II. IMPACTS OF STREAM FLOW ALTERATIONS ON THE NATIVE FISH ASSEMBLAGE AND THEIR HABITAT AVAILABILITY AS DETERMINED BY 2D MODELING AND THE USE OF FISH POPULATION DATA TO SUPPORT INSTREAM FLOW RECOMMENDATIONS FOR THE SECTIONS OF THE YAMPA, COLORADO, GUNNISON AND DOLORES RIVERS IN COLORADO

EXECUTIVE SUMMARY ...... II- 1 Gunnison River ...... II- 33 INTRODUCTION ...... II- 2 Dolores River ...... II- 33 STUDY AREAS ...... II- 4 CONCLUSIONS ...... II- 34 Yampa River Site Locations & Hydrology ...... II- 4 FLOW RECOMMENDATIONS ...... II- 37 Colorado River Site Locations & Hydrology ...... II- 5 Yampa River ...... II- 37 Gunnison River Site Locations & Hydrology . . . . . II- 6 Colorado River ...... II- 40 Dolores River Site Location & Hydrology ...... II- 7 Gunnison River ...... II- 44 Dolores River ...... II- 47 METHODS ...... II- 9 ACKNOWLEDGEMENTS ...... II- 51 RESULTS ...... II- 11 LITERATURE CITED ...... II- 52 Fish Analysis ...... II-11 Duffy, Yampa River ...... II-11 APPENDIX ONE - Site Hydrographs ...... II- 55 Sevens, Yampa River ...... II- 11 APPENDIX TWO - Fish Data Tables ...... II- 59 Lily Park, Yampa River ...... II- 12 TABLES Flow and Degree-Day Correlation ...... II- 12 II- 1 - Depth and velocity criteria used to define Corn Lake, Colorado River ...... II- 13 meso-habitat types ...... II- 10 Clifton, Colorado River ...... II- 13 II- 2. Species composition (%) for fish >15 cm for pre- and post 2002, for all study sites surveyed during project...... II- 16 Escalante, Gunnison River ...... II- 14 II- 3. Biomass estimates for fish >15 cm for pre- and post II- Delta, Gunnison River ...... 14 2002 for all study sites surveyed during project...... II- 17 Big Gypsum, Dolores River ...... II- 15 II- 4. Physical attributes of each study site (length, mean Geomorphic and Habitat Analysis – All Sites . . . . II- 18 velocity, mean width, and surface area) at modeled flows .... II- 19 DISCUSSION ...... II- 22 II- 5. Species composition (%) for fish >15 cm, Delta (2D site) and Escalante, Gunnison River, 2003, Base Flows ...... II- 22 2004 and 2005...... II- 21 Spring Runoff Flows ...... II- 24 FIGURES II- 1. Yampa River study site locations: Lily Park (RM 53), Nonnative Species ...... II- 25 Sevens (RM 64) and Duffy (RM 109)...... II- 4 Species Account ...... II- 26 II- 2. Colorado River study site locations at Corn Lake Bluehead Sucker ...... II- 26 and Clifton ...... II- 5 Flannelmouth Sucker ...... II- 27 II- 3. Gunnison River study site locations at Escalante and Delta (Uncompadgre River confluence)...... II- 6 Roundtail Chub ...... II- 28 II- 4. Dolores River study site location in the Big Gypsum Endangered Species ...... II- 29 valley...... II- 8 Colorado Pikeminnow ...... II- 29 II- 5. Modeled bluehead sucker biomass (kg/ha) as a function White Sucker ...... II- 30 of discharge, Lily Park, Sevens and Duffy, Yampa River ...... II- 37 Nonnative Predators ...... II- 30 II- 6. Percent of maximum bluehead sucker modeled II- biomass as a function of discharge, Lily Park and Northern Pike ...... 30 Sevens, Yampa River ...... II- 38 Smallmouth Bass ...... II- 31 II- 7. Modeled flannelmouth sucker biomass (kg/ha) as a Channel Catfish ...... II- 31 function of discharge, Lily Park, Sevens and Duffy, Yampa River ...... II- .39 Supplement Data ...... II- 31 II- 8. Percent of maximum flannelmouth sucker modeled Colorado River ...... II- 31 biomass as a function of discharge, Lily Park, Yampa River ...... II- 32 Sevens and Duffy, Yampa River ...... II- 39 II- 9. Modeled bluehead sucker biomass (kg/ha) as a function A1-14.Colorad River (Palisade gage) annual minimum flows of discharge, Corn Lake and Clifton, Colorado River ...... II- 41 1998 to 2005...... II- 58 II- 10. Percent of maximum bluehead sucker modeled biomass A1-15. Gunnison River (Delta plus Uncompraghe gages) as a function of discharge, Corn Lake and Clifton, annual minimum flows 1998-2005) ...... II- 58 Colorado River ...... II- 41 A1-16. Dolores River (Bedrock gage) annual minimum flows II- 11. Modeled flannelmouth sucker biomass (kg/ha) as a 1998-2005...... II- 58 function of discharge, Corn Lake and Clifton, Colorado River ...... II- 43 APPENDIX TWO (Fish Data Tables) II- 12. Percent of maximum flannelmouth sucker modeled A2-1. Species composition, density and biomass estimated biomass as a function of discharge, Corn Lake and at Duffy, Yampa River, 1998 to 2004...... II- 59 Clifton, Colorado River ...... II- 43 A2-2 Species composition, density and biomass estimated II- 13. Modeled bluehead sucker biomass (kg/ha) as a at Sevens, Yampa River, 1998 to 2004 ...... II- 60 function of discharge, Delta and Escalante, A2-3. Species composition, density and biomass estimated Gunnison River ...... II- 44 at Lily Park, Yampa River, 2000 to 2004...... II- 61 II- 14. Percent of maximum bluehead sucker modeled A2-4 Species composition, density and biomass estimated biomass as a function of discharge, Delta and Escalante, at Corn Lake, Colorado River 1999 to 2004...... II- 62 Gunnison River ...... II- 45 A2-5. Species composition, density and biomass estimated II- 15. Modeled flannelmouth sucker biomass (kg/ha) as a function at Clifton, Colorado River, 2000 to 2003...... II- 63 of discharge, Delta and Escalante, Gunnison River ...... II- 46 A2-6. Species compositon, density and biomass estimated II- 16. Percent of maximum flannelmouth sucker modeled at Escalante, Gunnison River, 2004 and 2005...... II- 64 biomass as a function of discharge, Delta and Escalante, A2-7. Species composition, density and biomass estimated Gunnison River ...... II- 46 at Delta, Gunnison River, 2004 and 2005 ...... II- 65 II- 17. Modeled bluehead sucker biomass (kg/ha) as a function of A2-8. Species compositon, density and biomass estimated discharge, Big Gypsum, Dolores River ...... II- 47 at Big Gypsum, Dolores River, 2000 to 2005...... II- 66 II- 18. Percent of maximum bluehead sucker modeled biomass as a function of discharge, Big Gypsum, Dolores River ...... II- 48 CD (DISK) APPENDICES II- 19. Modeled flannelmouth sucker biomass (kg/ha) as a Fish- Histograms: Figures 1-99. Length frequency histograms function of discharge, Big Gypsum, Dolores River ...... II- 48 for Yampa, Colorado, Gunnison and Dolores Rivers sites, 1998 to 2005...... CD II- 20. Percent of maximum flannelmouth sucker modeled biomass as a function of discharge, Big Gypsum, Fish- Gunnison River 1992 -93: Tables 1-4. Comparison for fish Dolores River ...... II- 49 collected by electrofishing in the Gunnison River in 1992 & 1993 (Burdick 1995) to collections in 2003, 2004 and 2005 (Anderson and Stewart 2006) ...... CD APPENDIX ONE (Site Hydrograph) Habitat- 2D modeling: Figures 1-16. Meso-habitat and Shannon A1-1. Yampa River (Maybell gage) hydrograph 1998 to 2004 ...... II- 55 Diversity at modeled flows for Corn Lake, Clifton, Lily Park, A1-2 Colorado River ( Palisade gage) hydrograph 1998 to 2004 ... II- 55 Sevens, Duffy, Big Gypsum, Delta and Escalante...... CD A1-3. Gunnison River (Delta plus Uncompaghre gages) Reports –Telemetry: Byers et al. (2001) and Rees and hydrograph 1998 to 2005 ...... II- 55 Miller (2001)...... CD A1-4 Dolores River (Bedrock gage) hydrograph 1998 to 2005...... II- 55 Reports - F-288-R Federal Aid: Anderson (2005), Anderson A1-5. Yampa River (Maybell gage) annual peak flows and Stewart (2003), Anderson and Stewart (2006), 1998 to 2005 ...... II- 56 Stewart (2000), Stewart et al. (2005), Stewart and Anderson (2006, in press) ...... CD A1-6. Colorad River (Palisade gage) annual peak flows 1998 to 2005...... II- 56 Site Topography: Plotfiles for each modeled flow at each study sites: Clifton and Corn Lake (Colorado River), A1-7. Gunnison River (Delta plus Uncompraghe gages) Big Gypsum (Dolores River), Delta and Escalante annual peak flows 1998-2005) ...... II- 56 (Gunnison River), Duffy, Lily Park and Sevens A1-8. Dolores River (Bedrock gage) annual peak flows (Yampa River)...... CD 1998-2005...... II- 56 A1-9. Yampa River (Maybell gage) hydrograph below 400 cfs 1998 to 2004 ...... II- 57 A1-10 Colorado River ( Palisade gage) hydrograph below 800 cfs 1998 to 2004 ...... II- 57 A1-11. Gunnison River (Delta plus Uncompaghre gages) hydrograph below 1,000 cfs 1998 to 2005 ...... II- 57 A1-12 Dolores River (Bedrock gage) hydrograph below 60 cfs 1998 to 2005...... II- 57 A1-13. Yampa River (Maybell gage) annual minimum flows 1998 to 2005 ...... II- 58 IMPACTS OF STREAM FLOW ALTERATIONS ON THE NATIVE FISH ASSEMBLAGE AND THEIR HABITAT AVAILABILITY AS DETERMINED BY 2D MODELING AND THE USE OF FISH POPULATION DATA TO SUPPORT INSTREAM FLOW RECOMMENDATIONS FOR THE SECTIONS OF THE YAMPA, COLORADO, GUNNISON AND DOLORES RIVERS IN COLORADO

EXECUTIVE SUMMARY

In 1999 the Colorado Water Conservation Board Abundances of bluehead sucker, (Catostomus (CWCB) asked the Colorado Division of Wildlife discobolus ), flannelmouth sucker ( Catostomus (CDOW) to provide biologically justified instream latipinnis ) and roundtail chub ( Gila robusta ) were flow recommendations for the Yampa and Colorado much higher in moderate to high base flow rivers. Rivers. This research project, ‘Riverine Fish-Flow Higher base flows were associated with greater Investigation’, evaluated the use of two-dimensional availability of riffle habitats and increased fish biomass. (2D) flow models for determining habitat preferences Spring runoff flows were secondary to base flows in of native fish and for developing instream flows for maintenance of native species diversity and biomass. the Yampa River. Some evidence suggested that spring runoff flows were The first paper of this project completion report, related to the reproductive success of bluehead sucker. Stewart and Anderson (2007), described the approach, Reduced spring runoff flows resulted in obvious methods and results of using 2D modeling for relating geomorphic impacts on the Dolores River, but not on flow and fish habitat availability. The current report the Gunnison River. summarized and evaluated fish community structures at Reduced base flows on the Yampa River were each of the 2D modeling sites. Trends in native fish associated with dramatic increases in certain nonnative abundance along sections of four rivers in western species, decreases in total fish biomass, increased rates Colorado – the Yampa, Colorado, Gunnison and Dolores of predation and increased rates of white sucker – were related to stream flow and habitat characteristics hybridization with flannelmouth and bluehead sucker. at eight study sites. The purpose of this report is to Increased abundance of nonnative species was usually provide community scale biological justifications for associated with negative impacts to the native species the instream flow recommendations that utilized a 2D assemblage. In fact, by 2004 all native fish species modeling process that focused on two common native were rare in the upper Yampa River and nonnative sucker species, the flannelmouth and bluehead sucker. species were proving highly problematic to the Changes in hydrograph over time have been linked recovery of native fish. to declines in native fish abundance, suggesting that The abundance of bluehead sucker was a reliable alterations in native fish assemblages could be indicator for adequate base flows and habitat consistent with altered hydrographs for each river. maintenance for the native fish assemblage. Bluehead From 2000 to 2004 drought conditions led to severe sucker habitat peaked at flows of 600 to 1,200 cfs for reductions in flow on the Yampa and Dolores Rivers. the Colorado, Gunnison and Yampa Rivers. That flow Fish sampling during those years quantified changes in range was also associated with high habitat diversity fish communities following severely reduced flows. and high native fish biomass. The research findings Pre- and post- drought fish data documented the validated the assumption that flows that maintained persistence of the native fish assemblage under a wide adequate bluehead sucker abundance (about 25 percent range of flow conditions. Although the four rivers of the community) will also maintain adequate habitat exhibited very different flow regimes, the fish data for flannelmouth sucker and roundtail chub habitat and were consistent with model projections made using the also probably for rare and endangered native fish. 2D methodology (Stewart and Anderson 2007).

II-1 INTRODUCTION

Maintaining stream flows for native fish abundance would strongly suggest that changes in the management has become a priority for both state native fish assemblage were a response to altered base (Espegren 1998) and federal agencies (McAda 2003) flow regimes. in Colorado. Instream flow recommendations serve The primary objectives of this analysis were to to identify the maximum amount of water that can be identify patterns in native fish persistence and to removed from a stream without adversely altering its relate native fish abundance in each site to its flow ecosystem and natural processes (Annear et al. 2002). and habitat conditions. The basic assumption of the Extended periods of reduced flows can alter fish research was that difference in bluehead sucker and assemblages and lower carrying capacity for native flannelmouth sucker biomass under different flow fishes (Travnichek et al. 1995). scenarios would conform to 2D model predictions A primary objective of this research project was to given habitat suitability criteria were correctly determine habitat suitability criteria for three native identified. species still common to western Colorado – bluehead Efforts to promote the ecological integrity of river sucker, flannelmouth sucker and roundtail chub – for ecosystems have relied on mimicry of the natural flow use in developing instream flow recommendations regime (Burdick 1995, Modde and Smith 1995, Poff (Stewart and Anderson 2006, Special Report A). The et al. 1997). Tyus and Karp (1989) speculated that the research in this report (Part I) identified habitat higher persistence of native fishes in the Yampa River preferences for bluehead and flannelmouth sucker was associated with maintenance of habitats sustained from populations in the Yampa and Colorado Rivers by a relatively unaltered regime of fluctuating using a GIS approach that employed two-dimensional seasonal and annual flows. This hypothesis – that (2D) flow models of depth and velocity. Bluehead native fish persistence will be higher in rivers with sucker were highly associated with the availability of lesser degrees of altered hydrographs – is not very moderately deep riffle habitats whereas flannelmouth useful in identifying specific flow management sucker were associated with deep runs (Anderson and options. Some flow alterations may be beneficial and Stewart 2003). Roundtail chub were found to occupy not all hydrographic phases are equivalent in their multiple habitats, but the research did not include a biological impacts. method for developing habitat suitability criteria for Periodic drops in base flows have been a natural species that utilize multiple habitats. occurrence in Colorado. Quantifying the impacts of The Yampa and Colorado rivers exhibited similar sustained periods of reduced flow requires data for channel morphology but the Yampa had lower base both fish abundance and habitat composition under flows. Together the two rivers were in the low and altered and unaltered flow conditions. Long term moderate flow ranges for native fish habitat flow reductions may facilitate establishment of availability. Habitat suitability criteria for bluehead nonnative fish species. Nonnative fish have been and flannelmouth sucker, developed from the found to negatively affect the structure and biomass of Colorado and Yampa Rivers (Anderson and Stewart native fish assemblages (Courtenay and Moyle 1992, 2003), were also valid when used to predict bluehead Scoppettone 1993). Nonnative fishes impact native and flannelmouth sucker biomass at two sites on the fish populations through competition for limited Gunnison River. resources, predation and hybridization. Evidence of valid suitability criteria is obtained The complexity of examining impacts of reduced when biomass estimates from field observations flows on native fish is greatly amplified when match the model results. Because pre-drought fish nonnative fish are a major component of a fish biomass data were available, the 2002 Colorado of assemblage. When nonnative fishes are only a minor record drought provided an opportunity to assess component of the fish community, temporary or even impacts on native fish biomass during periods of long term flow reductions may not impose lasting severely reduced flows. Declines in native fish consequences on native fish, because their abundance

II-2 will likely increase when their habitats are restored. nonnative species. The persistence of native fish in However, when severe and prolonged flow reductions these rivers should provide insights into the result in a major expansion of nonnative fish species, consequences of these altered hydrographs, the recent the native fish assemblage may not be able to recover drought and species introductions. More specifically, even if natural flows return. the persistence of bluehead and flannelmouth sucker The Yampa, Colorado, Gunnison and Dolores in rivers with different flow regimes should validate rivers originally had typical snowmelt-driven annual the reliability of the 2D modeling results. hydrographs. Water development projects have altered the hydrographs of all four rivers, which are now distinctive for their runoff and base flow periods. Originally, the four rivers had the same native fish assemblage, but today each now includes several

II-3 STUDY AREA

Yampa River Site Hydrology Site Locations The Yampa River maintains relatively natural spring flows, with volume reduced by an average of The Yampa River, the largest tributary of the six percent during the months of April, May and Green River, flows westward through northwestern June (Modde et al. 1999). At the Deerlodge gage Colorado. Three study sites were established (RM 50), which is downstream of the Little Snake upstream of the confluence with the Little Snake River confluence, the Yampa drains an area of 7,660 River (Figure II-1). From its confluence with the mi 2 (19,839 km 2) and the annual average flow was Green River to the town of Craig, the Yampa River 2,049 cfs (58.0 cms) from 1983 to 2004. We used has been designated critical habitat for four federally the Maybell gage (09251000) at RM 85.8 to endangered fish species. represent flows for the three study sites. The The Duffy study site began at RM 109 and was drainage area at the Maybell gage is 3,410 mi 2 7.2 km in length. The Sevens site began at RM 64 (8,832 km 2) and the mean annual flow was 1,548 cfs and was 2.9 km long. Electro-fishing crews sampled (44.35 cms) from 1917 to 2005. both Duffy and Sevens in 1998, 1999, 2000, 2001, Bankfull flow determined from bed profile data at 2003 and 2004. A third site, Lily Park began at RM the Maybell gage was approximately 9,000 cfs (258 52 and was 3.1 km in length. Fish were sampled at cms) (Andrews 1980). Recent channel surveys from Lily Park in 2000, 2001, 2003 and 2004. No Sevens and Lily Park indicate a bankfull flow of 11,000 sampling took place on the Yampa in 2002. cfs (315 cms) (Richard and Anderson 2007). Peak

FIGURE II-1. Yampa River study site locations: Duffy, Sevens (RM 64) and Lily Park.

II-4 flows at the Maybell gage in 1998, 1999 and 2000 were at about RM 170 (Figure II-2). Two major upstream similar to the median peak flow of 9,930 cfs (281.2 diversions divert flow from the river during the cms) from 1917 – 2004 (Figure A1-1). During 2001, irrigation season (April 1 to November 1) and flows 2002 and 2004 peak flows were lower. In 2002 peak during irrigation season at the Palisade Gage are flow was only 3,420 cfs (96.9 cms) (Figure A1-5). typically 1,200 to 1,600 cfs (34 and 45 cms) less than Yampa River base flows are generally low relative those upstream of the diversions. Winter (November to peak and mean annual flows. Base flows below 250 to March) flows in the 15-Mile Reach typically cfs (7.1 cms) were infrequent in 1998 and 1999 but exceed 2,000 cfs (56 cms). High flows in winter base flows below 100 cfs were common from 2000 to result from deliveries to senior water rights. The 15- 2004 (Figure A1-9). The median minimum mean daily Mile Reach of the Colorado River is included in flow from the Maybell gage is 119 cfs (3.37 cms). critical habitat for endangered Colorado pikeminnow Annual minimum flows for 1998 through 2004 were (Ptychocheilus lucius ) and razorback sucker 115, 166, 30, 50, 1.8, 43 and 22 cfs, respectively (3.2, (Xyrauchen texanus ). 4.7, 0.8, 1.4, 0.05, 1.2 and 0.6 cms) (Figure A1-13). Two study sites were in the 15-Mile Reach: Clifton and Corn Lake. Clifton extended from RM 177.7 to RM 180.4 and has a total length of 4.2 km Colorado River (Figure II-2). Corn Lake was from RM 177.5 downstream to RM 175.3 and was 3.9 km in length. Site Locations Field crews sampled fish in 1999, 2000, 2001, 2003 The 15-Mile Reach of the Colorado River and 2004 at Corn Lake, but Clifton was sampled in extends from Palisade, Colorado (RM 185) 2000, 2001 and 2003. downstream to the confluence of the Gunnison River

FIGURE II-2. Colorado River study site locations in the 15-Mile Reach for Corn Lake and Clifton, Colorado.

II-5 Site Hydrology Gunnison River The Colorado River upstream of the confluence of the Gunnison River has a drainage area of 8,753 Site Locations mi 2 (22,670 km 2) and mean annual flow of 2,822 cfs The Gunnison River is the largest tributary to the (79.9 cms) during 1991 to 2004. Pitlick et al. (1999) upper Colorado River and its confluence with the determined bankfull flow for the 15-Mile Reach to Colorado River is located in the city of Grand be near 22,000 cfs (621 cms). Junction, CO. From its mouth upstream to the The median annual peak flow for the 14-year confluence of the Uncompahgre River at the town of Palisade gage (09106150) history was 13,250 cfs Delta, the Gunnison has been designated critical (375.2 cms) (Figure A1-2). Drought conditions habitat for endangered Colorado pikeminnow and reduced peaks flows in 2002 and 2004. The annual razorback sucker (McAda 2003) . peak in 2002 was only 2,780 cfs (78.7 cms), the Two study sites were on the Gunnison: Delta lowest for the period of record (Figure A1-6). and Escalante. The Delta study site extended from Base flows above 800 cfs (22.7 cms) are normal the Uncompahgre River confluence (RM 56.3) in the 15-Mile Reach. The 14-year median downstream 3.9 km to the county road bridge minimum flow for the Palisade gage is 543 cfs (15.4 (Figure II-3). The Escalante site was from Escalante cms). Flows were very low in 2002, typically under Bridge (RM 42.7) downstream about 4.4 km to Hail 100 cfs (2.8 cms) and less than normal in 2003 and Mary rapids (Figure II-3). 2004 (Figure A1-10). Minimum summer/fall flows (mean-daily) recorded at the Palisade Gage for 1998 through 2004 were 980, 1240, 543, 477, 58, 342 and 341 cfs, respectively (27.7, 35.1, 15.4, 13.5, 1.6, 9.7 and 9.7 cms) (Figure A2-14).

