Volume 152, Issue 5 p. 577-593
ARTICLE
Open Access

Evaluating Muskellunge catch-and-release mortality at elevated summer water temperature

Ian Taylor Booth

Corresponding Author

Ian Taylor Booth

West Virginia University, Division of Forestry and Natural Resources, Wildlife and Fisheries Program, Morgantown, West Virginia, USA

Correspondence

Ian Taylor Booth

Email: [email protected]

Search for more papers by this author
Kyle J. Hartman

Kyle J. Hartman

West Virginia University, Division of Forestry and Natural Resources, Wildlife and Fisheries Program, Morgantown, West Virginia, USA

Search for more papers by this author
Derek Crane

Derek Crane

Coastal Carolina University, Conway, South Carolina, USA

Search for more papers by this author
Jeff Hansbarger

Jeff Hansbarger

West Virginia Division of Natural Resources, Alum Creek, West Virginia, USA

Search for more papers by this author
Jordan Weeks

Jordan Weeks

Wisconsin Department of Natural Resources, La Crosse, Wisconsin, USA

Search for more papers by this author
Josh Henesy

Josh Henesy

Maryland Department of Natural Resources, Thurmont, Maryland, USA

Search for more papers by this author
Heather Walsh

Heather Walsh

U.S. Geological Survey Eastern Ecological Science Center, Kearneysville, West Virginia, USA

Search for more papers by this author
Jeff Williams

Jeff Williams

Virginia Department of Wildlife Resources, Marion, Virginia, USA

Search for more papers by this author
First published: 15 April 2023
Citations: 3

Abstract

Objective

Fisheries managers and anglers have expressed concerns regarding warmwater angling mortality, representing a need to evaluate mortality rates at various water temperatures and multiple latitudes. Up to 97% of Muskellunge Esox masquinongy caught by anglers are released, and previous research on catch and release (C&R) for Muskellunge has suggested relatively low mortality rates (0–5%). However, those studies were all conducted within the range of water temperatures that are thermally optimal for Muskellunge and generally at water temperatures less than 25°C. As many Muskellunge populations routinely experience temperatures greater than 25°C during the summer months our objective was to quantify warmwater (>25°C) C&R mortality rates in Muskellunge.

Methods

We collected 102 adult Muskellunge (>760 mm) and stocked them into and identify factors influencing mortality by using experimental ponds. Adult Muskellunge (n = 102) were stocked into eight earthen or plastic-lined flow-through ponds (0.06–0.71 ha) at densities less than 16 fish/ha. Muskellunge (n = 50) were angled by utilizing specialized Muskellunge fishing gear at water temperatures of 19.6–32.6°C, with 32 fish being caught at temperatures exceeding 25°C. After being angled, fish were closely monitored for 2 weeks to assess mortality; fish that remained uncaught during the experiment were used as controls (n = 53).

Result

Mortality was greater for angled fish (30.0%) than for control fish (11.3%). Differences in C&R mortality were compared across a range of temperature regimes using Firth logistic regression. Five-day cumulative temperature and net time were positively related to the probability of mortality, but size and sex were unrelated to mortality. Increasing C&R mortality with temperature was mitigated somewhat by lower catch rates at higher temperatures. Mortalities per 100 angler-hours were 0 at <25°C, 4.98 at 25.00–27.49°C, 2.48 at 27.5–30.0°C, and 1.17 at >30°C.

Conclusion

Recent field studies have identified the importance of thermal refuge in mitigating summer C&R mortality of Muskellunge. This study identified specific temperature conditions responsible for elevated mortality in the absence of refugia. Although increasing temperatures above 25°C led to increasing C&R mortality in our ponds, lower catchability seemingly provided some mitigation. The interactive effects of thermal refugia and catch rates with temperature warrant further investigations into population-level effects at varying levels of exploitation.

Abstract

Impact Statement

Muskellunge catch-and-release mortality is positively correlated with water temperature and handling times, posing risks to populations in the species' southern range and where summer water temperatures exceed 25°C.

INTRODUCTION

Recreational angling is practiced by millions across the globe, and its contribution to declining stocks of fish has been well documented (Post et al. 2002; Cooke and Cowx 2004; Food and Agriculture Organization of the United Nations 2012). Historically, recreational angling was harvest oriented, but today the practice has shifted to become release oriented, with up to 60% of recreationally caught fish being released (Cooke and Cowx 2004) in the hopes of improving size structures and producing trophy individuals (Quinn 1996; Arlinghaus et al. 2007). The motivations for participating in this practice can be due to ethical concerns or because of restrictive regulations (Arlinghaus et al. 2007), but anglers generally assume that release will have no measurable negative impact on survival and that the fish can be caught again in the future (Wydoski 1977; Brousseau and Armstrong 1987).

The Muskellunge Esox masquinongy is a recreational sport fish whose native range extends from the St. Lawrence River in Quebec to southeastern Manitoba and stretches south along the Appalachian Mountains to Tennessee and North Carolina (Page and Burr 1991). Historically, anglers believed that Muskellunge were predators with harmful impacts on other fish populations; therefore, Muskellunge were heavily harvested (Crossman 1986). High harvest rates persisted into the 1970s (Richards 2017), and the combined effects of recreational harvest, commercial fishing (Crossman 1986), and loss of spawning habitat in the 1900s (Dombeck 1979) contributed to the decline of natural Muskellunge populations. In the 1940s, anglers first began to express concern over high exploitation rates and subsequent population-level effects (Farrell et al. 2007). In the 1980s and 1990s, Muskellunge fisheries began to improve due to the increasing popularity of conservation-oriented catch-and-release (C&R) angling practices, introductions into new waters to create recreational fisheries, and a shift in management tactics designed to increase catch rates and improve size structure (Farrell et al. 2007; Richards 2017; Simonson 2017). By 2010, most management agencies had decreased bag limits and increased minimum size limits (Kerr 2011), resulting in increased abundances and average size of Muskellunge across the species' range (Kerr 2007; Casselman et al. 2017).

