Conservative regulations and stocking in inland recreational fisheries have been used as conservation tools to rehabilitate poorly recruiting, overexploited, or collapsed populations (Rypel et al. 2018), to keep fisheries in a safe operating space (Carpenter et al. 2017), in biomanipulations to control invasive species (e.g., rusty crayfish Orconectes rusticus; rainbow smelt Osmerus mordax) (Krueger and Hrabik 2005;Hein, Vander Zanden, and Magnuson 2007;Gaeta et al. 2015), to increase angler catch rates, and to create put-and-take fisheries. Voluntary catch-and-release practices have also been promoted by managers, stakeholder groups, and anglers for certain species (e.g., muskellunge Esox masquinongy, black basses Micropterus spp.) with the goal of conservation. Likewise, anglers may be obligated to catch-andrelease (CR) in certain fisheries managed for trophy potential because only a small subset of the population is legally harvestable. Often, CR practices are promoted under the assumption that catch rates, population size structure, and trophy potential will improve as a result. Because fisheries management is aimed to maintain a balance between sustainability, catch rates, and trophy potential in harvested fisheries, fish community imbalances may occur when anglers choose to CR certain fish species and/or species-specific harvest regulations are ultra-conservative (Walters et al. 2000). Further, regulations requiring harvest are not typically used in North America; however, this practice is more common in Europe (Arlinghaus 2007). Angler preferences for CR are generally out of managerial control, therefore these ideals may lead to imbalances in fish communities and not meet angler expectations in the long-term; particularly in fisheries managed for multiple species (Hansen et al. 2015;Miranda et al. 2017). Here, a review is provided on the influences of CR practices in north temperate inland lakes fisheries with primary emphasis on largemouth bass M. salmoides, smallmouth bass M. dolomieu, muskellunge, walleye Sander vitreus, and panfish (sunfishes Lepomis spp., crappies Pomoxis spp., yellow perch Perca flavescens). A brief literature review on CR research is provided, trends in CR behavior over time are identified, consequences of CR behavior on catch rates, recruitment, relative abundance, size structure, growth, and trophy potential are discussed, and recommendations for better balancing fish communities in the future is provided. This review suggests that nearly exclusive CR behavior and ultra-conservative harvest regulations may not achieve the perceived benefits of such practices to meet angler, stakeholder, and manager expectations.
Angler CR is a relatively recent socially-driven phenomenon for inland fisheries that has been promoted to conserve charismatic and previously harvested species, and to improve population size structure and trophy potential. For example, stakeholder groups such as the Bass Anglers Sportsman Society (B.A.S.S.) and Muskies, Inc. have promoted CR as a mechanism to increase trophy potential in the black basses and muskellunge. Prevalence of CR has also been commensurate with the increasing popularity and commercialization of fishing tournaments over time because tournament sponsors often require and promote CR. Nevertheless, many inland fisheries and anglers remain harvest-oriented (e.g., walleye, bluegill L. macrochirus, black and white crappie P. nigromaculatus and P. annularis, yellow perch) (Gaeta et al. 2013). Catch-and-release is assumed to be beneficial for population conservation and increasing population size structure and catch rates in line with angler expectations. As a result, most studies of CR have focused on measuring and reducing hooking mortality to determine under what conditions and for what species CR will be most beneficial. Countless studies have been conducted on this topic across a range of species and habitats (e.g. Cooke and Suski 2005;Arlinghaus et al. 2007;Cooke and Schramm 2007;Arlinghaus et al. 2012;Kerns, Allen, and Harris 2012). In general, most studies examining hooking mortality of black basses and muskellunge have focused on hooking location on the fish, type of hook, water temperature, physiological stress associated with angling and tournament weigh-in procedures, and handling procedure effects. Another body of literature has focused on CR influences on nesting black basses. Overall, estimates of hooking mortality in the black basses have been highly variable (range 0-98%). Studies conducted in laboratories or net pens have shown higher and more variable hooking mortality (Schramm et al. 1987;Kwak and Henry 1995;Weathers and Newman 1997;Wilde 1998;Pollock and Pine 2007;Wilde and Pope 2008), whereas mark-recapture studies (where released fish are able to choose an unconfined natural habitat during the refractory period) have shown lower and less variable hooking mortality (Cox 2000;Cline et al. 2012;Sass et al. 2018). Tournament weigh-in procedures (holding in livewells for up to eight hours, exposure to air) have been shown to cause significant physiological stress in the black basses (Edwards et al. 2004a, Suski et al. 2005). Elevated water temperatures and fight time have also been shown to increase black bass hooking mortality rates (Edwards et al. 2004a, 2004b, Keretz et al. 2018). Along with physiological and genetic effects, numerous studies have also suggested that CR of nesting black basses has negative numerical population-level consequences (e.g. Phillip et al. 2009;Sutter et al. 2012;Parkos et al. 2013). Although individual nest success may be compromised by CR of nesting black basses (Kieffer et al. 1995;Ostrand, Cooke, and Wahl 2004;Suski and Philipp 2004;Trippel et al. 2017), no numerical population-level negative consequences of this practice have been documented. For example, Jackson et al. (2015) found that CR for smallmouth bass had no effect on recruitment, Trippel et al. (2017) observed no effect of bed-fishing on recruitment and overall reproductive success of Florida bass M. salmoides floridanus, and Zipkin et al. (2008) showed strong compensatory recruitment of smallmouth bass subjected to experimental overharvest. Shaw and Allen (2016) found that successful Florida bass broods had to drop below about 5/ha before direct declines in recruitment were observed. In a northern Wisconsin lake, Sass et al. (2018) observed an increase in largemouth bass abundance when the population was subjected to CR for five years suggesting increased recruitment. Therefore, hooking mortality and bed fishing do not appear to have direct negative consequences on the overall abundance of black bass populations. For muskellunge, hooking mortality was found to be low in specialized anglers (Landsman et al. 2011). Further, Shaw, Sass, andEslinger (2019) showed that muskellunge hooking mortality was likely negligible using data from a long-term compulsory creel census on a northern Wisconsin lake. Few studies have focused on the potential negative implications of CR on population size structure and trophy potential.
