North American Journal of Fisheries Management 24:648±653, 2004 ᭧ Copyright by the American Fisheries Society 2004

Factors Affecting Condition of Flannelmouth Suckers in the , Grand Canyon,

CRAIG PAUKERT*1 U.S. Geological Survey, Southwest Biological Science Center, Grand Canyon Monitoring and Research Center, 2255 North Gemini Drive, Flagstaff, Arizona 86001, USA

R. SCOTT ROGERS Arizona Game and Fish Department, 2221 West Greenway Road, Phoenix, Arizona 85023, USA

Abstract.ÐThe impoundment of the Colorado River by Glen Canyon Dam, Arizona, in 1963 created a highly regulated environment in the Grand Canyon that altered the native ®sh populations, including the ¯annelmouth sucker latipinnis. Flannelmouth suckers were sampled from 1991 to 2001 to determine seasonal, annual, and spatial trends in ®sh condition (i.e., relative

weight [Wr]). Mean Wr peaked during the prespawn and spawning periods and was lowest in summer and fall, but it was never lower than 93. Condition was variable throughout the Grand Canyon but was typically greatest at intermediate distances from Glen Canyon Dam, possibly because of the increased number of warmwater tributaries in this reach. Flannelmouth sucker condition in September was positively correlated with Glen Canyon Dam discharge during summer (June±August); this result may be due to the larger euphotic zone and greater macroinvertebrate abundance observed during higher water ¯ows. Increased dam discharge that stimulates river productivity may provide bene®ts for this native ®sh.

The Colorado River in the Grand Canyon his- McKinney et al. 1999), and predation (Ward et al. torically had eight endemic ®shes (Minckley 2002). Therefore, continued monitoring of ¯an- 1991), of which three are now extirpated (Colo- nelmouth sucker populations is needed to assess rado pikeminnow Ptychocheilus lucius, bonytail changes in the status of this native ®sh. Gila elegans, and roundtail chub Gila robusta) and Fish body condition is a common ®sheries as- two are federally endangered (humpback chub sessment tool. Measures of ®sh condition have Gila cypha and razorback sucker Xyrauchen tex- been linked to general ®sh health (Coughlan et al. anus). The ¯annelmouth sucker Catostomus lati- 1996), fat and lipid content (e.g., Neumann and pinnis, also endemic to the Colorado River, is rel- Murphy 1992), prey availability (e.g., Paukert and atively common in the Grand Canyon compared Willis 2003), reproductive potential (e.g., Neu- to other native ®shes (Minckley 1991). mann and Murphy 1992), and environmental con- Although previous research has suggested that ditions (e.g., productivity [DiCenzio et al. 1995] the Grand Canyon population of ¯annelmouth and water level ¯uctuations [Johnson et al. 1992]). suckers is relatively stable (Douglas and Marsh In general, higher condition is associated with 1998), there is concern that recruitment may be higher energy (fat) content, increased food base or declining because of the effects of lower water reproductive potential, or more favorable environ- temperature caused by hypolimnetic releases from mental conditions. Glen Canyon Dam (Weiss et al. 1998; Clarkson Our objectives were to evaluate long-term (Ͼ10 and Childs 2000; Ward et al. 2002), loss of suitable years) temporal and spatial trends in ¯annelmouth rearing habitat (Thieme et al. 2001), blockage of sucker condition and to evaluate the biotic and spawning migrations (Chart and Bergersen 1992; abiotic relationships that might in¯uence condi- tion. The goal of the study was to assess condition on a large scale (i.e., canyonwide) to aid in the * Corresponding author: [email protected] evaluation of trends in the ¯annelmouth sucker 1 Present address: U.S. Geological Survey, Biological population in the Colorado River system. Resources Division, Kansas Cooperative Fish and Wild- life Research Unit, Division of Biology, 205 Leasure Hall, Kansas State University, Manhattan, Kansas Study Area 66506, USA. The study area encompassed the Colorado River Received May 9, 2003; accepted September 12, 2003 from river kilometer (RKM) 26 (the