Muskellunge can reach lengths up to 1.5 m and weights up to 27 kg, can live up to 30 years (Casselman 2007), and exist at relatively low densities (<1 adult fish/ha; Jennings et al. 2010). The popularity of Muskellunge angling has increased substantially since C&R practices were popularized in the 1980s (Richards 2017), with agencies reporting large increases in the number of anglers specifically targeting Muskellunge (Simonson 2012; Meerbeek 2014; Richards 2017). Even with increased angler participation, Muskellunge harvest rates are low and release rates are estimated to be greater than 97% (Fayram 2003; Margenau and Petchenik 2004; Kerr 2007; Landsman et al. 2011; Simonson 2012; Richards 2017). Because of the Muskellunge's life history characteristics (low population density and long life span) and historical exploitation, small increases in mortality rates can affect abundance and size structure (Faust and Hansen 2016). Therefore, research identifying the factors that influence mortality and resultant population-level effects is important to management agencies and angling groups (Arlinghaus et al. 2007). Water temperature increases induced by climate change constitute one such factor that has become increasingly apparent. It is imperative to inform managers of the possible effects of increased water temperatures on aquatic ecosystems. The physiological consequences that accompany increasing temperatures can be detrimental to ectotherms (e.g., fish), especially under exhaustive stress, such as C&R angling (Clarke and Johnston 1999). A literature review by Gale et al. (2013) found that warmer temperatures resulted in higher mortality in 49 (70%) of the 70 papers that quantified postcapture survival for 17 different fish species. Boyd et al. (2010) also observed this trend in Rainbow Trout Oncorhynchus mykiss and Mountain Whitefish Prosopium williamsoni, which exhibited C&R mortality rates of 16% and 28%, respectively, when temperatures were 23°C or greater, whereas these species exhibited no C&R mortality at temperatures less than 20°C. In a meta-analysis of C&R mortality studies on Striped Bass Morone saxatilis, Bettinger and Wilde (2013) estimated that 70% of fish caught using live bait and 38% of fish caught using artificial baits died after release during the summer months. Previous work has suggested low mortality (0–5%) for angled and released Muskellunge when artificial lures are utilized (Strand 1986; Frohnauer et al. 2007; Landsman et al. 2011; Hessenauer et al. 2021), whereas mortality is higher (22% in 50 days, 83% in 1 year; Margenau 2007) when single J-hooks and live bait are used. However, these studies were all conducted in the northern extent of the species' range in near-optimal water temperatures of 15–26°C (22.3 ± 1.8°C; Cole and Bettoli 2014), with only one fish being caught when water temperature was greater than 25°C (Strand 1986; Frohnauer et al. 2007; Margenau 2007; Landsman et al. 2011). The southern extent of the Muskellunge's distribution includes North Carolina and Tennessee, where water temperatures routinely exceed 25°C in the summer months (Crossman 1978; National Oceanic and Atmospheric Administration 2021; U.S. Geological Survey [USGS] 2021). Henesy et al. (2022) found that when daily average temperatures reached 24°C, over 50% of radio-tagged Muskellunge occupied thermal refugia, and occupation of refugia increased to 90% when temperatures reached 26°C. The occupation of thermal refugia at 24°C and 26°C further compounds the need to explore the effects of C&R mortality at elevated temperatures. Lack of information on the effects of C&R angling at temperatures greater than 25°C represents a science gap that is of interest to managers and anglers.

Some Muskellunge anglers have adopted a practice to voluntarily cease angling when surface water temperatures exceed 26.7°C (80°F), as popularized in an article (Heyob 2013) that suggested an increased frequency of floating or dead Muskellunge (from unknown causes) when surface water temperatures exceeded this threshold. Evaluating the effects (or lack thereof) of C&R angling at elevated temperatures is essential to properly educate anglers and inform managers of potential threats to Muskellunge populations. We chose to investigate C&R mortality of adult Muskellunge at water temperatures greater than 25°C due to the lack of information from previous studies at elevated temperatures, the threat of climate change-induced increases in water temperature, and the shift in angler sentiments and fish behaviors when temperatures exceed this threshold.

Effects of C&R angling on fish are influenced by biotic factors as well as abiotic factors and angler experience (Twardek et al. 2018). Prolonged air exposure, fight time, and severe hooking injury have been demonstrated as detrimental to physiology after release (Barton and Iwama 1991; Muoneke and Childress 1994; Bartholomew and Bohnsack 2005; Arlinghaus et al. 2007). Muskellunge angling has become highly specialized, with anglers adopting unique fishing equipment and handling techniques to mitigate the sublethal and lethal consequences of C&R (Gasbarino 1986; Chipman and Helfrich 1988; Quinn 1996; Cooke and Suski 2005; Campbell et al. 2010). Large rods, heavy line, wire or fluorocarbon leaders, and large nets are utilized by experienced Muskellunge anglers (Landsman 2008). Regardless of how prepared and careful an angler is, Muskellunge still experience a degree of stress associated with angling. After being angled and released, Muskellunge may experience greater disease susceptibility, reduced body condition, lower reproductive potential, and even death (Suski et al. 2003, 2007; Arlinghaus et al. 2007; Klefoth et al. 2008). The threats of climate change in concert with the dangers of warmwater angling have made it increasingly important to provide relevant information on how fish and fisheries will react under elevated temperatures and increased levels of exploitation. The goals of this study were to use experimental ponds to (1) quantify C&R mortality of Muskellunge at various summer water temperatures and (2) identify factors (environmental, biotic, and handling) that influence C&R mortality in Muskellunge. Information from this study can be used by anglers and managers to evaluate the effects of C&R angling on Muskellunge populations.

METHODS

Study sites

Angling took place in eight ponds spread across four cooperating hatchery facilities and one state park during May–September of 2020 and 2021. Ponds were located at Palestine State Fish Hatchery (Elizabeth, West Virginia; n = 2 ponds; 0.61 ha), USGS Eastern Ecological Science Center (EESC; Kearneysville, West Virginia; n = 1 pond; 0.2 ha), Table Rock State Fish Hatchery (Morganton, North Carolina; n = 2 ponds; 0.06 ha), Richard Bong State Recreation Area (Kansasville, Wisconsin; n = 1 pond; 0.71 ha), and Buller Fish Hatchery (Marion, Virginia; n = 1 pond; 0.61 ha). The number of fish in each pond ranged from 6 (North Carolina) to 26 (West Virginia) individuals. Maximum depth in the ponds ranged from 1.5 to 4.6 m. Mechanical aerators were utilized to maintain proper dissolved oxygen (DO) levels (>4.15 mg/L) and to keep pond temperatures homogenized. Water temperature data were collected on site by Onset HOBO MX2201 water temperature data loggers at two or more locations and depths in each pond. To maintain proper fish health and reduce confounding environmental variables, weekly water quality checks were conducted to ensure that ponds were at acceptable nitrate (<40 mg/L), nitrite (<2 mg/L), pH (6.7–8.0), DO (>4.15 mg/L), and chlorine (<0.1 mg/L) levels. Water quality data were collected using a YSI Pro 10 multiparameter meter and a Hach Surface Water Test Kit. All ponds were plastic lined or had an earthen bottom, and a continuous flow of freshwater was maintained throughout the duration of the experiment.

