Diet of Burbot and Implications for Sampling

Kathryn E. McBaine, Department of Fish and Wildlife Sciences, University of Idaho, 875 Perimeter Dr. MS 1141, Moscow, Id 83844 Zachary B. Klein, Idaho Cooperative Fish and Wildlife Research Unit, Department of Fish and Wildlife Sciences, University of Idaho, 875 Perimeter Dr. MS 1141, Moscow, Idaho 83844 Michael C. Quist, U.S. Geological Survey, Idaho Cooperative Fish and Wildlife Research Unit, Department of Fish and Wildlife Sciences, University of Idaho, 875 Perimeter Dr. MS 1141, Moscow, Idaho 83844 Darren T. Rhea, Wyoming Game and Fish Department, 432 East Mill Street, Pinedale, Wyoming 82935 Abstract Burbot (Lota lota) are an apex piscivore that were illegally introduced to the Green River drainage, Wyoming, raising concerns for the conservation and management of fishes throughout the basin. However, relatively little is known about the diet of non-native burbot. The objectives of this research were to characterize diet composition of burbot and identify differences in diet composition as a function of sampling gear. Diet composition was characterized using frequency of occurrence, percent by number, and percent by weight to identify the importance of each prey type to burbot. Diet composition was compared across gears to identify the relationship between gear and diet. Fishes were present in the stomach contents of nearly all burbot sampled and composed 62–100 percent of the stomach contents of burbot greater than 300 mm. Prey diversity was greatest in diets of burbot sampled with small-mesh hoop nets. Results from the current study provide important information on the diet of non-native burbot and highlight the potential influence of gear on diet studies. Key Words: Burbot, Gear bias, Diet, Green River, Wyoming their native distribution (Cott et al. 2011), Introduction relatively little is known about how burbot Burbot (Lota lota) are the only function in food webs in systems where they freshwater species of the family Gadidae are non-native. and have a circumpolar distribution In Wyoming, burbot represent native throughout Europe, Asia, and North America and non-native populations and are a (Stapanian et al. 2010). Burbot have been primary management concern for state, categorized as opportunistic piscivores federal, and tribal natural resource agencies. (Rudstam et al. 1995, Amundsen et al. 2003) Burbot are native to the Tongue and Wind- with fish typically dominating the diet of Bighorn river drainages, but are considered burbot greater than 400 mm (Rudstam et either rare (Wind-Bighorn River drainage) al. 1995, Schram et al. 2006). Fratt (1997) or extirpated (Tongue River, Krueger and evaluated prey consumption of burbot in Hubert 1997). In the Green River drainage, Green Bay, Lake Michigan, Wisconsin, burbot were illegally introduced into Big and reported that 55 percent of stomach Sandy Reservoir in the 1990s (Gardunio et contents (by volume) of burbot less than al. 2011). Since their initial introduction, 400 mm were fishes. Bailey (1972) reported burbot have been found from Flaming Gorge greater than 90 percent occurrence of fishes Reservoir (FGR) to the confluence of the in the diet of 119–742 mm burbot in Lake New Fork and Green rivers. The rapid Superior, Wisconsin and Michigan. The expansion of burbot in the Green River has author suggested that burbot were important increased concern for the management of competitors with other large piscivores in sport fishes and conservation of native fishes the system due to their non-selective diet in the system. The Green River supports and high consumption rates. Although economically, socially, and ecologically burbot are apex piscivores throughout important fishes including brown trout

