Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
- s Fishes of the Upper I(lamath Basin The upper I(lamath basin is an ancient, isolatecT, ancT unusual environ- ment for fish. Most, or possibly all, of the native species that live in the upper basin are endemic to it. The distinctiveness of the upper basin ancT its fishes has been recognized since the first ichthyologists explorecT it in the late lath century (Cope 1879, Gilbert 1898), but the application of new kincTs of genetic analysis to the fishes is revealing even more diversity ancT complexity than was previously known (e.g., Docker et al. 1999, Tranah 2001 ). Since the shortnose ancT Lost River suckers were listecT as enciangerecT species in 1988, a great clear of attention has been paicT to their biology, especially in Upper I(lamath Lake, whereas the rest of the species ancT the rest of the basin have received comparatively little attention. The other endemic fishes, some of which may be consiclerecT for listing in the future, interact with the enciangerecT suckers ancT thus complicate management practices intenclecT to benefit them. In acicTition, nonnative fishes, which are abundant in the basin, affect the enciangerecT suckers. Overall, the upper basin's lancT ancT water shouicT be managed through an ecosystem-basecT approach with all native fishes in mincT uncler the assumption that manage- ment favoring native fishes is likely to have positive effects on other ecosys- tem components. Failure to clo this is likely to result in listing of acicTitional species as threatened or enciangerecT. The purposes of this chapter are to describe the factors that lecT to the high endemism of fishes in the upper I(lamath basin ancT to its invasion by nonnative fishes, to give a brief summary of the biology ancT welfare of each of the native fishes with special 179
180 FISHES IN THE KLAMATH RIVER BASIN attention to the listecI suckers, an overview of the nonnative fishes, ancI to identify gaps in knowlecige about all the fishes. NATIVE FISHES The fishes of the upper I(lamath basin originated when a large river draining the western interior of North America flowecI through the I(la- math region on its way to the ocean (Minckley et al. 1986, Moyle 20021. Uplift ancI erosion have since caused the water in the region to flow at different times into the Great Basin to the east, into the Columbia River via the Snake River to the northwest, into the Sacramento River via the Pit River to the south, ancI into the lower I(lamath River to the west. As the connections to large drainage basins shifted back ancI forth, fishes from each of the basins entered the upper I(lamath basin (Minckley et al. 1986, Moyle 20021. Species that persisted through periods of change, which incluclecI drought ancI volcanism, evolvecI into the endemic fauna of the upper I(lamath basin (Table 5-11. These fishes are aciaptecI to the shallow lakes, meandering rivers, ancI climatic extremes of the upper I(lamath basin. The closest relatives of moclern fishes of the upper I(lamath basin are now found in the Great Basin, Columbia River, Pit River, ancI lower I(lamath River. The present connection of the upper I(lamath basin to the lower I(la- math basin probably is fairly recent (Pleistocene, less than 1.8 million years BP), but the connection formecI ancI was blockecI more than once. Connec- tion of the upper ancI lower basins lecI to colonization of the upper basin by anaciromous Chinook salmon, steelheacI, ancI Pacific lampreys. Repeated isolation of anaciromous fishes, which occurred when the connection be- tween the two parts of the I(lamath basin was broken, left behind resident populations that now differ from parent stocks (such as recibancI trout ancI I(lamath River lamprey). The lower basin contains mainly fast-flowing, cool-water rivers ancI streams that are icleally suited for anaciromous fishes but inhospitable to fishes of the upper basin, which are aciaptecI to lakes or warmer streams ancI rivers of lower gradient. Thus, the two basins have remarkably different fishes. The absence of major physical barriers to movement of fish before installation of clams explains the former use of the upper basin by anaciro- mous fishes ancI the apparent occasional entry into the upper basin of the I(lamath smaliscale sucker, which is abundant in the lower basin. Only five families of fishes Petromyzonticiae, Cypriniciae, Catosto- miciae, Salmoniciae, ancI Cotticiae are native to the upper basin, ancI the species in these families have many unusual adaptations to the environment of the basin. The lampreys ancI suckers of the upper basin show some interbreeding (hybriclization) among species.
