Transactions of the American Fisheries Society 151:453–473, 2022 © 2022 American Fisheries Society. This article has been contributed to by U.S. Government employees and their work is in the public domain in the USA. ISSN: 0002-8487 print / 1548-8659 online DOI: 10.1002/tafs.10362 ARTICLE Attraction, Entrance, and Passage Efficiency of Arctic Grayling, Trout, and Suckers at Denil Fishways in the Big Hole River Basin, Montana Ben Triano* Fish and Wildlife Ecology and Management Program, Ecology Department, Montana State University, Post Office Box 173460, Bozeman, Montana 59717, USA Kevin M. Kappenman* U.S. Fish and Wildlife Service, Bozeman Fish Technology Center, 4050 Bridger Canyon Road, Bozeman, Montana 59715, USA Thomas E. McMahon Fish and Wildlife Ecology and Management Program, Ecology Department, Montana State University, Post Office Box 173460, Bozeman, Montana 59717, USA Matt Blank Western Transportation Institute, Montana State University, 2327 University Way, Bozeman, Montana 59717, USA Kurt C. Heim U.S. Fish and Wildlife Service, Western New England Complex, 11 Lincoln Street, Essex Junction, Vermont 05452, USA Albert E. Parker Center for Biofilm Engineering, Department of Mathematical Sciences, Montana State University, 304 Barnard Hall, Bozeman, Montana 59717, USA Alexander V. Zale U.S. Geological Survey, Montana Cooperative Fishery Research Unit, Fish and Wildlife Ecology and Management Program, Department of Ecology, Montana State University, Post Office Box 173460, Bozeman, Montana 59717, USA Nolan Platt U.S. Department of Agriculture Forest Service, Lolo National Forest, 24 Fort Missoula Road, Missoula, Montana 59804, USA Katey Plymesser Department of Civil Engineering, Montana State University, 223 Cobleigh Hall, Bozeman, Montana 59717, USA *Corresponding authors: benjamintriano@gmail.com; kevin_kappenman@fws.gov Received September 24, 2021; accepted March 21, 2022 453 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 454 TRIANO ET AL. Abstract The Big Hole River basin in southwestern Montana supports the only indigenous, self-sustaining fluvial population of Arctic Grayling Thymallus arcticus in the conterminous United States, but the basin is fragmented by numerous low-head irrigation diversion dams. Denil fishways at 63 diversion dams provide Arctic Grayling and other fishes opportunities for year-round access to critical habitats; however, their efficiency has not been evaluated. We quantified attraction, entrance, and passage for hatchery-reared Arctic Grayling, wild trout (Brook Trout Salvelinus fontinalis and Brown Trout Salmo trutta), and wild suckers (White Sucker Catostomus commersonii and Longnose Sucker C. catostomus) during 14 field trials conducted at six Denil fishways over a representative range of fishway slopes and hydraulic conditions using passive integrated transponder telemetry. Attraction (60.4–84.3%) and entrance (44.3– 78.6%) efficiencies were variable across test conditions and reduced overall fishway efficiencies (19.1–55.8%). In con- trast, upon entry, passage efficiencies were high (96.2–97.0%) for all taxa across all test conditions. Attraction of hatchery-reared Arctic Grayling increased with upstream depth (a surrogate for fishway discharge) and attraction flow, but attraction of wild fish was less affected by these conditions. Entrance of Arctic Grayling, Brook Trout, and Brown Trout decreased with upstream depth and fishway slope, especially when plunging entrance conditions associ- ated with shallow downstream depths were present. However, entrance of Arctic Grayling and both trout species increased with downstream depth, and submerged fishway entrances demonstrated promise for increasing entrance effi- ciency at fishways with high discharges and steep slopes. We demonstrate that comprehensive evaluations of fishway efficiency components can identify specific solutions that improve fishway efficiency; application of these engineering solutions at individual fishways (as needed) could improve their efficiency and further enhance aquatic connectivity for fishes in the Big Hole River basin and elsewhere. Connectivity in rivers facilitates migration of fish to percentage of approaching fish that are attracted to, enter, meet life history requirements and respond to biological and successfully pass a fishway (Baker et al. 2019). and environmental cues (Northcote 1984; Baras and Lucas Quantifying these metrics individually provides impor- 2001). However, anthropogenic barriers such as dams, tant insight as to how each component limits overall fish- diversions, culverts, and road crossings can impede fish way efficiency (Bunt et al. 2012). For example, large migration and negatively affect populations (Morita et al. numbers of approaching fish may be attracted to a fish- 2000; Alò and Turner 2005). Barrier removal can restore way, but few may enter. Alternatively, few approaching fish populations (Roni et al. 2008) but is not always feasi- fish may be attracted to the entrance, but those that are ble because of overriding socioeconomic benefits and high attracted may have high entrance and passage success. costs (O’Hanley and Tomberlin 2005; Januchowski- The time it takes a fish to become attracted to, enter, and Hartley et al. 2013). Alternatively, fishways are often pass a fishway is also of interest because delays at barriers installed at potential migration barriers to restore connec- and passage structures can detrimentally affect reproduc- tivity (Bunt et al. 1999; Schmetterling et al. 2002), but rel- tion and fish health (Mesa and Magie 2006; Newton et al. atively few fishways have been evaluated in the field to 2018). Understanding how fish are limited in using a fish- determine their efficiency in passing fish (Noonan et al. way provides a basis for strategic improvements to fish- 2012; Cooke and Hinch 2013). way design that address specific limiting factors and A comprehensive evaluation of fishway efficiency enhance overall efficiency. requires the systematic assessment of three distinct effi- Denil fish ladders (Katopodis 1992) are commonly used ciency components (attraction, entrance, and passage) of to restore connectivity in rivers (Clay 1995; Bunt et al. approaching fish, any of which can reduce overall fishway 1999; Haro et al. 1999; Schmetterling et al. 2002); how- efficiency (Bunt et al. 2012, 2016). However, previous ever, their efficiency for passing many fish species under studies have typically defined success rates by combining different slopes and hydraulic conditions is not well under- multiple efficiency components (Forty et al. 2016; Hodge stood (Haro et al. 1999; Mallen-Cooper and Stuart 2007). et al. 2017), precluding assessment of each distinct compo- Seasonal hydrologic variation alters entrance and exit nent. “Approach” describes the number of fish that water depths (hereafter downstream and upstream depths, encounter a potential barrier and is an index of potential respectively, or collectively fishway depths) at Denil fish- population use (Hodge et al. 2017). “Attraction efficiency” ways (Platt 2019), and fishway depths markedly affect is the percentage of approaching fish that locate the fish- hydraulic conditions inside a Denil fishway, thereby affect- way entrance, “entrance efficiency” is the percentage of ing passage success (Haro et al. 1999; Blank et al., In attracted fish that enter the fishway, and “passage effi- press). Fishway depths are directly influenced by fishway ciency” is the percentage of entering fish that successfully slope, which has variable effects on passage success pass (Cooke and Hinch 2013). “Overall efficiency” is the through Denil fishways; slope had little effect on passage 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 455 success of Arctic Grayling Thymallus arcticus through an lewisi, Mountain Whitefish Prosopium williamsoni, Burbot Alaska Steeppass fish ladder (a type of Denil fishway) Lota lota, White Sucker Catostomus commersonii, Long- (Tack and Fisher 1977) but increasing slope negatively nose Sucker C. catostomus, Mountain Sucker C. platyr- affected passage of nonsalmonid fishes though Denil fish- hynchus, Longnose Dace Rhinichthys cataractae, and ways (Haro et al. 1999; Mallen-Cooper and Stuart 2007). Rocky Mountain Sculpin Uranidea sp. cf. bairdii) and Denil fishways were installed throughout the Big Hole nonnative (Brook Trout Salvelinus fontinalis, Brown Trout River basin in Montana to improve aquatic connectivity Salmo trutta, and Rainbow Trout O. mykiss) fishes for imperiled Arctic Grayling. Big Hole River Arctic (Oswald 2000). Grayling are the last remaining indigenous, self-sustaining Seasonal hydrologic variation and irrigation practices fluvial Arctic Grayling population in the conterminous in the Big Hole River basin result in highly variable fish- United States (Shepard and Oswald 1989; Kaya 1992) and way depths at Denil fishways (Platt 2019). The river is fed have been considered for protection under the Endangered by spring snowmelt that results in peak stream discharges Species Act since 1982 (USOFR 2014). Big Hole River during June and early July and low base flows during Arctic Grayling make seasonal migrations exceeding 80 August and September (Sladek 2013; Vatland 2015). Land km to access critical main-stem and tributary habitats for use in the basin is predominantly agricultural, with about spawning, feeding, wintering, and thermal refuge (Shepard 1,000 water rights allocated for seasonal diversion of sur- and Oswald 1989; Lamothe and Magee 2003); however, face water to irrigate hay fields and support livestock the basin is fragmented by numerous low-head irrigation (MTFWP and USFWS 2006). Low base flows are often diversion dams (pin-and-plank style; Schmetterling et al. exacerbated by irrigation withdrawals (Vatland 2015) and 2002) that impound and divert water for agriculture. The potentially limit connectivity through