Catch-and-release practices have increased in prevalence over time for black bass and muskellunge anglers in Wisconsin. Wisconsin statewide largemouth bass release rates have exceeded 80% since the early 1990s and have averaged about 95% since 2000 (Hansen et al. 2015) (Figure 1). Catch-and-release for muskellunge anglers in Vilas County, Wisconsin, increased in the mid-1980s and release rates have exceeded about 90% since the mid-2000s (Gilbert and Sass 2016) (Figure 1). On Escanaba Lake, Wisconsin, release rates have exceeded 75% since the mid-1980s despite no size limit, bag limit, or closed season on muskellunge. Escanaba Lake muskellunge release rates have averaged >95% since 2010 (Eslinger et al. 2017) (Figure 1). Further, Shaw, Sass, and Eslinger (2019) defined two distinct breakpoints (1995,2011) in muskellunge harvest on Escanaba Lake during 1956-2016 illustrating a transition in the fishery from harvest-to catch-and-release-oriented. During the periods of 1956-1994, 1995-2010, and 2011-2016, muskellunge harvest declined on Escanaba Lake and averaged about 30, 7, and 1 fish per year, respectively (Shaw, Sass, and Eslinger 2019). The reduction in muskellunge harvest was unrelated to directed muskellunge angler effort, which was consistent during these time periods (Shaw, Sass, and Eslinger 2019). Using an angler diary survey, Gaeta et al. (2013) recorded release rates of 99% for muskellunge and 97% for black bass in northern Wisconsin, with CR being the most common reason for release. Increases in CR of muskellunge in Canada have also been documented (Mosindy and Duffy 2007;Landsman et al. 2011). Similar patterns of increasing CR for black bass over time have been observed across much of their range in North America including the states of Alabama, California, Connecticut, Florida, Georgia, Iowa, Idaho, Michigan, Minnesota, Mississippi, Tennessee, Texas, and Wisconsin (Allen, Walters, and Myers 2008;Isermann, Maxwell, and McInerny 2013;Kerns, Allen, and Hightower 2016;Miranda et al. 2017;Hessenauer et al. 2018). Considerable evidence exists to suggest that black bass and muskellunge anglers have transitioned from harvest-oriented to nearly exclusive CR regardless of regulations. This current "social fishery norm" has rarely been evaluated to test whether the practice of CR is achieving manager, stakeholder, and angler goals and/or creating imbalances in fish communities under an ecosystem-based fisheries management framework (but see Hansen et al. 2015;Miranda et al. 2017;Hessenauer et al. 2018).
One perceived goal of species-specific CR practices is an improvement in angler catch rates. The achievement of this goal appears to be species-specific and related to fish learning behavior. For the black basses, the increasing prevalence of CR has resulted in improvements in fishery-independent and angler catch rates over time (Hansen et al. 2015;Rypel et al. 2016) (Figure 2). Increases in catch rates have also coincided with a decline in largemouth bass individual growth rates, population size structure, and trophy potential (Hansen et al. 2015;Rypel et al. 2016) (Figure 2). Fishery-independent and -dependent catch rate data for muskellunge during the transition to primarily CR is sparse due to the inherent elusiveness of the species and a lack of targeted fishery-independent surveys. Nevertheless, muskellunge compulsory creel census angler data and fishery-independent surveys on Escanaba Lake have suggested that muskellunge angler catch rates (1 muskellunge/33 hours of directed angler effort) have not changed despite stable directed angler effort and a reduction in muskellunge abundance over time (Eslinger et al. 2017;Shaw, Sass, and Eslinger 2019). Statewide fishery-independent surveys of muskellunge in Wisconsin have shown opposing trends in fyke net and electrofishing catch per unit effort (Rypel et al. 2016). Fyke net CPUE has increased, whereas electrofishing CPUE has decreased during the recent CR period (Rypel et al. 2016). This may suggest that larger, mature individuals, which are more vulnerable to spring fyke netting, have increased in abundance, whereas smaller muskellunge, which are more vulnerable to fall electrofishing, may have declined in abundance. A similar pattern for smaller muskellunge has been observed in Escanaba Lake; however, adult abundance has declined over time (Eslinger et al. 2017). Walleye anglers are harvest-oriented (Kempinger and Carline 1977;Gaeta et al. 2013;Hansen et al. 2015). During 1946-2002, there was no size limit, bag limit, or closed season on walleye in Escanaba Lake (Kempinger and Carline 1977;Haglund, Isermann, and Sass 2016). In 2003, a 711 mm minimum length limit and a daily