© Intermountain Journal of Sciences, Vol. 24, No. ??, December 2018 1 (Salmo trutta), rainbow trout (Oncorhynchus of the Green River fish assemblage, little mykiss), Colorado River cutthroat trout (O. empirical data on diet are available for non- clarkii pleuriticus), roundtail chub (Gila native burbot. robusta), bluehead sucker (Catostomus Information on diet is fundamental discobolus), and flannelmouth sucker C.( for understanding how a given species latipinnis). Managers have hypothesized may influence the food web of a system that burbot compete with and (or) directly (Garvey and Chipps 2012). Although consume native fishes and economically a number of analytical techniques are important trout species. However, relatively available to quantify dietary information little is known about how non-native burbot (e.g., bioenergetics modeling, stable isotope may affect the trophic dynamics of recipient analysis), identification of gut contents is systems. a commonly used technique. Gut contents Negative effects of introduced are ideally quantified over extensive spatial species are often not exclusive to a single and temporal scales to capture seasonal mechanism. For example, both predation and temporal variation in diet (Hyslop and resource competition may occur 1980, Garvey and Chipps 2012). However, between native and non-native species. short-term diet studies can provide valuable Mills et al. (2004) evaluated interactions data that answer narrow questions (e.g., between non-native western mosquitofish fish- versus invertebrate-dominated diet, (Gambusia affinis) and native least chub consumption of native fishes) and can be (Iotichthys phlegethontis) in Walter Spring, used to guide future management decisions California and found that adult mosquitofish and research foci. greater than 30 mm fed extensively on Sampling techniques are an important 9–13 mm least chub. Once least chub consideration when describing the diet were too large to be consumed by western of fishes (Bowen 1996). Active gears, mosquitofish, they were negatively such as electrofishing often select for influenced by resource competition. sedentary individuals (Reynolds and Similarly, non-native burbot likely influence Kolz 2012). Sedentary individuals are native fishes through multiple mechanisms. not actively foraging, and studies using Cott et al. (2011) investigated the trophic information predominantly from sedentary ecology of burbot relative to lake trout individuals may underestimate the amount (Salvelinus namaycush), northern pike (Esox or inaccurately describe the types of lucius), and lake whitefish Coregonus( food consumed by fish in the population. clupeiformis) in four boreal Canadian lakes Alternatively, fish captured with passive using stable isotope analysis. Burbot and gears often contain greater amounts of food lake trout were both described as top-level than those caught by active gears. For piscivores in the lakes, and burbot were instance, Hayward et al. (1989) reported that thought to play a particularly important the amount of food in yellow perch (Perca role in structuring fish assemblages flavescens) stomachs was greater in fish via predation and competition. In the caught with gill nets than those caught by Green River drainage, burbot have been trawling in Lake Erie, Ohio. Furthermore, hypothesized to alter the system through passive entrapment gears can sample non- resource competition (i.e., habitat, food) target prey species increasing the potential and predation (Gardunio et al. 2011). For for post-capture consumption by piscivorous instance, Gardunio et al. (2011) suggested species. Breen and Ruetz (2006) examined that burbot outcompete smallmouth bass the diets of two bowfin (Amia calva) and (Micropterus dolomieu) for available prey in eight yellow bullhead (Ameiurus natalis) FGR as evidenced by declining catch rates captured in fyke nets stocked with round of smallmouth bass following establishment goby (Neogobius melanostomus), banded of burbot. Despite concerns regarding the killifish (Fundulus diaphanous), and influence of burbot on the trophic dynamics bluntnose minnow (Pimephales notatus).

2 McBaine et al. The authors reported that a single bowfin burbot, we sought to evaluate the influence consumed 35 percent of the fish stocked of passive-entrapment (hoop nets) and active in a fyke net, suggesting that piscivory is (electrofishing) gears on diet composition of likely high in entrapment gears. Therefore, burbot in the Green River. In addition, we the choice of sampling technique has the provide a short-term description of non- potential to influence diet composition by native burbot diet. Although we understand either sampling active or sedentary fish or that short-term diet studies do not capture by confounding diet composition by post- the spatio-temporal variability in diet, we capture piscivory. argue that any description of diet of non- Although the influence of sampling native burbot will be useful for directing gear on diet analysis has been recognized management actions and future research. for decades (Hayward et al. 1989), certain For instance, information on diet of non- instances (e.g., target species, habitat) native burbot is invaluable for understanding dictate when a particular sampling gear is if targeted suppression of the species is used. For instance, burbot are cold-water needed in the Green River. stenotherms that prefer deep habitats (Klein et al. 2015a) and are most often sampled Methods using passive gears such as hoop nets, cod The Green River originates in the Wind traps, and gill nets (Bernard et al. 1991, River Range of western Wyoming and is Spence 2000). Considering the need to a primary tributary of the Colorado River accurately describe the diet of non-native (Fig. 1). The Green River basin covers