FISHES OF THE UPPER KLAMATH BASIN TABLE 5-1 Native Fishes of the Upper I(lamath Basin 181 Adult Species Habitata Statusb Comments Pacific lamprey, Lampetra tridentata Klamath River lamprey, L. similis Miller Lake lamprey, L. miller) Pit-Klamath brook lamprey, L. Iethophaga R. L R C R. L, W U W. C C? C? Same species as in lower river but land-locked and probably distinct Also in lower river Once thought extinct Shared with Pit River Klamath tui chub, L, R. W A Abundant and widespread Siphatales bicolor bicolor Blue chub, Gila coerulea R. W C Special concern species in California Klamath speckled dace, W. C, R. L C? May be more than one Rhinichthys osculus klamathensis form Shortnose sucker, Chasmistes brevirostris Lost River sucker, Deltistes luxatus Klamath largescale sucker, Catostomus snyderi Klamath smallscale sucker, C. rimiculus Klamath redband trout, Oncorhynchus mykiss subsp. L, R L R.L L R. L, W C? Listed as endangered Listed as endangered May be more than one form; declining? R. W. C R R. L C? Coastal steelhead, O. mykiss irideus R. C E Chinook salmon, O. tshawytscha R E 13ull trout, Salvelinus confluentus C L Upper Klamath marbled sculpin, C, W. R C Cottus klamathensis klamathensis Klamath Lake sculpin, Cottus princeps Slender sculpin, Cottus tenuis L. R R Common in lower basin Fishery; may be more than one form: lake and stream Anadromous, common in lower basin Anadromous, common in lower basin Threatened species Widespread in basin L, R A Abundant in Upper Klamath Lake Gone from much of former range aAdult habitat: L, lakes; R. river; W. warm-water creeks; C, cold-water creeks. bStatus in upper basin: A, abundant; C, common; E, extirpated; L, listed under federal Endan- gered Species Act; R. rare; U. unknown. Petromyzonticiae: Lampreys The lampreys of the upper I(lamath basin are all clerivecI from anaciro- mous Pacific lampreys that became lancI-lockecI, perhaps multiple times
182 FISHES IN THE KLAMATH RIVER BASIN over millions of years. Their evolution ancI ecology are poorly unclerstoocI. Four species are recognized (Docker et al. 1999, Lorion et al. 2000, Moyle 2002), but aciclitional species may be uncovered as genetic studies proceed. There are two basic life cycles among lampreys: one with preciatory aclults ancI one with nonpredatory aclults. Both types spencI the first 3-7 yr of their lives living in mucI ancI sancI as eyeless, wormlike larvae (ammocoetes) that feecI on algae ancI organic matter. The ammocoetes metamorphose into silvery, eyecI aclults. Aclults of the preciatory forms attach to other fish with their sucking-clisc mouths, through which they remove bloocI ancI body fluicis. Typically the prey survives the attack of a preciatory lamprey, but the attack may impair growth ancI survival (Moyle 20021. Preciatory lampreys engage for about a year in this feeding behavior, which enables them to grow to produce a larger number of gametes than clo nonpredatory lam- preys. The aclults of the nonpredatory form clo not feecI; they live only long enough to reproduce (Moyle 20021. The Pacific lamprey is regarclecI as a lancI-lockecI version of the precia- tory anaciromous species, but the form native to the upper I(lamath basin probably shouicI be a separate taxon. Nothing is known about its ecological differences from the slightly smaller (14-27 cm) I(lamath River lamprey. The I(lamath River lamprey is a nonmigratory preciatory species that is wiclespreacI in the upper ancI lower basins. Little is known about its biology except that it preys on native suckers ancI cyprinicis, especially in reservoirs (Moyle 20021. The Miller Lake lamprey is the smallest (less than 15 cm) preciatory lamprey known anywhere in the woricI; it occurs mainly in the Sycan ancI Williamson rivers, where resident prey species are abundant (Lorion et al. 20001. The Miller Lake lamprey is closely relatecI to the nonpredatory Pit-I(lamath brook lamprey, which is abundant ancI wicle- spreacI in small streams in the upper I(lamath ancI Pit River basins. Because of the long (about 1 million years) separation of the Pit ancI I(lamath basins, genetic studies will probably show that the two populations belong in different taxa. The exact distribution of the four species in the watershed is not known, because most collections are of ammocoetes, which are clifficult to identify in the fielcI. Cypriniciae: Minnows The I(lamath tui chub is wiclespreacI in the interior basins of the west- ern United States anti is cliviclecI into a number of subspecies (Moyle 20021. Some, inclucling the I(lamath tui chub, may eventually be recognized at the species level. Tui chubs are chunky, omnivorous minnows that can become large (about 30 cm) ancI have high longevity (20-35 yr), especially in large lakes. In the I(lamath basin, they are the most abundant ancI wiclely clistrib-
FISHES OF THE UPPER KLAMATH BASIN 183 utecI native fish. They occur in streams, rivers, reservoirs, anti lakes (Simon anti Markle 1997a,b; Buettner anti Scoppettone 1991) anti in a wicle array of habitats (BoncI et al. 19881. They are tolerant of high temperature (over 30°C), low clissolvecI oxygen (below 1 mg/L), anti high pH (10-11; Falter anti Cech 1991, Castleberry anti Cech 1992, Moyle 20021. Despite those tolerances, they typically are among the most abundant species in fish kills of Upper I(lamath Lake (Perkins et al. 2000b), although counts of cleacI chubs usually clo not distinguish tui chubs from blue chubs. In the last 30 years, the tui chub has cleclinecI in abundance in the Lost River, where it has changed from a dominant member to a minor component of the fish fauna (Shivery et al. 2000a). In contrast with tui chub, the blue chub is confined largely to the I(lamath basin anti a few adjacent basins into which it may have been introclucecI (Moyle 20021. It is especially abundant in lakes, reservoirs, anti other warm, still habitats (BoncI et al. 1988, Buettner anti Scoppettone 19911. It may be the most abundant native fish in Upper I(lamath Lake, although it may also be in clecline there, along with most other native fishes (Simon anti Markle 1997b, Moyle 20021. It clearly is in clecline elsewhere in the upper I(lamath basin. For example, Contreras (1973) founcI that the blue chub was the most abundant species in the upper part of the Lost River but that the tui chub was the most abundant in the lower half of the river. More recent sampling indicates that both species have been largely replacecI by fathead minnows, brown buliheacis, anti other nonnative species that tolerate poor water quality (Shivery et al. 2000a). Not much is known about the biology of the blue chub except that it is omnivorous, schools, anti reaches a length of about 25 cm. It is somewhat less tolerant of high temperatures anti low clissolvecI oxygen than the tui chub (Castleberry anti Cech 1992) anti is common in fish kills of Upper I(lamath Lake (Perkins et al. 2000b). The specklecI ciace is even more wiclespreacI than the tui chub in western North America anti probably consists of a complex of species (Moyle 20021. Dace from both the upper anti lower I(lamath basins are recognized as just one subspecies, but the two forms probably are distinct, anti the upper basin probably supports more than one taxon (M. E. Pfrencler, Utah State University, personal communication, 20021. The specklecI ciace is common in the upper basin but is most abundant in coo! streams associated with rocks anti grave! (Buettner anti Scoppettone 1991, BoncI et al. 19881. Even so, Castleberry anti Cech (1992) founcI that it couicI withstand high tem- peratures (28-34°C) anti low concentrations of clissolvecI oxygen (1-3 ma/ L). The status of the specklecI ciace in the basin is not known, because collections are biased toward the larger fishes. It apparently has become very uncommon in the Lost River, however (Shivery et al. 2000b).