FIGURE II-3. Gunnison River study site locations: Delta and Escalante (gages are 1 km upstream of Delta and on Uncompahgre upstream of confluence).

II-6 Site Hydrology gage is from 1939 to present, with a mean annual Flow alterations for the lower Gunnison River flow of 301 cfs (8.5 cms). are best described by examination of Whitewater From 1977 to 2005 the median peak (mean gage (09152500) flow records, which is the most daily) flow for the combined Delta and downstream gage located about 12 km upstream of Uncompahgre gages was 6,192 cfs (175.4 cms) its confluence with the Colorado River. The (Figure A1-3) and it was 7,510 cfs (212.7 cms) for drainage area at the Whitewater gage was 7,928 mi 2 the Whitewater gage. The statewide drought started (20,534 km 2) and mean annual flow was 2,556 cfs in 2000 and peak and spring runoff flows were very (72.4 cms) for 1897 to 2004. The Aspinall Unit low in both 2002 and 2004 (Figures A2-3 and 5). reservoirs (Blue Mesa, Morrow Point and Crystal) Peak flow was only 1,454 cfs (41.2 cms) in 2002 at were completed by the United States Bureau of the Delta and Uncompahgre gages (Figure A1-7). Reclamation between 1966 and 1976. Mean annual Gunnison River base flows in the study reach flow was little changed by the project: the average generally exceed 800 cfs (22.7 cms), but they were mean annual flow from 1897 to 1965 was 2,611 cfs as low as 600 cfs (17.0 cms) in 2002 and 2003 (73.9 cms); it was 2,477 cfs (70.1 cms) from 1966 to (Figures A1-11 and A1-15). From 1998 to 2005 2004. The average mean annual flow from 1977 to summer/fall minimum mean daily flows for the 2004 was 2,567 cfs (72.7 cms). summed data from the Delta and Uncompahgre Gunnison River peak flows at Whitewater were gages were 1062, 1303, 1026, 832, 543, 622, 633 cfs substantially reduced after completion of the and 1,012 cfs, respectively (30.1, 36.9, 29.1, 23.6, Aspinall project. The median peak (mean daily) 15.4, 19.7, 18.0, 25.4 cms) (Figure A1-15). flow from 1897 to 1965 was 15,000 cfs (424.8 cms) Pitlick et al (1999) reported the mean bankfull compared to 7,355 cfs (208.3 cms) from 1966 to width was 73.4 m and mean slope was 0.12 percent 2005. Minimum flows have substantially increased at the Whitewater gage; the mean bankfull width since 1965. The Whitewater gage median minimum was 73 m and mean slope was 0.19 percent at the (mean daily) flow from 1897 to 1965 was 456 cfs Delta site; mean bankfull width was 68 m and mean (12.9 cms) compared to 864 cfs (24.5 cms) from slope was 0.12 percent slope at the Escalante site. 1966 to 2005. Pitlick et al. (1999) determined bankfull flow Dolores River was 14,600 cfs (414 cms) at Whitewater, which was common prior to 1965 but was rarely exceeded after 1966. The bankfull flow for the Delta and Escalante Site Location portions of the Gunnison River was 13,300 cfs (377 The Dolores River headwaters are in the San cms) (Pitlick et al. 1999). Juan Mountains of southwestern Colorado. The river The hydrographs for both the Delta and flows northward about 320 km to its confluence with Escalante sites were obtained by summing daily the Colorado River in Utah. From its mouth flows from two gages (the Gunnison River’s Delta upstream to the Bradfield Bridge, the river has a gage and the Uncompahgre River’s Delta gage), large roundtail chub population, making it instead of utilizing Whitewater gage flow records. potentially important for conservation of this The Delta and Uncompaghre gages were nearer the species. Although Colorado pikeminnow have been study sites than the downstream Whitewater gage. collected in the Dolores near its mouth (Valdez et al. The USGS Delta Gage (09144250) is located 1992), the Dolores is not critical habitat for about three km upstream of the Uncompahgre River endangered fish species. confluence. The Gunnison River drainage area at is The study site, located in the Big Gypsum Valley, 5,626 mi 2 (14,571 km 2). The period of record for began at the Bureau of Land Management boat launch the Delta gage is from 1977 to the present and the and ended just upstream of the county road bridge, mean annual flow is 1,978 cfs (56.0 cms). The which is 3.3 km downstream (Figure II-4). Uncompahgre gage (09149500) has a drainage area The site is about 72 river miles (116 km) of 1,115 mi (2,888 km). The period of record for this downstream from McPhee Reservoir and about 16

II-7 miles (26 km) downstream of Disappointment McPhee Reservoir has a storage capacity of Creek. Disappointment Creek puts a high volume of 381,000 AF (470,000 dam 3), which is similar to the fine sediments into the Dolores during runoff and 30-year average annual Dolores River inflow of storm events. 397,000 AF (490,000 dam 3). Total user demand is 3 Site Hydrology about 240,000 AF (296,000 dam ) per year. McPhee’s regulated outflow is a volume of water Two canals diverted virtually the entire Dolores called ‘the fish pool’ which is approximately 29,300 River flow (up to approximately 1,400 cfs or 40 AF (36,100 dam 3), equivalent to a mean annual flow cms) during the irrigation season (April to October) of 40 cfs (1.1 cms). Inflow that cannot be stored or from 1886 to 1984 (Dolores River Dialogue 2006a). diverted is called “the spill”. Since 1985 there have In 1985 McPhee Dam began storing water for the been seven years when demand exceeded the inflow Dolores River Project, a system of canals, tunnels and there was no spill. Spring outflow peaks ranged and laterals used to deliver water for irrigation. from 34 to 177 cfs (1.0 to 5.0 cms) during years with Before the damming the Dolores River, the no spill, while inflows ranged from 563 to 2,159 cfs hydrograph showed very low base flows (about 2 cfs (15.9 to 61.1 cms). The median inflow and outflow or 0.05 cms) during the irrigation season. Since peaks after 1985 were 2,941 and 2,009 cfs, 1985 McPhee Reservoir outflows have usually been respectively (83.3 and 56.9 cms). Demand reduces above 25 cfs (0.7 cms). On the other hand, spring outflow relative to inflow by about 40 percent during runoff has been much lower in magnitude and high flow years and by about 85 percent during dry duration than before completion of the dam. years.

D O L O R E S R IV E R

S a n M ig u e l R iv e BEDROCK # r N

W E

S

Montrose County

BIG GYPSUM study site San Miguel County R

E

V I

R

S

E

R

O

L

O

D

FIGURE II-4. Dolores River study site location: Big Gypsum.

II-8 The USGS Bedrock gage (09169500), used to During the drought of the early 2000s peak and represent flow conditions at the Big Gypsum site, is spring runoff flows were captured by reservoir located about 58 km downstream of the Big Gypsum storage. Peaks (mean daily flow) for 1998 through study site and about 174 km downstream of McPhee 2005 were 3,560, 3,100, 1,170, 522, 54, 323, 307 dam. The Dolores has a drainage area at Bedrock of and 5,060 cfs, respectively (100.8, 87.8, 33.1,14.8, 2,024 mi 2 (5,242 km 2). Mean annual stream flow at 1.5, 9.1, 8.6 and 143.3 cms) (Figure A1-8 and A1- Bedrock was 516 cfs (14.6 cms) before 1985 and 16). Lowered mainstem runoff flows during the 284 cfs (8.0 cms) from 1985 to 2004. We did not drought meant sediments from tributaries attempt to determine natural flow levels by adding accumulated in the channel instead of being flushed. back diversions and reservoir storage to gage The “fish pool” shares allocation shortages with records. other water project users. During the drought years The highest annual peak (mean daily) recorded 2002 and 2003, reservoir or “fish pool” releases at Bedrock was 8,150 cfs (230.8 cms) in 1973. The were greatly reduced. Annual minimum (mean median annual peak flow for the 32-year Bedrock daily) flows recorded at the Bedrock gage for 1998 gage history was 3,095 cfs (87.7 cms) (Figure A1-4). through 2005 were 21, 32, 25, 24, 1.4, 6.4, 20 and 31 Bankfull flow at Big Gypsum was determined to be cfs respectively (0.6, 0.9, 0.7, 0.7, 0.04, 0.2, 0.6 and near 2,800 cfs (79.3 cms) (Richard and Anderson 0.9 cms) (Figure A1-12). 2006).

METHODS

Fish Collections total-fish estimate combined all species. Recapture probabilities differ by species, which means the Field crews collected fish using a raft fitted with total-fish recapture probability may fluctuate pulsed DC current electrofishing gear and a single between years if there are changes in species forward anode array. The raft was maneuvered by composition. For rare species or in the event of one either oars or a battery powered trolling motor. or no recaptures, we calculated a density estimate by Stunned fish were collected by two netters. Fish dividing the number fish collected by the mean were identified to species. White sucker recapture probability determined for less common (Catostomus commersoni ) and hybrid suckers ( C. c. species. We based the biomass estimates on length- x C. latipinnis and C. c . x C. discobolus ) were weight regressions determined from samples taken identified based on distinguishing characteristics of in each study river. parental species. All fish were measured to the The z-test with an alpha of 0.05 was used to test nearest millimeter for total length. Fish over 150 for differences in density estimates across years at mm were marked and released for recapture data. A each study site. For sites with three or more years of series of fish were weighed to the nearest gram sampling, we used the Bonferroni inequality allowing length-weight relationships to be correction to control the overall significance level determined for each common species in the sample. (.05) for the simultaneous comparison of all pairs of Density estimates with ninety-five percent years (D. Bowden, Colorado State University, pers. confidence intervals were made for all fish species comm.). For example, at Duffy and Sevens, stations based on recapture frequency of fish over 150 mm. with four years of data, the z value (2.631) The Darroch multiple mark method was used to corresponds to an alpha of 0.05. estimate density (Everhart and Youngs 1981). The

II-9 Habitat Quantification We determined the surface area of each of these 16 mesohabitat types for a series of flows using We quantified habitat availability as a function ArcInfo. Solution files (2D model output) were of flow by creating 16 mesohabitat categories (Table imported into ArcInfo and mesohabitat types were II-1). Meso-habitats are usually defined in general based on 1-meter x 1-meter depth and velocity categories as pools, runs, riffles and rapids and is grids. Each mesohabitat’s surface area was based on the mean velocity of the habitat type. Pools determined by summing the number of grids that fell have low velocities, runs are moderate, riffles have into each habitat category (Anderson and Stewart swift currents and rapids are fastest velocities. We 2003 and CD Appendix, Habitat – 2D modeling). defined pool velocities from zero to 0.149 m/sec, Using a mesohabitat’s surface area as a measure of runs from 0.15 to 0.6 m/sec, riffles from 0.61 to 1.5 habitat abundance, we used the Shannon Diversity m/sec and rapids had velocities over 1.5 m/sec. method (Stewart 2000) to indicate habitat diversity Habitat usability is also a function of depth. We as a function of flow. divided pools and runs into five depth categories, riffles into four and rapids into two (Table II-1).

TABLE II-1. Depth and velocity criteria used to define meso-habitat types.

II-10 RESULTS

Fish Analysis mean velocity and sand and gravel substrates. Pool and run habitats were 94 percent of the total area and Duffy, Yampa River riffles only 6 percent at 250 cfs (7.0 cfs) (Table II-5). Native fish (>15 cm TL) species were uncommon During the drought, when flows were less than 150 cfs to rare at Duffy during all years of sampling, (4.3 cms), preferred native sucker habitat was composing only 13 to 16 percent of all fish collected negligible during the first three years ( 1998 -2000) and only seven to nine percent in 2003 and 2004 (Table A2-1, Sevens, Yampa River top panel). Duffy had the lowest prevalence of native The proportion of native fish (>15 cm TL) at fish of any study site (Table II-2). Pure bluehead and Sevens was stable from 1998 to 2001, with a four-year flannelmouth sucker were about 10 percent of the mean of 72 percent, which was relatively high for the catch in 1998 and 1999 (pre-drought) but only five Yampa River (Table II-2). Native fish declined to 42 percent in 2003. Duffy had the highest occurrence of percent in 2003, the third year of the drought and to 29 the nonnative white sucker. White suckers, including percent in 2004. hybrids with flannelmouth and bluehead suckers were Before the drought Sevens had a very high over 70 percent of the catch during the first three years flannelmouth sucker percentage, ranging from 48 to but declined during drought years. The nonnative 53 percent between 1998 and 2001 (Table A2-2, top smallmouth bass ( Micropterus dolomieu ) increased panel). Flannelmouth sucker dropped to 37 percent in during the drought years and became the most 2003 and 26 percent in 2004. Bluehead sucker common species in 2004 at 44 percent. averaged 21 percent of the catch for the first three The total fish density estimates at Duffy were years, indicating that this species was stable prior to significantly (alpha = 0.05) higher before 2002 than the drought. During the subsequent years with after 2002 (Table A2-1, middle panel). Density reduced flows bluehead sucker dropped to 13 percent estimates for flannelmouth sucker, bluehead sucker in 2001 and only two percent in 2003 and 2004. and roundtail chub were highest in 1998 and 1999 White sucker + hybrids were around 15 percent for the (pre-drought) and lowest in 2003 and 2004. White first four years but grew to 36 percent in 2003 and sucker + hybrids, common carp ( Cyprinus carpio ) and 2004. Smallmouth bass were uncommon (about two northern pike ( Esox lucius ) also declined in density percent) before 2002 but rose to 13 percent in 2003 after 2002. Smallmouth bass was the only species to and 23 percent in 2004. increase in density during the dry years of 2001 to Total fish density estimates for Sevens were 2004. significantly (alpha = 0.05) higher in 1998 (179/ha) Total fish biomass was poor at Duffy even prior to and 1999 (179/ha) than in subsequent drought years the drought (Table II-3). Biomass for most species (Table A2-2). Density and biomass for all native fish was lower after 2002 (Table A2-1). White sucker + were lowest in 2003 and 2004. Flannelmouth sucker hybrids comprised about 72 percent of the total fish biomass was about 66 kg/ha in 1998 and 1999 but biomass before 2002 at about 40kg/ha and dropped to dropped to about 24 kg/ha by 2004 (Table II-3). 11 kg/ha in 2004. Bluehead sucker biomass was Bluehead sucker biomass dropped at the greatest rate, negligible in 2004 at only 0.1 kg/ha. from about 16 kg/ha (1998 to 2000) to 5 kg/ha in 2001 Flows and channel geomorphology determine and then 0.6 kg/ha in 2003. Native sucker > 35 cm habitat availability. Typically pre-drought base flows became rarer during the drought years of 2000 to 2003 had been near 250 cfs in the Yampa. The channel (see Appendix Length Frequency, on included data geomorphology at Duffy was generally wide and disk). shallow relative to other sites (Table II-4). Duffy had The density and biomass estimates for white the highest width/depth ratio, a low gradient, slow sucker + hybrids and common carp were highest in

II-11 2003 (Table A2-2), because of exceptionally high Although smallmouth bass estimates and catch rates recruitment of yearling (10 to 20 cm) white sucker and rose in successive years, they were not significant carp. The strong 2002 white sucker and carp year class (alpha = 0.05) due to low recapture rates. suggested relaxed predation that year, since age-1+ The biomass estimate for flannelmouth sucker was white sucker and common carp were not common in very high at Lily Park in 2000, comparable to the other years. Smallmouth bass was the only species to estimates for the Colorado River (Table II-3). exhibit increased density and biomass in 2003 and 2004 Biomass estimates of total fish, flannelmouth sucker (Table A2-2), indicating a positive response to reduced and bluehead sucker declined progressively and were flows and altered thermal conditions (Table II-4). lowest in 2004 (Table A2-3). Channel catfish biomass Sevens had the lowest gradient (0.05 percent) of was highest in 2000 (224 kg/ha) but was 54 kg/ha in all study sites (Table II-4). The river was also wide 2004. In spite of increasing relative abundance and and shallow relative to the higher volume Colorado density estimates for smallmouth bass, biomass and Gunnison rivers. Sevens had a low mean velocity remained fairly constant over the years because of and high width/depth ratio. At 250 cfs only two fewer fish > 29 cm. percent of the site was riffle and rapid habitat; two- Bluehead sucker and flannelmouth sucker mean thirds of the site was in the shallower run habitat types lengths and the proportion of fish > 40 cm were higher (Table II-5). The substrate at Sevens was mostly sand before 2002. In 2000 about 70 percent of and gravel. Field crews observed active bank cutting flannelmouth sucker were > 40 cm and 21 percent and enlargement of sand dune substrates during the were > 45 cm. In 2003 those figures fell to 36 percent survey period. and two percent, respectively. Bluehead sucker > 35 cm were abundant in 2000 at 49 percent but less Lily Park, Yampa River common in 2003 at 29 percent. Native species comprised about 57 percent of the Larger channel catfish were also rarer after 2002. fish assemblage at Lily Park in 2000, the first year of Seventy-two percent of channel catfish were from 30 the drought (Table A2-3). Flannelmouth sucker was to 39 cm in 2000, but only 48 percent in 2001 and 13 the most common species (fish > 15 cm) at Lily Park percent in 2003. In 2000 only one percent of the from 2000 to 2004 at nearly 50 percent (Table II-2). channel catfish fish collected were < 27 cm, but 71 Bluehead sucker was the third most common species percent were < 27 cm in 2003. in 2000, 2001 and 2003 at about eight percent in those Lily Park, located just downstream of Cross years (Table A2-3). Lily Park had by far the highest Mountain Canyon, had a very different channel composition of channel catfish ( Ictalurus punctatus ), geomorphology from that of the two upstream sites. at any site, which varied between 40 and 18 percent. Its gradient was much steeper (0.2 percent versus 0.06 Smallmouth bass had a strong increase in relative and 0.05 percent at Duffy and Sevens), its channel was abundance by 2004. Four species at Lily Park were narrower, its mean velocity was much higher and its less than or near one percent of the catch: roundtail substrate was dominated by cobble (Table II-4). Runs chub, Colorado pikeminnow, northern pike and white were swifter at Lily Park and riffle and rapid habitat sucker + hybrids. types were more common, at 10 percent (Table II-5). Density estimates for total fish, flannelmouth Field crews often observed large mayfly hatches sucker, bluehead sucker and channel catfish were during sampling, suggesting that invertebrate highest in 2000 (Table A2-3) and were significant productivity was also much higher than at the other (alpha = 0.05) for total fish and flannelmouth sucker. two Yampa sites. The 2004 flannelmouth sucker density estimate was significantly lower than the estimates for 2000 and Flow and Degree-Day Correlation 2001. Bluehead sucker estimates were significantly lower for 2004 than for 2000. Channel catfish density Age-1 smallmouth bass in the Yampa River estimates exhibited high variability across years and increased in abundance at Sevens and Lily Park in were significantly different (alpha = 0.05) between 2003 and 2004, years in which temperature units 2000 and 2001 and between 2001 and 2004. (degree-days over 12 C) were elevated the preceding

II-12 years. We found a positive correlation between the hybrids biomass all appeared to have higher number of days flow was below 200 cfs and degree- abundance in the years that followed reduced flow of days-12C and (Anderson 2005). These data (not 2002 and 2003, but the differences were not shown here) indicate that reduced flows not only statistically significant. alter habitats but also thermal regimes. In general, Colorado River base flows are typically 800 to recruitment of age-1 smallmouth bass was less in 1,000 cfs in the irrigation period and near 2,000 cfs years following cooler growing seasons (1998, 1999 in the winter. The gradient at Corn Lake was 0.16 and 2000) and higher when the previous years had percent and the width-depth ratio was relatively low higher degree-days. The highest abundance of age- (Table II-4). The mean width at Corn Lake was less 1 smallmouth bass was in 2004; the preceding year than the mean width of the 15-Mile Reach overall. had the warmest summer water temperatures The mean velocity (0.54 m/s) was similar to Lily (Anderson 2005). Park, but riffle and rapid habitats were 53 percent of the area (Table II-5). Corn Lake, Colorado River Native fish composition at Corn Lake was Clifton, Colorado River somewhat higher (79 percent) under pre-drought Sampling was done at Clifton in only three flow conditions than post- drought conditions (70 years: 2000, 2001 and 2003. Native fish percent) (Table II-2). Flannelmouth sucker and composition at Clifton was similar to that at Corn bluehead sucker had similar percentages of Lake in the two years before the 2002 drought, but in approximately 36 to 38 percent in 1999, the highest 2003 native fish were less prevalent at Clifton, flow year. Bluehead sucker percentage dropped to having dropped from 78 to 64 percent (Table II-2). 28 percent in 2004, while flannelmouth sucker and Flannelmouth sucker and roundtail chub roundtail chub composition was fairly consistent composition was stable over the four year period, but during the study period (Table A2-4). The percent bluehead sucker composition dropped from 41 composition of white sucker, white-hybrids and percent before 2002 to 25 percent in 2003 (Table A2- channel catfish increased during the lower flow 5). At 12 percent Clifton had the highest percentage years of 2003 and 2004. Centrarchid (sunfish) of carp of any site, which may have been more the composition was similar in 1999, 2000, 2001 and result of sewage treatment effluents than habitat 2004, but was highest in 2003, indicating better availability. The percentages of white sucker + recruitment or survival during 2002. hybrids and channel catfish were higher in 2003 than The lower flows of 2002 and 2003 apparently in the earlier years, conforming to the Corn Lake did not affect the density estimates for total fish and trend. Centrarchid composition was low in 2000 and flannelmouth sucker, since their estimates were 2001 and increased in 2003 after the reduced flows similar across the sample period (1999 to 2004). in 2002. However, the density estimates for bluehead sucker Density estimates for flannelmouth sucker and in 2004 was significantly less (alpha = 0.05) than in roundtail chub were similar over the three years, years before 2002 (Table A2-4). Channel catfish indicating stability through 2003 (Table A2-5). density was significantly higher (alpha = 0.05) in Bluehead sucker density was lowest in 2003, 2004 than 1999- 2001. whereas the channel catfish density estimate was Flannelmouth sucker biomass was stable at Corn highest in 2003. Lake throughout the study period (Table A2-4) and One year of reduced flows did not appear to was second highest among the study sites (Table II- affect flannelmouth sucker biomass, but bluehead 3). Bluehead sucker had an apparent decline in sucker biomass appeared reduced (Table A2-5). biomass in 2004, but Corn Lake was still a high Channel catfish biomass was higher in 2003 than in biomass site for this species. Channel catfish the earlier years, indicating that recruitment could biomass was much higher in 2004, indicating have been improved during the reduced-flow period. improved recruitment during the reduced flow Roundtail chub, carp and white sucker + hybrids period. Roundtail chub, carp and white sucker + appeared to have increasing trends in biomass the

