Tributary Use by Imperiled Flannelmouth and Bluehead Suckers in the Upper Colorado River Basin
Abstract
Habitat alterations and establishment of nonnative fishes have reduced the distributions of Flannelmouth Sucker Catostomus latipinnis and Bluehead Sucker C. discobolus to less than 50% of their historical ranges in the Colorado River basin. Tributaries are sometimes less altered than main-stem habitat in the basin and may be important to support various life history processes, but their role in the maintenance of Flannelmouth Sucker and Bluehead Sucker populations is poorly understood. Using mark–recapture techniques, we show tributaries are important habitat for native suckers in the upper Colorado River basin and report three main findings. First, both Flannelmouth and Bluehead suckers likely respond to a thermal cue that initiates spawning movement patterns. Suckers moved into Coal Creek from the White River beginning in mid-May of 2012 and 2013 to spawn. The majority of sucker spawning movements occurred when water temperatures in White River exceeded 11–14°C and those in Coal Creek were 2.5–4°C warmer, while flows varied between years. Second, based on PIT tag detection arrays, 13–45% of suckers showed spawning site fidelity. Sampling only with fyke nets would have resulted in the conclusion that site fidelity by native suckers was only 1–17%, because nets were less efficient at detecting marked fish. Third, most suckers of both species emigrated from Coal Creek within 48 h after being captured while suckers that were detected only via arrays remained resident for 10–12 d. The posthandling flight response we observed was not anticipated and to our knowledge has not been previously reported for these species. Remote PIT tag antenna arrays allowed for a stronger inference regarding movement and tributary use by these species than what could be achieved using just fyke nets. Tributaries are an important part of Flannelmouth Sucker and Bluehead Sucker life history and thus important to conservation strategies for these species.
Received October 7, 2016; accepted March 20, 2017 Published online June 28, 2017
Tributaries are important sources of habitat diversity in a drainage network because they increase channel complexity, substrate diversity, and depth variation (Benda et al. 2004; Moyle and Mount 2007; Pracheil et al. 2009). Tributaries also influence main-stem river water temperature, flow, and sediment loads (Benda et al. 2004; Moyle and Mount 2007; Sabo et al. 2012). Tributaries and associated abiotic processes may increase species diversity and are important for many riverine fishes to complete their life histories, especially in systems where main-stem habitat is highly altered and degraded (Stevens et al. 1997; Pracheil et al. 2009, 2013; Sabo et al. 2010; Webber et al. 2013). Tributaries are increasingly important for conservation of riverine fish because they are often less degraded than main-stem habitat, and conservation of smaller tributaries may be more practical than main-stem habitats altered for hydropower or transportation (Moyle and Mount 2007; Pracheil et al. 2013).
Tributaries provide essential habitat for reproduction and early life history stages of many riverine fishes. In the Colorado River basin, impoundments have drastically altered the main-stem river environment, thereby increasing the importance of tributaries for rare or endangered fishes. For example, most Humpback Chub Gila cypha in the lower Colorado River basin presently migrate into the tributary Little Colorado River because it has a natural thermal regime, flow pattern, and sediment load (Gorman and Stone 1999; Coggins et al. 2006; Sabo et al. 2010; Van Haverbeke et al. 2013). Tributary spawning presumably occurs because similar habitat is rare or unavailable in the main-stem Colorado River because of flow and temperature alterations by the upstream Glen Canyon Dam (Gorman and Stone 1999; Coggins et al. 2006). Colorado Pikeminnow Ptychocheilus lucius and Razorback Sucker Xyrauchen texanus, both also native and endangered in the Colorado River basin, were once abundant in the main-stem Colorado River and larger tributaries (Bestgen et al. 2007; Zelasko et al. 2010; Bestgen 2015; Marsh et al. 2015). Habitat degradation in main-stem rivers extirpated these species from much of the lower Colorado River basin, and now the largest populations are found in upstream sections of the Colorado River and tributary habitat such as the Green and White rivers (Bestgen 1990; Modde et al. 1996; Osmundson and Burnham 1998; Bestgen et al. 2007). Colorado Pikeminnow and Razorback Sucker make long seasonal migrations to spawning locations, and such movements may be motivated by both flow and water temperature cues (Tyus 1990; Tyus and Karp 1990; Irving and Modde 2000; Zelasko et al. 2010; Bestgen et al. 2011). Recent research documented the importance of smaller tributaries for reproduction by those species (Bottcher et al. 2013; Fresques et al. 2013; Webber et al. 2013; Cathcart et al. 2015).