184 FISHES IN THE KLAMATH RIVER BASIN Catostomiciae: Suckers The four species of suckers in the I(lamath basin (Table 5-1) have an interesting and long evolutionary history (Moyle 2002) and probably once were the most abundant fishes, in terms of biomass, in the lakes and large rivers. The listed shortnose sucker and Lost River sucker, which have been the focus of most fish studies in the upper basin, will be treated in the last part of this chapter. The I(lamath smaliscale sucker is rare (or perhaps absent since the construction of Copco Dam) in the upper basin, although it is found in upper lenny Creek, a tributary to Copco Reservoir. It is abun- dant in the lower basin (see Chapter 71. The I(lamath largescale sucker is resident in the upper basin. All four species show some evidence of hybrid- ization with each other (Tranah 20011. The I(lamath largescale sucker, which becomes large (about 50 cm) and has a long lifespan (31 yr or more), as do the shortnose and Lost River suckers, is one of the least understood fish in the basin (Moyle 20021. It appears to be mainly a resident of large rivers, although a small population exists in Upper I(lamath Lake, and it is rare or absent in the Lost River (I(och et al. 1975, Buettner and Scoppettone 1991, Shively et al. 2000a). It apparently is common and widely distributed in the Williamson, Sprague, and Wood rivers (Reiser et al. 20011. In Upper I(lamath Lake, the I(lamath largescale sucker is found mainly near inflowing streams; this suggests a low tolerance for lake conditions, but it has been found at temperatures near 32°C in environments of dissolved oxygen at 1 mg/L and pH over 10 (Moyle 20021. Lake populations of largescale suckers migrate for spawning in March and April; peak spawning activity occurs a month or so earlier than that of shortnose and Lost River suckers. Radio-tagged fish have migrated as far as 128 km upstream, presumably to find grave! for spawn- ~ 1 ~ A A ~ ~ ~1 _ ~ 1 1 1 1 1 1 1 · 1 · · 1 1 ~ 7 ~ J C7 ~ng (Ke~ser et al. ZUU1). lhe l~lamath largescale sucker hybr~d~zes w~th the shortnose and Lost River suckers. Genetic studies by Tranah (2001 ~ suggest that the largescale suckers in the Sprague River belong to a different taxon from other largescale suckers in the basin. The status of the I(lamath largescale sucker is poorly understood. The lake populations probably are similar to those of the shortnose and Lost River suckers in having declined in abundance. The status of stream popu- lations is not known, although they are assumed to be widespread and abundant (Reiser et al. 20011. Salmoniciae: Salmon anc! Trout The bull trout is a predatory char that is widely distributed in the northwestern United States but is considered a relict species in the I(lamath basin. It apparently entered the I(lamath basin when it was connected to the
FISHES OF THE UPPER KLAMATH BASIN 185 Snake River but then became isolatecI. Genetic evidence reflects isolation ancI suggests that the bull trout of the upper I(lamath basin couicI be as- signecI to a distinct taxon or evolutionarily significant unit (Ratliff ancI Howell 19921. The bull trout is known from only 10 creeks in the upper I(lamath basin four tributaries to the Sprague River, four to the Sycan River, ancI two to Upper I(lamath Lake (Ratliff ancI Howell 1992, Buchanan et al. 1997) although it has been extirpated or is at risk of extirpation in most of these creeks. An important characteristic of streams containing bull trout is high water quality; temperatures clo not exceed 18°C in these streams (Moyle 20021. The bull trout tencis to disappear from streams with clegraclecI water quality even if the streams can support other kincis of trout. The bull trout also cleclines when the brook trout invades its habitat. Hy- briclization between the bull trout ancI the brook trout has taken place in some I(lamath basin streams (Markle 19921. Threats to the existence of the bull trout are not peculiar to the I(lamath basin; they occur throughout its range. Thus, the bull trout of the upper I(lamath basin was listecI by the U.S. Fish ancI WilcIlife Service (USFWS) in 1998 as threatened. The bull trout, like the enciangerecI suckers of the upper basin, clemancis special attention in the future. Unlike the suckers, however, the bull trout is spatially separated from the I(lamath Project ancI most other water man- agement because its distribution is restricted primarily to heac~waters that are remote from Upper I(lamath Lake or the lower reaches of tributaries that are so important to suckers. At present a goocI clear of attention is being given to the welfare of bull trout, but much work remains to be clone. The recibancI trout is a resident rainbow trout whose ancestors en- terecI the upper I(lamath basin when it was connected to the Columbia Basin via the Snake River (Behnke 19921. Coastal rainbow trout (steel- heacI) later entered the upper basin, but the recibancI trout clerivecI from the Columbia Basin maintained its identity ancI is recognizable by its morphology ancI color. Behnke (1992) indicates that there are two types of recibancI trout in the basin: a small form resident in isolatecI streams ancI the form present in Upper I(lamath Lake; he suggests that the lake form is so distinctive (for example, it has large numbers of gill rakers, an adaptation to life in lakes) that it deserves subspecies designation (as 0. m. newberrii). The Oregon Department of Fish ancI WilcIlife (ODFW), however, regards all recibancI trout in the interior basins of Oregon as belonging to one taxon, even though it states that the I(lamath Lake recibancI trout is "unique in terms of life history, meristics, disease resis- tance, ancI allozyme variation" (Bowers et al. 19991. The various stream populations in the basin also show genetic evidence of isolation from one another (Reiser et al. 20011. RegarcIless of taxonomic position, these fish have persisted because of their ability to thrive in lake ancI stream concli- tions that wouicI be lethal to most salmonicis.
