Spatial and Temporal Consumption Dynamics of ...

Transactions of the American Fisheries Society

ISSN: 0002-8487 (Print) 1548-8659 (Online) Journal homepage: http://www.tandfonline.com/loi/utaf20

Spatial and Temporal Consumption Dynamics of Trout in Catch-and-Release Areas in Arkansas Tailwaters Jon M. Flinders & Daniel D. Magoulick To cite this article: Jon M. Flinders & Daniel D. Magoulick (2017) Spatial and Temporal Consumption Dynamics of Trout in Catch-and-Release Areas in Arkansas Tailwaters, Transactions of the American Fisheries Society, 146:3, 432-449, DOI: 10.1080/00028487.2017.1281169 To link to this article: http://dx.doi.org/10.1080/00028487.2017.1281169

Published online: 21 Mar 2017.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=utaf20 Download by: [University of Arkansas Libraries - Fayetteville]

Date: 22 March 2017, At: 08:46

Transactions of the American Fisheries Society 146:432–449, 2017 © American Fisheries Society 2017 ISSN: 0002-8487 print / 1548-8659 online DOI: 10.1080/00028487.2017.1281169

ARTICLE

Spatial and Temporal Consumption Dynamics of Trout in Catch-and-Release Areas in Arkansas Tailwaters Jon M. Flinders*1 Department of Biological Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA

Daniel D. Magoulick U.S. Geological Survey, Arkansas Cooperative Fish and Wildlife Research Unit, Department of Biological Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA

Abstract

Restrictive angling regulations in tailwater trout fisheries may be unsuccessful if food availability limits energy for fish to grow. We examined spatial and temporal variation in energy intake and growth in populations of Brown Trout Salmo trutta and Rainbow Trout Oncorhynchus mykiss within three catch-and-release (C-R) areas in Arkansas tailwaters to evaluate food availability compared with consumption. Based on bioenergetic simulations, Rainbow Trout fed at submaintenance levels in both size-classes (≤400 mm TL, >400 mm TL) throughout most seasons. A particular bottleneck in food availability occurred in the winter for Rainbow Trout when the daily ration was substantially below the minimum required for maintenance, despite reduced metabolic costs associated with lower water temperatures. Rainbow Trout growth rates followed a similar pattern to consumption with negative growth rates during the winter periods. All three size-classes (400 mm TL) of Brown Trout experienced high growth rates and limited temporal bottlenecks in food availability. We observed higher mean densities for Rainbow Trout (47–342 fish/ha) than for Brown Trout (3–84 fish/ha) in all C-R areas. Lower densities of Brown Trout coupled with an ontogenetic shift towards piscivory may have allowed for higher growth rates and sufficient consumption rates to meet energetic demands. Brown Trout at current densities were more effective in maintaining adequate growth rates and larger sizes in C-R areas than were Rainbow Trout. Bioenergetic simulations suggest that reducing stocking levels of Rainbow Trout in the tailwaters may be necessary in order to achieve increased catch rates of larger trout in the C-R areas.

Catch-and-release (C-R) regulations have become readily adopted in many waters as a fisheries management tool for a diverse array of fishes (Arlinghaus et al. 2007). If properly applied, C-R regulations can reduce angling mortality and lead to increased residence times of fish and higher densities of larger fish (Anderson and Nehring 1984; Carline et al. 1991; Lucy and Studholme 2002). Implicit in C-R regulations are the assumptions that the fish will survive and grow in the absence of harvest and fish in river systems will remain within the section of designated C-R restrictions (i.e., limited movement) (Wydoski 1977; Schill et al. 1986; Muoneke and Childress 1994; High and Meyer 2009). Despite the rapid incorporation of C-R regulations into many salmonid fisheries management programs, limited data often exist

to evaluate the biological success (i.e., increased survival) of such regulations on fish populations (Matlock 2002; Cooke and Schramm 2007). Studies examining C-R regulations have addressed factors that affect immediate or delayed mortality rates in fish populations (Muoneke and Childress 1994; Pollock and Pine 2007), sublethal effects (Meka and Margraf 2007; Pope et al. 2007), and effects of fish released at elevated temperatures (Boyd et al. 2010; Havn et al. 2015). A component lacking in many C-R studies is the evaluation of growth and production of fish populations in response to increased density-dependent factors from a food availability perspective (Arlinghaus et al. 2007; Cooke and Schramm 2007). Growth is a function of food availability (i.e., proportion of potential prey detected, captured, and

*Corresponding author: jon.fl[email protected] 1 Present address: Idaho Department of Fish and Game, 4279 Commerce Circle, Idaho Falls, Idaho 83401, USA. Received August 31, 2016; accepted January 9, 2017

