filled with aerated seawater (temperature=20.5 –22.3 jC, salinity=20.3 –25.2x). Just before deployment, each fish was placed with seawater in a 100-ml whirl-pak bag that was twisted closed so that the free end of the tether was accessible. We attached a fish in a bag by its tether to the magnet on each chronological tethering device. The devices (N = 5) were attached at 3m intervals to 3-mm-diameter chain. Two-kilogram anchors and float lines were attached to the ends of each chain. Two chains with a total of 10 tethered fish were positioned in each adjacent shallow and deep habitat at appropriate depths using a meter stick or sounding lead. Since our null hypotheses were designed to test the effect of absolute depth on predation risk, tethered fish were maintained at depths < 50 cm (10 –50 cm) in the shallow habitats during high and low tide. Tethers were set at the greatest possible depth in the deep habitats. Differences in depth between adjacent habitats averaged 186 cm (50 – 357 cm) at low tide and 255 cm (73 –453 cm) at high tide (Table 1). The tethering devices were staked to the substratum by a diver who released the fish from the bags and started the timing mechanisms. The fish remained in the bags < 12 min. After 4 h, we retrieved the devices and recorded the number of remaining fish and watch times for absent fish. We performed tethering at one site per day on 3 consecutive days each week for 4 weeks (July 9 to August 3, 2001). Tethers were deployed 2 h before high or low tide and f1.5 h after sunrise (0.5 –2 h). Thus, we measured predation risk (fmortality) in the habitats during two high and two low tides coinciding with the morning crepuscular period. We used a generalized linear model with a binomial link function to examine the effects of site, tide stage (high vs. low) and depth (shallow vs. deep) on predation risk for tethered flounder. The relationship between predation risk and depth measured as a continuous variable was examined using a GAM with a binomial link function and cubic spline smooth (4 df; Hastie and Tibshirani, 1990). 2.2.4.3. Collections of potentially important predators and available prey. We collected predators and prey in the habitats where tethering was performed with trammel nets (100 m long  2 m deep, 5-cm nylon inner mesh, 36-cm nylon walls) and a 1-m beam trawl (3-mm mesh liner). To avoid disturbing predators during tethering, the gears were fished at each site f24 h after tethering. Thus, predator and prey sampling was also conducted on 3 consecutive days each week for 4 weeks. Trammel nets were set in the adjacent habitats within 2 h of high or low tide and f2.5 h after sunrise and fished for 2 h. After each trammel net was retrieved, a beam trawl tow (speed = 1m s 1, duration = 30 s) was made in the habitat. We measured (mm) and counted all fish and invertebrates. Catch data were standardized as number of individuals collected per hour in trammel nets and 10 m 2 of tow in beam trawls. Stomachs of fish collected in trammel nets were removed and diet items were identified to species, counted, weighed ( F 0.1 g) and measured (mm). Effects of site, habitat depth and tide stage on predator and prey abundance were

144

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

assessed with three-way ANOVAs. Abundances were log transformed [ln(x + 1)] to stabilize variances.

3. Results 3.1. Changes in juvenile winter flounder depth distribution with body size In May, large flounder were more strongly associated with shallow habitats than small fish. Size-dependent changes in depth distribution were indicated by the significant interaction between depth and length class in the generalized additive model (depth  length: df = 1, v2 = 9.94, P = 0.018; depth: df = 1, v2=83.38, P b 0.001; length: df = 1, v2 = 149.83, P b 0.001). Median depths exceeded 2.5 m for recently settled fish < 25 mm SL which were collected over broad depth ranges (Fig. 3). As fish increased in size, the median and range of depth of occurrence gradually decreased. Large fish > 35 mm SL were concentrated in habitats f1 m deep. Sand shrimp >30 mm TL and capable of

Fig. 3. In May, recently settled winter flounder < 25 mm SL were collected over a broader range of deeper depths than larger sized individuals > 35 mm which were associated with shallow habitats. Sand shrimp > 30 mm TL and capable of consuming recently settled flounder, were associated with shallow habitats f2 M deep during the same period. Vertical lines range from 25 to 75th quantiles while dots indicate median depths. The number of individuals in each size class is indicated above the vertical lines.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

145

consuming flounder < 20 mm SL were associated with shallow habitats (f2 m) while median depths for small shrimp exceeded 4 m (Fig 3). 3.2. Shallow water predation risk 3.2.1. Laboratory studies Tethered flounder did not suffer mortality from handling stress, but were less active than untethered individuals. All fish (N = 20) survived the 72-h handling stress experiment. However, tethered fish were significantly less active than untethered fish (Separate variance t test; t = 4.99 df = 11.6, P b 0.001; Fig. 4a). Tethered fish spent most

Fig. 4. In the laboratory (A) tethered winter flounder were less active and spent more time buried or resting on the substratum and (B) when exposed to summer flounder predators showed slightly lower mortality than untethered fish. Means F 1S.E. are plotted [Ten replicates were used in behavioral observations (A), while five replicate of each treatment were performed in the predation experiment (B)].