Less is known about movements and tributary use by other increasingly rare large-bodied native fishes of the Colorado River basin such as the Flannelmouth Sucker Catostomus latipinnis and Bluehead Sucker C. discobolus (Bezzerides and Bestgen 2002). Flannelmouth and Bluehead suckers have experienced large declines in distribution and abundance in the last century and each now occupies less than 50% of historical habitat (Bezzerides and Bestgen 2002; Budy et al. 2015). Declines in main-stem rivers are partially due to habitat and abiotic alterations caused by dams and nonnative fish introductions (Martinez et al. 1994; Propst et al. 2008; Sweet and Hubert 2010; Dauwalter et al. 2011; Budy et al. 2015). Bezzerides and Bestgen (2002) suggested that extensive tributary use by Flannelmouth and Bluehead suckers may be one reason these species did not experience more dramatic declines, such as those seen in the four federally listed (as endangered), large-bodied fishes that were once widespread in main-stem Colorado River basin habitat. Thus, understanding movement patterns and tributary use by Flannelmouth and Bluehead suckers will be critical for formulating practical conservation strategies that include protection of tributaries, particularly when associated main-stem reaches of larger rivers are regulated by dams (Moyle and Mount 2007; Pracheil et al. 2009, 2013).
Flannelmouth and Bluehead suckers have been classified as both migratory and sedentary because some individuals move distances greater than 200 km while others have restricted movements (Chart and Bergersen 1992; Douglas and Marsh 1998; Beyers et al. 2002; Bezzerides and Bestgen 2002; Compton et al. 2008; Sweet and Hubert 2010). Spawning movements have been noted for Flannelmouth Suckers in both lower and upper Colorado River basin tributaries, but low recapture rates have limited inferences regarding the importance of tributaries and the role of flow and water temperatures to cue movements (Chart and Bergersen 1992; Weiss et al. 1998; Compton et al. 2008; Cathcart et al. 2015). Use of stationary PIT tag antenna arrays may detect tags continuously throughout the year and increase recaptures, allowing better inferences regarding movement patterns, especially those related to environmental variables (Greenberg and Giller 2000; Connolly et al. 2008; Johnston et al. 2009). Thus, our study had two main objectives. First, we evaluated tributary use by native Flannelmouth and Bluehead suckers and spawning site fidelity in the upper White River, Colorado, using mark–recapture methods and stationary PIT tag antenna arrays. Second, we evaluated timing of sucker spawning movements into these tributaries and the role of water temperature and flow in cueing these movements. We also report posttagging movements of suckers that may affect capture–recapture tagging studies that handle fish. Our study adds to the limited ecological knowledge about tributary use and movements for reproduction of Flannelmouth and Bluehead suckers in the upper Colorado River basin and will allow better-informed decisions to improve the status of these declining species.
METHODS
Study area
The headwaters of the White River are in the mountainous and high elevation Flat Tops Wilderness in northwestern Colorado. The White River flows west through Colorado and Utah, draining approximately 13,000 km2 until it joins the Green River near Ouray, Utah (Lanigan and Berry 1981; Chart and Bergersen 1992; Martinez et al. 1994). Taylor Draw Dam, the only main-stem dam on the White River, impounded Kenney Reservoir beginning in October 1984 (Figure 1). Taylor Draw Dam is a barrier to fish migrating upstream. The White River upstream from Kenney Reservoir to the confluence of the North Fork and South Fork of the White River, about 150 river kilometers (rkm), was chosen as the study site because it is one of the last, relatively large, unobstructed, free-flowing river sections in the upper Colorado River basin. It also supports relatively robust native fish populations including Flannelmouth and Bluehead suckers (Chart and Bergersen 1992). Snowmelt is the dominant water source in the White River, and runoff usually occurs from mid-April through late June; flow typically peaks in early June with an average maximum flow of 91.5 m3/s (range, 23.7–186.5 m3/s; Figure 2). Base flows occur from July through March. Thunderstorms periodically increase flow and turbidity from June through August.
The upper White River drainage, Colorado, including the location (arrow) of U.S. Geological Survey stream gauge 09304800 near Meeker, and an expanded map (lower) of Coal Creek depicting the placement of the two remote PIT tag antenna arrays and their proximity to the main-stem White River, 2012–2013.
Mean daily flow (m3/s) of the White River, Colorado, measured at U.S. Geological Survey gauge 09304800 for the period of record (1962–2013) and study years 2012 and 2013.
Flows in our main study tributary, Coal Creek, were not measured by a gauge and we made only occasional measurements of streamflow. We did not make more regular flow measurements in Coal Creek because we felt fish resident to the White River and then moving into Coal Creek were responding largely to flows in the former stream. In general, flows rose and fell during spring runoff periods in Coal Creek similar to that in the White River, because each was fed by snowmelt. Flows in Coal Creek were low (mean channel width, about 5 m) during base flow and based on observations were typically less than 10% of those in the White River, which had a much larger channel (width, 30–40 m). Base flows in Coal Creek are discussed in a following section.