II-13 year after reduced flows. Flannelmouth sucker and very high at Escalante relative to Delta, the other bluehead sucker density and biomass (carrying Gunnison site and relative to the Colorado and capacity) were much higher in the Colorado River Yampa River sites. than in the Yampa River (Table II-3). Escalante’s density estimates were lower in 2005 Clifton had a steeper gradient at 0.2 percent (Table than in 2004 (Table A2-6); these differences were II-4) than did Corn Lake and the 15 Mile Reach significant (alpha = 0.05) for flannelmouth and overall. The Colorado River at Clifton was wider than bluehead sucker but not for roundtail chub or white at Corn Lake and along the 15-Mile Reach overall. sucker + hybrids. We suspect that the dramatically This additional width was primarily the result of an old reduced flannelmouth and bluehead sucker estimates diversion structure about midway through the site that in 2005 resulted from different catch and emigration formed a very large upstream pool. Clifton had 22 rates between the two years because of higher flows in percent more surface area per kilometer than Corn 2005. The highly different estimates for flannelmouth Lake, which was comprised of pool habitat. Surface sucker between years indicated either a population area of riffle/rapid habitats per kilometer was similar crash or a strong sampling bias. We believed the for Clifton and Corn Lake, indicating these sites had flannelmouth sucker density was more stable than similar amounts of swift current habitats. However, indicated by the estimates and considered the 2004 surface area of riffle/rapid habitat per hectare was less data more reliable. We believed a more reliable at Clifton, 41 percent compared to Corn Lake (53 estimate of population size for bluehead sucker was percent ) because of a wider channel at Clifton (Table obtained by the average of the 2004 and 2005 II-5). The native sucker density estimates at Clifton estimates. (Anderson and Stewart 2007). and Corn Lake were very similar when based on fish Bluehead sucker and roundtail chub biomass was per kilometer or fish per riffle area. The density high at Escalante relative to sites on other rivers (Table estimates based on fish/ha were lower at Clifton, II-3). Escalante’s flannelmouth sucker biomass was because of its larger surface area. intermediate to those from the Yampa and Colorado rivers sites. The biomass of white sucker + hybrids Escalante, Gunnison River was very high at Escalante. The Gunnison River fish community was The Gunnison River at Escalante was much sampled for the first time in 2003, mark and narrower than the Colorado and Yampa Rivers (Table recapture sampling was done in 2004 and 2005. The II-4). Its moderate gradient (0.09 percent) and narrow Gunnison River did not experience reduced base channel provided a very low width/depth ratio during flows during the 2002 drought because of regulated normal flow conditions. At 39 percent, riffle and rapid flow from upstream reservoirs (Anderson and habitats were less prevalent at Escalante than at Delta, Stewart 2006). but deeper run habitats were much more prevalent at Escalante had the highest native fish Escalante than at Delta or in the Yampa River (Table composition of all the study sites (Table II-2). II-5). The relatively slower and deeper habitats Native species (fish > 15 cm TL) composed 83 appeared conducive for supporting a high proportion percent and 78 percent of the catch in 2004 and of juvenile fish. 2005, respectively. Bluehead sucker was the most common species at about 47 percent - the highest of Delta, Gunnison River all the sites. Flannelmouth sucker was common at Native species (fish > 15 cm) comprised 64 about 18 percent. Escalante had the second highest percent and 74 percent of the catch at Delta in 2004 composition of roundtail chub at 14 percent. White and 2005, ranking the site the third highest behind sucker + hybrids was the only other significant Escalante and Corn Lake (Table II-2). Flannelmouth member of the community. Carp were only three sucker was the most common species at Delta in the percent. Channel catfish, northern pike and 2003 and 2004 surveys at around 40 percent (Table smallmouth bass were not collected during the A2-7). Roundtail chub were about 16 percent, nearly Gunnison River surveys. equal to Escalante. White sucker + hybrids and carp The productivity of juvenile native sucker was were more prevalent at Delta than at Escalante.

II-14 Density estimates for most species varied between 2005. Black bullhead ( Ameiurus melas ) was the most 2004 and 2005, with carp the only species with density common species in 2004 at 45 percent, but was estimates not significantly (alpha = 0.05) different uncommon in prior years. Big Gypsum was the only between years (Table A2-7). The density estimate for site where white sucker were not collected. bluehead sucker increased from 130/ha in 2004 to Density estimates for flannelmouth sucker were 593/ha in 2005. We attribute that increase to very high significantly different (alpha = 0.05) between all four recruitment of age1+ or Age2+ bluehead sucker in years. Roundtail chub estimates were not 2005, as well as differing emigration rates between significantly different (alpha = 0.05) between years, years (more information in Anderson and Stewart except for 2000 and 2004. The number of fish caught 2006). Flannelmouth sucker density decreased in each year (total catch) approximates catch rates per 2005, which we attribute to differing catch and year, since three passes were the standard for each emigration rates between years. The 2005 roundtail year. Field crews caught a total of 580 and 514 chub estimate for Delta was very similar to the flannelmouth suckers in 2001 and 2005, respectively, Escalante estimate. White sucker + hybrid abundance but only 25 in 2004. There were 383 bluehead sucker was higher at Delta than Escalante. in 2001, but only five and four in 2004 and 2005, The 2005 biomass estimate for bluehead sucker of respectively. Black bullhead numbers were 197 in 180 kg/ha (Table A2-7) at Delta was the highest of any 2004, but much fewer in prior years. Crews collected site (Table II-3). In Table II-3 the value for bluehead more roundtail chub, channel catfish and green sunfish sucker biomass at Delta was the average of the 2004 in the first two years than in 2004 and 2005. and 2005 estimates, because it seemed to be more Total fish biomass was very poor at Big Gypsum accurate for indicating habitat potential (carrying (Table II-3). Big Gypsum had the lowest biomass for capacity) at this site. In Table II-3 the value for the flannelmouth sucker and bluehead sucker, but 2004 flannelmouth sucker biomass estimate was used, roundtail chub biomass was higher than the Yampa because the 2005 estimate appeared biased. Delta had River sites. Channel catfish, carp and flannelmouth the highest roundtail chub and white sucker + hybrids sucker biomass were higher before 2002 than biomass of any site. afterward (Table A2-8). Black bullhead biomass was The mean velocity at the Delta site was 0.69 m/s, highest in 2004, the year following the heavily reduced the highest of all the sites (Table II-4). The moderate 2002 and 2003 flows. gradient (0.16%), the narrow channel (42 m) and high Big Gypsum had a moderate slope of 0.15 percent base flows produced a site that was primarily and a low width-depth ratio at flows above normal at comprised of riffle and rapid habitat types (Table II-5). 50 cfs (Table II-4). The mean velocity of the Dolores Delta’s habitat composition was very different from at Big Gypsum was the lowest of the study sites. Pool Escalante’s, suggesting that habitat differences were a habitat composed 63 percent of the site, and riffle and factor for the differing species compositions at these rapid habitats composed only three percent (Table II- two locations on the Gunnison River. 5). The high percentage of pool habitat meant that fish that occupy pool habitat (roundtail chub) were more Big Gypsum, Dolores River prevalent than the species that tend to inhabit runs Native fish prevalence (fish > 15 cm) varied from (flannelmouth sucker) and riffle (bluehead sucker). 87 percent to 43 percent over the years 2000, 2001, 2004 and 2005 (Table A2-8). Native fish prevalence was highest in 2001 and lowest in 2004. Roundtail chub was the most common species in 2000 (55 percent), but was considerably less common (around 25 percent) in 2001 and 2005. Flannelmouth sucker was the most common species in 2001 and 2005 (55 and 59 percent, respectively), but was rare (two percent) in 2004. Bluehead sucker composition ranged from six percent in 2001 to zero percent in

II-15 TABLE II-2. Species composition of fish >15 cm collected from the Yampa, Colorado, Gunnison and Dolores rivers during the pre- and post-2002 (severest drought year) periods.

II-16 TABLE II-3. Biomass estimates (kg/ha) for fish >15 cm collected in the pre- and post-2002 (severest drought year) periods.

II-17 Geomorphic and Habitat Analysis – All Sites Lily Park, on the Yampa, and Clifton, on the Colorado, had similar slopes and channel widths, but Our 2D modeling (GIS) produced two versions their base flows were very different, suggesting that of habitat quantification. First, we quantified the their differences in habitat and fish composition preferred habitats of flannelmouth and bluehead were mainly due to their different flow regimes. sucker for use in the instream flow methodology (see Lily Park’s base flow was similar to the other two the Flow Recommendations section below). Yampa sites, Duffy and Sevens, but Lily Park’s Second, we also quantified the surface area of the 16 higher gradient and narrower width produced more mesohabitats delineated in Table II-1 in order to riffle and rapid habitat per flow volume than at the estimate habitat diversity as a function was flow. other Yampa sites (Table II-4). Habitat availability is primarily a function of The Yampa River had a wider channel than the flows and secondarily a function of channel Colorado and Gunnison Rivers, even though the geomorphology. When base flows are similar, Yampa has less flow volume (Table II-4). The wider differences in habitat availability are due to such channel resulted in higher width-depth ratios and, in geomorphic properties like slope and width. general, lower mean velocities on the Yampa. Determining the relationship between stream flows Typically a wider channel results in a higher and native fish abundance requires the ability to composition of shallow, slow habitats, which are less quantify habitat availability as a function of base suitable for large native sucker than deeper, swifter flows. This was performed by quantifying the habitats. prevalence of 16 mesohabitat types as a function of On the Yampa, Duffy and Sevens were low flow (see Habitat Appendix in the attached CD, gradient, low velocity and high width/depth ratio Anderson and Stewart 2003). sites (Table II-4). They also had the poorest native Mean velocities are directly related to bed slope fish productivity and among the highest incidences or gravitational forces, as indicated in Table II-4. of hybridization with white sucker and abundance of Among sites with comparable flows, those with nonnative predators (Tables II-2 and II-3). These higher bed slopes (Delta, Clifton and Lily Park) had negative interactions with nonnative species became a higher amount of riffle and rapid habitat than did more prevalent during the four years of drought at the lower gradient sites (Table II-5). The Clifton these sites. data is confuscated due to its altered channel width When measured at similar flows, the Gunnison, as a result of the diversion dam; riffle and rapid Colorado and Yampa Rivers have relatively similar habitat area was high at Clifton when measured by habitat composition. At 250 cfs (7.1 cms) riffle and station length, but reduced when measured by rapid habitats were about six to seven percent for station area. Sites with more riffle and rapid Delta and Escalante, about 12 percent at Clifton and availability had higher total fish biomass and native Corn Lake, two percent for Sevens, six percent for sucker biomass (Table II-3). The abundance of Duffy and 10 percent at Lily Park (Habitat appendix bluehead and flannelmouth sucker was strongly in CD). Differences in riffle and rapid availability at correlated with availability of their preferred the same discharge are due to differing slopes and habitats. In this study, availability of preferred widths. habitats appeared to explain about 88 percent of The fact that the Gunnison, Colorado and Yampa bluehead sucker and 83 percent of flannelmouth Rivers have similar habitat composition at similar sucker biomass when fish data was standardized. flows rates strongly suggests that the fish As a further example base flows and slopes were communities in these rivers would be more alike if fairly similar for Corn Lake, on the Colorado and at flows were more alike. Delta, on the Gunnison. The narrower width at Delta We mapped the availability of the 16 mesohabitat produced a higher proportion of riffle and rapid types and calculated the Shannon diversity index of habitats (Table II-5). The higher availability of riffle each one over a range of flows (see the habitats at Delta explained the larger bluehead accompanying CD appendix Haibtat – 2D modeling sucker biomass. Figures II-1 to II-16). Stewart (2000) determined the

II-18 Shannon diversity index is appropriate for indicating diversity (see the CD Habitat appendix). Base flows habitat diversity based on the 16 meso-habitat types near 1,000 cfs are common in the 15-Mile Reach and mapped for each study site. since we can infer that habitat diversity is near- Shannon diversity peaked at 1,000 cfs (28.3 cms) optimal, we can also infer that the native fish at Clifton and at 1,200 cfs (34.0 cms) at Corn Lake, assemblage approaches near-optimal diversity in suggesting that flows near 1,100 cfs (31.2 cms) on these flow conditions. the Colorado River support near-optimal habitat

TABLE II-4. Physical attributes measured at each study site and surface areas for a set of modeled flows.

II-19 Shannon diversity peaked at 600 cfs (17.0 cms) lacks riffle and rapid habitats (Table II-4). Most at Delta and at 800 cfs (22.7 cms) at Escalante, native species can likely tolerate temporary low flow suggesting flows near 700 cfs (19.8 cms) provide conditions by switching to alternative habitats. near-optimal habitat diversity on the Gunnison River Nevertheless, several nonnative species were more (see the CD Habitat appendix). The Gunnison has a competitive at exploiting low flow habitats on the high base flow regime (1,000 to 1,200 cfs). Base Yampa. The low flow regime of the Yampa flow over 1,000 cfs had reduced habitat diversity produced a more depauperate native fish assemblage because of high amounts of riffle and rapid habitats than was observed in the rivers with higher flow (Table II-4). A dominance of riffle and rapid habitats rivers. may be detrimental to the young life stages of the At Big Gypsum Shannon diversity peaked near native fish assemblage and small-bodied fish in 200 cfs (5.7 cms) and fell sharply at flows less than general. 70 cfs (2.0 cms)(CD Habitat Appendix). The Shannon diversity peaked at 900 cfs (25.5 cms) Dolores River flow regime has been very low for at Lily Park (see CD Habitat Appendix), but was not both runoff and base flows. The interaction of both identified at Sevens and Duffy. Base flows on the altered runoff and altered base flows has resulted in Yampa River were about 250 cfs (7.1 cms) in ‘wet’ very low riffle and rapid habitat availability (Table years and less than 100 (2.8 cms) cfs during the II-5). The Dolores River native fish assemblage had drought. The low flow regime of the Yampa River poor diversity and biomass (Tables II-2 and II-3).

II-20 . e t i s y d u t s h c a e r o f w o l f e s a b l a c i p y t e h t t a r e t e m o l i k r e p y t i l i b a l i a v a t a t i b a h o s e M . 5 - I I E L B A T

II-21 DISCUSSION

The 2D modeling effort described by Stewart collected during the study and wild bonytail, ( Gila and Anderson accomplished two objectives: first, el egans ) and humpback chub, ( Gila cypha ), were not determining the habitat preferences of bluehead and collected. The three remaining larger bodied native flannelmouth and correlating biomass estimates for species were the bluehead sucker, flannelmouth the two species with habitat availability; and sucker and roundtail chub. These three species were second, modeling native fish biomass under the primary representatives of the native fish different flow regimes. The findings on flow and assemblage and the focus of our sampling efforts. habitat relationships formed the basis of our Base Flows instream low recommendations for the four study rivers, which we present at the end of this current report. Originally, the four rivers had a natural This report has presented descriptions of native snowmelt hydrograph with high spring runoff flows fish and hydrographs found in the Yampa, Colorado, and a base flow period from August to April. Gunnison and Dolores rivers during and after the Recent hydrographs of all four rivers, however, severe drought of 2002. The pristine native fish show alterations by human activities and unique assemblage arose and was regulated by a natural differences across rivers. The Yampa River retains a flow regime. Flow is a “master variable” strongly relatively natural spring runoff flow but has lower correlated with many critical physical-chemical base flows. The Colorado River in the 15-Mile attributes of rivers, such as water temperature, Reach has moderately reduced spring flows, but channel geomorphology and habitat diversity (Poff retains near natural base flows. The Gunnison has et al. 1997). As we have seen here, changes in the highly reduced spring flows but increased base natural flow regime can have significant impacts flows. The Dolores River’s spring and base flows onecological integrity dependent on how flow are both lower than they were originally. characteristics have changed relative to natural We found base flow volume was the most conditions (Poff et al. 1997). important variable explaining differences in native Natural flow regimes cannot be restored in these fish abundance among the four rivers. The rivers study rivers. Even efforts to mimic the natural flow with higher base flow had higher native fish biomass regimes would have to be integrated into current and composition. Habitat availability for water use programs and may be difficult to achieve flannelmouth and bluehead sucker maximized in the on a consistent basis. Nevertheless, maintenance flow range of 800 to 1,200 cfs for the Gunnison, and enhancement of extant native fish are realistic Colorado and Yampa Rivers and dropped sharply to goals for both fish management and flow very low levels at flows less than 300 cfs. management. The data presented here have Habitat diversity was also related to base flow documented the extant native fish in the four rivers volume. As might be expected, pools and shallow and the modified stream flow regimes that supported runs were the typical habitats available during low and did not support persistence of native fish during flows, whereas deep and fast-current habitat types the first half of this decade. were common when base flows were high. Base Colorado’s original native fish assemblage in the flows for the Colorado River during 2000 and 2001 Colorado River basin at elevations below 6,200 ft were 800 to 1,000 cfs (23 to 28 cms). Habitat comprised about nine species. Speckled dace, availability and habitat diversity for flannelmouth (Rhinichthys osculus) and mottled sculpin, (Cottus and bluehead sucker estimated using the Shannon bairdi), are small-bodied species that were Diversity Index, were almost ideal in this flow incidentally collected but not quantitatively sampled range. The nearly natural base flows of the Colorado as part of this study. The endangered Colorado River suggest that its native fish assemblage pikeminnow and razorback sucker were rarely (flannelmouth sucker, bluehead sucker and roundtail

II-22 chub) may be, of the four rivers, the most The Colorado River in the 15-Mile Reach had representative of an unaltered base flow hydrograph. one year of drastically reduced flows, 2002. Changes to the fish assemblage during the drought Nonetheless, the native fish assemblage changed period provided empirical evidence that base flows are only minimally between 1999 and 2004. We the most important factor maintaining the native fish detected no significant changes in flannelmouth assemblage. The Dolores and Yampa had severely sucker or roundtail chub biomass. The biomass of reduced base flows for four consecutive years. Over bluehead sucker, a species dependent on riffle most of the summer of 2002 flows on the Dolores habitat area, was slightly lower in 2004. Moderate River were less than two cfs (0.06 cms) (Bedrock increases in white sucker, channel catfish and gage) and on the Yampa flows were less than 12 cfs centrachid abundance did occur. These data imply (0.3 cms) (Maybell gage). The Colorado River that one year of reduced flows is not particularly experienced one year of drastically reduced base flow, detrimental to the native fish assemblage. 2002 (Palisade gage). The Gunnison River did not Anderson (2005) has observed, however, experience reduced base flows (Delta and improved reproductive success among nonnative Uncompahgre gages). species (white sucker, channel catfish and The native fish community in the Dolores River smallmouth bass) in the Colorado River following appeared stable before the recent drought period. the low flows in 2002. On the Yampa the severity of Valdez et al. (1992) reported no significant changes in the reductions in native fish was cumulative after species composition in the Dolores River between consecutive years of reduced flows. These data 1990-91 surveys and similar surveys made ten years suggest two or more consecutive years of reduced earlier by the U. S. Fish and Wildlife Service (Valdez flows would have increasingly negative effects on et al. 1982). The authors concluded that the the native fish assemblage. ichthyofaunal community in the Dolores had remained The Gunnison River had the highest base flow relatively stable over that ten-year period. regime and the highest native fish biomass. Because The Dolores River fish community at the Big the Gunnison River did not experience low base Gypsum site changed dramatically between 2000 and flows during the 2000 to 2004 drought period and 2004. Among the changes were drastic declines in since the Gunnison had high native fish biomass and bluehead sucker abundance; increased abundance of composition in 2003, 2004 and 2005, we assumed black bullhead, a species associated with stagnant pool that this river appeared to be a base flow control site. habitats; reduced biomass of roundtail chub, channel Nevertheless, the Gunnison native fish catfish and carp; and a reduction in the frequency of assemblage measured in this study differed from native fish larger than 20 cm (see the CD appendix 1992 and 1993 data (Burdick 1995). White sucker Fish- Histograms). and white sucker hybrids were 9.6 percent (n = The Yampa River fish assemblage also changed 11,148) of all fish collected (boat electrofishing) in dramatically in 2003 and 2004 relative to pre-drought Reach 5 (Delta site) in 1992-93. In contrast, white conditions in 1998 and 1999. Notable changes sucker + hybrids made up 23.3 percent (n = 6,717) included very low occurrences of speckled dace, of the catch at Delta in 2004-05. White sucker + mottled sculpin and bluehead sucker (Anderson 2005); hybrids composition in Reach 4 (Escalante site) was reduced total fish biomass, both native and nonnative; 2.6 percent (n = 5,463) in 1992-93 and 14.8 percent increased white sucker hybridization with native (n = 8,612) in 2004-05. Flannelmouth sucker were sucker; and increased crayfish abundance that was less prevalent at Escalante and bluehead sucker were detrimental to native invertebrate species (P. Martinez, less prevalent at Delta in our surveys than in CDOW researcher, pers. comm.). Low flows during Burdick’s (1995 and see our CD appendix Fish the drought years appeared to explain the explosion in Gunnison). smallmouth bass abundance and an expanding Burdick (1995) compared his data to those of smallmouth bass population appears more problematic Valdez et al. (1982) and concluded that the native for reestablishing the native fish assemblage than fish assemblage had not changed over the 12-year reduced base flows (Anderson 2005). period from 1981 to 1993. The higher incidence of