432

CONSUMPTION DYNAMICS OF TROUT

consumed), temperature-dependent metabolic costs (e.g., activity energy, food processing), and the assimilation efficiency of the food (Fausch 1984). Thus, a decrease in food availability affects growth and therefore population size structure. Food limitation can occur in both regulated rivers (Filbert and Hawkins 1995; Weiland and Hayward 1997) and unregulated streams (Cada et al. 1987; Ensign et al. 1990; Huryn 1996). In Arkansas tailwaters, nonnative Rainbow Trout Oncorhynchus mykiss and Brown Trout Salmo trutta provide economically important fisheries and often experience high fishing pressure (>1,000 angler-hours per hectare annually) (Bowman et al. 1996). The use of C-R regulations in rivers and streams that receive high fishing pressure have been effective in sustaining high numbers of large trout and higher catch rates (EngstromHeg 1981; Anderson and Nehring 1984; Carline et al. 1991). Several C-R areas were implemented in Arkansas tailwaters in 1995 with the goal of providing increased catch rates of larger trout. Tailwaters may be particularly limited in food resources (e.g., Gammarus, Lirceus) for larger trout due to the lack of larger food items available in the drift (Dodrill et al. 2016). As fish size increases the time required for foraging drift organisms increases resulting in larger fish selecting larger prey to offset the energetic costs of increased foraging (Keeley and Grant 1997; Hayes et al. 2000). Increasing the density and size of trout in tailwaters may result in limited growth, decline in average size, and a reduction in the forage base (Filbert and Hawkins 1995; Weiland and Hayward 1997; McKinney and Speas 2001; Yard et al. 2016). To estimate the consumption required to satisfy growth, bioenergetics models are a commonly used tool (Kitchell et al. 1977) and are ideal for addressing potential food limitation within special regulation C-R areas. Individual consumption rates can be compared with maintenance rations to examine potential spatial–temporal bottlenecks in prey supply (Utz and Hartman 2006). The objectives of this study were to determine whether (1) food supply regulated the growth of any size-classes of Brown Trout and Rainbow Trout and (2) estimated levels of consumption in Brown Trout and Rainbow Trout were sufficient to meet their metabolic demands. Assessing food availability in C-R areas may assist managers in developing stocking strategies in the tailwaters aimed at achieving trout densities sufficient to provide adequate growth rates and ultimately providing increased catch rates of larger trout.

METHODS Study site.—The study was conducted on Bull Shoals and Norfork tailwaters in the Ozark Highlands of Arkansas. Bull Shoals tailwater on the White River is located in Marion and Baxter counties, Arkansas (36°21′N, 92°34′W), and Norfork tailwater is located on a tributary of the White River in Baxter County, Arkansas (36°14′N, 92°14′W) (Figure 1). Bull Shoals tailwater supports a trout fishery for approximately 164 km downstream from Bull Shoals Dam. Norfork tailwater supports trout for approximately 7 km from the Norfork Dam until the

433

FIGURE 1. The tailwater catch-and-release (C-R) areas (shown in gray) below Bull Shoals and Norfork reservoirs, Arkansas.

confluence of the tailwater with the White River. Annual water temperatures in Bull Shoals (mean = 10.1°C, range = 7.4–13.8°C) and Norfork (mean = 11.6°C, range = 6.1–14.8°C) C-R areas were fairly similar and were much cooler than water temperatures observed in Sylamore (mean = 15.1°C, range = 4.3–23.2°C) in 2005 and 2006 (Figure 2). Water discharges from the Bull Shoals Dam during this study averaged 50.5 m3/s (SE, 2.84) and ranged from 1.4 to 230.4 m3/s, and releases from Norfork Dam averaged 28.5 m3/s (SE, 1.12) and ranged from 1.7 to 122.0 m3/s (U.S. Army Corps of Engineers, unpublished data). The amount of discharge was lowest in the winter, while the highest discharge occurred in the spring. Alternating shoal and pool areas characterize the study reaches. Substrate was mostly gravel, with some bedrock in hydraulically scoured areas to sand and silt in pools. Filamentous algae, Cladophora, was found attached to the substrate in both tailwaters, and a nuisance diatom, Didymosphenia geminata, was also present in high abundance in Bull Shoals C-R area and often formed thick, mucilaginous mats covering the substrate. Bull Shoals Dam C-R area begins 0.09 km below Bull Shoals Dam extending downstream 1.5 km, and the surface area is approximately 22.0 ha. Sylamore C-R area is located approximately 124 km downstream from Bull Shoals Dam. Sylamore C-R area is 4.1 km long and has a surface area of 60.3