186 FISHES IN THE KLAMATH RIVER BASIN Behnke (1992) wrote of observations he macle on Upper I(lamath Lake in September 1990 (p. 1811: "In clear-water sections influencecI by spring flows, huncirecis of large, robust trout from about 1 to 5 kg couicI be reaclily observed. In shallow (2 m) Pelican Bay, in the midst of a bloom Fof bluegreen algae] (I estimated a Secchi clisk clarity of about 40 cm), I caught a magni- ficent trout of 640 mm ancI 2.3 kg." This is consistent with continuing reports of a strong summer fishery for trout, especially in Pelican Bay (e.g., HogluncI 20031. Water temperatures in Upper I(lamath Lake in summer are 20-25°C, occasionally spiking to 27°C, ancI clissolvecI oxygen may cirop below 4 mg/L for several clays (Perkins et al. 2000b). Springs ancI the mouths of streams in Pelican Bay, which apparently have higher water quality than the lake, may serve as refuges for the trout, especially cluring episodes of very poor water quality in the lake. Trout have been reported in the lake's summer fish kills, but the only example of mass mortality was in 1997, when about 100 large trout were founcI cleacI (Perkins et al. 2000b). The lake population of recibancI trout is acifluvial; it migrates up into the WoocI, Williamson, ancI Sprague rivers for spawning cluring spring. The rivers also support resident populations of these trout, as cloes the river below Upper I(lamath Lake, mostly above Boyle Dam (Bowers et al.19991. IsolatecI populations, which are genetically distinct from the I(lamath Lake ancI river populations, exist in the upper Williamson ancI Sprague rivers ancI in lenny Creek, which flows into Iron Gate Reservoir (Bowers et al. 19991. Hatchery rainbow trout (coastal stock) in the past have been stocked in I(lamath basin streams, ancI some interbreeding with native recibancI trout was noted. Stocking now is limitecI to Spring Creek, which flows into the lower Williamson River. The hatchery fish apparently have poor survival because they are not resistant to endemic disease ancI are not aciaptecI to high pH (Bowers et al. 19991. RecibancI trout are cloing surprisingly well in the I(lamath basin, con- siclering all the changes that have taken place. The fishery for the lake ancI river populations is an important recreational resource. The populations of small streams are vulnerable, however, to habitat clegraciation by roacis, grazing, ancI other activities. The lake ancI river populations will neecI pro- tection from adverse water quality ancI nonnative species ancI probably wouicI benefit from improved habitat in the rivers ancI improved access to upstream habitat (Bowers et al. 19991. Cotticlae: Sculpins The scuipins are a poorly stucliecI group in the I(lamath basin despite the presence of at least three endemic species (I(lamath Lake scuipin, slen- cler scuipin, ancI Upper I(lamath marblecI scuipin). There may be aciclitional taxa in the watershed as well (Bentivoglio 19981.
FISHES OF THE UPPER KLAMATH BASIN 187 The I(lamath Lake scuipin apparently is the most abundant scuipin in Upper I(lamath Lake. It is caught in large numbers in the lake with bottom trawls (D. Markle, Oregon State University, personal communication, 2001) ancI in smaller numbers with beach seines ancI trap nets (Simon ancI Markle 1997b). The abundance of this scuipin is estimated to be in the millions (Simon et al. 19961. It is present only in Upper I(lamath ancI Agency lakes ancI in springs ancI creeks that flow into the west sicle of Upper I(lamath Lake (Bentivoglio 19981. The present distribution coincides with the known historical distribution of the species. Little is known about its environmen- tal requirements, but it lives mainly in offshore areas with bottoms of sancI ancI silt ancI appears to be able to withstand wiclely varied lake conditions. No I(lamath Lake scuipins have been reported in the fish kills of Upper I(lamath Lake, but cleacI fish of this species wouicI not float ancI so wouicI be easy to overlook. The apparent ability of the I(lamath Lake scuipin to live in conditions of poor water quality (especially low clissolvecI oxygen) is similar to that of prickly scuipin (Cottus asper) in Clear Lake of central California which, like Upper I(lamath Lake, is subject to massive blooms of cyanobacteria (Moyle 20021. The slencler scuipin apparently once was common in the Williamson, Sprague, Sycan, ancI Lost rivers ancI in Upper I(lamath Lake (Bentivoglio 19981. Bentivoglio (1998) collectecI scuipins throughout the upper basin in 1995-1996, however, ancI found slencler scuipins only in the lower William- son River ancI a few in Upper I(lamath Lake. Simon ancI Markle (1997b) also recorclecI small numbers in Upper I(lamath Lake. Little is known about the ecology of this fish, although it seems to require coarse substrates ancI high water quality; it is especially characteristic of coicI springs. Its closest relative is the rough scuipin (C. asperimmus) of the Fall River in California (Robins ancI Miller 1957), which requires coicI, spring-fecI streams (Moyle 20021. It is fairly long-livecI for a scuipin (7 yr) but is small (rarely longer than 75 mm; Bentivoglio 19981. Overall, the