648 FLANNELMOUTH SUCKER CONDITION 649 in¯ow) below Glen Canyon Dam (near the Ari- year, month, and river location, and therefore the zona±Utah border and considered RKM 0.0) analysis was limited to ®sh between 200 and 599 downstream to at RKM 363.2. The mm TL. This length range was the one most com- study area is strongly in¯uenced by Glen Canyon monly collected during the study. Dam, which has reduced ¯ood frequency, seasonal Because condition changes with ®sh length, we ¯ood variation, ¯ood magnitude, and sediment categorized ¯annelmouth suckers into 100-mm transport in the river (Stevens et al. 1997). In ad- length-groups (i.e., 200±299 mm TL, 300±399 mm dition, hypolimnetic releases from Glen Canyon TL, etc.) for purposes of analysis. The Colorado Dam have stabilized water temperature at about River in the Grand Canyon was also categorized 10ЊC (with little warming throughout the 363- into 16-km river reaches (based on the upstream- RKM-long study area), whereas pre-dam temper- most RKM) to ensure that several ®sh were col- ature conditions varied from near 0ЊC to 29.4ЊC lected in most reaches. Three separate three-way (Stevens et al. 1997). analyses of variance (ANOVAs) with interactions Several major warmwater tributaries of the Col- were conducted: (1) to determine whether mean orado River in the Glen and Grand canyons are Wr differed among length-groups, river reaches, important to native ®shes, including the Little Col- and months during 1993, (2) to determine whether orado River (RKM 98.7) and the Paria River mean Wr differed among length-groups, years, and (RKM 1.4). These tributaries are known spawning river reaches during March, and (3) to determine locations for the ¯annelmouth sucker and other whether mean Wr differed among length-groups, native species (Valdez and Ryel 1995; Weiss et al. years, and reaches during September. In all anal-

1998; Douglas and Douglas 2000; Meretsky et al. yses, the Wr data were log10 transformed to better 2000), and unlike the main-stem Colorado River, meet the ANOVA assumptions (Zar 1996). the tributaries have remained relatively unchanged Correlation analyses were used to determine the in terms of ¯ow and temperature regimes. relation between ¯annelmouth sucker mean Wr and water ¯ows recorded at three U.S. Geological Sur- Methods vey (USGS) gauging stations (Paria River at Lee's Fish collection.ÐFlannelmouth suckers were Ferry [RKM 1.4], Colorado River above Lee's Fer- collected with a variety of techniques, including ry [RKM 0.2], and [RKM hoop nets, trammel nets, and electro®shing (e.g., 98.7] near Cameron, Arizona). Correlation anal-

Valdez and Ryel 1995). Sampling took place from ysis was also used to compare mean Wr with mean March 1991 to September 2001 in the main-stem macroinvertebrate density (number/m2) deter- Colorado River (Glen and Grand canyons) be- mined throughout the study reach with a Hess sam- tween the Paria River and Diamond Creek. These pler in June 1991±2001 (Benenati et al. 2002). We collections were part of continued long-term mon- used only ¯annelmouth suckers collected in Sep- itoring of Grand Canyon ®sh populations and were tember (the most common sampling month) to not sampled speci®cally for this study. Each ®sh avoid bias associated with seasonal differences. was measured for total length (TL) to the nearest Mean water ¯ows for the 3 months prior to the millimeter and weighed to the nearest gram. Be- September ®sh collections (June±August) were cause of the uncertainty of sex determinations, we used in correlation analyses because changes in did not distinguish between male and female ¯an- condition might not respond instantaneously to nelmouth suckers in the analysis. changes in environmental factors.

Analysis.ÐRelative weight (Wr; Blackwell et al. 2000) was used to assess trends in condition. The Results standard weight (W ) equation used to determine s Monthly Differences in Condition ¯annelmouth sucker Wr was described by Didenko et al. (in press). Only ®sh collected in 1993 were A total of 673 ¯annelmouth suckers were used to determine monthly (i.e., seasonal) trends weighed and measured during January±November in condition because this was the most complete 1993. We found no signi®cant two- or three-way annual data set (®sh were collected throughout the interactions among length-group, month, or river