434

FLINDERS AND MAGOULICK

FIGURE 2. Mean monthly water temperatures (°C) in Bull Shoals, Norfork, and Sylamore C-R areas from January to December in 2005 and 2006.

ha. Norfork C-R area is located approximately 4 km downstream from the dam. Norfork C-R area was 1.8 km long with a surface area of 11.2 ha. Catch-and-release trout fishing regulations were implemented by the Arkansas Game and Fish Commission (AGFC) on January 1, 1995, at the Bull Shoals, Norfork, and Sylamore catch-and-release (C-R) areas. All trout caught in C-R areas must be released immediately, and tackle is restricted to the use of one artificial lure with a single, barbless, hooking point. While the C-R areas were not directly stocked, the surrounding areas were highly augmented by a put-and-take fishery for Rainbow Trout (~279 mm TL at stocking) and a put-grow-and-take fishery for Brown Trout (~150 mm TL at stocking). Rainbow Trout were stocked year round, whereas Brown Trout were only stocked in the fall and winter. Approximately 1.18 million and 92,000 Rainbow Trout were stocked annually at Bull Shoals and Norfork tailwaters, respectively (AGFC, unpublished data). Annual stocking of Brown Trout in 2005 and 2006 was approximately 60,000 in Bull Shoals and 10,000 in Norfork tailwaters. Cutthroat Trout O. clarkii and Brook Trout Salvelinus fontinalis were also stocked in low numbers within both tailwaters. Reproductive success has been highly variable among Arkansas tailwater reaches, and as a result trout populations are primarily maintained by intensive stocking programs (Bowman et al. 1996; Pender and Kwak 2002). Abundance and density of trout.—Sampling was conducted on a seasonal basis at Bull Shoals and Norfork C-R areas from May 2005 to June 2006. Sylamore C-R area was sampled seasonally from October 2005 to October 2006; however, no sampling was conducted in the summer of 2006 at Sylamore due to high water releases from Bull Shoals and Norfork dams. Seasons were spring (April–June), summer (July–September), fall (October–December), and winter (January–March). On each sampling date, trout were collected at night using two crews, one using an electrofishing boat and the other a

processing boat. The fiberglass electrofishing boats were equipped with Smith-Root 5.0 GPP electrofishing units and boom-mounted steel cable electrotrodes. Settings for the GPP unit were as follows: mode = DC, voltage = high range (50–1,000 V), pulses per second = 30, amperage ≈ 2.0–2.5 A. All sampling was conducted on two consecutive nights at low flows during periods of no power generation. We marked fish on the first night and recaptured fish on the second night of sampling for the abundance estimates. Boat electrofishing started at the upstream end of the C-R area and proceeded downstream to the lower end of the C-R area. At the end of a sampling run, all trout collected were transferred from live wells on the electrofishing boat to live wells on the processing boat. On the first night of sampling, all Brown Trout and Rainbow Trout were anesthetized with a clove oil mixture (1:10 clove oil : ethanol) at 10 mL solution per 20 L water (Prince and Powell 2000), measured for TL, and weighed to the nearest 0.1 g wet weight. Fish were then tagged below the dorsal fin with individually numbered yellow Hallprint TBA T-bar anchor tags (2 in, TL; 11/4 in, colored; Hallprint, Holden Hill, South Australia, Australia), and released. On the second night Brown Trout and Rainbow Trout collected were measured, weighed, checked for tags, and released. Subsamples of untagged trout from each species were euthanized with a concussive blow to the cranium for gut content analysis (GCA). The stomachs were removed and placed in a 10% buffered formalin solution. Two size-classes of Rainbow Trout and three sizes of Brown Trout were chosen for GCA based on size-frequency data (Stan Todd, AGFC, unpublished data). Attempts were made to collect 60 Brown Trout from small (400 mm TL, n = 20) sizeclasses and 60 Rainbow Trout from small (≤400 mm TL, n = 40) and large (>400 mm TL, n = 20) size-classes at each site per season. A Peterson single mark–recapture population estimate with the Chapman modification was used to estimate trout abundance in the catch-and-release areas (Ricker 1975): ^ ¼ ðn1 þ 1Þðn2 þ 1Þ 1; N ðm2 þ 1Þ where n1 = number caught and marked in the first sampling period, n2 = number caught in the second sampling period, and m2 = number of marked fish in the second sampling period. We converted abundance estimates to standard area units by dividing the estimates by wetted area (ha) at base flow per site to provide density estimates. Growth rates.—We calculated growth from the change in length of tagged individuals recaptured from seasonal population estimate surveys. Only fish collected at the beginning of a seasonal tagging interval (e.g., spring–summer) were used for instantaneous daily rate of growth estimates for that season. Growth rates were estimated for each season, and an average of the seasonal changes in lengths was used to estimate growth rates per year for each size-