Figure 1. River sections used for Burbot sampling in the Green River, Wyoming during the summer and autumn (2013). Boxes depict each section in detail, with sites sampled in the summer (solid black circles) and autumn (open black circles). Diet of Burbot and Implications for Sampling 3 parts of Wyoming, Colorado, and Utah. The standardized with a frequency of 45Hz and headwaters are characterized by high-gradient duty cycle of 45 percent at 2,750–3,250 runs interspersed with pool-riffle habitat W (Miranda 2009). A 2.4-m long dip net (Kurtz 1980). From its headwaters, the with 6-mm bar knotless mesh was used by Green River flows for approximately 235 a single netter positioned on the bow of the km before entering Fontenelle Reservoir. boat. Electrofishing was initiated at the From Fontenelle Reservoir downstream uppermost point of each 150-m reach and to the confluence of the Big Sandy River, preceded downstream until the entire reach the Green River is characterized by long had been sampled. runs averaging 450 m (Wiley 1974). All captured burbot were weighed From the Big Sandy River confluence to (nearest 1.0 g) and measured for total length Flaming Gorge Reservoir, the Green River (nearest 1.0 mm). On the final sampling is relatively low gradient (Wiley 1974). event for each reach, all captured burbot Sampling was conducted in the Green River, were euthanized with an overdose of MS- Wyoming, from August through November 222 (tricaine methanesulfonate, Western 2013. The river was divided into four Chemical, Inc., Ferndale, Washington). The sections to allocate sampling effort (Klein anterior portion of burbot stomachs were et al. 2015b). Each river section was then removed, preserved in 10 percent formalin, divided into 150-m long reaches. Reaches and returned to the University of Idaho for were sampled using night electrofishing, diet analysis. small-mesh hoop nets (6.4-mm bar mesh), Stomachs were opened and rinsed to and large-mesh hoop nets (19-mm bar ensure the removal of all contents. Seventy- mesh). A total of 28 reaches was sampled five stomachs were empty and removed over a 9-day period such that each reach from further diet analysis. Prey items were was sampled three times with each gear. An enumerated and weighed to the nearest additional 12 reaches were opportunistically 0.01 g by taxonomic category. Non-fish sampled with either night electrofishing, categories included Insecta, (Orconectes small-mesh hoop nets, or large-mesh hoop spp.), Gastropoda, Amphipoda, rocks, nets to obtain additional diet samples. and unknown material. Fish categories Small-mesh hoop nets were 3.0 m consisted of longnose dace (Rhinichthys long, had seven 0.6-m diameter hoops, and cataractae), speckled dace (R. osculus), constructed of 6.4-mm bar mesh. Large- redside shiner (Richardsonius balteatus), mesh hoop nets had an overall length of utah chub (Gila atraria), white sucker × 2.9 m with four 0.91-m diameter hoops flannelmouth sucker hybrid, mountain and were constructed of 19-mm bar mesh. whitefish (Prosopium williamsoni), rainbow Cod ends were anchored upstream and nets trout, burbot, mottled sculpin (Cottus were positioned parallel to the current. A bairdii), unknown catostomid, and unknown single net was baited with dead white sucker salmonid. Prey items identified as fish, (Catostomus commersonii) (a non-native but not assigned to taxonomic group were species common in the system) and fished categorized as unidentified fish. Diagnostic for approximately 24 hours in a given reach. structures were used when whole items were Bait was placed in a perforated plastic unavailable. For example, Orconectes spp. container attached in the cod end of each prey items were counted using the number net. Equal effort was used at each reach and of identifiable heads. catch was recorded as the number of fish per Burbot were grouped into 50-mm sampling event. length bins. Proportions of diet categories A drift boat equipped with a Smith- by number and weight were calculated for Root VVP-15B electrofisher (Smith- individual burbot and averaged for each Root, Vancouver, Washington) powered 50-mm length group. Diet composition was by a 5,000 W generator was used for also categorized as frequency of occurrence, night electrofishing. Power output was percent by number, and percent by weight

4 McBaine et al. for each gear type (night electrofishing, averaged 418 mm (± SE; ± 11 mm) in small-mesh hoop net, and large-mesh length; whereas, burbot sampled with hoop hoop net). Frequency of occurrence was nets had a mean length of 334 mm (± 12 calculated as the number of individuals mm). Burbot sampled using large-mesh with prey items of a particular category hoop nets averaged 340 mm (± 40 mm) in divided by the total number of individuals total length. with stomach contents. Percent by number Fish were observed in nearly all burbot was calculated as the number of items of stomachs (n = 211) and varied from 25–100 each prey type divided by the total number percent by number across lengths (Fig. of food items enumerated for each fish 2). Unidentified fish accounted for the and then averaged across individuals with greatest proportion of stomach contents stomach contents. Similarly, percentage among length bins, except for 200–249 by weight was calculated as the average mm. Non-fish contents were observed in proportional weight of each prey category all length bins, except for 700–749 mm (n across individuals with stomach contents. = 1). Diversity of prey items was greatest Standard error was calculated for both for 300–349 mm and 450–499 mm burbot. percent by number and percent by weight Proportions of prey items varied little for each category. between percent by number and weight for A multivariate analysis of variance all burbot length categories. Fish made up (MANOVA) was used to identify differences 62–100 percent of the diet by number for in diet composition by gear type (Johnson burbot greater than 300 mm. Fish in the diet 1998, Chipps and Garvey 2007). Analysis of burbot 150–300 mm represented 25–86 of variance (ANOVA) was then used to percent by number. Of the identified fishes, test for differences between gear types for salmonids were 2–25 percent by weight a given diet category (Ott and Longnecker of the contents for 250–699 mm burbot. 2010). If differences in count data were Burbot less than 350 mm consumed a higher observed between gears for a given diet proportion of non-fish prey items relative category, a Tukey-pairwise comparison was to burbot greater than 350 mm. Of these used to detect differences between gears. non-fish prey items, insects were 14–67 All tests were considered significant at α = 0.05. percent by number for burbot 150–349 mm (Figure 2). Orconectes spp. were observed Results in stomach contents of 250–699 mm burbot, In total, 231 burbot were sampled for but did not account for more than 15 percent diet analysis (Table 1). Night electrofishing by number or 11 percent by weight. sampled 156 burbot, small-mesh hoop Overall diet composition varied by gear. nets sampled 68 burbot, and seven burbot Diversity of ingested prey items was greatest were sampled using large-mesh hoop nets. for burbot captured in small-mesh hoop nets Burbot sampled using night electrofishing (Table 2). White sucker × flannelmouth

Table 1. Summary statistics for burbot (Lota lota) sampled from the Green River, Wyoming in August–November 2013. Burbot were sampled using night electrofishing, small-mesh hoop nets, and large-mesh hoop nets. Total length (mm) Sampling gear n x̄ SE Minimum Maximum Night electrofishing 156 418 11 31 719 Small-mesh hoop net 68 334 12 125 606 Large-mesh hoop net 7 340 40 178 497 All gears 231 391 9 31 719