Fish collection and care

Adult Muskellunge (>760 mm total length) were collected from wild populations in the vicinity of hatchery locations in Maryland, North Carolina, Virginia, West Virginia, and Wisconsin. In total, 102 fish were collected during March–June 2020 (n = 12) and 2021 (n = 90; Table 1). Muskellunge were collected by cooperating state agencies through boat electrofishing and trap-netting; they were then held in live wells to maintain proper DO levels (>7 mg/L) while being transported to the ponds. Before being released into the ponds, fish were checked to determine whether they were previously tagged with a passive integrated transponder (PIT) tag. If a fish was not previously tagged, it was injected with a PIT tag (Biomark) in the dorsal musculature or directly behind the operculum, depending on the discretion of state agencies. Muskellunge were then measured (total length, mm; wet weight, kg), and sex was determined by expulsion of gametes or by examination of external morphology of the urogenital region (LeBeau and Pageau 1989). Ponds were stocked with forage on a regular basis with locally collected and donated forage fish (American Shad Alosa sapidissima, Smallmouth Buffalo Ictiobus bubalus, Bluegill Lepomis macrochirus, Green Sunfish L. cyanellus, Grass Carp Ctenopharyngodon idella, and suckers [Catostomidae]) or hatchery-reared Rainbow Trout. Acclimation periods ranged from 2 weeks to 2 months before angling events were conducted. Upon conclusion of angling and observation periods, ponds were drained and the remaining fish were identified, returned to their respective water bodies, or kept as future broodstock by the hatchery. Wisconsin fish were euthanized due to biosecurity concerns.

TABLE 1. Names and locations of hatchery sites, collection sites, and coordinates where Muskellunge were captured for use in the mortality study during 2020–2021. Overall, 102 Muskellunge were included in analyses. Fish were sourced from 12 waters and were housed with individuals from their own state across eight ponds.
Pond site Collection site Latitude, longitude Number of fish Average total length (mm)
Palestine State Fish Hatchery, Elizabeth, WV East Lynn Lake, WV 38.108894°, −82.353145° 18 928.4
Stonecoal Reservoir, WV 38.979718°, −80.360653° 13 937.6
North Bend Lake, WV 39.222909°, −81.083614° 11 894.8
Monongahela River, WV 39.622507°, −79.967214° 10 976.5
U.S. Geological Survey Eastern Ecological Science Center, Kearneysville, WV Potomac River, MD 39.515613°, −77.862987° 6 1034.0
Richard Bong State Recreation Area, Kansasville, WI Lake Waubesa, WI 43.013583°, −89.317683° 5 966.2
Wisconsin River, WI 43.308380°, −89.725994° 6 941.5
Twin Valley Lake, WI 43.026785°, −90.086010° 5 913.4
Pewaukee Lake, WI 43.067136°, −88.308609° 5 958.6
Table Rock State Fish Hatchery, Morganton, NC New River, NC 36.461084°, −81.335328° 9 1055.1
Lake Rhodhiss, NC 35.781681°, −81.485165° 3 922.7
Buller Fish Hatchery, Marion, VA New River, VA 37.083519°, −80.578670° 12 1021.0

Angling gear

The majority of Muskellunge are caught by anglers specifically targeting Muskellunge (Bauerlien et al. 2021, 2022); therefore, only specialized Muskellunge angling equipment was used in this study. Angling equipment consisted of 2-m or longer heavy-action rods, large reels, main line braid and fluorocarbon or wire leaders rated to at least 22.7-kg test, and large artificial lures. We utilized common Muskellunge lures up to 51 cm long that possessed both single and treble hooks and ranged from one to nine points per lure (Figure 1). All hooks used were barbed to increase catch rates and decrease the chances of losing a fish during the event. No live bait was used for this study, as it makes up a small percentage of angled Muskellunge and results in higher mortality rates relative to artificial lures (Margenau and Petchenik 2004). Large-mesh, knotless landing nets were utilized because they have been shown to lower the incidence of injury and mortality when netting angled fish (Barthel et al. 2003; Brownscombe et al. 2017; Lizée et al. 2018) and are the most common type of net used by Muskellunge anglers. Anglers were fitted with head-mounted action video cameras that continuously recorded angling and handling procedures.

Details are in the caption following the image
Examples of some of the artificial lures that were used to catch Muskellunge in this study. Lure lengths ranged from 51 to 508 mm and had up to three treble hooks.

Angling protocols

Fish were angled from the hatchery ponds by or under the direct supervision of experienced Muskellunge anglers once surface temperatures exceeded 21.1°C. We aimed to angle half of the fish in each pond, leaving the remainder to act as controls. Data collection for each angling event began when the hook was set and concluded upon successful release of the fish back into the pond. Angling was conducted in groups of two or more people to reduce handling times and stress on the fish. Upon setting the hook, a timer was started and the fish was netted as quickly as possible, as is typically done by Muskellunge anglers. Fight time (s) was recorded as the time it took to reel in the fish, and the timer was stopped once the fish was captured in the net. Once landed, the fish was kept in the water for dehooking and processing. Hooking location(s) was identified as the mouth, gills, esophagus, and/or tongue, and hooks were removed with long needle-nosed pliers. Because there are multiple points per lure, fish were often hooked in multiple locations. In these instances, the primary hooking location was used in analyses. Primary hooking location was defined as the “weight-bearing” location, or the location in which the most force was exerted on the fish as it was reeled in. If unhooking times exceeded 2 min, then the hooks were cut below the barb with handheld bolt cutters. The fish was then scanned for its PIT tag and recorded for individual identification. Net time (s) was recorded as the time from initial landing until the fish was removed from the net for the admiration period. Because Muskellunge are highly prized, many anglers hold them out of water for pictures and admiration. To emulate this, each fish was held out of water for 30 s in a horizontal orientation to simulate how an angler would take a picture or admire the catch. Fish were either held by hand at the base of the caudal fin and with a hand supporting the body near the pectoral fins or by hand on the operculum and with a hand supporting the body near the anal fins. The fish was immediately released back into the pond after the 30-s handling period.

After release, reflex impairment was observed and given a total score ranging from 0 to 3 (Davis and Ottmar 2006). We looked at three reflex indicators: the ability to maintain equilibrium, burst swimming away upon release, and descent from the water surface after release. Each individual reflex was given a score of either 0 or 1. A score of 0 indicated that the reflex was not present (e.g., the fish was unable to maintain equilibrium), while a score of 1 indicated that the reflex was present. The sum of the three scores was the total reflex action mortality predictor (RAMP) score for each fish. We evaluated RAMP scores because they have the potential to provide a noninvasive, rapid assessment of fish health that could allow anglers to determine the likelihood of postrelease mortality. If a lower RAMP score is associated with an increase in the probability of postrelease mortality, then anglers may consider harvesting those individuals to reduce wanton waste.