Grand Canyon from January to November). To de- reach (ANOVA, P Ն 0.166). However, mean Wr scribe annual trends, we selected the two months differed across months (F ϭ 4.50; df ϭ 10, 531; (March and September) for which the data set was P Ͻ 0.001), length-groups (F ϭ 3.04; df ϭ 3, 531; most complete from 1991 to 2001. Not all lengths P ϭ 0.029), and river reaches (F ϭ 2.54; df ϭ 18, of ¯annelmouth sucker were collected in every 531; P ϭ 0.0005). Mean monthly Wr was greatest 650 PAUKERT AND ROGERS

FIGURE 1.ÐMonthly mean relative weight (Wr) values (ϮSEs) for ¯annelmouth suckers collected in the Col- FIGURE 2.ÐMean September relative weight (Wr) val- orado River, Grand Canyon, Arizona, in 1993. ues (ϮSEs) for ¯annelmouth suckers collected in the Colorado River, Grand Canyon, Arizona, 1991±2001.

(Ͼ100) during January±May, peaking at 107 in

March; mean Wr later decreased to 93 in August (Figure 1). Mean W was highly variable among September sampling revealed no signi®cant r two- or three-way interactions among river reach, river reaches. The highest mean Wr values were observed at river reaches RKM 225 (mean ϭ 108; year, or length-group (P Ն 0.284). Mean Wr did SE ϭ 4.5), RKM 177 (mean ϭ 107; SE ϭ 2.3), not differ among length-groups (F ϭ 0.24; df ϭ and RKM 193 (mean ϭ 107; SE ϭ 1.9). The lowest 3, 411; P ϭ 0.869), but did differ signi®cantly mean values were observed at river reaches RKM among years (F ϭ 3.30; df ϭ 9, 411; P ϭ 0.0007) 306 (mean ϭ 94; SE ϭ 6.1) and RKM 338 (mean and among river reaches (F ϭ 2.18; df ϭ 17, 411; P ϭ 0.004). In general, mean Wr decreased in ϭ 94; SE ϭ 3.0). Mean Wr was never lower than 94, and all reaches upstream of reach RKM 225 downriver reaches; 9 of 10 river reaches below reach RKM 209 had mean W values lower than had mean Wr values greater than 100. Overall, r 100, and 5 of 8 river reaches above RKM 209 had mean Wr was highest for 400±499-mm ®sh (mean ϭ 101; SE ϭ 0.8) and lowest for 200±299-mm ®sh mean Wr values greater than 100. Although the (mean ϭ 97; SE ϭ 0.8). Although signi®cantly highest mean Wr was observed in river reach RKM different (F ϭ 3.04; df ϭ 3, 531; P ϭ 0.029), these 354 (mean ϭ 106; SE ϭ 6.4), only three ®sh were two Wr values were very similar. collected in that segment. The lowest mean Sep- tember Wr occurred in 1993 and 1995 (mean ϭ 96 Annual Differences in Condition for each year; Figure 2). Mean Wr increased in A total of 193 ¯annelmouth suckers were col- 1996 and 1997 and began a decline in 1998, but lected in March 1991±1993 and 1995±1997, and was never lower than 95 (Figure 2). a total of 554 ¯annelmouth suckers were collected in September 1991±1993 and 1995±2001. Analy- Correlation between Condition and Water Flows ses of March data indicated no signi®cant two- or three-way interactions among length-group, year, Flannelmouth sucker condition was strongly re- lated to water ¯ows from Glen Canyon Dam. Mean or river reach (P Ն 0.555). Mean Wr did not differ signi®cantly among years (F ϭ 0.20; df ϭ 5, 131; Wr of ¯annelmouth suckers collected in September P ϭ 0.964) or length-groups (F ϭ 1.31; df ϭ 3, 1991±2001 was positively correlated with mean monthly summer ¯ow (June±August) in the Col- 131; P ϭ 0.274). Mean Wr was variable among river reaches (F ϭ 2.20; df ϭ 12, 131; P ϭ 0.015), orado River above Lee's Ferry (r ϭ 0.71, P ϭ 0.02, but was highest near reach RKM 177 (mean ϭ N ϭ 10; Figure 3A), but not with Paria River ¯ow 118; SE ϭ 8.3) and lowest in river reach RKM (r ϭ 0.44, P ϭ 0.20, N ϭ 10) or Little Colorado