slencler scuipin appears to have clisappearecI from much of its native range ancI is uncommon in most areas where it is found today. The Upper I(lamath marblecI scuipin is the most wiclely clistributecI scuipin in the I(lamath basin (A. Bentivoglio, USFWS, personal communi- cation, 20021. It is founcI in most streams ancI rivers in the basin in a wicle range of conditions, inclucling summer temperatures over 20°C (BoncI et al. 19881. It is most abundant among coarse substrates in the larger streams where water velocities are moderate to low (BoncI et al. 19881. In the Lost River basin, it is known mainly from riffles in Willow ancI Boles Creeks (I(och et al. 1975) but has become scarce in recent years (Shivery et al. 2000a). It is largely absent from the reservoirs in the basin, at least in California Data in Buettner ancI Scoppettone 1991), but is fairly common in Upper I(lamath Lake (Simon et al. 1996, Simon ancI Markle 1997b). It
188 FISHES IN THE KLAMATH RIVER BASIN occurs mostly on soft bottoms in the lake ancI apparently enters the water column to feecI at night (Markle et al.19961. It has been recorclecI in at least one of the fish kills of Upper I(lamath Lake (Perkins et al. 2000b). The marblecI scuipin, like most stream scuipins, generally hicles uncler or among rocks, where it feecis on benthic invertebrates (Moyle 20021. Females glue their eggs to the bottoms of rocks ancI logs where cleveloping embryos are tenclecI by males until they hatch. The larvae are benthic ancI clo not move far from their natal site. They become mature in their second summer ancI live 4-5 yr (Moyle 20021. The cletails of their ecology ancI life history in the upper I(lamath basin have not been clescribecI. NONNATIVE FISHES In the last century, the upper I(lamath basin has been invaclecI by 17 nonnative species (Table 5-2), 15 of which were introclucecI for sport fish- ing or for bait. Most of the 17 are not particularly common in the basin, but some are abunciant ancI wiclespreacI (or are spreacling), ancI their effects on native fishes are poorly unclerstoocI. One of the most recent invaclers is the fatheacI minnow, which is now one of the most abunciant fishes in Upper I(lamath ancI Agency lakes (Simon ancI Markle 1997a). The Sacramento perch, which was introclucecI into Clear Lake in the 1960s, has the potential to become very abunciant in other lakes of the basin (Moyle 20021. Other introclucecI species especially yellow perch, brown buliheacI, ancI pump- kinseecI are locally abunciant, especially in reservoirs ancI sloughs or poncis (Buettner ancI Scoppettone 1991, Simon ancI Markle 1997b). Brook trout, brown trout, ancI nonnative strains of rainbow trout are common in coicI- water streams ancI have replacecI native recibancI trout ancI bull trout in many areas. One concern is that future changes in water quality in the basin may promote further expansion of nonnative species. The fatheacI minnow, which is native to eastern North America, ap- pearecI in the I(lamath basin in the early 1970s, perhaps as a result of release of fish usecI in bioassay work (Simon ancI Markle 1997a). By 1983, it was common in Upper I(lamath Lake ancI by the early 1990s it hacI spreacI to the Lost River system (Simon ancI Markle 1997a, Shively et al. 2000a). It was collectecI in the lower I(lamath River in 2002 (M. Belchik, unpublishecI memo). FatheacI minnows often are the most abunciant species at sampling sites. Their effects on other fishes are not well unclerstoocI, although cleclines in catches of tui chub ancI blue chub have been associatecI with their ascenciance. The Sacramento perch is native to central California, where it has largely clisappearecI from its native habitats. It survives mainly when intro- clucecI into alkaline waters outsicle its native range (Moyle 20021. It was introclucecI by the California Department of Fish ancI Game into Clear Lake in the 1960s ancI spreacI throughout the Lost River ancI into the I(lamath
FISHES OF THE UPPER KLAMATH BASIN TABLE 5-2 Nonnative Fishes of the Upper Klamath Basin 189 S. pecles Adult Habitata Statusb Comments Goldfish, Carassius auratus Golden shiner, Notemigonus chrysoleucas Fathead minnow, Pimephales promelas Brown bullhead, Ameiurus nebulosus Black bullhead, A. melas Channel catfish, Ictalurus punctatus Kokanee, Oncorhynchus nerka Rainbow trout, O. mykiss Brown trout, Salmo trutta Brook trout, Salvelinus fontinalis Sacramento perch, Archoplites interruptus L, P. R. W White crappie, Pomoxis annularis L, R Black crappie, P. nigromaculatus Green sunfish, Lepomis cyanellus Bluegill, L. macrochirus Pumpkinseed, L. gibbosus Largemouth bass, Micropterus salmoides Yellow perch, Perca flavescens L, R. P U L, R. P R L, P P. L, W P. L L, R L, R. C A A U ? U C C, R C C U C U U C U L, R. P A L, P P. W Locally common Bait fish Probably still spreading Widespread Localized populations May not be established Localized populations ? Widely planted, hatchery strains Localized in headwaters Spreading Abundant in a few reservoirs Recorded in Lost River Widespread in reservoirs P. W U Locally abundant L, R. P C Widespread P. L, R C Common in reservoirs Abundant in large