Diet of Burbot and Implications for Sampling 5

Figure 2. Diet composition of Burbot sampled from the Green River, Wyoming in August- November 2013. Diet composition presented as average percent by number (upper panel) and average percent by weight (lower panel) for Burbot by 50-mm length bin. Asterisks indicate diet samples obtained from a single fish sucker, unknown catostomid, burbot, and for night electrofishing, 84 percent for mottled sculpin were only observed in small-mesh hoop nets, and 87 percent for the diet of burbot captured in small-mesh large-mesh hoop nets (Table 2). Results hoop nets. Utah chub was only observed of the MANOVA revealed a significant in stomachs from night electrofishing. difference in diet composition among gear

Cyprinids represented nearly 10 percent types (F2,153 = 1.72; P < 0.02). The ANOVA of the diet of burbot caught in small-mesh identified four diet taxa that were different hoop nets. Invertebrates composed 27 among gears. The diet of burbot captured percent by number of stomach content in with small-mesh hoop nets contained burbot caught by night electrofishing. The significantly higher numbers of redside percent by weight of fish in the diet of shiner (F2,153 = 3.83; P < 0.03), white sucker burbot was similar among gears: 67 percent × flannelmouth sucker hybrid F( 2,153 = 3.53;

6 McBaine et al. % W 1.24 (0.87)1.24 2.47 (1.10) 2.47 4.01 (1.44) 4.01 5.88 (1.79) 1.45 (0.76) 1.45 3.05 (1.24) 4.49 (1.66) 2.62 (1.20) 2.52 (1.23) 2.46 (1.22) 2.64 (1.09) 1.49 (0.88)1.49 0.52 (0.52) 0.66 (0.47) 0.69 (0.64) 0.64 (0.64) 51.12 (3.81) 51.12 12.04 (2.60) % N All gears 0.71 (0.46) 0.71 2.42 (1.18) 2.37 (1.15) 2.24 (1.15) 0.98 (0.71) (1.88) 6.79 1.86 (1.07) 4.49 (1.66) 2.24 (0.91) 1.81 (0.86) 1.81 0.75 (0.53) 0.75 2.28 (1.05) 3.53 (1.30) 0.32 (0.32) 1.84 (0.85) 0.64 (0.64) 11.86 (2.58) 11.86 52.87 (3.73) FO 0.02 0.02 0.12 0.09 0.04 0.02 0.03 0.01 0.04 0.04 0.60 0.05 0.01 0.04 0.01 0.03 0.04 0.03 0.00 % W 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 (13.19) 13.19 (17.27) 71.88 (14.93) 14.93 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 % N (10.00) 10.0 10.00 (10.00) 80.00 (12.25) Large mesh hoop net FO 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.20 0.20 0.00 0.00 0.00 0.00 0.00 Other Vertebrates Invertebrates % W 0.00 0.00 0.00 4.03 (2.80) 2.10 (2.08) 2.10 1.64 (1.64) (3.78) 8.78 7.98 (3.87) 7.98 8.20 (3.91) 0.45 (0.32) 0.93 (0.93) 2.08 (2.08) 2.08 (2.08) 5.60 (3.00) 6.25 (3.53) 3.05 (2.20) 3.23 (2.36) 43.60 (6.94) % N 0.00 0.00 0.00 2.31 (1.47) (3.74) 7.87 (2.31) 3.13 (2.31) 3.13 2.43 (1.72) (1.04) 1.04 (1.04) 1.04 7.29 (3.33) 7.29 7.29 (3.64) 7.29 6.59 (2.79) 2.08 (2.08) 2.08 (2.08) 0.94 (0.66) 6.25 (3.53) 46.53 (6.56) Small mesh hoop net FO 0.06 0.10 0.04 0.02 0.00 0.02 0.04 0.00 0.13 0.06 0.04 0.56 0.00 0.02 0.02 0.04 0.08 0.10 0.00 0.00 0.00 0.00 % W 1.99 (1.14) (1.54) 3.10 3.88 (1.91) (0.79) 0.79 2.50 (1.24) 2.25 (1.33) 3.93 (1.80) 0.07 (0.07)0.07 3.00 (1.55) 6.30 (2.33) 0.53 (0.53) 0.57 (0.57) (3.76) 17.48 53.62 (4.68) sampled from the Green River, Wyoming in August–November 2013. Diet composition is presented in Wyoming Lota lota) sampled from the Green River, 0.00 0.00 0.00 0.00 % N (1.75) 3.72 2.82 (1.61) 3.88 (1.91) 2.97 (1.52) (0.72) 1.33 2.62 (1.45) 2.30 (1.26) 0.03 (0.03) 0.32 (0.32) 0.49 (0.49) 6.89 (2.39) 0.65 (0.65) (3.76) 17.48 (4.67) 54.51 Night electrofishing FO 0.00 0.09 0.01 0.03 0.17 0.01 0.04 0.01 0.04 0.05 0.06 0.60 0.04 0.00 0.01 0.00 0.04 0.00 r

cker × ite Su Flannelmouth Sucke Flannelmouth Orconectes spp. Gastropoda Amphipoda Insecta Utah Chub Longnose Dace Speckled Dace Redside Shiner Wh Taxon Table 2. Diet composition of burbot ( Table as frequency of occurrence (FO), mean percent by number (%N), and weight (%W) for night electrofishing, small-mesh hoop net, hoop net, and all gears combined. Numbers in parentheses represent one standard error large-mesh Mountain Whitefish