The White River study area represents a continuum from an upstream coldwater environment to a downstream warmwater one. Other studies used the confluence of the White River with the Green River as rkm 0; this location is 177 rkm downstream from the Kenney Reservoir–White River inflow area (Chart and Bergersen 1992; Martinez et al. 1994). In our study, rkm 0 was at the Kenney Reservoir–White River inflow area. The upstream coldwater portion of the study area begins at the confluence of the North Fork and South Fork of the White River and continues downstream through the town of Meeker, Colorado (rkm 151–100). This section is typical of other coldwater environments where salmonids are the dominant fishes (Rahel and Hubert 1991; Griffith 1993). The channel form is pool–riffle–run with predominantly cobble substrate. Downstream from rkm 100 near Meeker to rkm 68, the river transitions to cool water and supports both warmwater and coldwater species. The downstream portion of the study area (0–68 rkm) is a warmwater environment where salmonids are few and warmwater species such as Flannelmouth and Bluehead suckers dominate the fish community. The channel form in the downstream portion is a uniform run with substrate composed of predominantly sand in low velocity areas and cobble in higher velocity locations. The main-stem White River is 30–40 m wide throughout the study area.
In April 2011, we deployed five HOBO temperature loggers in the main-stem White River and an additional 15 in tributaries. Temperature loggers were tethered to the shore and weighted down in moving water so they remained on the river bottom unaffected by direct sunlight. Temperature was recorded once per hour, and data were downloaded in October of each year, 2011–2013. These 20 temperature loggers were deployed to evaluate potential thermal habitat for suckers throughout the basin. However, temperature data are reported for only those sites used by each sucker species. Flow data were gathered from U.S. Geological Survey (USGS) gauge 09304800 on the main-stem White River.
Pilot study (site selection)
Surveys in April and May of 2011 revealed seven potential tributaries for native sucker spawning: Crooked Wash, Piceance Creek, Flag Creek, Curtis Creek, Coal Creek, Miller Creek, and North Elk Creek. Based on presence of mostly perennial flows and presence of Flannelmouth or Bluehead suckers, Piceance Creek and Coal Creek were selected for this study. All other streams surveyed were not suitable due to barriers, ephemeral reaches, or naturally cold water temperatures.
The confluence of Piceance Creek and the White River is at an elevation of 1,740 m above sea level (ASL) and 68 rkm upstream from the Kenney Reservoir–White River inflow area. Piceance Creek flows through a landscape dominated by sagebrush Artemisia spp., pinyon pine Pinus edulis, and juniper Juniperus osteosperma. The channel is approximately 2.5–5 m wide and the substrate is predominantly sand. The confluence of Coal Creek and the White River is located at 1,935 m ASL and rkm 118. Coal Creek flows through an agricultural valley dominated by irrigated fields. The channel is approximately 5 m wide and the substrate is a mix of cobble and sand.
Irrigation practices alter flows in both Piceance and Coal creeks but in different ways. During base flow periods, temporary dams (tarps) in Piceance Creek divert water into irrigation canals and reduce or eliminate flow in the stream channel. In contrast, flow in Coal Creek is enhanced beginning in spring by an upstream irrigation ditch that diverts water from the main-stem White River to the Coal Creek drainage. Thus, flow in Coal Creek is higher than historical levels due to irrigation returns from diversions of White River flow.
Fish capture by fyke net
Fyke nets were used to capture Flannelmouth and Bluehead suckers in Coal Creek in spring 2011, 2012, and 2013. Sampling in 2011 was exploratory but indicated that fyke nets could be used to capture relatively large numbers of native suckers. We deployed four double-winged nets (mesh diameter, 4 cm) that opened facing downstream to capture fish migrating upstream in spring and early summer, presumably to spawn. Net wings, 5 m long and 2 m high, were carefully anchored in the stream substrate and spanned the entire stream channel. We placed the cod end of nets in slow-moving water to minimize stress on captured fish. Preliminary sampling in the high flow and late runoff year 2011 occurred periodically for 2–3-d periods from late June through July. In 2012 and 2013 fyke nets remained in Coal Creek continuously from early May through mid-June. Nets were removed in summer when catch rates fell to less than one fish per day for a full week. Fish were removed from nets daily. Fyke nets were placed in Piceance Creek during May of 2012, but no fish were captured so sampling at that site was terminated.
Species and PIT tag data
Fish captured in fyke nets (see below) were identified, measured (TL, mm), and weighed (g), and sex was determined by the presence of gametes (eggs or milt) in mature adults. Each fish was scanned prior to PIT-tagging to assess whether it was previously tagged. We implanted a 32-mm, half-duplex, Oregon RFID PIT tag in all suckers > 230 mm TL, and all PIT tags were scanned prior to insertion to ensure functionality. We choose 32-mm PIT tags to maximize detection range for fish in the deeper main-stem White River; this choice is discussed elsewhere (Fraser 2015). The PIT tags (3.7 mm in diameter) were inserted into the body cavity behind the pelvic fin using a tag injector with a sterilized needle.
