Monday, September 22, 2003

Part III

Department of the Interior and Wildlife Service

50 CFR Part 17 Endangered and Threatened Wildlife and Plants; Notice of Remanded Determination of Status for the ( macrolepidotus); Final Rule

VerDate jul<14>2003 14:42 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00001 Fmt 4717 Sfmt 4717 E:\FR\FM\22SER2.SGM 22SER2 55140 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

DEPARTMENT OF THE INTERIOR (telephone: 916/414–6600; facsimile: since been drained and reclaimed. 916/414–6713). Information is available Splittail were present within Buena Fish and Wildlife Service in alternate formats upon request. Vista and Kern Lakes (Moyle 2002), both SUPPLEMENTARY INFORMATION: of which are reclaimed. 50 CFR Part 17 Some researchers (Sommer et al. Background [RIN 1018–AH73] (1997)) indicate that splittail still occur, The Sacramento splittail (hereafter at least during optimal conditions, Endangered and Threatened Wildlife referred to as splittail) is a fish species through as much as 78 percent of their and Plants; Notice of Remanded native to central and former range in terms of river reaches. Determination of Status for the represents the only extant species in its However, others (Moyle and Yoshiyama Sacramento splittail (Pogonichthys genus in North America. We have 1992) believe the species appears to be macrolepidotus) previously discussed the taxonomic restricted to a small portion of its former history of the splittail along with the range, with dams and diversions AGENCY: Fish and Wildlife Service, physical description of the taxon in our preventing access to upstream habitat in Interior. final listing rule (64 FR 5963). Please large rivers and streams beyond the ACTION: Final rule; revised refer to that document for a detailed valley floor (Moyle and Yoshiyama determination. discussion of these subjects. It is our 1992). The State of California indicates intent, in this document, to reiterate and that splittail still occur in a large portion SUMMARY: On January 6, 1994, we, the discuss only those topics directly of its range (80% in the Sacramento, and U.S. Fish and Wildlife Service (Service) relevant to this decision. 70% in the San Joaquin). There appears proposed to list the Sacramento splittail To assist the reader in understanding to be consensus that at least 20% and (Pogonichthys macrolepidotus), a fish terminology used in this determination, possibly more of the species range has species native to central California, as a we have provided below several terms been reduced. Baxter (2001b) found that threatened species under the with their corresponding definitions as the range of the splittail extends away Endangered Species Act of 1973, as they are used in this document. As used from the Delta, though detections on the amended (Act). We published a final in this determination, the term ‘‘Delta’’ periphery of its range appear to be part rule to list the species as threatened on refers to all tidal waters contained of a single, mobile, Sacramento and San February 8, 1999. Our final decision to within the legal definition of the San Joaquin River/Bay-Delta population that list the Sacramento splittail was Francisco Bay-Sacramento-San Joaquin includes fish from the Napa and subsequently challenged in the cases River Delta, as delineated by section systems. Their San Luis & Delta-Mendota Water 12220 of the State of California’s Water distribution in the Estuary suggests that Authority v. Anne Badgley, et al. and Code. Generally, the Delta is contained brackish water may characterize optimal State Water Contractors, et al. v. within a triangular area that extends rearing habitat for fish greater than 75 Michael Spear, et al. On June 23, 2000, south from the City of Sacramento to the millimeters (mm) (3.0 inches (in)) the Federal Eastern District Court of confluence of the Stanislaus and San standard length (SL) (Moyle et al. 2001). California found our final rule to be Joaquin Rivers at the southeast corner Suisun Marsh includes the largest areal unlawful and on September 22, 2000, and Chipps Island in at the extent of shallow water habitat available remanded the determination back to us southwest corner. The term ‘‘Estuary,’’ to the splittail and likely has the greatest for a re-evaluation of our final decision. as used in this determination, refers to concentrations of the species. However, because the District Court did tidal waters contained in the Splittail are relatively long-lived and not vacate our previous final decision, Sacramento and San Joaquin Rivers, the larger fish may be 8 to 10 years old the decision remained in place until we Delta, and San Pablo and San Francisco (Moyle 2002). Splittail reach about 110 issued a new determination. After a bays. ‘‘Export facilities,’’ as used in this mm (4.3 in) SL in their first year, 170 thorough review and consideration of determination, refers to the Bureau of mm (6.6 in) SL in their second year, and all the best scientific and commercial Reclamation (USBR) Central Valley 215 mm (8.4 in) SL in their third year information available, we are removing Project (CVP) and the California (Moyle 2002). Male and female splittail the Sacramento splittail from the list of Department of Water Resources (CDWR) may mature by the end of their second threatened species. In accordance with State Water Project (SWP) water export year (Daniels and Moyle 1983), but the Administrative Procedure Act, the facilities in the South Delta. some males mature in their first year Service has determined that this rule Splittail are native to California’s and some females do not mature until relieves an existing restriction, and good Central Valley. Historically, splittail their third year (Caywood 1974). cause exists to make the effective date were found as far north as Redding on The largest females can produce over of this rule immediate. the Sacramento River (Rutter 1908). 250,000 eggs per year (Daniels and EFFECTIVE DATE: In compliance with the Splittail were also found in the Moyle 1983). Other and more current Federal Eastern District Court of tributaries of the Sacramento River as estimates of splittail fecundity have California order, this rule is effective far as the current Oroville Dam site on shown high variability and occasionally, September 22, 2003. the Feather River and Folsom Dam site lower numbers. Caywood (1974) found ADDRESSES: Comments and materials on the American River (Rutter 1908). a mean of 165 eggs per mm (6.5 in) of received, as well as supporting Along the San Joaquin River, historic SL of fish sampled and reported a documentation used in the preparation distribution is unclear. Girard (1854) maximum of 100,800 eggs in one of this final decision, are available for reported two Pogonichthys species in female. Daniels and Moyle (1983) public inspection, by appointment, the San Joaquin River. These reports do observed approximately 17,500 to during normal business hours at the not make a distinction between which 266,000 eggs per female splittail. Feyrer Sacramento Fish and Wildlife Office, of the two species was found at and Baxter (1998) found a mean of 261 U.S. Fish and Wildlife Service, 2800 particular locations on the San Joaquin eggs per mm (10.2 in) of SL and Cottage Way, Suite W–2605, River. In the southern Central Valley, estimated maximum fecundity at Sacramento, CA 95825. Tulare Lake was likely to have 150,000 eggs. Bailey et al. (1999) FOR FURTHER INFORMATION CONTACT: supported many native fish species, examined fish held for a considerable Wayne White (see ADDRESSES), including splittail (Moyle 1976) but has time in captivity and found that

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00002 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55141

fecundity ranged from 24,753 to 72,314 hatch are required to reach this size (Micropterus salmoides) and other eggs per female, which agrees with class. centrarchids, and other native and non- Caywood’s (1974) observations. It is speculated that Suisun Marsh is native piscivores (Moyle 1976, Moyle Although primarily a freshwater the likely late stage rearing area for 2002a). Introduced, non-native benthic species, splittail can tolerate salinities juvenile splittail hatched and reared in foragers such as shokihaze goby as high as 10 to 18 parts per thousand the extensive spawning habitat found (Tridentiger barbatus), chameleon goby (ppt) (Moyle 1976; Moyle and within the Yolo Bypass, as a hydrologic (T. trigonocephalus), and yellowfin goby Yoshiyama 1992). Salinity tolerance in connection apparently exists between (Acanthogobius flavimanus), may feed splittail increases in proportion to these waters (N. Monsen, unpubl. data on splittail eggs. Introduced length; adults can tolerate salinities as referenced in Moyle et al. 2001). planktivorous, threadfin shad high as 29 ppt for short periods (Young Splittail use of Suisun Marsh varies (Dorosoma petenense) and inland and Cech 1996). Splittail populations with outflow (Baxter 1999a). silverside (Menidia beryllina), compete fluctuate annually, depending on Splittail are benthic foragers. In directly with larval and juvenile splittail spawning success, which is well Suisun Marsh, adults feed primarily on for food. Other non-native cyprinids, correlated with freshwater outflow and opossum shrimp (Neomysis mercedis, such as golden shiner (Notemigonus the availability of shallow water habitat and presumably, non-native shrimp crysoleucas), red shiner (Notropis with submerged vegetation (Daniels and species of the genus Acanthomysis as lutrensis), and fathead Moyle 1983; Sommer et al. 1997). Fish well), benthic amphipods ( (Pimephales promelas) are also likely to typically reach sexual maturity by the spp.), and other small crustaceans, compete with splittail. In recent years, end of their second year. The onset of although detrital material makes up a splittail have been found most often in spawning is associated with rising water large percentage of their stomach slow moving sections of rivers, sloughs, levels, increasing water temperatures, contents (Daniels and Moyle 1983). In and in dead end sloughs (Moyle 1976, and increasing day length. Peak the Delta, clams, crustaceans, insect Daniels and Moyle 1983). Reports from spawning occurs from February through larvae, and other invertebrates also are the 1950’s, however, mention May, although records of spawning exist found in the adult diet. More recently, Sacramento River spawning migrations for late January to early July (Wang research has indicated a shift in adult and catches of splittail during fast tides 1986). In some years, most spawning splittail diet towards the non-native in Suisun Bay (Caywood 1974). Current may take place within a limited period Asiatic clam (Potamocorbula amurensis) accounts place splittail as far upstream of time. For instance, in 1995, a year of in Suisun Marsh. as the Red Bluff Diversion Dam on the high spawning activity, most splittail Historically, Eurytemora affinis, the Sacramento River (Baxter 1999a). spawned over a short period in April, native euryhaline copepod, has been the Splittail have been recorded in recent even though larval splittail were most important food for larval in times from within Salt Slough and at the captured from February through early the Estuary. Three non-native species of Merced River confluence on the San July (Moyle et al. 2001). Within each euryhaline copepods (Sinocalanus Joaquin River, and within the Napa and spawning season older fish reproduce doerrii, Pseudodiaptomus forbesi, and Petaluma Rivers (Baxter 1999a, 1999b; first, followed by younger individuals Pseudodiaptomus marinus) became USACE 2002a, 2002b). (Caywood 1974). established in the Delta between 1978 Splittail are frequently found in areas Splittail spawning occurs over and 1987 (Carlton et al. 1990), while subject to flooding because they require flooded vegetation in tidal freshwater native E. affinis populations have flooded vegetation for spawning and and brackish water habitats of estuarine declined since 1980. It is not known if rearing. Historically, the major flood marshes and sloughs and slow-moving, the non-native species have displaced E. basins (e.g., Colusa, Sutter, American, shallow reaches of large rivers. affinis or whether changes in the and Yolo basins; Tulare, Buena Vista, Observations of splittail spawning have estuarine ecosystem now favor S. doerrii and Kern lakes) distributed throughout indicated spawning at depths of less and the two Pseudodiaptomus species. the Sacramento and San Joaquin valleys than 1.5 meters (m) (4.9 feet (ft)) in the Meng and Orsi (1991) reported that S. provided spawning and rearing habitat. Cosumnes River floodplain (Moyle et al. doerrii is more difficult for larval striped These flood basins have all been 2001), and at depths of less than 2 m bass to catch than native copepods reclaimed or modified for flood control (6.6 ft) in Sutter Bypass (Moyle et al. because it is fast swimming and has an purposes (i.e. as bypasses), and much of 2001). Sommer and Harrell (1999) effective escape response. It is not the floodplain area adjacent to the rivers postulated that individual splittail may known if this difference in copepod is now inaccessible behind levees. The not spawn in the year following a swimming and escape behavior has Yolo Bypass may approximate some of successful effort. affected the feeding success of young the Yolo Basin’s former role, and the Splittail larvae remain in shallow, splittail. Zeug et al. (2002) and Hieb Butte Creek, Butte Sink, Sutter Bypass weedy areas close to spawning sites for (2002) reported a high abundance of an system remains somewhat intact. Meng 10 to 14 days and move into deeper introduced, predatory Palaemonid and Moyle (1995) reported that the core water as they mature and swimming shrimp (Exopalaemon modestus) in the distribution of splittail extends from ability increases (Wang 1986; Sommer et Yolo Bypass and Delta. It is not known Suisun Bay and Marsh through the al. 1997). Bailey (1994) has documented what effect(s) this invasive species will western Delta. that splittail eggs hatch in 3 to 5 days have on the trophic (food) pyramid of The Yolo and Sutter bypasses and the at 18.5 degrees centigrade (°C), (65.3 the estuary, though Moyle (2002b) Cosumnes River floodplain serve as degrees Fahrenheit (°F)). Bailey (1994) speculates it is likely to prey on mysid important splittail spawning and early also found that at 5 to 7 days after shrimp and thus, may compete with rearing habitat (Sommer et al. 1997), as hatching, the yolk sac is absorbed and splittail for food. Juvenile feed mainly they approximate the large, open, the diet begins to include small rotifers. on plankton composed of small shallow water areas which have been Moyle et al. (2001) states that splittail of (zooplankton), and then small extensively reduced. The Yolo and 20 to 25 mm (0.8 to 1.0 in) total length crustaceans and insect larvae as body Sutter bypasses provide good habitat for (TL) ‘‘* * * are essentially small size increases. fish, particularly splittail, when flooded juveniles, capable of fairly active Predators of splittail include striped for several weeks in March and April. swimming’’ and that 4 to 5 weeks post- bass (Morone saxatilis), largemouth bass To provide the best spawning

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00003 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55142 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

conditions for splittail, water must University of California (UC) Davis’s areas) and then calculating an area remain on the bypasses until fish have Suisun Marsh Otter Trawl (Suisun weighted average catch from each area; completed spawning, and larvae are Marsh OT); (5) Service’s Chipps Island the index is the sum of these area able to swim out on their own, during Trawl survey (Chipps Is. Trawl); (6) fish weighted mean catches. The annual Fall the draining process. The Cosumnes salvage operations (which repatriate fish MWT Index is the sum of the four River also possesses natural and taken from water intake screens) at the monthly indices. Splittail lengths were restored floodplain features. This river CVP Tracy Fish Collection Facility not recorded until 1975, so for data is unique in that it is not dammed and (CVP); and (7) fish salvage at the SWP collected prior to 1975, Young Of Year the hydrograph is relatively natural. The Skinner Delta Fish Protective Facility in (YOY) (age 1) fish could not be contributions made by this habitat are the south Delta (SWP). differentiated from other age classes. somewhat limited by the fact that the Four other sampling programs Fall MWT data from 1967 through 2002 Cosumnes River watershed is lower in provide additional splittail information was used in our abundance analysis. elevation than most adjacent rivers. It is but the data are insufficient to support Studies therefore somewhat less dominated by useful indices. These are: (1) Service’s the extended spring peak flow Delta Beach Seine Survey; (2) CDFG’s The San Francisco Bay Studies characteristic of a higher altitude Summer Townet Survey; (3) U.S. Army sample waters west of the Delta seaward watershed with greater snowmelt Corps of Engineers’ (USACE) to south San Francisco Bay using both potential. Survey; and (4) CDFG’s Creel Census. a midwater trawl (Bay Study MWT) and In summary, the current distribution an otter trawl (Bay Study OT) (Baxter Surveys Employed in Abundance of splittail habitat is certainly reduced at 1999a). These programs capture Analyses least 20% and may be much more relatively few splittail, but are still reduced in extent from that which may The data available even today on considered important because they have historically been present. Clearly, splittail abundance are not optimal. involve two types of sampling perhaps the largest portion of the There are a number of survey programs equipment and frequent sampling splittail’s habitat is contained in the which generate data, each of which have (Baxter 1999a). Much of the sampling natural and newly restored floodplains more or less limiting factors. This has takes place in San Francisco Bay in of the Cosumnes River, managed made analysis of the status of the deep water channels that are not floodplains such as the Yolo and Sutter species based on this survey data characteristic splittail habitat. Monthly bypasses, disjunct segments adjacent to problematic. Descriptions of all fisheries indices are calculated as the sum of the Sacramento and San Joaquin rivers sampling programs that routinely detect regional volume-weighted average catch and in lower reaches of their respective splittail follow, and are differentiated per 10,000 cubic meters (m3) (353,147 tributaries. into two categories: those that were used cubic feet (cf)) for the Bay Study MWT In years where the Yolo and Sutter in the calculation of abundance indices and the sum of regional area-weighted bypasses are not sufficiently inundated, and those that were not. average catch per 10,000 m 3 (353,147 splittail spawning is confined primarily Fall Midwater Trawl Survey cf) for the Bay Study OT (Sommer et al. to the natural and newly restored 1997). During the 1997 index period, the floodplains of the Cosumnes River and The Fall MWT was initiated by CDFG Bay Study MWT collected only one the margins of rivers and other in 1967 to sample striped bass, a non- YOY, and the Bay Study OT collected floodplain features that are inundated at native sport fish. In addition to striped none at index stations. The tremendous lower river stages. These areas likely bass, CDFG has maintained records of variability in this survey’s catch is likely represent only a fraction of the area other fish species captured in the due to the rare or limited occurrence of which was historically subject to samples in most years. This monitoring individuals splittails at the periphery of inundation; levees preclude access to program currently samples 100 sites its range, which would result in limited reclaimed floodplains and basins. There from San Pablo Bay in the west to Rio detectability during sampling. Splittail are indications, based on presence of Vista on the lower Sacramento River can be expected to be captured in San larvae and juveniles, that spawning in and to Stockton on the San Joaquin Francisco and San Pablo bays only the Sacramento River occurs relatively River. Data are collected from during time of infrequent, high outflow, far upstream at Colusa (Baxter 1999a; September through December using a when captures appear to increase for all 1999b). Splittail appear to utilize the midwater trawl with a 3.7 square m net-based gear types. San Francisco Bay San Joaquin River in wet years when (39.8 square ft) wide mouth. Unlike the Studies data from 1980 through 2002 appreciable runoff exceeds the capacity summer townet survey, the Fall MWT was used in our abundance analysis. for storage and diversion of runoff. The survey catches all splittail size classes, Suisun Marsh Otter Trawl Tuolumne, Cosumnes, Feather, although larger fish are more likely to American, Napa, and Petaluma rivers, evade capture. Catches of splittail are The Suisun Marsh OT surveys began and numerous other smaller waters generally low in number because in 1979 and are conducted by the support splittail spawning activity. splittail generally reside and feed on the University of California (UC) Davis as Early indications are that the Napa River channel bottom. Furthermore, splittail part of a long-term study of the ecology may contain a robust subpopulation of apparently use shallow (less than 6 m of the entire fish community of the splittail (USACE 2002a, 2002b). (19.7 ft)) and near-shore waters to a marsh. Data from the 1979 survey have higher degree than open channels. The been excluded from our abundance Abundance Fall MWT does not sample edge waters, analysis as greater sampling effort was Seven sampling programs capture and the proportion of samples in employed in 1979 than in all splittail frequently enough to allow the shallow water stations varies by region: subsequent years (Dr. Peter Moyle, pers. calculation of useful abundance indices. 20 of 35 stations in San Pablo Bay; 1 of comm.). The survey is funded by These programs are: (1) CDFG’s Fall 18 in Carquinez Strait; 8 of 25 in Suisun California Department of Water Midwater Trawl (Fall MWT); (2) CDFG’s Bay/Marsh; and 1 of 38 in the Delta. A Resources (CDWR) in part to determine San Francisco Bay Midwater Trawl (Bay monthly abundance index for splittail if management actions in Suisun Marsh Study MW); (3) CDFG’s San Francisco captured by the Fall MWT is calculated are affecting fish communities. The Bay Otter Trawl (Bay Study OT); (4) by grouping the samples by area (17 program samples 21 sites monthly in