bag limit of one fish (i.e., essentially a CR regulation) was enacted for walleye on Escanaba Lake (Haglund, Isermann, and Sass 2016). Since 2003, no walleye have been legally harvested from Escanaba Lake by anglers, with only five walleye being legally harvested by a Chippewa tribal member in December, 2018. Despite an essentially CR regulation for walleye, adult density and angler catch rates did not differ between the exploited and unexploited time periods (Haglund, Isermann, and Sass 2016). Plausible mechanisms for angler catch rates not being or negatively correlated with CR stems from learning behavior in fishes. Learning behavior, lure avoidance, and the type of bait (artificial vs. live) have all been shown to reduce angler catch rates across a number of species including northern pike Esox lucius (Beukema 1970;Kuparinen, Klefoth, and Arlinghaus 2010), largemouth bass (Hackney and Linkous 1978;Hessenauer et al. 2016;Louison et al. 2019), rainbow trout Onchrhynchus mykiss (Askey et al. 2006), and walleye (Bailey et al. 2018). Therefore, effects of voluntary or mandatory CR practices on angler catch rates appear to be variable, species-specific, and related to fish learning behavior. (Hansen et al. 2015). Reprinted with permission as G.G. Sass was a coauthor of Hansen et al. (2015).
Catch-and-release angler behavior is presumed to increase species-specific abundances over time. According to generalized Ricker and Beverton-Holt stock recruitment relationships, juvenile recruitment would be assumed to decline or stabilize with increased adult stock size (Beverton and Holt 1957;Ricker 1975). For black basses, prevalence of CR has either improved or had no effect on recruitment as strong stock-recruitment relationships are lacking for the black basses (Zipkin et al. 2008;Allen et al. 2011;Parkos et al. 2013;Sass et al. 2018). In Wisconsin, fisheries-dependent and -independent data have shown that catch rates and relative abundances of largemouth bass have increased over time with associated declines in the mean length of age-6 fish (Hansen et al. 2015) (Figure 2). Sass et al. (2018) suggested that largemouth bass recruitment increased after a small, northern Wisconsin lake population was subjected to CR angling for five years. Jackson et al. (2015) and Trippel et al. (2017) showed that CR had no effect on smallmouth and Florida bass recruitment, respectively. Despite previous research suggesting that voluntary or mandatory CR of nesting black bass is detrimental to individual nest success, no evidence exists to support negative numerical population-level consequences. That said, climate change and associated increases in growing degree days have been shown to favor largemouth bass recruitment and relative abundance (Hansen et al. 2017). Muskellunge have been shown to exhibit a Ricker stock-recruitment relationship (Eslinger, Dolan, and Newman 2010). Catch-and-release would be expected to increase adult muskellunge abundances and therefore reduce juvenile survival unless the population was supplemented by stocking. Indeed, muskellunge recruitment has declined to nearly undetectable levels on Escanaba Lake over time (Eslinger et al. 2017) (Figure 3). Adult muskellunge abundance has also declined significantly over time suggesting that reproductive senescence may occur in muskellunge and/or that there is a defined carrying capacity for relatively unexploited muskellunge populations in the absence of stocking (Eslinger et al. 2017;Shaw, Sass, and Eslinger 2019) (Figure 3). Given observations from the relatively unexploited Escanaba Lake muskellunge population, CR may weaken resilience of the population to anthropogenic, abiotic, and biotic perturbations and/ or the population may be seeking a new low adult density equilibrium like other unexploited fisheries in the absence of stocking (Johnson 1976;Goedde and Coble 1981;Anderson et al. 2008). Like muskellunge, elimination of legal harvest of walleye on Escanaba Lake has resulted in low, stable recruitment (Haglund, Isermann, and Sass 2016). For walleye, boom and bust recruitment events seem to be driven by harvest; however, exploitation effects on recruitment appear to be population-specific (Tsehaye, Roth, and Sass 2016;Sass and Shaw 2018). Cultivation/depensation effects on recruitment may prevent populations from recovery due to food web changes even if exploitation is limited or eliminated (Walters and Kitchell 2001). Ultimately, reproductive strategy (nest guarding, centrarchids; broadcast spawning with no parental care, muskellunge/walleye), exploitation, and cultivation/ depensation influences on food web structure may dictate the relative influence of CR on recruitment.