Remote PIT tag antenna arrays
Prior to fish sampling, we installed two remote PIT tag antenna arrays in Coal Creek during the first week of April in 2012. The downstream antenna array was placed 200 m upstream from the confluence with the White River (downstream antenna array), and the other was 1,800 m upstream from the confluence (upstream antenna array). Individual antennas consisted of a 50-cm-wide loop formed by a 10-gauge wire that encompassed the entire width of Coal Creek. Each antenna array consisted of two antennas separated by 1–2 m (Figure 1) so that we could determine the direction of fish movement. Antenna arrays were anchored to the streambed in riffles that were 20–25 cm deep. Preliminary testing with floating and submerged PIT tags (surrogate fish) indicated antennas had a read range of 45–60 cm; therefore, tags were likely detected in the entire water column (antenna efficiency is discussed in Results). The arrays were powered with two, deep-cycle, 12-V batteries that were changed weekly when data were downloaded.
We condensed tag detection data captured by PIT tag arrays by using only the first and last detection of a fish each day at each array to reflect daily movement patterns. The antenna data clarified behavior patterns of suckers posthandling and enabled us to determine antenna efficiency because both the number of fish tagged and the antenna detections were known and compared. As tagged fish returned to Coal Creek in subsequent years, antenna detections were used to determine when fish returned, how long they remained in Coal Creek (excluding fish that were physically handled), and site fidelity over the study duration.
RESULTS
Tributary Use
Flannelmouth and Bluehead suckers entered Coal Creek in mid-May of both 2012 and 2013 when water temperatures were rising and flows were high or declining (Figure 3). Movement patterns were not documented during preliminary sampling in 2011 because we did not sample regularly. Fyke nets were deployed for three 2–3-d periods during July and 1 d in August 2011. However, fish tagged in 2011 that returned to Coal Creek in 2012 and 2013 were used to describe movement patterns and site fidelity in those years. Flannelmouth Suckers first entered Coal Creek on May 14 in 2012 when temperatures in the White River reached 11.0°C and Coal Creek temperatures reached 12.5°C (Figure 4). The majority of Flannelmouth Suckers had entered Coal Creek from the White River in 2012 when the 5-d mean river water temperature reached 10.7°C (Table 1). In 2013, Flannelmouth Suckers were first captured in Coal Creek on May 18 when the water temperature reached 8.2°C in the White River and 11.7°C in Coal Creek. The majority of suckers had moved from the White River to Coal Creek when the mean water temperature of the river reached 10.9°C. The average flow in the main-stem White River during the peak capture period for Flannelmouth Suckers in 2012 was 17.0 m3/s (range, 11.5–22.2 m3/s) when flows were declining. The average flow in the main-stem White River during the peak capture period for Flannelmouth Suckers in 2013 was 39.5 m3/s (range, 26.9–50.7 m3/s) during three flow peaks.
| Year | Flannelmouth Sucker | Bluehead Sucker |
|---|---|---|
| 2012 | 10.7 (10.3–13.1) | 13.8 (12.2–14.0) |
| 2013 | 10.9 (8.4–12.1) | 11.6 (9.1–12.3) |
Timing and number of Flannelmouth (FMS) and Bluehead (BHS) suckers captured in fyke nets (indicated by bars) in Coal Creek and mean daily flow and water temperature for the main-stem White River in 2012 (upper panel) and 2013 (lower panel). Flow and water temperature data for the White River near Coal Creek were from U.S. Geological Survey gauge 09304200). The arrows represent the day that the fyke nets were placed in and removed from Coal Creek; dates on the x-axes are month/day.
Mean daily water temperature of Coal Creek and the main-stem White River near Coal Creek (U.S. Geological Survey gauge 09304200) during the native sucker spawning period in 2012 (upper panel) and 2013 (lower panel). Dates on the x-axes are month/day.
Bluehead Sucker movement into Coal Creek from the White River largely overlapped that for Flannelmouth Suckers in each of 2012 and 2013. The first Bluehead Suckers entered Coal Creek on May 15 in 2012 when water temperatures in the White River adjacent to Coal Creek reached 10.3°C (Figure 3) and Coal Creek temperatures reached 13.3°C. The majority of Bluehead Suckers had moved from the White River to Coal Creek in 2012 when the 5-d mean water temperature of the river reached 13.8°C (Table 1). Bluehead Suckers first entered Coal Creek on May 25 in 2013 when the water temperature was 9.5°C in the White River and 13.5°C in Coal Creek. In 2013, the majority of Bluehead Suckers had moved from the White River to Coal Creek when the mean water temperature of the river reached 11.6°C. The flows during the peak capture period for Bluehead Suckers during both 2012 and 2013 were similar to those for Flannelmouth Suckers.