II-23 white sucker and white sucker hybrids in our survey strength in 2001 and 2002 would explain the lower appears to be a fairly recent change in the Gunnison bluehead sucker biomass in 2004 at Delta. Spring River fish community. flows were higher in 2003 and a strong 2003 year class could have recruited and increased bluehead Spring Runoff Flows sucker biomass in 2005. The relative importance of high spring flows in Water authorities often ask for spring flow maintaining native fish integrity was minor in the recommendations when requesting instream flow Yampa River. The Yampa had relatively high spring recommendations. The primary goal of spring flow flows in 2000, 2001 and 2003. Any apparent recommendations is to maintain channel increase in native fish reproductive success in those geomorphology with flushing flows. Bankfull flow years was negated by poor YOY survival related to is typically identified because it is the most efficient the severely reduced summer flows. flow for transporting sediment and maintaining the Spring flows in the Dolores River below McPhee channel. Spring flows are the mechanism for dam were heavily reduced during the drought, with flushing fines, sorting bed materials and scouring negative consequences to the native fish assemblage. pools- all important processes for maintaining fish Annual hydrographs of the Dolores before the habitats during the base flow period. Moreover, McPhee Dam showed a high frequency of flows several native fish species spawn during spring capable of transporting bed material (3,000 to 5,000 flows. Maintaining spring water temperatures and cfs) but very low annual base flows (two to five cfs). spawning habitats are therefore also considerations In the pre-McPhee period flows scoured riffles and in making spring flow recommendations. pools nearly every year, which maintained the Spring flows appeared to be related to native fish habitats used by native fish during the irrigation reproductive success and recruitment in this study. season. The pre-McPhee conditions appeared more There is evidence that native species recruitment conducive to maintaining native fish than those improves following moderate to high spring flows in observed in recent years (Anderson and Stewart the Gunnison River (Anderson and Stewart 2006). 2006). The lack of flushing flows for four Recruitment of nonnative species increased on the consecutive years allowed sediment to accumulate in Colorado River following the low spring flow years. riffles and pools, resulting in a net loss of habitat The importance of spring flows for maintaining a quality that was not replaced by higher base flow native fish assemblage appeared minor however, releases from the reservoir. compared to base flows. More data should be Pitlick (1999) reported the Colorado River collected to determine relationships between spring channel had narrowed slightly and was more or less flows to native fish recruitment. Efforts to sample in equilibrium with the present flow and sediment young of the year (YOY) in the fall would help transport regimes. That study also reported that determine correlations between spring runoff flows channel narrowing on the Gunnison River has been and year class strength. minor. A slightly downsized channel may be the Spring runoff flows on the Gunnison were low in natural consequence of reduced spring runoff flows. 2000, 2001, 2002 and 2004. Mean daily peak flow Despite higher annual flow volumes, both the was only 1,464 cfs (41.4 cms) in 2002 and it was Gunnison and Colorado had narrower widths than 2,769 cfs (78.4 cms) in 2004. Reduced spring flows the Yampa. have been associated with reduced reproductive Spring flows on the Yampa have been the most success of native species and improved reproductive natural of all four rivers, which suggests that success of nonnative species (Burdick 1995). The sediment transport and channel maintenance has increase in white sucker may have occurred during taken place fairly regularly in recent years. this recent period of reduced runoff flows (Anderson Geomorphic properties of the upper Yampa, and Stewart 2006). Bluehead sucker composition however, were the least conducive to maintaining and abundance were lower in 2004 than in 2005 and fish habitat during periods of low flows. The Yampa 1992-1993. Low recruitment due to poor year class River channel was wide and unstable, probably

II-24 because of local land use practices. Active bank Colorado pikeminnow in Dinosaur Canyon. cutting was evident at the Sevens site and field crews Recovering endangered species has been a higher observed that sand dune substrates enlarged at priority for native fish management and maximizing Sevens and upstream of Cross Mountain during the pikeminnow spawning potential is an important study period. The benefits of high spring flows for recovery objective. Nevertheless, the lesson of the channel maintenance and native species recruitment Yampa River is that a holistic approach for the native may have been negated by active bank erosion along fish assemblage would be to prioritize adequate base sections of the Yampa River. flows over high spring flows. Bed slope, stream width and the width/depth ratios of typical base flows were the geomorphic Nonnative Species properties associated with maintenance of adult native fish habitat during the base flow periods Competition, predation and hybridization by studied. These geomorphic variables should be nonnative species are other factors that can potentially identified in future instream flow investigations. affect native fish biomass and diversity independently Determination of mean depths and mean velocities of habitat availability. As a rule, most of the of representative riffles would also be informative. successful nonnative species are detrimental and are Riffles need to have a depth that allows for migration well adapted to survive drought and inhabit low- to of large fish. Adequate depths and velocities of riffle mild- velocity habitats. Species native to the Colorado habitats are also important in maintaining both fish River Basin have adapted to regular and often extreme and invertebrate habitats. Adult bluehead sucker flooding events each spring. Many tend to spawn preferred deep and fast riffles. These mesohabitats during the runoff period and all except the chub were plentiful on the Colorado and Gunnison rivers species utilize moderate- to swift- velocity habitats. but were negligible on the Yampa (< 200 cfs) and The site with the largest nonnative fish component Dolores (< 60 cfs). Poor invertebrate productivity was Duffy on the Yampa River. Even in 1998 and due to inadequate riffle quality and quantity is likely 1999 pure native sucker were rare at this site because the most direct cause of the poor total fish of previous hybridization with white sucker. In fact, productivity on the Yampa and Dolores. native sucker were so rare at Duffy that the site was An overly wide channel is counter productive for unsuitable for studying native fish habitat use maintenance of adult native fish habitats during low (Anderson and Stewart 2003). Moreover, Duffy and flow periods. Discharge (Q) equals width x mean the river reach from Maybell to Hayden had a very depth x mean velocity. Fish select habitats based on large population of nonnative predators. Impacts of depths and velocities. A narrower channel means northern pike predation included a near elimination of that either depths or velocities are increased – both all forage-sized (12 to 40 cm) fish from the site (see the are positive attributes for adult native fish habitat. A CD, appendix Fish-histograms). In the Yampa the relatively wider river will require relatively higher nonnative white sucker may have gained superiority flows to attain equivalent habitat availability. The over the native sucker because of better predation Yampa River was wider than the Colorado and avoidance adaptations instead of more efficient Gunnison, but had relatively lower base flows. The exploitation of available habitats. 2D modeling indicated that 400 cfs (11.3 cms) The negative impacts of nonnative species would be needed at Sevens and Duffy (Yampa interactions at Duffy have superseded the impacts of River) to yield a similar amount of bluehead sucker flow regimes in configuring the native fish habitat as the 250 cfs (7.1 cms) at Corn Lake on the assemblage. Once predation has reduced the native Colorado River yields (Anderson and Stewart 2003). fish assemblage, little additional pressure is necessary Maintaining the relatively natural runoff flows in to keep natives suppressed. A return to pre-drought the Yampa River appears to be more critical in the flows is therefore not likely to restore the native fish lower Yampa. Bestgen et al. (1998) reported that assemblage of the upper Yampa without concurrent moderate to high runoff flows played an important reductions in predator impacts. Nevertheless, the flow role in the reproductive success for endangered management at Duffy should not be ignored because

II-25 of these nonnative fish pressures. Instream flows are species. Low-head weirs appear to have potential critical for maintaining habitat diversity. Habitat for controlling undesirable nonnative species in diversity is simplified at flows less than 300 cfs. other rivers. Low flows and habitat simplification could mean that the hybridization and predation impacts were Species Accounts unavoidable. Unlike Duffy, the Sevens site on the Yampa had Bluehead Sucker a high native fish component prior to the drought Data on bluehead sucker provide the most period. Anderson (2005) speculated that information for justifying instream flows needed to flannelmouth sucker abundance was high and stable maintain the native fish assemblage. The bluehead at Sevens because of recruitment of fish from sucker is a relatively large-bodied fish with a strong spawning sites near Cross Mountain Canyon or the association for “medi-riffle” habitat (Anderson and Little Snake River confluence. Flannelmouth sucker Stewart 2003, Byers et al. 2001 and Rees and Miller abundance did not decrease until the third 2001; see CD Reports appendix). Sampling for this consecutive year of reduced flows. White sucker study confirmed a strong relationship between and smallmouth bass numbers also increased in the increased medi-riffle habitat and increased bluehead third year of the drought. Sevens appears to be an sucker biomass. example of reduced flows precipitating an altered Bluehead sucker biomass was highest on the native fish assemblage. Gunnison River. The 15-Mile Reach and near White sucker may be better than native sucker at Parachute on the Colorado River (Osmundson 1999, avoiding predation by smallmouth bass. If Anderson 1997) were other sites with high bluehead smallmouth bass persist in the Yampa, white sucker sucker biomass, high base flows and high and hybrids may completely replace native sucker availability of riffle habitat. Bluehead sucker upstream of Maybell because of these differences in biomass was extremely low at Duffy and Sevens predatory impacts. During a period of higher flows (Yampa River) and Big Gypsum (Dolores River) in the late 1990s, smallmouth bass were uncommon where optimal habitat availability was lacking. in the Yampa and were not expected to be a threat The primary objective of most cross-section given flows and thermal conditions of that period methodologies is to maintain the quality of riffles (Nesler 1995). During the 2000-2005 flow regimes, (Nehring 1979). Riffles are the most vulnerable however, smallmouth bass proliferated throughout habitat to dewatering, yet they are the most the river and their dominance will likely continue as important habitat for invertebrate productivity. long as base flows and temperatures are suitable for When riffle habitats are maintained, there should be their reproduction. sufficient habitats for perpetuating carrying capacity Channel catfish and carp were other nonnative (biomass) and composition for all members of the species commonly collected in the four rivers we native fish assemblage (Nehring 1979). studied. Neither species appeared to be a serious Bluehead sucker abundance is a very good threat to flannelmouth sucker, bluehead sucker and indicator of riffle habitat availability. Riffle habitat roundtail chub populations. Not enough data were availability appeared nearly optimal on the Colorado available, however, to indicate wheather channel River. Bluehead sucker were about 36 percent of all catfish and carp have been a threat for endangered fish collected and about 17 percent of the total species, such as Colorado pikeminnow and biomass at Corn Lake on the Colorado. At Lily Park razorback sucker. on the Yampa, where base flows were lower, Field crews found no nonnative predators in the bluehead sucker were only nine percent of the catch Gunnison River during the study period. Burdick and about seven percent of the biomass. These (1995) attributed the lack of channel catfish to the differences between the two rivers were most likely Redlands Diversion dam, a physical block to their due to differences in base flow volume, since slopes upstream migration. Redlands Dam has been an and widths were similar at these sites. An instream efficient method for controlling certain nonnative flow for the 15-Mile Reach (Colorado River) and

II-26 Lily Park (Yampa River) would be the same when correlations between habitat area and flannelmouth based on a standardized methodology that relied on sucker biomass. riffle habitat availability. Flow recommendations for Flannelmouth sucker was the most common these two rivers should differ only if native fish native fish collected during this study. It was the management objectives or water availability are most common species at the Lily Park, Corn Lake, notably different. Clifton, Delta, Sevens and Escalante sites, The reproductive success of bluehead sucker may composing about 47, 41, 38, 41, 40 and 31 percent of be related to spring runoff flows. The senior author total biomass at each site, respectively. The relatively observed spawning bluehead sucker in mid May near high abundance of flannelmouth sucker resulted from the mouth of Dominguez Creek. In low flow years the high availability of preferred habitat. In fact, the ova and larvae survival may be poor because of more only sites where flannelmouth sucker was not the efficient predation. Spring flows were higher in 2003 most common species were those in which their on the Gunnison River than in the other years studied optimal habitat had been reduced because of either and there appeared to be a strong 2003 year class of very low or very high base flows. In the low flow bluehead sucker (Anderson and Stewart 2006, CD). rivers (Dolores & Yampa < 200 cfs), pools and Bluehead sucker biomass on the Colorado and shallow runs were the dominant habitat and Yampa rivers was relatively stable in years of normal flannelmouth habitat was lacking. At high base spring runoff and base flows, but bluehead sucker flows, the Gunnison >1200 and the Colorado >2000, biomass dropped after years of reduced runoff flows. riffles and rapids were dominate habitats and Hybridization with white sucker is a potential flannelmouth habitat is reduced. problem for bluehead sucker. The Yampa had the Flannelmouth sucker occupy ‘deep semi-swift most extreme problem of white sucker hybridization. runs’, which is typically the most plentiful main During the drought medi-riffle habitat was lacking on channel habitat (Anderson and Stewart 2003-CD, the Yampa and pure bluehead sucker became rare. Byers et al. 2001-CD, Rees and Miller 2001-CD). Bluehead sucker - white sucker hybridization may Flannelmouth sucker should be the highest biomass prohibit recovery of pure bluehead sucker in the species in the community because of its large body upper Yampa River even if riffle habitat can be size (up to 60 cm). The species also inhabits the most restored. The Gunnison River had the next highest prevalent habitat type in normal flow conditions. hybridization rates; but medi-riffle habitat was Flannelmouth sucker were present over a wide abundant on the Gunnison and pure bluehead sucker range of base flows. A general observation was that biomass was very high. The Colorado River had higher flow rivers had a higher compliment of larger high bluehead sucker biomass and only a small white sized fish. Flannelmouth sucker over 45 cm were sucker and hybridization component in the 15-Mile common in the Gunnison and Colorado Rivers. In Reach. The Dolores had the lowest bluehead sucker the low-flow Dolores River, flannelmouth sucker population, but this river did not have white sucker. were typically less than 20 cm. Flannelmouth sucker Bluehead sucker should be the primary indicator size and discharge were intermediate at Lily Park on species for biologically-based instream flow the Yampa River. Achieving a large size allows recommendations. Bluehead sucker abundance is individuals to exploit swifter habitats and also avoid directly related to the availability and quality of riffle predation. habitats. Riffles are important habitats for speckled Flannelmouth sucker appear capable of tolerating dace and aquatic invertebrates. Optimal bluehead periods of reduced flows better than bluehead sucker sucker habitat availability occurred when riffles were because of their different habitat requirements. about 30 to 50 percent of the surface area. Possibly due to a higher degree of overlapping habitat and spawning periods, flannelmouth sucker Flannelmouth Sucker appeared more susceptible to hybridization with Flannelmouth sucker is another good indicator white sucker (WS x FMS). species for flow and habitat relationships. Stewart Hybridization with white sucker is a very serious and Anderson (2003) identified significant threat to flannelmouth sucker on the Yampa River.

II-27 Low base flows, habitat overlap and predation increase proportionately with decreasing flow due to pressure are selection forces for high hybridization reduced velocities, so pool habitat availability tends rates there. Flannelmouth x white sucker hybrids are to maximize at lower discharges. The fact that common on the Gunnison River as well, where base roundtail chub are a predator species means their flows have been very high. Reduced peak flows may biomass potential is more likely dependent on prey be associated with the increased white sucker availability (which is riffle associated) than on pool abundance on the Gunnison. White sucker numbers habitat availability. Attempts to correlate roundtail were low in the 15-Mile Reach on the Colorado River chub biomass to pool area did not result in and the species was absent in the Dolores River. The significant correlations (Anderson and Stewart Dolores River is an example of a persisting 2003). flannelmouth sucker population under conditions of In general roundtail chub were highly resistant to long term flow reductions, but a threat of white flow reductions, but they were vulnerable to sucker hybridization would likely imperil this species predation. Roundtail chub biomass was highest in under these flow conditions. the Gunnison River, probably because this river did Flannelmouth sucker is an important species for not have nonnative predators and because its high biologically justifying instream flow riffle content made it biologically productive. recommendations to maintain the native fish Among the four rivers studied, the Gunnison was assemblage. Flannelmouth sucker is the most unique in its lack of nonnative predators, which common native fish. An adequate flow regime would make it a very good location for testing relationships support an abundant flannelmouth sucker population of habitat and biomass. that includes both juveniles and large fish (>45 cm). The river with the highest percentage An abundant flannelmouth sucker population with a composition of roundtail chub was the Dolores. The normal size distribution is a very good indication that Dolores was the lowest flow river and had a very habitat diversity is being maintained for all members high relative availability of shallow pool habitats. of the native fish assemblage. Roundtail chub in the Dolores were small (10-20 cm), indicating a relationship between small body Roundtail Chub size and the quality of available habitats. Roundtail Roundtail chub abundance was not a good chub achieved a total length up to 45 cm in the high indicator for justifying instream flows. Roundtail volume Gunnison and Colorado Rivers. Reduced chub are a multi-habitat species. During the day body size and decreased age of maturity are apparent they occupied deep pools and at night they adaptations to long-term low flow conditions. frequented runs and riffles, presumably for feeding Roundtail chub abundance during the drought (Byers et al. 2001 and Rees and Miller 2001; see the was more related to predation impacts than habitat CD appendix). Roundtail chub appeared more availability. Roundtail chub were uncommon at concentrated in scour or eddy pools in areas with both Duffy and Sevens on the Yampa River. higher channel complexity. Boat electrofishing is Predation impacts likely increase during drought not as efficient in deep pools and roundtail chub periods. In times of low flows, predator avoidance numbers could have been underestimated in some would be more difficult because prey and predators surveys. Turbidity, which decreases light are confined to the remaining habitats. penetration during the day, also appeared to Roundtail chub were very rare at Lily Park on influence the catch rates of roundtail chub. In some the Yampa even though suitable habitat was surveys, however, the roundtail chub catch rates available. The lack of roundtail chub at Lily Park improved during turbid conditions. indicates no local spawning and a complete lack of Roundtail chub were not a good indicator species larvae drifting down from upstream spawning sites. for instream flow needs for two reasons: their habitat Channel catfish were extremely abundant at Lily preferences and their status as predators. Roundtail Park, however and may have been responsible for chub prefer deep pools habitats, which are the least the lack of roundtail chub recruitment. There was vulnerable habitats to dewatering. Pool habitats evidence of a catfish annual migration through Lily

II-28 Park (Anderson 2004) and therefore the Lily Park caused by factors other than lack of adult fish habitat catfish population could have been even larger than on the Colorado 15-Mile Reach. estimated. Rose and Hann (1989) believed the lack of base Roundtail chub were relatively common in the flow habitats were likely not the primary limiting Colorado River. The species’ percent composition factor for Colorado pikeminnow in the Colorado and biomass in the 15-Mile Reach were similar to River. As they explained “Formulating stream flow upstream at Parachute (Anderson 1997). The 15- recommendations for adults is meaningless if channel Mile Reach had a large channel catfish population, stability, spawning, larval transport, exotic competition whereas catfish were rare at Parachute. This and/or other factors are limiting to the population .” suggests channel catfish were not a negative impact Nevertheless, formulating holistic instream flow for roundtail chub in the 15-Mile Reach. recommendations with the goal of sustaining the native Roundtail chub abundance and habitat was not fish assemblage is an ecological imperative. very informative in making recommendations for Using radio telemetry, Modde et al. (1999) and instream flows. Roundtail chub should be a Osmundson et al. (1995) described Colorado common species at almost any range of flows. If pikeminnow habitat use in the Colorado and Yampa juvenile fish are abundant but adult fish are missing Rivers. With a modified inflection point (as in the Dolores River), adult habitat is methodology, Modde et al. (1999) identified 93 cfs insufficient. On the other hand, if large adults are (2.6 cms) as the minimum instream flow for the present but juveniles are missing (as at Duffy and Yampa River. Osmundson et al. (1995), using a Sevens on the Yampa), predation is likely the cause. videography method, recommended an instream If no roundtail chub are present (as at Lily Park on flow of 1,243 cfs (35.2 cms) for the 15-Mile Reach the Yampa), lack of recruitment is the suspected of the Colorado. Although both recommendations cause. The presence of all sizes of roundtail chub (5 were intended to protect adult Colorado pikeminnow to 45 cm) is good evidence that habitat diversity is habitat, they demonstrate that different adequate and predation by nonnative fish is minimal methodologies may yield widely different results. (as in the Colorado and Gunnison Rivers). The traditional approach for biologically justifying instream flow recommendations is to use Endangered Species an analysis methodology that integrates the needs of ..[A] holistic approach to endangered fish the native fish assemblage. Holistic instream flows management is essential and the species (or should be compatible with endangered fish recovery life stages) selected for study should reflect goals. Adult Colorado pikeminnow habitat in the the environmental constraints on the Colorado, Gunnison and Yampa River does not population as a whole. –Rose and Hann appear to be limited nor does it appear appropriate (1989) to use pool habitat to justify an instream flow. Recommendations for spawning flows, larval drift Colorado Pikeminnow flows and wet year flows are outside the normal Colorado pikeminnow is not a suitable species definition of an instream flow (base flow period) for determining instream flows for the same reasons and recommendations that address specific given for roundtail chub. During the day Colorado biological concerns need to be based on an pikeminnow primarily frequent deep pools (eddies). identified limiting factor. Again, pools are the least vulnerable habitats to The other three endangered species in the upper dewatering. At night Colorado pikeminnow are Colorado River Basin - razorback sucker, predators and occupy multiple habitats (Modde et al. humpback chub and bonytail - are also not good 1999). Correlations between habitat and Colorado indicator species for recommending instream flow. pikeminnow biomass would not be possible to These species are currently very rare and making determine, because appropriate data are rarely base flow recommendations for maintaining their collected. The very low abundance of Colorado adult habitats will not likely correct factors that are pikeminnow and other endangered fish is likely currently limiting their abundance.