435

CONSUMPTION DYNAMICS OF TROUT

class. Specific growth rates, G, were estimated seasonally using the mark–recapture data for each site, species, and size-class using the following growth model: G ¼ ðln Wt ln W0 Þ=Δt; where G is the mean daily growth, Wt is the final weight, W0 is the intial weight, and Δt is the growth period between recaptures (Jensen 1990). Instantaneous daily and annual growth rates were estimated across four tagging intervals at Bull Shoals from May 2005 to May 2006 and at Norfork from June 2005 to June 2006. At Sylamore, instantaneous daily and annual growth rates were estimated across three tagging intervals from October 2005 to October 2006. While fish tagging was conducted from October 2005 to 2006 at Sylamore, after May 2006 no Rainbow Trout and only one medium and one large Brown Trout were recaptured; therefore annual estimates were not possible for Rainbow Trout and were limited for Brown Trout. Water temperatures in the three C-R areas were monitored throughout the study period with HOBO data loggers (Onset, Pocaset, Massachusetts). Temperature loggers were anchored to the bottom of the substrate and placed at an upper, middle, and lower location within each C-R area. The mean water temperature (°C) for 1 d was calculated from data collected at 15-min intervals for each data logger. Temperatures were then averaged daily across the three data loggers to generate the average temperature. Trout diets.—A total of 1,387 trout stomachs were examined in the C-R areas. At Bull Shoals, Norfork, and Sylamore C-R areas we examined 551, 573, and 263 stomachs, respectively. Prior to examination in the laboratory, stomachs were transferred from thee formalin solution to containers with 95% ethanol. At the time of examination, stomachs were dissected and their gut contents were placed in a petri dish. Using a dissecting microscope prey items were identified to lowest practical taxon, counted, and measured to the nearest 0.1 mm with an ocular micrometer. Partially digested or broken macroinvertebrates were identified, counted, and measured based on head widths. Ingested fish prey still intact were identified and measured for TL. When prey fish were in later stages of digestion they were measured according to either vertebral length (VL; vertebral column was complete) or SL (fish missing only the caudal fin). We used the relationship between VL or SL for Ozark Sculpin Cottus hypselurus based on measurements of Ozark Sculpin that were found in the stomachs and ranged in TL from 58 to 101 mm to determine TL from VL [TL = 1.57902(VL), r2 = 0.93] or SL [TL = 1.11903(SL), r2 = 0.98]. Zooplankton (cladocerans) were readily digested in most stomachs, which made accurate length measurements difficult to obtain. In stomachs containing intact zooplankton, the zooplankton were measured from head to tail and an average length of 2.5 mm TL was obtained (n = 135; range = 2.0–3.2 mm; SE, 0.02). In stomachs where zooplankton were not intact the zooplankton were counted in a Ward counting wheel. Counts of zooplankton were then multiplied by the estimated average length from the intact zooplankton to estimate dry mass.

Length–dry mass or head width–dry mass equations from the literature were used to estimate the mass (mg) of each macroinvertebrate and fish (Dumont et al. 1975; Rogers et al. 1976; Sample et al. 1993; Weiland and Hayward 1997; Benke et al. 1999). Algae present in the stomach samples were dried in an oven at 50–60°C for 48–72 h and weighed to obtain dry weights (0.0001 mg). For GCA no distinction was made between Cladophora and D. geminata found in the trout stomachs at Bull Shoals and were combined together as algae for the analyses. In instances where certain taxa of macroinvertebrates were ingested in large numbers (i.e., >125 individuals) a subsampling method was employed to randomly select prey individuals for measuring. All individuals from a taxon were placed in an Imhoff cone and total volume was increased to 1 L with water (Wrona et al. 1982). The subsample was mixed for 2–5 min by bubbling air through the subsample by means of an air stone connected to the bottom of the cone. Subsamples were then removed using a 50-mL Hensen Stempel pipette and total lengths of the first 75 individuals of a taxon encountered were measured. The total counts of prey ingested were multiplied by the average length of prey measured from the subsample to estimate dry mass for the remaining macroinvertebrates in the sample. Stomach contents were expressed as a percent weight (%W), which is the total dry weight of each prey item expressed as percentage of the overall weight of the stomach contents of Brown Trout or Rainbow Trout for each season and size-class. We calculated %W for each prey taxon or group as follows: Wi ¼