Suckers captured in Coal Creek with fyke nets during 2011, 2012, and 2013 were entirely large-bodied adults (Figure 5). Flannelmouth Suckers were larger and outnumbered Bluehead Suckers by a ratio of 6:1 in 2012 and 7:1 in 2013. Mean TL of Flannelmouth Suckers was 472 mm (range, 385–570 mm; N = 75) in 2011, 455 mm (range, 381–554 mm; N = 224) in 2012, and 470 mm (range, 413–576 mm; N = 157) in 2013. Bluehead Sucker mean TL was 414 mm (range, 357–469 mm; N = 35) in 2011, 390 mm (range, 340–460 mm; N = 40) in 2012, and 412 mm (range, 365–480 mm; N = 23) in 2013.
Length frequency distributions for Flannelmouth (FMS) and Bluehead (BHS) suckers captured in fyke nets in Coal Creek during May and June in 2012 (upper panel) and 2013 (lower panel).
Gamete expression for suckers captured in Coal Creek indicated more males than females were captured for each species in both years. In 2012, 96 of the 224 Flannelmouth Suckers captured expressed milt and 35 expressed eggs. In 2013, 91 of the 157 Flannelmouth Suckers captured expressed milt and 10 expressed eggs. In 2012, 21 of the 40 Bluehead Suckers captured expressed milt and three expressed eggs. In 2013, 12 of 23 Bluehead Suckers captured expressed milt and two expressed eggs. The remainder of the suckers captured in Coal Creek did not express gametes, so gender was not identified. The relatively low number of fish with positive determinations made sex-specific comparisons of length difficult.
The PIT tag antenna detections of recaptured suckers indicated similar trends in tributary use compared with those for fyke-net captures. In 2012, Coal Creek PIT tag antennas first detected movement of Flannelmouth Suckers tagged in 2011 on May 16, 2 d after the first fyke-net capture of untagged fish. Bluehead Suckers were first detected on May 13, 1 d earlier than the first fyke-net capture in 2012. In 2013, antennas in Coal Creek first detected Flannelmouth Suckers on May 16, 2 d earlier than the first fyke-net capture. Bluehead Suckers were first detected on May 17, 1 d earlier than the first fyke-net capture. Antenna detections from suckers returning to Coal Creek in years after they were initially tagged indicated mean residence times of 10–12 d in 2012 and 2013 for fish that did not experience physical handling (range, 1–32 d; Table 2).
| Year | Bluehead Sucker | Flannelmouth Sucker |
|---|---|---|
| 2012 | 11.2 (1–27), N = 6 | 10.7 (1–16), N = 10 |
| 2013 | 11.8 (1–30), N = 20 | 10.5 (1–32), N = 74 |
Antenna Efficiency
Each antenna array in Coal Creek had two separate antennas that resulted in high detection rates of suckers tagged and released in Coal Creek. All fish in 2012 were accounted for with detections by at least one antenna array (264 of 264) and in 2013 99.5% were detected (179 of 180). By pairing antennas at each array we were able to determine the direction of movement. Of the 444 fish tagged and released during 2012 and 2013, 436 were detected by the most downstream antenna array as fish were presumably exiting Coal Creek (see Behavior section below; Figure 1). Seven fish were detected moving downstream past the upstream antenna array but were not detected at the downstream antenna array, and one fish was not detected at either antenna array. Considering the read range of the antennas (45–60 cm) and the depth of the water (20–25 cm), it is unlikely that these fish exited Coal Creek while the antenna arrays were operational. Therefore, it is likely that these eight fish either remained or died in Coal Creek between the downstream and upstream antenna arrays.
Site Fidelity
Flannelmouth and Bluehead suckers exhibited moderate levels of spawning site fidelity to Coal Creek from 1 year to the next, ranging from 13% to 45% (Table 3). Recaptures of tagged fish indicating site fidelity were from both fyke nets and antenna arrays, but in every instance antenna arrays detected a much higher percentage of fish returning. For example, fyke-net captures showed Flannelmouth Suckers had site fidelity rates of 2–3% and those of Bluehead Suckers were only slightly higher (Table 3). In comparison, antennas demonstrated site fidelity for Flannelmouth Suckers to Coal Creek of 45% and 22% for suckers initially tagged in 2011 and 2012, respectively. Fidelity rates from antenna detections for Bluehead Suckers were 40% and 13% for fish initially tagged in 2011 and 2012, respectively.