The practice of species-specific CR is assumed to increase the abundance of that species. Although this may occur for some species, abundance responses to CR practices may also be species-specific and dictated by aquatic ecosystem carrying capacity/productivity and the diversity of the fish community (Johnson 1976;Goedde and Coble 1981;Anderson et al. 2008;Eslinger et al. 2017). For largemouth bass, CR has generally led to increases in abundance regardless of ecosystem productivity, diversity of the fish community, and/or fishing for nesting males (Hansen et al. 2015;Rypel et al. 2016;Sass et al. 2018) (Figure 2). During 1991-2011 (a period of high CR for largemouth bass in Wisconsin), statewide electrofishing CPUE increased from about 15 to 30 > 203 mm bass/ hr and from about 2 to 5 > 356 mm bass/hr (Hansen et al. 2015) (Figure 2). Over a 48-year period beginning in the 1960s, Rypel et al. (2016) showed that largemouth and smallmouth bass electrofishing CPUE in Wisconsin increased from about 10 to 32 and 3 to 18 bass/hr, respectively. Sass et al. (2018) observed a significant increase in largemouth bass abundance in a northern Wisconsin lake subjected to five years of experimental CR fishing. Using mark-recapture techniques, largemouth bass density increased from about 60 to 115 bass/ha (Sass et al. 2018). For largemouth and smallmouth bass, CR (regardless of bed fishing) appeared to have resulted in improved recruitment leading to greater overall abundances (Rypel et al. 2016;Sass et al. 2018). Although CR has resulted in increased abundances of largemouth and smallmouth bass, abundance increases and overall carrying capacity appeared greater for largemouth than smallmouth bass. Despite the prevalence of CR for both species, abundance increases observed in largemouth bass were about double that of smallmouth bass (Rypel et al. 2016). This may suggest that smallmouth bass are more specialized in their spawning habitat preferences and diet, and may be more territorial during spawning than largemouth bass (Becker 1983;Rejwan et al. 1997). In small, northern Wisconsin lake basins, largemouth bass nest densities ranged from 52-129 nests/km of shoreline and nests were associated with coarse woody habitat (CWH), boulders, fine woody habitat, human structures, and beaver structures (Weis and Sass 2011). Lawson, Gaeta, and Carpenter (2011) also noted that largemouth bass nests were more associated with CWH in high CWH:low lakeshore residentially developed northern Wisconsin lakes, with nests found to be significantly deeper in low CWH:high lakeshore residentially developed lakes, and that docks did not serve as a surrogate for CWH. To support this assertion of smallmouth bass being more spawning habitat specialists than largemouth bass, Rejwan et al. (1997) found that smallmouth bass nests were patchily distributed with high density nesting areas remaining stationary across years in Lake Opeongo, Ontario, Canada. Smallmouth bass nest densities ranged from 0-25 nests/km of shoreline at the whole-lake scale and from 0-7 nests/100 m at the 100 m scale (Rejwan et al. 1997). These findings may suggest that largemouth bass are naturally more abundant than smallmouth bass, thus explaining differences in recruitment and the associated level of abundance increases observed when their fisheries are predominantly CR. In addition, smallmouth bass may be subjected to higher hooking mortality rates than largemouth bass (Edwards et al. 2004a(Edwards et al. , 2004b) ) and climate change may favor largemouth bass over smallmouth bass (Hansen et al. 2017).
Catch-and-release of muskellunge over the past 30 years does not appear to have caused an increase in Wisconsin statewide or Escanaba Lake muskellunge relative abundances. Rypel et al. (2016) 3). These results suggest that muskellunge abundance may be relatively unresponsive to CR. Muskellunge are a long-lived, apex piscivore and naturally exist at low densities in the absence of stocking. In Wisconsin CR muskellunge fisheries that are not stocked, average muskellunge density is about 0.1 fish/ha (Eslinger et al. 2017). Therefore, carrying capacities for muskellunge may be naturally low and CR may not result in increased total mortality rates in the absence of exploitation, particularly when hooking mortality is low (Landsman et al. 2011;Shaw, Sass, and Eslinger 2019). In contrast, natural mortality rates in older largemouth bass may increase in CR fisheries allowing recruitment and overall abundances to increase at the expense of reductions in population size structure (Sass et al. 2018). For muskellunge, CR may not result in elevated abundances because fishing mortality is the strongest component of total mortality (Shaw, Sass, and Eslinger 2019). Therefore, unexploited muskellunge populations tend to show improved size structure, a stable age distribution, high adult survival, and low recruitment (Eslinger et al. 2017;Shaw, Sass, and Eslinger 2019).