Wi Q P

;

Wi

i¼1

where i is the prey item, Wi is the dry weight of prey type i, and Q is the number of prey types. Only stomachs containing prey items were used for calculations and analyses of %W. Dry weights of prey types were converted to energy units (J) with established energy values obtained from our own data and published values for the models (Cummins and Wuycheck 1971; Luecke and Brand 1993; Madon and Culver 1993; Bryan et al. 1996; Hanson et al. 1997). To determine energy values of prey, we collected benthic macroinvertebrates with a Hess sampler and immediately picked the organisms from the samples while they were still alive and placed them on ice. Sculpin and crayfish were sampled seasonally using a 1.0-m2 quadrat sampler with 6-mm mesh by placing the quadrat sampler in riffles and kick-seining within the sampler to dislodge fish and crayfish and wash them into the attached sampler bag (Peterson and Rabeni 2001). In the laboratory all prey items were rinsed with filtered water (Millipore) and inspected for any debris. Sculpin were measured to the nearest TL and crayfish were measured for carapace length (CL). In order to achieve enough sample of macroinvertebrates for bomb calorimetry, multiple organisms (more than three individuals) of the same species were pooled to achieve the minimum

436


mass (i.e., 0.2–0.02 g). Prior to calorimetry, prey samples were thawed, blotted dry, and placed in a tared aluminum weigh boat to obtain wet weight (0.0001 mg). Samples were then dried in an oven at 50–60°C for 48–72 h and reweighed to obtain dry weights. After being dried and weighed, sculpin and crayfish were homogenized whole using a Wiley Mill (40 mesh) and reground, if necessary, into a fine powder to ensure homogeneity within each sample. Aquatic macroinvertebrates were homogenized using a mortar and pestle. Gastropods were extracted from their shells and weighed, and organisms analyzed whole. After drying and homogenizing, the sample was added to the calorimeter vessel to get a complete firing. Prey energy density values (cal/g dry weight) were estimated using a bomb calorimeter (Parr 6200 Calorimeter). Prey energy density values (cal/g dry weight) were then converted to the appropriate units (J/g wet weight) and were based on the percent water determinations from weighed organisms. We used the energy value for the season when available. However, when no energy values were available seasonally, energy values were assumed to be constant throughout the year. Daily energy intake and expenditure.—We used a bioenergetics modeling approach to assess spatial and temporal energy demands by each size-class of Brown Trout and Rainbow Trout. We calculated daily energy expenditure (DEE) or maintenance ration, which is the amount of energy required to obtain zero growth over the course of a day (J·g−1·d−1) and compared DEE to the estimated daily energy intake (DEI) (J·g−1·d−1) or daily ration. Estimates of consumption to determine DEI (J·g−1·d−1) were derived using the Eggers (1977) model: C24 ¼ E24 R; where C24 is consumption (i.e., DEI) over 24 h, E24 is the energy in stomach contents over 24 h, and R is the instantaneous gastric evacuation rate. Taxon-specific length–dry mass regressions of prey observed in the diets were used to convert to energy (J). We assumed no energy was obtained from Cladophora and D. geminata (Weiland and Hayward 1997). For each sampling event, gastric evacuation rates were calculated for different water temperatures (T; °C) using the equation of Elliott (1972) for Brown Trout (R = 0.053e0.112T) and Hayward and Weiland (1998) for Rainbow Trout (R = 0.0405e0.067T). We used stomachs collected from night sampling instead of those from day sampling. Weiland and Hayward (1997) found no differences between mean food weight of Rainbow Trout in day and night samples collected at baseflows in the White River system. Fish with empty stomachs were included in the DEI estimates. We estimated DEE (J·g−1·d−1) using the “Wisconsin” bioenergetics model, which is based on the balanced energy equation (Hanson et al. 1997): G ¼ C ðM þ F þ U Þ; where G = growth, C = consumption, M = metabolic rate (includes specific dynamic action, standard metabolism, and active