Rocks Unidentified fish Unknown material Unknown catostomid Rainbow Trout Burbot Unknown salmonid Mottled Sculpin

Diet of Burbot and Implications for Sampling 7

spatial overlap between predator and prey. P < 0.04), burbot (F2.153 = 4.82; P < 0.01), Burbot occupy deep habitats with rocky and mottled sculpin (F2,153 = 5.48; P < 0.006) in their stomachs than other gears. substrate (Dixon and Vokoun 2009, Klein et al. 2015a); whereas, many juvenile stream- Discussion dwelling fishes occupy shallow habitats to Our results suggest non-native burbot avoid predation (Schlosser 1987, Delbert- from the Green River have similar diets to Lobb and Orth 1990). Harrison et al. (2013) burbot found in their native distribution. suggested that burbot move into the littoral In the current study, 25-100 percent of the zone during the crepuscular period to forage. stomach content (by weight) for non-native However, the authors noted a size-structured burbot contained fish. Schram et al. (2006) pattern in depth distribution; whereby, reported that fishes constituted greater than small burbot did not exhibit pronounced 90 percent by weight of the diet of burbot diel movements compared to large burbot. greater than 400 mm in the Apostle Islands The authors suggested that small burbot of Lake Superior, Wisconsin. Similarly, avoid foraging in littoral zones to reduce Fratt et al. (1997) described the diet of interspecific and intraspecific predation. As burbot in Green Bay and western Lake such, small burbot in the Green River may Michigan and reported that 94 percent be constrained to a diet composed primarily by volume was fishes. Additionally, the of invertebrates until they are no longer consumption of fish by burbot is often gape limited or the threat of size-dependent reported as being positively related to fish predation is negligible. Regardless of length. Amundsen et al. (2003) observed the exact mechanism resulting in the diet that percent by number of fish prey items of non-native burbot, our results suggest increased from 30 percent in 100–200 mm burbot may negatively influence the trophic burbot to nearly 100 percent for burbot dynamics of the Green River. greater than 400 mm in the subarctic Pasvik Non-native species can negatively watercourse of northern Norway and Russia. influence recipient ecosystems through Similarly, Tolonen et al. (1999) concluded various mechanisms including predation, that the probability of burbot consuming competition, and hybridization (Vitule et al. fish was positively correlated with length in 2009). For example, Ruzycki et al. (2003) Kilipisjärvi, a lake in northern Finland. In suggested non-native lake trout negatively the current study, an average of 56 percent influenced the persistence of Yellowstone (by weight) of the diets of burbot less than cutthroat trout (Oncorhynchus clarkii 300 mm contained fish; whereas, diets bouvieri) in Yellowstone Lake, Wyoming, of fish greater than 300 mm contained an by consuming approximately 14 percent of average of 82 percent fish. The observed the vulnerable cutthroat trout population differences between the diets of small and in a single year. Similarly, Saunders et large non-native burbot is likely related al. (2014) described the diet of burbot in to ontogenetic diet shifts associated with FGR and concluded that Orconectes spp. behavior and gape limitation. Small burbot occurred in 78-85 percent of the stomachs. (<300 mm) are likely gape limited as The authors suggested non-native burbot evidenced by Damsgard’s (1995) commonly could negatively influence smallmouth bass used (Stockwell et al. 2010, Harrison et al. populations in FGR through competition 2013) prey vulnerability model (maximum for Orconectes spp. Although our results prey length (cm) = [0.535 × predator length indicate that non-native burbot do not (cm)] - 0.487). Based on Damsgard’s consume high proportions of Orconectes model, a 300 mm burbot could only ingest a spp., the abundance of fish in burbot diet 155 mm prey item. Although small burbot suggest the species could alter the trophic could theoretically consume fish roughly dynamics of the Green River through direct half their body size, prey items may not be predation and competition. Klobucar et available for some burbot due to a lack of al. (2016) concluded that burbot in FGR