Walleye anglers are harvest-oriented, therefore abundance data on unexploited or predominantly CR fisheries is rare (Gaeta et al. 2013). Escanaba Lake is well studied and transitioned from a no length limit, bag limit, and closed season walleye fishery to an unexploited walleye fishery beginning in 2003 (711 mm minimum length limit, daily bag limit of one fish, no closed season) (Haglund, Isermann, and Sass 2016). The expectation of increased walleye density with the removal of exploitation did not occur outside of an increase in female density; adult and male density did not change between the exploited and unexploited time periods (Haglund, Isermann, and Sass 2016). Despite observed reductions in adult walleye density and an associated release of densitydependent constraints on growth and maturation rates with exploitation (Schueller et al. 2005;Schueller, Hansen, and Newman 2008;Schmalz et al. 2011;Tsehaye, Roth, and Sass 2016;Sass and Shaw 2018), the elimination of exploitation may not lead to marked improvements in walleye density as carrying capacities for this species may be constrained similar to muskellunge (Mrnak et al. 2018). Low natural mortality rates may act to stabilize species-specific fish abundances despite CR when hooking mortality is low and fishing mortality is the primary component of total mortality (Payer, Pierce, and Pereira 1989;Hansen, Fayram, and Newman 2011;Landsman et al. 2011;Tsehaye, Roth, and Sass 2016;Mrnak et al. 2018;Sass et al. 2018;Shaw, Sass, and Eslinger 2019). Abiotic and biotic factors such as water temperature and forage availability have been shown to be predictors of walleye and muskellunge recruitment (Madenjian et al. 1996;Hansen et al. 1998;Beard, Hansen, and Carpenter 2003;Eslinger, Dolan, and Newman 2010;Shaw, Sass, and VanDeHey 2018), whereas stock-recruitment relationships for Florida, largemouth, and smallmouth bass remain elusive (Zipkin et al. 2008;Allen et al. 2011;Parkos et al. 2013). This may highlight an additional difference in species-specific abundance responses to CR. Parental guarding species such as largemouth bass may be more responsive to CR, whereas recruitment of broadcast spawners such as muskellunge and walleye may be driven by highly variable abiotic and biotic factors. Parental guarding smallmouth bass abundance responses to CR may lie in between those of largemouth bass and walleye/muskellunge due to the requirement of specific, and sometimes limited, spawning habitat (Rejwan et al. 1997). Nevertheless, intentional overharvest of a smallmouth bass population resulted in strong compensatory recruitment suggesting that CR (including bed fishing) may have little to no influence on population-level recruitment in this species (Zipkin et al. 2008;Jackson et al. 2015) like Florida and largemouth bass (Allen et al. 2011;Shaw and Allen 2016).
Catch-and-release and size structure, growth, and trophy potential Anglers, stakeholders, popular articles, the scientific literature, and managers have often promoted CR to improve size structure, growth rates, and trophy potential in fish populations. Species-specific levels of density-dependence and variability in aquatic ecosystem productivity can act in concert to decrease the probability of reaching these goals (Sass and Kitchell 2005;Hansen et al. 2015;Gilbert and Sass 2016;Pedersen et al. 2018;Noring et al. 2019). Increased abundances of largemouth bass due to CR, increased growing degree days, and lake warming due to climate change in north temperate inland lakes (Rypel et al. 2016;Hansen et al. 2017) have resulted in reductions in size structure and individual growth rates, with declines in maximum growth potential. Proportional size distributions (PSD-Q and PSD-P) (Neumann et al. 2012) declined in a small northern Wisconsin lake where largemouth bass were experimentally subjected to five years of CR angling (Sass et al. 2018). In heavily exploited largemouth bass populations with relative low population sizes, CR mortality has also been shown to negatively effect population size structure (Hessenauer et al. 2018). During 1944-2012, the mean length of largemouth bass in Wisconsin increased significantly, whereas mean maximum length was unchanged; this time period also corresponded with a statistically significant increase in largemouth bass electrofishing CPUE suggesting higher relative abundances over time (Rypel et al. 2016). According to Rypel et al. (2016), mean and mean maximum length of largemouth bass were about 250 and 450 mm, respectively. In contrast, mean length of age-6 largemouth bass declined from about 381 to 330 mm during 1992-2011 suggesting densitydependent effects on growth (Hansen et al. 2015) (Figure 2). Largemouth bass size structure and individual growth rates have declined as a result of CR, protection during the spawning season, and associated increases in abundance. As a result, maximum growth potential has also been reduced.
Although relative abundances of smallmouth bass in Wisconsin lakes have increased significantly over time, this species appears to be more resilient to density-dependent constraints on growth (Rypel et al. 2016). Mean length and mean maximum length for smallmouth bass in Wisconsin increased significantly during 1944-2012 (Rypel et al. 2016). Mean length for smallmouth bass increased from about 200 to 250 mm and mean maximum length has increased from about 350 to 400 mm (Rypel et al. 2016). As discussed above, smallmouth bass may naturally be a lower density species than largemouth bass. As such, density-dependent constraints on growth have not been as strong and CR (and potentially protection during the spawning season) has increased abundances and maximum growth potential for smallmouth bass in the midwestern USA.