| Antenna detection | Fyke-net recaptures | ||||
|---|---|---|---|---|---|
| Tagging year | Fish tagged | 2012 | 2013 | 2012 | 2013 |
| Flannelmouth Sucker | |||||
| 2011 | 40 | 16 (40) | 18 (45) | 2 (5) | 2 (5) |
| 2012 | 224 | 50 (22) | 3 (1) | ||
| Bluehead Sucker | |||||
| 2011 | 35 | 9 (26) | 14 (40) | 6 (17) | 5 (14) |
| 2012 | 40 | 5 (13) | 3 (8) | ||
Fyke-Net Efficiency
Preliminary sampling in 2011 with fyke nets indicated sucker captures occurred with minimal crew and minimized stress on captured fish. However, comparing fyke net captures to antenna detections indicated that fyke nets were not as efficient at sampling suckers as were the antennas. For example, during the entire study, 84 tagged Flannelmouth Suckers were detected by the antennas as they returned to Coal Creek, but only 7 of the 84 (8%) were recaptured in fyke nets. Fyke-net efficiency was better for Bluehead Suckers but still low; of the 28 tagged Bluehead Suckers detected by the antennas, only 14 (50%) were recaptured with fyke nets.
Posthandling Behavior
Antenna detections revealed that 70–88% of native suckers captured and handled in Coal Creek exited within 24 h (Table 4) and 83–96% left within 48 h. Thus, few fish were available for recapture in subsequent sampling within a year. The unexpected emigration occurred whether suckers were captured, handled, and tagged for the first time or had been previously tagged and were simply handled and scanned for the presence of a tag. Suckers recaptured in fyke nets in subsequent years exhibited the same behavior as those after the initial capture, because 15 of the 21 suckers recaptured exited Coal Creek within 24 h of being handled. In comparison, tagged suckers detected in years after the first tagging but not handled had residence times of 10–12 d (range, 1–32 d).
| Year | Captured | Exited Coal Creek | |
| Day 1 | Day 2 | ||
| Flannelmouth Sucker | |||
| 2012 | 224 | 189 (84) | 11 (5) |
| 2013 | 157 | 128 (82) | 9 (6) |
| Bluehead Sucker | |||
| 2012 | 40 | 35 (88) | 3 (8) |
| 2013 | 23 | 16 (70) | 3 (13) |
DISCUSSION
We have shown tributaries are important habitats for native suckers in the upper Colorado River basin and report three main findings. First, both Flannelmouth and Bluehead suckers are likely responding to a thermal cue that initiates migration, as suckers moved into Coal Creek under similar and increasing water temperatures in 2012 and 2013, while flows varied between years. Second, suckers showed moderately high spawning site fidelity, a conclusion that was facilitated by the use of antenna arrays. Sampling only with fyke nets would have resulted in fewer suckers captured and the conclusion that site fidelity by native suckers was low. Third, most suckers of both species emigrated from Coal Creek within 48 h after being captured. This flight response was not anticipated and has not to our knowledge been reported for these species. Below we discuss the nuances and implications of these findings, especially for conservation of Flannelmouth and Bluehead suckers.
Tributary Use
We found movement of Flannelmouth and Bluehead suckers into tributaries was cued by water temperature, verifying the findings of other investigators (Bezzerides and Bestgen 2002; Zelasko et al. 2011; Fraser 2015). The majority of Flannelmouth and Bluehead suckers moved from the White River into Coal Creek when river water temperatures, presumably a main driver of movements, were 11–14°C, and Coal Creek temperatures were 2.5–4°C warmer. The wide range of temperatures during movements and the short time series of 2 years did not allow for statistical analysis of more specific thermal cues that might initiate spawning movements. The location of suckers and the specific thermal regime they experienced in the White River prior to immigration was also unknown. Timing of arrival to Coal Creek may also depend on the distance traveled to Coal Creek. None of these hypotheses can be addressed with our data. Distributions of hatching dates of sucker larvae sampled over the longitudinal gradient of the White River and in Coal Creek were closely correlated with water temperature thresholds (Fraser 2015) and supported the notion that sucker reproduction is cued mainly by water temperature. The presence of only adult suckers and their increased reproductive condition, in addition to subsequent collection of larvae of both species in Coal Creek, indicated the suckers moved there to spawn (Fraser 2015). Tributaries are likely attractive sucker habitat as they are often warmer and more productive and harbor fewer predators than the main-stem habitat (Weiss et al. 1998; Douglas and Douglas 2000; Paukert and Rogers 2004; Pool et al. 2013; Datry et al. 2014; Cathcart et al. 2015). Coal Creek water temperatures that were, on average, 2.5–4°C warmer than main-stem temperatures may promote earlier spawning (Fraser 2015) and subsequently increased growth and development of larvae, which may increase their survival (Kaeding and Osmundson 1988; Thompson et al. 1991; Bestgen et al. 2006; Coleman and Fausch 2007).