146

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

of the time buried in the sand (mean proportion of time F S.E.: tethered = 0.91 F 0.04; untethered = 0.32 F 0.10). Tethered fish suffered slightly lower mortality (x¯ = 0.64, S.E. = 0.16) than untethered fish (x¯ = 0.88, S.E. = 0.05) when exposed to summer flounder in the laboratory (Fig. 4b). However, the mortality rates were not significantly different (Separate variance t test; t = 1.63, df = 6, P = 0.15). 3.2.2. Field study 3.2.2.1. Tethering experiment. Mortality for flounder tethered in the field (total N = 239) was 54%. Nearly all of the tethers (98%) were severed within the loops securing the fish indicating that they were probably consumed by predators. However, only 31 survival times were recorded for absent fish because the magnets were frequently not removed from the steel plates on the chronological devices. As a result, we used analysis of survival time to supplement analysis of survival frequencies. At all three study sites and during both high and low tides, predation risk (= mortality) for tethered flounder was lower in shallow (x¯ = 0.29, S.E. = 0.09) than in adjacent deep habitats (x¯ = 0.72, S.E. = 0.04; Fig. 5a,b,c). Depth accounted for the largest proportion of

Fig. 5. At all three study sites (A, B, C; see Fig. 1) and during high and low tides predation risk (= x¯ mortality F 1S.E.) for tethered flounder was lower in shallow than in adjacent deeper water habitats. (D) Predation risk (x¯ F 2S.E. confidence bands) increased nonlinearly with depth to a stable maximum of f0.78 at 175 cm based on the deviance plot derived from the GAM.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

147

explained deviance (fvariance) in the binomial GLM (Table 2a). More fish were consumed at site A than at sites B and C (x¯ F S.E.: A=0.63 F 0.06; B=0.40 F 0.05; C=0.48 F 0.11). Tidal variation in predation risk was inconsistent across depths and sites as evidenced by the significant 3 way interaction (Fig. 5a,b,c; Table 2a). Time to predator encounter for fish absent from tethers was more than twice as long in shallow (x¯ = 150 min, S.E. = 16, N = 12) than deep habitats (x¯=62 min, S.E.=12, N=19), varied with depth, but not tide stage or the interaction of depth and tide (Table 2b). Relative predation risk for flounder increased rapidly from 0.20 at a depth of 0.2 m to a stable maximum of f0.78 at 1.7– 5 m (Fig. 5d). The GAM indicated that the relationship between predation risk and depth was significantly nonlinear (v2 = 15.9; P < 0.001). 3.2.2.2. Distribution of potential predators and prey. Summer flounder (N = 48, 232 – 555 mm TL) and blue crabs (Callinectes sapidus; N = 53, 105 –180 mm CW) were common in trammel net collections, while blue fish (Pomatomus saltatrix; N = 2, 152,185 mm TL) and oyster toadfish (Opsanus tau; N = 1, 310 mm TL) were rare. Summer flounder and blue crab catches were not different among weeks of the experiment (ANOVA, F z 0.92, P z 0.43). Catches of summer flounder varied with depth (three-way ANOVA: MS = 1.31, df = 1, F = 16, P = 0.001) and site (MS = 0.97, df = 2, F = 12, P = 0.001) but not with tide stage and the interactions were not significant (MS V 0.34, F V 4, P z 0.06). At all three study sites, more fish were collected in deep (x¯ = 1.42 F 0.36 fish h 1) than shallow (x¯ = 0.38 F 0.13 fish h 1) habitats (Fig. 6). Catches were highest at site C

Table 2 Depth accounted for most of the explained variation in analyses of (A) predation risk (fmortality) which was lower and (B) survival times which were longer for tethered flounder in shallow than deep habitats (see Fig. 5) (A) Source of variation

df

Site Depth (deep, shallow) Tide stage (high, low) Site  Depth Site  Tide stage Depth  Tide stage Site  Depth  Tide stage Residual Total

2 1 1 2 2 1 2 227 238

(B) Source of variation

df

Depth (deep, shallow) Tide stage (high, low) Depth  Tide stage Error

1 1 1 27

% Total deviance

P value (v2)

2.5 15.3 0.8 1.8 0.6 0.1 2.8 76.1

0.015 b 0.001 0.01 0.05 0.38 0.54 0.01

MS

F

P

51988.6 1708.4 1.6 2769.6

18.8 0.6 0.0

b 0.001 0.44 0.98

Deviance 8.4 5.6 2.7 6.0 1.9 0.4 9.3 252.1 331.3

Model for analysis of mortality frequencies (A) used a binomial link function. Analysis of survival times (B) should be viewed with caution because sample sizes were unequal across sites and thus site was not used as a factor.

148

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

Fig. 6. At all three study sites (A, B, C; see Fig. 1), summer flounder predators were less abundant in shallow than in adjacent deep habitats. In contrast, blue crabs were equally abundant in shallow and deep habitats. Mean abundances F 1S.E. are plotted.

and lowest at site B (A: x¯ = 0.75 F 0.31; B: x¯ = 0.25 F 0.13; C: x¯ = 1.69 F 0.44). Blue crab catches did not vary with depth, tide or site and the interactions were not significant (x¯ = 1.2 crabs h 1 F 0.3; three-way ANOVA: F V 2.79; P z 0.11; Fig. 6). Predation risk was correlated the catch rate of summer flounder (R = 0.44, P=0.03) but not blue crabs (R = 0.38, P = 0.07). Winter flounder (N = 8, 18 – 48 mm SL) occurred in summer flounder stomachs (3 of 23 stomachs with prey). However, Atlantic silverside (Menidia menidia) was the most frequent prey ( F = 65%) and proportionally dominant by weight (46%) and number (48%). Sand shrimp and grass shrimp (Cr. septemspinosa, F = 35%; Palaemonetes spp., F = 26%, respectively) were also common in the predators diet. Bluefish and oyster toadfish stomachs were empty. Sand shrimp, grass shrimp, Atlantic silversides and winter flounder were common in beam trawls (Fig. 7). Sand shrimp were most abundant in shallow habitats (three-way ANOVA, Depth: MS = 17.7, df = 1, F = 5.2, P = 0.041; other main effects and interactions, F < 3.9, P > 0.05). Near times of low water, winter flounder (mean SL = 47 mm, 29 –147; N = 36) were common in the shallows while grass shrimp were somewhat more abundant in deep habitats. However, density variations for these two species as well as Atlantic silversides were not significant ( P > 0.05) in the ANOVAs. Winter flounder density was