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00004 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55143

nine sloughs with an otter trawl that abundance indices. The Age 0 index adequately describe the differential drags along the bottom and samples reached minor peaks in 1982 and 1986, variability in catch that may occur as much of the water column in the declined to low levels during the 1987– approach velocities at Clifton Court shallow sloughs. In small sloughs, the 1992 drought (based on total splittail Forebay (SWP) or Old River (CVP) trawl samples much of the cross catch), then increased sharply to a change. sectional area; in large sloughs, the record level in 1995; minor peaks Unlike the CVP and SWP salvage, sampling fraction is smaller. A monthly occurred in 1998 and 2000, and several surveys do not account for the abundance index is calculated as mean remaining data tracked water year volume per unit time sampled. Trawl catch per trawl. The annual abundance variability. For Age 1 splittail, the data, presented as fish captured per index is calculated as the mean of the Chipps Island index for the period 1976 volume of water sampled, do not monthly index values (Sommer et al. to 2001 shows high variability. describe the trawl speed, or the 1997). While the splittail catches are The Chipps Island trawl seems to perceived trawl approach speed when dominated by YOY, the sampling also sample splittail best in high outflow pulled against a current. Seine indices consistently catches larger fish. In this years when all age groups are more are expressed as catch of fish per haul regard, the Suisun Marsh OT sampling vulnerable to trawls due to increased and do not include factors for catch per of splittail is perhaps the most thorough turbidity, as is likely true for all gear unit volume and/or per unit time. of the various sampling programs. types and surveys. It is, however, Seines are employed at sites with low Splittail collection in the Marsh is difficult to discern actual abundance water velocities, but variation in enhanced by reduced gear avoidance in from year biases, and turbidity can be velocity within and between sampling narrow, relatively shallow sloughs high at Chipps Island regardless of locations likely exists. Trawls and sampled as part of the monthly survey. outflow. Regardless, because the trawl seines may be more effective when In such conditions, the net samples a captures fish only in the top couple of employed through higher velocity larger proportion of the channel cross meters (or yards) of water in open waters; splittail may be more vulnerable sectional area than in any other survey. channels, relatively low numbers of the to capture when already navigating Larger sizes of splittail, however, benthic-foraging splittail are caught. The swifter currents. Trawls, seines, and apparently become progressively less indices are probably less precise at low pumps therefore share a common vulnerable to the trawls, a limitation population levels due to the infrequent difficulty in expressing catch per unit shared by all trawl-based surveys. captures of splittail, a characteristic volume per unit time. Each of these Spawning occurs only sporadically in shared by all surveys. The Chipps Island techniques may also differentially detect the marsh, and in most years YOY Survey data from 1976 through 2002 splittail under turbid conditions. The recruit from upstream in the Sacramento was used in our abundance analysis. pumps differ from trawling and seining, River, including the bypasses (Sommer Central Valley and State Water Project however, in that the pumps may et al. 1997). Recent modeling studies (CVP and SWP) Fish Salvage differentially entrain (collect) weak indicate that the Yolo Bypass, a major swimming juvenile and fatigued post- The CVP and SWP operate fish spawning and nursery area, may be spawn adult splittail as velocities screening facilities to divert fish away hydrologically connected to Suisun towards the facilities vary. Regardless of from the pump intakes into holding Marsh (N. Monsen, Stanford University, boat or current speed, or turbidity, facilities where they are counted, unpubl. data) so juvenile trends in the trawls and seines do not draw fish measured, and released. Data collection marsh are likely to be heavily towards them, whereas the pumps may. influenced by upstream production in takes place at two hour intervals when the pumps are operating. Consequently, The SWP catch also does not account the Yolo Bypass during those years for the predation that occurs in the when inundated for a sufficient period the fish salvage operations provide the highest number of splittail caught per Clifton Court Forebay, nor the latent of time. Suisun Marsh Otter Trawl data mortality that may occur when salvaged from 1980 through 2001 was used in our survey, but the number of data points (annual indices) is comparable to the fish are released. abundance analysis. Comparisons between CVP and SWP other surveys. All splittail age groups Chipps Island Survey are collected, the surveys do not suffer salvage and other sampling operations have to be made with caution. The U.S. Fish and Wildlife Service from gear avoidance by fish, and Nevertheless, the general patterns are conducts a sampling program for sampling locations do not vary over similar to other studies, with juvenile salmon in the deep water time. Reliable CVP data and SWP data diminished catches of both adults and channel near Chipps Island at the both start in 1979. The salvage juveniles during periods of drought and western terminus of the Delta. A abundance index is calculated based on large catches of juveniles following wet midwater trawl is pulled at the surface the total number of fish salvaged in ten 20 minute hauls per day during divided by the volume of water pumped winters. The CVP and SWP fish salvage May and June (Sommer et al. 1997). (Sommer et al. 1997). However, the data from 1979 through 2002 was used Data are compiled to produce an index pumps are not operated as sampling in our abundance analysis. based on the catch per hour of trawling programs per se so the amount of Surveys Not Employed in Abundance for the months of May and June ‘‘sampling’’ is related to the amount of Analyses combined (Sommer et al. 1997). The water exported, which in turn is related program was initiated in 1975, but data to the amount of water available, water US Fish and Wildlife Service Beach before 1979 must be viewed with some demand, and, in recent years, changes Seine Survey caution as many splittail were not in pump operations to protect migratory The survey provides the broadest measured (Baxter 1999a); as only data salmon, splittail, and delta smelt geographical coverage of all of the related to the number of splittail caught (Hypomesus transpacificus) and to sampling programs but is focused on were recorded. Length data from 1987 maintain appropriate salinities in outmigrating juvenile salmonids. The through 1993 was recorded such that Suisun Bay and Marsh. Also, the beach seine primarily captures YOY determinations of age from the data Salvage index does not address catch splittail but any fish less than 25 mm (1 cannot be done, and is therefore per volume per unit time. Lacking a in) long are not identified. The limited inadequate to calculate age-specific time factor, the Salvage index may not data show low catches of splittail during

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00005 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55144 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

dry years and higher catches during wet initially conducted from August, 1989, recruitment relationship (relationship years, reinforcing the concept of a strong to December, 1994, and was resumed in between the number of adult fish and outflow-production relationship. This 1999 and 2000. Adult splittail catch the number of offspring typically general relationship may, however, be data were only recorded during 1991 expected to join the adult population) in due to other factors. For example, through 1994, and in 1999 and 2000. splittail (Sommer et al. 1997; Moyle turbidity may be higher in high outflow This survey collected angler count, 2002). Consequently, YOY abundance years, thus rendering fish more fishing effort and fish catch information may not describe the current of future vulnerable to capture. on the Sacramento River from Redding population sizes or trends. to Carquinez Bridge year round with the Summer Townet Survey Abundance Trend Analyses same effort, 4 week days and 4 weekend The CDFG summer townet survey days per month per section, so changes We initially evaluated and analyzed began in 1959 to provide an index of in catch can reflect fish presence related the aforementioned data series using a striped bass abundance. It samples YOY to angler effort. method published by Meng and Moyle fish twice monthly at 30 sites using To reflect only the presence of (1995) in the Transactions of the oblique tows in mid-channel. Starting migrating fish, Baxter (2001b) analyzed American Fisheries Society. This and ending dates vary from year to year. only catch data from Garcia Bend (RKM method was used during the initial Sample sites are located throughout the 80 (RM 50)) and upstream. Creel census status review for the splittail and was Delta, Suisun Bay, and San Pablo Bay. data from 1991 through 1994 indicated again employed during the development Data for species other than striped bass a total annual catch of 114, 266, 498, of the proposed rule to list the splittail were not regularly recorded until after and 110 splittail, respectively. The 1999 (59 FR 862). This same method was 1962, but were also not recorded in and 2000 censuses yielded an annual replicated during the development of 1966, 1967, and 1968 (Sommer et al. catch of 103 and 232 splittail, the final listing rule published on 1997). The survey catches only low respectively. These catches represent 96 February 8, 1999, (64 FR 5963) using numbers of YOY splittail, presumably days of survey effort each year and are abundance data provided and updated because it focuses on pelagic (open useful primarily to help establish the by CDFG, CDWR, and UC Davis. The water) habitats while splittail are periods in which adult splittail migrate Meng and Moyle (1995) methodology benthic in orientation. Not surprisingly, upstream. No abundance indices were (see 66 FR 2828 for complete splittail catch varies widely and the calculated by any agency, organization, description of methods) has been index reflects only gross changes in or individual from these data, as they superceded by more current models YOY splittail abundance. The index fail to meet the criteria established by employed by CDFG, and was not used peaked in 1982, was low during the Meng and Moyle (1995) and are to help us make this final 1987 to 1992 drought years, and generally considered inadequate to the determination. Further, this removal abruptly rebounded in 1995 and 1998 task of quantifying splittail abundance. does not discuss the more recently (Baxter 1999a, 1999b). available analytical methods such as Survey Summary permutation-based exact calculations of Napa River Survey All fish sampling methods may p-values for stratified (as opposed to This survey exists in association with inherently suffer from a selection bias. unstratified) Mann-Whitney U-tests, as a flood control and ecosystem This bias results from the particular appeared in the August 17, 2001, notice restoration project in the Napa River. It method and must be considered when (66 FR 43145) where we presented an is performed by consultants under interpreting results. Because none of the updated statistical analysis of contract to USACE, and involves a range surveys were designed specifically to abundance data for the Sacramento of sampling techniques including beach monitor splittail populations, the survey splittail and requested comments on it. seine, purse seine, otter trawl, fyke nets, equipment, survey locations, and While these stratified Mann-Whitney U- and a 20 mm (0.8 in) size class surveys. sampling frequency must all be taken tests represented an improvement on The Napa River Survey began sampling into consideration when interpreting the what essentially remained a Meng and in March 2001 and has detected splittail data. All the survey methodologies Moyle (1995) statistical approach, and (USACE 2002a, 2002b) but the data are appear to sample young of the year presented a major alternative to the too recent and of too short a term (two (YOY) most effectively. As a result categorical (i.e., ‘‘before’’ and ‘‘after’’) years, including 2002 unpublished data) conclusions regarding YOY abundance approaches of both Meng and Moyle to be useful for an abundance index. appear to be the most accurate and (1995) and Sommer et al. (1997), The survey is scheduled to be reliable. Combined information from all substantive scientific and statistical completed in 2007 or 2008, after 7 years survey efforts suggest that some issues raised during the August 17, of data collection. Additionally, the successful reproduction occurs every 2001, (66 FR 43145) public comment Napa River is less well understood in year, but large numbers of young are period resulted in our using an terms of relationships between outflow, produced only during years of relatively alternative statistical analysis to help us splittail habitat, and splittail high outflow (wet years). This suggests make this final determination. The production, than are the Central Valley that the majority of adult fish in the following details the history and rivers and the Delta. As such, the population result from spawning in wet findings of the current analysis. variables employed in our current years and lowest numbers are produced In an August 17, 2001, notice (66 FR analysis of abundance and trend (see during drought years. The distribution 43145) we presented an updated Abundance section, below) cannot be and timing of YOY in the surveys also statistical analysis of abundance data for applied to this distinct river system at indicates that most spawning takes the Sacramento splittail and invited this time. place in the bypasses, rivers or upper public comments on the analysis and Delta, although some sporadic spawning data, in specific technical review of the California Department of Fish and also takes place in Suisun Marsh. It information. We concurrently sought Game Creel Census must be recognized, however, that YOY peer review on the statistical analysis CDFG collects creel census data in abundance may not be an entirely from five subject-area experts affiliated association with the Sacramento River accurate indicator of adult abundance with a total of five agencies and System Angler Survey. This survey was because there exists no observed stock- organizations. Requests for peer review

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00006 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55145

were sent to: (1) Dr. Peter B. Moyle of be applied to all sets of splittail age The question of how to analyze the UC Davis, Davis, California; (2) Dr. class data from all seven applicable less-than-optimal data we have on Charles H. Hanson of Hanson abundance monitoring data sets (Fall splittail was vexing. In large part we Environmental, Inc., Walnut Creek, MWT, Bay Study OT, Bay Study MWT, have accepted the statistical model California; (3) Randall D. Baxter of Chipps Island, Suisun Marsh, CVP provided to us by CDFG and USBR. CDFG, Central Valley/Bay-Delta Branch, salvage, and SWP salvage). The seven However, while our approach was Stockton, California; (4) Michael surveys include a total of 20 discrete generally consistent with theirs, there Chotkowski of the USBR, Mid-Pacific sets of age-specific abundance are two major differences. First, we used Region, Sacramento, California; and (5) monitoring data. These 20 datasets all 20 data sets weighted equally; Ted R. Sommer of CDWR, consist of the 2 age classes (0 and 1 or whereas the BOR and CDFG Environmental Services Office, more) for the Suisun OT, in addition to recommended that the data sets be Sacramento, California. the 3 age classes (0, 1, and 2 or more) weighted by their relative importance. Following careful consideration of for each of the other 6 surveys. Second, we accepted a 20 percent risk comments received from numerous The CDFG/USBR MLR Model that we would wrongly conclude there respondents to the August 17, 2001, explicitly controls for potential is a downward trend in the population notice, including those provided confounding effects of hydrological year for each of the 20 data sets in order to through the peer review process, we type, the factor that is nearly reduce the risk that we would fail to concluded that the abundance indices unanimously viewed as the single detect a trend if, in fact, one exist. We and Multiple Linear Regression (MLR) strongest predictor of splittail year class used this approach in order to ensure model jointly developed and submitted strengths (e.g., Moyle et al. 2001), by our assumptions were conservative. The by CDFG (2001) and USBR (2001), utilizing the number of days total delta effect was to establish a ‘‘worse-case’’ hereafter referred to as the CDFG/USBR inflow (DAYFLOW, California scenario with respect to the status of the MLR Model, provided the best scientific Department of Water Resources’ populations when we conducted our data (method) available, for statistically mathematical hydrology model) exceeds threats analysis. As a result, our evaluating temporal trends of splittail 1,557 cubic meters per second (cms) interpretation of the model results abundance information. The CDFG/ (55,000 cubic feet per second (cfs)) differs from theirs. USGS MLR Model thus superceded the during the February through May Our model results indicate that fifteen permutation-based exact calculations of spawning/rearing period as a predictor of twenty data sets have a downward p-values for stratified (as opposed to (independent variable). The 1,557 cms trend, more downward trending data unstratified) Mann-Whitney U-tests. (55,000 cfs) variable was selected sets than we would expect based on On March 21, 2002, (67 FR 13095), we because it approximates the critical chance. Typically, statisticians decide reopened the public comment period inflow value above which Delta whether such trends are ‘‘statistically (67 FR 13095;67 FR 15337) to solicit floodplains, especially the key splittail significant’’ or not. Interpreting the comments on the CDFG/USBR MLR spawning area in the Yolo Bypass, model results using the classic Model. We again sought peer review on become inundated. The 1,557 cms statistical standard (p # 0.05) for the statistical analysis from the five (55,000 cfs) variable thus captures the determining significance, we find that individuals identified above. We have existence of appreciable bypass and five of the fifteen downward trends are retained the CDFG/USBR MLR Model, spawning habitat inundation. This is statistically significant. CDFG and USBR albeit in a slightly modified form, after conceptually comparable, yet superior, believe that this result is insufficient to consideration of all public comments to the stratified Mann-Whitney U tests make a determination that the splittail received, inclusive of this and preceding presented in the August 17, 2001, notice is declining in abundance. By adopting comment periods. (66 FR 43145), which also controlled for the more relaxed standard (p # 0.20), we The CDFG/USBR MLR Model hydrological year type. There is, increase the likelihood that a significant includes HYDROLOGY and TIME (year) however, one potentially important result will be identified, a conservative as independent variables and assumption associated with the CDFG/ approach. Taking this approach (p # ABUNDANCE INDICES as the USBR MLR Model that remains 0.20), we find nine significant dependent variable. We consider this untested, and that concerns the downward trending data sets and two statistical approach superior to the assumption of a lack of interaction significant upward trending data sets. previous practice of using Mann- between the HYDROLOGY and TIME We believe that the existing data sets Whitney U tests (Meng and Moyle 1995; variables. In essence, the CDFG/USBR constitute the best available scientific Sommer et al. 1997) because it does not MLR Model assumes that the long term information and that our more require arbitrarily dividing an probabilities of high and low flow water conservative approach indicates a inherently continuous data set into years are random. number of significant declining splittail ‘‘before’’ and ‘‘after’’ categories (see population trends exist. Coupled with Discussion of CDFG/USBR MLR Model previous discussion of this issue in the the CDFG and USBR results, we have results August 17, 2001, notice; 66 FR 43145). bracketed the range of possibility We consider the CDFG/USBR MLR The results addressed in this regarding the population status of the Model superior to the polynomial discussion differ somewhat from those species as a whole. We believe this regression model presented in the published previously (67 FR 13095) due range is the best context for us to use August 17, 2001, notice (66 FR 43145) to the inclusion of new data for 2001 when we conduct our threats analysis. because existing abundance index and 2002 in some of the indices as it has We fully concur with the statements monitoring programs have not been become available (see discussion of each of various respondents that abundance conducted for a sufficient duration to survey, above). We also removed from monitoring data for splittail have provide for reasonably conclusive the analysis data taken for the Suisun methodological weaknesses of one sort application of the polynomial model (as OT in 1979, based on comments or another; none of the surveys were concluded in the August 17, 2001, received from the USBR (2002) designed specifically to rigorously notice (66 FR 43145)). We also support indicating that differing survey measure splittail population numbers use of the CDFG/USBR MLR Model protocols were used in 1979 as (see Moyle et al. 2001; Meng and Moyle because of the facility with which it can compared to other years. 1995; and Sommer et al. 1997 for