II-29 White Sucker Nonnative Predators White sucker is a species that is detrimental to the native fish assemblage. Identification of white Northern Pike sucker habitat preference was not an objective of this Northern pike have been able to maintain a large project, but in general, pure white sucker were population in the Yampa River over a wide range of collected primarily from low velocity habitats such flow conditions. The species appeared to prefer as pools, backwaters and slow deep runs. White main channel pools during the base flow and winter sucker x flannelmouth hybrids achieved a large body periods (Nesler 1995). The preference for pool size (40-50 cm) and occupied faster run habitats than habitats would explain how northern pike have been pure white sucker. White sucker x bluehead sucker able to persist through the prolonged drought period hybrids achieved a large body size and were on the Yampa. common in swift habitats. Both hybrids were able to Nesler (1995) found no relationship between inhabit areas used by the native parent. White northern pike abundance and spring runoff sucker and hybrids were able to occupy a wide range magnitude. Suitable spawning habitat is very of pool and run habitat types. abundant in the Yampa River upstream of Craig. We did observe an apparent longitudinal trend in The large northern pike population is maintained by white sucker composition, with higher relative recruitment of fish from off-channel ponds, sloughs abundance at upstream sites. On the Colorado River and upstream reservoirs (Nesler 1995). In the white sucker and hybrids were 22 percent of the total Colorado, Gunnison and Dolores rivers lack of catch at Rifle, 12 percent at Parachute and six suitable spawning habitat appears to explain the low percent in the 15-Mile Reach, decreasing in a numbers of northern pike collected. downstream direction. On the Yampa River white sucker and hybrids were 72 percent at Duffy, 13 Smallmouth Bass percent at Sevens and less than one percent at Lily Smallmouth bass was the only species with a Park and white sucker are rare further downstream in strong positive change in abundance during the Dinosaur Canyon (Mark Fuller U.S. Fish and drought on the Yampa River. Nesler (1995) Wildlife Service, personal comm.). On the described suitable smallmouth bass habitat as low- Gunnison River white sucker were 28 percent at current velocity habitat with instream or shoreline Austin, 22 percent at Delta and 15 percent at cover and limited boulder-gravel substrates Escalante. A longitudinal distribution suggests that inundated year-round for cover and feeding. These cooler temperatures tend to benefit this species. We habitats are plentiful in the Little Yampa Canyon also observed, however, that white sucker and reach and are not diminished by drought conditions. hybrids increased in prevalence at Sevens (Yampa Recruitment of age-1+ smallmouth bass appeared River), the Gunnison River and the 15-Mile Reach negatively related to the magnitude of spring runoff during the drought period, a time of reduced runoff flows on the Yampa. Large numbers of YOY but flows and presumably warmer water temperatures. small numbers of age1+ smallmouth bass were White sucker and hybrid recruitment appeared to collected at Duffy on the Yampa in 1998, 1999 and increase relative to native sucker following the low 2000. This finding indicated poor YOY overwinter spring runoff flow years in the Gunnison and survival. During the drought age-1+ recruitment Colorado Rivers. Perhaps native sucker recruitment increased. Drought years tend to have higher thermal was poor in low runoff years. White sucker and units and a longer growing season that results in hybrids have been able to persist under regimes of bigger YOY smallmouth bass. Bigger YOY have both low and high base flows. The white sucker and better overwinter survival and much higher hybrids are species (taxa) well adapted to thrive over recruitment rates after their first winter season. A a wide variety of flow conditions including both the warmer and lengthier growing season in the other ‘dry’ and ‘wet’ hydrographs. White sucker has western Colorado rivers might also result in greater demonstrated it can persist in western Colorado recruitment of smallmouth bass in those rivers . rivers regardless of flows or flow alterations.

II-30 Channel Catfish Supplemental Data Channel catfish had high density populations in Lily Park (Yampa River) and in the Colorado River. Colorado River Nesler (1995) summarized the preferred habitat of A common assumption concerning stream flow is adult channel catfish from several authors. Channel that habitat availability increases with increasing catfish preferred deep pool habitat (Aadland et al flows. Our 2D modeling effort projected that habitat 1991) and usually select slow current velocities and availability for flannelmouth and bluehead sucker preferred cover (Peters et al. 1992). Sampling would rapidly increase with increasing flow from zero during this study took channel catfish from to about 600 cfs in the 15-Mile Reach of the Colorado backwaters and deep pools, but collected the River. Flannelmouth and bluehead sucker habitat majority from run habitats with gentle to moderate would be near maximum at flows of 800 to 1,200 cfs, velocities. As many as 20 to 60 channel catfish were but habitat would not increase substantially at flows collected from certain run habitats (usually < 1m above 1,400 cfs. The shape of the 15-Mile Reach depth and < 0.5 m/s) in Lily Park and 15-Mile Reach (Figure II-18) habitat-flow relationship was also surveys. Collection of large numbers of channel similar to the Gunnison and Yampa Rivers (Figures II- catfish in certain glides suggests that the catfish (30- 19 and II-20). 50 cm) may have been foraging in schools. We suggest that the 2D modeling habitat-flow The high-density channel catfish population at relationship that project bluehead and flannelmouth Lily Park (Yampa River) was maintained by sucker habitat does not substantially increase with recruitment of juveniles from downstream (Nesler flows over 1,400 cfs, can be confirmed from field 1995). We observed the same process at work in the observations in different sections of the Colorado current project at Lily Park and the 15-Mile Reach River. (Colorado River). In general, adult catfish migrate The 15-Mile Reach is located below 2 major downstream to spawn (temperature related). The diversion structures that reduce flows during the April smallest channel catfish collected at Lily Park and to November irrigation season. The upper diversion the 15-Mile Reach prior to 2002 was about 30 cm. structure is called the Government Highline Diversion In 2003 YOY and channel catfish less than 20 cm Dam or simply known as the Roller Dam and is were collected at Lily Park, which indicated located in Debeque Canyon about 20 km upstream of spawning sites were further upstream during low the 15-Mile Reach. flow years (Anderson 2004-CD). Spawning sites Pitlick and Cress (2000) found that the Colorado during normal base flow conditions were likely River near Una Bridge and the town of Parachute had further downstream because water temperatures take similar bed slopes and bankfull widths to those of the longer to warm in high flow years. 15-Mile Reach. Similar channel geomorphology for Downstream migration to spawning sites makes the 15-Mile Reach and Parachute indicates the main channel catfish vulnerable to removal by trapping difference in habitat potential for these two areas is efforts. The Gunnison River did not have channel their respective base flow regimes. catfish or other nonnative predator species. As noted The Colorado River upstream of the roller dam earlier, the Redlands Diversion Dam appears to be has had a much higher base flow regime than the 15- an effective block to upstream migration. Anderson Mile Reach. Flows recorded at the Cameo gage on (1997) did not collect channel catfish upstream of the Colorado River were rarely less than 1,800 cfs the Government Highline Diversion Dam on the between 1993 and 1996. Flows recorded at the Colorado River. Again, strategically located low- Palisade gage were about 1,200 cfs less than those at head diversion dams or weirs may offer an effective the Cameo gage during the irrigation season. technique for controlling or reducing channel catfish Anderson (1997) sampled the Colorado River from upstream reaches. near Parachute for fish in 1994, 1995 and 1996 using similar electrofishing and population estimation techniques as those used for the 15-Mile Reach in this report (1999 to 2004). The biomass estimates for

II-31 Parachute for flannelmouth sucker and bluehead of the 15-Mile Reach. Nevertheless, habitat sucker averaged 187 and 94 kg/ha, respectively. availability for native sucker was modeled to be Flannelmouth sucker and bluehead sucker biomass in higher at flows typically associated with the 15-Mile the 15-Mile Reach averaged 235 and 100 kg/ha, Reach. respectively. These data indicate a similar flannelmouth and bluehead sucker biomass for Yampa River Parachute and the 15-Mile Reach. Native fish have not persisted in the upper Yampa The composition of flannelmouth and bluehead River except at Lily Park. Lily Park was a very short sucker at Parachute was 38 percent and 36 percent of river section located between the confluence of the all fish, respectively (Anderson 1997). In the 15-Mile Little Snake River and Cross Mountain Canyon. The Reach (this study) flannelmouth sucker were 36 slope was steeper and the river narrower than at the percent and bluehead sucker, 38 percent. These data other two Yampa sites and thus Lily Park was not indicate that the Parachute reach has a native sucker representative of the river upstream of Cross and roundtail chub population size that is nearly Mountain. equivalent to that in the 15-Mile Reach. Similar Lily Park was the only Yampa site sampled where native fish biomass at Parachute and the 15-Mile pure flannelmouth sucker and bluehead sucker were Reach suggests that habitat availability was not common, because of the lack of white sucker at Lily different between the two locations. Similar native Park. White sucker have been only rarely collected in sucker biomass at Parachute and the 15-Mile Reach Dinosaur Canyon and therefore hybridization is not a confirmed that flows above 1,800 cfs did not increase problem in the lower Yampa River (Mark Fuller bluehead sucker or flannelmouth biomass in the USFWS Vernal, personal comm.) Colorado River. Perhaps one of the more insidious impacts of the These field observations confirmed our 2D model drought is a recent explosion in the crayfish projections that stream flows above 1,400 cfs would community in the Duffy site by 2003 (Anderson not result in greater availability of habitat for 2005). Martinez (2006) estimated that the current flannelmouth and bluehead sucker in the Colorado biomass and production of crayfish exceeds the fish River. The geomorphic and fish data also indicate that population. The nonnative virile crayfish ( Orconectes habitat modeling done in the 15-Mile Reach is also virilis ), appears to have increased during the recent applic able to the Colorado River upstream of the drought. This change has implications for fish Roller Dam. productivity. Although fish species that prey on Fish population surveys by the U. S. Fish and crayfish (such as smallmouth bass and channel Wildlife Service (USFWS) in the 15-Mile Reach and catfish) may benefit (Martinez 2006), suckers and the river upstream of the Roller Dam conform to chub may be in competition with crayfish for those reported by Anderson (1997). Osmundson invertebrate forage, which could explain concurrent (1999) reported higher flannelmouth and bluehead reductions in sucker biomass. sucker catch rates in the 15-Mile Reach than at Parachute. The USFWS data indicate that the 15- Gunnison River Mile Reach is more productive for native fish than Native fish have persisted in the reduced spring upstream of the Roller Dam, where base flows were flow and increased base flow regime of the higher. Gunnison River. High base flow regimes were not Two factors could be an influence for greater detrimental to native species and did not inhibit native fish biomass in the 15-Mile reach than in the white sucker reproduction and recruitment. The upstream reaches. First, the 15-Mile Reach is cooler water temperature associated with high base located near the confluence of the Gunnison River flow regimes may be beneficial for white sucker. and native fish from the Gunnison River could Increased water temperatures in the lower Gunnison supplement recruitment in the 15-Mile Reach. River could have a negative impact on white sucker, Second, the sewage treatment facilities located near but more data are needed on white sucker Clifton could be adding to the general productivity temperature preferences. In contrast, increased

II-32 water temperatures could benefit species like smallmouth bass. Reduced spring runoff flows in the Gunnison River could be detrimental to endangered fish such as Colorado pikeminnow and razorback sucker, since their spawn is associated with the spring runoff. Flow recommendations concerning the magnitude, duration and timing of peak flows are typically made for geomorphic purposes. Geomorphic processes are important for maintaining preferred habitats during the low flow period, but base flows are high in the Gunnison. Spring flow recommendations to ensure that geomorphic functions and processes remain intact need to be crafted in context with the existing base flow regimen. Spring flow recommendations are clearly desirable if reproduction or recruitment of endangered or other native fish is affected in poor runoff years. It appeared that bluehead sucker had improved recruitment during normal (> 4,000 cfs) runoff years on the Gunnison. Dolores River The fish data on the Dolores River demonstrate that roundtail chub and flannelmouth sucker can survive and adjust to low-flow conditions over the long term, albeit in very low abundance. Bluehead sucker, however, only barely survived the recent drought and further surveys are needed immediately to establish the status of this species. The Dolores River has been heavily diverted since 1886 (Vandas et al. 1990). The small size of flannelmouth sucker, bluehead sucker and roundtail chub at Big Gypsum may be adaptations by the native fish to the existing habitat conditions. Endangered Colorado pikeminnow and razorback sucker have not been observed in the Dolores River above Gateway or in other low-flow rivers. The lack of runoff flows in the Dolores River has had geomorphic consequences. Large sedimentation deposits had accumulated at Big Gypsum in 2004. The large spring runoff of 2005 then scoured these deposition sites. Although there was evidence of roundtail chub and flannelmouth sucker recruitment even during years of low spring runoff, there was no evidence of bluehead sucker recruitment in 2004 or 2005.

II-33 CONCLUSIONS

Based on the data reported here, we have found common nonnative fish at Big Gypsum in the that maintaining adequate base flows is the highest Dolores River, but we consider these impacts on priority for sustaining a thriving native fish native species to have been secondary to the impacts community, especially when negative impacts of of low flows. White sucker and carp were the most nonnative fish introductions intensify. Native fish common nonnative fish in the Gunnison River. abundance (density and biomass) was much greater White sucker is a high-impact species because of in the high base flow Colorado and Gunnison rivers hybridization with native sucker. Smallmouth bass than in the Yampa and Dolores rivers. Our habitat is another species with high negative impacts on modeling determined that most extant native fish native fish. The Yampa River had very high (bluehead sucker, flannelmouth sucker, speckled abundance of both white sucker and smallmouth dace and roundtail chub) prefer habitats with bass at Seven and Duffy. Channel catfish and moderate to swiftly flowing currents. Even smallmouth bass were abundant at Lily Park. roundtail chub, which occupies pool habitats during Among different species habitats, we found that the day, had its highest biomass in the Gunnison bluehead sucker habitat was the most indicative for River where they occupied eddy pools. Availability the habitat needs of the native fish assemblage of preferred habitat was directly and significantly overall. An abundant bluehead sucker population, related to base flow volume. composing about 25 percent of adult fish (> 15 cm) The Yampa River, with its low base flows, had and about 15 percent of the total biomass, indicated lower native fish abundance than the Colorado River that medi-riffle habitat was common. The and abundance decreased in years following a long availability of medi-riffle habitat was also associated term drought. Low base flows were associated with with peaking habitat diversity, based on the Shannon low native fish biomass, low availability of native diversity index. fish habitats and the lack of riffle habitat suitable for We validated the assumption that flows suitable invertebrate production. for adult bluehead sucker are also suitable for adult We found no overt trend in native fish abundance flannelmouth sucker using the 2D habitat analysis. varying with the magnitude or volume of runoff Flannelmouth sucker were usually the most common flows, making runoff flows secondary in importance native species collected. When bluehead sucker for maintaining native species (flannelmouth sucker, habitat and numbers were adequate, flannelmouth bluehead sucker and roundtail chub). The Colorado sucker habitat and numbers were high. River had moderately reduced runoff flows but a We confirmed the assumption that flows suitable near native base flow. The Gunnison River had for adult bluehead sucker are also suitable for heavily reduced runoff flows and greatly increased speckled dace by electrofishing under a wide range based flows. The Yampa River retains a relatively of base flow regimes. Speckled dace occupy natural spring runoff but has low base flows. The shallower riffles than do adult bluehead sucker. In Dolores River has low runoff and base flows relative rivers where bluehead sucker were common, to natural conditions. speckled dace were also common to abundant (the Increased abundance in nonnative species was Colorado and Gunnison Rivers). In the Dolores usually associated with declines in the native species Rivers bluehead sucker were rare but speckled dace assemblage. The impacts of nonnative species were were common but in the Yampa River both bluehead high to moderate in all study rivers. Channel catfish sucker and speckled dace were rare. and carp were the most common nonnative species We also used electrofishing surveys to confirm (fish > 15 cm TL) in the Colorado River, but these the assumption that flows suitable for adult bluehead species did not have substantial habitat overlap with sucker are also suitable for adult roundtail chub. the extant native species during normal base flows. Roundtail chub were common in the high base flow Channel catfish and carp were also the most rivers (Gunnison and Colorado) and persisted in low

II-34 abundance in low base flow rivers (Dolores and that initiated the wide scale losses in the native fish Yampa). Roundtail chub were vulnerable to assemblage upstream of Cross Mountain. predation and predation impacts appeared heightened during low flow periods. On the Yampa River low base flows may have exacerbated predation impacts Summary Statements on roundtail chub. The assumption that flows suitable for adult • Minimum base flows are needed to maintain the bluehead sucker are also suitable for spawning and existing native fish communities of the YOY life stages of the native fish assemblage also Colorado, Gunnison, Yampa and Dolores rivers. appears valid. High numbers of juvenile native fish were collected in the 15-Mile Reach on the Colorado • Base flows are necessary for maintaining adult and at Escalante on the Gunnison, whereas low habitats and invertebrate forage availability. numbers were collected at Lily Park where bluehead sucker habitat was poor. • Base flows that were too low appeared to We also examined the assumption that flows exacerbate interaction with nonnative fish, suitable for adult bluehead sucker are also suitable namely, predation and hybridization. for endangered adult Colorado pikeminnow and adult razorback sucker. Although these two endangered • Smallmouth bass increased on the Yampa River fish were rarely collected during the study, most of during the drought period. those found occupied habitats that were similar to roundtail chub and flannelmouth sucker habitats. It • Runoff flows were not correlated with native does not seem reasonable the adult Colorado fish abundance. pikeminnow and razorback sucker would require habitats or flows that are outliers to the natural flow • For community persistence base flows should regime. Thus, maintaining of bluehead sucker therefore be a higher priority over runoff flows. abundance by maintaining medi-riffle habitats will likely ensure the persistence of the habitat types • Spring runoff flows may be important for required by all adult native fish. reproductive success of certain native species. For the Colorado, Yampa and Gunnison rivers the 2D modeling analysis indicated adequate bluehead • In the short term the reduced spring runoff flow sucker habitat in the flow range of 600 to 1,200 cfs. may have less negative impacts on native fish Base flows in this range appear to be sufficient to than reduced base flows. maintain bluehead sucker abundance and reduce some of the nonnative fish problems identified with • Bluehead sucker abundance is a very good low base flows. indicator of riffle habitat availability. The 2D modeling analysis also validated the relationship between flow and habitat availability. • Bluehead sucker reproductive success may be Several other factors are associated with low base related to spring runoff flows. flows. Variations in flow also affect water temperature, cover, migration barriers and species • White sucker have had severe negative impacts interactions such as predation and hybridization. In on native sucker. general these associated factors are less detrimental in higher flow rivers, because native fish are well • Bluehead sucker should be the primary adapted to exploit swift habitats. During low-flow indicator species for biologically based instream conditions multiple factors act against the native fish flow recommendations. assemblage, because habitats and resources are more limiting. For example, 2002 was the lowest flow year on the Yampa River and appeared to be the year

II-35 • Flannelmouth sucker appeared capable of • The abundance of white sucker increased in an tolerating periods of reduced flows better than upstream direction in the Colorado, Gunnison bluehead sucker, because of basic differences in and Yampa Rivers. habitat requirements between the two species. • The abundance of white sucker may be related • Roundtail chub was not a good indicator to water temperatures. species for making instream flow recommendations because of its habitat use and • Channel catfish were not collected in the predatory foraging behavior. Gunnison, which indicated the Redlands Diversion Dam has blocked their migration. • Roundtail chub were highly resistant to flow reductions, but vulnerable to predation by • Low head dams may be an effective method for nonnative species (northern pike and controling channel catfish. smallmouth bass). • One of the more insidious impacts of the • Colorado pikeminnow is not a good indicator drought has been the recent explosion in species for instream flow recommendations crayfish at the Duffy site on the Yampa River. because of its habitat use and predatory foraging behavior.

II-36 FLOW RECOMMENDATIONS USING 2D HABITAT MODELING

We employed 2D modeling to develop habitat be determined for Duffy (Figure II-5). The analysis suitability criteria for two native sucker species: modeled Duffy flows to 600 cfs and Sevens flows to flannelmouth and bluehead sucker. Our methods 880 cfs. The biomass curves for bluehead sucker at and results appeared in Anderson and Stewart (2003) Duffy and Sevens were very similar at flows less and this report. The analysis yielded significant than 600 cfs. When the 2D modeling was contracted correlations between habitat availability and in 2001, we believed that 600 cfs would be a flannelmouth and bluehead sucker biomass. The reasonable upper limit for the flow simulations, relationship between discharge and projected because that limit was well above the usual base biomass (optimal habitat) was the primary tool for flow regime of about 250 cfs and a previous making biologically based instream flow instream flow study recommended 93 cfs for the recommendations. The inflection point of the instream flow (Modde et al. 1999). Our modeling biomass-to-discharge relationship determined the added simulations up to 2,000 cfs in 2004 for Lily recommended instream flow. Park, but additional simulations were not run for Duffy and Sevens. Yampa River Bluehead sucker biomass maximized at flows near 1,400 cfs at Lily Park. A flow of 600 cfs Flow and Habitat Relationships provided about 80 percent of maximum bluehead The inflection point of the discharge-to-biomass sucker biomass at Lily Park and at Sevens (Figure II- relationship for bluehead sucker at Lily Park and 6). A flow of 800 cfs is necessary to provide 85 Sevens was at 600 cfs, an inflection point could not percent of maximum bluehead sucker biomass at Lily

FIGURE II-5. Modeled biomass (kg/ha) of bluehead sucker as a function of discharge, Duffy, Sevens and Lily Park, Yampa River.

II-37 Park. The typical base flows of 250 cfs provided Lily Park. The modeling shows a steep decline in about 20 percent of maximum at Sevens and 40 biomass at flows less than 600 cfs. Modeled base percent of maximum at Lily Park. Flows of 100 cfs flows of near 250 cfs yielded a flannelmouth sucker provided only 1 percent of the maximum bluehead biomass of about 100 kg/ha at Lily Park and Sevens. sucker biomass at Sevens and 14 percent at Lily Park. At 150 cfs modeled biomass was only 15 kg/ha at The bluehead sucker biomass curve (Figure II-5) Sevens and 5 kg/ha at Duffy. was consistent with our sampling data. Bluehead Discharge at 700 cfs (the inflection point flow for sucker biomass was higher at Lily Park than at Sevens Lily Park) provided about 78 percent of the maximum and Duffy in the flow range below 250 cfs (Table II- modeled flannelmouth sucker biomass (Figure II-8). 2). Lily Park and Corn Lake on the Colorado River Base flows of 700 cfs or higher have not been (Figure II-9) had similar bluehead sucker biomass available for the Yampa River. Typical base flows projection curves of about 120 kg/ha at 1,000 cfs. have been about 250 cfs during the winter but much These curves indicate that Lily Park and Corn Lake less in the irrigation period. Both the measured and would have similar bluehead sucker biomass given the modeled biomass of flannelmouth sucker were similar base flow regimes. Lily Park (at 250 cfs) had substantially lower under a base flow regime below about 30 percent of the bluehead sucker biomass of 150 cfs. Corn Lake (at 1000 cfs) (Table II-2). For flannelmouth sucker the discharge-to-biomass Base Flow Recommendation inflection point at Lily Park was at 700 cfs (Figure II- Based on the 2D modeling results above, our 7). The Duffy and Sevens inflection points were instream flow recommendation for the Yampa River likely higher than the highest simulated flow. from the Little Snake River confluence upstream to Flannelmouth sucker biomass peaked at 1,400 cfs at the town of Craig is 650 cfs. The 650 cfs

FIGURE II-6. Percent of modeled maximum biomass (kg/ha) of bluehead sucker as a function of discharge at Duffy, Sevens and Lily Park, Yampa River.