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

149

Fig. 7. Trends in the abundance of dominant prey species collected in beam trawls in the shallow and deep habitats during the tethering experiment. Mean abundances F 1S.E. are plotted.

inversely correlated with the catch rate of summer flounder (Spearman’s Rho = 0.03) but not blue crabs (Rho = 0.29, P = 0.08).

0.40, P =

4. Discussion 4.1. Artifacts of tethering Prey vulnerability is a function of the abundance and foraging intensity of predators, the detectability of prey, as well as probabilities of attack and capture following prey detection. We believe that tethering allowed us to assess relative patterns of winter flounder predation risk resulting from variations in the abundance and foraging intensity of predators but probably not the other components of the predator –prey function. We probably eliminated unnatural increases in detectability to olfactory predators (e.g., crabs; see Curran and Able, 1998) by tying flounder to tethers rather than using hooks or needles to puncture the body walls of the animals. Our laboratory experiment indicated that tethering did not increase, but may have slightly reduced detectability to summer flounder. Mortality for tethered and untethered flounder was not statistically different, but tethered fish, which were less active and spent more time buried, were consumed by the predator at

150

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

slightly lower rates. Flounder sewn to tethers also exhibit reduced activity (Curran and Able, 1998), and many predator– prey studies, including one examining interactions of summer flounder with winter flounder, show that predation rates often covary with prey activity (Ware, 1973; Lawler, 1989; Werner and Anholt, 1993; Manderson et al., 2000). We do not know whether tethering affected attack probabilities following prey detection. However, predator capture success may have been elevated even though we used relatively long tethers in the field (75 cm). Although tethering cannot be used to estimate natural mortality, the technique is useful for measuring relative predation risk as long as associated artifacts are consistent across treatments (Peterson and Black, 1994; Aronson and Heck, 1995; Halpin, 2000; Aronson et al., 2001). Tethering artifacts probably did not vary across treatments in our experiment because we intentionally selected immediately adjacent and simple habitats unlikely to produce treatment-specific artifacts including those related to tangling. 4.2. Shallow water predation risk Our tethering experiment indicated that structurally simple shallow habitats f1 m deep serve as predator refuges for juvenile flounder > 30 mm SL. During both high and low tides, probabilities of predator encounter were lower and times to predator encounter were longer for flounder in shallow habitats than immediately adjacent deeper areas. Relative predation risk increased nonlinearly with depth and simple habitats appeared to serve as effective refuges only at depths ] 1 m. Predation risk also increases rapidly with depth for juvenile cod Gadus morhua in simple habitats (Linehan et al., 2001) and is low in shallow water for several round fish and decapod crustaceans (Ca. sapidus, Palaemonetes pugio, F. heteroclitus: Ruiz et al., 1993; Dittel et al., 1995; Hines and Ruiz, 1995; Clark et al., 2003). Results of our collection and dietary analysis of predators indicated that summer flounder were partially responsible for the depth-dependent variation in winter flounder predation risk we measured. The predators were more abundant in deep habitats and winter flounder occurred in their stomachs. Predation risk was also correlated with summer flounder abundance. Following the winter flounder settlement period, summer flounder are important predators of fish > 20 mm SL in the NSBES (Manderson et al., 2000, unpublished data) and other estuarine nursery grounds (Pearcy, 1962; Poole, 1964). Crabs, particularly Carcinus sp., feed on small flatfishes (Pihl, 1990; Fairchild and Howell, 2000) and blue crabs were relatively abundant in our trammel net collections. However, blue crabs were equally abundant in the shallow and deep habitats and partially consumed flounder were never found on tethers at the end of experiments. Thus, blue crabs did not appear to contribute to the depth-dependent variation in predation risk that we measured. Blue crabs rarely consume flounder in the laboratory (Curran and Able, 1998; Phelan, unpublished data) and fish prey are nearly absent in diets of blue crabs collected throughout the estuary in which our experiment was performed (Meise and Stehlik, 2003). Marine birds, particularly cormorants (Phalacrocorax sp.), are important flatfish predators (Birt et al., 1987; Leopold et al., 1998; Hjo¨rleifsson and Pa´lsson, 2001;) and may have contributed to habitat-specific patterns of winter flounder predation risk.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