354 (mean ϭ 94; SE ϭ 2.7). Mean Wr exceeded River ¯ow (r ϭ 0.13, P ϭ 0.72, N ϭ 10). Mean 100 in all river reaches between RKM 97 and RKM September Wr of ¯annelmouth suckers tended to 306, whereas samples collected above RKM 97 or increase with mean June macroinvertebrate den- below RKM 306 had mean Wr values less than sity between RKM 0.8 and RKM 328.8 (r ϭ 0.57, 100. P ϭ 0.08, N ϭ 10; Figure 3B). FLANNELMOUTH SUCKER CONDITION 651

spawning (Pope and Willis 1996), individual ¯an- nelmouth suckers may have exhibited increased condition. However, the large scale of our analysis suggests that many of the ®sh sampled did not spawn in the main-stem Colorado River in Octo- ber. Flannelmouth sucker condition was variable throughout the 363-RKM-long study area. How- ever, the greatest condition values were typically observed at intermediate distances from Glen Can- yon Dam for both the seasonal and annual data. Increased condition may be associated with fa- vorable environmental conditions (Blackwell et al. 2000) occurring in these intermediate reaches. Biomass of the alga Cladophora glomerata peaked at RKM 104 (similar to the location of peak con- dition indices in our study), but macroinvertebrate drift, a common food source for ¯annelmouth suckers (Minckley 1991), was typically greatest in upstream segments (Shannon et al. 1996). Warm- water tributaries can be important spawning and rearing areas for native ®shes (Valdez and Ryel 1995; Weiss et al. 1998; Douglas and Douglas 2000); seven major warmwater tributaries of the FIGURE 3.ÐCorrelations of mean September relative Colorado River occur at intermediate distances weight (Wr) of ¯annelmouth suckers with (A) mean monthly summer (June±August) ¯ows (cubic feet per from Glen Canyon Dam (RKM 141±152). There- second) from the Colorado River above Lee's Ferry, fore, an increased number of warmwater tributaries 1991±2001 and (B) mean June macroinvertebrate den- in a reach may be related to increased condition sity in the Colorado River (river kilometers 0.8±328.8), in ¯annelmouth suckers. 1991±2001 (Benenati et al. 2002). No ®sh were collected Flannelmouth sucker condition peaked in Sep- in September 1994. tember 1996 and 1997, but the condition of ®sh sampled in March did not differ among years. In contrast, McKinney et al. (1999) found no annual Discussion difference in condition (Fulton's condition factor, Flannelmouth sucker condition peaked during K) of ¯annelmouth suckers collected from Feb- the prespawn period and subsequently declined. ruary to December 1992±1997 in the Colorado Flannelmouth suckers spawn in March and April River at Lee's Ferry. Condition in our study was (Weiss et al. 1998), coincident with the period of strongly correlated to water ¯ows from Glen Can- peak ¯annelmouth sucker condition. Similarly, yon Dam. Increased ¯ows in the Grand Canyon McKinney et al. (1999) found ¯annelmouth sucker increase the size of the euphotic zone, particularly condition at Lee's Ferry also increased in spring in the upper canyon above RKM 189 (M. Yard, and was lowest in summer. Decreased condition USGS, unpublished data). This ®nding is likely after spawning is common in ®sh (Pope and Willis related to the increased macroinvertebrate densi- 1996) and likely occurred in ¯annelmouth suckers ties during peak ¯ows, which was documented by in the Grand Canyon. Our results suggest that ¯an- Benenati et al. (2002). nelmouth suckers do not rebound with increased Mean ¯annelmouth sucker condition never de- condition in late summer (October or November), clined below 93, despite the fact that these ®sh as has been observed for the humpback chub (Val- inhabit a cold, stenothermic environment in which dez and Ryel 1995; Meretsky et al. 2000). How- water temperatures are typically near 10ЊC, much ever, Douglas and Douglas (2000) documented Oc- lower than the species' preferred temperature of tober spawning of ¯annelmouth suckers in Havasu 25ЊC (Deacon et al. 1987). Although cold water Creek (RKM 252.2). Although our results showed temperatures in the main-stem Colorado River may no evidence of increased October condition sug- reduce recruitment (Thieme et al. 2001), swim- gestive of gonadal development and possible ming performance (Ward et al. 2002), and possibly 652 PAUKERT AND ROGERS growth (Valdez and Ryel 1995), ¯annelmouth Chart, T. E., and E. P. Bergersen. 