Angler effort for each trip was recorded and included the number of participating anglers, start time, end time, and the number of fish caught. Ponds were then observed daily for a period of 2 weeks after C&R events to detect any laboring, floating, or dead Muskellunge. If a Muskellunge perished within 2 weeks after its initial C&R event, that mortality was considered a direct consequence of angling. Fish that perished outside the 2-week window from initial C&R were considered a result of natural mortality. Researchers and hatchery personnel checked the ponds for deceased Muskellunge daily after angling events throughout the duration of the study. When a dead fish was found, the day and time of its discovery were recorded and the body was immediately recovered from the ponds for processing. Deceased fish were scanned for a PIT tag to identify the fish and its angling history (angled or control). Total length and weight were also measured if the body had not decayed or been preyed upon. Fish were observed for any abnormalities or lesions that could be inconsistent with angling events.

Variable preparation for modeling

Temperature was treated as a continuous variable in our analyses. Due to one of the ponds being deep enough to develop a thermocline (Wisconsin; 4.6 m; mean difference of 4.85°C between top and bottom temperature loggers), the average of surface and bottom temperature readings for each pond was used during analyses. Catch-and-release studies have generally used the average daily temperature or the temperature at the time of catch when assessing mortality rate. We hypothesized that the temperatures before and after a C&R event influenced the fate of released fish. Increasing temperatures have been shown to exacerbate the stress response in ectotherms like fish (Clarke and Johnston 1999), and a prolonged stress response can prove to be detrimental to fish health. Examining the temperatures before and after an angling event can help to elucidate whether fish are more susceptible to C&R mortality from prolonged thermal stress. Thus, we created multiple temperature metrics to evaluate this hypothesis: average daily temperature, temperature at the time of catch, 5-day cumulative total, 4-day cumulative total, post-3-day cumulative total, pre-3-day cumulative total, post-4-day cumulative total, and pre-4-day cumulative total. The time at which the fish was caught acted as the midpoint around which the 5-day and 4-day cumulative totals were centered (i.e., cumulative total of the 48 h before and 48 h after the time of catch = 4-day cumulative total). Temperature metrics were used to individually create candidate models, with only one temperature metric being used per model in combination with other predictor variables (i.e., we fitted models by varying the temperature metric while holding the other predictor variables constant for all possible combinations of predictor variables). Candidate models of individual predictor variables were also explored. Total length (mm) was treated as a continuous predictor variable and used in lieu of weight or age because it is generally the basis of most C&R regulations and provides the most relevant and easily obtained information to anglers and managers.

Because fish were collected from multiple populations and housed in separate ponds, we initially considered a categorical variable for pond location to be included with all candidate models. However, a logistic regression model using individual ponds (e.g., West Virginia pond 5, West Virginia pond 6, North Carolina pond 16, North Carolina pond 17, etc.) to predict the probability of mortality indicated that mortality did not differ among ponds. Additionally, a model using pond as a predictor variable did not perform better than a null (intercept-only) model or a fully parameterized model. Lacking any significant pond effect on response variables, we pooled data across all ponds for subsequent analyses. The acclimation period was also considered for inclusion in our candidate models, as fish had differing lengths of time between being stocked into ponds and subsequent capture. A logistic regression model using acclimation period (days) as a predictor variable was not significant and did not perform better than a null (intercept-only) model or a fully parameterized model. Acclimation period had no significant effect on our response variable of mortality and was thus excluded from the construction of candidate models.

Statistical analyses

Fisher's exact test was used to determine whether C&R mortality differed from control mortality, and a difference was considered significant at p-values less than or equal to 0.05 (Fisher 1922). To test which factors most influenced mortality, we utilized R version 4.0.3 through R Studio version 1.2.5033 (R Core Team 2021) to fit Firth logistic regression models using the brglm2 package (Kosmidis 2021). Firth logistic regression reduces the bias resulting from small sample size, rare occurrence of events (mortality), and cases of separation compared to standard logistic regression (Firth 1993). We estimated the probability of mortality for Muskellunge that were caught and released as a function of water temperature, fight time, net time, length, and hooking location. We created all possible combinations of candidate models that made ecological sense and in which the variables passed checks of multicollinearity (variance inflation factor < 2). Models that included more than one temperature regime were excluded from the list of candidate models, as it did not make ecological sense to observe more than one temperature regime per model. Candidate models were constructed based on additive and interactive relationships between the predictor variables, with up to two levels of interaction. The ability of RAMP to predict mortality was assessed independently using Firth logistic regression. We estimated the probability of mortality as a function of RAMP scores, with each RAMP score between 0 and 3 treated as a categorical variable. This was assessed independent of our candidate models to determine (1) the accuracy with which RAMP scores predict mortality and (2) whether RAMP scores could be an effective tool for anglers and managers to utilize. We also used Firth logistic regression to estimate the probability of catching a fish as a function of average daily water temperature. Finally, linear regression was used to describe the relationship between angler catch per unit effort (CPUE) and average daily temperature by using the stats package (R Core Team 2021). To assess whether changes in catch rates at different temperature regimes offset any increase in C&R mortality, we decided to provide a metric of mortalities per 100 angler-hours. Mortalities per 100 angler-hours were calculated for trips conducted at four different temperature regimes: less than 25°C, from 25.00°C to 27.49°C, from 27.5°C to 30.0°C, and greater than 30°C.

The Bayesian information criterion (BIC) was used to select the most likely model among the candidate models describing the relationship of temperature, angling duration, size, and hooking location with the probability of mortality for caught-and-released Muskellunge (Schwarz 1978). Top models were chosen based on the BIC score, the difference in BIC (ΔBIC), and Schwarz weights, and we assumed that variables included in the top models had the greatest influence on mortality. The odds of mortality were calculated using variable coefficients from our most likely model. Finally, odds of mortality and odds ratios (ORs) were calculated to further investigate the effects of the predictor variables that were included in the most likely models. We also thought it important to identify models that can be easily interpreted by recreational anglers. We did this to provide a digestible model that is most applicable to the type of data that recreational anglers can obtain at the time of catch. It is unreasonable to expect an angler to know the 5-day cumulative temperature or to predict the post-3-day cumulative temperature at the time of angling.