151

Cormorants were not observed in the study area, but wading birds (snowy and cattle egrets, Egretta thula, Bubulcus ibis; herons, Ardea herodias, Butorides virescens) were present during tethering. Egrets consumed at least two fish in shallow habitats during low tide based on survival times for tethered fish and birds observations. Shallow water foraging in wading birds is generally synchronized with low tide (Burger, 1983; Frederick, 2002) and we believe that wading bird predation had a minimal effect in our experiment because predation risk in shallow habitats did not vary systematically with tide stage. In another study wading bird predation did not elevate mortality rates of juvenile spot (Leiostomus xanthurus) along experimental depth gradients designed to test for the combined effects of fish and birds (Crowder et al., 1997). Although summer flounder were more abundant in deeper water where they appeared to represent a predation threat to winter flounder, other observations indicate that the predators use shallow habitats. Surveys and telemetry studies indicate that the predators may undertake flood tide foraging migrations into marsh creeks where they have been observed at depths < 15 cm (Szedlmayer and Able, 1993; Rountree and Able, 1992, 1997). Summer flounder occur at depths < 10 cm along an artificial depth gradient in our laboratory and preliminary telemetry studies indicate that the predators move from deep water onto adjacent sandbars during the morning crepuscular period in the area where our field experiment was performed (Manderson and Bejda, unpublished data). Summer flounder are clearly not behaviorally or morphologically constrained to deep water. However, the animals may avoid environmental conditions associated with very shallow habitats (e.g., high temperatures and/or light levels) when sufficient prey are available in deeper water. Atlantic silverside, sand shrimp and grass shrimp were dominant prey for summer flounder (see also Powell and Schwartz, 1979; Rountree and Able, 1992; Manderson et al., 2000) and abundant in the deep habitats in our study. Densities of these prey may have been high enough in deep water to eliminate the need for summer flounder to forage in the shallow habitats. The shallow habitats probably provided a spatial rather than a structural refuge for young flounder. The search and capture success of most predators were probably not inhibited substantially in the habitats where depths were z 10 cm and prey shelters were absent. In this study, piscivorus fish were less abundant in the shallow habitats where predator encounter rates were also low based on survival times and frequencies for tethered winter flounder. Probabilities of occurrence for four important fish predators of juvenile winter flounder (Striped searobin, Prionotis evolans; Summer flounder; Weakfish, Cynoscion regalis; and the hakes Urophycis regia and Urophycis chuss) increase with depth based on analyses of system wide trawl surveys of the Navesink River and Sandy Hook Bay (Manderson, unpublished MS). These data suggest that habitat overlap between piscine predators and winter flounder prey is relatively low in shallow habitats and support the hypothesis that small fish use shallows to avoid large predators which are less abundant in them (Ruiz et al., 1993; Linehan et al., 2001). 4.3. Size-dependent changes in winter flounder depth distribution Following larval settlement, juvenile winter flounder became concentrated in shallow habitats as they increased in size. In our analysis, newly settled fish occurred at the greatest

152

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

depths. As fish increased in size, depth of occurrence gradually decreased so that individuals > 35 mm SL were associated with habitats f1 m deep. Young North Sea plaice (Pleuronectes platessa) in Scotland show a remarkably similar inverse relationship between body size and depth (Gibson et al., 2002, see Fig. 7). The change in winter flounder depth distribution with size could have been produced by the high mortality of small fish in deep habitats. Alternatively, the relationship could reflect the onshore migration of fish shortly after larval settlement. Late-stage flounder larvae are associated with bottom waters (Pearcy, 1962; Crawford and Carey, 1985) and settle in large numbers in relatively deep habitats (2– 6 m; Manderson et al., 2003). Larger juveniles are photopositive (McCracken, 1963) and generally most abundant in shallow habitats as are a number of other north Atlantic flatfishes (Gibson, 1973; Riley et al., 1981; Allen and Baltz, 1997; Walsh et al., 1999; Stoner et al., 2001; Amara and Paul, 2003, but see Armstrong, 1997). Movement of winter flounder from deep settlement sites toward shallow shorelines is expected as a result of the photo-positive response as long as stressors that might override that response are absent in shallow water (e.g., temperature, Casterlin and Reynolds, 1982; competition, Armstrong, 1997; wading bird predation). We speculate that changes in requirements for survivorship and perhaps growth during the early juvenile ontogeny of winter flounder could make a shallow water migration advantageous. Newly settled flatfish ] 20 mm SL are vulnerable to crustacean predators, including shrimp of the genus Crangon, which may regulate many flatfish populations (Macer, 1967; Edwards and Steele, 1968; van der Veer and Bergman, 1987; Seikai et al., 1993; Bertram and Leggett, 1994; Witting and Able, 1995; Ansell et al., 1999; van der Veer et al., 2000). In our analysis, sand shrimp (Cr. septemspinosa) capable of consuming winter flounder ] 20 mm SL (Witting and Able, 1995; Taylor, 2003) were associated with shallow habitats where they probably represent a considerable predation threat. Deeper areas were dominated by shrimp too small to feed on newly settled flatfish. Sand shrimp in southern New England show similar changes in depth distribution with body size (Modlin, 1980). As flounder outgrow crustacean predators, they become increasingly vulnerable to piscivorous fishes including summer flounder and striped searobin (Pr. evolans). The feeding efficiencies of these predators are low at small prey sizes but increase to reach asymptotes at relatively large prey sizes (winter flounder k 45 mm SL, Manderson et al., 1999, 2000). Large piscivorous fish are typically more abundant in deeper habitats (Kneib, 1987; McIvor and Odum, 1988; Methven and Schneider, 1998; Paterson and Whitfield, 2000, this study). Thus, as early juvenile winter flounder increase in size, predation risk may shift rapidly from shallow to deep habitats exerting a strong selective pressure for a shallow water migration shortly after settlement. Similar size-dependent changes in predation risk are thought to drive ontogenetic habitat shifts in other juvenile fish (Werner and Gilliam, 1984; Dahlgren and Eggleston, 2000; Halpin, 2000; Bystro¨m et al., 2003), including early juvenile North Sea plaice (Gibson et al., 2002). Winter flounder also undergo rapid changes in sediment preference, burial capability and diet that coincide with the changes in depth distribution that we observed (Pearcy, 1962; Stehlik and Meise, 2000; Phelan et al., 2001). Burial capabilities increase with size and fish ~ 40 mm SL prefer courser sediments (Phelan et al., 2001) typical of shallow and intertidal areas that can serve as important juvenile feeding grounds (Tyler, 1971; Wells et al., 1973). Conceivably, ontogenetic changes in feeding requirements could reinforce selective pressures for