Movement into tributary habitat may also be initiated by flow, but because levels and patterns were different between years, specific mechanisms to induce movements were difficult to identify (Weiss et al. 1998; Bottcher et al. 2013; Cathcart et al. 2015). Flow rates in 2013 were more than double those in 2012 and exhibited multiple peaks when suckers entered Coal Creek, compared with lower flows in 2012 that were declining from the peak. Thus, reproduction during inconsistent flow patterns and magnitudes indicated flows may be only a secondary cue for sucker movement. Although some studies have noted that native fishes begin spawning movements as seasonal flow declines (Tyus 1990; Tyus and Karp 1990; Irving and Modde 2000; Osmundson 2011), additional longer-term data are needed to better understand the interplay of water temperatures, flow, and photoperiod on spawning of native suckers in the Colorado River basin.
Consistent with other studies, we showed Flannelmouth and Bluehead suckers in the upper White River used tributaries for spawning, based on the high proportion of ripe adults moving into Coal Creek and subsequent presence of larvae (Fraser 2015). For example, Flannelmouth Suckers were known to migrate into unregulated tributaries of the highly regulated main-stem lower Colorado River in Arizona (Weiss et al. 1998) as well as into McKinney Creek, a small spawning tributary to Muddy Creek in Wyoming (Compton et al. 2008). In the San Juan River basin ripe Flannelmouth Suckers made migrations into McElmo Creek (Cathcart et al. 2015). Our study is the first to document relatively high annual return rates, up to 45%, of fish to the same spawning tributary.
The upper White River has only two perennial tributaries that have the appropriate thermal requirements and access for reproduction by Flannelmouth and Bluehead suckers. In Piceance Creek where no sucker movement was observed, flows were greatly reduced due to water diversion for irrigation. In contrast, Coal Creek possesses a more natural snowmelt hydrograph, is less affected by upstream irrigation diversions, and collects irrigation water that returns to the White River resulting in increased flow throughout the adult spawning season and for subsequent early life stages for both sucker species. Increased flow results in higher connectivity with the White River, especially in low water years, which increases the utility of Coal Creek as a spawning area. Similarly, Cathcart et al. (2015) showed that McElmo Creek, a San Juan River tributary with supplemented flow, also attracted reproducing Flannelmouth Suckers in spring. Coal Creek provides an example of how flow maintenance or supplementation may provide a substantial benefit for fish conservation, especially if there are no barriers to upstream fish movement.
Site Fidelity
In addition to using tributaries for reproduction, Flannelmouth and Bluehead suckers first tagged in Coal Creek demonstrated annual site fidelity of up to 40–45%, as measured by antenna arrays. Other cypriniform fishes in the upper Colorado River basin move annually in spring to specific sites, presumably to spawn, but rates of spawning site fidelity are not known (Tyus 1990; Tyus and Karp 1990). Spawning by native suckers certainly occurred in the main-stem White River, based on the broad distribution of early life stages, especially in downstream reaches (Fraser 2015). We do not know the proportion of fish using Coal Creek versus main-stem habitat; however, in general, use of tributaries for spawning may alleviate competition for main-stem spawning sites, especially when tributaries are known productive sites for reproduction. A better understanding of the proportion of suckers from the White River basin that use tributaries such as Coal Creek to spawn would add perspective to tributary use information.
Compared with antenna detections of PIT tags, fyke-net recaptures indicated a lower fidelity of 5–17% of native suckers to Coal Creek. In other words, of the 112 suckers detected as returning to Coal Creek by antenna arrays a year or more after being tagged, only 21 were also captured in fyke nets. Most suckers detected by antennas bypassed fyke nets even though the entire channel was blocked at 3–4 locations. It is unclear how suckers bypassed multiple fyke nets because we carefully weighted net wings to the stream substrate and checked net integrity daily. Although fyke nets may induce less stress on fish and have lower mortality rates than other sampling gear (Hopkins and Cech 1992), antenna detections indicated fyke nets were inefficient in capturing suckers even in the relatively small Coal Creek, and their exclusive use would have resulted in flawed estimates of site fidelity.
Remote PIT tag antenna arrays placed in the main-stem White River would further enhance understanding of seasonal movement, home range, and movement patterns of Flannelmouth and Bluehead suckers. Main-stem antennas would be especially useful if arrays were operated year round, including when traditional sampling is not possible due to extreme flow conditions or presence of ice. Additionally, PIT tag antenna arrays can increase detections and recapture data useful to estimate demographic rates of riverine fish (Webber and Beers 2014; Cathcart et al. 2015). For example, a PIT tag antenna array placed in the Green River detected 569 endangered Razorback Suckers during 2012 and 2013, 93% of which had not been detected since their release from the hatchery up to 13 years previously, despite intensive annual sampling with electrofishing equipment (Webber and Beers 2014).