RESULTS

Angling and collection

During 2020 and 2021, 102 Muskellunge (50 males, 36 females, and 16 individuals of unknown sex) were collected by boat electrofishing or trap-netting. Total length (mm) of fish averaged (±SD) 960 ± 90 mm and ranged from 761 to 1208 mm (Figure 2). Wet weight (kg) of fish averaged 6.4 ± 2.3 kg and ranged from 3.1 to 15.0 kg. Fifty Muskellunge were caught and released during May–August 2020 and 2021. Average daily water temperatures in the ponds while angling was conducted ranged from 19.6°C to 32.6°C (Figure 3). All ponds (besides the deepest pond in Wisconsin) maintained average surface and bottom differentials of less than 1°C and lacked thermal refuge. The Wisconsin pond had a maximum depth of 4.6 m and contained thermal refuge throughout the summer. Average pond temperature when fish were angled was 26.2 ± 3.4°C (range = 19.57–31.15°C), and 30 fish (60%) were caught at temperatures of 25°C or greater. Mean fight time (±SD) for angled fish was 48.5 ± 63 s (range = 8–394 s), and mean net time was 142.6 ± 59 s (range = 43–360 s). All of the ponds maintained safe levels of nitrate (<40 mg/L), nitrite (<2 mg/L), pH (6.7–8.0), DO (>4.15 mg/L), and chlorine (<0.1 mg/L) throughout the duration of the study.

Details are in the caption following the image
The length frequency distribution of individual Muskellunge used for the pond study across all states (n = 102, min = 761 mm, max = 1208 mm, μ = 960 mm, bin width = 14.9 mm).
Details are in the caption following the image
Average daily temperature during acclimation and angling periods in all ponds utilized for the Muskellunge mortality study (MD = Maryland pond 7; NC 16 = North Carolina pond 16; NC 17 = North Carolina pond 17; VA = Virginia pond 9; WV 5 = West Virginia pond 5; WV 6 = West Virginia pond 6; WV P = West Virginia [2020] pond; WI S = Wisconsin pond surface temperature; WI B = Wisconsin pond bottom temperature).

Mortality

The total mortality rate of all Muskellunge was 20.6% (n = 21 deaths) and ranged from 8.3% at Table Rock State Fish Hatchery, North Carolina, and USGS EESC, West Virginia, to 25% at Palestine State Fish Hatchery, West Virginia, and the Richard Bong Recreation Area, Wisconsin (Table 2). Mortality rates from C&R angling ranged from 0% (Wisconsin; n = 0) to 42.9% (West Virginia; n = 12), and overall C&R mortality for the experiment was 30% (n = 15; Table 2). For fish that were caught and released at temperatures of 25°C or greater, mortality was 43.3% (n = 13). In contrast, when capture temperatures were 19–25°C, C&R mortality was 10% (n = 2). Control fish (n = 53) mortality was 11.3% (n = 6), and the proportion of fish that died after C&R was significantly greater than the proportion of control fish that died (Fisher's exact test: p = 0.03). Of the control deaths, 83.3% occurred in one pond in 2021 (Wisconsin; n = 5), with the other death occurring during the 2020 season in West Virginia. Five fish were captured two or more times throughout the study; recapture events all occurred outside the 2-week observation period from initial capture but were not included in analyses. One angled fish perished more than 2 weeks after the initial catch date and was included as a control mortality for analyses.

TABLE 2. Catch-per-unit-effort (CPUE; fish/h) data from 78 trips and over 525.38 h of angling during the Muskellunge catch-and-release (C&R) mortality study in 2020–2021. Total CPUE estimates are given alongside data from each pond location and split into trips that occurred at temperatures above 25°C or below 25°C. The total mortality and C&R mortality rates (%) are also provided for each state's respective ponds. “Observations” is the number of fish included in each regime. Abbreviation: N/A, not applicable.
Pond location or variable Temperature Mortality rates
All <25°C >25°C Total C&R
All ponds 0.09 0.27 0.07 20.6 30.0
WV 0.07 2.00 0.07 25.0 42.9
MD 0.12 N/A 0.12 16.7 16.7
WI 0.28 0.32 0.00 25.0 25.0
NC 0.14 0.16 0.08 8.3 8.3
VA N/A N/A N/A 8.3 8.3
Hours 525.38 52.02 473.36 N/A N/A
Observations 47 14 33 102 50

Modeling of factors influencing mortality

The top-10 models based on the BIC analysis as well as the results of null and fully parameterized models are provided in Table 3. The most likely model indicated that mortality was a function of the additive relationship between 5-day cumulative temperature and net time. When net time was held constant at its average, every 1°C increase in the 5-day cumulative temperature increased the odds of mortality by 18.3% (OR = 1.18; 95% confidence interval [CI] = 1.06–1.31; Figure 4). With 5-day cumulative temperature held constant at its average, every 1-s increase in net time increased the odds of mortality by 2.1% (OR = 1.02; 95% CI = 1.00–1.04; Figure 4). Despite not being a top model, a model in which mortality was a function of the additive relationship between temperature at the time of catch and net time was explored to provide relevant and easy-to-understand information for anglers. When net time was held at 60 s, the mortality rate was 0.73% at a catch temperature of 22°C, 1.95% at 24°C, 5.27% at 26°C, 14.19% at 28°C, and 38.23% at 30°C.

TABLE 3. Bayesian information criterion (BIC) table for the top-10 Firth logistic regression models fitted to predict the probability of Muskellunge catch-and-release (C&R) mortality from the predictor variables (temperature, fight time, net time, fish length, and sex) based on 50 C&R events in 2020 and 2021. The outcomes of fully parameterized and null models are also included. Results include the number of predictor variables (K), BIC score, difference in BIC score between the given model and the best performing model (ΔBIC), and individual model weight (BIC weight). Abbreviations: Avg, average; Cumul_Temp, cumulative temperature; Temp, temperature.
Mortality model K BIC ΔBIC BIC weight
5-day Cumul_Temp + Net Time 3 48.31 0.00 0.25
4-day Cumul_Temp + Net Time 3 48.89 0.58 0.19
Post-3-day Cumul_Temp + Net Time 3 49.96 1.65 0.11
5-day Cumul_Temp × Net Time 4 50.79 2.48 0.07
4-day Cumul_Temp × Net Time 4 51.35 3.04 0.05
Avg Daily Temp + Net Time 3 51.71 3.40 0.05
5-day Cumul_Temp + Net Time + Length 4 51.89 3.58 0.04
4-day Cumul_Temp + Net Time + Length 4 52.45 4.14 0.03
Post-3-day Cumul_Temp × Net Time 4 52.68 4.37 0.03
Post-3-day Cumul_Temp + Net Time + Length 4 53.68 5.37 0.02
Fully parameterized 9 57.06 8.74 0.00
Null 1 65.00 16.69 0.00
Details are in the caption following the image
Fitted model utilizing Firth logistic regression, showing the effects of 5-day cumulative temperature on the probability of Muskellunge catch-and-release mortality when net time is held constant (top left), the effects of net time when 5-day cumulative temperature is held constant (bottom left), the effects of temperature at the time of catch when net time is held constant (top right), and the effects of net time when temperature at the time of catch is held constant (bottom right). Shaded regions represent ±95% confidence intervals.