metabolism), F = egestion, and U = excretion. We calculated daily energy required to obtain zero growth using the model. The model required specific inputs on the temperature occupied by the fish, fish weight, and fish energy density. Physiological variables used in the model for Rainbow Trout were from Rand et al. (1993) with the exception of maximum consumption and respiration, which were taken from Railsback and Rose (1999). The Brown Trout physiological variables from Dieterman et al. (2004) have provided accurate predictions under various fish sizes, water temperatures, and ration levels (Whitledge et al. 2010). Fish energy densities were estimated using the dry-weight-to-energydensity equation for Salmonidae (Hartman and Brandt 1995). Average temperature from the date of sampling was used for the simulations. We compared estimates of DEI with DEE by season to determine whether fish were obtaining sufficient energy to maintain body weight and to evaluate bottlenecks in food availability compared with consumption. Statistical analysis.—We tested for differences in diets among seasons using a permutational multivariate analysis of variance (PERMANOVA), which tests the simultaneous response of one or more variables to factors in an ANOVA experimental design on the basis of a distance measure using permutation methods (Anderson 2001). The response variables were the proportion of the prey group by dry weight from the diet analysis and the predictor variable was season. Prey groups that represented 79%), with the exception of spring 2006, and the remainder of the diet included benthic fish (e.g., sculpin) (range = 7–21%) (Table 3). Smaller Brown Trout at Norfork exhibited an ontogenetic shift from macroinvertebrates to sculpin in the medium and large sizeclasses (16–92%). Amphipods were the most commonly consumed macroinvertebrate in the diets of Brown Trout at Norfork. At Sylamore, diets of smaller Brown Trout were dominated by Gastropoda and Decapoda, but they did exhibit some piscivory in the winter when sculpin were consumed. Diets of Brown Trout comprised various fish species at Sylamore, such as darters Etheostoma spp., River Redhorse Moxostoma carinatum, and Northern Hogsucker Hypentelium nigricans in the winter and darters and Striped Shiners Notropis chrysocephalus in the spring. Fish (e.g., sculpin) had greater prey energy density (J/g) than did macroinvertebrates (e.g., Amphipoda, Isopoda, Gastropoda) and crayfish at Bull Shoals and Norfork (Table 4). Amphipods had slightly greater energy density than isopods. Gastropods (Plueuroceridae) had the lowest observed energy density. At Bull Shoals, prey caloric values (J/g wet weight) were significantly different among prey types (ANOVA: F3, 11 = 14.307, P < 0.001). The lowest caloric values were found in gastropods (Pleuroceridae) and the highest in sculpin. We also found significant differences in the caloric values of prey at Norfork (ANOVA: F4, 15 = 29.861, P < 0.001), with the lowest caloric values occurring in decapods and the highest in sculpin. As sculpin increased in TL their caloric values decreased at Norfork (linear regression: F1, 8 = 15.145, P = 0.005, r2 = 0.654), while sculpin at Bull Shoals exhibited no relationship between TL with caloric values (linear regression: F1, 6 = 0.763, P = 0.416, r2 = 0.113).

Trout Diets Brown Trout and Rainbow Trout diets differed seasonally among each size-class (PERMANOVA: P < 0.01), with the exception of small (PERMANOVA: F2, 6 = 0.63, P = 0.84) and large Brown Trout (PERMANOVA: F2, 9 = 1.84, P = 0.08) at Sylamore. Isopods were the dominant macroinvertebrate prey item in the diets of Rainbow Trout during summer and fall

Daily Energy Expenditure and Intake We used the contents of 1,387 stomachs to determine the spatial and temporal DEI estimates. Water temperatures in Bull Shoals and Norfork exhibited similar seasonal fluctutations and were much cooler than water temperatures measured in Sylamore. As a result daily mean water temperatures used in the models at Sylamore were warmer than in Bull Shoals and

438


TABLE 1. Mark–recapture results indicating the number of fish marked in the first sampling event (n1), fish captured in second sampling event (n2), and fish ^ with SE and 95% CIs based on binomial distribution by size-class for with a mark (i.e., recapture) in second sampling event (m2). Abundance estimates (N) Brown Trout (BNT) and Rainbow Trout (RBT) and both species of trout (TRT) in the Bull Shoals, Norfork, and Sylamore C-R areas in 2005 (05) and 2006 (06). Mean water temperatures (°C) used for determining daily energy intake (J·g−1·d−1) and daily energy expenditure (J·g−1·d−1) are included. Spr = spring, Sum = summer, Win = winter.