reservoirs aHabitats are listed in order of importance for each species: C, cold-water streams; L, lakes; P. ponds and reservoirs; R. rivers; W. warm-water streams. bStatus in upper basin: A, abundant; C, common; R. rare; U. uncommon. River downstream to Iron Gate Reservoir (Buettner anti Scoppettone 19911. It is not particularly abundant in most areas where it is present. It has not yet establishecI itself in Upper Klamath Lake. If it cloes colonize Upper Klamath Lake, it will probably become abundant there, as it has in other shallow lakes (Moyle 20021. It feecis primarily on insect larvae (especially miciges), but aclults can be piscivorous (Moyle 20021. ENDANGERED SUCKERS OF THE KLAMATH BASIN All four native sucker species of the Klamath basin are endemic. The enciangerecI Lost River sucker anti shortnose sucker are part of a species
190 FISHES IN THE KLAMATH RIVER BASIN group of suckers that are large, long-livecI, late-maturing, ancI live in lakes but spawn primarily in streams; collectively, they are commonly referred to as lake suckers. Lake suckers populatecI much of the Snake River, Great Basin, anti Lahontan Basin region (Miller ancI Smith 1981, Scoppettone ancI VinyarcI 19911. Present-ciay species in the genus Chasmistes inclucle not only the shortnose sucker (C. brevirostrisJ but also the cui-ui (C. cujus) of Pyramid Lake, Nevada; the lune sucker (C. liorus); ancI a species that recently became extinct, the Snake River sucker (C. muriei) of Wyoming. Lost River suckers anti shortnose suckers (Figure 5-1) are closely relatecI to the more speciose ancI wiclely clistributecI sucker genus Catostomus; some recent taxonomic treatments place Lost River suckers in this genus (e.g., Moyle 20021. The lake suckers cliffer from most other suckers in having terminal or subterminal mouths that open more forward than clown, an apparent aciap- tation for feeding on zooplankton (small swimming animals) rather than suctioning foocI from the substrate (Scoppettone ancI VinyarcI 19911. Zoo- planktivory can also be linkecI to the affinity of these suckers for lakes, which typically have greater abundances of zooplankton than clo flowing waters. Historically, Lost River suckers ancI shortnose suckers occurred in the Lost River ancI upper I(lamath River ancI their tributaries, especially Tule i :; j ~ ~ ~ ~ u . i a . ,..~ ~ I ...... a-- - .. . . ~ A} Lost R'n'er Sucker ................ B) Shortnose Sucker FIGURE 5-1 Endangered suckers of the I(lamath River basin. (A) A Lost River sucker from Clear Lake; (~) a shortnose sucker from Clear Lake. Source: Moyle 2002, pp. 199, 203. Drawings by A. Marciochi. Reprinted with permission; copy- right 2002, University of California Press.
FISHES OF THE UPPER KLAMATH BASIN 19 Lake, Upper I(lamath Lake, Lower I(lamath Lake, Sheepy Lake, and their tributaries (Moyle 2002; USFWS 2002, Appendix D). Their current distri- bution (Table 5-3; Figures 5-1 and 5-2) reflects a combination of local extirpations and redistribution through water management. Suckers no longer occur in Lower I(lamath Lake or Sheepy Lake, which were exten- . . · · . . ~ . · sively drained in the 1920s; the populations in Tule Lake apparently do not reproduce successfully. luveniles of Lost River and shortnose suckers have been found in much of the Lost River, but they probably originate in Miller Creek (Shivery et al. 2000a). An additional population, probably consisting of shortnose suckers, was extirpated from nearby Lake of the Woods, Oregon, in 1952 when government agencies poisoned the lake to remove potential competition with trout (53 Fed. Reg. 27130 F198811. The endan- gered suckers also are found in the main-stem reservoirs of the I(lamath irrigation project (Chapter 3; Figure 1-4), but these populations appear to be nonreproducing (Desjardins and Markle 2000, USFWS 20021. Repro- ducing populations exist in Clear Lake and perhaps the Lost River. Short- nose suckers also have a reproducing population in Gerber Reservoir (Moyle 2002, USFWS 2002). Accounts of sucker distribution often are complicated by difficulties in distinguishing species, especially when the fish are young. Lost River suck- ers and shortnose suckers are partly distinguished from I(lamath largescale suckers and I(lamath smaliscale suckers by greater maximum size. The Lost River sucker can be 26-40 in. long, the shortnose sucker no longer than 21 in., the I(lamath largescale sucker no longer than 18 in., and the I(lamath smaliscale sucker, a poorly studied species, at least 18 in. The Lost River sucker differs from the shortnose sucker and the I(lamath largescale sucker with respect to some anatomical features of the head' mouth, lips, gill rakers, and body shape (Cunningham et al. 20021; it can generally be distinguished by its longer head and narrower, smaller mouth (see Figure 5-11. The life histories of Lost River suckers and shortnose suckers are in some ways similar to those of anadromous salmon. Salmon spawn in fresh- water and live most of their lives at sea before returning to their natal (birth) rivers to spawn and die. Similarly, the adults of the endangered suckers commonly ascend from lakes to rivers to spawn, the eggs hatch in gravel, and the larvae float or swim downstream to lakes, where they grow and mature before returning to rivers or springs to spawn. Unlike salmon, lake suckers spawn repeatedly. It is not known which individuals return consistently to their natal rivers to spawn, but at least 50°/O do return at least one time to a river in which they have previously spawned (Cunning- ham et al. 20021. There are many exceptions to these generalizations. For example, some individuals or subpopulations spawn in lakes, whereas oth- ers live their entire lives in rivers or streams. The repeated spawning of the