Denil fishways. Candidate Conservation Agreement with Assurances Fishways typically operate at deep upstream depths (45.0– (CCAA) for Fluvial Arctic Grayling in the upper Big 60.0 cm) and high fishway discharges (0.075–0.20 m3/s) Hole River was established in 2006 (MTFWP and during peak stream discharges, but fishway discharges can USFWS 2006), with a primary focus being mitigation of decrease to nearly zero during base flows and periods of barriers to Arctic Grayling migration. Since 2001 and fol- irrigation withdrawal. lowing the establishment of the CCAA, standard-type Denil fishways have been installed at 63 irrigation diver- Study Sites sions in the basin to improve aquatic connectivity; how- Fishway efficiency was evaluated in 14 field trials at ever, their efficiency for passing Arctic Grayling and other six Denil fishways in the upper Big Hole River basin fishes has not been evaluated comprehensively. from June to October of 2018. Seventeen fishways were We evaluated the efficiency of Denil fishways for facili- initially evaluated in 2017 to assess seasonal hydrologic tating upstream passage of Arctic Grayling and other spe- and hydraulic variation and physical differences among cies in the Big Hole River basin using passive integrated fishway installations (Platt 2019). Three primary study transponder (PIT) telemetry. Our main objectives were to sites (multiple field trials) and three secondary study sites (1) quantify all three efficiency components (attraction, (single trials) were selected for fishway efficiency evalua- entrance, and passage) to determine which components tions in 2018 (Figure 1; Table 1) to best represent the limit overall efficiency and (2) evaluate how fishway range of fishway depths and slopes observed in 2017. Pri- depths and slope affect each efficiency component. Achiev- mary sites were on Steel Creek, upper Warm Springs ing these objectives allowed us to provide specific recom- Creek, and the main-stem Big Hole River, with respec- mendations for improving the overall efficiency of existing tive slopes of 5.0, 10.7, and 15.6%; three or four trials and future Denil fishway installations, which could con- were conducted at each primary site over a range of fish- tribute to enhancing aquatic connectivity in the Big Hole way depths. Secondary sites were included to represent River basin and elsewhere. other conditions observed among Big Hole River Denil fishways, such as submerged fishway entrances at Rock Creek and Swamp Creek (Figure 2) and a near 0.0% METHODS slope and plunging entrance conditions at lower Warm Springs Creek (Figure 3). All Denil fishways tested were Study Area standard type with identical width and depth dimensions The Big Hole River originates in the Beaverhead of 61 × 61 cm (Katopodis et al. 1997). Mountains of southwestern Montana and flows 250 km to its confluence with the Beaverhead River (DNRC 1979). Hydraulic Conditions The river and its tributaries provide a variety of critical Upstream and downstream depths and water tempera- habitats for a diverse assemblage of native (Arctic Gray- tures were recorded every 5 min by data loggers (Model ling, Westslope Cutthroat Trout Oncorhynchus clarkii U20L-04; Onset Computer Corporation, Bourne, 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 456 TRIANO ET AL. FIGURE 1. Map of the upper Big Hole River basin study area, depicting locations of the 63 Denil fishways currently installed under the CCAA program. Massachusetts) installed in stilling wells positioned at the and slope (Platt 2019). Total stream discharge downstream exit and entrance of the fishway. Fishway depths were of the diversion structure was estimated daily using stage– measured relative to the bottom (or “invert”; Figure 4) of discharge relationships developed by Platt (2019). Fishway the fishway exit and entrance (i.e., upstream and down- attraction flow (%) was calculated as the relative contribu- stream depths describe how full the fishway is at the exit tion of fishway discharge to the total stream discharge and entrance, respectively). Fishway discharge was calcu- downstream of the diversion dam. lated every 5 min as the average estimate of five laboratory-derived rating curves for standard-type Denil Test Fish fishways (Katopodis 1992; Katopodis et. al 1997; Rajarat- Field trials were run with wild fish present at each site, nam et al. 1997; FAO and DVWK 2002; Odeh 2003) that including Brook Trout (mean TL  SD = 231 58 mm) predicted discharge from upstream depth (measured rela- and other taxa as available: Longnose Sucker (194 34 tive to the v-notch of the most upstream baffle; Figure 4) mm), White Sucker (230 57 mm), Brown Trout (285 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 457 TABLE 1. Average physical and hydraulic conditions in trials 1–14; the variation in hydraulic conditions observed during each trial is reported in parentheses, and overall means repre- sent the average conditions observed across all trials. Abbreviations are as follows: slope = fishway slope, and length = fishway length. Water Slope Length Upstream Fishway Downstream temperature Attraction Study site Trial (%) (m) depth (cm) discharge (m3/s) depth (cm) (°C) flow (%) Steel Creek 4 5.0 2.77 47.1 (45.0–50.1) 0.067 (0.061–0.077) 19.7 (18.9–20.7) 17.5 (12.1–23.8) 74.0 6 5.0 2.77 25.0 (23.6–26.6) 0.013 (0.011–0.016) 12.1 (11.3–13.1) 16.6 (11.2–22.2) 77.0 9 5.0 2.77 36.6 (33.3–38.8) 0.037 (0.029–0.043) 15.8 (14.9–16.8) 13.0 (8.6–17.9) 68.0 10 5.0 2.77 20.5 (19.7–21.4) 0.007 (0.006–0.008) 12.6 (11.9–13.7) 12.2 (7.7–17.7) 30.0 Warm Springs 3 10.7 3.66 65.4 (58.4–69.6) 0.209 (0.165–0.239) 44.7 (41.8–53.9) 16.2 (13.0–19.8) 29.0 Creek (upper) 7 10.7 3.66 23.8 (22.8–25.2) 0.019 (0.017–0.022) 26.5 (24.7–27.7) 14.0 (10.0–18.6) 21.0 12 10.7 3.66 18.1 (15.0–21.5) 0.008 (0.004–0.014) 28.9 (27.4–30.8) 6.4 (3.5–10.1) 6.0 14 10.7 3.66 37.1 (35.7–42.4) 0.059 (0.054–0.081) 26.4 (25.0–28.0) 4.3 (2.0–7.4) 80.0 Big Hole River 8 15.6 3.66 22.1 (17.5–26.8) 0.022 (0.011–0.035) 20.7 (19.2–24.4) 10.8 (8.0–14.3) 10.0 11 15.6 3.66 52.8 (50.2–55.1) 0.165 (0.148–0.180) 23.8 (21.0–25.9) 9.1 (5.3–11.8) 52.0 13 15.6 3.66 36.9 (35.4– 38.6) 0.075 (0.068–0.083) 19.8 (18.3–21.9) 8.4 (4.8–12.2) 46.0 Rock Creek 1 13.4 3.66 55.0 (51.4–61.5) 0.167 (0.143–0.210) 80.4 (76.7–86.4) 14.1 (10.7–18.5) 105.0 Warm Springs 5 0.6 3.66 54.1 (47.9–57.5) 0.083 (0.057–0.099) 5.7 (4.0–9.1) 17.3 (13.2–21.9) 33.0 Creek (lower) Swamp Creek a 2a 15.7a 3.66 49.4 (45.6–52.5)a 0.145 (0.123–0.165)a 70.7 (69.8– 71.7)a 19.0 (14.0–23.5)a 80.0a Overall means 9.5 3.39 38.0 0.072 25.9 12.3 49.0 aTrial 2 was not included in overall mean calculations because of a considerable change in hydraulic conditions that occurred 20 h into the trial; hydraulic conditions reported here are from the first 20 h of trial 2. Trial 2 test fish were not included in data analyses and are only discussed in anecdotal comparisons. 458 TRIANO ET AL. FIGURE 2. Photo of the submerged fishway entrance during trial 2. The submerged entrance demonstrated promise for increasing entrance efficiency under the otherwise-limiting high fishway discharge and steep slope (Table 1) by creating presumably more favorable entrance conditions than those in trial 11 (Figure 10). FIGURE 3. Photo of plunging entrance conditions during trial 5. The deep upstream depth, shallow downstream depth, and shallow slope (Table 1) resulted in an entrance plunge that presumably restricted the entrance of Arctic Grayling (0.0% entrance of 14 attracted Arctic Grayling). 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 459 FIGURE 4. Denil fishway schematic identifying key components and depth measurements. 73 mm), Burbot (282 55 mm), and Mountain Whitefish were randomly netted from the raceway, scanned for tags, (206 mm). We included 50 Brook Trout in all trials except anesthetized, measured, and transported about 350 km to trial 1 (n= 36) and up to 30 fish combined of other taxa. study sites in a 340-L insulated, oxygenated holding tank Wild fish were collected on the first day of each trial by at 12°C. Test Arctic Grayling remained in the holding backpack electrofishing the stream segment immediately tank for 18–24 h prior to transfer to holding pens at each upstream of the diversion structure. Short-term displace- study site; temperature differences between the holding ment trials in which fish are captured upstream of a tank and study streams were <2°C. potential barrier and relocated downstream are effective for rapid evaluations of passage structures by invoking a Test Protocol homing response that increases motivation and participa- All test fish were kept in holding pens for at least 1 h tion of fish (Armstrong and Herbert 1997; Schmetterling before simultaneous release downstream of the fishway et al. 2002; Burford et al. 2009). Captured fish were anes- and diversion dam. Release times were near midday thetized with AQUI-S anesthetic (Aquatactics Fish (1145–1430 hours) except in trial 1 when release time was Health, Kirkland, Washington), measured (TL), and 1750 hours. Release locations were in the second pool tagged with half-duplex (HDX) PIT tags (Oregon RFID, downstream (15–40 m) of the diversion dam (Figure 5) so Portland, Oregon). Fish longer than 130 mm were tagged test fish had to volitionally leave the “release pool” and with 23-mm × 3.65-mm tags (0.6 g); shorter fish were move upstream (“approach”) through a riffle to reach the tagged with 12-mm × 2.12-mm tags (0.1 g) (Larsen et al. “approach pool” directly below the fishway. Plastic fenc- 2013; Forty et al. 2016). Tags were inserted into the abdo- ing (6.35-mm mesh) was positioned at the downstream men through a 2–5-mm incision on the ventral surface end of the release pool to prevent emigration from the anterior to the pelvic girdle (Forty et al. 2016); sutures study area; downstream emigration of test fish has been were not necessary for tag retention with such