II-38 FIGURE II-7. Modeled biomass (kg/ha) of flannelmouth sucker as a function of discharge at Duffy, Sevens and Lily Park, Yampa River.

FIGURE II-8. Percent of maximum modeled biomass (kg/ha)of flannelmouth sucker as a function of discharge at Duffy, Sevens and Lily Park, Yampa River.

II-39 recommendation is the average of the inflection to 2004 and flow exceeding 11,100 cfs occurred in points for the curves representing bluehead and only one of those years (Figure A1-1). The Yampa flannelmouth sucker biomass as a function of River has had a higher frequency of bankfull flows discharge. Nevertheless, the 650 cfs recommendation than the Colorado and Gunnison rivers in recent is not realistic for the Yampa because of the lack of years. The magnitude, duration and frequency of available water. We make the 650 cfs spring runoff flows do not appear to be a primary recommendation as a formality, since it is necessary limiting factor for the native fish assemblage in the to maintain consistency when using an instream flow upper reaches of the Yampa River. methodology. At flow less than 500 cfs, the geomorphology of Population sampling in the Yampa River has the Yampa River was poor for maintaining fish documented that the post-1999 base flow regimes habitat. A much narrower channel (akin to the were not adequate to sustain the native fish Dolores River at Big Gypsum) would be required to assemblage. The base flow regime before 1999 was provide substantial native sucker habitat in the range more conducive to sustaining native fish abundance of 200 to 250 cfs. Base flows substantially higher and species. Prior to 1999 typical summer/fall base then 250 cfs do not currently appear to be a possibility flows were near 250 to 300 cfs with minimum flow of for the Yampa River. Bank stabilization or channel at least 200 cfs. The fish data suggests that a narrowing efforts designed to improve native fish reasonable minimum flow recommendation is at least habitat are not highly feasible. Elimination of 200 cfs. undesirable nonnative species (smallmouth bass, northern pike, channel catfish and white sucker) Spring and Channel Maintenance Flow would be effective for restoring native fish and some Recommendations removal efforts are currently under way as part of the Upper Colorado River Recovery Program. We did not formulate spring flow recommendations as part of this analysis. Colorado River Maintaining high spring flows appears to be important for native fish management in the lower Flow and Habitat Relationships Yampa River (Dinosaur Canyon), however, and for The inflection point of the discharge-to-biomass mitigating altered flows from the Green River. relationship for bluehead sucker on the Colorado Bestgen et al. (1998) concluded that moderate spring River was 600 cfs at Clifton and 700 cfs at Corn Lake flows produce the largest Colorado pikeminnow YOY (Figure II-9). Modeled bluehead sucker biomass numbers. Maintaining high spring flows in the peaked at a discharge of 1,000 cfs at Corn Lake and Yampa upstream of Cross Mountain Canyon has not 1,200 cfs at Clifton. The modeled biomass did not been associated with maintenance of the native fish increase at flows over 1,200 cfs. Sampled bluehead assemblage in that river reach. In fact, bank sucker biomass was not higher in sections of the instability appears to be a problem in the upper Yampa Colorado River with flows over 1,400 cfs (see the River (Richard and Anderson 2006) and high spring Supplemental Data for Each River section above). flows may be exacerbating bank erosion. Modeled biomass at both sites declined rapidly as Bankfull flows are directly related to sediment flows decrease below 500 cfs (Figure II-9). transport and therefore channel maintenance. At Corn Lake a flow of 700 cfs (inflection point) Channel maintenance flows are necessary to maintain maintained about 90 percent of the modeled channel geomorphology and habitats used by fish maximum biomass of bluehead sucker (Figure II-10). during base flow periods. Andrews (1980) At Clifton the inflection point flow of 600 cfs determined bankfull flow to be about 9,000 cfs for the maintained 73 percent of the maximum and 800 cfs Yampa River near the Maybell gage. Richard and maintained 85 percent of the maximum. These Anderson (2006) identified bankfull flow of 11,100 differences between Corn Lake and Clifton are an cfs for Sevens and Lily Park. Flows exceeding 9,000 artifact of their different channel widths. Corn Lake cfs have occurred in four of the seven years from 1998 and Clifton had similar optimal habitat area and

II-40 FIGURE II-9. Modeled biomass (kg/ha) of bluehead sucker as a function of discharge at Corn Lake and Clifton, Colorado River.

FIGURE II-10. Percent of modeled maximum biomass (kg/ha) of bluehead sucker as a function of discharge at Corn Lake and Clifton, Colorado River.

II-41 biomass estimates per distance (km) but not for appear to be representative of the natural flow regime surface area. Clifton had a wider channel and a larger (Pitlick 1999). Flows of 800 to 1,000 cfs also have a surface area, so its biomass (kg/ha) was lower than high habitat diversity rating (Shannon index). The Corn Lake’s. The average width of Corn Lake and fish surveys in 1999, 2000 and 2001 were in the flow Clifton was close to the mean width of the 15-Mile range of 800 to 1,200 cfs. Large numbers of age-1+ Reach overall. Thus, the average of the Corn Lake native fish were collected at flows of 800 cfs and and Clifton biomass curves is probably the best above (see the CD Appendix –Fish- Histograms). The representation for the reach. fish sampling confirmed that a flow of 700 cfs would For the flannelmouth sucker in the Colorado the maintain adequate nursery habitats for native fish inflection points for the biomass curve were 600 cfs at species. Typically, fry and nursery habitats Corn Lake and 800 cfs at Clifton (Figure II-11). availability are expected to be maximized at a lower Modeled maximum biomass of flannelmouth was at base flow than that needed for adult fish habitat, since about 1,000 cfs. Flows above 1,000 cfs did not juveniles occupy shallower and lower velocity increase modeled flannelmouth sucker biomass in the habitats. 15-Mile Reach (Figure II-12). Sampled flannelmouth sucker biomass was not higher in sections of the Spring and Channel Maintenance Flow Colorado River with higher base flow regimes (see Recommendations the Supplemental Data section above). Modeled flannelmouth sucker biomass declined sharply as The 2D modeling did not include spring or flows declined below 500 cfs. channel maintenance flow recommendations since The inflection point flow of 600 cfs for there is little promise of relating peak flows or flannelmouoth sucker at Corn Lake maintained about recurrence of bankfull flows with biologically based 85 percent of the modeled maximum biomass (Figure metrics. Bankfull flows are directly related to II-12). The 800 cfs inflection point at Clifton sediment transport and therefore channel maintained about 94 percent of the modeled maintenance. Channel maintenance flows maintain maximum. Modeled flows of 400 cfs maintained channel geomorphology, which is the template for about 60 to 70 percent of the flannelmouth sucker habitat availability during base flow periods. biomass at both sites. Pitlick (1999) identified bankfull flow to be 22,000 cfs for the 15-Mile Reach of the Colorado Base Flow Recommendation River. Bankfull flows have not been achieved since Based on the 2D modeling results above, our 1997 (eight years). In 2002 the mean daily peak flow instream flow recommendation for the 15-Mile Reach was 2,780 cfs; in 2004 it was 5,510 cfs. Although of the Colorado River is 700 cfs. the fish collections showed higher recruitment of The 700 cfs recommendation was the bluehead nonnative fish in years of low peak flows, overt sucker inflection point at Corn Lake and the average problems with channel morphology or sedimentation of the flannelmouth sucker inflection points for the were not observed during our fish or habitat surveys two sites. Typical base flows in the 15-Mile Reach in the years of low spring flows. have ranged from 800 cfs to 1,200 cfs. This suggests Recommendations for the magnitude, frequency that the bluehead sucker and flannelmouth sucker and duration of spring runoff flows have low biomass estimates made in this flow regime also potential for implementation and spring flows have represent the maximum biomass of these sites. The been driven almost entirely by snow-pack conditions. 700 cfs flow maintained 85 percent of the modeled As noted earlier, we have concluded that base flows maximum bluehead sucker biomass and 89 percent of should be a higher priority than spring flows. For the maximum flannelmouth sucker biomass. these reasons it seems more appropriate to focus on Maintenance of 85 percent of the maximum value maintaining adequate base flows. Adequate base appears adequate for ensuring the long-term survival flows are maintained in the 15-Mile Reach with of native fish in the 15-Mile Reach. flows of 700 to 1,200 cfs. Base flows do not need to Base flows in the range of 800 cfs to 1,200 cfs be in excess of 1,200 cfs. Flows that exceed 1,200

II-42 FIGURE II-11. Modeled biomass (kg/ha) of flannelmouth sucker as a function of discharge at Corn Lake and Clifton, Colorado River.

FIGURE II-12. Percentage of modeled maximum biomass (kg/ha) of flannelmouth sucker as a function of discharge at Corn Lake and Clifton, Colorado River.

II-43 cfs during the base flow period should be stored for a flows decreased below 400 cfs (Figure II-13). spring release. Maximum bluehead sucker biomass was modeled Our spring flow recommendation is to maintain at flows near 1,000 cfs at both Delta and Escalante. the current 15 year median peak of about 13,000 cfs. About 95 percent of maximum remained at 700 cfs We also recommended that biological sampling be and 91 percent of maximum remained at 600 cfs conducted to determine relationships between spring (Figure II-14). These findings suggests that base flows and biological processes such as reproductive flows of 600 cfs are sufficient for perpetuation of success of native and nonnative fish species. A existing bluehead sucker biomass. biological data base is necessary to determine the For flannelmouth sucker the inflection points for habitat suitability indices associated with spring the discharge-to-biomass curve was at 600 cfs at Delta flows and, in turn, to model the optimal conditions and Escalante (Figure II-15). Flannelmouth sucker associated with spring flow regimes. projected biomass peaked between 600 to 800 cfs at Delta and between 1,000 to 1,400 cfs at Escalante. Gunnison River The flannelmouth sucker curves indicated reduced abundance when flows exceed 1,000 cfs at Delta and Flow and Habitat Relationships 1,400 cfs at Escalante. Modeled flannelmouth sucker The inflection points of discharge-to-biomass abundance dropped sharply when flows are less than relationships curves for bluehead sucker at both Delta 400 cfs (Figure II-15). and Escalante were at 600 cfs (Figure II-13). The projected maximum flannelmouth sucker Bluehead sucker biomass peaked at about 1,000 cfs at biomass peaked near 800 cfs at Delta and 1,200 cfs at both study sites on the Gunnison. Flows over 1,200 Escalante (Figure II-16). At 600 cfs about 98 percent cfs resulted in decreasing habitat. The curves for of maximum biomass was retained at Delta and 87 both Delta and Escalante dropped sharply as base percent persisted at Escalante. This modeling

FIGURE II-13. Modeled biomass (kg/ha) of bluehead sucker as a function of dischargeat Delta and Escalante, Gunnison River.

II-44 FIGURE II-14. Percentage of modeled maximum biomass (kg/ha) of bluehead sucker as a function of discharge at Delta and Escalante, Gunnison River.

identified that a 600 cfs base flow regime is sufficient The fish surveys in 2004 and 2005 empirically to maintain existing flannelmouth sucker biomass. determined that fry and juvenile native fish were common at the base flows in those years, 900 to 1,200 Base Flow Recommendation cfs. Typically, fry and nursery habitat availability are Based on the 2D modeling results above, our expected to maximize at lower base flows than those instream flow recommendation for the Gunnison maximizing adult fish, because juveniles prefer River from its Colorado River confluence upstream to shallow low-velocity habitats. the town of Delta is 600 cfs. The recommendations of 600 cfs represents the Spring and Channel Maintenance Flow inflection points for bluehead and flannelmouth Recommendations sucker biomass as a function of discharge at both study sites on the Gunnison. The 2D modeling Again, the 2D modeling did not include indicated that about 90 percent of the projected developing a spring or channel maintenance flow maximum bluehead and flannelmouth sucker biomass because there is little promise of relating peak flows would be maintained at 600 cfs. or recurrence of bankfull flows with biologically The 600 cfs recommendation is supported by the based metrics. Bankfull flows are directly related to Shannon habitat diversity values, which were highest sediment transport and therefore channel at flows of 600 to 800 cfs. The maximum Shannon maintenance. Channel maintenance flows are diversity suggests that the habitat types required by necessary to maintain channel geomorphology and other species and by younger life stages of native habitats used by fish during base flow periods. species would also be available at a 600 cfs base flow Pitlick (1999) identified bankfull flow on the (Anderson and Stewart 2006). Gunnison River to be 14,500 cfs, but flows have not

II-45 FIGURE II-15. Modeled biomass (kg/ha) of flannelmouth sucker as a function of discharge at Delta and Escalante, Gunnison River.

FIGURE II-16. Percentage of modeled maximum biomass (kg/ha) of flannelmouth sucker as a function of dischargeat Delta and Escalante, Gunnison River.

II-46 been that high since 1995. In 2002 the mean daily Our recommendation is to strive to maintain an peak was flow was 1,464 cfs, in 2004 it was 2,769 cfs. average peak of 6,000 cfs for a spring peak flow. Reduced reproductive success for native species is a Sediment transport studies are necessary to determine potential negative consequence in years with low accurate sediment transport rates under the current runoff flows. Two or more consecutive years of low hydrograph and potentially altered hydrographs. spring flows could negatively impact native sucker biomass. The lack of flushing flows in 2002 and 2004 Dolores River did not appear to result in negative impacts to channel geomorphology. Field crews did not notice excessive Flow and Habitat Relationships sedimentation or channel deterioration during the At Big Gypsum the discharge-to-biomass 2003, 2004 and 2005 fish and habitat surveys. inflection point for bluehead sucker was 300 cfs Since 1965 the mean annual peak on the (Figure II-17). Bluehead sucker biomass was still Gunnison has been about 6,000 cfs. If 6,000 cfs has increasing at the highest modeled flow of 500 cfs. been functional for sediment transport equilibrium in Modeled bluehead sucker biomass fell sharply as the last 40 years, it should continue to be functional in flows decreased below 100 cfs. the future. Base flows have been quite high since The modeled maximum for bluehead sucker 1965 (1,000 to 1,200 cfs), however, these flows are biomass was at the highest modeled flow of 500 cfs certainly capable of transporting fine sediment from (Figure II-18). About 85 percent of maximum riffles and runs. Spring runoff or flushing flows may biomass was available at 300 cfs and 68 percent of need to be higher than 6,000 cfs to maintain current maximum biomass remained at 200 cfs. Outlet sediment transport rates if the base flow regime drops flows from McPhee Dam normally range from 25 to to 600 cfs or less. 60 cfs. Only one percent of the bluehead sucker

FIGURE II-17. Modeled biomass (kg/ha) of bluehead sucker as a function of discharge at Big Gypsum, Dolores River.

II-47 FIGURE II-18. Percentage of modeled maximum biomass (kg/ha) of bluehead sucker as a function of discharge at Big Gypsum, Dolores River.

FIGURE II-19. Modeled biomass (kg/ha) of flannelmouth sucker as a function of discharge at Big Gypsum, Dolores River.

II-48 biomass maximum remained at 25 cfs and 12 Base Flow Recommendation percent remained at 60 cfs. Measured bluehead sucker biomass at Big Gypsum was very poor in all Based on the 2D modeling results above, our sample years, but it was extremely low in years after instream flow recommendation for the Dolores 2002, the severest year of the drought. River from the San Miguel confluence upstream to The inflection point for the flannelmouth sucker the Disappointment Creek confluence is 300 cfs. biomass curve was at 300 cfs (Figure II-19). The The recommendation is not a possibility for an projected maximum biomass for flannelmouth was instream flow downstream of McPhee dam because at the highest modeled flow of 500 cfs. Projected of historic diversions and current water availability. flannelmouth sucker biomass dropped sharply when Current releases from McPhee reservoir are flows fell below 100 cfs. consistent with its operational plans and constraints. The modeled maximum biomass for The habitat modeling indicates that flows near 300 flannelmouth sucker peaked at 500 cfs at Big cfs are required to produce a native fish assemblage Gypsum. At 300 cfs about 85 percent of maximum biomass comparable in biomass to those currently biomass was retained (Figure II-20). The modeling occupying the Colorado River in the 15-Mile Reach indicated that flannelmouth sucker habitat is and the Gunnison River between Delta and severely limited at base flows less than 60 cfs at Big Whitewater. (The Colorado and Gunnison are rivers Gypsum. The electrofishing surveys found that that appear to be most representative of naturally flannelmouth sucker biomass was in fact very poor occurring standing stocks of native sucker.) The 2D in all years of the survey. modeling effort clearly identifies low flows as the

FIGURE II-20. Percentage of modeled maximum biomass (kg/ha) of flannelmouth sucker as a function of discharge at Big Gypsum, Dolores River.

II-49 cause for a nearly complete lack of adult native instream flow recommendation on long- term fish sucker at Big Gypsum in recent years. sampling data. Before 2000 adult flannelmouth and The reason the 2D instream flow methodology bluehead sucker were commonly collected in the did not produce a reasonable instream flow Dolores from McPhee dam to Dove Creek (Nehring, recommendation for the Dolores River has to do DOW unpublished data), but very few adult native with several underlying factors. The 300 cfs sucker have been collected from this reach after recommendation is based on channel 2002. Clearly, Dolores River flows at the Bedrock geomorphology and not current water availability. gage in 2002, 2003 and 2004 were inadequate for The model identified the instream flow based on sustaining the native fish assemblage. Native fish habitat criteria for adult bluehead and flannelmouth were more plentiful from 1986 to 1999, however, sucker, which occupy the deeper and swifter which suggests that river flows during that period habitats. In addition, the 2D model simulations were were adequate. conducted only up to 500 cfs, which was not high Our analysis of fish and habitat data from the enough to capture the range of flows where the curve recent drought period raises a major concern in the peaked. Flow simulations up to a 1,000 cfs appear management of flow for ecological or native fish necessary to capture the full bluehead and objectives. The “fish pool” (reservoir water reserved flannelmouth sucker biomass potential for the for downstream releases) is not a fixed amount of Dolores River. water but varies depending on the water yield each Moreover, the analysis modeled only one site on year. The fish pool is a fixed quantity in years when the Dolores River. Additional data on fish and there is a ‘spill’, but in years without spills (when the habitat upstream of Disappointment Creek would entire runoff is stored or diverted), the fish pool can have provided useful information. In general, the be much lower that usual. This was the case on the reach upstream of Disappointment Creek has a Dolores in 2002 and 2003. This practice is poor steeper gradient and does not have a sedimentation flow management for sustaining native fish and the problem. A steeper gradient and clean cobble ‘shared shortage’ policy appears to have had severe substrates were geomorphic factors that improved negative consequences for native fish abundance habitat and biomass at Lily Park relative to upstream during 2002 and 2003. Yampa River sites. On the Dolores these Even in spill years the downstream releases on geomorphic factors would also result in higher the Dolores appear to be only minimal for providing invertebrate and fish productivity upstream of native fish habitat. Native fish evolved under a Disappointment Creek relative to Big Gypsum, snowmelt-driven hydrograph. Spring flows have under similar flows. several geomorphic processes (Richard and Two approaches would provide an instream flow Anderson 2006). In years when McPhee Reservoir recommendation that is based on current water use fully captures all the spring flows, base need to be as practices. One is to rerun the 2D model and using high as normal or higher to mitigate the habitat criteria for juvenile bluehead and consequences of lost runoff flows. The Dolores flannelmouth sucker. The fish surveys at Big River Dialogue (2006B) estimated that no spill years Gypsum (2000 to 2005) found that juvenile-sized have a frequency of about 45 percent. native sucker far outnumbered adult fish. Juvenile The minimum flow necessary to sustain a bluehead and flannelmouth sucker (12- 23 cm) prefer modest flannelmouth and bluehead sucker biomass habitats with lower depth and velocity ranges than during spill years appears to be in the range of 50 to those preferred by adults and use of juvenile criteria 60 cfs. The minimum flow necessary during “no would produce a lower inflection point that the adult spill” years may need to be higher than 60 cfs. At 60 criteria did in the results presented above. This cfs about 61 percent of the habitat at Big Gypsum is approach would makes sense if the native fish in pool, 34 percent in run and only five percent in management objective is to provide long-term habitat riffles (see the CD appendix- 2D Modeling, Figure suitable for sustaining small or stunted native fis h. II-11). At 80 cfs about 50 percent is pool, 42 percent The second approach would be to base the is run and eight percent is riffles. At Big Gypsum a

II-50 base flow of 80 cfs is necessary to achieve a 50:50 Gypsum site were 2,600 cfs. Excessive pool-to-run/riffle ratio. sedimentation and channel encroachment was In summary, based on habitat features identified obvious during the 2004 fish surveys and highly at Big Gypsum, we recommend an instream flow of detrimental to maintaining native fish. 60 cfs is recommended for years with spills and an The Big Gypsum site is located downstream of 80 cfs flow for years without spills. Disappointment Creek, which is a major source of sediment (Richard and Anderson 2006). Upstream Spring and Channel Maintenance Flow of Disappointment Creek the Dolores does not Recommendations appear to have severe sedimentation problems. Spring flows of 2,000 cfs for a period of seven days Bankfull flows are directly related to sediment at a frequency of every other year would likely transport and channel maintenance. Channel maintain channel configuration (Vandas et al. 1990). maintenance flows are necessary to maintain Observations made during our study are supportive channel geomorphology and habitats fish require of a spring instream flow recommendation of 2,000 during base flow periods. Richard and Anderson cfs at a frequency of every other year. (2006) determined that bankfull flows at the Big

ACKNOWLEDGEMENTS

The Colorado Division of Wildlife under Federal Aid Project F-289 funded this study. We would like to thank Dr. Craig Addley and Jennifer Ludlow at Utah State University for their help with the two- dimensional modeling. Report drafts were carefully reviewed and I greatly appreciate comments received from Dr. John Woodling, David Graf and Mike Japhet of the CDOW and Michele Garrison of the CWCB.