92 FISHES IN THE KLAMATH RIVER BASIN TABLE 5-3 Current ancI Former Distribution of Aclult Lost River Suckers anti Shortnose Suckers in the I(lamath Basin Habitatsa Lost Map River Shortnose Code Suckers Suckers Reference Upper I(lamath Lake Peripheral Springs Boulder Springs Cinder Flats Ouxy Springs Silver Bldg. Springs Sucker Springs Harriman Springs Barkley Springs Tributaries Wood River Lower Williamson River Upper Williamson River Sprague and Sycan Lake of the Woods, OR Lower I(lamath Lake, CA Clear Lake, CAd + + 3 4 6 7 Spawn Spawn Spawn Spawn 5 Spawn Spawn Spawn 8 9 10 11 12 0 13 + 14 Willow Creek 15 Boles Creek 16 Gerber Reservoir 17 Sheepy Lake 18 Sheepy Creek 19 Tule Lake 20 Lost River 21 J.C. Boyle Reservoir 22 Copco Reservoir 23 Iron Gate Reservoir I(lamath River Moyle 2002 Spawn Spawn Spawn Spawn Spawn Spawn - b Spawn Spawn Spawn ob Spawn o + Spawn + .-c + .. +e Spawn Spawn Spawn o + + +e Spawn - (+) (+) spawnf Spawn (+) (+) (+) (+)g 24 25 (+) (+) Hayes et al. 2002 Hayes et al. 2002 Hayes et al. 2002 Hayes et al. 2002 Hayes et al. 2002 59 Fed. Reg. 61744  59 Fed. Reg. 61744  Markle and Simon 1994 Cunningham et al. 2002 Janney et al. 2002 Moyle 2002 Scoppettone and Vinyard 1991 59 Fed. Reg. 61744 , USFWS 2002 Moyle 2002 Spawn Moyle 2002 59 Fed. Reg. 61744  Moyle 2002 Moyle 2002 USFWS 2002 59 Fed. Reg. 61744  53 Fed. Reg. 27130  Scoppettone 1988, Scoppet- tone and Vinyard 1991 Moyle 2002 59 Fed. Reg. 61744  (+) (+) aTributary streams and springs are listed under lakes into which they flow. bR. S. Shively, U. S. Geological Survey, I(lamath Falls, Oregon, personal communication, 2002. CAn extirpated population of C1)asmistes in Lake of the Woods, Oregon, originally referred to as C. stomias (Andreasen 1975), may have been another population of shortnose suckers (Moyle 2002). dDrainage for Clear Lake includes numerous small reservoirs and tributary streams that contain both species (USFWS 2002, Appendix D). eShortnose suckers in Clear Lake and Gerber Reservoir may have been confused with I(lamath largescale suckers or with shortnose suckers and I(lamath largescale sucker hybrids (D. F. Markle, Oregon State University, personal communication 2002), although genetic information indicates that hybridization is rare (D. Buth, University of California at Los Angeles, and T. Dowling, Ari- zona State University, personal communications, July, 2002). fLarvae in Lost River apparently do not survive (Moyle 2002). gShortnose suckers in Copco Reservoir may have hybridized with I(lamath smallscale suckers ( Scoppettone and Vinyard 1991). Abbreviations: +, currently present; +-, previously present; (+), small population, probably non- breeding; Spawn, current or previous spawning; Spawn -, spawning inferred from fish in spawning condition; 0, not known ever to occur; -, lack of information.
FISHES OF THE UPPER KLAMATH BASIN 193 .~ | Crater ~ T 1 (Wood River By (era n ch (BEM) ! ~: Agency Lake \~` ~ Ranch (USER)\ ~ ~ C h i loquin Dam Upper Klamath Lo National Wildlife Refuge K Aft ~ Upper Lake of We Woods of. 1 ~ 5~ Klamath Marsh National Wildlife; ~ Ret :- ~ .f~ ~I' _ \~ Alto Pry ~ of' Sprague Rived Upper Klamath National Wildlife ~ Refuge Math Falls ..... BearV 11 y ~ ............ At. National WiIdlifeLL~ . ~--~...~ - .... Refuge ~ . ~ ................ N,.,.,.,~mlll~ ...~................ ~ ~ ............... . ~ 6 Refuge ~ = ....... Gerber ( Reservoir \ ,,, ~ Lower Klamath . National Wildlife 131 t ~~ Tule Lake National Wildlife Refuge ~.`'Z'e' Clear Lake ,~- ~ W Ci;'~ ~ ~ ~ ' Clear Lake National Wildlife Refuge ~ J ~ Oregon_ California Klamath Project Service Area O s 10 1s 20 Miles FIGURE 5-2 Locations of current and past populations of Lost River suckers and shortnose suckers. Numbers indicate current or former locations of suckers; light gray shows the area of the Klamath Project; dark gray shows standing water. See Table 5-3 for additional information. enciangerecI suckers, combined with their exceptional longevity, allows in- cliviclual aclults to contribute to multiple year classes. Successful year classes are crucial to survival of both species, as explainecI below. The requirements of the two species of enciangerecI suckers are best unclerstoocI in the context of their life-history stages, as clescribecI below. Unless a species-specific difference is inclicatecI, the description of any given life-history feature is assumed to apply to both species. The quantity anti quality of information on the species have increased substantially since the fishes were listecI as enciangerecI in 1988.