small inci- hypothesized as a cause for low participation in passage sions (Bolland et al. 2009; Larsen et al. 2013). Tagged fish studies (Hodge et al. 2017). Fencing was also installed were placed in 19-L buckets of clean stream water for along the diversion dam spillway such that the fishway about 15 min to regain equilibrium and then transferred to was the only path upstream. After release, fish movement holding pens in the stream prior to release. was monitored for 72 h using PIT telemetry. Previous pas- Thirty-two age-1, hatchery-reared Arctic Grayling (212 sage studies have shown movement of displaced fish 23mm) were released simultaneously with wild fish in all within hours of release (Armstrong and Herbert 1997; trials because of the low availability of wild Arctic Gray- Burford et al. 2009), and we observed 60.8% participation ling. Test Arctic Grayling originated from a population in (n= 74 fish) during a 72-h pilot study in 2017. Axolotl Lake, Montana, that was established from Big Hole River Arctic Grayling stock. Arctic Grayling were Antenna Construction and Operation spawned at Axolotl Lake, and embryos were transported We used four stationary PIT antennas (modified from to the Yellowstone River Trout Hatchery in Big Timber, Hodge et al. 2017) to track the approach, attraction, Montana, where they were incubated and reared at 12°C. entrance, and passage of fish (Figure 5). Antennas were Test Arctic Grayling were exercised in a flowing raceway constructed as vertically oriented swim-through loops; at about 0.3 m/s for at least 1 month prior to use in trials; antenna 1 consisted of one or two loops of 8-gauge exercise training of 4–6 weeks increases swimming perfor- copper-clad aluminum cable, and antennas 2–4 consisted mance in captive-reared fish (Davison 1989, 1997). Test of four loops of 12-gauge copper wire attached to a 0.6-m× Arctic Grayling were PIT-tagged (23 mm) at least 1 month 0.6-m wooden frame. Approaching fish were detected prior to field testing. On the day before each trial, 32 fish at antenna 1 (A1), which spanned the stream channel 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 460 TRIANO ET AL. FIGURE 5. General schematic of the study site and PIT array configuration, including locations of four PIT antennas used to record the approach (A1), attraction (A2), entrance (A3), and passage (A4) of fish; the additional antenna (A5) is not pictured. Approximate detection ranges of A1–A4 are depicted by gray boxes. Schematic is not to scale. between the release pool and approach pool. Attracted calculated as the percentage of total test tags successfully fish were detected at antenna 2 (A2), which was positioned detected at each antenna. Estimates from single-tag tests 1–3 cm downstream of the fishway entrance and detected were averaged among 13 trials to generate a single esti- fish that staged within 0.6 m of the entrance. Entering fish mate of detection probability at each antenna position were detected at antenna 3 (A3), which was positioned on (A1, A2, etc.) across all trials. A separate estimate was the third baffle inside the fishway and detected fish as they calculated similarly for each antenna position from mul- first entered the fishway. Passing fish were detected at titag tests. antenna 4 (A4), which was positioned just upstream of the We also analyzed the encounter history of each test fish fishway exit and detected fish as they passed the final baf- to generate antenna-specific detection probability estimates fle of the fishway. Each antenna was connected to a stan- from observed fish movements (Hodge et al. 2017). For dard remote tuner board and all four tuner boards were example, if a fish was released and only detected at A5 (a linked to a multiantenna HDX reader (Oregon RFID, “recorded” detection), its encounter history would be Portland, Oregon) that recorded the tag number, antenna, 100001 (the first “1” representing release, the 0s represent- date, and time of each detection. The array was powered ing missed detections at A1–A4, and the last “1” repre- by two 12-V 80-Ah absorbed glass mat batteries connected senting detection at A5). Because this fish was detected at in parallel. A fifth antenna (A5), which spanned the A5, it was assumed to have passed through all down- stream channel 30–50m upstream of the diversion dam, stream antennas (A1–A4), and “adjusted” detections were was linked to a separate multiantenna HDX reader and inserted to complete the adjusted encounter history for was used to estimate detection probability at A4. this fish (111111). Adjusted encounter histories were built for every fish, including both “recorded” and “adjusted” Detection Probability detections for all upstream and downstream movements. We conducted two in-situ detection tests prior to each The number of “recorded” detections was compared to trial to ensure proper function of each PIT antenna. First, the number of expected detections (recorded plus adjusted) separate plastic pipes (3 m long, 1.9 cm in diameter) were at each antenna to calculate antenna-specific detection placed through both A1 and the fishway (A2–A4). A sin- probability estimates for each trial (Heim et al. 2016). gle 23-mm PIT tag attached to parachute cord was pulled Antenna-specific estimates were averaged among the 13 through each pipe with a fishing rod at about 1.5 m/s to trials to generate a single estimate of detection probability simulate a single fish moving through the PIT array at each antenna position. (single-tag test). A similar test was conducted with three 23-mm PIT tags spaced 0.4 m apart to simulate multiple Data Analyses fish moving through the PIT array simultaneously (mul- Species groups.—We considered three species groups titag test) and evaluate the potential for missed detections (Arctic Grayling, trout, and suckers) to increase sample due to tag collision from overlapping signals (http:// sizes within groups for data analyses. Brown Trout, White support.oregonrfid.com). Ten single-tag and ten multitag Suckers, and Longnose Suckers were not tested in all tri- tests were conducted prior to each trial, and antenna- als, so we combined Brook Trout and Brown Trout specific detection probability estimates for each trial were (“trout” species group) and White Suckers and Longnose 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 461 Suckers (“suckers” species group). We expected similar 30 entering fish passed, passage efficiency was 50.0%. performance within these groups because the swimming Overall efficiency in this example was 15.0%, with 15 of abilities of Brook Trout and Brown Trout and that of 100 approaching fish successfully passing the fishway. White Suckers and Longnose Suckers are comparable (Jones et al. 1974; Castro-Santos et al. 2013). Further- Statistical Analyses more, similar life histories within each group provided We conducted three separate statistical analyses to (1) expectations of similar motivation and behavior during tri- estimate attraction, entrance, and passage efficiency for als. each species group (efficiency estimates and species com- Transit times.—We calculated transit time between parisons); (2) predict attraction, entrance, and passage effi- each pair of sequential antennas to evaluate how long it ciency over the range of upstream depths tested (fishway took fish to approach, attract, enter, and pass through fish- efficiency versus upstream depth); and (3) determine what ways. Approach time was calculated as the time between factors affected attraction, entrance, and passage efficiency release and first approach, attraction time as the time (multiple regression analysis). All statistical modeling was between first approach and first attraction, entrance time performed using mixed-effects logistic regressions, with as the time between first attraction and first entrance, and individual fish as the sampling unit and attraction, passage time as the time between first entrance and first entrance, and passage as three separate binary outcomes. passage. Only “recorded” detections with known detection In all analyses, the log odds of each outcome (attraction, times were considered in these analyses. entrance, and passage) were modeled as linear combina- Summarized fishway efficiency.—We summarized tions of fixed effects; random effects for study site and attraction, entrance, and passage efficiencies for each spe- trial were included in all models to account for noninde- cies group across 13 trials (Table 2) to generally assess pendent observations of test fish within each trial (Zuur which components were most limiting to overall efficiency. et al. 2009). Fish were frequently detected passing anten- Attraction, entrance, and passage efficiencies were quanti- nas multiple times during a trial; however, we only fied for each species group during each 72-h trial by ana- included a single observation per fish for each outcome lyzing adjusted encounter histories. We first omitted fish (attraction, entrance, or passage) that included its farthest that did not volitionally reach the approach pool; the upstream progress relevant to that outcome by the end of number of approaching fish was used only as a metric of the trial (Goerig et al. 2016). Hydraulic conditions fluctu- participation to calculate attraction efficiency. Attraction, ated minimally during each trial, and fish were assigned entrance, and passage efficiencies were then calculated as the average hydraulic conditions present during their the percentages of fish reaching an antenna that later respective trials (Table 1). reached the next upstream antenna. For example, if 100 Efficiency estimates and species comparisons.