II-51 LITERATURE CITED

Aadland, L. P., C. M. Cook, M. T. Negus, H. G. Andrews, E.D., 1980. Effective and bankfull Drewes and C. S. Anderson. 1991. discharges of streams in the Yampa river basin, Microhabitat preferences of selected stream Colorado and Wyoming. Journal of Hydrology fishes and a community-oriented approach to 46, pp. 311-30. instream flow assessments. Investigational Annear, T., I. Chisholm, H. Beecher, A. Locke, P Report No. 406. Minnesota Division of Fish Aarrestad, N. Burkhart, C. Coomer, C. Estes, J. and Wildlife, St. Paul. 142 pp. Hunt, R. Jacobson, G. Jobsis, J. Kauffman, J. Anderson*, R. M. 2005. Riverine fish-flow Marshall, K. Mayes, C. Stalnaker and R. investigations: Quantification of impacts of Wentworth. 2002. Instream Flows for the 2002 drought on native fish population in Riverine Resource Stewardship. Instream the Yampa and Colorado Rivers. Flow Council, Cheyenne, WY. (see Annual Progress Report to Colorado Division www.instreamflowcouncil.org) of Wildlife, F-289-R. Fort Collins, CO 89pp. Bain, M. R., J. T. Finn. And H. E. Booke. 1988. Anderson, R.M. 2004, Riverine Fish-Flow Stream flow regulation and fish community Investigations: Quantification of impacts of structure. Ecology 69(2):382-392. the 2002 drought on native fish populations in Bestgen. R. R., R. T. Muth and M. A. Trammell. 1998. the Yampa and Colorado rivers. Annual Downstream transport of Colorado squawfish Progress Report to Colorado Division of larvae in the Green River drainage: temporal Wildlife, F-289-R7. Fort Collins, CO 96pp. and spatial variation in abundance and Anderson, R. M. 1997. An evalulation of fish relationships with juvenile recruitment. Final community structure and habitat potential for report to the Recovery Program for Colorado squawfish and razorback sucker in Endangered Fishes in the Upper Colorado the unoccupied reach (Palisade to Rifle) of the River, Project Number 32. Larval Fish Lab, Colorado River, 1993 -1995. Colorado River Colorado State University, Fort Collins . Recovery Implementation Program Project Burdick, B.. D.. 1995. Ichthyofaunal studies of the No. 18. Colorado Division of Wildlife Federal Gunnison River, Colorado, 1992 - 1994 . Final Aid Project SE-3. Fort Collins, Co. Report prepared for the Recovery Anderson*, R. M. and G. Stewart. 2006. Riverine Implementation Program for Endangered fish-flow investigations: Quantification of Fishes in the Upper Colorado River Basin. habitat availability and instream flows on the U.S. Fish and Wildlife Service, Grand Gunnison River and impacts of long term Junction, Colorado. 66 pp. drought on native fish populations in the Byers*, D.W., C. Sodergren, J. M. Bundy and K. Dolores River. Annual Progress Report to Bestgen. Habitat use and movement of Colorado Division of Wildlife, F-289-R. Fort bluehead sucker, flannelmouth sucker and Collins, CO 89pp. roundtail chub in the Colorado River. Anderson*, R. M. and G. Stewart. 2003. Riverine Contribution 121, Larval Fish Laboratory, fish-flow investigations: Using 2D modeling Dept. of Fishery and Wildlife, CSU, Fort to determine relationship between flow and Collins Co. habitat availability for warn-water riverine fish Chapman, D. G. 1954. The estimation of biological in Colorado. Annual Progress Report to populations. Annals. Mathematical Statistics Colorado Division of Wildlife, F-289-R. Fort 25:1-15. Collins, CO 106pp. Courtenay, W. R. and P. B. Moyle. 1992. Crimes against Biodiversity: The lasting legacy of fish introductions. Trans. N. A. Wildlife & Nat. Res. Conference No. 57. Vol. 57:365-372.

II-52 Dolores River Dialogue (DRD) Hydrology Report Osmundson, D. B. 1999. Longitudinal variation in (draft). 2006a. DRD Report, Cortez CO 61 pg. fish community structure and water (David Graf CDOW editor, Grand Junction). temperature in the upper Colorado River. Dolores River Dialogue (DRD) Correlation Report Final Report. U. S. Fish and Wildlife Service, (draft). 2006b. DRD Report, Cortez CO 61 pg. Grand Junction, Colorado. Espegren, G. D. 1998. Evaluation of the standards and Osmundson, D. B., P. Nelson, K. Fenton and D. W. methods used for quantifying instream flows Ryden. 1995. Relationships between flow and in Colorado. Colo. Water Con. Board. rare fish habitat in the ’15-Mile reach’ of the Denver, CO. 18 pp. upper Colorado River. Final Report. U. S. Everhart, W. H. and W. D. Youngs. 1981. Principles Fish and Wildlife Service, Grand Junction, of fishery science. Cornell University press. Colorado. Martinez, P. J. 2006. Westslope Warmwater Fisheries. Peters, E, J., R. S. Holland and B. C. Chapman. 1992. Federal Aid in Fish and Wildlife Restoration Studies of the channel catfish in the lower Project F-325-r Progress Report. Colorado Platte River, Nebraska. Final Report, Federal Division of Wildlife, Fort Collins, 121 pages. Aid Project F-78-R. Nebraska Game and Parks McAda, C.W. 2003. Flow recommendations to benefit Commission, Lincoln, 198 pp. endangered fishes in the Colorado and Pitlick, J., M. Van Steeter, B. Barkett, R. Cress and M. Gunnison Rivers. Recovery Program Project Franseen. 1999. Geomorphology and Number 54. Final Report July 2003. U. S. Fish hydrology of the Colorado and Gunnison and Wildlife Service, Grand Junction CO. rivers and implications for habitats used by Modde, T., W. J. Miller, R M. Anderson and D. Irving. endangered fishes. Final Report to the 1999. Determination of habitat availability, Recovery Program for Endangered Fishes in habitat use and flow needs of endangered the Upper Colorado River, Project Number 44, fishes in the Yampa River between August and Department of Geography, University of September. Final Report to the Recovery Colorado, Boulder. Program for Endangered Fishes in the Upper Pitlick, J., M. and R. Cress. 2000. Longitudinal trends Colorado River. U. S. Fish and Wildlife in channel characteristics of the Colorado Service, Denver CO. 134 pp. Rivers and implications for food web Modde, T. and G. Smith. 1995. Flow dynamics. Final Report. University of recommendations for endangered fishes in the Colorado, Boulder. Yampa River. Final Report to the Recovery Poff, N. L.. J. D. Allan, M. B. Bain, J. R. Karr, K. L. Program for Endangered Fishes in the Upper Prestegaard, B. D. Richter, R. E. Sparks and J. Colorado River, U. S. Fish and Wildlife C. Stromberg. 1997. The Natural Flow Service, Denver, CO. 44 pp. Regime, A paradigm for river conservation Nehring, R. B. 1979. Evaluation of instream flow and restoration . BioScience Vol. 47 No. 11. methods and determination of water quantity Rees*. D. E. and W. J. Miller. 2001. Habitat selection needs for stream in the state of Colorado. and movement of native fish in the Colorado Report to the U.S. Department of interior, U.S. River, Colorado. Miller Ecological Fish and Wildlife Service. Colorado Division Consultants, Inc. Fort Collins, CO. of Wildlife Final Report, Fort Collins, CO. Richard G. A. and R. M. Anderson. 2006. Channel- Nesler, T. P. 1995. Interactions between endangered forming discharge on the Dolores River and fishes and introduced gamefishes in the Yampa Yampa River, Colorado. Colorado Division River, Colorado , 1987-1991. Colorado River of Wildlife Technical Report #80, Fort Recovery Implementation Program Project Collins, CO. No. 91-29. Colorado Division of Wildlife Federal Aid Project SE-3, Fort Collins, CO.

II-53 Rose, K. L. and D. R. Hann. 1989. Consolidated Vandas, S., Whittaker, D., Murphy, D., Prichard, D., instream flow report habitat modeling on the MacDonnell, L., Shelby, B., Muller, D., Fogg, Green River using the physical habitat J.\ and Van Havern, B., 1990. Dolores River simulation system. Final Report. U. S. Fish Instream Flow Assessment, Project Report . US and Wildlife Service, Grand Junction, CO. Department of the Interior, Bureau of Land Scoppettone, G. G. 1993. Interactions between Native Management. and Nonnative fishes of the Upper Muddy River, Nevada. Transactions of the American Fisheries Society: 122:599-608. Author* = citation.pdf is included provide in CD in Stewart* G. 2000. Two-dimensional hydraulic Appendix - Report-Project F-288. modeling for making instream flow recommendations. M. S. Thesis Colorado State University, Fort Collins CO. Stewart* G., R. M. Anderson and E. Wohl. 2005. Two-Dimensional modeling of habitat suitability as a function of discharge on two Colorado Rivers. River Research and Applications: 21:1061-1074. Stewart* G. and R. M. Anderson, 2005. Two- Dimensional modeling of habitat suitability as a function of discharge on two Colorado Rivers. River Research and Applications (In press). Travnichek, V. H., M. B. Bain and J. J. Maceina. 1995. Recovery of a warmwater fish assemblage after the initiation of a minimum-flow release downstream from a hydroelectric Dam. Transactions of the American Fisheries Society: 124:836-844. Tyus H. M. and C. A. Karp. 1989. Habitat use and streamflow needs of rare and endangered fishes, Yampa River, Colorado, U. S. Fish and Wildlife Service, Biological Report 89 (14). Valdez, R. A., P. G. Mangan, M. McInerny and R. P. Smith. 1982. Tributary Report: Fishery Investigations of the Gunnison and Dolores Rivers. Pages 321-365 in W. H. Miller et al., editors. Colorado River Fishery Project, Final Report; Part Two, Field Studies. U. S. Fish and Wildlife Service and Bureau of Reclamation. Salt Lake City Utah. . Valdez, R. A., W. J. Masslich and A.Wasowicz. 1992. Dolores River native fish habitat suitability study (UDWR Contract No. 90-2559). BIO/WEST Inc. Logan Utah. 118 pp.

II-54 APPENDIX ONE Site Hydrographs 4 0 , 0 ) 2 , / e ) 1 / g 5 e 3 0 a / g 1 / g a 3 3 g 0 e s f 0 c d k 2 / 5 4 a c 9 1 0 / / 0 s o , 1 i 3 / 3 s l r 3 f s c a d i 0 2 e n P , 5 0 a ( e i 2 0 B , 3 d g ( 3 2 0 a e h / / 1 1 g 1 / m p h / s 3 , i e 3 r a p d e r n a v a i a s g i r 1 R d 2 d i 0 o g l 0 s e / 0 r a e 1 o 2 / r M P / d r 3 o l 1 y / d o 3 D h y , e r h 1 g 0 0 / e a r 1 0 / g v e 0 3 i k 2 v . . c / i R o 1 5 4 r / R d 0 0 3 o 0 e 0 0 0 / B d s 1 2 2 / a e 9 3 r r – – 9 9 8 8 o o 1 l l / 9 9 o o 1 9 9 9 / 9 / 3 1 1 C D 1 / 3 8 9 . . 9 2 4 1 8 - - / 9 / 1 1 1 0 0 0 0 0 0 0 0 / 1 / 0 0 0 0 0 0 0 3 3 0 0 0 0 0 0 A A 0 0 0 0 0 0 0 0 0 0 0 0 8 6 4 2 0 2 4

0 0 0 0 0 e e 1 1 1

5 4 3 2 1 r r

flow in cfs in flow u cfs in Flow u g g i i F F – 8 9 9 1 5 , 0 4 ) 0 0 2 e / s 0 1 2 / g u / 3 l 1 a / p 3 g 4 0 l a . l 0 t 3 2 5 l / e 0 1 0 0 / e b 3 2 0 / D y 1 ( 2 / a . 3 3 – 5 0 h M 0 0 8 0 2 ( p / 2 9 2 1 0 / a o 9 h 0 3 t r 2 1 / p 8 g 9 1 , 2 a / 9 o ) 0 r 3 1 r 0 s 2 g s / d e e 1 1 / o g y 0 3 g r a 0 h g a s d 2 f / e 1 c r g y 1 r 0 / 0 e g 0 h 3 0 e h 2 v / 8 r a , i r 1 s / 9 p f 0 g e 3 R c s 0 m i v h 0 0 o i n 0 c 2 n 0 a a / 2 0 n i , R 1 o p 0 U d 6 / 2 e / 3 s s a s i i 1 m / u m l . p 3 n , n o p 9 a 4 i e 9 n a c m d t 0 g 9 l 9 u e a n 1 a 9 e 0 / g 9 M D l 1 2 Y G U 1 l / / e 1 3 / b 3 y a 8 . . 9 M 8 1 3 9 9 - - 9 1 / 1 1 1 / 1 0 0 0 0 0 0 0 0 1 / / 0 0 0 0 0 0 0 0 0 0 A A 0 0 0 0 0 0 0 3 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 4 6 8 0 2 4 e e

9 8 7 6 5 4 3 2 1 1 1 1

r r

Flow in cfs in Flow flow in cfs in flow u u g g i i F F

II-55 1 e d k 5 2 5 2 9 , 0 2 3 . 5 4 0 a 0 c 0 1 0 0 4 0 0 s - 0 7 2 o i 4 l r - - a d 2 2 , 0 7 e 3 2 P 8 0 . 8 2 ( 2 0 0 0 - B 8 2 3 0 , 0 ( 0 3 s 0 1 1 0 - - s w 7 o w . l 0 2 5 f o 0 2 l 0 2 f k 1 2 1 a - k 0 1 0 7 6 e . a 0 0 0 1 1 p 3 9 e y - , 9 0 c 9 l 0 n p 0 - e a u l q e u a r 2 1 5 . 0 F 3 0 A 0 , n u e 5 0 n 0 c . . n 0 - n n n 3 1 1 M u e 3 4 5 , 9 e a d , a 3 M 1 2 n l e d 9 1 1 0 0 p 5 0 0 i e a 9 4 e 0 e 9 r a c n 9 , 9 0 a 4 0 0 c x d 8 4 k - 0 . R 5 - f E e i r s - 0 0 f 2 2 u a l o v n n e w – – i o E 1 3 s - v f 7 8 9 , x , f i 5 R P 9 c p a 9 9 2 2 6 l 8 e i 0 0 e R s 0 3 . , 0 9 9 o a a 3 e 0 3 - d 0 k e 0 - 1 1 d d s f l g e o , , e a a w ) ) n g r r e 1 2 c s , 9 6 e e o o 1 9 , e , 5 9 2 l l . 7 B 0 g g 9 0 - 0 F 1 e o o a a t d r o r e g g C D 2 o 0 q c 0 k u 4 . e g 1 . a 0 n . . g c e 6 8 y , 5 2 - - 8 6 4 2 1 0 0 0 0 0 0 9 1 1 6 0 7 0 0 0 0 0 5 2 0 0 0 0 0 A A 0 - - 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 e e

0 0 0 0 0 0 0

0 5 0 5 0 5 r r 0

3 (cfs) Peak Annual 2 2 1 1

5 Annual Peak (cfs) Peak Annual u u g g i i 0 . F F 1 0 . 2 0 ) 1 . 1 3 e s - g 4 u 2 5 l 0 a 4 , 0 2 3 p 0 , 0 4 g 1 2 0 9 - 2 9 . . 2 . 0 4 9 l 4 a 4 - 0 0 l 0 6 t 0 0 7 l , 0 e 2 2 2 e 5 2 b o 9 0 t - 0 5 0 8 D y 0 0 4 7 . . 0 - 8 ( 6 5 . 7 0 a 0 2 0 9 , 0 - 1 s . 4 2 5 M , 1 7 5 ( e 0 w 3 r , 0 0 0 7 4 g 2 o s 7 2 . . , 0 l 6 h 0 6 7 - 0 . 0 5 a f 1 9 w 0 0 2 3 p y - 7 0 o – 3 k c , m 0 l 5 2 o n 8 a f 0 6 - c e . . 9 e 6 n 7 9 k 0 7 u y 6 U 9 9 . p c 6 , 9 q a - 0 & n 1 5 1 l e e e n a , r a t a u f i l ) p 0 5 2 q . d e u 9 . e s 9 2 l e 8 e D r 0 1 , 0 8 , n , 0 5 c e F 3 a . 6 M 0 0 0 n w e - n g u c o e l a - n a f 9 1 n 0 4 3 d , 9 e 8 9 M 9 . 8 k 9 r . g 8 9 e n d , 9 9 0 3 a 0 9 e e , 9 - e c 3 d e R a e 4 8 1 e 0 i . x u p c a v 0 r n - i l r x n c E o 1 1 3 5 a f E s 0 9 g f e 8 0 . 3 u , 9 f R 4 . 4 0 8 1 0 p n v h 4 8 2 0 - 0 e i n n a a 0 A k 3 R . o p 0 f l 0 o s 2 a w i 2 m . s – p , n 0 o 7 M n c m a 9 2 y . u n b a 0 9 e 1 l 1 Y U G . 1 2 l 2 0 g 0 , 0 9 a 3 0 - g 0 e 1 , . . . 1 0 9 5 7 1 0 - - 7 t 1 1 1 1 o 0 0 0 0 0 0 0 0 0 6 9 , 9 2 0 0 0 0 0 0 0 0 4 A A 0 7 0 0 0 0 0 0 0 0 0 0 0 - 4 0 4 1 8 5 2 9 6 3 . e e

0 0 0 0 0 0 2 2 1 1 1

r r 0 0 0 0 0 0 (cfs) Peak Annual 0 0 0 0 u u

5 0 5 0 5

1 1 2 2 Annual Peaks (cfs) Peaks Annual g g i i F F

II-56 , ) e , g ) a e 4 0 g g 0 5 2 0 a / e / 1 1 g / d / 3 3 a k s c i l 3 o 4 0 a r 0 0 / 1 2 d P / / ( 3 e 1 / 3 s B f ( c 3 s 2 0 / f 0 0 1 0 / c 0 3 2 / 8 0 1 / 6 3 w 2 0 o w / l 1 1 / o 0 e l 3 0 b 2 e / s f 1 r b / c 3 e 0 1 r 0 0 v / e 8 i 1 / v . . 0 3 w s i R 0 f 4 5 o l c 0 R e 0 0 2 o 0 / b 6 0 0 1 s d / 0 d 2 2 w 0 3 e a o / i o r r – – 1 l r / e e 3 8 8 o o b p 9 l l , 9 9 , 9 e o o e 9 9 9 g g 1 a 1 1 C D / a 9 g 1 g 9 / / k 1 e 3 / . . c a 3 o d 0 2 r s i d 1 1 8 l e - - 9 a B 9 P 1 1 8 1 9 / / A A 1 1 0 0 0 0 0 / / 3 0 0 0 0 3 0 0 0 0

e e

8 6 4 2 6 4 2 r r

flow in cfs in flow u cfs in Flow u g g i i F F , s ) u 5 e l 0 4 0 g 0 p 2 / 0 a 1 2 / a / g 3 t 1 l / l l 3 e 4 e 0 . D 0 b ( 5 3 2 / y 0 1 0 / s 0 a 3 f 0 2 / c 2 1 M / ( 3 – 3 0 0 0 8 s 0 2 / f 9 2 1 0 / s , c 0 9 f 3 c 0 1 1 2 0 0 / , 0 2 0 1 ) 0 w / 0 1 4 0 3 s o 2 / l w e 1 / o w e l 1 g 3 e 0 o b a l b 0 2 r 1 g s e / s 0 e f e 1 0 b c / g e 2 v / a 3 r 0 i r 1 g 0 / g 4 e 3 . e R r 0 w v h 4 g 0 i o n l a 0 h 0 0 e 0 a 2 R o p / 0 0 b p 2 1 s / 2 d / a m i m 1 o 3 / o – i p n 3 c r o 8 n e n c m U p 9 9 9 , 9 u n a s 9 9 e 9 9 u l g 1 1 U Y G 1 / / a p 1 1 g / a / l t 3 l . l 3 e e . 1 b D 8 y 9 1 8 9 a - - 9 9 M 9 1 1 1 / 1 1 / / A A 0 0 0 0 0 0 1 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 / 5 0 5 0 5 0 5 0

0 8 6 4 2 3 e e

1 1 2 2 3 3 4 1 r r

flow in cfs in flow flow in cfs in flow u u g g i i F F

II-57 1 2 5 0 8 1 0 2 - 9 s . 0 5 s w 0 9 0 o . w 2 3 2 l 0 4 0 - f o 1 0 l 2 4 8 f - . 7 - 0 9 m 2 1 0 , 8 u m 4 . . 0 e 3 2 0 g 1 2 u 4 0 m a 2 0 i g 3 7 m . k - n i 0 c i o 7 n r - . i d m 0 3 e - . . 0 l B m 4 l 4 , . 0 y 4 5 s 0 l 6 6 2 c . a l 0 0 0 0 0 f w n y 2 2 6 / a o 0 - 0 l . e c r f f 4 2 n 0 u 2 / 8 7 0 2 e e 7 0 m 9 q r u – 1 – u q 9 1 e - e e m r r 7 1 2 m 8 i 5 F . n F i 0 - 9 9 e 5 m - . c m m e 5 M n 9 9 n l 1 0 4 s c e a a e u m f i 0 3 1 1 d d u n 2 c e c 4 d 0 i s - n 0 a , u f e e , 2 e 2 0 s n n 3 5 c ) ) d 0 r s - x 4 . 2 0 A M M - E e e 0 4 e 1 0 e 5 . m r e 4 3 2 g 0 g 3 i v c n e i i a a m x v E u i g 9 1 R g - - m 8 9 0 9 3 R 9 0 s . 3 8 e o k 2 5 . u 0 9 0 - m 3 2 0 d 9 0 c d s m 1 2 a a e o e r r r r s / f i a o o d l l l 2 l l 2 f . . e l a o 0 1 1 0 o o w , 9 2 P B 9 4 P 9 ( ( C D 0 a 1 1 - l i , 9 s 7 9 a 1 7 1 d 0 . . . e - 1 . 0 4 6 0 g a 1 1 g - - e , 1 1 1 9 9 0 A A 1 0 t o 0 0 0 0 0 0 0 e e 0 0 0 0 0 0 0 0 0 0 2 6 5 4 3 2 1 0 0 0 0 0 0 0 0 0

r r 0

8 6 4 2 0 8 6 4 2 0

1 (cfs) Minimum Annual 1 1 1 1 4 u u Annual Minimum (cfs) Minimum Annual g g i i F F 1 . 1 5 s 4 0 2 1 0 0 . 0 w 8 0 0 2 2 - o s 2 9 9 - . o l . 0 3 2 t 2 – f 0 0 0 w 3 0 7 0 8 4 0 0 7 o 5 2 9 - l 9 m 1 f , 9 2 u 8 0 8 e . . 1 0 r 0 4 0 m 5 2 g m , - - 0 0 i h ) 2 0 u - 2 3 a 1 n s - 2 0 p i - 4 m 2 0 e 0 3 7 m i 1 6 2 7 . . 0 3 o m 2 2 0 g 0 0 n 0 2 6 c 3 0 i 0 l a n 3 8 2 l U - . g y m 4 a c & 3 4 f l 6 / n - e a . 0 l 2 t 6 r 5 . r e l 0 1 0 0 a 0 e u e 0 y g f 0 , D c q 2 / 2 1 , n h s - e ) r e m – f r r u n f c a e q 5 e 7 a e . i m 8 b 1 1 r p 5 e d . 0 9 1 9 1 F m 0 c e 9 5 u 0 m e 9 8 c n 1 M m e - s - n c m 1 e 6 e 0 o 1 M e d 2 d r 1 , e 0 u e S 2 D 4 9 c 0 e e ) d u 0 . , 0 e c s i m 0 c o a 4 0 1 2 x . n e t 4 n v E m 0 x i r - - e g e E U r 2 n e / R 8 f a 6 u a 9 s v l J 0 l 3 l g i ( , 9 m . n u 1 1 1 1 0 l i 3 l n . R w 6 9 o l i 0 6 9 m o p l 9 e s u f a - i m b a p f m n 2 t l o y . u l w n 0 m a s 2 e m . - i , u 0 a M 3 n i D M 9 a 0 ( Y G ( y 9 M 3 b , 9 1 e . 1 1 l l 0 g . . 3 1 a 1 2 9 . g 3 5 0 9 0 e 7 , 1 1 1 - 9 - - 1 7 1 1 0 t o A A 0 0 0 0 0 0 0 0 0 0 0 0 0 6 2 8 4 e e 0 0 0 0 0 0