1992. Impact of main- suckers can become tolerant of cold water (Chart stem impoundment on the distribution and move- and Bergersen 1992; McKinney et al. 1999) and ment of the resident ¯annelmouth sucker (Catos- tomidae: Catostomus latipinnis) population in the still increase in weight. In addition, the Ws equa- White River, Colorado. Southwestern Naturalist 37: tion used to calculate Wr was likely created from 9±15. populations in regulated systems. Therefore, Wr Clarkson, R. W., and M. L. Childs. 2000. Temperature values in our study may re¯ect condition relative effects of hypolimnial-release dams on early life to ®sh in other regulated systems, but not historical stages of Colorado River basin big-river ®shes. (i.e., pre-dam) ¯annelmouth sucker condition. Copeia 2000:402±412. Coughlan, D. J., B. K. Baker, D. G. Gloutman, and W. Flannelmouth suckers apparently are able to tol- M. Rash. 1996. Application and modi®cation of the erate the highly regulated, stenothermic conditions ®sh health assessment index used for largemouth in the Colorado River below Glen Canyon Dam. bass in the Catawba River, North Carolina±South Mean Wr values were at or near 100, corresponding Carolina. Pages 73±84 in L. E. Miranda and D. R. to the 75th percentile of weights at a given length DeVries, editors. Multidimensional approaches to (Blackwell et al. 2000), which suggests that con- reservoir ®sheries management. American Fisheries dition was moderate to high compared to that of Society, Symposium 16, Bethesda, Maryland. Deacon, J. E., P. B. Schumann, and E. L. Stuenkel. 1987. other ¯annelmouth sucker populations. However, Thermal tolerances and preferences of ®shes of the the greatest condition values were associated with system (Utah, Arizona, Nevada). Great periods of higher ¯ows and productivity. Increased Basin Naturalist 47:538±546. condition in other ®shes has been associated with DiCenzio, V. J., M. J. Maceina, and W. C. Reeves. 1995. increased food availability (e.g., Marwitz and Factors related to growth and condition of the Al- Hubert 1997; Paukert and Willis 2003), and this abama subspecies of spotted bass in reservoirs. likely also occurred with ¯annelmouth suckers in North American Journal of Fisheries Management 15:794±798. the Colorado River. Efforts to increase dam dis- Didenko, A., S. A. Bonar, and W. J. Matter. In press. charge during the summer months may increase Standard weight (Ws) equations for roundtail chub, food availability and bene®t the native ¯annel- ¯annelmouth sucker, razorback sucker, and hump- mouth suckers in this large river system. back chub. North American Journal of Fisheries Management. Acknowledgments Douglas, M. E., and P. C. Marsh. 1998. Population and survival estimates of Catostomus latipinnis in north- We thank all the people who collected ¯annel- ern Grand Canyon, with distribution and abundance mouth suckers in the Grand Canyon, including the of hybrids with Xyrauchen texanus. Copiea 1998: Arizona Game and Fish Department (AGFD), 915±925. SWCA Environmental Consultants, Inc., BIO/ Douglas, M. R., and M. E. Douglas. 2000. Late-season WEST, Humphrey Summit Support, and the U.S. reproduction by big-river in Grand Fish and Wildlife Service (USFWS). David Ward Canyon (Arizona). Copiea 2000:238±244. Johnson, S. L., F. J. Rahel, and W. A. Hubert. 1992. (AGFD), Lew Coggins (USGS), Dennis Stone Factors in¯uencing the size structure of brook trout (USFWS), Kevin Pope (Texas Tech University), populations in beaver ponds in Wyoming. North and two anonymous reviewers provided helpful American Journal of Fisheries Management 12: comments on this paper. Scott Bonar of the Ari- 118±124. zona Cooperative Fish and Wildlife Research Unit Marwitz, T. D., and W. A. Hubert. 1997. Trends in rel- ative weight of walleye stocks in Wyoming reser- provided the Ws equations prior to their publica- tion. voirs. North American Journal of Fisheries Man- agement 17:44±53. McKinney, T., W. R. Persons, and R. S. Rogers. 1999. References Ecology of ¯annelmouth sucker in the Lee's Ferry Benenati, E. P., J. P. Shannon, G. A. Haden, K. Straka, tailwater, Colorado River, Arizona. Great Basin and D. W. Blinn. 2002. Monitoring and research: Naturalist 59:259±265. the aquatic food base in the Colorado River, Ari- Meretsky, V. J., R. A. Valdez, M. E. Douglas, M. J. zona, during 1991±2001. U.S. Geological Survey, Brouder, O. T. Gorman, and P.C. Marsh. 2000. Spa- Grand Canyon Monitoring and Research Center, Fi- tiotemporal variation in length±weight relationships nal Report, Cooperative Agreement Number 1452- of endangered humpback chub: implications for 98-FC-225590, Flagstaff, Arizona. conservation and management. Transactions of the Blackwell, B. G., M. L. Brown, and D. W. Willis. 2000. American Fisheries Society 129:419±428.