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

153

shallow water migration that are associated with size-dependent changes in habitat-specific predation risk and increases in the range of soft bottom habitats that young flounder find burial refuge. 4.4. Conclusions Our results indicate that juvenile winter flounder emigrate from or suffer higher mortality in deep water settlement areas to become concentrated in shallow habitats that serve as effective refuges even when they lack complex physical structures. Many small fish and invertebrates are most abundant in shallow areas that comprise a considerable amount of habitat area available in coastal bays and estuaries. Shallow habitats, defined from subtidal depths V 1 m at mean low water to mean high water, account for an average of 35% of total surface area available in north east Atlantic bays and estuaries and can exceed 80% in bar built estuaries based on analysis of NOS digital elevation models (Table 3). Natural fluctuations in environmental factors (e.g., high summer temperature, low dissolved oxygen) that temporarily eliminate shallow habitats by forcing juveniles into deeper waters where predation risk is high could reduce year class strength in local populations during some years. Human activities such as dredging and bulkheading could

Table 3 Shallow habitats (depth V 1 m at mean low water to mean high water) account for an average of 35% of the total habitat area available in East Coast US bays and estuaries based on analysis of National Ocean Service Bathymetric Digital Elevation Models (30 m resolution; see http://seaserver.nos.noaa.gov/bathy/) across the geographic range of winter flounder Coastal Bay or Estuary

Chesapeake Bay, Virginia/Maryland Delaware Bay, Delaware/New Jersey New Jersey Coastal Bays, New Jerseya Barnegat Bay, New Jerseya Raritan Bay, New Jersey Navesink River/Sandy Hook Bay, New Jerseyb Long Island Sound, New York/Connecticut Great South Bay, New Yorka Gardiners Bay, New York Narragansett bay, Rhode Island Buzzards Bay, Massachusetts Cape Cod Bay, Massachusetts Massachusetts Bay, Massachusetts Great Bay, New Hampshire Casco Bay, Maine Penobscot Bay, Maine Muscongus Bay, Maine Sheepscot Bay, Maine Saco Bay, Maine a b

Bar built estuaries. Area where data collected in present study.

Maximum depth (m)

Surface area estimate (hectares) Total

Shallow habitat

50 47 16 14 34 15 113 23 31 59 45 59 71 25 72 165 70 81 32

1,119,370 191,548 27,231 17,682 42,231 3747 325,638 34,674 51,215 40,822 63,695 143,537 93,819 4674 42,434 97,644 20,084 7715 4739

313,102 34,083 23,684 16,126 12,083 1488 33,395 27,812 11,311 7144 7265 16,881 10,029 3111 10,039 19,476 4683 1688 1135