2 1 1

0 0 0 0 0 r r

1 2 3 4 5 Annual Minimum (cfs) Minimum Annual Minimum flow (cfs) flow Minimum u u g g i i F F

II-58 APPENDIX TWO

Fish Data Annual Species Composition, Density and Biomass Estmates for Each Study Site l a 5 3 4 a 9 6 5 4 0 4 5 9 4 6 6 a 6 4 8 ...... h 7 1 9 t 1 h / 5 8 2 n / 2 5 1 9 0 8 6 5 7 2 0 3 2 2 o 5 , , , g 8 5 4 6 6 5 6 4 2 6 6 4 2 2 n T 1 1 1 k n r e e 8 9 1 7 6 5 6 5 5 5 7 2 3 0 h k ...... 3 2 3 1 i t 0 1 0 1 2 0 0 0 1 2 0 0 0 0 r p o N h . t 4 u 0 o s 1 0 9 0 0 3 7 9 6 3 8 9 9 . . . s 0 3 8 4 ...... m 6 6 7 3 1 2 l a 1 3 1 4 5 6 6 1 2 2 4 1 1 l 2 1 1 – b a 8 m 9 S 9 1 , r p e r 5 8 3 8 5 5 3 2 3 3 1 8 1 5 7 ...... v 2 2 1 a i 1 0 0 1 1 0 1 0 0 7 6 2 6 1 1 ) C R d e a t p l c e e h l m l n s a i 6 5 3 9 8 3 5 3 1 9 4 n o ...... f 3 4 3 4 4 6 3 t Y a c 2 5 2 3 1 4 6 3 4 3 3 a , h ) h c y e C s ) f r i f f e a r t u l l . a c r t D a e e c f r s h k e o o d r c i h f r 2 4 0 7 8 2 9 7 9 7 7 e e u r ...... 1 2 2 0 6 8 5 b s p g s 9 9 3 9 0 1 9 9 6 2 0 e 7 7 7 5 6 3 3 y a e 3 3 3 2 2 1 3 3 2 1 1 p s e t t h t i r n a m + e e h a m c b r i r W t g m e s o u l p i e ( n k w ( ( n s o o y o s s d t n i i t a s a n i 6 9 8 6 5 8 9 8 6 1 4 5 6 9 7 2 s r i a ...... 0 0 s m n o 1 0 0 0 0 2 2 1 2 1 1 2 1 0 0 1 l m o m e o e o i p o D i k b C i m B p o d c n l s a i e a i y t t c b i d 3 9 6 5 1 5 7 6 2 1 3 e u ...... s 4 3 4 3 2 5 0 n p h 1 0 0 0 0 2 2 1 2 1 0 n u c S e o d R , n o d i r a t i e e 5 6 7 0 7 5 1 8 3 2 5 2 1 k s h ...... 4 7 4 4 1 c o e 0 3 3 4 2 0 0 1 1 2 1 0 0 u u p l s m B o c h t s u e i r o c e m e 0 4 8 7 9 0 7 7 2 8 1 2 k l ...... 6 5 5 2 4 3 c p e 4 2 1 0 0 1 4 2 2 0 1 1 u n S s n a l F . 1 - 2 A 8 9 0 1 3 4 8 9 0 1 3 4 8 9 0 1 3 4 9 9 0 0 0 0 9 9 0 0 0 0 9 9 0 0 0 0 r e l 9 9 0 0 0 0 9 9 0 0 0 0 9 9 0 0 0 0 a b 1 1 2 2 2 2 1 1 2 2 2 2 1 1 2 2 2 2 e a T Y

II-59 l a 1 0 4 1 1 7 1 2 4 7 a 5 4 5 5 ...... a 7 6 2 1 . . . . h 9 4 t h / 9 9 2 4 3 2 5 0 0 7 3 3 n / 2 8 7 3 3 0 o , , 7 7 2 0 7 6 6 1 g 8 6 8 7 8 9 8 7 n T 1 1 1 1 1 1 1 1 1 1 k n r e e 3 8 2 3 4 3 6 6 3 4 0 2 4 6 4 8 4 5 h k ...... i t 1 1 0 0 0 0 3 4 0 0 1 0 1 2 0 0 0 0 r p o . N 4 0 0 2 h – t 8 u 9 o s 8 5 0 5 5 4 6 5 8 2 9 2 9 3 7 . . 9 s 3 3 ...... m 5 2 4 l a 1 1 2 1 2 0 2 5 0 5 1 1 0 0 5 4 l 3 2 b , a r m e v S i R ) a d p e p 6 5 4 8 6 0 0 5 t 1 1 6 9 ...... r . . . . m c 4 5 4 4 4 5 2 0 5 0 1 7 8 0 a 9 4 5 6 e a 1 1 2 3 1 1 2 2 l C l Y o , c s ) h n e s ) e r i l f e a v e r t h l e l n a c s 8 8 5 4 2 7 8 t i ...... 0 5 5 4 2 a S n e . . . . . f 7 7 2 5 3 6 c 4 9 1 2 0 4 1 t f a 2 7 9 6 3 h r e 1 1 2 2 1 1 1 a o h r o h c f e C e r p g s e a e p s t t r r n m a e e e a s k b c r m r i c d g i t m e u r 4 2 7 8 8 3 9 2 4 3 6 0 o s ...... l u p 1 5 7 6 6 6 S b i ( 0 9 6 0 7 0 2 3 4 2 4 0 e n 1 1 1 1 3 3 y k ( 3 2 1 2 7 2 2 2 1 1 2 1 e ( n t h s i y o s t s i h + i t s a i s a W s n m o m e o p o i D i l i b m B a o t d b c 6 d 0 4 7 5 5 3 9 9 0 8 1 . n u ...... 5 4 4 3 4 2 n s 0 a h 6 6 3 3 1 6 3 3 2 1 1 e u 1 i c y o c t i e R s p n S e d d r a , e e 8 9 9 9 2 5 8 n 6 4 0 6 0 ...... k 3 8 2 3 h . . . . . 2 2 o 2 6 6 7 7 4 4 c 2 1 2 1 e i 1 2 5 0 1 4 3 4 1 1 1 1 u t u i l s s B o p m h t o u c r o e s 2 5 4 3 6 0 1 1 0 2 4 1 m ...... k 8 6 0 3 6 6 l e i 3 0 7 2 0 9 6 6 0 9 5 4 c 4 4 5 5 3 2 e c 6 6 4 4 3 1 6 6 5 3 3 2 u n e s n p a l S F . 2 - 2 8 9 0 1 3 4 8 9 0 1 3 4 8 9 0 1 3 4 9 9 0 0 0 0 9 9 0 0 0 0 9 9 0 0 0 0 A 9 9 0 0 0 0 9 9 0 0 0 0 9 9 0 0 0 0 r 1 1 2 2 2 2 1 1 2 2 2 2 1 1 2 2 2 2 e a l e b a Y T

II-60 5 l . a 9 8 9 4 9 2 1 8 3 a . . . . . a 0 3 1 h 5 8 5 4 t / 8 7 1 6 0 h 8 5 0 n 1 / 0 9 2 o , , , , 2 4 2 5 4 g 6 2 0 n , T 4 2 2 2 5 5 5 2 2 k 1 n r e e 2 2 1 1 6 4 4 0 8 8 6 7 h k ...... i t 0 0 0 0 2 1 1 1 3 2 2 1 r p o . N 4 0 0 h t 2 u – o s 0 6 5 7 . . . 8 8 0 2 2 s 9 0 . . . . . m 5 5 9 0 1 7 l a 3 0 6 8 7 8 0 l 1 5 5 b a 2 , m r S e v i R p 8 1 6 1 5 2 7 ...... r a 6 2 2 3 3 6 3 4 9 1 8 1 a 2 p 1 3 2 2 2 3 3 ) C m d e a t Y c l , e e l h l k 3 3 6 2 2 n 9 s 7 . . . . . r o . i . 0 8 8 7 4 n 3 7 5 3 4 f a c 1 3 t 4 1 2 2 6 a 1 3 7 9 2 6 5 P a ) 5 1 2 1 2 h h c e s ) y C r i l f e a i r t l L l a c t a e r c f h o e + s f o r h r e d t e e i s e i r 3 2 6 3 9 3 2 4 2 5 1 r ...... p g 4 e k e h b t 0 0 0 1 0 0 2 0 0 2 1 a c p s t y a W u r n h m s e m e a i b c r t r g s m e o e u l p i ( n s w k ( ( n o s o y o d s a n t i i t a 3 s n i 1 1 5 4 3 6 7 6 3 r s i m ...... 0 a 0 0 s . o n 0 0 0 0 0 2 0 0 5 o l m 0 o i m e e o p o b D i k C i m B d p o n c a s l i y e i t a i t c b 2 3 1 s d 3 4 3 6 2 6 e u ...... 0 0 0 0 0 0 n n . . . p h 0 0 0 3 0 0 u 0 0 0 e S c o d R , n o i d t r i a e s e 8 8 1 1 2 8 9 1 ...... k o h 9 7 8 6 6 4 7 7 7 9 8 7 c e p 9 5 5 4 3 1 1 1 u u l s m B o c s h t e i u c r o 5 5 5 3 4 3 4 3 e e ...... m k p 8 8 5 2 l 7 7 0 8 8 3 2 0 c 4 6 5 5 e 4 7 5 0 1 4 2 2 S u n 3 2 2 2 2 1 1 1 s n a l . F 3 - 2 A r 0 1 3 4 0 1 3 4 0 1 3 4 e a 0 0 0 0 0 0 0 0 0 0 0 0 l e 0 0 0 0 0 0 0 0 0 0 0 0 b Y 2 2 2 2 2 2 2 2 2 2 2 2 a T

II-61 l 9 9 4 7 8 a a a 3 1 7 7 1 0 3 8 1 7 7 9 8 6 8 t h h / 4 9 5 6 8 9 6 4 1 9 n / 2 4 7 6 8 o , , , , , 8 6 8 9 7 5 4 5 7 6 g n T 3 2 3 4 3 k d a e 3 6 3 7 6 . . . . . h l 1 0 1 0 0 l u B l l ) a ( s e h i 9 5 2 7 4 ...... s c i 4 0 1 1 2 1 f e 0 n p 0 u s 2 S – 9 9 9 1 p 6 0 5 6 7 2 6 0 9 r . . . . , 1 4 2 2 7 7 4 1 3 7 0 r 8 9 9 3 a 1 1 1 3 1 1 1 1 1 8 9 4 7 e C v i ) d R e t l o c e h d e n l s 9 1 4 3 a 1 2 l i . . . . 4 1 1 0 r n f 4 6 5 7 9 7 9 o 6 4 4 8 t 3 4 3 7 a o 2 1 c 3 5 7 7 a l h c ) o h C e s ) C r i f e a , r t l e l a c r t k a e e c f a s k h e o L d c r i h u r 3 1 0 7 e e 5 . . . . r n 0 1 0 3 1 2 s b 6 4 6 8 0 p g r 8 3 5 4 e 1 1 2 2 3 3 y 1 a e 4 2 9 7 o p s t t h i r n C m h + e e a r c b r W o r g f m e o u l p s i ( n e k ( t ( n w a y o o o s t i i d t n s m i s a 4 3 2 3 i n a 1 4 6 6 0 0 s r t i ...... n 0 0 0 0 . . . . o o s m 0 1 1 0 1 1 e l m 0 0 0 0 p e o D e o i k m s C B i o s p c a s m e i o l i c i e a b t p b 9 9 2 3 d . . . . d 1 8 7 u S 3 4 3 5 5 9 7 n 0 6 9 9 n 1 3 1 h 3 5 2 5 u a c o y R t i s n e d r a d e e , 3 3 2 8 4 0 1 3 k 5 6 7 1 8 9 0 h 6 1 3 5 3 6 2 0 n c 3 3 3 3 2 9 8 e 3 2 2 2 1 1 1 1 o u u i l s t i B s o p h m t o u r c o e 1 5 9 9 4 3 5 4 9 7 s m k 8 1 0 2 8 l 1 1 6 7 0 5 6 8 9 6 e c 3 3 4 3 3 e i 3 2 3 3 3 2 1 2 2 2 u n c s n e a p l F S . 4 - r 2 9 0 1 3 4 9 0 1 3 4 9 0 1 3 4 a 9 0 0 0 0 9 0 0 0 0 9 0 0 0 0 A e 9 0 0 0 0 9 0 0 0 0 9 0 0 0 0 Y 1 2 2 2 2 1 2 2 2 2 1 2 2 2 2 e l b a T

II-62 l 6 5 8 a a a 7 9 8 4 4 5 7 8 5 t h h 1 8 1 5 2 0 n / / 2 4 5 o , , , 6 7 7 4 5 6 n g T 3 4 3 k d a e 2 4 0 . . . h l 0 0 1 l u B l l ) a ( s e h i 2 1 0 . . . s c i 1 1 4 f e n p . u s 3 S 0 0 2 – 0 p 4 9 3 3 r 0 2 4 2 2 1 4 2 8 9 a 1 1 1 9 8 0 1 1 1 1 C 2 , ) r d e e v l t i e c h R n e s 5 8 5 l i . 1 7 6 5 7 6 l n . . f o 3 2 0 t 8 8 8 6 a o 5 5 1 1 d 1 a c h a c C r h ) o s ) e i l r f e o r a l t r l a C t c e a , s c e k f e d c n h o i h u r r o 7 1 0 e 5 3 6 4 0 6 t . . . s r b e f 4 3 4 1 1 1 g 3 4 8 i y e p e l a t h p t i s C r n h + e m r e W a b c o r r f m g e s u o p l ( e i n t ( w k n a ( o o y o t d i n s i m t a 3 9 i i n s s 1 4 3 6 7 r t i . . . . . 0 0 s 0 0 a n . . o s 0 0 0 0 0 o l m e 0 0 m e p e o D o k i s C i m s B p o a c s m e o i l i i c a b e t b p d d 1 9 0 3 9 3 3 4 4 u . . . n S n 6 5 5 1 1 1 h 5 5 6 u a c o y R t i s n e d r a d e e , 7 1 7 k 1 7 5 7 9 5 h 8 0 5 n c 4 2 2 6 7 6 e 1 2 1 o u u i l s t i B s o p h m t o u r c o e 6 1 2 0 1 2 s m k 3 2 3 l 9 7 5 9 7 9 e c 3 4 3 e i 2 2 2 1 1 1 u n c s n e a p l F S . 5 - r 2 0 1 3 0 1 3 0 1 3 a 0 0 0 0 0 0 0 0 0 A e 0 0 0 0 0 0 0 0 0 Y 2 2 2 2 2 2 2 2 2 e l b a T

II-63 l 5 1 5 9 a a a 4 7 4 7 8 7 1 t h h 9 6 8 n / / 4 3 5 1 o , , , , 8 5 3 n g T 1 4 3 1 k d a e h l l u B l l ) a ( s e h i . s c i 5 f e 0 n p 0 u s 2 S – 3 0 0 2 p 2 3 6 2 r . . . , 2 0 0 . . . 4 r 5 8 5 a 6 3 3 1 e 7 3 7 C v i ) R d e n l t e c o h n e s s l i i l n f t n a o a c n h c u C h ) s ) e G i r f e , r a l e t r l a t t c e a n s c e k f e d a c h o i l h u r 9 1 3 9 r 9 2 6 . . . . a e . s r b e 4 3 2 7 5 4 c g 9 y e 1 1 p e 1 1 6 6 s a t h p t i s E r n h + e m r e W a b c o r r f m g e s u o p l ( e i n t ( w k n a ( o o y o t i d n s i m t a i i n s s 1 r t i . s a n o s 0 o l m e m e p e o D o k i s C i m s B p o a c s m e o i l i i c a b e t b 5 4 3 0 1 p 3 0 d . . . . . d u 0 6 n S 2 3 5 0 7 n h 3 2 1 1 1 4 3 u a c o y R t i s n e d r a d e e 8 5 9 5 , 2 8 6 . . . . k h 0 5 3 n 1 8 5 0 c e 5 3 1 4 4 4 9 o u u i l s t i B s o p h m t o u r c o e 7 4 0 7 1 4 5 s m . . . k l 6 0 7 1 e 8 0 6 c e i 2 2 1 1 2 2 1 u n c s n e a p l F S . 6 - r 2 3 4 5 3 4 5 3 4 5 a 0 0 0 0 0 0 0 0 0 A e 0 0 0 0 0 0 0 0 0 Y 2 2 2 2 2 2 2 2 2 e l b a T

II-64 l a 2 3 9 6 a a 9 4 7 h 2 7 3 9 t h / 3 7 7 n / 6 7 4 4 o , , , , g 7 6 6 n T 1 2 3 1 k d a e h l l u B l l ) a ( s e h i s c i f e n p u s . S 5 0 0 2 – p 4 9 9 0 3 r . . 5 7 0 . . . 7 5 a 4 6 0 6 9 6 1 1 9 8 0 C 2 ) , r d e e l t v e i c h n e s R l i l n f t n a o a c h o c s ) C h i e s ) r n i f e a n r t l r l u a c t e a e G s c k f h , e d c o i r a h r u 7 5 0 0 4 3 2 . . . . . t e e s r b l 5 1 2 4 9 0 0 p g y e 1 2 e e 2 2 1 8 9 a t h p s t i D r n h + m r e e a W o b c r f r g m e o s u l p i e ( n t k ( w ( n a o o y o s t d i n m i t s i a i n s 3 t r a i . s n s o 0 o l m m e e p e o o D i k s C i m B s p o a c m s e o i l i i c b a e t b 1 0 0 p 8 d d . . . 2 7 1 u . . . 6 n S 6 5 0 n h 6 3 8 2 1 2 4 u a c o y R t i s n e d r a d e e , 7 1 5 8 0 3 0 . . . . k h n 3 9 8 2 0 0 1 c e 1 5 1 o 2 2 4 6 u i u l t s i B s o p h m t o u c r o e 2 1 3 s 7 8 8 3 m . . . k l e 6 5 4 1 1 0 6 c i e 3 2 3 2 4 4 1 u c n s e n a p l S F . 7 - r 2 3 4 5 3 4 5 3 4 5 a 0 0 0 0 0 0 0 0 0 A e 0 0 0 0 0 0 0 0 0 Y e 2 2 2 2 2 2 2 2 2 l b a T

II-65 l 7 3 5 8 a a 9 2 3 9 . . . . a 3 2 0 4 . . . . t h h 9 4 5 3 0 6 2 8 n / / 1 9 2 7 o 1 5 1 8 5 5 4 4 4 1 2 2 n g T 1 1 1 2 k d a 5 1 1 e . . . 0 5 5 1 9 . . . . . h 4 1 6 l 5 0 9 8 0 l 4 6 1 6 1 8 u 9 . . . . B 0 0 2 0 l l ) a ( s e h i 0 4 2 . . . s 0 c i 2 1 0 f e . n p 5 u s 0 S 0 2 – 0 0 p 5 2 0 r . . 4 0 1 4 7 2 3 9 6 4 ...... 2 7 4 a 3 2 7 1 8 3 8 2 2 4 1 1 , C r e ) v d i e l t R e c h s n e s 7 7 2 8 3 3 9 l i ...... 1 2 0 7 2 e l n . . . . . f 6 0 1 8 6 0 0 r t a o 9 5 5 8 3 1 4 3 2 1 1 1 a o c h l c C o h ) s ) e D i r f e , r a l t r l a m t c e a s c e u k f e d c h s o i h u r r p e s r b e y g y e p e a G t h p t i s r n g h + i e m e W a b c B r r r m g e o u o p l f ( i n ( w k n s ( o o y o e t i d n t s i t a i n s a s r i s a n o m o l m e i m p e t o D o k i s C i m e B p o c s s s a e i l i c m a e t o b 7 8 5 8 3 9 6 4 i p d ...... 0 4 6 2 u . . . . b n S 4 5 9 4 3 3 4 8 h 3 2 1 5 5 2 3 2 5 4 3 7 u c d o n R a y t i d s r a n e e e 2 2 2 6 5 0 2 6 1 k h ...... 0 0 0 c d e 2 6 1 1 6 1 0 0 0 u , u l s n B o i t i s o h t p u r o m e 7 0 1 0 o 0 m 4 2 2 2 4 6 . . . . k 6 l ...... c 5 5 5 9 0 c 7 e 2 3 4 3 0 6 1 1 5 5 2 u s n s e n i a l c F e p S . 8 - r 2 0 1 4 5 0 1 4 5 0 1 4 5 a 0 0 0 0 0 0 0 0 0 0 0 0 A e 0 0 0 0 0 0 0 0 0 0 0 0 Y 2 2 2 2 2 2 2 2 2 2 2 2 e l b a T

II-66