In general, CR in muskellunge fisheries has led to improvements in size structure (Rypel et al. 2016;Eslinger et al. 2017); however, further improvements in size structure appear to be capped by densitydependent constraints on growth, continued stocking, high minimum length limits, and low daily bag limits to reduce maximum growth and trophy potential (e.g., muskellunge >1219 mm and >1270 mm) (Gilbert and Sass 2016;Shaw, Sass, and Eslinger 2019). For example, muskellunge PSD-762 mm, PSD-965 mm, and PSD-1067 mm increased significantly on Escanaba Lake during 1987-2014; a period when CR increased (Eslinger et al. 2017;Shaw, Sass, and VanDeHey 2018) (Figure 4). Similarly, mean length and mean maximum length of muskellunge in Wisconsin have rebounded over time presumably due to CR (Rypel et al. 2016). In Wisconsin, mean length and mean maximum length of muskellunge captured by electrofishing has increased from about 600 to 800 mm and 1000 to 1200 mm, respectively. Faust et al. (2015) found that mean asymptotic lengths of female and male muskellunge in northern Wisconsin were 1265 mm and 1102 mm, respectively, whereas minimum ultimate lengths were lower at 1143 mm for females and 864 mm for males. Faust et al. (2015) concluded that female maximum growth potential was commensurate with trophy muskellunge management in Wisconsin (1270 mm minimum length limit and a daily bag limit of one fish) and adequate to meet angler desires. This review suggests that the frequency of trophy muskellunge has been compromised over time by CR, trophy regulations, density-dependent constraints on growth, and continued stocking in naturally reproducing muskellunge populations (Gilbert and Sass 2016;Shaw, Sass, and Eslinger 2019). Using data from a long-term muskellunge fishing contest in Vilas County, Wisconsin, Gilbert and Sass (2016) observed a 102 mm and 4.2 kg decline in the mean length and weight of the top ten muskellunge registered in the contest during 1964-2010, which encompassed the transition of muskellunge anglers from harvest-oriented to CR (Eslinger et al. 2017;Shaw, Sass, and Eslinger 2019) (Figure 4). Mean length of the top ten muskellunge registered in the contest declined from about 1327 mm in 1964 to about 1233 mm in 2010 (Gilbert and Sass 2016) (Figure 4). Similarly, mean weight of the top ten muskellunge registered declined from about 18 kg to 14 kg between 1964 and 2010, respectively (Figure 4). Therefore, CR, highly conservative regulations, and continued stocking may be artificially overinflating muskellunge densities, reducing trophy growth potential, and not achieving angler expectations in these fisheries (Gilbert and Sass 2016;Shaw, Sass, and Eslinger 2019). Given the current CR in muskellunge fisheries, adoption of liberal regulations (e.g., no closed season, no bag limit, and no minimum length limit) and reductions or elimination of stocking would likely better serve angler desires (Gilbert and Sass 2016;Eslinger et al. 2017;Sass, Rypel, and Stafford 2017;Shaw, Sass, and Eslinger 2019).
Walleye anglers are predominantly harvest-oriented (Gaeta et al. 2013); however, certain anglers also desire and focus on trophy walleye opportunities. In instances where there has been little to no exploitation on a walleye fishery (e.g. Escanaba Lake, 2003-present), size structure, growth, and trophy growth potential responses have been like those observed for muskellunge under nearly exclusive CR (Haglund, Isermann, and Sass 2016). Walleye angler and stakeholder expectations would anticipate a quality fishery with high potential for producing trophies. Because walleye growth rates have been shown to be weakly density-dependent and more related to ecosystem productivity (Sass and Kitchell 2005;Pedersen et al. 2018;Noring et al. 2019), reductions in walleye exploitation through CR may improve size structure, but would not be predicted to improve maximum growth potential. Mean length of males increased significantly after the elimination of legal harvest of walleye on Escanaba Lake; however, mean length of females and PSD-510 mm did not differ between the exploited and unexploited time periods (Haglund, Isermann, and Sass 2016). Interestingly, density-dependent responses of walleye in growth (increased), maturation schedule (reduced), and productivity (reduced) have been clear under high exploitation scenarios; however, responses to a lack of exploitation appear to be capped based on the carrying capacity of the system similar to muskellunge (Scheuller et al. 2005(Scheuller et al. , 2008;;Schmalz et al. 2011;Rypel et al. 2018;Sass and Shaw 2018). Collectively, fish responses to CR appear to be species-specific and may not align with angler desires.
Recently, attempts to reduce panfish (i.e., bluegill, crappie, yellow perch) harvest through conservative daily bag limits and seasonal restrictions have been experimentally implemented on several Wisconsin lakes as a tool to improve size structure and provide the same amount of fillet yield to harvest-oriented anglers in these fisheries (Rypel 2015;Lyons et al. 2017). This review questions whether the promotion of CR for panfish and the implementation of more conservative harvest regulations is warranted given the observed 20 mm mean length improvement in bluegill population size structure (Rypel 2015) in lieu of the unintended potential consequences of stunting and imbalances in fish communities. Evaluation of Wisconsins experimental panfish regulations are ongoing and will shed light on the question of whether size structure improvements were achieved or if undesired consequences (stunting, imbalances in fish communities) were more prevalent. Again, promotion of CR and the perceived benefits are likely species-specific and angler desires and single-species management goals may conflict with overall ecosystem-based management goals in fisheries managed for multiple species (Hansen et al. 2015).
Responses of fishes to CR are species-specific Acknowledgement of species-specific differences to CR have been discussed previously; however, it has been in the context of what anglers can do to minimize hooking mortality and sublethal effects when information for a species is lacking (Cooke and Suski 2005). This review has shown that realized hooking mortality is lower than previously thought for several species when mark-recapture studies are used for evaluation (Landsman et al. 2011;Cline et al. 2012;Sass et al. 2018;Shaw, Sass, and Eslinger 2019). Catch-and-release can have variable species-specific consequences with some responses meeting manager and angler expectations, whereas others may not meet the defined goals of promoting this practice. Because of species-specific responses to CR, the promotion of this angling practice should not be treated as a "one size fits all" strategy to achieve angler, stakeholder, and manager desires in fisheries. At the same time, this review acknowledges that this perspective is Wisconsin-and north temperate lake-centric and species in different environments and habitats may respond differently to CR.