Hooking location and RAMP scores

A RAMP score of 3 (all reflexes present) was associated with survival (OR = 0.02; 95% CI = 0.00–0.45), but a RAMP score of 1 or 2 (loss of some reflexes) was not a good predictor of survival or mortality, as the 95% CIs of their ORs overlapped 1.0 (RAMP score of 1: OR = 1.67, 95% CI = 0.49–5.65; RAMP score of 2: OR = 0.44, 95% CI = 0.10–2.03). There was also no association between hooking location (mouth: n = 44; gills: n = 3; tongue: n = 3) and the odds of mortality (the 95% CI for each hooking location overlapped 1.0). No fish were hooked in the esophagus.

Catch per unit effort and mortalities per 100 angler-hours

The CPUE data for the Virginia fish were incomplete and hence were excluded from analyses. The CPUE data were then based on 78 trips totaling 525.4 h, with 47 Muskellunge caught, resulting in an overall CPUE of 0.09 fish/h. The CPUE was lowest for the West Virginia ponds at 0.07 fish/h and highest for the Wisconsin pond at 0.28 fish/h (Table 2). When temperatures were less than 25°C, anglers caught 14 fish and the CPUE was 0.27 fish/h. The chances of catching a fish were almost four times greater for trips at temperatures less than 25°C than for trips conducted at warmer temperatures (CPUE = 0.07 fish/h; n = 33 fish; Table 2). This trend of reduced CPUE at temperatures exceeding 25°C was observed across all ponds at which catches above this threshold occurred. There was moderate evidence that the probability of catching a Muskellunge decreased as water temperatures increased (OR = 0.85; 95% CI = 0.72–1.01; Figure 5). The CPUE was negatively related to average daily water temperature (F1, 76 = 4.81, p = 0.018), but water temperature only explained a small proportion of the variance in CPUE (adjusted R2 = 0.05; Figure 5).

Details are in the caption following the image
Fitted Firth logistic regression (top) and linear regression (bottom) models, showing the relationship between average daily temperature (°C) and either the probability of catching a Muskellunge (%) or angler CPUE (fish/h). Data were collected from 78 trips conducted in 2020 and 2021. Shaded regions represent ±95% confidence intervals.

Mortalities per 100 angler-hours were calculated at four different temperature regimes. Trips conducted at temperatures less than 25°C had zero mortalities per 100 angler-hours. Trips conducted at temperatures between 25.00°C and 27.49°C had 4.98 mortalities/100 angler-hours. Trips conducted at temperatures between 27.5°C and 30.0°C had 2.48 mortalities/100 angler-hours. Trips conducted at temperatures greater than 30°C had 1.17 mortalities/100 angler-hours.

DISCUSSION

We utilized hatchery ponds as mesocosms to monitor post-C&R mortality of Muskellunge for 2 weeks and natural mortality of control fish for up to 4 months. Muskellunge were collected from waters across five different states (North Carolina, Maryland, Virginia, West Virginia, and Wisconsin). We noted a total mortality rate of 20.6%, a C&R mortality rate of 30%, and control mortality of 11.3% (Table 2). The C&R mortality value is in stark contrast to C&R mortality ranging from 0% to 5% as reported in previous studies, all of which were conducted in the northern range of the Muskellunge's distribution (i.e., Minnesota, Michigan, and Ontario; Strand 1986; Frohnauer et al. 2007; Landsman et al. 2011; Hessenauer et al. 2021). A portion of the C&R fish (n = 50) in our study were likely to have perished due to natural causes, regardless of being caught. If we apply the control mortality (11.3%) to our C&R group, then the adjusted C&R mortality would be 18.7%, which is still markedly higher than previous reports (Strand 1986; Frohnauer et al. 2007; Landsman et al. 2011; Hessenauer et al. 2021). We expected this because the previous studies were conducted at temperatures less than 25°C, similar to temperatures within the Muskellunge's realized thermal niche (22.3 ± 1.8°C; Cole and Bettoli 2014), and warm water has been demonstrated to increase the stress response up to death in some angled fish (Wilde et al. 2000; Arlinghaus et al. 2007; Boyd et al. 2010; Landsman et al. 2011; Gale et al. 2013). The ponds in the present study were shallow (maximum depth = 1.5–4.6 m) and had continuous flow, resulting in average differences of less than 1°C between surface and bottom temperatures (except in Wisconsin, where the mean temperature differential was 4.85°C). Our results suggest that water bodies with significant thermal refuge or that do not approach temperatures above 25°C would experience lower summer C&R mortality than our ponds.

We demonstrated increasing rates of Muskellunge C&R mortality with increasing temperatures, perhaps confirming the caution put forth by Heyob (2013) and other anglers who cease fishing at temperatures greater than 26.7°C. We discovered that mortality was strongly associated with the 5-day cumulative total temperature, suggesting that the temperatures experienced by fish prior to and after the angling event can affect the probability of mortality. This information can aid conservation-oriented anglers in determining when Muskellunge C&R mortality may limit their conservation goals. We encourage future studies to validate these results and to continue research on the possible consequences for Muskellunge populations.

The West Virginia ponds experienced the highest C&R mortality (42.9%), the lowest CPUE (0.03 fish/h in 2020; 0.15 fish/h in 2021), and the highest water temperatures. In contrast, the Wisconsin pond exhibited the lowest C&R mortality (0%) and the highest CPUE (0.28 fish/h) and developed a distinct thermocline (Figure 3). These results align with a concurrent C&R study on Stonewall Jackson Lake, West Virginia (thermally stratified), where C&R mortality rates were low (12%) and exploitation rates were high (33% of tagged fish) despite high surface water temperatures (>25°C) comparable to those experienced in our hatchery ponds (Jenkins 2022). Additionally, low catch rates and high mortality rates (33%) were observed in another concurrent study of C&R mortality of Muskellunge in the James River, Virginia. Even though fish had access to thermal refuge areas (creek mouths), the refuges were smaller and the temperatures were still greater than 25°C (Bauerlien et al. 2022). The results of these two studies in combination with the present paper suggest that Muskellunge with access to thermal refugia (<25°C) could be more susceptible to angling but less likely to die than fish caught in systems with no refuge or where refuge temperatures exceed 25°C. Thus, managers should treat their respective Muskellunge waters on a case-by-case basis and consider the habitat and environmental characteristics when implementing management decisions.