% Shallow habitat 28 18 87 91 29 40 10 80 22 17 11 12 11 67 24 28 23 22 24

154

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

permanently reduce population sizes of species dependent on shallow habitats as predation refuges. Acknowledgements Special thanks to Jessica Strehlner, Fred Scharf and Linda Stehlik for thier help in the field. We also thank Don Kroodsma, Tom Noji, John Boreman, Mary Fabrizio and anonymous reviewers for their valuable comments on earlier drafts of the manuscript. [SS] References Adams, A.J., Ebersole, J.P., 2002. Use of back-reef and lagoon habitats by coral reef fishes. Mar. Ecol. Prog. Ser. 228, 213 – 226. Allen, F.L., Baltz, D.M., 1997. Distribution and microhabitat use by flatfishes in a Louisiana estuary. Environ. Biol. Fisches 50, 85 – 103. Amara, R., Paul, C., 2003. Seasonal patterns in the fish and epibenthic crustaceans community of an intertidal zone with particular reference to the population dynamics of plaice and brown shrimp. Estuar. Coast. Shelf Sci. 56, 807 – 818. Ansell, A.D., Comely, C.A., Robb, L., 1999. Distribution, movements and diet of macrocrustaceans on a Scottish sandy beach with particular reference to predation on juvenile fishes. Mar. Ecol. Prog. Ser. 176, 115 – 130. Armstrong, M.P., 1997. Seasonal and ontogenetic changes in distribution and abundance of smooth flounder Pleuronectes putnami, and winter flounder, Pleuronectes americanus, along estuarine depth and salinity gradients. Fish. Bull. US 95, 414 – 430. Aronson, R.B., Heck, K.L.J., 1995. Tethering experiments and hypothesis testing in ecology. Mar. Ecol. Prog. Ser. 121, 307 – 309. Aronson, R.B., Heck, K.L.J., Valentine, J.V., 2001. Measuring predation with tethering experiments. Mar. Ecol. Prog. Ser. 214, 311 – 312. Bertram, D.F., Leggett, W.C., 1994. Predation risk during the early life history periods of fishes: separating the effects of size and age. Mar. Ecol. Prog. Ser. 109, 105 – 114. Birt, V.L., Birt, T.P., Goulet, D., Cairns, D.K., Montevecchi, W.A., 1987. Ashmole’s halo: direct evidence for prey depletion by a seabird. Mar. Ecol. Prog. Ser. 40, 205 – 208. Blaber, S.J.M., Brewer, D.T., Salini, J.P., 1995. Fish communities and the nursery role of the shallow inshore waters of a tropical bay in the Gulf of Carpentaria, Australia. Estuar. Coast. Shelf Sci. 40, 177 – 193. Black, R., Miller, R.J., 1991. Use of the intertidal zone by fish in Nova Scotia. Environ. Biol. Fisches 31, 109 – 121. Burger, J., 1983. Jamaica Bay studies: 2. Effect of tidal, temporal and weather variables on distribution of ibises, egrets and herons on a coastal estuary. Acta Oecol. 4, 289 – 297. Bystro¨m, P., Persson, L., Wahlstro¨m, E., Westman, E., 2003. Size- and density-dependent habitat use in predators: consequences for habitat shifts in young fish. J. Anim. Ecol. 72, 156 – 168. Casterlin, M.A., Reynolds, W.W., 1982. Thermoregulatory behavior and diel activity of yearling winter flounder, Pseudopleuronectes americanus (Walbaum). Environ. Biol. Fisches 7, 177 – 180. Clark, K.L., Ruiz, G.M., Hines, A.H., 2003. Diel variation in predator abundance, predation risk and prey distribution in shallow-water estuarine habitats. J. Exp. Mar. Biol. Ecol. 287, 37 – 55. Crawford, R.E., Carey, C.G., 1985. Retention of winter flounder larvae within a Rhode Island salt pond. Estuaries 8, 217 – 227. Crowder, L.B., Squires, D.D., Rice, J.A., 1997. Nonadditive effects of terrestrial and aquatic predators on juvenile estuarine fish. Ecology 78, 1796 – 1804. Curran, M.C., Able, K.W., 1998. The value of tethering fishes (winter flounder and tautog) as a tool for assessing predation rates. Mar. Ecol. Prog. Ser. 163, 45 – 51. Dahlgren, C.P., Eggleston, D.B., 2000. Ecological processes underlying ontogenetic habitat shifts in a coral reef fish. Ecology 81, 2227 – 2240.

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

155

Dittel, A.I., Hines, A.H., Ruiz, G.M., Ruffin, K.K., 1995. Effects of shallow water refuge on behavior and density-dependent mortality of juvenile blue crabs in Chesapeake Bay. Bull. Mar. Sci. 57, 902 – 916. Edwards, R.R.C., Steele, J.H., 1968. The ecology of 0-group plaice and dabs at Loch Ewe: I. Population and food. J. Exp. Mar. Biol. Ecol. 2, 215 – 238. Fairchild, E.A., Howell, W.H., 2000. Predator – prey size relationship between Pseudopleuronectes americanus and Carcinus maenas. J. Sea Res. 44, 81 – 90. Frederick, P.C., 2002. Wading birds in the marine environment. In: Scheiber, E.A., Burger, J. (Eds.), Biology of Marine Birds. CRC Press, LLC, Boca Raton, FL, pp. 617 – 655. Gibson, R.N., 1973. Intertidal movements and distribution of young fishes on a sandy beach with special reference to the plaice (Pleuronectes platessa L.). J. Exp. Mar. Biol. Ecol. 12, 79 – 102. Gibson, R.N., 1997. Behavior and the distribution of flatfishes. J. Sea Res. 37, 241 – 256. Gibson, R.N., Pihl, L., Burrows, M.T., Modin, J., Wennhage, H., Nickell, L.A., 1998. Diel movements of juvenile plaice Pleuronectes platessa in relation to predators, competitors, food availability and abiotic factors on a microtidal nursery ground. Mar. Ecol. Prog. Ser., 165. Gibson, R.N., Robb, L., Wennhage, H., Burrows, M.T., 2002. Ontogenetic changes in depth distribution of juvenile flatfishes in relation to predation risk and temperature on a shallow-water nursery ground. Mar. Ecol. Prog. Ser. 229, 233 – 244. Goldberg, R., Phelan, B., Pereira, J., Hagan, S., Clark, P., Bejda, A., Calabrese, A., Studholme, A., Able, K.W., 2002. Variability in habitat use by young-of-the-year winter flounder, Pseudopleuronectes americanus, in three northeastern U.S. estuaries. Estuaries 25, 215 – 226. Ha, D.S., 1996. An investigation of timed-tethering: a method for study of coral reef piscivory. MS thesis Dept. Biol., University of Windsor, Windsor, Ontario, Canada, pp. 80. Halpin, P.M., 2000. Habitat use by an intertidal salt – marsh fish: trade-offs between predation and growth. Mar. Ecol. Prog. Ser. 198, 203 – 214. Hastie, T.J., Tibshirani, R.J., 1990. Generalized Additive Models. Chapman & Hall, London. XI-329 pp. Hines, A.H., Ruiz, G.M., 1995. Temporal variation in juvenile blue crab mortality: nearshore shallows and cannibalism in Chesapeake Bay. Bull. Mar. Sci. 57, 884 – 901. Hjo¨rleifsson, E., Pa´lsson, J., 2001. Settlement, growth and mortality of 0-group plaice (Pleuronectes platessa) in Icelandic waters. Neth. J. Sea Res. 45, 321 – 324. Holbrook, S.J., Schmitt, R.J., 2002. Competition for shelter space causes density dependent predation mortality in damselfishes. Ecology 83, 2855 – 2868. Howell, P.T., Molnar, D.R., Harris, R.B., 1999. Juvenile winter flounder distribution by habitat type. Estuaries 22, 1090 – 1098. Kneib, R.T., 1987. Predation risk and use of intertidal haitats by young fish and shrimp. Ecology 68, 379 – 386. Lawler, S.P., 1989. Behavioral responses to predators and predation risk in four species of larval anurans. Anim. Behav. 38, 1039 – 1047. Leopold, M.F., van Damme, C.J.G., van der Veer, H.W., 1998. Diet of cormorants and the impact of cormorant predation on juvenile flatfish in the Dutch Wadden Sea. J. Sea Res. 40, 93 – 97. Linehan, J.E., Gregory, R.S., Schneider, D.C., 2001. Predation risk of age-0 cod (Gadus morhua) relative to depth and substrate in coastal waters. J. Exp. Mar. Biol. Ecol. 263, 25 – 44. Macer, C.T., 1967. The food web in Red Wharf Bay (North Wales) with particular reference to young plaice (Pleuronectes platessa). Helgol. Mar. Res. 15, 560 – 573. Macpherson, E., Duarte, C.M., 1991. Bathymetric trends in demersal fish size: is there a general relationship? Mar. Ecol. Prog. Ser. 71, 103 – 112. Manderson, J.P., Phelan, B.A., Bejda, A.J., Stehlik, L.L., Stoner, A.W., 1999. Predation by striped searobin (Prionotis evolans, Triglidae) on young-of-the-year winter flounder (Pseudopleuronectes americanus, Walbaum): examining prey size selection and prey choice using field observations and laboratory experiments. J. Exp. Mar. Biol. Ecol. 242, 211 – 231. Manderson, J.P., Phelan, B.A., Stoner, A.W., Hilbert, J., 2000. Predator – prey relations between age 1+ summer flounder (Paralichthys dentatus, Linnaeus) and age-0 winter flounder (Pseudopleuronectes americanus, Walbaum): predator diets, prey selection, and effects of sediments and macrophytes. J. Exp. Mar. Biol. Ecol. 251, 17 – 39. Manderson, J.P., Pessutti, J., Meise, C., Johnson, D., Shaheen, P., 2003. Winter flounder settlement dynamics and