8 McBaine et al. consume an estimated 45,400 kg of fish suggested that reduced activity due to colder annually assuming a population size of water temperatures in the winter subjected 80,000 burbot. Although a population resident fishes to increased levels of estimate of burbot is not available for the predation by burbot. If burbot in the Green Green River, the results of Klobucar et al. River exhibit similar seasonal diet shifts, (2016) suggest burbot in the Green River piscivory may increase in the winter further could negatively influence fish populations threatening native and sport fish species. though direct predation. Furthermore, Our results suggest diet composition is Klobucar et al. (2016) estimated that burbot likely influenced by gear type. Specifically, diets in FGR contained an average of 32 selectivity for small-bodied fishes in percent (by weight) fish; whereas, our results entrapment gears could bias diet analysis indicate that fish constituted an average of due to post-capture piscivory by target 75 percent (by weight) of burbot diet in the species. Merriner (1975) used multiple Green River. As such, burbot may have a sampling techniques (i.e., gill nets, higher per capita rate of predation in the haul seines, trawls, and pound nets) to Green River compared to predation rates characterize the diet of weakfish Cynoscion( in FGR. In addition to direct predation, regalis) in Pamlico Sound and waters non-native burbot may negatively influence near Morehead City, North Carolina. The fish populations and species assemblages author reported contrasting occurrences through indirect effects. Knudsen et al. of diet items in relation to gear type and (2010) reported that burbot negatively only observed thread herring (Opisthonema influenced Arctic charr Salvelinus( alpinus) oglinum) in stomachs of weakfish captured populations through direct predation and with pound nets. Interestingly, diet of predation-induced shifts in resource use (i.e., burbot captured by small-mesh hoop nets habitat, diet). Although additional research in the current study contained a majority of is likely needed to understand the influence small-bodied fishes such as redside shiners of non-native burbot on the food web of and mottled sculpin. Alternatively, larger- the Green River, the presence of an apex bodied fish such as mountain whitefish were piscivore in the system is a concern for the observed in the diet of burbot captured by conservation and management of native and large-mesh hoop nets. The fact that burbot sport fishes. consumed higher proportions of small- Although we describe diet composition bodied fishes in small-mesh hoop nets may of non-native burbot, inferences should be be the result of the size selectivity of small- made with caution. Burbot were sampled and large-mesh hoop nets. For example, from August to November and the diet 1,258 redside shiners were captured in data presented here likely do not reflect small-mesh hoop nets over the course of seasonal variations in diet. Rudstam et al. the study; whereas, no redside shiners were (1995) concluded that the diet of burbot in caught in large-mesh hoop nets. Bowen Green Bay, Lake Michigan, was dominated (1996) cautioned that large fish captured in by alewives (Alosa pseudoharengus) in entrapment gears may feed on prey types winter and spring, and shifted to rainbow disproportional to natural occurrences or smelt (Osmerus mordax) in summer and consume prey not normally in the diet. autumn. Similarly, Chisholm et al. (1989) Duffy et al. (2011) quantified the number reported that largescale sucker (Catostomus of juvenile salmonids (Oncorhynchus spp.) macrocheilus) were most important to that had been consumed by piscivores in burbot in Libby Reservoir, Montana, in downstream migrant traps in Prairie Creek, autumn and winter; whereas, yellow perch California. Adult Coastal cutthroat trout dominated burbot diets in spring. Burbot in (O. clarkii clarkii) captured in live boxes FGR primarily consumed northern crayfish consumed five to six times as many juvenile in the autumn and increase piscivory in the salmonids as those sampled using other winter (Klobucar et al. 2016). The authors techniques. Thus, greater occurrence of

Diet of Burbot and Implications for Sampling 9 redside shiner, white sucker × flannelmouth baseline data on diet of non-native burbot in sucker, burbot, and mottled sculpin in diets the Green River. of burbot captured in small-mesh hoop nets may have been an artifact of opportunistic Acknowledgements feeding behavior associated with sampling We thank J. Johnson, S. Opitz, and gear. Based on our data, a single sampling A. Senecal for assistance with field work. technique would have yielded a different We also thank R. Keith, A. Senecal, H. diet composition for burbot in the Green Sexauer, M. Smith, and T. Neebling of River. White sucker × flannelmouth sucker, Wyoming Game and Fish Department unknown catostomid, burbot, and mottled for their cooperation in planning and sculpin were only observed in the diet of implementation of field research. Funding burbot captured in small-mesh hoop nets, for the project was provided by Wyoming and utah chub was only observed in stomach Game and Fish Department. Additional contents of burbot captured by night support was provided by the U.S. Geological electrofishing. Burbot are often sampled Survey, Idaho Cooperative Fish and Wildlife using passive entrapment gears (e.g., hoop Research Unit. The Unit is jointly sponsored nets, cod traps) due to the habitat use by the U.S. Geological Survey, University (e.g., deep water) of the species. As such, of Idaho, and Idaho Department of Fish and diet collected from burbot sampled using Game, and Wildlife Management Institute. entrapment gears may not adequately reflect This project was conducted under the what burbot would consume under normal University of Idaho Institutional Animal conditions. Care and Use Committee Protocol 2011-33. Our results suggest the diet of non- The use of trade, firm, or product names is native burbot was similar to the diet of for descriptive purposes only and does not burbot within their native distribution. Non- imply endorsement by the U.S. Government. native burbot are a functional apex piscivore and have the potential to influence trophic Literature Cited dynamics in the Green River. As such, Amundsen, P. A., T. Bøhn, O. A. Popova, managers of the Green River may want to F. J. Staldvik, Y. S. Reshetnikov, N. focus efforts on understanding how an apex A. Kushulin, and A. A. Lukin. 2003. piscivore may influence species interactions Ontogenetic niche shifts and resource in the system. As additional research will partitioning in a subarctic piscivorous likely require further diet analysis, managers fish guild. Hydrobiologia 497:106–119. should be cognizant of the potential biases associated with using entrapment gears. Although entrapment gears are commonly Bailey, M. M. 1972. Age, growth, used to sample burbot, alternative sampling reproduction, and food of the Burbot, techniques should be used for diet studies Lota lota (Linnaeus), in southwestern focused on the species. Gill nets or similar Lake Superior. Transactions of the passive sampling techniques (e.g., trammel American Fisheries Society 101:667– nets) are effective for sampling benthic 674. species (e.g., burbot) in lentic systems Beauchamp, D. A., D. L. Parrish, and R. A. (Beauchamp et al. 2009). However, gill nets Whaley. 2009. Coldwater fish in large are not effective in high-current velocities standing waters. Pp. 97–138 in S. A. typical of many small, non-wadebale Bonar, W. A. Hubert, and D. W. Willis, rivers; therefore, active techniques such as editors. Standard methods for sampling electrofishing may be the best alternative North American freshwater fishes. (Klein et al. 2015a). Collectively, our American Fisheries Society, Bethesda, results highlight the importance of gear Maryland. selection for diet studies while providing