Previous C&R studies have focused on angling (fight) time as a physiological stressor, whereas few have reported the effects of time spent in the net (Barton and Iwama 1991; Muoneke and Childress 1994; Bartholomew and Bohnsack 2005; Arlinghaus et al. 2007). Our model suggests that within the range of our observations, fight time is a poor predictor of mortality but net time is a good predictor of mortality. It is possible that fight time was not included in a top model because the majority of longer fight times occurred in ponds/at times with the lowest temperatures. Thermally stressed fish may not have fought as hard or as long as the fish in cooler waters, as we found a negative correlation between fight time and temperature at the time of catch. Increased mortality rates for fish that were held in nets longer may be due to the suppression of the fish's ability to move water across the gills (i.e., by restricting opercular movement). The likelihood that a fish may injure itself increases with time, especially if the fish thrashes about during unhooking. Additionally, water temperatures are highest at the surface, where fish handling and processing occur. Minimization of both angling time and net time remains a best practice and is recommended for C&R angling (Brownscombe et al. 2017). Gasbarino (1986) suggested that Muskellunge captured by nonspecialized anglers (inexperienced, ill-equipped anglers) may exhibit higher levels of mortality due to the fish's large size, power, and rows of teeth. Experienced Muskellunge anglers possess large nets, hook cutters, fish gloves, and long pliers to aid them in safely landing and unhooking the fish. Without these tools, handling (net) times and angling duration can lead to increased fish stress and mortality. Generalist, nonspecialized anglers are effective at releasing smaller target species but may lack the handling skills and tools necessary to successfully release Muskellunge that are caught inadvertently. Therefore, utilizing the extreme observations of net time in the model may offer a sufficient proxy for the C&R mortality rates associated with nonspecialized anglers.

Reflex impairment (RAMP) has been shown to be related to mortality in several commercial fish species (Davis and Ottmar 2006); however, we were unable to accurately predict mortality with the loss of reflexes. There is evidence that survival increases with RAMP score (presence of reflexes), and only one fish that had a RAMP score of 3 (all reflexes present) was observed to perish after being released. There were no cases of fish exhibiting zero reflexes postrelease, indicating that the fish may not have been exhausted to extremes, which may explain why the RAMP scores were not related to mortality. Future studies evaluating the predictive power of RAMP scores should aim to (1) standardize the reflexes used when assessing impairment and (2) validate their applicability. Recreational anglers and professionals would benefit from having a reliable and rapid measure of fish condition that is easy to obtain and requires no special tools. As a measure of fish condition, RAMP has the potential to reduce instances of wanton waste if it is proven to be a reliable indicator of mortality. Additionally, anglers can keep diaries to report CPUE data, biotic and abiotic factors, RAMP scores, and recapture data if fish are equipped with external or PIT tags.

We found no significant relationships between hooking locations and mortality. This was most likely due to the low number of fish that were hooked in sensitive areas, such as the gills (n = 3), tongue (n = 3), and esophagus (n = 0). Therefore, our results are most applicable to fish that are hooked in the nonvital areas of the mouth. More studies that aim to understand the effects of hooking location on postrelease mortality of Muskellunge should be conducted.

Implementing species- and season-specific fishing regulations may often be problematic, especially on a case-by-case basis. Although some members of the Muskellunge angler community cease fishing at temperatures exceeding 26.7°C, this custom is not practiced by all anglers. Generalist anglers, harvest-oriented anglers, and other specialized angler groups like those targeting Largemouth Bass Micropterus salmoides are less likely to adhere to these rules due to little concern over C&R mortality with their target species (Kerns et al. 2016). Therefore, educating anglers who utilize Muskellunge fisheries—or other fisheries where Muskellunge may be caught incidentally—about the potential for increased mortality associated with high temperatures (>25°C) and the benefits of proper release equipment should be a high priority. Proper education and preparedness can help both specialized and generalist anglers to hopefully mitigate the factors influencing C&R mortality of Muskellunge.

The combination of C&R angling and management has resulted in significant increases in the abundance and size structure of Muskellunge populations (Casselman et al. 2017; Simonson 2017). As one of the goals of C&R management is producing trophy individuals, fully understanding the consequences of the practice is paramount to ensuring future success of the fishery. Brousseau and Armstrong (1987) noted that C&R mortality must be low for regulations to be effective and produce desirable results for the fishery. We observed elevated C&R mortality rates of Muskellunge during summer months, which could counteract the goals of C&R management strategies and negatively affect populations. However, increasing mortality with temperature coincided with decreasing catch rates, suggesting that the potential effects of warmwater angling may be offset by low catchability. Our metric of mortalities per 100 angler-hours elucidated which temperatures are most harmful to Muskellunge populations in the context of C&R mortality and angling susceptibility. At temperatures between 25.00°C and 27.49°C, we observed 4.98 mortalities/100 angler-hours. This is double the mortality observed at temperatures between 27.5°C and 30.0°C (2.48 mortalities/100 angler-hours) and over triple the mortality observed from trips with temperatures greater than 30°C (1.17 mortalities/100 angler-hours). No mortalities occurred when temperatures were less than 25°C. These discrepancies between mortalities per 100 angler-hours at different regimes highlight how fish susceptibility to angling and low catchability offset the increases in C&R mortality at higher temperatures. When waters are between 25.00°C and 27.49°C, Muskellunge catch rates remain relatively high but warmwater C&R mortalities are likely. This range coincides with the caution that Heyob (2013) and his contemporaries put forward when they suggested cessation of fishing at temperatures greater than 26.7°C. Volkoff and Rønnestad (2020) noted that feeding activity decreases when fish are thermally stressed and that feeding ceases at temperatures substantially below the critical maximum temperature for a given species. These increased rates of C&R mortality during warmwater periods may suggest that most of the annual mortality of Muskellunge occurs in the summer months during periods of extended thermal stress. We suggest that managers consider the characteristics (available refugia, number of high-degree days, temperature and DO profiles, and levels of exploitation) of each water body under their management to assess any potential summer C&R regulation needs for Muskellunge.