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00007 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55146 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

descriptions of surveys). However, did not meet the definition of a determination to the Federal Register existing data sets do constitute the best threatened species as set forth in the for publication on or before September available scientific information for the Act. 15, 2003. This final determination is in species. On June 23, 2000, the Court rendered compliance with that joint stipulation While our conservative approach to summary judgment in the two cases in agreement. analyzing that information is more favor of the plaintiffs, finding that our likely to produce results indicating that promulgation of the final rule listing the Summary of Comments and significant declining splittail population splittail as threatened was unlawful. On Recommendations trends exist, we believe that using this September 22, 2000, the court remanded During the five comment periods ‘‘worst case’’ scenario in analyzing the the determination of whether or not the following the remand, we contacted all impacts reported in the section entitled splittail is a threatened or endangered appropriate State and Federal agencies, Summary of Factors Affecting the species to us. The court ordered us to Tribes, county governments, elected Species is most likely to result in a re-evaluate our final determination and officials, and other interested parties listing finding that is robust. publish a new finding within 6 months and invited them to comment. We have Because we have chosen to adopt the of the date of the remand order, and requested that all interested parties CDFG/USBR MLR Model jointly kept the rule in effect during that submit factual reports or information submitted by CDFG and USBR (as our period. The court used its equitable that might contribute to the primary basis for abundance analyses), powers to retain the protections of the development of a final determination. In and are no longer using our analysis in Act for the species during the remand of addition, we have invited public our August 17, 2001 notice (66 FR the rule to the Service. comment through the publication of 43145), specific comments on our On January 12, 2001, we reopened the notices in various newspapers. We analysis in our August 17, 2001 notice comment period for 30 days to seek published notice of the January 12, (66 FR 43145) will not be addressed in information regarding the splittail’s 2001, reopening of the comment period the section entitled Summary of status, abundance and distribution, as in the Sacramento Bee, Fresno Bee and Comments and Recommendations. well as information regarding issues Contra Costa Times newspapers. For the identified by the District Court in its May 8, 2001, notice, we invited public Previous Federal Action June 23, 2000, judgment (66 FR 2828). comment through publication of notices On February 8, 1999, we published a At that time, we were subject to a court- in the Antioch Ledger-Dispatch, the final rule listing the splittail as ordered deadline of March 22, 2001. On Marysville Appeal-Democrat, the Fresno threatened under the Act (64 FR 5963). March 16, 2001, we received an Bee, and the Sacramento Bee. For the Please refer to the final rule for a extension from the District Court until August 17, 2001, reopening notice we discussion of Federal actions prior to June 22, 2001, so that we could reopen invited public comment through the publication of the final rule. At the the comment period. Subsequent to that publication of notices in the Marysville time of our final determination of extension, we reopened the comment Appeal-Democrat, the Fresno Bee, and threatened status for the splittail, the period for the second time since the the Sacramento Bee. An electronic mail splittail population had declined in remand, from May 8, 2001 to June 7, address for submission of comments both numbers and range and was 2001 (66 FR 23181). On June 28, 2001, was provided in the May 8, 2001, and primarily threatened by changes in we received an additional extension August 17, 2001, notices and was posted water flow and water quality resulting from the court so that the comment on the Sacramento Fish and Wildlife from the export of water from the period could be reopened and we could Office’s official web site. For the March Sacramento and San Joaquin Rivers, have additional time to obtain reviews 21, 2002 reopening notice, we invited periodic prolonged drought, loss of of the revised statistical analyses which public comment through publication of shallow water habitat, introduced we employed in response to prior notices on March 27, 2002, in the aquatic species, and agricultural and comments. The comment period was Marysville Appeal-Democrat, the industrial pollutants. then opened on August 17, 2001 (66 FR Sacramento Bee, and the Fresno Bee. An Subsequent to the publication of the 43145); while the court ordered decision electronic mail address was not final rule, plaintiffs in the cases San date was established as January 31, provided for the March 21, 2002, Luis & Delta-Mendota Water Authority 2002. We later received an additional reopening due to uncertainties regarding v. Anne Badgley,° et al. and State Water extension from the court until October our internet access. An electronic mail Contractors, et al. v. Michael Spear, et 15, 2002, so that we could seek address was, however, provided with al. commenced action in the Federal comments on the MLR Model submitted our April 1, 2002, correction, and with Eastern District Court of California, by CDFG and USBR during the August our October 31, 2002, reopening. We challenging the listing of the splittail as 17, 2001, comment period. On March also sent out notices of each reopening threatened, alleging various violations 21, 2002, we reopened the comment of the comment period to all parties on of the Act and of the Administrative period for the fourth time since the a mailing list for Sacramento splittail Procedure Act (5 U.S.C 551 et seq.), remand (67 FR 13095) and on April 1, information. specifically that we: (1) Failed to use the 2002, we corrected the duration of the During the five comment periods best scientific and commercial data comment period to reflect 60 days (67 opened since the remand, we received available; (2) ignored all pre-1980 and FR 15337). On October 31, 2002, we a total of 33 written comment letters post-1992 data available and that we received an additional extension from representing 1 Federal agency, 2 State used only selected data from the 1980 the court so that the comment period agencies, 2 local governments, and 13 to 1992 period; (3) did not publish a could be reopened for a fifth time since private individuals or organizations. We summary of the available data, which the remand (67 FR 66344) to solicit reviewed all comments received for data we considered, and the comments on the revised statistical substantive issues and new information relationship between the data and our analysis we had done, as described in regarding the status of the Sacramento decision on the final rule; and (4) our March 21, 2002 document (67 FR splittail. Of the comments we received, promulgated the final rule in a manner 13095). Finally, on February 28, 2003, only 3 supported listing. Information that was arbitrary, capricious, and not in the court approved a joint stipulation contained in these comments was accordance with law, in that the splittail requiring us to submit our final reviewed to determine if it raised any

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00008 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55147

new substantive issues that had not current population declines. However, peer reviewer specifically cited Manly been raised in comments previously even considering our conservative (2002) which states ‘‘The Service claims submitted, and subsequently addressed analysis of the apparent splittail that lack of power to detect a trend gives in this final determination. population declines and the threats a reason for using a 20 [percent] level The following is a summary of analysis, we believe the conservation of significance in assessing whether or comments we received during the 197 elements of the California State and not there is evidence of a trend with days associated with the five comment Federal cooperative program (CALFED) individual series. This then allows [the periods opened since the remand of the and the Central Valley Project Service] to claim evidence for a trend final listing rule. For additional Improvement Act (CVPIA) programs for 7 of the 20 series. Although this information on comments received adequately mitigate for these threats sounds impressive, it is less so when it during three previous comment periods (please refer to Summary of Factors is realized that by chance alone 4 of the before the current litigation, please see Affecting the Species section for a 20 series (i.e., 20 [percent] of them) are the previous final listing rule (64 FR detailed discussion of CALFED and the expected to give a significant result if 5963). Substantive comments and CVPIA). this level of significance is used.’’ The information raised or provided during Peer Reviewer Comment 3: A peer peer reviewer also asserted that the the public comments periods have reviewer submitted comments that weak nature of the MLR Model either been incorporated directly into included an analysis using a modified regression coefficients will be this notice or addressed below. version of Meng and Moyle’s (1995) pre- demonstrated with the calculation of decline and post-decline method. The Peer Review splittail abundance indices for 2000, peer reviewer also divided the data by 2001, and 2002 and their inclusion into As previously discussed in the above year class and used data available from the models. abundance section, we requested 5 all years and requested we consider Our Response: Using the most recent biologists to provide scientific review of these analyses. data, our analysis now indicates that 9 the proposed listing of the splittail as Our Response: As discussed earlier in of 20 indices show significant negative threatened. Technical data provided by this notice, we acknowledge that there trends at the 20 percent level of the peer reviewers have been are other methods by which to analyze significance, while 2 of 20 show incorporated into or addressed in this the available data, but that we have now significant positive trends at that document, while other issues raised by employed an analysis using the CDFG/ significance level. As we noted earlier the peer reviewers are addressed below. USBR MLR Model data series to in the analysis, we achieved these Peer Reviewer Comment 1: A peer describe population trends of the results by a conscious choice of a reviewer cited the ‘‘White Paper’’ splittail. We refer the peer reviewer to variable that accepted a higher risk of (Moyle et al. 2001) for splittail as raising our Abundance section for a discussion incorrectly identifying downward the possibility that abundance may not of our most recent statistical analysis of trends in population in order to take a be a reliable measure of population the species population trends. conservative position in our threats status for the splittail. Peer Reviewer Comment 4: A peer analysis. Our Response: We acknowledge that reviewer criticized us for evaluating the Peer Reviewer Comment 6: A peer abundance may not be the most reliable results of the CDFG/USBR MLR Model reviewer criticized our acceptance of the measure of population status, but assert for all 20 data series of splittail ‘‘sign’’ (i.e., positive or negative) results that it is the best scientific measure abundance index data, instead of of the CDFG/USBR MLR Model available. The utility of abundance as a limiting the evaluation to the nine data coefficients at face value because in measure of splittail population status is series that the respondents view as most most cases (16 of 20) the true signs (i.e., reflected in its continued use by the representative of overall splittail positive or negative) were just as likely scientific community including populations. Another peer reviewer to be positive as negative. researchers (Meng and Moyle 1995, stated that Bay Study OT and Fall MWT Our Response: We cannot apply the Sommer et al. 1997) and agencies data were more indicative of splittail respondent’s reasoning to the available (CDFG, CDWR, USBR). abundance trends, rather than the trends data. The p-value for a coefficient is Peer Reviewer Comment 2: A peer made evident by data collected at the what statistical analysis has indicated it reviewer cited the ‘‘White Paper’’ SWP Salvage facilities, Chipps Island, should be; simply because a given p- (Moyle et al. 2001) for splittail as and in Suisun Marsh, which the value does not rise to the level of 95 reporting a tentative population model respondent felt were narrow in percent significance criterion, does not result that stated, ‘‘* * * a long series geographic scope. indicate that the p-value automatically of dry years is unlikely to drive the Our Response: We note that these and reverts to 50 percent. splittail to extinction, even if the other respondents have previously Peer Reviewer Comment 7: It was population is greatly reduced.’’ Another criticized us, while employing different noted by a peer reviewer that in half the peer reviewer asserted that if the analysis, for not treating all 20 data CDFG/USBR MLR Model runs the splittail were truly going extinct, all series equally and for not including all dependent variable was significantly surveys would show a decline. available data series in statistical non-normal and that as a consequence Our Response: A species warrants evaluations of abundance trends. We probability statements will be ‘‘slightly’’ listing as threatened under the Act if is refer the commentor to the section in error. in danger of becoming endangered in entitled Abundance for a discussion of Our Response: We believe that the the foreseeable future throughout all or our treatment of the data series. peer reviewer’s comment is correct. This a significant portion of its range (16 Peer Reviewer Comment 5: A peer type of error alone, however, would not U.S.C. 1532(20)). It is possible for the reviewer reiterated his assessment that necessarily invalidate our evaluations of splittail to be undergoing threats or the statistical evidence for a declining the signs and magnitudes of the declines in a significant portion of its trend in splittail abundance is weak, regression coefficients. The error would range without declines showing in all and cited an analysis that asserted that have to be of a nature that creates bias. surveys. Alternatively, threats to the evidence for a time trend in 7 of 20 data The peer reviewer did not provide any splittail may support listing even in the series is not a compelling factor in statistical or other argument to explain absence of our ability to document determining that declines exist. The why such error would necessarily result

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00009 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55148 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

in bias. The unknown statistical effects drought unlikely to influence the Our Response: We agree that USBR’s of non-normality in half the model runs analytical findings from an ever- submission was labeled A Sample constitutes just the sort of uncertainty lengthening period of record. Most Alternative Model of Sacramento that leads us to be cautious about giving importantly, we now employ the CDFG/ Splittail Abundance. However, USBR undue weight to any conclusions USBR MLR Model, which explicitly (2001) included no language in their regarding the abundance index data for controls for potential confounding agency comment letter and peer review splittail. effects of hydrological year type. The submission to suggest their intent was to Peer Reviewer Comment 9: A peer respondent’s concern would be more have us retain the polynomial regression reviewer believes that the extended applicable to abandoned analytical analysis (66 FR 43145), revert to the drought of 1984 to 1992 created only a techniques. The arbitrary pre- and post- Meng and Moyle (1985) analysis, adopt perception of decline and that it was the decline cut point approach of Meng and the Sommer et al. (1997) analysis, or ‘‘* * * accidental juxtaposition of a Moyle (1985) was driven by trends employ any other analytical technique series of wet, strong splittail years with noticed during the 1987 to 1992 until the CDFG/USBR MLR Model and a series of dry, weak years that drought, as was a formerly touted results reached a greater state of prompted [our] interest in the first alternative analysis that involved the refinement. place.’’ use of 1987 (the beginning of the To the contrary, USBR’s peer review Our Response: We disagree with the drought) as a cut point (Sommer et al. and comment letter states, ‘‘Results peer reviewer’s claims that the period of 1997) for determining percent declines. presented in Table 1, include actual p- extended drought has been ignored, as We also disagree with the contention values for the Service’s inspection.’’ well as with the contention that the that the 1987 to 1992 drought serves as (USBR 2001). To advocate we abandon splittail’s drought-driven declines are the only factor which triggered our the model is to advocate we abandon the sole factor under consideration in investigations of the splittail’s status. analysis of p-values. Furthermore, USBR our determination. We first note that the Our interest in the splittail was scientifically derived and submitted period of continuous drought is prompted initially by the statement in multiple conclusions in their peer considered by most authoritative Daniels and Moyle (1983) that the review and comment letter, such as, ‘‘In sources to have begun in 1987 (Moyle et splittail’s and delta smelt’s ‘‘*** summary, the results [of the CDFG/ al. 2001; Baxter 1999a; Sommer et al. abundance could decline rapidly if USBR MLR Model] presented here 1997), not 1984 as reported by the environmental conditions become clearly indicate that hydrologic respondent. We note, however, that variability strongly affects YOY splittail unfavorable for them, possibly making 1985 and 1986 were dry years (Cannon indices, and also affects some adult them candidates for listing as threatened 2001 in prep.). indices in succeeding years as cohorts species.’’ We subsequently included the The declines noted during the 1987 to propagate through the population.’’ Sacramento splittail as a category 2 1992 drought were the likely result of a (USBR 2001). These conclusions were candidate species for possible future paucity of spawning habitat being not accompanied by any disclaimers listing as endangered or threatened in available. The drought decreased the that the conclusions should be the January 6, 1989, Notice of amount of floodplain (i.e. Yolo Bypass disregarded because the model was not Review (54 FR 554). The candidate and mainstem river margins) available yet sufficiently developed or that the category system was abandoned on for spawning and thus, spawning output conclusions should not be applied to was lower. Low splittail population February 28, 1996 (61 FR 7457), and the review of the splittail’s population densities were aggravated by the CVP species meeting the definition of the trends. and SWP’s diversion of a greater former category 2 (such as splittail) were The CDFG/USBR MLR Model was proportion of water from the Delta than no longer considered candidates. Our also submitted to us by CDFG. in prior years; fish were entrained at the administrative proceedings on splittail Consistent with the USBR peer review facilities and the entrapment zone resumed on November 5, 1992, when and comment letter, CDFG also derived (location where fish become vulnerable we received a petition from the Natural and submitted multiple conclusions to the export facilities’ effect on currents Heritage Institute to add the Sacramento based on the specific runs of the CDFG/ in the Delta), was located well upstream splittail to the List of Endangered and USBR MLR Model that the USBR is now of Suisun Marsh in increasingly Threatened Wildlife and to designate criticizing us for accepting. CDFG suboptimal habitat. These events are critical habitat for this species in the advocated the use of the CDFG/USBR described in detail in our February 8, Sacramento and San Joaquin Rivers and MLR Model (as submitted) in their peer 1999, final listing rule (64 FR 5963). associated estuary. review and comment letter (CDFG 2001) The basis for the peer reviewer’s Peer Reviewer Comment 10: A peer with the statement, ‘‘Our response is claim that we are disproportionately reviewer, in response to our March 21, composed of two parts: a discussion of concerned with splittail declines noted 2002 (67 FR 13095) notice, believed that individual analyses presented in our during the 1984 (or 1987) to 1992 we should not have adopted the CDFG/ August 17, 2001, notice (66 FR 43145), drought is unclear. True, the ‘‘accidental USBR MLR Model which was jointly and a summary of the results of a juxtaposition’’ of wet and dry years submitted in CDFG’s and USBR’s multiple regression analysis [the CDFG/ resulted in abundance data that respective peer review and comment USBR MLR Model] that we believe is appeared to illustrate a precipitous drop letters. The CDFG/USBR MLR Model more useful in evaluating trends in in the splittail population. There are, was advocated by its submitting survey indices.’’ Again, consistent with however, up to 10 years of pre-drought agencies as an approach superior to our the USBR’s peer review and comment as well as up to 8 to 10 years of post- Meng and Moyle (1985) method utilized letter, CDFG’s peer review and comment drought data. The data collected during in our 1994 proposed listing (59 FR 862) letter did not qualify any of the six years of continuous drought are but and 1999 final listing (64 FR 5963) conclusions they derived from the a subset of the nearly 20 years of extant rules, the polynomial regression CDFG/USBR MLR Model with splittail data. The splittail’s relatively technique discussed in our August 17, disclaimers about the inappropriateness long life span and resilience following 2001 (66 FR 43145) notice, or the of employing the model. unfavorable conditions renders the Sommer et al. (1997) technique formerly We independantly evaluated the declines exhibited during a discrete forwarded by CDFG and CDWR. structure and findings of the CDFG/

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00010 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55149

USBR MLR Model and determined that to reflect trends and stated that adult The Transactions of the American it represented the best available and juvenile indices should not be Fisheries Society does not use the scientific and commercial information. combined. The peer reviewer also double-blind method for peer review. We retain our conclusions regarding our suggested that Bay Study OT and Fall This issue was raised in regards to our analysis and meta-analysis of the MWT were more representative past use of the Meng and Moyle (1995) model’s results, regardless of its measures of abundance. methodology to determine splittail developers’ current desire to Our Response: While we concur that abundance. secondarily qualify its application. declining catch efficiency may be a Our Response: We agree with this Peer Reviewer Comment 11: A peer characteristic of trawls, we do not agree assertion. Each piece of scientific work, reviewer commented that it was unclear that it should be used to exclude a trawl whether a peer reviewed published whether we had independently re- survey’s data. Declining catch efficiency paper or an unpublished, unreviewed, derived the CDFG/USBR MLR Model within a given trawl survey is expected draft report, must be objectively results submitted jointly by CDFG and to be uniform from year to year, thus evaluated for the scientific merit of its USBR. rendering inter-annual analysis valid. content alone. Peer reviewed Our Response: We did not Although an age bias will make data publication provides no guarantee of independently re-derive those results. series for older age-class fish less scientific merit. The test of time, We accepted the results presented in sensitive for detecting change, it will following publication, provides the CDFG (2001) and USBR (2001) at face not produce a long-term directional bias ultimate measure of scientific merit. value, as they were developed by (i.e., we have no reason to believe that Indeed, subsequent iterative subject area experts within CDFG and the capture efficiency for older age class examination of the splittail’s status has USBR during a peer review and public splittail is becoming progressively resulted in our abandonment of Meng comment process. worse over time). Thus, any trends in and Moyle (1995), Sommer et al., (1997) Peer Reviewer Comment 12: A peer the older age class data series with a and our permutation-based exact reviewer believed that our that our substantive p-value can be viewed to be calculations of p-values for stratified (as statement, ‘‘* * * [the] traditional roughly as accurate and reliable as for opposed to unstratified) Mann-Whitney [alpha-value] criteria assume a much the Age-0 class of splittail showing U-tests (66 FR 43145). higher standard of statistical power than trends at comparable p-values. Peer Reviewer Comment 16: A peer the splittail data are able to meet * * *’’ We also concur that trawls’ declining reviewer claimed we ignored the draft in our March 21, 2002, notice (67 FR catch efficiency does preclude the ‘‘White Paper’’ published by Moyle et 13095) is erroneous. combination of age class data. We report al. (2001, in prep.) Our Response: We agree with the peer each index separately herein and do not Our Response: We use the various reviewer that in a strictly literal sense, combine adult and juvenile indices findings and hypotheses found in the the choice of an alpha-value criterion other than for meta-analytical purposes. draft and revised White Paper (Moyle et can be made without any regard for We also acknowledge that, in certain al. 2001 in prep.) extensively statistical power. However, in practice, situations, adult abundance for different throughout this document. researchers are concerned with both age classes (of adults) is combined and Peer Reviewer Comment 17: A peer type I error (determined by the choice reported because the data are collected reviewer stated that the range of the of an alpha value) and type II error in that manner, i.e., salvage data are splittail is wider than was previously (directly related to the statistical power reported as Age-1 and as Age-2 and thought. of a study). When conducting our greater with no differentiation made for Our Response: The greater range of analysis, we made a conscious choice to individuals greater-than Age-2 classes. the splittail was acknowledged in the use the more conservative Situations such as this represent a relict January 12, 2001, notice (66 FR 2828). nontraditional approach of using an of the sampling methodology but remain The above Background section of this alpha value of 0.20. the best available information. We final document contains a discussion of Peer Reviewer Comment 13: A peer continue to believe that as long as the the range of the splittail. reviewer asserted that the purpose of degree of age-based capture bias is Peer Reviewer Comment 18: Several statistical hypothesis testing in the case constant over a survey period, all age peer reviewers felt that we should not of the MLR Model is to decide whether classes should show approximately the classify the Yolo and Sutter bypasses as trends do or do not exist, not to evaluate same trends, and that combining age a threat to the splittail, as we did in the gradients of reliability in evaluating classes for meta-level statistical analyses January 12, 2001, reopening of comment trends. is not problematic. period (66 FR 2828), based primarily Our Response: The CDFG/USBR MLR We reiterate that the Suisun Marsh upon the data found in Sommer et al. Model is a probabilistic approach to OT, which combines an efficient, (1997) and Sommer (2001a). The examining time trends, it is not a bottom trawling technique with focused bypasses have demonstrated the categorical ‘‘either/or’’ approach (as the surveys in a small habitat at the core of capability of producing large numbers of respondent appears to assert). We chose the splittail’s range, is the most likely to splittail when inundated. One peer to evaluate the probabilities associated detect a trend and likely suffers from reviewer also felt that the bypasses with competing hypotheses concerning less sampling inefficiency than the Bay cannot be considered a threat simply the abundance status of splittail. It is for Study OT (low detection of splittail at because the conditions could be better. this reason that we stated that all trends, periphery of range) and Fall MWT Another peer reviewer claimed that not just trends meeting an arbitrary (unlikely to detect benthic fish and does current operations in the bypasses do traditional confidence criterion (95 not sample shallow water or near-shore not harm splittail or their habitat. percent confidence, or alpha-value of areas). Another peer reviewer felt that the 0.05) were evaluated. Peer Reviewer Comment 15: A peer bypasses are not to be considered a Peer Reviewer Comment 14: A peer reviewer asserted that the peer review threat because even though their reviewer believes that a trawl’s process for scientific publications splittail habitat conditions are not declining catch efficiency for adult doesn’t necessarily ensure that optimal, they are still sufficient to splittail as compared to juvenile and published papers are unbiased, provide substantial benefits to the YOY rendered trawl surveys less likely scientifically sound, and without errors. species. Finally, another peer reviewer