Posthandling Behavior
Antenna detections showed that the majority of Flannelmouth and Bluehead suckers moved out of Coal Creek within 24 h of being captured and handled, and none returned the same year. In contrast, suckers that were detected but not handled remained for an average of 10–12 d. Antenna detection data indicated that handling alone was sufficient to cause emigration of suckers from Coal Creek, because fish recaptured in fyke nets that were tagged in previous years exhibited the same behavior. Rapid emigration was not expected and would not have been apparent without PIT tag antenna arrays.
Rapid emigration had at least three consequences. First, movement of fish out of Coal Creek after handling explained low recapture rates of suckers in the year they were initially tagged and made estimates of abundance or other vital rates difficult. Second, high emigration rates limited inferences on natural behavior patterns of suckers occupying Coal Creek. For example, most suckers were captured before reaching the second antenna, and that may have limited how far upstream they could have traveled or how long they would have stayed in Coal Creek. Finally, emigration of captured suckers may have lowered site fidelity rates by altering the behavior of other individuals entering Coal Creek to spawn, causing them to abandon spawning. Our study cannot address whether suckers captured in Coal Creek subsequently spawned in the main-stem White River in the year they were captured. However, abandoning spawning could be detrimental to declining fish populations, and those effects should be considered when sampling during spawning movements. Continued operation of remote antennas in Coal Creek in subsequent years without deploying fyke nets is recommended to determine whether fyke nets affected site fidelity rates.
Our movement behavior results were similar to those for American Shad Alosa sapidissima, which showed that more than 50% of fish abandoned their upstream spawning movements less than 24 h after being captured, handled, and tagged (Bell 1985; Barry 1986; Moser 2000). Salmonids have also shown unexpected posthandling movement downstream, up to 100%, when fish were captured during an upstream spawning movement (Bernard et al. 1999; Makinen et al. 2000; Young et al. 2006). However, in Roubideau and Cottonwood creeks, small tributaries in the Gunnison River basin, Colorado, Flannelmouth and Bluehead suckers were trapped and tagged at an instream weir, and no posthandling flight response was noted (K. G. Thompson, unpublished data). In Roubideau and Cottonwood creeks any flight response would have only been detected by trapping tagged fish in the weir on their outward migration, as no antenna arrays were used to detect movement posthandling. Differences in spawning habitat availability, timing of spawning, and detection probability, or distance from stream mouth (Coal Creek and Gunnison River tributaries are similar in size) are among the potential reasons for the differences in posthandling responses between our studies, but they highlight the need for further understanding this response. Posthandling movement behavior needs to be considered when evaluating movement patterns or tag-recapture rates, especially when sampling-reach lengths are short, because fish movements out of the study area could bias abundance estimates based on the assumption of no emigration.
Posthandling emigration of suckers from Coal Creek may also give insight into the imprecise population abundance estimates we obtained from White River mark–recapture sampling (Fraser 2015). If marked fish left our relatively short White River sampling sections after being handled and were unavailable for recapture, abundance estimates would likely be biased high because of reduced recapture probabilities. Low recapture rates of suckers have been observed in previous studies and have limited the assessment of both long-term and short-term movement patterns (Chart and Bergersen 1992; Douglas and Douglas 2000; Webber and Beers 2014); these results could also be due to handling-related movements. The use of remote PIT tag antenna arrays in the main-stem White River upstream and downstream from the sampling locations to evaluate the movement patterns of suckers posthandling could be used to assess the importance of this issue.
Our research suggests that tributaries are important spawning habitats for native Flannelmouth and Bluehead suckers and that some proportion of the population returns each year. Thus, maintaining connectivity from the main-stem river to tributary habitats via flow maintenance or barrier removal is critical and should be a conservation focus for Flannelmouth and Bluehead suckers (Douglas and Douglas 2000; Bottcher et al. 2013; Pool et al. 2013; Cathcart et al. 2015). Similar to Cathcart et al. (2015), we demonstrate that flow supplementation in Coal Creek likely increased the utility of that habitat for native suckers, especially when compared with a flow-impaired tributary such as Piceance Creek. Our study also increased understanding of the movement and reproductive ecology of Flannelmouth and Bluehead suckers, information that should be useful to inform habitat protection and management strategies and improve the status of these declining species.
ACKNOWLEDGMENTS
Funding for this project was provided by the U.S. Geological Survey and Colorado Parks and Wildlife. We thank the Strang and Nelson families of Meeker, Colorado, for granting us access to Coal Creek. We thank E. Kluender, D. E. Snyder, S. Seal, N. Shannon, B. Avila, C. Bryant, E. Pettigrew, and J. Smith and others we may have forgotten for contributing to field work, larval fish identification, and other technical assistance. P. Budy and three anonymous reviewers provided valuable insight to improve the paper. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. This study was performed under the auspices of Colorado State University Institutional Animal Care and Use Committee protocol number 12-3365A.