Although these results do provide useful information for managers and anglers, our findings must be interpreted within the constraints of the study design. Muskellunge were collected from 12 different waters across five states and were housed in ponds with individuals from their respective states. We pooled data across all ponds to increase our statistical power, disregarding site variability for both collection sites and ponds. Populations of fish used in the study may have unaccounted-for pre-existing conditions or comparatively poor fitness that could impact their relative probability of C&R mortality. Five of our six control deaths occurred in the Wisconsin pond, whereas zero C&R deaths were recorded for that pond. This could indicate that the Wisconsin pond contained conditions that were detrimental to fish health, such as a disease outbreak or an algal bloom, thus skewing the results. Additionally, 61.9% of all mortalities occurred in the West Virginia ponds (n = 13), with one being a control fish. However, West Virginia fish made up 51% (n = 52) of the total fish in the study; therefore, while there was a high percentage of deaths in the West Virginia ponds, it was not necessarily disproportionate. West Virginia ponds did reach and sustain higher temperatures than the other locations, perhaps providing stronger evidence that thermal stress is an important factor in determining C&R mortality. Nonetheless, we cannot rule out the possibility that the West Virginia populations held a greater propensity to suffer C&R mortality than other populations in the study or that the West Virginia ponds otherwise provided poorer habitat.

Fish were given access to different forage species dependent on facility biosecurity and availability, which may have affected the results. Different species of forage fish could require larger energy expenditures to capture and consume. Muskellunge in ponds where forage was easier to consume could have had larger energy reserves to combat any detrimental effects of C&R angling, which would have confounded the results. Additionally, fish may have preferred one prey species over another and may have been less likely to attack lures, thereby skewing CPUE results between ponds. Standardizing the forage base to one species can combat this and could reduce confounding variables in future studies.

Another factor to consider is the acclimation period. Muskellunge were acclimated at different times of the year, and the acclimation period extended from 2 weeks to 2 months depending on the collection dates. Muskellunge caught by anglers early in the study may have had less time to acclimate to the new pond environments and may been more susceptible to C&R mortality than their counterparts who had more time to adjust. This could result in either a higher or lower probability of mortality based on whether the fish had regained homeostasis after the stress from being collected, transported, and housed in a foreign environment. Such a result would be most prominent during the first month of angling events, as fish caught toward the end of the experiment would have had sufficient time to acclimate after being housed in the ponds for multiple months. In our preliminary exploration of model parameters, we concluded that acclimation period was not a significant predictor of mortality based on logistic regression. However, it is still worth considering in future experiments to standardize and control for differences in acclimation periods to prevent confounding elements.

Over two-dozen different anglers participated in the C&R angling portion of the study. A random or fixed predictor variable of angler was considered to account for this, but it was not included in final analyses. Anglers varied by pond and, with the exception of the lead author, no anglers participated in angling events at multiple ponds included in the study. In our preliminary analysis, we found that pond was a poor predictor of mortality and did not perform better than a null (intercept-only) model or a fully parameterized model. Therefore, pond as a categorical variable already accounted for differences in Muskellunge C&R mortality between anglers and was shown to not be significant. Additionally, the lead author accounted for 54% of all catches, whereas the other participants generally caught one to two fish. With such a small sample size and large discrepancies between anglers, we decided not to explore the use of angler as a predictor variable and reduce our statistical power. Future studies can control for this by only using data from a single angler.

Lastly, the use of nonangled individuals as a control population, although logical, does have drawbacks. There exists a wide range of studies demonstrating that vulnerability to angling is influenced by factors such as locomotor activity, hormonal responsiveness to stress, metabolic activity, and individual behavior of fish species (Cooke et al. 2007; Louison et al. 2017; Louison 2018; Koeck et al. 2019). The C&R and control individuals included in our pond study may have had differences that affected their vulnerability to angling and how they handled abiotic and biotic stressors. Fish with high metabolic rates and increased hormonal responses to stress would likely be more susceptible to the detrimental effects of elevated temperatures and C&R angling. Assessing baseline physiology prior to stocking fish into the ponds and observing changes in physiology after a C&R event could elucidate whether factors associated with vulnerability to angling were driving mortalities. A physiological component was considered early in the study's proposal but was not included due to the elevated temperatures at which the experiment was performed. Respirometry and drawing of blood at such elevated temperatures would almost certainly initiate a stress response resulting in death. Additionally, the methods required to assess these factors are not representative of what a fish experiences during a C&R event in the wild and would make our findings less applicable to recreational C&R angling. The potential for mortality to vary based on physiological differences or behavior was beyond the scope and design of our study. Our nonangled control fish experienced the same environmental conditions within the ponds as our C&R fish and thus remained the best proxy for a control that we could accomplish. We implore researchers in future studies to build on our work and address the constraints outlined above.

Our study was primarily focused on Muskellunge populations in the southern part of the species' distribution, except for Wisconsin. However, the results of this study are broadly applicable across the entire range of the Muskellunge when considering the probable increase in water temperatures induced by climate change (Maberly et al. 2020; Woolway et al. 2020). As our planet warms, the frequency of water temperatures exceeding 25°C and the duration of these warmwater periods will increase. Understanding how fish populations will react to increased average temperatures—and the implications of these temperature increases—is imperative for successful fisheries management. The results of this study suggest that C&R mortality will increase alongside increasing temperatures and may necessitate regulatory or management interventions. Studies exploring the ability of Muskellunge populations in the northern extent of the species' distribution to withstand elevated temperatures are necessary to prepare for future climate scenarios.

ACKNOWLEDGMENTS

This project would not have been possible without the cooperation of volunteer anglers and fisheries management agency personnel. This work was supported by the U.S. Department of Agriculture's National Institute of Food and Agriculture (McIntire–Stennis Project WVA00048, Accession Number 0195351) and the West Virginia Agricultural and Forestry Experiment Station. Funding and in-kind support were also provided by the West Virginia Division of Natural Resources, Hugh C. Becker Foundation, Trooper Eric Workman Foundation, Muskies Inc. (National), Jim Moore and West Virginia Chapter 09 of Muskies Inc., Toby Tester and Virginia Chapter 76 of Muskies Inc., Christine Densmore and the USGS EESC, Scott Loftis and the North Carolina Wildlife Resources Commission, and Tim Boyer and the Western North Carolina Musky Club.

    CONFLICT OF INTEREST STATEMENT

    We declare there is no conflict of interest associated with this work.

    ETHICS STATEMENT

    We declare there are no ethical concerns associated with this work.

    DATA AVAILABILITY STATEMENT

    The data from this study is available from the lead author upon request.