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00011 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55150 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

stated that the Yolo and Sutter bypasses trends, and that combining age classes Our Response: We agree with and use are a ‘‘net benefit’’ to the splittail in that for meta-level statistical analysis will many of the various findings and without their existence, the species not be a problem. hypotheses found in the draft and might not have persisted to the present revised White Paper (Moyle et al. 2001) State Agencies day. extensively throughout this document. Our Response: We have determined, We received comments from the We believe that the White Paper is a based on consideration of scientific data following California State agencies: useful resource and contributes to the and information provided by Department of Fish and Game (CDFG) knowledge on splittail biology. The respondents, that the Yolo and Sutter and Department of Water Resources paper has been referenced throughout bypasses are not, in and of themselves, (CDWR). Technical data provided by the this document. a threat to the splittail. Our reevaluation CDFG and CDWR have been State Agency Comment 4: CDWR of this issue is discussed in Factor E of incorporated into or addressed in this stated that the hypothetical analytical the section entitled Summary of Factors document, while other issues raised by model presented at the January 29, Affecting the Species. State agencies are addressed below: 2001, CALFED Bay-Delta Program Peer Reviewer Comment 19: A peer State Agency Comment 1: CDFG (CALFED Program) Splittail Science reviewer felt that our determination that submitted comments that included an Conference and described in the White the Sutter and Yolo bypasses would analysis using a modified version of Paper (Moyle et al., 2001) indicates that require inundation for at least 30 Meng and Moyle’s (1995) pre-decline the splittail, even during severe and continuous days between March and and post-decline method. CDFG also lengthy drought, is unlikely to be driven April in order for them to be considered divided the data by year class and used to extinction. a beneficial splittail spawning habitat data available from all years and Our Response: We ultimately arrive at was inaccurate and could affect water requested we consider these analyses. the same conclusion as Dr. Moyle, that supply and flood management. Our Response: As discussed earlier in the splittail is unlikely to be driven to Our Response: We have not proposed this notice, we acknowledge that there extinction. However, at this point we inundation of the bypasses for any are other methods by which to analyze are unwilling to accept that premise specific interval, duration, or frequency. the available data, but that we have now solely on the basis of the White Paper. Rather, we have speculated that the employed an analysis using the CDFG/ To date, there remains no proven bypasses would have their greatest USBR MLR Model data series to scientific method for determining the benefits to splittail if they became describe population trends of the current splittail population size inundated at a frequency and duration splittail. We refer CDFG to our primarily because no extant survey was that as closely as possible mimics the Abundance section for a discussion of designed specifically to monitor splittail natural, precipitation-driven our most recent statistical analysis of populations or to determine their hydrograph. The reference to 30 days is the species population trends. absolute numbers. Further, the splittail a statement regarding how the State Agency Comment 2: CDFG exhibits relatively wide variation in inundation patterns of the bypasses at reiterated their assessment that the annual abundance in response to times do not meet the life history statistical evidence for a declining trend prevailing hydrologic conditions; it is requirements of the splittail. Inundation in splittail abundance is weak. CDFG likely that the population size exhibits of bypasses in dry years would reduce cited an analysis that asserted that appreciable year to year variability the effects of drought on the splittail. evidence for a time trend in 7 of 20 data which would confound size estimates. We also speculate that if the bypasses series is not a compelling factor in Calculating the current population’s were inundated at a frequency and determining that declines exist. CDFG risk of and/or time to extinction would duration that as closely as possible specifically cited Manly (2002) which require estimates of absolute population mimics the natural, precipitation-driven states ‘‘The Service claims that lack of size, rate of decline, and minimum hydrograph, then the numbers of non- power to detect a trend gives a reason viable or sustainable population size, native fish would be reduced, as non- for using a 20 [percent] level of none of which currently exist in a native fishes favor ponded and significance in assessing whether or not scientifically defensible form. Moreover, continuously inundated habitats. there is evidence of a trend with it must also be noted that the statutory Peer Reviewer Comment 20: A peer individual series. This then allows [the and regulatory standard for ascertaining reviewer believed that full Service] to claim evidence for a trend threatened status is not to determine implementation of the CALFED Program for 7 of the 20 series. Although this whether or why a species will become would preclude the need to list the sounds impressive, it is less so when it extinct in the near future, but if, splittail and indicated that over $10 is realized that by chance alone 4 of the pursuant to section 3(19) of the Act, it million had been spent on actions that 20 series (i.e., 20 [percent] of them) are ‘‘* * * is likely to become an could improve conditions for splittail. expected to give a significant result if endangered species within the Our Response: We refer the peer this level of significance is used.’’ CDFG foreseeable future throughout all or a reviewer to the section entitled also asserted that the weak nature of the significant portion of its range’’. An Summary of Factors Affecting the MLR Model regression coefficients will endangered species, pursuant to section Species. be demonstrated with the calculation of 3(19) of the Act, is that ‘‘* * * which Peer Reviewer Comment 21: A peer splittail abundance indices for 2000, is in danger of extinction throughout all reviewer asserted that the age-based 2001, and 2002 and their inclusion into or a significant portion of its range capture bias argues against combining the models. ***’’. Our analysis, including a data from different age groups. Our Response: As we note in our nontraditional conservative approach to Our Response: We assume this earlier analysis we made a conscious estimating population trends examines comment refers to the pooling of data decision to use the more conservative, the factors identified in the Act and in series from all age classes for meta-level nontraditional 0.20 alpha for analysis fact we find that the splittail does not statistical evaluation. We believe that as purposes. warrant listing at this time. long as the degree of age-based capture State Agency Comment 3: CDWR State Agency Comment 5: CDWR felt bias is constant over a survey period, all claimed we ignored the ‘‘White Paper’’ that we should not classify the Yolo and age classes should show about the same published by Moyle et al. (2001). Sutter bypasses as a threat to the

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00012 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55151

splittail, as we did in the January 12, emigrate prior to stranding reduces the trend analysis.’’ CDWR, in comments 2001, reopening of comment period (66 risk of stranding. CDWR also felt that also submitted under the August 17, FR 2828), based primarily upon the data the magnitude of the entrainment 2001, comment period, stated: ‘‘A more found in Sommer et al. (1997) and threats presented by the bypasses was defensible alternative would be to Sommer (2001a). The bypasses have overestimated when we cited in the develop a multivariate model demonstrated the capability of January 12, 2001, notice (66 FR 2828), incorporating the effects of both flow producing large numbers of splittail the death of a number of juvenile and time.’’ CDWR also made reference when inundated. splittail in an approximately 0.8 hectare to the USBR application of regression Our Response: We have determined, (ha) (2 acre (ac)) borrow pit as techniques, which also were provided based on consideration of scientific data statistically-significant and that the in USBR’s comments. We have and information provided by classification of ‘‘natural sinks’’ as a considered the CDFG, CDWR, and USBR respondents, that the Yolo and Sutter threat was in error. recommendations to employ a bypasses are not, in and of themselves, Our Response: We have considered multivariate, regression based model a threat to the splittail. Our reevaluation these data and now agree that and have incorporated an analysis using of this issue is discussed in Factor E of entrainment in the Yolo Bypass is less the CDFG/USBR MLR Model data series the section entitled Summary of Factors than was originally thought. Information as described in the section entitled Affecting the Species. presented at the January 29, 2001, Abundance. We will therefore forego State Agency Comment 6: CDWR felt CALFED Splittail Science Conference providing responses to specific that our determination that the Sutter indicates that a modest degree of comments on the perceived bias of the and Yolo bypasses would require topographic variability within an Meng and Moyle (1995) and alternate inundation for at least 30 continuous inundated area may be beneficial, as it methodologies previously employed days between March and April in order may create a diversity of flow patterns because our analytical tools have been to for them to be considered a beneficial and velocities which in turn may allow upgraded to utilize the modified splittail spawning habitat was juvenile splittail to evade predation and methodology employed by CDFG and inaccurate and could affect water forage more effectively during egress. USBR. supply and flood management. State Agency Comment 9: CDWR The CDFG/USBR MLR Model Our Response: We have not proposed believed that full implementation of the provided in CDFG and USBR comments inundation of the bypasses for any CALFED Program would preclude the addresses the shortcomings of other specific interval, duration, or frequency. need to list the splittail and indicated methods, thus allowing our analysis Rather, we have speculated that the that over $10 million had been spent on using the CDFG/USBR MLR Model data bypasses would have their greatest actions that could improve conditions series to supercede abundance analyses benefits to splittail if they became for splittail. based upon methods appearing in prior inundated at a frequency and duration Our Response: We refer CDWR to the rules. In combination with meta- that as closely as possible mimics the section entitled Summary of Factors analyses to analyze the distribution of natural, precipitation-driven Affecting the Species. MLR Model results across the 20 hydrograph. The reference to 30 days is indices, statistical inferences based on Other Public Comments and Responses a statement regarding how the the CDFG/USBR MLR data series are inundation patterns of the bypasses at We address other substantive informative. times do not meet the life history comments and accompanying Our analysis using the CDFG/USBR requirements of the splittail. Inundation information in the following summary. MLR Model data series incorporates the of bypasses in dry years would reduce Relatively minor editing changes and results of seven surveys (Fall MWT, Bay the effects of drought on the splittail. reference updates suggested by Study OT, Bay Study MWT, Chipps We also speculate that if the bypasses commenters have been incorporated Island, Suisun Marsh OT, CVP salvage, were inundated at a frequency and into this document, as appropriate. and SWP salvage), and includes duration that as closely as possible Comment 1: The court directed that separate indices of YOY, age 1 (juvenile) mimics the natural, precipitation-driven we provide a more thorough response to and age 2+ (adult) age class abundance. hydrograph, then the numbers of non- the California Resources Agency The independent examination of native fish would be reduced, as non- comments, specifically comments abundance of all age classes throughout native fishes favor ponded and submitted by CDFG and CDWR in July these surveys helps mediate continuously inundated habitats. 1998. The court also directed that we discrepancies among survey results, State Agency Comment 7: CDWR address the perceived biases from the discrepancies that are a likely indication commented that our classification of the Meng and Moyle (1995) method. We that splittail populations are not very Yolo Bypass as a threat in the January also received specific comments on evenly distributed over space and time 12, 2001, notice (66 FR 2828) would issues related to prior statistical and/or that different sampling undermine potential ecosystem analyses of abundance. methodologies are not very comparable. restoration actions that would benefit Our Response: We have adopted a The model also does not require the splittail. multiple linear regression approach uninterrupted data; all available data Our Response: In this notice, we proposed by CDFG and U.S. Bureau of from each survey’s period of record is determine that the Sutter and Yolo Reclamation (USBR). CDFG, in included. Further, our analysis controls bypasses are not in and of themselves comments submitted in association with for the confounding effects of threats. the August 17, 2001, comment period, hydrology, and involves no inherent or State Agency Comment 8: CDWR stated: ‘‘Although CDFG reported Mann- intentional bias towards either wet or objected to our statements regarding the Whitney U test results in previous dry water year types. Strict adherence to entrainment risks present in the comments (February 8, 2001), we now uniformity among all data series is also bypasses based upon Sommer et al.’s suggest greater reliance on a multiple inconsistent with the precautionary (1997) findings that entrainment is not regression approach to trend analysis, nature of section 4 of the Act. a significant threat within the bypasses. described in a following section of our We recognize a distinct danger in It is thought that the splittail’s comments. We no longer support use of controlling for hydrological effects in evolutionarily-derived ability to the Mann-Whitney U procedure of time our analyses, because systematic

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00013 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55152 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

changes in hydrological regimes, due to for these threats (refer to Factor A for a We currently support use of the human manipulation or long term detailed discussion of the CALFED CDFG/USBR MLR Model because of the climate change, could just as feasibly be program and the CVPIA programs). facility with which it can be applied to a causative factor as a confounding Comment 4: A respondent informed all sets of splittail age class data from all source of ‘‘noise.’’ If systematic changes us that CDFG re-analyzed the striped seven applicable abundance monitoring in hydrological regimes were occurring, bass egg and larval survey and found data sets (a total of 20 discrete sets of it would not be prudent to control for that splittail spawn in the mainstem of age-specific abundance monitoring that factor. Our since-superceded the Sacramento River, especially in dry data). This approach therefore includes polynomial regression analysis of years. This indicates that splittail occur consideration of YOY splittail without abundance data (See Abundance section in the Sacramento River upstream from granting undue analytical weight to any of the August 17, 2001, notice) the Delta. single survey or age class or controlled for influences of hydrological Our Response: CDFG and our survey inappropriately combining different cycles without discarding hydrology as results confirm that splittail use river survey equipment types. Regardless of a potential directional factor margin habitat in the mainstem the strengths and weaknesses of year determining long term trends of splittail Sacramento River. Indeed, recent 2000 YOY abundance, these data were abundance. We expect that the indications are that river margin habitat considered in our analysis using the polynomial regression analysis is where splittail spawning occurs CDFG/USBR MLR Model data series presented in the August 17, 2001, notice through periods of drought. (see our Abundance section for a may eventually inform the Comment 5: A respondent stated that discussion of our most recent statistical understanding of the effects of changed young of the year (YOY) abundance was analysis of the species population hydrology on the splittail, once the at near record levels in 2000, thus trends). future, cumulative hydrologic analyses inferring the splittail is not in decline. Comment 6: The Court and numerous for potential water development projects Our Response: Data presented in the commenters requested that we address have been developed by the responsible Spring 2001 Interagency Ecological and clarify the issue of splittail agencies. Program Newsletter (Baxter 2001a), resiliency and that the species may be Comment 2: The court directed us to provided as an attachment to public able to withstand drought and produce show the relationship between the data comment submitted on this rulemaking, high numbers of young of year (YOY) used in our decision-making analysis do indicate that splittail spawning was during wet periods. and the original final rule and how we Our Response: We concur that reached the conclusion that the splittail highly successful in 2000. This spike of splittail are a resilient species and that was threatened. juvenile fish is to be expected given the they can reproduce effectively in wet Our Response: We have provided a relatively wet conditions of 2000, and more detailed analysis in the section the splittail’s ability to exploit suitable years. Sacramento splittail populations entitled Summary of Factors Affecting habitat when available. Also, YOY are fluctuate annually depending upon the the Species. The threats to the species generally the most reliably sampled fish availability of shallow water habitat have also been summarized in an in any given survey, since their raw with submerged vegetation (Daniels and additional section entitled Conclusion abundance is temporarily high and YOY Moyle 1983). Meng and Moyle (1995) Regarding Abundance, Distribution, and splittail are likely less effective at and Sommer et al. (1997) have found Factors Affecting the Species. We have evading sampling equipment. that splittail year-class abundance is also included in the Abundance section Population level conclusions drawn positively correlated with freshwater of this notice a discussion of our most from such a spike must be made with outflow occurring during the species’ recent statistical analysis of the species caution because, though extremes in late winter and spring spawning season. population trends. YOY abundance appear to be reflected The evolutionary strategy of the splittail Comment 3: Several respondents cited in 2 to 3 year subsequent adult therefore appears to be one of the draft White Paper (Moyle et al. 2001 abundances, the splittail appears to opportunism, whereby the population in prep) for splittail as reporting a exhibit no stock-recruitment collectively invades and exploits tentative population model result that relationship (Sommer et al. 1997). spawning habitats if and when they stated, ‘‘ * * * a long series of dry years Possible reasons for the lack of a stock- become available. Historically, this is unlikely to drive the splittail to recruitment relationship may be resilience is likely to have maintained extinction, even if the population is variation in female growth, survivorship the population of splittail through greatly reduced.’’ and fecundity from such causes as inter- extended droughts. This resilience also Our Response: A species warrants and intra-annual hydrologic variation, has allowed the splittail to persist in listing as threatened under the Act if is environmental contaminants, years of spite of the significant loss of habitat in danger of becoming endangered in non-spawning, predation, etc., which that has occurred since the species was the foreseeable future throughout all or may be exerting independent or first described by Ayres. a significant portion of its range (16 synergistic influences on recruitment of Comment 7: A respondent wished to U.S.C. 1532(20)). It is possible for the splittail into the population. Regardless know why the Bay Study and CVP and splittail to be undergoing threats or of cause, large portions of YOY fail to SWP salvage data showed an increase in declines in a significant portion of its survive to the adult, spawning splittail abundance, and the commenter range without declines showing in all population age class. Juvenile requested that we explain the variation surveys. Alternatively, threats to the abundance may therefore be inadequate in the study results. splittail may support listing even in the to fully describe the size of the standing Our Response: This comment pertains absence of our ability to document or future adult populations and may to the Meng and Moyle (1995) current population declines. Finally, we also be inadequate to describe the methodology employed in our previous believe the conservation elements of the ability of the population to persist. analyses of splittail population. We refer California State and Federal cooperative Population abundance cannot be the respondent to our Abundance program (CALFED) and the Central accurately predicted based upon section for a discussion of our most Valley Project Improvement Act examination of juvenile abundance recent statistical analysis of the species (CVPIA) programs adequately mitigate alone. population trends.