Mark–recapture results Season

Temperature (°C)

Spr 05

8.9

Species

Size-class

BNT

Small Medium Large Small Large

RBT

Sum 05

11.0

TRT BNT

RBT

Fall 05

12.1

TRT BNT

RBT

Win 06

8.4

TRT BNT

RBT

Spr 06

8.9

TRT BNT

RBT

Small Medium Large Small Large Small Medium Large Small Large Small Medium Large Small Large Small Medium Large Small Large

TRT Spr 05

12.1

BNT

RBT

Sum 05

13.7

TRT BNT

RBT

Fall 05

10.5

TRT BNT

Small Medium Large Small Large Small Medium Large Small Large Small

n1 Bull Shoals 26 74 142 355 96 693 10 111 184 507 151 963 1 85 237 456 186 965 12 128 233 693 122 1,188 11 92 179 688 96 1,066 Norfork 24 170 191 397 79 861 2 154 202 593 103 1,054 1

Abundance

n2

m2

^ N

31 75 99 296 59 560 9 100 180 397 152 838 2 84 213 456 153 908 10 86 173 543 95 907 20 67 143 542 80 852

5 21 16 29 4 75 4 37 60 67 20 188 1 13 43 81 28 166 6 36 54 142 28 266 6 21 52 135 16 230

143 258 840 3,523 1,163 5,122 21 297 548 2,972 1,106 4,278 2 521 1,157 2,546 992 5257 19 302 739 2,639 406 4,042 35 286 488 2,750 461 3,939

43.3 38.3 169.6 574.3 443.1 512.2 4.7 30.6 46.5 303.3 203.6 245.0 0 112.7 138.9 229.4 150.1 333.3 3.0 31.5 71.6 168.3 54.3 182.7 6.7 43.1 44.4 182.3 87.9 195.5

75–303 187–375 559–1,312 2,532–4,966 536–2,684 4,173–6,330 13–45 235–388 452–679 2,407–3,705 756–1,659 3,783–4,861 1–9 333–853 898–1,516 2,103–3,109 726–1,390 4,594–6,041 14–36 242–394 601–929 2,301–3,048 306–560 3,663–4,480 21–69 210–413 400–612 2,387–3,191 311–716 3,537–4,407

24 109 127 225 48 533 1 114 126 838 91 1,170 7

7 30 49 27 7 120 0 38 54 120 15 227 1

77 606 491 3,211 489 3,803 5 456 468 4,118 597 5,417 7

17.7 82.2 46.2 538.4 141.7 280.8 2.4 50.8 40.3 307.8 121.3 284.5 0

47–144 458–828 401–617 2,296–4,567 275–939 3,264–4,463 1–116 360–597 389–579 3,501–4,865 393–945 4,830–6,096 2–27

SE

95% CI

439

CONSUMPTION DYNAMICS OF TROUT TABLE 1. Continued.

Mark–recapture results Season

Temperature (°C)

Species

RBT

Win 06

9.4

TRT BNT

RBT

Spr 06

11.4

TRT BNT

RBT

Size-class Medium Large Small Large Small Medium Large Small Large Small Medium Large Small Large

TRT Fall 05

Win 06

Spr 06

Fall 06

18.0

BNT

8.7

RBT TRT BNT

18.3

RBT TRT BNT

17.8

RBT TRT BNT

RBT TRT

Small Medium Large Small Small Medium Large Small Small Medium Large Small Small Medium Large Small

n1 80 95 636 38 850 25 106 107 869 73 1,180 12 113 152 756 35 1,068 Sylamore 1 35 9 285 330 10 54 9 910 983 8 14 6 139 167 1 8 2 44 55

Norfork (Table 1). The DEI and DEE of Brown Trout and Rainbow Trout varied between seasons and sites (Figures 5, 6). The DEE in trout was highest at Sylamore, except in the winter, compared with the other sites due to the elevated water temperatures. The DEI was often much higher in Brown Trout than in Rainbow Trout. Generally the medium size-class of Brown Trout experienced the highest seasonal DEI for both species in the C-R areas. The mean DEI for Brown Trout at Norfork either exceeded

Abundance

n2

m2

^ N

86 73 717 47 930 22 92 115 814 74 1,117 14 112 163 662 33 984

15 21 142 10 189 6 26 37 220 23 312 1 25 51 131 6 214

439 322 3,197 169 4,169 84 368 329 3,207 230 4,217 97 494 482 3,801 174 4,897

86.4 49.6 210.0 36.5 237.2 21.5 50.7 34.9 158.8 31.3 173.2 48.2 73.5 44.5 268.1 49.4 263.3

290–694 234–467 2,769–3,711 106–292 3,680–4,742 50–166 274–514 259–433 2,873–3,597 170–327 3,846–4,641 33–347 360–702 390–609 3,273–4,440 97–347 4,360–5,521

1 31 7 258 297 11 45 11 955 1022 7 14 3 211 235 0 13 1 76 90

0 6 2 30 38 2 15 2 179 198 1 1 0 11 13 0 0 0 1 1

3 164 26 2,388 2,528 43 157 39 4,837 5,057 35 112 27 2,472 2,831 1 125 5 1,732 2,547