—Attrac- fish approached the fishway and 60 were attracted, attrac- tion, entrance, passage, and overall efficiencies were esti- tion efficiency was 60.0%. If 30 of 60 attracted fish entered mated for each species group using logistic regression, the fishway, entrance efficiency was 50.0%. Lastly, if 15 of with species group as the single fixed effect in these TABLE 2. Summarized fishway efficiencies over 13 trials; total counts (n) and fishway efficiencies (%) are reported for each species group tested (trout = Brook Trout and Brown Trout, suckers = White Sucker and Longnose Sucker). Entrance Passage Overall Species Released Approached Attracted Entered Passed Attraction efficiency efficiency efficiency group (n) (n) (n) (n) (n) efficiency (%) (%) (%) (%) Arctic 416 335 (332)a 178 86 (83)a 79 53.1 48.3 95.2 23.8 Grayling Trout 668 618 (601)a 468 350 (333)a 319 75.7 74.8 95.8 53.1 Suckers 175 154 (151)a 79 60 (57)a 55 51.3 75.9 96.5 36.4 Burbot 8 7 4 2 2 57.1 50.0 100.0 28.6 Mountain 1 1 1 1 1 100.0 100.0 100.0 100.0 Whitefish All species 1,268 1,115 (1,092)a 730 499 (476)a 456 65.5 68.4 95.8 41.8 combined aTwenty-three fish with unknown passage fate (i.e., fish that entered fishways but were not detected exiting at A2 or A4) were omitted from passage and overall effi- ciency analyses; total counts of entering and approaching fish listed in parentheses represent the number of entering and approaching fish used in passage and overall effi- ciency analyses, respectively, to account for uncertainty about those 23 fish. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 462 TRIANO ET AL. models. From the odds of success for each outcome, the (Dormann et al. 2013). As expected, the two variables probability of success was calculated by describing fishway discharge (upstream depth and fishway discharge) were strongly correlated (r= 0.91). We included ð expðβ Þ P successÞ ¼ i , upstream depth in our models because of its ease of mea- ½1þ expðβiÞ surement in the field; however, we discuss upstream depth and fishway discharge synonymously. Explanatory vari- where P(success) is the efficiency estimate for each outcome ables for attraction included upstream depth, downstream and βi is the model intercept coefficient for each species depth, attraction flow, water temperature, and fish length. group (Zuur et al. 2009). Wald 95% confidence intervals Explanatory variables for entrance included upstream were calculated for each βi. We primarily discuss the effi- depth, downstream depth, fishway slope, water tempera- ciency estimates generated by these mixed-effects models as ture, and fish length. Explanatory variables for passage opposed to the summarized efficiencies in Table 2 because included upstream depth, fishway slope, water tempera- these estimates included random effects to account for non- ture, and fish length. independent observations of test fish within each individual We compared four types of models for each outcome trial. Efficiency estimates for each species group are and species group using Akaike information criterion reported as percentages, and odds ratios with P-values corrected for small sample sizes (AICc): (1) a main adjusted for multiple tests by the “single step method” effects model that included the additive combination of (Hothorn et al. 2008) were calculated to identify statistically explanatory variables plus random effects (for study site significant differences in the odds of attraction, entrance, or and trial), (2) a full model that included all main effects passage among species groups (Zuur et al. 2009). and all two-way interactions of the variables plus ran- Fishway efficiency versus upstream depth.—We used dom effects, (3) a reduced model obtained by backward logistic regression to predict attraction, entrance, and pas- selection of the interactions from the full model (P> sage efficiencies over the range of upstream depths tested 0.05), and (4) an intercept model that included only an without considering potential effects of additional explana- intercept term and random effects. Main effects of vari- tory variables. We focused on upstream depth in these sin- ables were not removed from reduced models regardless gle covariate models because upstream depth is an easily of their significance to explicitly account for potential measurable surrogate for fishway discharge, and the effects of all relevant explanatory variables (Wasserman amount of water needed to facilitate passage through et al. 1996; Ramsey and Schafer 2002). Hence, all vari- Denil fishways is a key management question in the Big ables were always included in our models (except the Hole basin. The models included fixed effects for upstream intercept models); importantly, we did not perform vari- depth and species group and an upstream depth × species able selection, which has several limitations (Lukacs group interaction to account for potentially different asso- et al. 2010). We considered models with ΔAICc less than ciations with upstream depth among species groups. The or equal to 2 to be competitive (Burnham and Anderson probability of success for attraction, entrance, and passage 2002) and selected the most parsimonious of competitive was predicted for each species group over the continuous models as the top model for each analysis. Explanatory range of upstream depths tested (18.1–65.4 cm) by variables in top models were centered on their overall   mean values (Table 1). Exponentiation of the model ð Þ ¼ exp β0,i þ β P success   1,i US  , coefficient of each statistically significant explanatory 1þ exp β0,i þ β1,i US variable predicted the effect on the odds ratio of a one- unit increase in that variable, while all other variables where P(success) is the predicted efficiency for each out- were held constant at their overall mean values (Sen and come, β0,i is the model intercept for the ith species group, Srivastava 1990). US is upstream depth, and β1,i is the coefficient for We identified outliers and assessed goodness of fit for upstream depth for the ith species group (Zuur et al. 2009). all top models. Outliers were defined as observations >4 Multiple regression analyses.—We evaluated the effects SDs from the mean (Ramsey and Schafer 2002). We tested of slope and hydraulic conditions (and two-way interac- the influence of outliers on parameter and standard error tions) on the odds of attraction, entrance, and passage for estimation in all models; no outliers were removed before each species group using separate, multiple regression any analysis because none were identified as influential in analyses. Each model included the explanatory variables any model. Goodness of fit of the mixed-effect logistic hypothesized to be most relevant for that outcome. A regression models was confirmed by residual plots, chi- Pearson’s correlation analysis was performed on all pairs squared tests of deviance, and Hosmer–Lemeshow tests of the explanatory variables; for strongly correlated vari- (P < 0.05). All models were fit using the statistical software ables (r> 0.70) we included the single variable with the R (R Core Team 2018) package lme4 (Bates et al. 2015), most likely functional significance for that response adjusted P-values were calculated using R package 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 463 multcomp (Hothorn et al. 2008), and AICc was calculated infrequent during trials. Potential tag collision occurred using R package AICcmodavg (Mazerolle 2019). at A2 and A4 as a low percentage of entering fish (4.6%; n = 23 of 499) were not subsequently detected exiting the fishway at either end. These 23 fish were RESULTS omitted from passage and overall efficiency analyses because we could not confirm their success or failure Detection Probability in passing fishways. Detection probability of PIT arrays was high throughout the study. Detection probability estimates Transit Times from single-tag tests ranged from 99.3% to 100.0% at Transit times varied among efficiency components and A1–A4, suggesting near-perfect detection if a single fish taxa (Figure 6). In general, approach and attraction times passed through the array. In contrast, estimates from were longer than entrance and passage times. Wild fish multitag tests ranged from 76.2% to 86.4%, illustrating approached fishways more quickly than hatchery-reared potential for missed detections resulting from tag colli- Arctic Grayling; over 91.0% of approaching trout and sion. However, detection probability estimates gener- suckers approached within 24 h after release compared ated from encounter histories of test fish ranged from with 67.9% of approaching Arctic Grayling. Trout were 92.9% to 98.4%, suggesting that tag collision was attracted to the fishway more quickly than Arctic FIGURE 6. Transit times (approach, attraction, entrance, and passage times) for Arctic Grayling, trout (Brook Trout and Brown Trout), and suckers (White Sucker and Longnose Sucker). 