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00014 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55153

We believe that trends noted in the believe while unusual it is conservative, should be given greater weight for Bay Study are likely due to the large and results in a more robust making population-scale numbers of YOY fish that were collected determination of whether the species determinations. Weighting such a select during certain wet years. High outflows should be listed. group of surveys necessarily could may transport juveniles from the Comment 10: Several respondents require inappropriately combining their Estuary to locations where Bay Study cited an analysis that criticized our indices. The nine surveys are a samples are collected. It is unclear what treatment of separate surveys of splittail composite of appreciably different gear happens to these fish once they are abundance indices as statistically types, some of which suffer from the transported to these areas. Fish independent. same detection limitations as were used transported to San Pablo Bay may Our Response: We followed a long by other respondents to advocate against survive to join, if not sustain, the Napa established practice in the peer- accepting certain other surveys. Mid- River and Petaluma River and Marsh reviewed literature on splittail of water trawling is an inappropriate subpopulations. Once located in these treating these surveys as statistically match to splittail habitat preferences areas, it is not known what contribution independent (e.g., Meng and Moyle and other aspects of splittail biology, so is made to the Central Valley population 1995; Sommer et al. 1997) including even geographically extensive mid- as a whole. papers repeatedly cited by the water surveying would not necessarily In regard to trends in CVP and SWP respondents in previously submitted be any more representative of overall salvage data, we believe that these too comments. We accept at face value splittail populations than geographically are driven by seasonal variation in Manly’s (2002) conclusion that an more restricted surveys better matched hydrology. Though it is true that analysis of corrections among residuals to splittail biology. hydrology and production are strongly provides evidence for some degree of We disagree with respondents’ claims correlated, and that salvage would be interdependence among the different that Bay Study MWT, Bay Study OT, expected to rise as populations rise, sets of survey data (Manly 2002:4–6). and Fall MWT data are more indicative there are concerns with the data’s We also accept at face value Manly’s of splittail abundance trends than are application (see discussion of surveys (2002) attempt to correct our meta- those found in data collected at the SWP under Abundance section, above). analysis of survey results to account for and CVP salvage facilities, Chipps In the case of splittail salvage, the interdependence in the data sets. We Island, and in Suisun Marsh because entrainment is likely influenced by the have consistently stated that the they each suffer from gear or location rate, or volume per unit of time, of abundance index data for splittail suffer difficulties. We postulate that each of export. As stated before, salvage data are from several fundamental inadequacies these surveys is, to varying degrees, expressed as fish captures per acre foot that make them far from ideal for unsuited to the task of assessing splittail and lack a time value. At higher rates of decision-making purposes (an opinion abundance. The Bay Study OT employs export, splittail are likely to be with which the respondents and their the efficient otter trawling technique but disproportionately entrained because of statistical consultant concur (Manly only infrequently captures splittail; higher velocities in the channels 2002:3,8)). surveys are conducted on the periphery adjoining or approaching the facilities Comment 11: Several respondents of the species’ range. The Bay Study and thus, abundance could be criticized us for evaluating the results of MWT employs an inefficient (at overestimated. All sampling gears may the CDFG/USBR MLR Model for all 20 capturing splittail) mid-water trawl. The be more effective at capturing splittail data series of splittail abundance index Fall MWT fails to sample near-shore during high outflows due to increased data, instead of limiting the evaluation areas and the benthos (bottom), where velocity and turbidity, but only the to the nine data series that the splittail are most likely to occur. The pumps have the ability to draw fish respondents view as most representative Fall MWT does not sample shallow towards them at different rates. The rate of overall splittail populations. waters; in Suisun Bay/Marsh 8 of 25 at which fish may become pulled Our Response: We note that these and sites are shallow, 1 of 38 in the Delta are towards the pumps cannot be described other respondents have previously shallow. We acknowledge that the using existing data. Differing rates of criticized us, while employing different Chipps Island Survey is a midwater export also introduce variability, which analysis, for not treating all 20 data trawl of deep channels and that it too cannot be discerned without a time series equally and for not including all would suffer from a similar bias. The factor. Salvage data, as mentioned available data series in statistical CVP and SWP salvage data may suffer previously, do not effectively sample a evaluations of abundance trends. from an unquantifiable differential large extent of the splittail population, We are aware of no other party who entrainment based on export rates (see as fish reared in the Sacramento River has rigorously evaluated abundance Abundance section, above). and/or Yolo Bypass are likely to largely index data (e.g., Sommer et al. 1997; We also do not believe it is avoid the pumps. Salvage data do Meng and Moyle 1995; Moyle et al. 2001 necessarily correct to infer that the however collect the largest number of in prep.) that has deemed it appropriate wider geographical coverage of the nine splittail of any survey. to limit the evaluation to the nine data surveys in question, alone, is sufficient Comment 9: Several respondents cited series favored by the respondents. to guarantee that those surveys are more an analysis that took issue with us for Further, CDFG and USBR elected to representative of overall splittail adopting a non-traditional alpha-value include all 20 data series in the CDFG/ populations. The Bay Study MWT, Bay of 0.20 (instead of 0.05) for evaluating USBR MLR Model applications Study OT, and Fall MWT are results of the CDFG/USBR MLR Model. submitted to us as part of earlier geographically wider in distribution, but Our Response: Available literature comments. given that estuarine conditions are customarily demands a rigid adherence We disagree with the respondent’s specifically managed to maintain to the traditional alpha value of 0.05. In suggestion that only data from a select optimum habitat conditions within this particular analysis, we chose to take group of nine survey indices that Suisun Marsh, the wider survey areas of a far more conservative approach in sample a wide geographic area (we the Bay Study MWT, Bay Study OT, and terms of how we evaluated the splittail’s assume the respondent is referring to Fall MWT are not likely to contribute to abundance. Accordingly, we used the three age classes each of the Bay Study a more informed trend analysis. Surveys non-traditional alpha value of 0.20. We MWT, Bay Study OT, and Fall MWT) need not cover large areas if a fixed

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00015 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55154 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

point is likely to result in detection of techniques used by Meng and Moyle appeared to illustrate a precipitous drop an appreciable number of individuals of (1995), Sommer et al. (1997), or the in the splittail population. There are, a migratory species; splittail are as permutation-based exact calculations of however, up to 10 years of pre-drought likely to arrive at a static survey point p-values for stratified Mann-Whitney U- as well as up to 8 to 10 years of post- in a key location as they are to be tests, published on August 17, 2001 (66 drought data. The data collected during captured by a mobile survey of varied FR 43145). six years of continuous drought are but habitats. Comment 13: A respondent claimed a subset of the nearly 20 years of extant We understand the respondent’s logic that we were incorrect in departing splittail data. The splittail’s relatively in formulating a hypothesis that the significantly from the analysis of CDFG long life span and resilience following nine surveys in question might be most and USBR. unfavorable conditions renders the representative of the overall splittail Our Response: We did not depart at declines exhibited during a discrete population due to geography, but note all from the statistical analysis provided drought unlikely to influence the that at this point such an opinion is by CDFG and USBR in the form of the analytical findings from an ever- only a working hypothesis with no CDFG/USBR MLR Model. We have fully lengthening period of record. Most actual data available to either support or accepted the model results submitted by importantly, we now employ the CDFG/ refute it. Until such data become CDFG and USBR. We have noted earlier USBR MLR Model, which explicitly available, we believe it is most in our analysis where we have departed controls for potential confounding conservative to follow the practice of from the CDFG and BOR analysis and effects of hydrological year type. The evaluating all the data series rather than our reasons for doing so. respondent’s concern would be more combining or rejecting discrete sets. We Comment 14: Several respondents applicable to abandoned analytical continue to believe that, of the stated that the extended drought of 1984 techniques. The arbitrary pre- and post- individual indices, the Suisun Marsh to 1992 created only a perception of decline cut point approach of Meng and Otter Trawl should be the most decline and that it was the ‘‘*** Moyle (1985) was driven by trends appropriate sampling method because it accidental juxtaposition of a series of noticed during the 1987 to 1992 samples core splittail habitat, utilizes an wet, strong splittail years with a series drought, as was a formerly touted effective, bottom-trawling gear, and of dry, weak years that prompted [our] alternative analysis that involved the samples a greater relative proportion of interest in the first place.’’ use of 1987 (the beginning of the the habitat at the sampling site. Our Response: We disagree with the drought) as a cut point (Sommer et al. Comment 12: A respondent claimed respondent’s claims that the period of 1997) for determining percent declines. we employed ‘‘Shifting approaches to extended drought has been ignored, as We also disagree with the contention the splittail listing’’ in regard to well as with the contention that the that the 1984 to 1992 drought serves as statistical testing of available data. splittail’s drought-driven declines are the only factor which triggered our Our Response: Since we have the sole factor under consideration in investigations of the splittail’s status. published one listing notice for the our determination. We first note that the Our interest in the splittail was splittail, on February 8, 1999 (64 FR period of continuous drought is prompted initially by the statement in 5963), we assume that this respondent considered by most authoritative Daniels and Moyle (1983) that the is actually referring to our evolving sources to have begun in 1987 (Moyle et splittail’s and delta smelt’s ‘‘*** evaluations of data relevant to the issue al. 2001; Baxter 1999a; Sommer et al. abundance could decline rapidly if of whether the splittail should be listed 1997), not 1984 as reported by the environmental conditions become or not, as have appeared in the January respondent. We note, however, that unfavorable for them, possibly making 12, 2001 (66 FR 2828); May 8, 2001 (66 1985 and 1986 were dry years (Cannon them candidates for listing as threatened FR 23181); August 17, 2001 (66 FR 2001 in prep.). species.’’ We subsequently included the 43145); and March 21, 2002 (67 FR The declines noted during the 1987 to Sacramento splittail as a category 2 13095); and October 31, 2002 (67 FR 1992 drought were the likely result of a candidate species for possible future 66344), notices reopening public paucity of spawning habitat being listing as endangered or threatened in comment periods. available. The drought decreased the the January 6, 1989, Animal Notice of It is common practice in science to amount of floodplain (i.e. Yolo Bypass Review (54 FR 554). The candidate continually formulate and revise and mainstem river margins) available category system was abandoned on hypotheses in response to new for spawning and thus, spawning output February 28, 1996 (61 FR 7457), and information. We have applied this was lower. Low splittail population species meeting the definition of the scientific process during the review of densities were aggravated by the CVP former category 2 (such as splittail) were the splittail’s status, as have certain and SWP’s diversion of a greater no longer considered candidates. Our respondents (i.e. CDFG, CDWR, USBR). proportion of water from the Delta than administrative proceedings on splittail The evolving results of our various in prior years; fish were entrained at the resumed on November 5, 1992, when statistical analyses and the background facilities and the entrapment zone we received a petition from the Natural information describing the bases for (location where fish become vulnerable Heritage Institute to add the Sacramento those analyses have each appeared in to the export facilities’ effect on currents splittail to the List of Endangered and successive notices. Notices are in the Delta), was located well upstream Threatened Wildlife and to designate solicitations for public comment and of Suisun Marsh in increasingly critical habitat for this species in the information, not final agency actions. As suboptimal habitat. These events are Sacramento and San Joaquin Rivers and a result of new scientific information described in detail in our February 8, associated estuary. and comments received during the 1999, final listing rule (64 FR 5963). Comment 15: A respondent many comment periods, we have The basis for the respondent’s claim questioned how the data collected relate updated our analytical methodology that we are disproportionately to a conclusion that the species is based on the best scientifically and concerned with splittail declines noted threatened. We had not provided commercially available information. during the 1984 (or 1987) to 1992 analyses of population level outcomes Note also that neither we, nor drought is unclear. True, the ‘‘accidental that could be linked to threats analyses. respondents, have advocated nor juxtaposition’’ of wet and dry years Another respondent believed that our implemented a return to the superceded resulted in abundance data that threats analysis is speculative,

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00016 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55155

imprecise, and meaningless absent any mentioned previously in this document. excluded entirely. Our we believe our data or analysis concerning population CDFG conducts the Fall Midwater analysis using the CDFG/USBR MLR level effects and that the threats analysis Trawl, summer townet, and the Bay Model data series (see section entitled does not show why the species is study; we conduct the beach seine and Abundance) incorporates all applicable threatened because of the factors, as Chipps Island Survey; UC Davis YOY and adult abundance data, though required under section 4 of the Act conducts the Suisun Marsh OT, and excludes the beach seine due to its lack Our Response: We refer the USBR and CDFG collect the salvage and of a reliable catch per unit time respondent to the sections entitled creel census data. indicator (seine hauls do not accurately Summary of Factors Affecting the Comment 19: A respondent felt the account for time, or unit area per time Species and Conclusion Regarding 2000 Service beach seine survey data sampled). Beach seine data are best Abundance, Distribution, and Factors supported the respondent’s earlier employed with regard to the splittail for Affecting the Species. We believe that comments that splittail were not determining range and timing of the splittail does not qualify for declining. The respondent stated that occurrence. threatened status at this time based on new insights include: (1) YOY Comment 20: A respondent stated that our analysis of the threats. abundance was at a near record level in while splittail are able to persist in a Comment 16: Several respondents 2000; (2) distribution data show that in few key areas, such as Suisun Marsh asserted that the peer review process for years of low spring outflow (e.g., 1992, and the lower Sacramento River, during scientific publications doesn’t 1994, and 1997), the largest catches of periods of low flow, the relatively necessarily ensure that published young splittail occurred upstream in the smaller populations would be papers are unbiased, scientifically Sacramento River, upstream of many vulnerable to a large scale disaster (e.g., sound, and without errors. The sampling programs; and (3) splittail toxic spill), habitat loss, entrainment Transactions of the American Fisheries spawn and recruit even in dry years. mortality, reduced outflows, non-native Society does not use the double-blind Our Response: YOY abundance for a species predation, and contaminants. method for peer review. This issue was species with naturally high juvenile Our Response: We refer the raised in regards to our past use of the mortality does not necessarily equate respondent to the section entitled Meng and Moyle (1995) methodology to with high recruitment. The respondent’s Summary of Factors Affecting the determine splittail abundance. statement that distributional data show Species. Our Response: We agree with this that in years of low spring outflow (e.g. Comment 21: The respondent stated assertion. Each piece of scientific work, 1992, 1994, and 1997), the largest that the court requested that we provide: whether a peer reviewed published catches of young splittail occurred (1) An estimate of the current paper or an unpublished, unreviewed, upstream in the Sacramento River, is population size of the splittail; (2) draft report, must be objectively inaccurate for two of the three years determine whether or why the current evaluated for the scientific merit of its referenced, and faulty conclusions are populations size is inadequate to content alone. Peer reviewed drawn from the data. prevent extinction in the near future; (3) publication provides no guarantee of Water year 1992 exhibited similar determine the rate of population decline scientific merit. The test of time, abundances of splittail in upper of splittail; and (4) identify the following publication, provides the Sacramento River and Far North Delta minimum viable population size. In ultimate measure of scientific merit. locations, and moderate abundance addition, a respondent stated that the Indeed, subsequent iterative overall. Water year 1994 did exhibit hypothetical analytical model presented examination of the splittail’s status has relatively higher abundance in upstream at the January 29, 2001, CALFED Bay- resulted in our abandonment of Meng locations, but abundance was low Delta Program (CALFED Program) and Moyle (1995), Sommer et al., (1997) throughout all locations. Water year Splittail Science Conference and and our permutation-based exact 1997 was wet, not dry as stated by the described in the White Paper (Moyle et calculations of p-values for stratified (as respondent. Also, regardless of being a al., 2001) indicates that the splittail, opposed to unstratified) Mann-Whitney wet year, water year 1997 exhibited low even during severe and lengthy drought, U-tests (66 FR 43145). splittail abundance in all locations. is unlikely to be driven to extinction. Comment 17: Several respondents Further, we expect that YOY spawned Our Response: There remains no claimed we ignored the ‘‘White Paper’’ higher in the Sacramento River to suffer proven scientific method for published by Moyle et al. (2001) higher mortality, relative to fish determining the current splittail Our Response: We agree with and use spawned in the Delta, as they migrate population size primarily because no the various findings and hypotheses downstream through progressively- extant survey was designed specifically found in the draft and revised White worsening habitat conditions to rejoin to monitor splittail populations or to Paper (Moyle et al. 2001) extensively the core population. Increased mortality determine their absolute numbers. throughout this document. We believe among splittail spawned upstream may Further, the splittail exhibits relatively that the draft White Paper is a useful explain why YOY tend to be captured wide variation in annual abundance in resource and contributes to the less frequently in downstream trawl- response to prevailing hydrologic knowledge on splittail biology, though it based surveys in certain dry years. The conditions; it is likely that the has not yet been finalized. The paper final statement, that splittail spawned population size exhibits appreciable has been referenced throughout this upstream exhibit successful spawning year to year variability which would document. and recruitment in dry years, is not confound size estimates. Comment 18: A respondent requested supported by survey data. While Calculating the current population’s that we acknowledge that the spawning success can be inferred from risk of and/or time to extinction would Interagency Ecological Program (IEP) YOY abundance, YOY fish do not require estimates of absolute population provides oversight for fisheries data necessarily recruit to the adult size, rate of decline, and minimum collection. population. There is some evidence that viable or sustainable population size, Our Response: We concur that the IEP high or low YOY abundance is none of which currently exist in a has oversight of the various fishery correlated, with a two to three year time scientifically defensible form. Moreover, programs. However, various agencies lag, with adult abundance. For this it must also be noted that the statutory collect the data for the surveys reason, YOY abundance cannot be and regulatory standard for ascertaining

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00017 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55156 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

threatened status is not to determine demonstrating a decline, might be due agree with the respondent that the whether or why a species will become to a shift in the splittail’s distribution, splittail had evolved subject to the extinct in the near future, but if, rather than a decline in numbers. variability of California’s climate and pursuant to section 3(19) of the Act, it Our Response: Data do suggest that has adapted to survive this variability. ‘‘* * * is likely to become an splittail shift their distribution in Comment 26: One respondent stated endangered species within the response to salinity conditions, and that that pesticide application is not a threat foreseeable future throughout all or a they are quick to respond and move into to the splittail because no data were significant portion of its range’’. An an area when conditions become presented to support the assumption endangered species, pursuant to section favorable (see Background section). that pesticides bioaccumulate in fish to 3(19) of the Act, is that ‘‘* * * which However, we believe our survey the point of causing morbidity, is in danger of extinction throughout all information is robust enough to detect a mortality, or reduced reproduction. or a significant portion of its range decline (see Abundance section) Several respondents took similar ***’’. Comment 23: One respondent exception to our statements regarding As stated above, analytical techniques objected to our determination in the the need for pesticide use on crops to do not exist to determine the rate of January 12, 2001, notice (66 FR 2828), be assessed and possibly regulated. The splittail population decline with current that rock revetment, or riprap, as it respondent also claimed pesticides were splittail data. Again, the absence of presently exists or is proposed, would no more of an environmental problem survey methodologies specifically have any significant impact upon the within the bypasses than in other areas designed to monitor splittail splittail. and that there was no reason to justify populations is a limiting factor in Our Response: While a general separate or additional regulatory determining rate of decline. An estimate dismissal of riprap and other types of programs that would apply only to the of splittail population decline, in the levee and bank protection is likely bypasses. A respondent stated that form of an exponential decay model, overly broad, the application of riprap pesticides may be present, but that they was included by in our August 17, 2001, and other bank treatments that has have been flushed from the bypasses notice (66 FR 43145) but was not used occurred throughout the splittail habitat prior to spawning. Another respondent in this document because of has resulted in the decreases in habitat stated that much of the pesticide respondents’ concerns that it is that have led to this examination of the loading in the Yolo Bypass was due to insufficient to describe the interactions status of the species. Bank protection runoff from upstream sites. in a complex aquatic ecosystem. Further can be placed on levees and riverbanks Our Response: Please see our our exponential decay model relied without damaging habitat, but it must be discussion under threats. In general, upon the results of the CDFG Mann- done so with explicit considerations for there are findings that have heightened Whitney U test results. The CDFG the habitat needs of the affected species. our concern regarding these substances. Mann-Whitney U test results have since Our analysis in this rule accepts that However, there is little data on the been superceded by the CDFG/USBR premise as part of our underlying direct affects to splittail. MLR Model. Lastly, there exists no review of the CALFED and CVPIA Comment 27: A respondent felt that method to determine the splittail’s contemplated actions. we were inconsistent when it was stated minimum viable population because, Comment 24: A respondent asked if in the January 12, 2001, reopening of again, no current survey was designed we would address the impacts of comment period (66 FR 2828), that specifically to monitor splittail boating and other activities affecting wetland rehabilitation could be population size. near-shore habitat. deleterious to the splittail, but that Since the publication of the Final Our Response: The impacts of boating wetland habitat improvements within Rule listing the splittail as threatened, a are not considered a significant source the species’ range would be beneficial. hypothetical analytical model was of habitat loss. In many regions of the The respondent felt we had not ‘‘*** developed and presented at the January Delta, wave wash is a natural integrated its concepts and concerns in 29, 2001, CALFED Bay-Delta Program phenomenon related to winds crossing a manner that weighs relative risks and (CALFED) Splittail Science Conference. areas of great fetch (open areas). The concepts.’’ The model is described in detail in the splittail evolved with the effects of wave Our Response: We agree with the White Paper (Moyle et al., 2001). wash within near-shore habitat. respondent that wetland restoration Service staff attended the Comment 25: One respondent differed projects are generally beneficial to aforementioned conference and are with the determination in our January splittail. aware of the model. A second review 12, 2001, notice (66 FR 2828), that Comment 28: A respondent felt that draft was provided to us on June 18, California’s variable Mediterranean our statement that the present operation 2001. climate is a threat to native fish, and of Federal, State and private water We believe that the model is, at contended instead that it favors native development projects, that entail water present, only a tool for testing existing fish over non-native fish. The storage, diversions, re-diversions, and hypotheses and for generating new respondent also stated that the splittail agricultural return flows, destroy hypotheses. Certain findings may be had evolved subject to the vagaries of splittail habitat was incorrect. interpreted to support listing and others California’s climate and was adapted to Our Response: We refer the may counter it, but we have determined survive them. respondent to the section entitled that neither is sufficiently robust to be Our Response: Our notice stated that Summary of Factors Affecting the included in this final document. Indeed, ‘‘The variability of California’s Species. once refined by the incorporation of Mediterranean climate exacerbates the Comment 29: A respondent felt that more accurate data, the model may be threats (emphasis added) * * *’’ to the we did not adequately acknowledge the useful for determining those mitigation splittail. The Mediterranean climate positive environmental effects of the and restoration efforts likely to have the includes periods of extended normal CVP and SWP. The respondent greatest benefit to the splittail. and above-normal precipitation but may specifically noted that the inland extent Comment 22: A respondent claimed also include periods of extended of saltwater intrusion into the Delta is that any decline evident in the Suisun drought. Splittail evolved under these currently lower than with the ‘‘without- Marsh OT data, or in any other survey conditions and are adapted to them. We project’’ condition.