156

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

the modification of settlement patterns by postsettlement processes in a north west Atlantic estuary. Mar. Ecol. Prog. Ser. 253, 253 – 267. McCracken, F.D., 1963. Seasonal movements of winter flounder, Pseudopleuronectes americanus (Walbaum), on the Atlantic Coast. J. Fish. Res. Board Can., 20. McIvor, C.C., Odum, W.E., 1988. Food, predation risk, and microhabitat selection in a marsh fish assemblage. Ecology 69, 1341 – 1351. Meise, C.J., Stehlik, L.L., 2003. Habitat use, temporal abundance variability, and diet of blue crabs from a New Jersey estuarine system. Estuaries 26, 731 – 745. Methven, D.A., Schneider, D.C., 1998. Gear-independent pattaerns of variation in catch of juvenile Atlantic cod (Gadus morhua) in coastal habitats. Can. J. Fish. Aquat. Sci. 55, 1430 – 1442. Modlin, R.F., 1980. The life cycle and recruitment of the sand shrimp, Crangon septemspinosa, in the Mystic River Estuary, Connecticut. Connecticut Estuaries 3, 1 – 10. Nagelkerken, I., Roberts, C.M., van der Velde, G., Dorenbosch, M., van Riel, M.C., Cocheret de la Morinie`re, E., Nienhuis, P.H., 2002. How important are mangroves and seagrass beds for coral-reef fish? The nursery hypothesis tested on an island scale. Mar. Ecol. Prog. Ser. 244, 299 – 305. Norcross, B.L., Holladay, B.A., Muter, F.-J., 1995. Nursery area characteristics of pleuronectids in coastal Alaska, USA. Neth. J. Sea Res. 34, 161 – 175. Norcross, B.L., Blanchard, A., Holladay, B.A., 1999. Comparison of models for defining nearshore flatfish nursery areas in Alaskan waters. Fish. Oceanogr. Transl. 8, 50 – 67. Paterson, A.W., Whitfield, A.K., 2000. Do shallow-water habitats function as refugia for juvenile fishes. Estuar. Coast. Shelf Sci. 51, 359 – 364. Pearcy, W.G., 1962. Ecology of an estuarine population of winter flounder Pseudopleuronectes americanus (Walbaum) Parts I – IV. Bull. Bingham Oceanogr. Collect. Yale Univ. 18, 1 – 77. Peterson, B.J., Thompson, K.R., Cowan, J.H. Jr., Heck, K.L.J. Jr., 2002. Comparison of predation pressure in temperate and subtropical seagrass habitats based on chronographic tethering. Mar. Ecol. Prog. Ser. 224, 77 – 85. Peterson, C.H., Black, R., 1994. An experimentalist’s challenge: when artifacts of intervention interact with treatments. Mar. Ecol. Prog. Ser. 111, 289 – 297. Phelan, B.A., Manderson, J.P., Stoner, A.W., Bejda, A.J., 2001. Size-related shifts in the habitat associations of young-of-the-year winter flounder (Pseudopleuronectes americanus): field observations and laboratory experiments with sediments and prey. J. Exp. Mar. Biol. Ecol. 257, 297 – 315. Pihl, L., 1990. Year class strength regulation in plaice (Pleuronectes platessa L.) on the Swedish west coast. Hydrobiologia 195, 79 – 88. Poole, J.C., 1964. Feeding habits of summer flounder in Great South Bay. N.Y. Fish Game J. 11, 28 – 34. Powell, A.B., Schwartz, F.J., 1979. Food of Paralichthys dentatus and P. lethostigma (Pisces: Bothidae) in North Carolina. Estuaries 2, 276 – 279. Riley, J.D., Symonds, D.J., Woolner, L., 1981. On the factors influencing the distribution of 0-group demersal fish in coastal waters. Rapp. P.-V. Reun.-Cons. Int. Explor. Mer 178, 223 – 228. Rountree, R.A., Able, K.W., 1992. Foraging habits, growth, and temporal patterns of salt – marsh creek habitat use by young-of-the-year summer flounder in New Jersey. Trans. Am. Fish. Soc. 121, 765 – 776. Rountree, R.A., Able, K.W., 1997. Nocturnal fish use of New Jersey marsh creek and adjacent bay shoal habitats. Estuar. Coast. Shelf Sci. 44, 703 – 711. Ruiz, G.M., Hines, A.H., Posey, M.H., 1993. Shallow water as a refuge habitat for fish and crustaceans in nonvegetated estuaries: an example from Chesapeake Bay. Mar. Ecol. Prog. Ser. 99, 1 – 16. Seikai, T., Kinoshita, I., Tanaka, M., 1993. Predation by crangonid shrimp on juvenile Japanese flounder under laboratory conditions. Nippon Suisan Gakkaishi 59, 321 – 326. Sheaves, M., 2001. Are there really few piscivorous fishes in shallow estuarine habitats? Mar. Ecol., Prog. Ser. 222, 279 – 290. Stehlik, L.L., Meise, C.J., 2000. Diet of winter flounder in a New Jersey Estuary: ontogenetic change and spatial variation. Estuaries 23, 381 – 391. Stoner, A.W., 1982. The influence of benthic macrophytes in the foraging behavior of pinfish, Lagodon rhomboides (Linneaus). J. Exp. Mar. Biol. Ecol. 58, 271 – 284. Stoner, A.W., Manderson, J.P., Pessutti, J., 2001. Spatially-explicit analysis of estuarine habitat for juvenile