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 464 TRIANO ET AL. FIGURE 7. Cumulative proportions of times between release and passage of 445 passing fish. Nights 1–3 are depicted as shaded areas bounded by solid vertical lines representing the average time from release until sunset and sunrise calculated over 12 trials (trial 1 was not included in these calculations because of its delayed release time); dashed vertical lines represent the variation in time after release until sunset and sunrise over those 12 trials, given the different release times. Grayling and suckers, with 47.0% of attracted trout, Summarized Fishway Efficiency 36.6% of attracted Arctic Grayling, and 31.1% of Of 1,092 approaching fish with known passage fates, 456 attracted suckers being attracted to fishways within 1 h (41.8%) successfully passed through fishways (Table 2). after their first approach. About 80.0% of all attracted fish Reductions in overall efficiency were mostly due to attraction were attracted within 24 h after their first approach. (65.5%; n= 730 of 1,115) and entrance (68.4%; n= 499 of Entrance times were shorter than attraction times for all 730) failures because nearly all fish that entered fishways suc- taxa. The majority of entering trout (79.0%), Arctic Gray- cessfully passed (95.8%; n= 456 of 476). We observed only 4 ling (64.6%), and suckers (71.2%) entered fishways within failed passage attempts by Arctic Grayling (n= 83), 2 by 1 h after their first attraction. Of those fish, mean entrance suckers (n= 57), and 14 by trout (n= 333). Moderate attrac- times were 5.2, 5.3, and 3.1 min for trout, Arctic Grayling, tion (53.1%) and entrance (48.3%) efficiencies accounted for and suckers, respectively. Upon entrance, passage times the low overall efficiency of Arctic Grayling (23.8%), whereas were brief, with 82.7% of all passes (n= 354) occurring in suckers (36.4% overall efficiency) were limited primarily by less than 1 min and 57.7% of all passes (n= 247) occurring attraction (51.3%). Trout had the highest overall efficiency in less than 10 s. Passage times were particularly brief for (53.1%) and were less limited by attraction (75.7%) and Arctic Grayling; 78.4% of passing Arctic Grayling passed entrance (74.8%) than Arctic Grayling and suckers. in less than 10 s compared with 53.7% of passing trout and 51.9% of passing suckers. Efficiency Estimates and Species Comparisons Trout passed fishways more quickly after release than Efficiency estimates derived from mixed-effects logistic suckers and Arctic Grayling, and diel periodicity appeared regressions exhibited a consistent pattern when compar- to affect timing of passage (Figure 7). The majority of isons were made across species groups (Figure 8). They passing trout (73.3%) and 47.3% of passing suckers passed were always highest for trout, intermediate for suckers, within 24 h after release, whereas Arctic Grayling passed and lowest for Arctic Grayling for each efficiency compo- at a slower and more consistent rate throughout the 72-h nent considered (attraction [trout, suckers, and Arctic trials (37.3, 33.3, and 29.3% on days 1–3, respectively). Grayling, respectively] = 84.3, 72.8, and 60.4%; entrance= Passage by all species peaked during sunset and at night 78.6, 69.3, and 44.3%; passage= 97.0, 96.4, and 96.2%; (70.3% of all passes), which was most evident for suckers, overall= 55.8, 38.5, and 19.1%). Most comparisons were with 89.1% of passing suckers passing during sunset or at statistically significant (P < 0.05; Figure 8), except for pas- night compared with 66.9% of passing trout and 70.6% of sage efficiency; nearly all fish that entered Denil fishways passing Arctic Grayling. successfully passed. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 465 FIGURE 8. Efficiency estimates (attraction, entrance, passage, and overall) and Wald 95% confidence intervals for trout (TR; Brook Trout and Brown Trout), suckers (SU; White Sucker and Longnose Sucker), and Arctic Grayling (GR) from mixed-effects logistic regression. Different letters (A, B, C) denote significant differences (P< 0.05) among species groups for each efficiency component. Fishway Efficiency versus Upstream Depth effects (Table 4). Attraction of Arctic Grayling increased Upstream depth was positively associated with attraction with upstream depth and attraction flow. A 1-cm increase and negatively associated with entrance, whereas passage in upstream depth was associated with a 7.8% increase in was predicted to be high across all upstream depths tested the odds of attraction of Arctic Grayling, and a 1.0% (Figure 9). Predicted attraction efficiencies of all species increase in attraction flow was associated with a 2.9% groups increased with upstream depth, particularly for Arctic increase in odds of attraction of Arctic Grayling. In con- Grayling, with attraction efficiency more than doubling from trast, upstream depth and attraction flow had varied about 30.0% at an 18-cm upstream depth to about 80.0% at effects on attraction of trout. A 1-cm increase in upstream 65 cm. Attraction efficiencies differed among species groups depth was associated with a 5.9% decrease in the odds of at shallower upstream depths; however, attraction efficiencies attraction of trout, and a 1.0% increase in attraction flow of all species groups were about 80.0% at a 65-cm upstream was associated with a 5.7% decrease in the odds of attrac- depth, which corresponded to a full fishway. Predicted tion of trout. However, an interaction between upstream entrance efficiencies of trout and suckers were between 60.0% depth and attraction flow indicated that increased attrac- and 80.0% across all upstream depths tested. In contrast, tion flow had a positive effect on attraction of trout at entrance efficiency of Arctic Grayling was predicted to fishways with shallow upstream depths; alternatively, decrease considerably at deeper upstream depths, from about increased upstream depth had a positive effect on attrac- 80.0% at an 18-cm upstream depth to about 20.0% at 65 cm. tion of trout at fishways with low attraction flows. Predicted passage efficiency of all species groups across all Attraction of all species groups increased with water upstream depths tested was high (>85.0%). temperature, and attraction of trout increased with down- stream depth (Table 4). A 1°C increase in water tempera- Multiple Regression Analyses ture was associated with increased odds of attraction of Attraction.— The top models for attraction of Arctic Arctic Grayling (by 28.9%), trout (by 31.1%), and suckers Grayling, trout, and suckers (Table 3) included significant (by 44.5%). A 1-cm increase in downstream depth was effects of upstream depth, attraction flow, water tempera- associated with a 22.5% increase in odds of attraction of ture, and downstream depth as well as several interactive trout. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 466 TRIANO ET AL. FIGURE 9. Relationships between fishway efficiencies (attraction, entrance, and passage) and upstream depth for Arctic Grayling, trout (Brook Trout and Brown Trout), and suckers (White Sucker and Longnose Sucker). Lines represent predicted efficiencies from mixed-effects logistic regressions, and each point represents an observed efficiency for one species group during an individual trial. Observed efficiencies are calculated on 1 to 58 fish; passage efficiencies ≤50.0% are based on 1 or 2 fish. Entrance.— The top models for entrance of Arctic (by 34.8%) and trout (by 37.0%). In contrast, entrance of Grayling and trout (Table 3) included significant effects of Arctic Grayling and trout increased with downstream upstream depth, downstream depth, slope, and water tem- depth. A 1-cm increase in downstream depth was associ- perature as well as several interactive effects (Table 4). ated with increased odds of entrance of Arctic Grayling Entrance of Arctic Grayling and trout decreased with (by 12.2%) and trout (by 16.3%). The positive effect of upstream depth, and this association was strongest at fish- downstream depth was stronger at fishways with gradual ways with steep slopes. A 1-cm increase in upstream depth slopes than at fishways with steep slopes, probably was associated with decreased odds of entrance of Arctic because fishways with gradual slopes often had shallow Grayling (by 9.8%) and trout (by 5.0%). For trout, deeper downstream depths that limited entrance. Entrance of upstream depths were more limiting to entrance at fish- trout increased with water temperature (Table 4). A 1°C ways with steep slopes than at fishways with gradual increase in water temperature was associated with a 9.4% slopes. Entrance of Arctic Grayling and trout also increase in the odds of entrance of trout. decreased with slope; a 1% increase in slope was associ- Passage.— The top models for passage of Arctic Gray- ated with decreased odds of entrance of Arctic Grayling ling and suckers (Table 3) showed no statistically 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 467 TABLE 3. Akaike information criterion (AICc) model selection for attraction, entrance, and passage of Arctic Grayling, trout (Brook Trout and Brown Trout), and suckers (White Sucker and Longnose Sucker) at Denil fishways in the Big Hole River basin, Montana. The number of parameters (K), AICc scores (AICc), ΔAICc describing the difference between AICc of modeli and the best model, and the Akaike weights (wi) describing the relative likelihood of being the best model among those tested are reported for each model; all tested models included random effects. The top models from each analysis are highlighted in bold. Abbreviations are as follows: U = upstream depth, A = attraction flow, D = downstream depth, T = water temperature, L = fish total length, and S = fishway slope. Efficiency Species group Model Parameters K AICc ΔAICc wi Attraction Arctic Grayling (n= 335) Reduced U +A + T + L + D +U×A +A×T 10 356.71 0.00 0.81 Arctic Grayling Full U+A + T +L + D + all two-way interactions 18 360.22 3.51 0.14 Arctic Grayling Main effects U+A + T +L + D 8 363.1 6.39 0.03 Arctic Grayling Intercept Random effects only 3 365.09 8.38 0.01 Trout (n= 618) Reduced U +A + T + L + D +U×A +U×L +U×D +A×T +A×D 13 547.77 0.00 0.77 Trout Full U+A + T +L + D + all two-way interactions 18 550.23 2.46 0.23 Trout Main effects U+A + T +L + D 8 569.88 22.11 0 Trout Intercept Random effects only 3 571.66 23.89 0 Suckers (n= 154) Full a U+A + T +L + D + all two-way interactions 16 168.27 0 Suckers Main effects U +A + T + L + D 8 169.00 0.73 0.91 Suckers Intercept Random effects only 3 173.53 5.26 0.09 Entrance Arctic Grayling (n= 178) Reduced U + S + T + L + D + S×D 9 215.9 0.00 0.99 Arctic Grayling Main effects U+ S + T +L + D 8 226.18 10.28 0.01 Arctic Grayling Full U+ S + T +L + D + all two-way interactions 18 230.02 14.12 0 Arctic Grayling Intercept Random effects only 3 234.52 18.63 0 Trout (n= 468) Reduced U + S + T + L + D +U×S +U×A + S×A + S×D 12 476.52 0.00 0.83 Trout Main effects U+ S + T +L + D 8 480.59 4.07 0.11 Trout Full U+ S + T +L + D + all two-way interactions 18 482.38 5.86 0.04 Trout Intercept Random effects only 3 484.13 7.61 0.02 Suckers (n= 79) Main effects U+ S + T +L + D 8 86.62 0 0.69 Suckers Fulla U+ S + T +L + D + all two-way interactions 16 87.63 1.01 Suckers Intercept Random effects only 3 88.19 1.57 0.31 Passage Arctic Grayling (n= 83) Fulla U+ S + T +L + all two-way interactions 11 31.52 0.00 Arctic Grayling Main effects U+ S + T +L 7 32.03 0.51 0.68 Arctic Grayling Intercept Random effects only 3 33.52 2.00 0.32 Trout (n= 333) Main effects U + S + T + L 7 108.61 0.00 0.91 Trout Full U+ S + T +L + all two-way interactions 13 113.21 4.60 0.09 Trout Intercept Random effects only 3 121.34 12.73 0.00 Suckers (n= 57) Intercept Random effects only 3 23.31 0.00 0.98 Suckers Main effects U+ S + T +L 7 31.01 7.70 0.02 Suckers Fulla U+ S + T +L + all two-way interactions 11 31.38 8.07 aModels could not be fit with random effects for study site and trial; no interactions were significant in these models so reduced models were identical to main-effects models for these analyses. Akaike weights were not compared between fixed and mixed-effects models. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 468 TRIANO ET AL. TABLE 4. Parameter estimates from top models for attraction, entrance, and passage of Arctic Grayling, trout (Brook Trout and Brown Trout), and suckers (White Sucker and Longnose Sucker). Parameter estimates (β), standard errors (SEs), P-values (P), and odds ratios (ORs) are reported for the top models in each analysis. All explanatory variables have been centered on their overall mean values (Table 1). Significant parameter estimates (P< 0.05) are denoted by an asterisk. Odds ratios describing the effect of a one-unit increase in each parameter are computed for all statistically significant main effects by exponentiation of parameter estimates (β). Explanatory variables are abbreviated as in Table 3. Trout Arctic Grayling Suckers Efficiency β Parameter SE P OR β SE P OR β SE P OR Attraction β0 Intercept 3.841* 0.675 <0.001 –0.483 0.602 0.422 1.970 1.091 0.071 β1 U (cm) –0.061* 0.024 0.009 0.941 0.075* 0.022 <0.001 1.078 –0.004 0.032 0.888 β2 A (%) –0.059* 0.015 <0.001 0.943 0.029* 0.010 0.003 1.029 0.011 0.015 0.444 β3 D (cm) 0.203* 0.038 <0.001 1.225 –0.011 0.026 0.667 0.112 0.073 0.125 β4 T (°C) 0.271* 0.094 0.004 1.311 0.254* 0.087 0.004 1.289 0.368* 0.094 <0.001 1.445 β5 L (cm) –0.005 0.020 0.787 0.090 0.065 0.167 0.016 0.043 0.711 β6 U (cm) × A (%) –0.007* 0.002 <0.001 0.003* 0.001 <0.001 β7 A (%) × T (°C) –0.005* 0.002 0.006 –0.010* 0.003 0.002 β8 D (cm) × U (cm) –0.013* 0.004 <0.001 β9 U (cm) × L (cm) 0.004* 0.002 0.008 β10 D (cm) × A (%) 0.003* 0.001 0.004 Entrance β0 Intercept 1.633* 0.281 <0.001 0.776* 0.314 0.014 1.461* β1 U (cm) –0.051* 0.015 <0.001 0.950 –0.103* 0.022 <0.001 0.902 β2 S (%) –0.462* 0.101 <0.001 0.630 –0.427* 0.136 0.002 0.652 β3 D (cm) 0.151* 0.035 <0.001 1.163 0.115* 0.030 <0.001 1.122 β4 L (cm) 0.028 0.022 0.205 0.146 0.086 0.088 β5 T (°C) 0.090* 0.044 0.042 1.094 0.031 0.056 0.577 β6 S (%) × D (cm) –0.025* 0.007 <0.001 –0.019* 0.007 0.004 β7 U (cm) × S (%) –0.006* 0.003 0.046 β8 U (cm) × T (°C) –0.010* 0.004 0.025 β9 S (%) × T (°C) –0.073* 0.028 0.010 Passage β0 Intercept 4.176* 0.555 <0.001 8.566* 4.182* β1 U (cm) –0.971 0.730 0.184 β2 S (%) –0.113 0.094 0.231 β3 T (°C) 0.221* 0.087 0.011 1.247 β4 L (cm) 0.179* 0.063 0.005 1.196 ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 469 significant effects by any explanatory variable; however, of hatchery-reared Arctic Grayling was limited at low fish- the top model for passage of trout included significant way discharges and low attraction flows, presumably positive effects of water temperature and fish length because fish are attracted to areas of high discharge (Bunt (Table 4). et al. 2012) and discharge occurring away from the fish- way entrance can distract approaching fish (Bunt et al. 1999). In contrast, attraction of wild test fish was less lim- DISCUSSION ited by low discharges and low attraction flow. Attraction Knowledge of fishway efficiency is limited by a lack of can be driven by biological characteristics of fish, behav- comprehensive field evaluations and the use of nonstan- ior, and motivation (Bunt et al. 2012), and our test Arctic dardized efficiency components (Bunt et al. 2012; Kemp Grayling were naïve to study streams and lacked the hom- 2016). Attraction efficiency estimates may be biased by ing motivation expected from displaced wild test fish. crowding of test fish to induce participation (Haro et al. Ours is the first study of Denil fishways to quantify 1999; Mallen-Cooper and Stuart 2007), and studies that entrance as a distinct efficiency component, and we found quantify attraction and passage efficiencies often fail to that entrance failure contributed significantly to reductions consider entrance efficiency. Hodge et al. (2017) defined in overall efficiency. Our entrance efficiency estimates were “attraction” as the probability that an approaching fish between 44.3% and 78.6%, and entrance of Arctic Gray- located the entrance and defined “passage” as the proba- ling and trout was limited at fishways with deep upstream bility that an entering fish passed the structure; however, depths and steep slopes. Deep upstream depths and steep they did not distinguish between attracted and entering slopes are correlated to high velocities and turbulence that fish. Forty et al. (2016) used “proportion of displaced fish have limited passage efficiency (Bunt et al. 1999; Haro attempting passage” as a metric of motivation but com- et al. 1999; Mallen-Cooper and Stuart 2007); however, bined approach, attraction, and entrance into a single effi- low passage efficiencies previously reported may have been ciency component. Such inconsistent methodology and a result of both entrance and passage failure (Bunt et al. terminology have prevented comparisons among studies 2012). Previous studies did not distinguish between and limited our understanding of the distinct components entrance and passage, and we provide evidence that deep of overall fishway efficiency (Bunt et al. 2012, 2016; Kemp upstream depths and steep slopes limited entrance but did 2016; Williams and Katopodis 2016; Hershey 2021). not affect passage. These results highlight entrance as a We identified where fish were limited in using fishways distinct limiting component of overall efficiency and direct by quantifying approach, attraction, entrance, and passage attention to entrance conditions related to downstream as four distinct components of overall fishway efficiency. depth. Our PIT array allowed us to distinguish between four Entrance of Arctic Grayling and trout was limited at types of potential failure: (1) released fish that never shallow downstream depths, presumably because shallow approached the fishway (approach failure), (2) approach- downstream depths can create plunging conditions at fish- ing fish that never located the entrance (attraction failure), way entrances (Figure 3). When upstream depth is greater (3) attracted fish that never entered the fishway (entrance than downstream depth, water accelerates and plunges failure), and (4) entering fish that failed to pass the fish- through the fishway and out of the entrance, a condition way (passage failure). By explicitly evaluating each com- that limits upstream passage of Arctic Grayling through ponent of overall efficiency, we determined that attraction Denil fishways (Blank et al., In press) and appeared to and entrance failure reduced overall efficiency, whereas limit entrance of Arctic Grayling (6.3%; n= 16) and trout nearly all fish that entered fishways successfully passed. (25.6%; n= 43) during trial 11 (Figure 10). Entrance dur- High detection probabilities of our and similar PIT arrays ing trial 11 may have also been limited by high discharge (Goerig et al. 2016; Hodge et al. 2017) strengthen these and steep slope (Table 1), but entrance efficiencies during results and confirm PIT telemetry as an accurate means to trial 2 at a nearly identical steep slope and high discharge distinguish between potential causes of failure. Further, were higher (37.5% and 73.3% by Arctic Grayling [n= 16] we ameliorated less than perfect detection probabilities by and trout [n= 30], respectively), probably because the incorporating known missed detections into our encounter entrance was submerged. A submerged entrance prevents histories to improve the accuracy of our results. plunging and presumably improves entrance conditions Attraction failure reduced overall efficiency; however, (Figure 2), emphasizing the potential benefits of deep our attraction efficiency estimates (60.4–84.3%) were simi- downstream depths at steep fishways during high dis- lar to, and considerably higher than, the mean attraction charges. efficiency (61.0%) reported in a multispecies meta-analysis Our passage efficiency estimates (96.2–99.0%) were by Bunt et al. (2012). Further, we think our estimates may higher and less variable than previous Denil passage effi- be conservative compared with those expected from natu- ciencies (mean = 51.0%, range = 0.0–97.0%; Bunt et al. rally motivated wild fish (Hodge et al. 2017). Attraction 2012) despite testing a wide range of conditions. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 470 TRIANO ET AL. FIGURE 10. Photo of the plunging entrance conditions during trial 11. The deep upstream depth, shallow downstream depth, and steep slope (Table 1) resulted in this entrance plunge that presumably limited the entrance of Arctic Grayling, Brook Trout, and Brown Trout during this trial. Consistently high passage success was not unexpected et al. 2018), and swimming performance of hatchery- because Denil fishways create favorable low-velocity zones reared Arctic Grayling and captive wild Brook Trout and inside the fishways over a wide range of discharges Brown Trout (Castro-Santos et al. 2013; Dockery et al. (USFWS 2017). We observed highly efficient passage at 2020) are comparable. Therefore, the low entrance effi- discharges between 0.007 and 0.209 m3/s (i.e., a nearly ciency of test Arctic Grayling probably reflects behavioral empty fishway to a completely full fishway), demonstrat- differences associated with hatchery-rearing conditions or ing their versatility for providing year-round passage the innate behavioral preferences of Arctic Grayling, opportunities. Moreover, we identified entrance as a sig- which may avoid turbulent fishway entrances because they nificant bottleneck separate from passage. Therefore, our prefer to swim around barriers rather than to jump over passage efficiency estimates more accurately depict actual them (Cutting et al. 2018). Future testing of wild Arctic passage than do earlier estimates, which combined Grayling at Big Hole Denil fishways is important to entrance and passage failure. address these uncertainties. We suspect that the attraction and entrance estimates Denil fishways provide upstream passage opportunities for our test Arctic Grayling were conservative compared through artificial barriers in the Big Hole River basin; with what could be reasonably expected from wild Big however, we determined that enhancing attraction flows Hole River Arctic Grayling. Moderate attraction efficiency and maintaining deep downstream depths could improve (60.4%) of test Arctic Grayling may have been due to attraction and entrance efficiency. Attraction flows could behavioral differences stemming from hatchery rearing, be enhanced by ensuring that any discharge escaping over naïveté to study streams, lack of homing motivation, and and through the diversion dam is redirected through the short trial durations. Test Arctic Grayling were typically fishway (Triano 2020). Modifications to improve entrance slower to approach fishways and approached in lower could be prioritized at fishways with shallow slopes numbers than wild test fish, which appeared to have a (≤5.0%), where shallow downstream depths and plunging strong homing motivation. Low entrance efficiency entrance conditions were typically observed, and at fish- (44.3%) of test Arctic Grayling could reflect behavioral or ways with steep slopes (≥15.0%), where deep downstream physiological limitations; however, their high passage effi- depths could increase entrance efficiency during high fish- ciency (96.2%) and brief passage times suggest they had way discharges. Fluctuations in fishway depths could be swimming abilities comparable to wild fish. Arctic Gray- estimated by seasonal monitoring or hydraulic modeling ling are strong burst swimmers (Northcote 1995; Cahoon to guide fishway modifications and installations; the 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 471 downstream invert elevation could be set to provide deep product names is for descriptive purposes only and does downstream depths and favorable entrance conditions not imply endorsement by the U.S. Government. This throughout the year (Clay 1995; Platt 2019). study was performed under the auspices of MSU Institu- Our results could be applied to a basinwide connectiv- tional Animal Care and Use Protocol 2018-71. There is ity assessment for Arctic Grayling in the Big Hole River. no conflict of interest declared in this article. Access to critical habitats could be essentially precluded considering the potential compounding effects of passing multiple sequential barriers with limited overall efficiencies REFERENCES (Cutting et al. 2018; Keefer et al. 2021); our results could Alò, D., and T. F. Turner. 2005. Effects of habitat fragmentation on be extrapolated to barriers throughout the basin based on effective population size in the endangered Rio Grande Silvery Min- their hydraulic characteristics to estimate these effects. now. Conservation Biology 19:1138–1148. Armstrong, J. D., and N. A. Herbert. 1997. Homing movements of Designation of critical habitats and determination of displaced stream-dwelling Brown Trout. Journal of Fish Biology 50:445– specific Arctic Grayling movement patterns would help 449. prioritize where fishway modifications would be most use- Baker, N., A. Haro, B. Watten, J. Noreika, and J. D. Bolland. 2019. ful. Lastly, effective downstream passage and prevention Comparison of attraction, entrance and passage of downstream of entrainment in irrigation ditches are critical for overall migrant American Eels (Anguilla rostrata) through airlift and siphon deep entrance bypass systems. Ecological Engineering 126:74–82. connectivity (Gale et al. 2008; Calles and Greenberg 2009) Baras, E., and M. C. Lucas. 2001. Impact of man’s modifications of river and our comprehensive methods can be applied to down- hydrology on the migration of freshwater fishes: a mechanistic per- stream passage. spective. International Journal of Ecohydrology and Hydrobiology 1:291–304. Bates, D., M. Maechler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1– ACKNOWLEDGMENTS 48. This study was funded by the U.S. Geological Survey Blank, M., K. M. Kappenman, E. Ryan, and K. Banner. In press. The Science Support Partnership Program, the U.S. Fish and effect of water depth on passage success of Arctic Grayling through Wildlife Service (USFWS) Region 6 Fish Passage Pro- two Denil fishways. Journal of Ecohydraulics. DOI: 10.1080/ gram and Bozeman Fish Technology Center, the George 24705357.2021.1978346. Bolland, J. D., I. G. Cowx, and M. C. Lucas. 2009. Evaluation of VIE Grant Chapter of Trout Unlimited, and Montana State and PIT tagging methods for juvenile cyprinid fishes. Journal of University’s (MSU) Department of Ecology, College of Applied Ichthyology 25:381–386. Engineering, and Western Transportation Institute. This Bunt, C. M., T. Castro-Santos, and A. Haro. 2012. Performance of fish research involved participation from Montana Fish, Wild- passage structures at upstream barriers to migration. River Research life and Parks (MTFWP), the USFWS, the Montana and Applications 28:457–478. Bunt, C. M., T. Castro-Santos, and A. Haro. 2016. Reinforcement and Department of Natural Resources and Conservation, the validation of the analyses and conclusions related to fishway evalua- Natural Resources Conservation Service, the MSU Col- tion data from Bunt et al.: “Performance of fish passage structures at lege of Engineering, and the U.S. Forest Service. We upstream barriers to migration”. River Research and Applications extend thanks to Bill Rice and Jim Magee (USFWS) for 32:2125–2137. their help in establishing this project, Austin McCullough Bunt, C. M., C. Katopodis, and R. S. McKinely. 1999. Attraction and passage efficiency of White Suckers and Smallmouth Bass by two (MTFWP), Mike Roberts (Montana Department of Natu- Denil fishways. North American Journal of Fisheries Management ral Resources and Conservation), Jacqueline Knutson 19:793–803. (Natural Resources Conservation Service), and Adam Burford, D. D., T. E. McMahon, J. E. Cahoon, and M. Blank. 2009. Braddock (USFWS) for their assistance in developing and Assessment of trout passage through culverts in a large Montana executing the objectives of this project, Joel Cahoon drainage during summer low flow. North American Journal of Fish- eries Management 29:739–752. (MSU) for providing valuable hydraulic and hydrologic Burnham, K. P., and D. R. Anderson. 2002. Model selection and multi- expertise, Chris Phillips and staff at the Yellowstone River model inference: a practical information-theoretic approach, 2nd edi- Trout Hatchery (MTFWP) for rearing the Arctic Grayling tion. Springer Verlag, New York. tested in this study, and the staff at the U.S. Forest Ser- Cahoon, J., K. Kappenman, E. Ryan, A. Jones, K. Plymesser, and M. vice Ranger Station in Wisdom, Montana, for providing Blank. 2018. Swimming capabilities of Arctic Grayling. Northwest Science 92:234–239. housing accommodations. Furthermore, this research was Calles, O., and L. Greenberg. 2009. Connectivity is a two-way street–the conducted thanks to the voluntary participation of need for a holistic approach to fish passage problems in regulated riv- landowners involved in the CCAA program. The manu- ers. River Research and Applications 25:1268–1286. script benefited from comments by Dr. Chris Myrick, Castro-Santos, T., F. J. Sanz-Ronda, and J. Ruiz-Legazpi. 2013. Break- Matthew Keefer, and two anonymous reviewers. The ing the speed limit–comparative sprinting performance of Brook Trout (Salvelinus fontinalis) and Brown Trout (Salmo trutta). Cana- Montana Cooperative Fishery Research Unit is jointly dian Journal of Fisheries and Aquatic Sciences 70:280–293. sponsored by the U.S. Geological Survey, MTFWP, Clay, C. H. 1995. Design of fishways and other fish facilities, 2nd edition. MSU, and the USFWS. Any use of trade, firm, or Lewis Publishers, Boca Raton, Florida. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 472 TRIANO ET AL. Cooke, S. J., and S. G. Hinch. 2013. Improving the reliability of fishway Katopodis, C. 1992. Introduction to fishway design. Fisheries and Oceans attraction and passage efficiency estimates to inform fishway engineer- Canada, Freshwater Institute, Winnipeg, Manitoba. ing, science, and practice. Ecological Engineering 58:123–132. Katopodis, C., N. Rajaratnam, S. Wu, and D. Tovell. 1997. Denil fish- Cutting, K. A., J. M. Ferguson, M. L. Anderson, K. Cook, S. C. Davis, ways of varying geometry. Journal of Hydraulic Engineering and R. Levine. 2018. Linking beaver dam affected flow dynamics to 123:624–631. upstream passage of Arctic Grayling. Ecology and Evolution Kaya, C. M. 1992. Review of the decline and status of fluvial Arctic 8:12905–12917. Grayling (Thymallus arcticus) in Montana. Proceedings of the Mon- Davison, W. 1989. Training and its effects on teleost fish. Comparative tana Academy of Sciences 52:43–70. Biochemistry and Physiology Part A: Physiology 94:1–10. Keefer, M. L., M. A. Jepson, T. S. Clabough, and C. C. Caudill. 