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00018 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55157

Our Response: We do not consider the final document contains a discussion of stranding in these basins, no data were pre-SWP and CVP extent of saltwater the range of the splittail. provided to indicate that these basins, intrusion to be detrimental to splittail. Comment 31: Nearly all respondents in their unaltered state, were a source of Saltwater intrusion was defined by the felt that we should not classify the Yolo mortality sufficient to cause a decline of respondent as the location of the and Sutter bypasses as a threat to the the species. There were no hydrologic chloride concentration of 1000 splittail, as we did in the January 12, data provided to indicate when the milligrams per liter (mg/L)(1000 parts 2001, reopening of comment period (66 historic basins would have become per million (ppm)), measured 90 FR 2828), based primarily upon the data connected or isolated from the minutes after high tide. It is not clear if found in Sommer et al. (1997) and Sacramento River in a typical year. the inference is that brackish water such Sommer (2001a). The bypasses have These basins, being situated lower than as this is detrimental to the splittail. demonstrated the capability of the adjoining river and likely Splittail occupy brackish water at producing large numbers of splittail maintaining an alluvial (stream bed various stages of their life and such when inundated. sediment) water connection, may have habitat may actually be essential to the Our Response: We have determined, existed as perennial marshes wherein species’ life history. The 1000 parts per based on consideration of scientific data splittail could persist until inundation million value is equivalent to 1 part per and information provided by was restored. Indeed, the White Paper thousand (ppt), which differs little from respondents, that the Yolo and Sutter (Moyle et al. 2001) states that splittail the 2 ppt standard identified as X2. The bypasses are not, in and of themselves, historically occurred in alkaline lakes White Paper (Moyle et al. 2001) a threat to the splittail. Our reevaluation on the valley floor. The Butte Basin includes numerous references to the use of this issue is discussed in Factor E of remains connected to the Sacramento of brackish water near X2 by splittail, the section entitled Summary of Factors River via the Sutter Bypass and Butte indicating that it may actually Affecting the Species. Creek; splittail are known to spawn in characterize optimal rearing habitat for Comment 32: Some respondents this area (Baxter 1999a). fish greater than 75 mm (3.0 in) in stated that the bypasses, the Sacramento It is also possible that, for the standard length (typically late year 0 or River Flood Control System, and other American River, Feather River, and early year 1 fish). Non-reproductive reclamation and flood control efforts are other eastside streams, pre-European (rearing juvenile and adult) splittail are beneficial to the splittail because they habitat conditions contained more most abundant in shallow brackish tidal redirect water into the Sacramento River complete and/or longer duration sloughs, such as those found in Suisun that, prior to the 1920s, would have surficial (surface water) hydrologic Marsh. Growth of splittail in brackish spilled into the Colusa, Yolo, Butte, connections between rivers and sinks sloughs is rapid in the first year of life, Sutter, and American basins, thus than they did following the period of with fish reaching a size of 12 to 14 cm entraining significant numbers of fish. massive hydraulic mining. Hydraulic (4.7 to 5.5 in) TL. Further, historic, pre- Our Response: Splittail evolved in the mining resulted in massive deposition reclamation conditions in the Delta Central Valley and we postulate that the of sediments in the beds of many would have allowed the ‘‘natural’’, non- species is likely evolutionarily equipped eastside streams. The streambeds then SWP and CVP manipulated X2 location to exist in the presence of natural flood became elevated. Rivers began to to exist within extensive flooded basins inundated during unaltered meander, as gradient and sinuosity are wetlands. Also note that splittail have hydrologic conditions. The splittail’s inversely related. When hydraulic wide salinity tolerance (10 to 18 ppt) high salinity tolerance (see Background mining ceased, the rivers began to (Moyle 1976; Moyle and Yoshiyama section, above) also indicates its ability straighten, eroding back through the 1992), with an absolute observed to persist in detached, increasingly deposits, and leaving elevated banks as tolerance of 29 ppt for short periods saline waters. The number of effective barriers for the basins’ receding (Young and Cech 1996). Inland brackish confounding factors as well as lack of flood waters. These elevated banks water intrusion may have thus been at any historic data severely limits our could have exacerbated the tendency for tolerable or even desirable ability to assess with any real authority the rivers to become disconnected from concentrations for the species. We do the ultimate effect of the Sacramento the natural basins. not consider the changes in estuarine River Flood Control System, the CVP Comment 33: Several respondents felt hydrology induced by the SWP and CVP and the SWP. Following is our assumed that our determination that the Sutter to be beneficial to the splittail and scenario regarding the effects on splittail and Yolo bypasses would require traditionally the Service and other of past reclamation and flood control inundation for at least 30 continuous wildlife agencies have accepted as fact efforts. However, we acknowledge that days between March and April in order the supposition that splittail habitat was alternative assumptions and for them to be considered a beneficial degraded as a result of the operation of conclusions could be drawn from splittail spawning habitat was these projects (see the section entitled existing information. inaccurate and could affect water Summary of Factors Affecting the Reclamation activities, including the supply and flood management. Another Species). Sacramento River Flood Control Project respondent indicated that constant Comment 30: The court directed that and similar efforts to prevent flooding of flows, related to inundation of the we respond to the issue that splittail urban and agricultural lands, have bypasses, would favor non-native fish. have a broader distribution than resulted in the confinement of the Our Response: We have not proposed previously thought, including a broader Sacramento River primarily to a single, inundation of the bypasses for any range in the Sacramento and San leveed or otherwise artificially-confined specific interval, duration, or frequency. Joaquin Rivers. Another respondent channel, with much of the former Rather, we have suggested that the noted that larval, Age 0 and Age 1 American and Colusa basin habitat no bypasses would have their greatest splittail have all been collected above longer available to fish occupying the benefits to splittail if they became the Delta. mainstem river. The respondent claimed inundated at a frequency and duration Our Response: The greater range of this was a benefit in that splittail were that as closely as possible mimics the the splittail was acknowledged in the no longer subject to entrainment in natural, precipitation-driven January 12, 2001, notice (66 FR 2828). these basins. While it is true that hydrograph. The reference to 30 days is The above Background section of this splittail are no longer subject to a statement regarding how the

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00019 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55158 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

inundation patterns of the bypasses at Our Response: The ESA does not The respondent discouraged times do not meet the life history allow us to consider cumulative impacts assessments of effects to splittail that requirements of the splittail. Inundation of multiple species listings and critical would occur based upon of bypasses in dry years would reduce habitat designations when making a implementation of projects that will be the effects of drought on the splittail. listing determination. constructed and/or operated in manners We also speculate that if the bypasses Comment 38: A respondent stated that that cannot be substantially verified at were inundated at a frequency and sport fishing take of other listed species, present, such as those in CALFED and duration that as closely as possible specifically salmonids, is a significant the CVPIA. mimics the natural, precipitation-driven source of mortality of splittail caught Our Response: We agree and have hydrograph, then the numbers of non- unintentionally and asked if the listing revised and reevaluated the threats native fish would be reduced, as non- of splittail would include measures to presented by existing conditions and native fishes favor ponded and protect the species from this threat. projects (see Summary of Factors continuously inundated habitats. Our Response: We concur that sport Affecting the Species section). Comment 34: Certain respondents felt fisheries can be a source of mortality for Comment 42: Several respondents that compensation should be provided splittail caught unintentionally. believed that full implementation of the to land owners when habitat However, since we have determined CALFED Program would preclude the restorations affected land use. that listing as a threatened species is not need to list the splittail and indicated Our Response: If habitat restorations warranted for the splittail, this notice that over $10 million had been spent on affect land use, there is a separate does not include restrictions on actions that could improve conditions process available to landowners for sportfishing. for splittail. Comment 39: Several respondents Our Response: We agree that actions redress. While we do not anticipate that objected to our statements regarding the taken under the CALFED program have efforts to restore the habitat will result entrainment risks present in the contributed to the current in substantial changes in the land use bypasses based upon Sommer et al.’s improvements in habitat that affects the practices in the bypasses, the (1997) findings that entrainment is not splittail and anticipate that other actions regulations governing listing [50 CFR a significant threat within the bypasses. of that type are forseeably likely to § 424.11(b)] state that listing of a species It is thought that the splittail’s occur. (We refer the respondent to the as threatened or endangered is made evolutionarily-derived ability to sections entitled Summary of Factors ‘‘* * * solely on the basis of the best emigrate prior to stranding reduces the Affecting the Species.) available scientific and commercial risk of stranding. Respondents felt that Comment 43: Various respondents information regarding a species’ status, the magnitude of the entrainment informed us of the contents of an April without reference to possible economic threats presented by the bypasses was 24, 2001, Sacramento Bee article or other impacts of such a overestimated when we cited in the wherein Dr. Peter B. Moyle, a determination.’’ Accordingly, we do not January 12, 2001, notice (66 FR 2828), recognized expert in aquatic ecology, consider or address this issue in our the death of a number of juvenile fisheries science, and the splittail, listing decision. splittail in an approximately 0.8 hectare discussed the February 8, 1999, listing Comment 35: Several respondents (ha) (2 acre (ac)) borrow pit as of the splittail as threatened. commented that our classification of the statistically-significant and that the Respondents related Dr. Moyle’s Yolo Bypass as a threat in the January classification of ‘‘natural sinks’’ as a statement that ‘‘Things were getting 12, 2001, notice (66 FR 2828) would threat was in error. better’’ and argued that it constituted an undermine potential ecosystem Our Response: We have considered opinion that the species should not have restoration actions that would benefit these data and now agree that been, and by inference, should not now the splittail. entrainment in the Yolo Bypass is less be listed. Our Response: In this notice, we have than was originally thought. Information Our Response: We have read the determined that the Sutter and Yolo presented at the January 29, 2001, article in question. We cannot conclude bypasses are not in and of themselves CALFED Splittail Science Conference that Dr. Moyle was making a statement threats. indicates that a modest degree of on the listing status of splittail. The bypasses remain important topographic variability within an However, we do note that ecosystem splittail spawning and rearing habitat inundated area may be beneficial, as it improvements are a primary reason why during wet periods. Sommer et al. may create a diversity of flow patterns we are removing the listing. We have (1997) and Sommer et al. (2001a, 2001b) and velocities which in turn may allow cited several of Dr. Moyle’s scientific found that the bypasses as they exist juvenile splittail to evade predation and publications and conclusions within today, and when flooded, already forage more effectively during egress. this document. provide substantial amounts of habitat. Comment 40: A respondent described Summary of Factors Affecting the Comment 36: A respondent claimed that many of the non-native species of Species that this determination could not be the Delta have arrived via the discharge promulgated because it was not likely to of ballast water from seagoing vessels After a thorough review and include the required critical habitat and asked if the listing of the splittail consideration of all the best scientific designation or the preparation of a would result in the regulation of and commercial information available, recovery plan. maritime trade. we have determined that the listing of Our Response: We have determined Our Response: As we have the Sacramento splittail as a threatened that listing as a threatened species is not determined that listing as a threatened species should be removed. We warranted for the splittail, and therefore species is not warranted for the splittail, followed procedures found at section the designation of critical habitat is not this notice does not include restrictions 4(a)(1) of the Act and regulations (50 warranted. on maritime trade. CFR part 424) implementing the listing Comment 37: A respondent claimed Comment 41: A respondent stated that provisions of the Act. A species may be that we must consider the cumulative we should consider only project- determined to be endangered or impacts of multiple species listings and induced effects associated with existing threatened due to one or more of the critical habitat designations. projects and their associated operations. five factors described in section 4(a)(1).

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00020 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55159

These factors, and their application to Pumping Plants (Cannon 2001 in prep.) from Federal funding), to make progress our decision to remove from the list the states that ‘‘* * * lower population toward achieving its goals which Sacramento splittail as threatened, are levels occurring as a consequence of include restoration and enhancement of as follows: salvage-entrainment related mortality splittail habitat (CALFED 2000a, 2000b). A. The present or threatened may be reducing population resilience While CALFED is not meeting the destruction, modification, or (e.g., less dependence on a single age expected schedules, the individual curtailment of its habitat or range. We class) and jeopardizing the long-term actions are occurring generally within have identified, as threats to the viability and ecological role of splittail the scope of their own schedules splittail, the present operation of in the estuary.’’ If entrainment mortality (CALFED 2000a, 2000b). With respect to Federal, State, and private water increases further, it could be expected to splittail actions, CALFED has identified development projects entailing water have even greater adverse effects on the the plan to be implemented, as well as storage, diversions and re-diversions, splittail. In addition, reservoir the funding level, funding sources, and releases, flood control, and export and operations and ramping rates for flood other resources necessary to implement agricultural return flows, which control inadvertently drain shallow it (CALFED 2000a, 2000b). In addition, destroyed splittail habitat (59 FR 682, 64 water spawning habitat along river CALFED has identified the appropriate FR 5963, 66 FR 2828). Each is discussed corridors and exacerbate stranding of authorities as well as the legal, briefly below as are the beneficial effects splittail. regulatory, and procedural requirements of CALFED and the CVPIA, which offset necessary to implement the some of these threats. Beneficial Actions Offsetting Adverse Affects conservation effort. Importantly, Habitat Loss: The Bay Institute (1998) CALFED has completed the has estimated that intertidal wetlands in A number of beneficial actions offset environmental reviews and the Delta have been diked and leveed so the above described adverse affects. consultations necessary to proceed with extensively that only approximately Below are some of the specific actions its proposed actions. CALFED describes 3,237 ha (8,000 ac) remain of the or programs describing the beneficial the nature and extent of threats being 161,875 ha (400,000 ac) that existed in actions. addressed, and addresses the threats to CALFED Habitat Restoration: The 1850, and that 90 percent of the riparian the splittail through its tidal and CALFED Bay-Delta Program (CALFED) forest and riparian wetlands of the riparian habitat restoration projects, fish exists as a multi-purpose (water supply, Sacramento Valley have been cleared, screen projects, environmental water flood protection, and conservation) filled, or otherwise eliminated. Diking, program, water quality program and program with significant ecosystem dredging, filling of wetlands, and numerous other programs (CALFED restoration and enhancement elements, reduction of freshwater flows through 2000a, 2000b). CALFED defines its and is well into its implementation more than half of the rivers, distributary conservation objectives in terms of sloughs, and the estuary for irrigated phase (CALFED 2000a, 2000b). The recovery of targeted species, including agriculture and urban use have widely stated mission of CALFED is to develop the splittail, and has identified the steps reduced fish habitat and resulted in a long-term comprehensive plan that necessary to implement the program extensive fish losses (Moyle et al., 1995; will restore ecological health and (CALFED 2000a, 2000b). The goal of Nichols et al., 1986). improve water management for all There has been loss and degradation beneficial uses of the Bay-Delta system CALFED to recover the splittail will of the near-shore habitat required by (CALFED 2000a, 2000b). The plan remain whether the splittail is listed or splittail. Riparian and natural bank specifically addresses ecosystem not (CALFED 2000a, 2000b). CALFED habitats are features that historically quality, water quality, water supply, and has identified and employed provided natural function to the stream levee system integrity (CALFED 2000a, quantifiable, scientifically valid banks and flood plains for splittail by 2000b). CALFED encompasses eight parameters to demonstrate achievement providing spawning substrate, organic separate program elements; each having of objectives and the standards by material, food supply, and cover from disparate potential effects to the splittail which progress is to be measured predators. Vast stretches of the (CALFED 2000a, 2000b). (CALFED 2000a, 2000b). CALFED Sacramento and San Joaquin Rivers, CALFED is a cooperative effort of the monitors and reports on progress their tributaries, and distributary U.S. Department of the Interior, the U.S. towards implementation (based on sloughs in the Delta have been Department of Commerce, the compliance with the implementation channelized and the habitat converted Environmental Protection Agency, the schedule) and effectiveness (based on or destroyed. California Environmental Protection evaluation of quantifiable parameters) of Delta water diversions and exports Agency, and the California Resources the conservation effort (CALFED 2000a, currently total 1.1 hectare-meters (ha-m) Agency, as well as other State and 2000b). Adaptive management has been (9 million acre-feet (MAF)) per year. Federal agencies, with the involved incorporated into CALFED (CALFED These diversions and exports also harm public formally participating originally 2000a, 2000b). the splittail. The Federal and State through the Bay-Delta Advisory Although the splittail reared in the water projects presently export as much Council, and currently through the Bay- Sacramento River and/or Yolo Bypass as approximately 740,000 ha-m (6 MAF) Delta Public Advisory Committee are likely to largely avoid the CVP and per year from the Delta when sufficient (CALFED 2000a, 2000b). CALFED is a SWP pumps, in the absence of any water is available. Agricultural long term effort with an initial, shorter consideration of the splittail in the diversions for lands within the Delta term implementation strategy (CALFED CALFED process, the splittail’s status range from 7,400 to 160,000 ha-m 2000a, 2000b). The Record of Decision could be adversely affected by program (60,000 acre-feet to 1.3 MAF); (ROD) for CALFED was signed in elements to increase water storage in the approximately 123,000 ha-m (1.0 MAF) August, 2000. Central Valley upstream of the Delta; per year in the long term period, CALFED has has received sufficient modify Delta hydrologic patterns to 136,000 ha-m (1.1 MAF) in critical and funding (approximately 80 percent of convey additional water south, and dry years (CALFED 2000b). The draft funding required from the State of upgrade and maintain Delta levees. White Paper entitled Factors Relating to California, from CVP and SWP water However, as noted previously CALFED Salvage of Splittail at South Delta project users and local entities, and has an explicit goal to balance the water