Relative weight (Wr) status and current use in ®sh- Minckley, W. L. 1991. Native ®shes of the Grand Can- eries assessment and management. Reviews in Fish- yon region: an obituary? Pages 124±177 in Colo- eries Science 8:1±44. rado River ecology and dam management: proceed- FLANNELMOUTH SUCKER CONDITION 653

ings of a symposium. National Academy of Sci- Thieme, M. L., C. C. McIvor, M. J. Brouder, and T. L. ences, Washington, D.C. Hoffnagle. 2001. Effects of pool formation and Neumann, R. M., and B. R. Murphy. 1992. Seasonal ¯ash ¯ooding on relative abundance of young-of- relationships with relative weight of body compo- year ¯annelmouth suckers in the Paria River, Ari- sition in white crappie, Pomoxis annularis Ra®n- zona. Regulated Rivers: Research and Management esque. Aquaculture and Fisheries Management 23: 17:145±156. 243±251. Valdez, R. A., and R. J. Ryel. 1995. Life history and Paukert, C. P., and D. W. Willis. 2003. Population char- ecology of the humpback chub (Gila cypha) in the acteristics and ecological role of northern pike in Colorado River, Grand Canyon, Arizona. Final Re- shallow natural lakes in Nebraska. North American port to the Bureau of Reclamation, Contract Number Journal of Fisheries Management 23:313±322. 0-CS-40-09110, Salt Lake City, Utah. Pope, K. L., and D. W. Willis. 1996. Seasonal in¯uences Ward, D. L., O. E. Maughan, S. A. Bonar, and W. J. on freshwater ®sheries sampling data. Reviews in Matter. 2002. Effects of temperature, ®sh length, Fisheries Science 4:57±73. and exercise on swimming performance of age-0 Shannon, J. P., D. W. Blinn, P. L. Benenati, and K. P. ¯annelmouth sucker. Transactions of the American Wilson. 1996. Organic drift in a regulated desert Fisheries Society 131:492±497. river. Canadian Journal of Fisheries and Aquatic Weiss, S. J., E. O. Otis, and E. O. Maughan. 1998. Sciences 53:1350±1369. Spawning ecology of ¯annelmouth sucker, Catos- Stevens, L. E., J. P. Shannon, and D. W. Blinn. 1997. tomus latipinnis (Catostomidae), in two small trib- Colorado River benthic ecology in Grand Canyon, utaries of the lower Colorado River. Environmental Arizona, USA: dam, tributary, and geomorpholog- Biology of Fishes 52:419±433. ical in¯uences. Regulated Rivers: Research and Zar, J. H. 1996. Biostatistical analysis, 3rd edition. Pren- Management 13:129±149. tice Hall, Upper Saddle River, New Jersey.