10 McBaine et al. Bernard, D. R., G. A. Pearse, and R. H. Dixon, C. J. and J. C. Vokoun. 2009. Conrad. 1991. Hoop traps as a means to Population structure and diet of Burbot capture burbot. North American Journal (Lota lota) in small streams near the of Fisheries Management 11:91–104. southern extent of the species’ range. Bowen, S. H. 1996. Quantitative description Journal of Freshwater Ecology 25:49–58. of the diet. Pp. 513–532 in B.R. Murphy Duffy, W. G., E. P. Bjorkstedt, and C. S. and D.W. Willis, editors. Fisheries Ellings. 2011. Predation on juvenile techniques, 2nd edition. American Pacific SalmonOncorhynchus spp. in Fisheries Society, Bethesda, Maryland. downstream migrant traps in Prairie Breen, M. J. and C. R. Ruetz III. 2006. Creek, California. North American Gear bias in fyke netting: evaluating Journal of Fisheries Management soak time, fish density, and predators. 31:151–164. North American Journal of Fisheries Fratt, T. W., D. W. Coble, F. Copes, and R. Management 26:32–41. E. Bruesewitz. 1997. Diet of Burbot in Chipps, S. R. and J. E. Garvey. 2007. Green Bay and western Lake Michigan Assessment of diets and feeding with comparison to other waters. Journal patterns. Pp. 473–514 in C. S. Guy and of Great Lakes Research 23:1–10. M. L. Brown, editors. Analysis and Gardunio, E. I., C. A. Myrick, R. A. interpretation of freshwater fisheries Ridenour, R. M. Keith, and C. J. data. American Fisheries Society, Amadio. 2011. Invasion of illegally Bethesda, Maryland. introduced Burbot in the upper Colorado Chisholm, I., M. E. Hensler, B. Hansen, River basin, USA. Journal of Applied and D. Skaar. 1989. Quantification Ichthyology 27:36–42. of Libby Reservoir levels needed to Garvey, J. E. and S. R. Chipps. 2012. Diets maintain or enhance reservoir fisheries: and energy flow. Pp. 733–779 in A. V. summary report 1983–1985. U.S. Zale, D. L. Parrish, and T. M. Sutton, Department of Energy, Bonneville Power editors. Fisheries techniques, 3rd edition. Administration, Division of Fish and American Fisheries Society, Bethesda, Wildlife, Portland, Oregon. Maryland. Cott, P. A., T. A. Johnston, and J. M. Gunn. Harrison, P. M., L. F. G. Gutowsky, E. G. 2011. Food web position of burbot Martins, D. A. Patterson, A. Leake, S. J. relative to lake trout, northern pike, and Cooke, and M. Power. 2013. Diel vertical lake whitefish in four sub-Arctic boreal migration of adult Burbot: a dynamic lakes. Journal of Applied Ichthyology trade-off among feeding opportunity, 27:49–56. predation avoidance, and bioenergetic Damsgard, B. 1995. Arctic Charr, Salvelinus gain. Canadian Journal of Fisheries and alpinus (L.) as prey for piscivorous fish: Aquatic Science 70:1765–1774. a model to predict prey vulnerabilities Hayward, R. S., F. J. Margraf, C. T. Knight, and prey size refuges. Nordic Journal of and D. J. Glomski. 1989. Gear bias in Freshwater Research 71:190–196. field estimation of the amount of food Delbert-Lobb, M. III. and D. J. Orth. 1990. consumed by fish. Canadian Journal of Habitat use by an assemblage of fish in a Aquatic Sciences 46:874–876. large warmwater stream. Transaction of Hyslop, E. J. 1980. Stomach contents the American Fisheries Society 120:65– analysis–a review of methods and their 78. application. Journal of Fish Biology 17:411–429.