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00021 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55160 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

supply program elements with these the through additional pumping at times Francisco Bay-Sacramento-San Joaquin restoration of the Bay-Delta and safe for fish (CALFED 2000a, 2000b). Delta Estuary.’’ We also note that the tributary ecosystems and recovery of the Already the EWA has demonstrated CVPIA is a mitigative effort for past splittail and other species. Because some success. In its first year, the impacts of the CVP, and like CALFED, achieving the diverse goals of the account provided 35,400 ha-m (287,000 is a multi-purpose program that, at full program is iterative and subject to acre-feet) of water for environmental implementation, will include both annual funding by diverse agencies, purposes without reducing allocations beneficial ecosystem restoration CALFED has committed to maintaining to agricultural and urban users. The elements as well as water supply, water balanced implementation of the EWA thus has functioned as a conveyance, and flood control projects, program within an adaptive mechanism for providing for improved all of which are required to be management framework (CALFED Delta conditions for splittail. implemented in a manner that considers 2000a, 2000b). Within this framework of A review of the CALFED ERP projects the needs of the environment, rather implementation, it is intended that the shows that as of June 2002, the ERP has than just maximizing flood control and storage, conveyance, and levee program funded: 58,300 acres of habitat proposed water supply and delivery which was elements would only be implemented in for protection, including 12,000 acres the case in the past. such a way that the splittail’s status dedicated to wildlife friendly The CVPIA exists as a multi-purpose would be maintained and eventually agriculture and 16,000 acres of (water supply, flood protection, and improved (CALFED 2000a, 2000b). The floodplain; 39,000 acres of habitat conservation) program with significant restorative components of CALFED will proposed for restoration, including ecosystem restoration and enhancement positively influence the status of the 9,500 acres of shallow water tidal and elements and has been approved by all splittail; these are the Ecosystem marsh habitat; 63 miles of upstream the affected parties including the FWS. Restoration Program (ERP), the Multi- habitat proposed for protection and/or It is well into its implementation phase Species Conservation Strategy (MSCS,) restoration; 93 miles of riparian corridor and is fully funded. While the CVPIA is and the Environmental Water Account proposed for protection and/or not meeting the expected schedules, the (EWA) (CALFED 2000a, 2000b). restoration; 72 fish screens accounting individual actions are occurring CALFED has identified 29 species for an additional 2,565 cfs of diversion generally within the scope of their enhancement conservation measures for capacity screened; 15 fish ladders and schedules. The CVPIA has identified the splittail (CALFED 2000a, 2000b). These 10 dam removals to provide better plan to be implemented, as well as the measures include a variety of actions upstream passage; 31 projects involving funding level, funding source, and other consistent with our conservation analysis of environmental water and resources necessary to implement it. In strategy. sediment quality; 18 projects intended addition, the authorities, and the legal, CALFED’s Ecosystem Restoration to specifically address nonnative regulatory and procedural requirements Program includes the development and invasive species; and 75 projects necessary to implement the implementation of a program to address supporting local watershed stewardship conservation effort have been identified. flows resulting from the present and environmental education (CALFED Finally the necessary environmental operation of Federal, State, and private 2002). Clearly substantial efforts are reviews and consultations have been water development projects, entailing underway to continue to restore and completed. The CVPIA describes the water storage, diversions and re- develop optimum splittail habitat. nature and extent of threats being diversions, releases, export and Full implementation of the 30 year addressed, and addresses the threats to agricultural return flows (CALFED program will require both State and the splittail through its tidal and 2000a, 2000b). This includes the Federal funding and is expected to riparian habitat restoration projects, fish development of a methodology for require both annual appropriations by screen projects, environmental water evaluating Delta flow and Congress and continued funding by the programs and numerous other programs. hydrodynamic patterns and State of California. To date, the federal The CVPIA’s conservation objectives are implementation of an ecologically based government has spent over $700 million defined in terms of recovery of targeted plan to restore conditions in the rivers on CALFED, and the overall species, of which the splittail is one, and sloughs of the Delta sufficient to expenditures for the first 3 years of the and has identified the steps necessary to support targets for the restoration of program exceeds $2 billion; all of which implement the program. The program aquatic resources, including splittail has been spent for environmental has identified and employed (CALFED 2000a, 2000b). restoration. quantifiable, scientifically valid The EWA’s stated purpose is to CVPIA Habitat Restoration: The parameters to demonstrate achievement provide benefits to threatened or Central Valley Project Improvement Act of its objectives and the standards by endangered fish without causing (CVPIA) (Public Law 102–575) signed which progress is to be measured. The additional adverse impacts on water October 30, 1992, amends previous CVPIA monitors and reports on progress deliveries from diversions and the authorizations of the Central Valley towards implementation (based on export facilities (CALFED 2000a, Project (CVP) (16 U.S.C 695d-695j) to compliance with the implementation 2000b). The EWA, not analyzed in the include fish and wildlife protection, schedule) and effectiveness (based on February 1, 1999, final rule (64 FR restoration, and mitigation as project evaluation of quantifiable parameters) of 5963), or in the January 12, 2001, notice purposes having equal priority with the conservation effort. (66 FR 2828), purchases water from irrigation and domestic water supply, Provisions of the CVPIA to benefit willing sellers, then banks, stores, and fish and wildlife enhancement fish and wildlife habitat include transfers and releases it as needed to having equal priority with power protection and restoration of natural protect fish and compensate water users generation. Two of the stated purposes channel, riparian, and wetland habitats (CALFED 2000a, 2000b). The EWA has of the CVPIA are to ‘‘protect, restore, (sections 3406(b)(1) and 3406(d)), set a goal of acquiring at least 23,400 ha- and enhance fish, wildlife, and dedication and management of 98,680 m (190,000 acre-feet) of water each year associated habitats in the Central Valley ha-m (800,000 ac-ft) of CVP yield through purchases, but also expects to * * * of California’’ and ‘‘to contribute (section 3406(b)(2)), acquisition of obtain additional 23,400 ha-m (190,000 to the State of California’s interim and additional water supplies to supplement acre-feet) of water on average each year long-term efforts to protect the San the amount dedicated (section

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00022 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55161

3406(b)(3)), modification of CVP Delta in 2001 (Coastal America 2000). sheds, etc.) followed by replacement operations (sections 3406(b)(1) and 3406 The project is likely to result in the with native vegetation appropriate to (b)(19)), removal of fish migration restoration of approximately 243 ha (600 each site (USFWS 2002a). A portion of barriers (sections 3406(b)(10) and ac) of open water, 134 ha (330 ac) of these actions are expected to benefit 3406(b)(17)), screening of water tidal emergent marsh, and 95 ha (235 ac) splittail through the improvement of diversions (section 3406(b)(21)), and of mud flat within Prospect Island’s vegetative conditions on floodplains acquisition of land and associated water approximately 486 ha (1,200 ac) and the eventual creation of large rights (section 3408(h)), among others. interior. These may represent habitat woody debris (via riparian tree mortality Funding sources for CVPIA mitigation enhancements for splittail. and entrainment). and restoration actions include the Restoration efforts have been The Vic Fazio Yolo Bypass Wildlife CVPIA Restoration Fund; State funds undertaken at the Cosumnes River Area (Wildlife Area), located within the provided to meet CVPIA cost share Reserve under management by the Yolo Bypass, will increase in size from requirements; and additional Federal Bureau of Land Management (BLM), its current approximately 1,497 ha funds appropriated by Congress. The Nature Conservancy, and a number (3,700 ac) to approximately 5,261 ha Two programs, the CVP Conservation of other agencies and private (13,000 ac) (The Nature Conservancy Program, and the CVPIA Habitat organizations (The Nature Conservancy 2002b). This increase was not analyzed Restoration Program, were created to 2002a). Restoration activities that in the February 1, 1999, final rule (64 proactively restore and improve the benefit splittail include riparian FR 5963), or in the January 12, 2001, Central Valley environment that was or enhancement and intentional breaching notice (66 FR 2828). Though the is being impacted by the operations of of levees to restore floodplain function. Wildlife Area does contain entrainment the CVP. These two programs have Restoration is ongoing and splittail are hazards, and is located along the provided funding to a number of likely to benefit from any efforts, as the slightly less infrequently inundated projects which collectively would area has also been described as among western edge of the Yolo Bypass, it will double the acres of riparian forest on the the most important floodplain habitats incorporate opportunities to restore the Sacramento River (from approximately still available to the species (Moyle et al. lower reaches of Putah Creek. The 8,093 ha (20,000 ac) to 16,188 ha 2001). added area may allow restorations to (40,000 ac)) and to contribute to the CDWR has also completed an proceed that benefit splittail to a greater recovery of threatened and endangered ecosystem restoration on Decker Island, degree than possible with the current species (Carlton 2003 in prep.). located on the Sacramento River, shorebird and waterfowl-intensive Combined efforts of Federal, state, and adjoining Sherman Island near the management regime. nonprofit partnerships have reforested confluence with the San Joaquin River Other State efforts may contain almost 1,619 ha (4000 ac) between Red (CDWR 1998). The project has restored actions beneficial to the splittail which Bluff and Colusa during the last 15 years approximately 4.45 ha (11 ac) of shallow were not analyzed in the February 1, (Carlton 2003 in prep.). Riparian forest water habitat that is likely to be utilized 1999, final rule (64 FR 5963), or in the restoration would, over time, also by the splittail. The California January 12, 2001, notice (66 FR 2828). increase the amount of large woody Department of Transportation has Assembly Bill (AB) 360, the State Delta debris habitat available to splittail. committed to restore 190 ha (470 ac) of Flood Protection Act, has a primary Section 3406(b)(2) of the CVPIA tidal marshes within the range of purpose of strengthening Delta levees dedicates 98,680 ha-m (800,000 ac-ft) of splittail for the benefit of splittail as with various ‘‘hard’’ measures, CVP yield annually to implement fish, compensation for impacts resulting from including riprap. Habitat restoration wildlife, and habitat restoration, and to the construction of the Benicia Martinez components of AB 360, more properly help federally listed species. A portion New Bridge (USFWS 2003a). considered mitigation for concurrent of the 98,680 ha-m (800,000 ac-ft) USACE and CDFG are currently in the State projects’ impacts to aquatic and identified in the CVPIA may be used to final stages of planning the Napa River terrestrial ecosystems in the Delta do meet the Department of the Interior’s Salt Marsh Restoration Project (USFWS require improvement rather than a strict obligations under the Bay-Delta Accord 2003b). Approximately 1,262 ha (3,120 mitigation approach which results in an (discussed below). The rest of the water ac) of diked salt ponds would be increased habitat benefit and a net can be used for instream flows, restored to tidal marshes usable by increase in habitat. The State Senate Bill additional Delta outflow, and the other splittail. (SB) 1086-funded Sacramento River purposes of the CVPIA. Management of The 44 ha (109 ac) Kimball Island Conservation Area is an interagency dedicated, supplemental, and Mitigation Bank reestablished riverine group chartered to promote and guide reoperated CVP yield will benefit aquatic bed, riparian forest, shaded protection and enhancement of riparian splittail when water releases are made at riverine aquatic, and tidal marsh habitat resources and fluvial function the reach times and locations that coincide with at the mouth of the Delta usable by of the lower Sacramento River between splittail spawning and rearing, and in splittail (Wildlands, Inc. 2002). Red Bluff and Colusa. The Nature such a manner that the releases are In early 2002, our Sacramento River Conservancy, working with the adequate to flood vegetated areas National Wildlife Refuge Complex Sacramento River Conservation Area adjacent to stream channels. The (SNWRC) began implementation of the and local stakeholders, has acquired provisions of section 3406(b)(2) are to be Environmental Assessment for Proposed appreciable amounts of land for implemented for five years and involve Restoration Activities on the restoration. This and other future not only upstream actions but also Sacramento River National Wildlife Sacramento River Conservation Area actions in the Delta which may benefit Refuge. The restoration activities will actions may be beneficial to splittail. splittail. result in the reestablishment or Conclusion: The loss of spawning and Other Habitat Restoration Projects: enhancement of approximately 960 ha rearing habitat remains a potential Ecosystem restoration efforts have been (2,372 ac) of land on 11 units or threat the splittail. However, the undertaken within the splittail’s range. subunits of the SNWRC. Restoration and implementation and magnitude of the USACE began implementation of an enhancement will involve the removal CALFED, and CVPIA programs, and ecosystem restoration project on of crops, orchards, and related other habitat restoration activities, Prospect Island in the northwestern infrastructure (pumping units, barns, which focus on the restoration of

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00023 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55162 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

habitats which directly and indirectly facilities in the south Delta. The south water for the benefit of splittail. In benefit splittail go far beyond any Delta is dominated by water from the addition, we believe that some benefits forseeable habitat losses (particularly in San Joaquin River, where pesticides will accrue from efforts associated with the context of the state’s Environmental (e.g., chlorpyrifos, carbofuran, and these programs (see Factor A above for Quality Act (CEQA) which explicitly diazinon), salts (e.g., sodium sulfates), a discussion on CALFED and the requires mitigation for habitat loss. The trace elements (boron and selenium), CVPIA). overall effect of such habitat restoration and high levels of total dissolved solids We also note that splittail’s habitat, activities is also expected to continue to are prevalent in agricultural runoff (59 the loss of which constitutes the single be beneficial for splittail at present and FR 862, 64 FR 5963). We are unwilling largest threat to the species, is protected into the foreseeable future. to dismiss the potential that disease is by the State under CEQA and by state B. Overutilization for commercial, related to the presence of environmental statutes specific to Delta levees which recreational, scientific, or educational contaminants. Of specific concern are protect levee habitat. Finally, plittail are purposes. We believe that the threats posed by metals, mercury, listed as a Species of Special Concern overutilization (i.e., recreational and selenium, and pesticides. We speculate requiring special considerations for commercial harvest) is not a factor that there is some possibility that mitigation and protection under CEQA. affecting the splittail. As noted in the disease in splittail could be a function To the extent that projects may January 6, 1994, proposed rule (59 FR of increased contaminant loading and sometimes be constructed without 862) and the 1999 final rule (64 FR subsequent immune system depression. proper authorization under section 10 of 5963), some scientific collecting is However, offsetting this concern is the Rivers and Harbors Act and section conducted for splittail, but these information found in the White Paper 404 of the Clean Water Act, this could activities do not adversely affect the (Moyle et al. 2001) indicating that result in threats to the splittail. species. In addition, striped bass anglers disease and parasite infestation may be Implementation of the unpermitted report occasional use of splittail as bait, a natural function related to the heavy projects could have negative effects on but we think this usage has little affect cost of migration and spawning. Post- near-shore splittail habitat similar to on the species. spawn adult splittail, and male fish in those described under Factor A , and In the January 6, 1994, proposed rule, particular, are substantially weakened would not necessarily include and the 1999 final rule, we also noted when outmigrating. We have considered mitigative features. that the small splittail fishery (Daniels whether selenium exposure can In summary, there is a slight potential and Moyle 1983; Caywood 1974) was reasonably be expected to exacerbate that some residual threats still face poorly documented and that no this condition. No research is known to splittail due to of inadequate evidence suggested it was a threat to be conducted on disease occurrence in application or enforcement of RHA and splittail. At present, we do not consider splittail; the only information we found CWA regulatory mechanisms. However, the threat of recreational fishing to be on disease in splittail was in the White we have been unable to document these significant. Baxter (2001b) analyzed Paper (Moyle et al. 2001). Therefore, threats in other than the most nebulous 1999 and 2000 creel census data from given the lack of available information, and anecdotal manner. Notwithstanding the Sacramento River from Garcia Bend we are unable to determine that splittail this potential, as the CALFED program to Redding. Monthly catch amounted to are impacted by disease. is designed to improve habitat for the 103 and 232 splittail, respectively. In the past, we have considered splittail as well as offset any adverse However, no abundance indices were threats of predation to be minor because effects of its own actions and provide calculated by any agency, organization, striped bass had coexisted with splittail for recovery of a number of species or individual from these data, as they for decades and because CDFG had including splittail, we believe it fail to meet the criteria established by forgone hatchery rearing and release of ameliorates the bulk of the minor threats Meng and Moyle (1995) and are striped bass (59 FR 862, 64 FR 5963). associated with this factor. generally considered inadequate to the We have determined that predation may E. Other natural or manmade factors task of quantifying splittail abundance. be a minor factor in the decline of the affecting its continued existence. In our The largest splittail are the first to splittail. Additionally, CALFED past rules and notices concerning the engage in the spawning migration includes numerous studies on the splittail (59 FR 682, 64 FR 5963, 66 FR (Caywood 1974; Moyle et al. 2001). The threats posed by predators (CALFED 2828), we identified the risk of drought, early season fishery thus targets and 2000a, 2000b) (see Factor A for a invasive species (including interference removes females with high reproductive discussion of CALFED). in CVP and SWP salvage operations by potential. The effect of this fishery in D. The inadequacy of existing the introduced Chinese mitten crab the Sacramento River may be relatively regulatory mechanisms. In the past (59 (Eriocheir sinensis)), detrimental flood greater in dry years, when splittail FR 682, 64 FR 5963), we did not bypass operations, the lack of screened spawning is largely confined to river consider the suite of available regulatory water diversions, poor water quality and margins. However, at present, there is mechanisms to be adequate to protect environmental contaminants including no evidence of any trend in the the splittail. Our primary concerns mercury, selenium and pesticides, available data suggesting that larger fish involved the likelihood that the CVPIA, bioaccumulation of selenium in the are being removed from the population the Bay-Delta Accord and CALFED, introduced Asiatic clam (Potamocorbula or that the size structure of the though not regulatory programs, would amurensis) as threatening the splittail. population have been altered by this or be sufficient to control water movement These topics and our current viewpoint other fisheries. in a way that would protect splittail. At of their affect on the splittail are further C. Disease or predation. In our 1994 that time, the funding and discussed below. proposed rule we indicated that this implementation of the Bay-Delta Accord Drought: The variability of factor was not applicable to splittail (59 and CALFED had just begun, and it was California’s Mediterranean climate is FR 862). Since that time, we have too early to know if their funding and not a threat to the species; it represents questioned whether that disease may be implementation would continue. We a baseline condition. This climate, a threat due to high incidences of adult now believe that progress to date however, may exacerbate the effects of splittail in poor health being captured in indicates that these mechanisms are the threats discussed above. Since the the State and Federal water project likely to allow effective management of proposal to list the splittail, California

VerDate jul<14>2003 14:42 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00024 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55163

has had relatively wet hydrologic CALFED includes numerous studies active stream channels, floodplains, conditions that benefit fish species, on the threats of non-native competitors shallow ponds, and borrow areas though water year 2001 was below (CALFED 2000a, 2000b)(see Factor A for (CALFED 2000a, 2000b) (see Factor A normal. Because the splittail is a a discussion of CALFED). for a discussion of CALFED). In floodplain adapted species, a dramatic Detrimental flood bypass operations: addition, the program will conduct decline in abundance was observed It has been documented that splittail instream flow studies to determine the during the 1987 to 1992 drought. make use of the Sutter Bypass, and flows necessary to support all life stages Similarly, abundance peaks during particularly heavy use of the Yolo of anadromous and estuarine fish years when there is extensive floodplain Bypass for spawning under certain species, including splittail (CALFED inundations, and of the Yolo and Sutter hydrologic conditions and that the 2000a, 2000b). bypasses in particular (Sommer et al. shallow, vegetated waters provide Entrainment as a result of water 1997) (see below for a discussion of excellent rearing conditions for juvenile diversions: We conclude that diversion Yolo and Sutter bypasses). When fish (Sommer et al. 1997, 2001a, 2001b). of water from any river or stream or another drought occurs, splittail indices The bypasses are primarily flood control other water course that results in the will again invariably drop. We have facilities and secondarily agricultural entrainment, injury or death of speculated the drought cycle may at lands, and are passively operated as Sacramento splittail, including some point stress the species to such. Splittail using the bypasses are stranding of eggs, larvae, juveniles or extinction if populations are too subject to many of the same threats adults; or diversions and subsequent depressed. However, we have no direct found elsewhere, such as habitat loss, runoff that results in the degradation of evidence this is the case, and in the environmental contamination, harmful waters containing splittail is no longer context of the significant habitat reservoir operations, pesticide loading, a threat to splittail. Entrainment of improvements being undertaken, are far competition with and predation by non- splittail at diversions is reduced if fish less concerned that populations will fall native fish, etc. screens are installed at diversions in to levels that makes this a concern. The flood bypasses are only flooded splittail habitat areas. Two programs when flows in the Sacramento River Invasive species: Chinese mitten crabs implemented under CVPIA, particularly reach a certain level. This inundation (Eriocheir sinensis) could reach the Anadromous Fish Restoration tends to occur at the correct time of year concentrations sufficient to Program (AFRP) and allied Anadromous for splittail spawning, but may be intermittently impede the operation of Fish Screen Program (AFSP), which reduced in frequency and duration were not analyzed in the January 12, fish screens and salvage facilities, thus (Yates 2001), with direct implications 2001, notice (66 FR 2828), have had a reducing the effectiveness of splittail for splittail spawning. This constitutes a net benefit to the splittail. Removal of salvage and repatriation efforts. Since threat in that adult fish, having migrated migration barriers and placement of fish the January 12, 2001, notice (66 FR to suitable spawning habitats on a screens on water diversions is ongoing 2828), USBR has installed a device, floodplain, could be denied the under the AFRP and AFSP, and several known as Crabzilla, to remove the opportunity to spawn. In those cases actions with adjunct benefits to splittail Chinese mitten crab from their CVP fish where adult splittail have successfully have been completed. Removal of salvage facilities. In addition, Chinese spawned, the resulting eggs or larvae migration barriers can provide mitten crabs have not appeared in large could become trapped and killed. additional splittail habitat where numbers at either of the fish salvage Insufficient floodplain inundation could potential habitat is blocked, and facilities in recent years. Therefore, the also force egress of juvenile splittail entrainment of splittail at diversions can Chinese mitten crab does not appear to before they have attained a size and be reduced if fish screens are installed be a current threat to splittail, as they swimming ability sufficient to avoid in splittail habitat areas. Though many have not appeared in large numbers at predation. small diversions remain unscreened, the fish salvage facilities and those that Since the publication of our January approximately 95 percent of water do are efficiently removed and 12, 2001, notice (66 FR 2828), we have annually diverted has been or is in the destroyed before they are able to clog determined, based on consideration of process of being screened, including all the pipes and intakes at the fish salvage scientific data and information provided water diversions greater than 40 cubic facilities. by the public, that the Yolo and Sutter feet per second, and many of the Of some concern is the presence of bypasses are not, in and of themselves, remaining unscreened diversions are Brazilian pondweed (Egeria densa) and a threat to the splittail. A threat is that small and intermittently operated water hyacinth (Eichhornia crassipes), which, if removed, will result in (O’Leary 2003 pers. comm.). CALFED’s both of which tend to form dense near- improvements in a species’ status. The Ecosystem Restoration Program includes shore and slough-wide mats of removal of the Yolo and Sutter bypasses a program to consolidate and screen the vegetation which serves as a retreat, would be highly detrimental to the remaining small agricultural diversions foraging, and ambush site for splittail splittail, as the bypasses constitute a in the Delta, and the Sacramento and predators and which may divert substantial portion of the species San Joaquin rivers. The NOAA Fisheries upstream- and downstream-migrating available spawning habitat. We agree Restoration Center has also begun to splittail into channels rather than the that the bypasses are presently fund small fish screen projects in the more-productive bankside habitat important to the splittail when Sacramento River within the range of (Moyle et al. 2001 in prep). The inundated and that they produce more the splittail. This represents a near-total California Department of Boating and fish than they harm. The bypasses likely reduction in the threat of entrainment in Waterways (CDBW) and the United have helped this resilient species to unscreened diversions to the splittail, States Department of Agriculture persist through over a century of largely and thus removal of the threat. (USDA) Agricultural Research Service unmitigated habitat destruction. Water quality and environmental (ARS) are presently and have been for CALFED’s ERP includes the contaminants: Metals such as copper, at least 10 years, engaged in a program development of a program to eliminate zinc and cadmium (Environmental to control these invasive plant species. fish stranding in the Sacramento, Protection Agency (EPA) 1976) can be To date, the control effort has not had Feather, and Yuba rivers and the Colusa directly toxic to fish, and presumably to a measureable effect on splittail. Basin Drain and Sutter Bypass in the splittail, especially in their sensitive