This review has provided clear examples where CR has not resulted in the desired outcome for anglers, stakeholders, and managers outside of conservation. For example, CR of largemouth bass (including bed fishing) has resulted in greater abundances at the expense of improved size structure, individual growth, and maximum growth potential (Cline et al. 2012;Hansen et al. 2015;Rypel et al. 2016;Sass et al. 2018). Outside of conservation, which has not been an issue for this species, largemouth bass responses to CR have been in stark contrast to what anglers, stakeholders, and managers have desired. Similarly, anglers and stakeholders have desired trophy muskellunge fisheries. The promotion of CR, stocking, and ultraconservative regulations have been used as tools to increase trophy potential of muskellunge populations. Although population size structure may be improved (Rypel et al. 2016;Eslinger et al. 2017), over 30 years of CR has not resulted in a greater frequency of trophy muskellunge (Gilbert and Sass 2016) (Figure 4). In contrast, CR for smallmouth bass (including bed fishing) has resulted in increased abundances, size structure, and trophy growth potential in this naturally lower-density, long-lived, and slow-growing species, which is aligned with angler, stakeholder, and manager desires (Rypel et al. 2016). In harvested species such as walleye, the promotion of CR to increase abundances, size structure, and trophy growth potential has had little effect to achieve these goals (Haglund, Isermann, and Sass 2016;Mrnak et al. 2018). Therefore, if conservation is not an issue, it would make little sense to promote CR and the use of conservative regulations, which would thereby reduce harvest opportunities for anglers for no reason. In the case of such species, system productivity and the fish community appear to dictate the abundance response, or lack thereof, to CR (Haglund, Isermann, and Sass 2016;Mrnak et al. 2018). Anglers, stakeholders, and managers should use the best species-specific science available to predict the outcomes of promoting CR. When species-specific information is not available, appropriate well-designed experiments and modeling should be conducted to test for responses (Carpenter 1998;Cooke and Suski 2005;Haglund, Isermann, and Sass 2016).
Catch-and-release is aimed to conserve fish populations, and many studies have been conducted to reduce hooking mortality and sublethal effects of the angling process (Cooke and Suski 2005;Arlinghaus et al. 2007;Cooke and Schramm 2007;Keretz et al. 2018); however, the ecology of fishes must not be overlooked. Nearly exclusive CR (e.g. black bass, muskellunge; Gaeta et al. 2013;Gilbert and Sass 2016) has the potential to create imbalances in fish communities managed for multiple species. Unnecessary stocking on top of natural recruitment can also overinflate natural abundances of certain species compounding these imbalances. Declines in walleye natural recruitment and adult abundances in the Ceded Territory of Wisconsin provide an example of where such imbalances may have occurred (Hansen et al. 2015;Hansen et al. 2017;Rypel et al. 2018;Embke et al. 2019;Sullivan et al. 2019). Although increases in growing degree days due to climate change have contributed to the increasing abundances of largemouth bass in northern Wisconsin (Hansen et al. 2017), nearly exclusive CR has also greatly aided this species (Gaeta et al. 2013;Hansen et al. 2015;Rypel et al. 2016;Miranda et al. 2017;Hessenauer et al. 2018;Sullivan et al. 2019). Anecdotal, followed by empirical, observations have confirmed reciprocal trends in walleye (negative) and largemouth bass (positive) abundances in northern Wisconsin prompting collaborative efforts to restore balance between both recreationally important species through walleye stocking, directed research (e.g., whole-lake centrarchid removal experiment), and allowing harvest of largemouth bass during the spawning period (Schnell 2014;Hansen et al. 2015;Kelling et al. 2016;Hansen et al. 2017;Sullivan et al. 2019). Nevertheless, restoring balance may be difficult to achieve if CR for bass remains pervasive and walleye and largemouth bass interact under cultivationdepensation dynamics (Walters and Kitchell 2001;Schnell 2014;Miranda et al. 2017;Hessenauer et al. 2018;Sullivan et al. 2019). These changes observed in community composition related to changes in fishing practices have been referred to as community overfishing. This circumstance occurs when selective and heavy fishing is applied to only a few species in the community, for example, walleye (Ontario Ministry of Natural Resources 1983). The species being harvested experiences depensatory mortality from the species that are favored by the new set of circumstances (e.g., largemouth bass and CR) (Ontario Ministry of Natural Resources 1983). Continued stocking on top of natural reproducing muskellunge populations and the pervasive CR of this species may further hinder efforts to rebalance these fish communities (Gilbert and Sass 2016;Shaw, Sass, and Eslinger 2019). Ultimately, anglers, stakeholders, and managers should consider unintended consequences of CR by using ecosystem-based fisheries management approaches in lieu of single species assessments in systems where multiple species are managed to avoid potential imbalances (Walters et al. 2005;Hansen et al. 2015).