1.4 46.2 8.8 374.2 350.2 16.2 26.1 14.5 290.2 286.7 15.9 56.3 15.9 637.1 679.0 0 80.9 2.4 965.0 1,428.6

1–77 92–326 13–72 1,734–3,340 1,902–3,403 20–123 111–241 18–111 4,250–5,528 4,474–5,739 14–125 39–400 9–545 1,453–4,284 1,706–5,004 1–38 28–2,455 2–116 500–6,291 774–4,969

SE

95% CI

or achieved metabolic demand across all seasons for small (five of five), medium (five of five), and large (five of five) sizeclasses. Although DEI was typically not as high for Brown Trout in Bull Shoals (250 mm) to grow at faster rates and supported higher Brown Trout densities than that found in Bull Shoals and Sylamore C-R areas. Feeding efficiencies and metabolic costs of fish operate as a function of temperature to influence fish growth rates (Elliott

1976; Wurtsbaugh and Davis 1977). Temperature also functions to influence macroinvertebrate production at various spatial and temporal scales (Hogg and Williams 1996) and may limit fish consumption rates if prey densities are low (Huryn 1996). The maximum water temperatures observed in Bull Shoals (13.8°C) and Norfork (14.8°C) were near the optimal range for trout growth throughout the summer and fall. Spring water temperatures at Sylamore (23.2°C) were above optimal for Brown Trout and Rainbow Trout. The optimal reported range for growth in Brown Trout is 12–13°C and in Rainbow Trout is 17–18°C (Hokanson et al. 1977; Jobling 1991; Elliott and Hurley 1998). Upper lethal temperatures for Brown Trout are slightly higher (29–30°C) than those reported for Rainbow Trout (25–27°C) (Hokanson et al. 1977; Elliott and Elliott 1995; Bear et al. 2007). The elevated water temperature at Sylamore C-R area resulted in higher metabolic costs and poor growth of trout.

446


FIGURE 6. Seasonally observed daily energy intake (DEI) or consumption (J·g−1·d−1) compared with daily energy expenditure (DEE) or maintenance ration (J·g−1·d−1) with 95% CIs by size-class for Rainbow Trout in the Bull Shoals, Norfork, and Sylamore C-R areas.

Benthic macroinvertebrate production at Sylamore was also the lowest observed in the three C-R areas (Flinders 2012). The higher water temperatures influencing trout growth, inability to collect any large Rainbow Trout, and low densities of Brown Trout (~3 fish/ha) all suggest marginal trout habitat existed at Sylamore. Thus, the likelihood of Sylamore C-R in achieving any management success as a C-R area (i.e., provide high catch rates of large fish) seems highly unlikely. We found both growth and bioenergetics modeling provided useful insights about the population dynamics of Brown Trout and Rainbow Trout in Arkansas tailwater C-R areas. There was good correspondence between daily energy intake and growth suggesting Rainbow Trout, particularly the smaller size-class, were consuming below the maintenance level, and diets contained high algal consumption suggesting periods of low food availability (Weiland and Hayward 1997; McKinney and Speas 2001). Brown Trout growth was generally positive, and energy intake was at the minimum or above the maintenance ration indicating their ability to be an ideal species for management in the C-R areas at the densities observed in this study (6–111 fish/ ha). Rainbow Trout densities in the three C-R areas in Bull Shoals and Norfork tailwaters (47–342 fish/ha) were substantially lower during this study than in Taneycomo tailwater (1,400 fish/ha), also an Ozark tailwater (Weiland and Hayward 1997). Despite the lower densities, a further reduction in trout densities

in the C-R areas would likely allow this species to increase in size (Yard et al. 2016). Although no trout were directly stocked into the C-R areas, hatchery stockings of trout nearby (~1 km) move into the C-R areas (Cushing 2007) and likely influence trout densities in the C-R areas. Rainbow Trout experience limited or no reproductive success based on stable isotope analysis (J. M. Flinders and D. D. Magoulick, unpublished data). Since Rainbow Trout have limited success in recruiting in the tailwaters, densities in the C-R areas are dependent largely on stockings outside C-R areas. Decreased densities of Rainbow Trout in the C-R areas could be accomplished by decreasing the number of trout stocked in the tailwaters. However, harvest-oriented anglers targeting catchable trout in areas outside the C-R may not be willing to experience decreased catch rates in an effort to improve trout growth in C-R areas. Increased bag limits outside the C-R areas (i.e., higher harvest rates) would reduce trout densities if harvest rates increased and would not require a reduction in trout stockings. Allowing harvest of smaller Rainbow Trout in C-R areas through the implementation of a minimize size limit restriction (