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00025 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55164 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

larval stages, with the effects including the Yolo Bypass. Recent there in wet years. All major rivers that particularly deleterious near inputs of samples of splittail from Montezuma are tributary to the Estuary are exposed acid mine drainage within the Slough collected by USGS scientists to large volumes of agricultural and Sacramento River watershed and in the (Stewart et al. 2000, Stewart et al. industrial chemicals that are applied in vicinity of highly industrialized near- unpubl. data) have revealed elevated the Central Valley watershed (Nichols et shore areas of the lower San Francisco muscle selenium concentrations ranging al. 1986) as agricultural chemicals and Bay Estuary. These metals damage gills as high as 4 to 5 mg/kg (5 ppm), and their residues, as well as chemicals and alter liver and nervous system liver concentrations ranging as high as originating in urban runoff find their functions causing death, behavioral 20 mg/kg (20 ppm). The relationship way into the rivers and estuary. changes, and reduced growth and between the bioaccumulation of In addition, re-flooding of the Sutter reproduction (EPA 1976). These metals selenium in the Asiatic clam and its and Yolo Bypasses and the use of other can have the same effects on food items predation by splittail could become flooded agricultural lands by splittail for of the splittail, reducing their prey base significant in the near-term future spawning can result in agricultural- and placing additional stress on the because the clam, via its predation on related chemical exposures depending splittail (EPA 1976). However, we are typical splittail prey items such as on the circumstances. not aware of any evidence suggesting estuarine copepods (Eurytemora affinis, Toxicology studies of rice field that splittail are at any higher risk of and Acartia spp.) (Kimmerer and irrigation drain water of the Colusa suffering direct or indirect adverse Penˇ alva 2000), is creating conditions Basin Drainage Canal have documented effects from metals exposure than other that promotes increasing reliance of significant toxicity of drain water to fish species within the Sacramento splittail on the clam as an alternate food striped bass (Morone saxatilis) embryos River and San Francisco Bay estuary source (Feyrer and Matern 2000). Thus, and larvae, Oryzias latipes larvae (in the systems. For all such species, the a potential scenario for the future is Cyprinodontidae family), and opossum potential for at least periodic adverse greater reliance of splittail on Asiatic shrimp, which is the major food impacts from exposure to metals is of clams as a food supply and possibly organism of striped bass larvae and substantive concern, but poorly further increases of selenium juveniles (Bailey et al. 1991), as well as understood. concentrations in both Asiatic clams all age classes of splittail. This drainage Three other potential contaminant and splittail. Selenium threats to canal flows into the Sacramento River threats are of concern specifically with splittail are not confined to the Yolo just north of the City of Sacramento. The respect to the splittail: (1) mercury; (2) Bypass/Suisun Marsh systems. We majority of drain water samples selenium; and (3) pesticides (persistent speculate that when splittail are collected during April and May 1990 organochlorines and currently used exposed to this level of selenium, there were acutely toxic to striped bass larvae organophosphates). In part, these is potential that a reduction in (96 hour exposures); this was the third contaminant threats are of concern reproductive performance will occur, consecutive year rice irrigation drain because they may be focused, to varying which would then result in poor post- water from the Colusa Basin was acutely degrees, on habitat features and hatch survivorship. This means that less toxic (Bailey et al. 1991). Splittail may biological characteristics tentatively splittail young would be able to recruit be similarly affected by agricultural and identified as particularly relevant to to adulthood. There are 1998 splittail industrial chemical runoff, particularly, splittail conservation (Moyle et al. data which confirm that these fish are because like striped bass, adults migrate 2001). being exposed to harmful levels of upriver to spawn and young rear upriver Recent analytical data indicate that selenium in their range along the San until waters recede in late spring. mercury concentrations in aquatic biota Joaquin River. While we have considered these in the San Joaquin River are exceeding Splittail apparently experience contaminants as possible threats to the screening thresholds and may pose substantial post-spawning stress, and splittail, it must also be noted that we ecological and human health risks are subject to substantial stress during have no information on the splittail’s (Davis et al., 2000). A benthic-foraging, salvage operations at the State and thresholds for metals and pesticides. We longer-lived fish such as splittail would Federal pumping facilities. In addition are unwilling to accept the use of a be likely to acquire higher and more to weakening the immune defenses of surrogate species to determine toxic levels of whole body mercury fish and wildlife, excessive acceptable thresholds for splittail. While concentration. We are concerned the environmental selenium can also trigger there are abundant non-native cyprinids combined data from these monitoring pathogen and toxin challenges that available (fathead [Pimephales and research efforts may indicate that would not otherwise have occurred. At promelas] and golden shiners mercury in the San Joaquin River poses this point, we have no direct [Notemigonus crysoleucas]), we assert a threat to ecological health in general, information on the potential effects of the splittail is behaviorally unlike these and the splittail, as a benthic forager, in selenium with respect to splittail. non-native fishes and most likely particular. Some findings have linked However we have considered the physiologically distinct from them as elevated mercury to the Consumnes and selenium-mediated vulnerability to non- well. Further, potential surrogate native Yolo Bypass (Slotten et al. 2000), which chemical stressors when assessing the cyprinids (hardhead [Mylopharodon are both primary spawning areas for threats presented by exposure of conocephalus], blackfish [Orthodon splittail (Moyle et al. 2001). splittail to selenium. microlepidotus], pikeminnow Furthermore, the Yolo Bypass may be Several of the pesticides present in [Ptychocheilus grandis]) are piscivorous hydrologically connected to Suisun the rivers of the Central Valley have (fish-eating) when adults, and therefore Marsh, the likely core rearing area for been documented to have adverse likely distinct from splittail. Splittail splittail (Moyle et al. 2001). Suchanek et effects on animal life. However, we have may have its closest relative in the al. (2000) is investigating the role of no direct evidence that pesticides are a Rhinichthys complex (speckled dace wetland restoration involving re- pervasive threat to the splittail [Rhinichthys osculus] and others) but flooding of mercury-contaminated soils. throughout its range. If there is a threat use of these diminutive, short-lived, Significant exposure to selenium it may be relatively greater in the small-stream species would be similarly could potentially pose a threat to bypasses due to the large amount of unadvisable. Lastly, we would have splittail throughout much of its range, spawning and early rearing that occurs serious concerns with results obtained

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00026 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations 55165

from non-cyprinids surrogate species, as copper, cadmium, and zinc, through available regarding the abundance and such as white sturgeon, bluegill, inland source control at inactive and distribution of; and the past, present, silverside, mosquito fish, and lake trout, abandoned mine sites, urban storm and future threats faced by the splittail as they would certainly be both water programs and agricultural BMPs; in this listing determination. The physiologically and behaviorally (4) reduce mercury levels in rivers and following narrative will summarize the distinct from splittail and therefore the estuary by source control at inactive pertinent data regarding abundance and useless in determining thresholds for and abandoned mine sites; (5) reduce threats. the splittail. We therefore have selenium impacts through reduction of Based upon our statistical analysis determined that the above mentioned loads at their sources and through using a relaxed standard for thresholds for other fish species are not appropriate land fallowing and land significance, we conclude that splittail indicative of the thresholds of the retirement programs; (6) reduce salt populations may have declined over the splittail. For all fish species, the sources in urban and industrial period of analysis. We recognize that potential for at least periodic adverse wastewater and facilitate development other agencies, including USBR and impacts from exposure to metals and of successful water recycling, source CDFG, believe that the available data do pesticides is of potentially substantive water blending, and groundwater not indicate a population decline. concern, but poorly understood and storage programs; (7) manage Delta However, the magnitude, certainty, and poorly documented. Thus we have no salinity by limiting salt loadings from its ecological significance of the apparent real basis for concluding that these tributaries and through managing population decline remain unclear. substances represent a particular threat seawater intrusion by such means as We believe that above all else, the to the splittail. using storage capacity to maintain Delta primary threat to splittail is the loss of Finally, Moyle et al. (2001) outflow and adjust timing of outflow, spawning and rearing habitat. Past hypothesize that success of juvenile and by export management; (8) reduce habitat losses are offset by the downstream migration is strongly linked turbidity and sedimentation; (9) reduce implementation programs of CALFED to the size that juvenile splittail achieve the impairment of rivers and the estuary and the CVPIA which are restoring prior to exiting the spawning areas. It from substances that exert excessive significant amounts of habitat was suggested that a minimum size of demand on dissolved oxygen; and, (10) previously lost. In addition, those 25 mm (1 in) greatly enhances success through research and monitoring, to programs ensure that future water of downstream migration. Moyle et al. identify parameters of concern in the operations and development will (2001) have already presented data water and sediment and impairment protect and improve existing habitats. demonstrating statistically-significant actions, to reduce their impacts to The many additional ongoing and future declining growth rates in Suisun Marsh aquatic resources. habitat restoration projects throughout splittail between 1980 and 1995. The Conclusion: Splittail are no longer the range of the splittail include, either apparent declines in growth rate appear threatened by interference in CVP and as direct or indirect effects, spawning to correlate to the invasion of the SWP salvage operations by the and rearing habitat for the splittail, or estuary by the Asiatic clam, and the introduced Chinese mitten crab and enhancement of such habitat. The subsequent shift of splittail to an Asiatic unscreened diversions. The Yolo and restoration of splittail habitat enables clam-dominated diet. Moyle et al. Sutter Bypasses are a net benefit to the greater spawning and rearing (2001) suggested that this trend might splittail. CALFED’s Ecosystem opportunities and thus increases the reflect poorer energetics of a non-mysid Restoration Program (discussed in population size, ameliorating all of the shrimp-dominated diet, but it can just as Factor A above) will conduct instream remaining threats to a level below the plausibly be suggested that it reflects the flow studies to determine the flows point at which the splittail would meet cachexia (contaminant-induced weight necessary to support all life stages of the definition of a threatened species. loss despite calorically sufficient dietary anadromous and estuarine fish species, We therefore have determined that the intake) that is a classic symptom of non- including splittail, which will offset the splittail is not in danger of extinction lethal selenium poisoning. However we threat of drought and flow regime through all or a significant portion of its have no particular basis for finding the changes resulting from water project range either now or in the foreseeable growth rates are the result of any operations. The threats of poor water future. It therefore does not meet the contaminent induced mechanism. quality from contaminants including definition of an endangered or CALFED’s Water Quality Program, mercury, selenium and pesticides, and threatened species. As a result, we have which was not analyzed in the January bioaccumulation of selenium in the determined that listing the splittail as 12, 2001, notice (66 FR 2828), will have introduced Asiatic clam, appear to be endangered or threatened under the Act a net benefit for the splittail when reduced by CALFED’s Water Quality is not warranted. implemented (see Factor A for a Program (discussed in Factor E above). In making this finding, we recognize discussion of CALFED). The Water At present, although environmental that the Sacramento splittail may be Quality Program includes the following contaminants are pervasive throughout experiencing a decline in population actions: (1) Reduce the impacts of the range of the splittail, and many size based upon our conservative pesticides through development and contaminants have the potential to pose statistical analysis, and that the species implementation of Best Management a significant threat to splittail, there is continues to face potential threats from Practices (BMPs) for both urban and insufficient scientific evidence at this habitat loss. We also recognize that the agricultural uses, through support of time to indicate that environmental full implementation of CALFED and the pesticide studies for regulatory agencies, contaminants impair splittail growth CVPIA restoration programs are not 100 and through providing education and and reproduction at all; much less to a percent certain. Finally, we recognize assistance in implementation of control magnitude that would warrant listing other threats to the species, its habitat, strategies for the regulated pesticide splittail due to that threat alone or in and its prey exist, including effects of users; (2) reduce the load of combination with others. drought and climate change on habitat; organochlorine pesticides in the system non-native competitors and predators; by reducing runoff and erosion from Finding and possible threats of disease and agricultural lands through BMPs; (3) We have carefully assessed the best environmental contaminants. We will reduce the impacts of trace metals, such scientific and commercial information continue to monitor the status and

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00027 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2 55166 Federal Register / Vol. 68, No. 183 / Monday, September 22, 2003 / Rules and Regulations

management of the species. We will under both section 3(20) of the Act and Endangered Species Act. We published continue to accept additional CESA (Fish and Game Code section a notice outlining our reason for this information and comments from all 2067). determination in the Federal Register concerned governmental agencies, the CDFG submitted comments regarding on October 25, 1983 (48 FR 49244). scientific community, industry, or any the status of the splittail during the Paperwork Reduction Act other interested party concerning this January 12, 2001, May 8, 2001, and finding. If we find that circumstances August 17, 2001, comment periods (66 This rule does not contain any change to the point that any of these FR 2828, 66 FR 23181, and 66 FR 43145, information collection requirements for threats change significantly, we will respectively) subsequent to the court’s which OMB approval under the reexamine the status of the splittail. June 23, 2000, summary judgement. Paperwork Reduction Act is required. Further, CDFG staff were involved in an An agency may not conduct or sponsor, Coordination With the State of interagency peer review effort California and a person is not required to respond undertaken concurrent with the August to, a collection of information unless it The State of California administers, 17, 2001, comment period. CDFG displays a valid OMB control number. via CDFG, the California Endangered comments were limited only to alternate For additional information concerning Species Act (CESA) (Fish and Game analyses of species abundance (see the permits and associated requirements for Code sections 2050 to 2116, et seq.). The Summary of Comments and threatened wildlife species, see 50 CFR purposes of the CESA are to conserve, Recommendations section). 17.21 and 17.22. protect, restore, and enhance any bird, We are actively coordinating with mammal, fish, amphibian, reptile, or California Environmental Protection References Cited plant meeting CESA criteria for Agency (CalEPA), the State Water A complete list of all references cited threatened or endangered status, and to Resources Control Board (SWRCB), and in this final rule is available upon acquire lands for habitat for these the Regional Water Quality Control request from the Sacramento Fish and species. Boards (RWQCBs) through public Wildlife Office (see ADDRESSES). Procedures governing the submission comment periods on their regulatory and review of petitions for listing, program actions (USFWS 2002b). The Authors uplisting, downlisting, and delisting of CalEPA, SWRCB, and OEHHA provided The primary authors of this document CESA endangered and CESA threatened no comments regarding the listing, are staff of the Sacramento Fish and species of plants and animals are however. The CDWR and the Wildlife Office. described in section 670.1, Title 14, Reclamation Board did comment to a California Code of Regulations. certain degree regarding the factors List of Subjects in 50 CFR Part 17 Under CESA, a State ‘‘threatened’’ affecting the splittail (see the Summary species is a California native species of Comments and Recommendations Endangered and threatened species, that, although not presently threatened section). Exports, Imports, Reporting and record- with extinction, is likely to become an We have given full consideration to keeping requirements, Transportation. endangered species in the foreseeable CDFG as well as CDWR Regulation Promulgation future in the absence of special recommendations to employ an protection and management efforts (Fish alternate abundance analysis (see ■ For the reasons given in the preamble, and Game Code section 2067). A State Abundance and our response to we amend part 17, subchapter B of ‘‘endangered’’ species is that which is in Comment 1). Indeed, we used the chapter I, title 50 of the Code of Federal serious danger of becoming extinct CDFG/USBR MRF model, the result of a Regulations, as set forth below: throughout all, or a significant portion, joint State and Federal scientific of its range due to one or more causes, undertaking, to determine if a trend PART 17—(AMENDED) including loss of habitat, change in exists for the species. Based on our ■ 1. The authority citation for part 17 habitat, overexploitation, predation, evaluation of conservation efforts continues to read as follows: competition, or disease (Fish and Game completed, currently underway, and Code section 2062). The splittail is not likely to stem from CALFED and the Authority: 16 U.S.C. 1361–1407; 16 U.S.C. listed as threatened or endangered by CVPIA, we now agree with the State that 1531–1544; 16 U.S.C. 4201–4245; Pub. L. 99– the State of California under the listing of the splittail as a threatened 625, 100 Stat. 3500; unless otherwise noted. authority of CESA. There appears to be species is not warranted at this time. ■ 2. Amend § 17.11(h) by removing the substantive similarity between the National Environmental Policy Act entry ‘‘Sacramento splittail’’ under Federal requirement under section ‘‘FISHES’’ from the List of Endangered 4(a)(1) of the Act and the State We have determined that we do not and Threatened Wildlife and Plants. requirement under section 14(i)(1)(A) of need to prepare an Environmental the of the California Code of Regulations Assessment and/or an Environmental Dated: September 15, 2003. to consider all factors affecting a Impact Statement as defined by the Marshall P. Jones, Jr., species. There also appears to be a high National Environmental Policy Act of Acting Director, Fish and Wildlife Service. degree of similarity between the 1969 in connection with regulations [FR Doc. 03–23919 Filed 9–18–03; 12:01 pm] definition of a ‘‘threatened species’’ adopted pursuant to section 4(a) of the BILLING CODE 4310–55–P

VerDate jul<14>2003 14:16 Sep 19, 2003 Jkt 200001 PO 00000 Frm 00028 Fmt 4701 Sfmt 4700 E:\FR\FM\22SER2.SGM 22SER2