J.P. Manderson et al. / J. Exp. Mar. Biol. Ecol. 304 (2004) 137–157

157

winter flounder (Pseudopleuronectes americanus): combining generalized additive models and geographic information systems. Mar. Ecol. Prog. Ser. 213, 253 – 271. Szedlmayer, S.T., Able, K.W., 1993. Ultrasonic telemetry of age-0 summer flounder, Paralichthys dentatus, movements in a southern New Jersey estuary. Copeia 1993, 728 – 736. Taylor, D.L., 2003. Size-dependent predation on post-settlement winter flounder Pseudopleuronectes americanus by sand shrimp Crangon septemspinosa. Mar. Ecol. Prog. Ser. 263, 197 – 215. Tupper, M., Boutilier, R.G., 1997. Effects of habitat on settlement, growth, predation risk and survival of a temperate marine fish. Mar. Ecol. Prog. Ser. 151, 225 – 236. Tyler, A.V., 1971. Surges of winter flounder, Pseudopleuronectes americanus, into the intertidal zone. J. Fish. Res. Board Can. 28, 1727 – 1732. van der Veer, H.W., Bergman, M.J.N., 1987. Predation by crustaceans on a newly settled 0-group plaice Pleuronectes platessa population in the western Wadden Sea. Mar. Ecol. Prog. Ser. 35, 203 – 215. van der Veer, H.W., Geffen, A.J., Witte, J.I.J., 2000. Exceptionally strong year classes in plaice Pleuronectes platessa: are they generated during the pelagic stage only, or also in the juvenile stage. Mar. Ecol. Prog. Ser. 199, 255 – 262. Walsh, H.J., Peters, D.S., Cyrus, D.P., 1999. Habitat utilization by small flatfish in a North Carolina estuary. Estuaries 22, 803 – 813. Ware, D.M., 1973. Risk of epibenthic prey to predation bay rainbow trout (Salmo gairdneri). J. Fish. Res. Board Can. 30, 787 – 797. Wells, B., Steele, D.H., Tyler, A.V., 1973. Intertidal feeding of winter flounders (Pseudopleuronectes americanus) in the Bay of Fundy. J. Fish. Res. Board Can. 30, 1374 – 1378. Werner, E.E., Anholt, B.R., 1993. Ecological consequences of the trade off between growth and mortality rates mediated by foraging activity. Am. Nat. 142, 242 – 272. Werner, E.E., Gilliam, J.F., 1984. The ontogenetic niche and species interactions in size-structured populations. Ann. Rev. Ecolog. Syst. 15, 393 – 425. Wilson, K.A., Able, K.W., Heck, K.L.J., 1990. Predation rates on juvenile blue crabs in estuarine nursery habitats: evidence for the importance of macroalgae (Ulva lactuca). Mar. Ecol. Prog. Ser. 58, 243 – 251. Witting, D.A., Able, K.W., 1995. Predation by sevenspine bay shrimp Crangon septemspinosa on winter flounder Pleuronectes americanus during settlement: laboratory observations. Mar. Ecol. Prog. Ser. 123, 23 – 31.