Diet of Burbot and Implications for Sampling 11

Johnson, D. E. 1998. Applied multivariate Miranda, L. E. 2009. Standardizing methods for data analysis. Duxbury electrofishing power for boat Press, Belmont, California. electrofishing. Pp. 223–230 in S. A. Klein, Z. B., M. C. Quist, D. T. Rhea, and Bonar, W. A. Hubert, and D. W. Willis, A. C. Senecal. 2015b. Habitat use of editors. Standard methods for sampling non-native Burbot in a western river. North American freshwater fishes. Hydrobiologia 757:61–71. American Fisheries Society, Bethesda, Maryland. Klein, Z. B., M. C. Quist, D. T. Rhea, and A. C. Senecal. 2015a. Sampling techniques Ott, R. L., and M. Longnecker. 2010. An for burbot in a western non-wadeable introduction to statistical methods and river. Fisheries Management and data analysis. Brooks/Cole, Cengage Ecology 22:213–223. Learning, Belmont, California. Klobucar, S. L., W. C. Saunders, and P. Reynolds, J. B. and A. L. Kolz. 2012. Budy. 2016. A Lota lota consumption: Electrofishing. Pp. 305–361in A. V. trophic dynamics of nonnative Burbot Zale, D. L. Parrish, and T. M. Sutton, in a valuable sport fishery. Transactions editors. Fisheries techniques, 3rd edition. of the American Fisheries Society American Fisheries Society, Bethesda, 145:1386–1398. Maryland. Knudsen, R., P. A. Amundsen, and A. Rudstam, L. G., P. E. Peppard, T. W. Klemetsen. 2010. Arctic Charr in Fratt, R. E. Bruesewitz, D. W. Coble, sympatry with Burbot: ecological and F. A. Copes, J. F. Kitchell. 1995. Prey evolutionary consequences. Hydrobiolgia consumption by the burbot (Lota lota) 650:43–54. population in Green Bay, Lake Michigan, based on a bioenergetics model. Krueger, K. L. and W. A. Hubert. 1997. Canadian Journal of Fish and Aquatic Assessment of lentic burbot populations Sciences 52:1074–1082. in the Big Horn/Wind River drainage, Wyoming. Journal of Freshwater Ruzycki, J. R., D. A. Beauchamp, and D. Ecology 12:453–463. L. Yule. 2003. Effects of introduced Lake Trout on Native Cutthroat Trout Kurtz, J. 1980. Fishery management in Yellowstone Lake. Ecological investigations - a study of the upper Application 13:23–37. Green River fishery, Sublette County, Wyoming 1975–1979. Wyoming Game Saunders, C., S. Klobucar, and P. Budy. and Fish Department, Cheyenne. 2014. Food web effects, seasonal movements, and spawning ecology of Merriner, J. V. 1975. Food habits of the Burbot, a large invasive piscivore, in Weakfish, Cynoscion regalis, in North Flaming Gorge Reservoir, Utah and Carolina waters. Chesapeake Science Wyoming. Utah Division of Wildlife 16:74–76. Resources Sportfish Research: Project ID Mills, M. D, R. B. Rader, and M. C. Belk. 121041. Logan, Utah. 2004. Complex interactions between Schlosser, I. J. 1987. The role of predation native and invasive fish: the simultaneous in age-and size-related habitat use by effects of multiple negative interactions. stream fishes. Ecology 68:651–659. Oecologia 141:713–721.

12 McBaine et al. Schram, S. T., T. B. Johnson, and M. J. Stockwell, J. D., T. R. Hrabik, O. P. Jensen, Seider. 2006. Burbot consumption and D. I. Yule, and M Balge. 2010. Empirical relative abundance in the Apostle Islands evaluation of predator-driven diel region of Lake Superior. Journal of Great verticle migration in Lake Superior. Lakes Research 32:798–805. Canadian Journal of Fisheries and Spence, C. R. 2000. A comparison of Aquatic Sciences 67:473–485. catch success between two styles of Tolonen, A., J. Kjellman, and J. burbot trap in lakes. Pages 165–170 Lappalainen. 1999. Diet overlap between in V. L. Paragamian and D. W. Willis, burbot (Lota lota (L.)) and whitefish editors. Burbot: biology, ecology, (Coregonus lavaretus (L.)) in a subarctic and management, American Fisheries lake. Annales Zoologici Fennici 36:205– Society, Bethesda , Maryland. 214. Stapanian, M. A., V. L. Paragamian, Vitule, J. R. S., C. A. Freire, and D. C. P. Madenjian, J. R. Jackson, J. Simberloff. 2009. Introduction of non- Lapppalainen, M. J. Evenson, and M. native freshwater fish can certainly be D. Neufeld. 2010. Worldwide status of bad. Fish and Fisheries 10:98–108. burbot and conservation measures. Fish Wiley, R. W. 1974. Tributary stream and Fisheries 11:34–56. inventory 1972–1974. Wyoming Game and Fish Department, Cheyenne.

Received 07 March 2018 Accepted 04 April 2018

Diet of Burbot and Implications for Sampling 13