2021. Davison, W. 1997. The effects of exercise training on teleost fish, a Technical fishway passage structures provide high passage efficiency review of recent literature. Comparative Biochemistry Physiology Part and effective passage for adult Pacific salmonids at eight large dams. A: Physiology 117:67–75. PLOS (Public Library of Science) ONE 16(9):e0256805. DNRC (Montana Department of Natural Resources and Conservation). Kemp, P. S. 2016. Meta-analyses, metrics and motivation: mixed mes- 1979. River mile index of the Missouri River. DNRC, Helena. sages in the fish passage debate. River Research and Applications Dockery, D. R., E. Ryan, K. M. Kappenman, and M. Blank. 2020. 32:2116–2124. Swimming performance of Arctic Grayling (Thymallus arcticus Pallas) Lamothe, P., and J. Magee. 2003. Movement and habitat selection of in an open-channel flume. Journal of Ecohydraulics 5:31–42. Arctic Arctic Grayling, Brook Trout, and Mountain Whitefish during Dormann, C. F., J. Elith, S. Bacher, C. Buchmann, G. Carl, G. Carré, J. drought conditions in the Big Hole River, Montana. Montana R. G. Marquéz, B. Gruber, B. Lacfourcade, P. J. Leitão, T. Department of Fish, Wildlife and Parks, Bozeman. Münkemüller, C. McClean, P. E. Osborne, B. Reineking, B. Larsen, M. H., A. N. Thorn, C. Skov, and K. Aarestrup. 2013. Effects Schröder, A. K. Skidmore, D. Zurell, and S. Lautenbach. 2013. of passive integrated transponder tags on survival and growth of juve- Collinearity: a review of methods to deal with it and a simulation nile Atlantic Salmon Salmo salar. Animal Biotelemetry 1:article 19. study evaluating their performance. Ecography 36:27–46. Lukacs, P. M., K. P. Burnham, and D. R. Anderson. 2010. Model selec- FAO (Food and Agricultural Organization of the United Nations) and tion bias and Freedman’s paradox. Annals of the Institute of Statisti- DVWK (Deutscher Verband für Wasserwirtschaft und Kulturbau cal Mathematics 62:117–125. e.V.). 2002. Fish passes–design, dimensions and monitoring. FAO, in Mallen-Cooper, M., and I. G. Stuart. 2007. Optimizing Denil fishways arrangement with Deutscher Verband für Wasserwirtschaft und Kul- for passage of small and large fishes. Fisheries Management and Ecol- turbau e.V., Rome. ogy 14:61–71. Forty, M., J. Spees, and M. C. Lucas. 2016. Not just for adults! Evaluat- Mazerolle, M. J. 2019. AICcmodavg: model selection and multimodel ing the performance of multiple fish passage designs at low-head bar- inference based on (Q)AIC(c). R package version 2.2-1. Available: riers for the upstream movement of juvenile and adult trout Salmo https://cran.r-project.org/package=AICcmodavg. trutta. Ecological Engineering 94:214–224. Mesa, M. G., and C. D. Magie. 2006. Evaluation of energy expenditure Gale, S. B., A. V. Zale, and C. G. Clancy. 2008. Effectiveness of fish in adult spring Chinook Salmon migrating upstream in the Columbia screens to prevent entrainment of Westslope Cutthroat Trout into irri- River basin: an assessment based on sequential proximate analysis. gation canals. North American Journal of Fisheries Management River Research and Applications 22:1085–1095. 28:1541–1553. Morita, K., S. Yamamoto, and N. Hoshino. 2000. Extreme life history Goerig, E., T. Castro-Santos, and N. E. Bergeron. 2016. Brook Trout change of White-spotted Char (Salvelinus leucomaenis) after damming. passage through culverts. Canadian Journal of Fisheries and Aquatic Canadian Journal of Fisheries and Aquatic Sciences 57:1300–1306. Sciences 73:94–104. MTFWP (Montana Fish, Wildlife and Parks) and USFWS (U.S. Fish Haro, A., M. Odeh, T. Castro-Santos, and J. Noreika. 1999. Effect of and Wildlife Service). 2006. Candidate agreement with assurances for slope and headpond on passage of American Shad and Blueback Her- fluvial Arctic Grayling in the upper Big Hole River. U.S. Fish and ring through simple Denil and deepened Alaska steeppass fishways. Wildlife Service, Tracking #TE104415-0, Washington, D.C. North American Journal of Fisheries Management 19:51–58. Newton, M., J. A. Dodd, J. Barry, P. Boylan, and C. E. Adams. 2018. Heim, K. C., M. S. Wipfli, M. S. Whitman, C. D. Arp, J. Adams, and J. The impact of a small-scale riverine obstacle on the upstream migra- A. Falke. 2016. Seasonal cues of Arctic Grayling movement in a tion of Atlantic Salmon. Hydrobiologia 806:251–264. small Arctic stream: the importance of surface water connectivity. Noonan, M. J., J. W. A. Grant, and C. D. Jackson. 2012. A quantita- Environmental Biology of Fishes 99:49–65. tive assessment of fish passage efficiency. Fish and Fisheries 13: Hershey, H. 2021. Updating the consensus on fishway efficiency: a meta- 450–464. analysis. Fish and Fisheries 22:735–748. Northcote, T. G. 1984. Mechanisms of fish migrations in rivers. Pages Hodge, B. W., E. R. Fetherman, K. B. Rogers, and R. Henderson. 2017. 317–355 in J. D. McCleave, G. P. Arnold, J. J. Dodson, and W. H. Effectiveness of a fishway for restoring passage of Colorado River Neill, editors. Mechanisms of migration in fishes. Plenum Press, New Cutthroat Trout. North American Journal of Fisheries Management York. 37:1332–1340. Northcote, T. G. 1995. Comparative biology and management of Arctic Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference in and European Arctic Grayling (Salmonidae, Thymallus). Reviews in general parametric models. Biometrical Journal 50:346–363. Fish Biology and Fisheries 5:141–194. Januchowski-Hartley, S. R., P. B. McIntyre, M. Diebel, P. J. Doran, D. Odeh, M. 2003. Discharge rating equation and hydraulic characteristics M. Infante, C. Joseph, and J. D. Allan. 2013. Restoring aquatic of standard Denil fishways. Journal of Hydraulic Engineering ecosystem connectivity requires expanding inventories of both dams 129:341–348. and road crossings. Frontiers in Ecology and the Environment O’Hanley, J. R., and D. Tomberlin. 2005. Optimizing the removal of 11:211–217. small fish passage barriers. Environmental Modeling and Assessment Jones, D. R., J. W. Kiceniuk, and O. S. Bamford. 1974. Evaluation of 10:85–98. the swimming performance of several fish species from the MacKen- Oswald, R. A. 2000. Inventory and survey of the salmonid populations zie River. Journal of the Fisheries Research Board of Canada of the Big Hole River of southwest Montana, 1981–1999. Montana 31:1641–1647. Department of Fish, Wildlife and Parks, Bozeman. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ATTRACTION, ENTRANCE, AND PASSAGE EFFICIENCY AT DENIL FISHWAYS 473 Platt, N. C. 2019. Designing and assessing the effectiveness of Denil fish- Tack, S. L., and J. R. Fisher. 1977. Performance of Arctic Grayling in a ways using hydraulic modeling-based approaches. Master’s thesis. twenty foot section of model “A” Alaska steeppass fish ladder. Final Montana State University, Bozeman. Report to the U.S. Army Corps of Engineers, Alaska Division, R Core Team. 2018. R: a language and environment for statistical com- Anchorage. puting. R Foundation for Statistical Computing, Vienna. Available: Triano, B. T. 2020. Attraction, entrance, and passage efficiency of Arctic https://www.R-project.org/. Grayling, trout, and suckers at Denil fishways in the Big Hole River Rajaratnam, N., C. Katopodis, S. Wu, and M. A. Sabur. 1997. Hydrau- basin, Montana. Master’s thesis. Montana State University, Bozeman. lics of resting pools for Denil fishways. Journal of Hydraulic Engi- USFWS (U.S. Fish and Wildlife Service). 2017. Fish passage engineering neering 123:632–638. design criteria. USFWS, Northeast Region R5, Hadley, Mas- Ramsey, F., and D. Schafer. 2002. The statistical sleuth—a course in meth- sachusetts. ods of data analysis, 2nd edition. Brooks Cole, Belmont, California. USOFR (U.S. Office of the Federal Register). 2014. Endangered and Roni, P., K. Hanson, and T. Beechie. 2008. Global review of the physical threatened wildlife and plants; revised 12-month finding on a petition and biological effectiveness of stream habitat rehabilitation tech- to list the upper Missouri River distinct population segment of Arctic niques. North American Journal of Fisheries Management 28:856– Grayling as endangered or threatened; proposed rule. Federal Regis- 890. ter 79:161(20 August 2014):49384–49422. Schmetterling, D. A., R. W. Pierce, and B. W. Liermann. 2002. Efficacy Vatland, S. J. 2015. Effects of stream temperature and climate change on of three Denil fish ladders for low-flow fish passage in two tributaries fluvial Arctic Grayling and non-native salmonids in the upper Big to the Blackfoot River, Montana. North American Journal of Fish- Hole River, Montana. Doctoral dissertation. Montana State Univer- eries Management 22:929–933. sity, Bozeman. Sen, A., and M. Srivastava. 1990. Regression analysis–theory, methods, Wasserman, W., J. Neter, C. J. Nachtsheim, and M. H. Kutner. 1996. and applications. Springer Verlag, New York. Applied linear statistic models, 4th edition. McGraw-Hill Higher Shepard, B. B., and R. A. Oswald. 1989. Timing, location, and popula- Education, Chicago, Illinois. tion characteristics of spawning Montana Arctic Grayling Thymallus Williams, J. G., and C. Katapodis. 2016. Commentary–incorrect applica- arcticus in the Big Hole River drainage, 1988. Montana Fish, Wildlife tion of data negates some meta-analyses results in Bunt et al. (2012). and Parks, Bozeman. River Research and Applications 32:2109–2115. Sladek, H. 2013. Trend analysis of water temperatures relative to air tem- Zuur, A., E. N. Ieno, N. Walker, A. A. Saveliev, and G. M. Smith. peratures and flow in the Big Hole River. U.S. Fish and Wildlife Ser- 2009. Mixed effects models and extensions in ecology with R. vice, Final Report, Dillon, Montana. Springer Science+Business Media, New York. 15488659, 2022, 4, Downloaded from https://afspubs.onlinelibrary.wiley.com/doi/10.1002/tafs.10362 by Montana State University Library, Wiley Online Library on [31/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License