Catch-and-release practices inherently couple social and ecological systems and this recognition has been acknowledged previously (Arlinghaus et al. 2007;Lynch et al. 2016Lynch et al. , 2017;;Hessenauer et al. 2016;Ward et al. 2016). Because CR is aimed to conserve fisheries, and many studies have been conducted to further reduce hooking mortality and the sublethal effects of angling, this practice has become a social norm in many fisheries and generally assumed to be beneficial to the target species. This review suggests that this is a narrow and generalized view (i.e., fisheries utopia) of the practice that may be short-sighted by not taking a long-term and ecosystem-based approach (Sass, Rypel, and Stafford 2017). When this social norm is pervasive for a species, CR has the potential to create imbalances in fish communities and angler, stakeholder, and manager desires may not be met. In such circumstances, how does one then change the social norm to swing the pendulum back toward balance and desirable multi-species fisheries? This review suggests that this is currently one of the biggest challenges facing inland fisheries management. In general, fisheries management is intended to regulate commercial, subsistence, and angler exploitation to conserve fisheries. Managers use harvest of fish by angler groups to structure fish populations, meet management objectives, and ensure sustainability of fisheries. Common tools used to regulate harvest include quotas, seasonal restrictions, length limits, and bag limits. Mandatory harvest regulations are not generally used in North American recreational fisheries outside of invasive species because animal welfare is not considered as strongly as it is in Europe (Arlinghaus 2007). Social norms that differ from management tools and objectives create a situation out of managerial control and dominated by pervasive angler and stakeholder beliefs that CR is always a good thing for fisheries. As such, changing the social norm of CR where the practice is creating imbalanced fisheries and not satisfying angler and stakeholder desires must derive from these groups themselves and rely on the best science available. Scientists and managers can contribute to these efforts by conducting deliberate CR experiments that are objective and indifferent to the advocacy of CR and by providing reviews such as the one presented here. Ultimately, when CR is not achieving fisheries management goals, this must be effectively communicated to anglers and stakeholders to restore a healthy level of harvest because pervasive CR can hinder regulation effectiveness (Miranda et al. 2017;Hessenauer et al. 2018). Further, management agencies generally lack the people power and funding to conduct wholelake fish removals to rebalance fisheries themselves and thus rely on angler harvest to achieve such goals.
CR can be used to maintain fish in a safe operating space (SOS) This review is not aimed to downplay the importance of CR in fisheries management. Catch-and-release, whether through conservative harvest regulations or angler behavior, has resulted in many fisheries conservation success stories. The purpose here is to highlight situations where CR has trumped efforts to structure fish populations and to meet angler desires through more commonly used output and input controls (Miranda et al. 2017;Hessenauer et al. 2018). An important instance where the promotion of CR can be particularly helpful is in situations where mechanisms causing fishery declines are outside of managerial control, such as climate change or angler effort in open access fisheries. Climate change cannot be managed in the short-term and has negatively affected some north temperate inland lake fishes (e.g. cisco Coregonus artedi, walleye) (Jacobson et al. 2012;Hansen et al. 2017). In such situations, the promotion of CR may provide resilience in conserving effected species by keeping populations within a "Safe Operating Space" (Carpenter et al. 2017). This form of CR would be independent of angler desires and solely focused on the true conservation of a species. Catch-and-release may also be important for sustainable management in high effort, open access fisheries even when CR mortality and harvest is low.
In some cases, pervasive CR by anglers has caused imbalances in fish communities and not met angler, stakeholder, and manager desires. When CR dominates in a fishery, commonly used input and output controls may be rendered useless to structure fisheries and meet angler desires. Studies of CR in recreational fisheries have almost exclusively focused on hooking mortality rates and methods to reduce these rates. Because natural mortality rates can be plastic depending on the level of fishing mortality, it is recommended that future studies focus on deliberate experiments to test for the influence of CR practices on the target species, fish community, and aquatic ecosystem. Such studies should not be conducted in laboratories, mesocosms, or with the use of net pens, but as deliberate ecosystem experiments (Carpenter 1998;Cline et al. 2012;Sass et al. 2018). Whole-lake studies (e.g., ecosystem experiments) that collect preand post-manipulation baseline data, have a reference system, and are monitored for enough time to detect potential effects are the only type of study that is relevant to the scale of management (Carpenter 1998).
Future research could also use modeling as a tool to describe the contagious or viral social behavior that CR has become in order to find a way to reverse these social norms where needed. Such modeling efforts may provide a better understanding of the behavior and subsequently identify mechanisms that have been important in establishing it as a social norm and identifying where this behavior can be reversed when needed. Promotion of harvest may also borrow from European sentiments of fish welfare to mandate harvest (Arlinghaus 2007;Arlinghaus et al. 2007); however, this mechanism should be viewed cautiously as to not swing the pendulum too far toward potential overharvest situations and the elimination of CR as a conservation and management tool (Rypel et al. 2016). Fisheries management is aimed to conserve fisheries through the control of exploitation in open access systems (i.e., to avoid the tragedy of the commons) and is not designed to dictate or control angler beliefs. Thus, CR, which is outside of managerial control, may be one of the biggest challenges facing fisheries management in the 21 st century.