94 FISHES IN THE KLAMATH RIVER BASIN Spawning Spawning occurs in tributary streams, in springs caused by upwelling of grounc~water in lakes, ancI around springs in rivers. The suckers may mi- grate as little as 2-4 mi up a stream from a lake (for example, up Willow Creek from Clear Lake), or over 20 mi (for example, up Boles Creek from Clear Lake ancI up the Sprague River to RM 74 from Upper I(lamath Lake; R. S. Shively, U. S. Geological Survey, I(lamath Falls, Oregon, personal communication, 2002~. Upstream migrations commence when snowmelt leacis to increases in river discharge from early February through early April for Lost River suckers ancI from late February to late May for short- nose suckers (Moyle 2002~. Spawning can occur at temperatures of 5.5- 19°C (Moyle 2002~. For example, migrations of Lost River suckers up the Williamson River in 2001 were concentrated in April ancI May ancI showed a peak in mid-April. Spawning of chortnoce c,~ckerc necked in miA-M~v _ ~~ ~ / ~ · 1 ~()()1 <(:unnlngham et al. 2002~. In anv given veer some temnora1 senara- . . . . ----a v- --- ~ ----a ~ _---r~---- --r----- t~on ot spawning among species may occur. I(lamath largescale suckers migrate first ancI are followecI by Lost River suckers ancI then shortnose suckers (Coleman et al. 1988, cited in Scoppettone ancI VinyarcI 1991), although migrations of the three may overlap (USGS 2002~. 1 ·1 1 1 1 ~ V V ~ . _ . . Shortnose suckers were numerically dominant in the lower Williamson River in 2001, but Lost River suckers outnumbered shortnose suckers by more than 10 to 1 at Chiloquin Dam, about 9 mi farther upstream (Cun- ningham et al. 2002, lanney et al. 2002~. Thus, the Lost River suckers may be more likely than shortnose suckers to migrate upriver to spawn, or perhaps the two species react clifferently to clams. In 2001,30 shortnose suckers were collectecI at lakeshore sites, compared to 900 founcI in the Williamson River, whereas Lost River suckers were five times more abun- ciant at spawning sites in the lake than in the Williamson River system (Hayes et al. 2002~ Cunningham et al. 20021. This suggests that spawning ~ ~ ~ v , by shortnose suckers in Upper I(lamath Lake is relatively rare at present. Shortnose suckers that clo spawn in the lake use the same spawning sites as Lost River suckers. In flowing water, the suckers spawn in riffles or runs with moderate current (less than 3.3 ft/s) over cobble or grave! bottoms at depths of 0.7-6.6 ft (Scoppettone ancI VinyarcI 1991, Perkins ancI Scop- pettone 1996, Markle ancI Cooperman 2002~. Grave! appears to be pre- ferrecI; patches of grave! aciclecI to a spawning area will be used if flow ancI clepth are appropriate (Goiclen 1969, Scoppettone ancI VinyarcI 1991, Moyle 2002~. Spawning in the upper Sprague River appears to be concentrated around springs (L. Dunsmoor, cited in USFWS 2002~. Spawning behavior is similar to that of other suckers in that one female spawns with several males anti the fertilizecI eggs, which are 2.5-3.2 mm in diameter, cirop into spaces in the gravel.
FISHES OF THE UPPER KLAMATH BASIN 195 Sampling at six known spawning sites along the eastern shoreline of Upper I(lamath Lake (Sucker Springs, Silver Building Springs, Ouxy Springs, Boulder Springs, Cinder Flats, and Modoc Point) indicates that Lost River suckers spawning in the lake are slightly larger than those ascending the Williamson River (lake fish were 150-200 mm longer: Hayes et al. 2002, p < 0.051. Nearly 80% of the fish captured at lake spawning sites occurred at three of the six sites (Sucker Springs, Silver Building Springs, and Ouxy Springs). As is common among spawning suckers, males outnumber fe- males at spawning sites. Sex ratios at nonspawning sites in Upper I(lamath Lake indicate a predominance of females; males tend to remain at spawning sites, whereas females do not (Coen et al. 20021. Lake spawning occurs in 0.5-3.7 ft of water; 95°/O of successful spawn- ~ngs occur In water deeper than 1.0 tt, and about And occurs at 1-2 ft (I(lamath Tribes, in USFWS 20021. Spawning aggregations were present from mid-March to early May. Peak abundances at all sites occurred during the first 2 wk of April, and a second peak occurred at Sucker Springs, the most heavily used site, in late April. The relative spawning condition (pre- spawn, ripe, postspawn) of fish captured in Upper I(lamath Lake from February to lune 2001 suggests that some eastern regions near spawning sites, such as Modoc Point and Goose Bay, are staging areas for spawning and that some western bays are used more heavily after spawning (Coen et al. 20021. The temporal sequence of capture of the sexes during the spawn- ing season also suggests that males move to staging and spawning areas ahead of females. Evidence from Hayes et al. (2002) is consistent with earlier conclusions by Perkins et al. (2000b) that river spawners and lake spawners constitute subpopulations of Lost River suckers in Upper I(lamath Lake, but does not prove complete segregation of populations. Of 201 Lost River suckers tagged during previous years and recaptured at springs in the lake in 2001, with some recaptures separated by as much as three yr, 198 (98.5%) were captured both times at eastern shore spawning sites. The other three fish had been tagged in the Williamson River. Also, 76% of the fish recaptured at the Chiloquin Dam fish ladder in 2001 had been tagged originally at the ladder in previous years, and 20% of the fish had been tagged at other sites on the Williamson River (lanney et al. 20021. About half the Lost River suckers caught in Upper I(lamath Lake were from sites other than those where they were tagged, either for within-year or between-year recaptures; this indicates that lake-spawning fish do not restrict their breeding activities to a single lacustrine spawning site. Ten shortnose suckers captured in 2001 were recaptures from previous years; all had originally been captured at shoreline sites. Movement between lake sites was apparent, as with the Lost River sucker.
196 FISHES IN THE KLAMATH RIVER BASIN Female Lost River suckers contain 44,000-236,000 eggs, ancI female shortnose suckers contain 18,000-72,000 eggs. Larger females bear more eggs, as is typical of most fishes (USFWS 20021. It is unknown whether inclivicluals of either species spawn more than once each year or whether inclivicluals spawn every year. Recapture ciata on lake spawners (Perkins et al. 2000b, Hayes et al. 2002) suggest that some Lost River suckers spawn every year. Cui-ui (Chasmistes cujus) are known to spawn several huncirecI times over a period of 3-5 clays (Scoppettone ancI VinyarcI 19911; Lost River suckers ancI shortnose suckers might behave similarly. Coen et al. (2002) founcI that 75°/O of male but only 40°/O of female Lost River suckers ancI 69% of male but only 46% of female shortnose suckers captured in February-lune 2001 were in spawning condition (see also Coen ancI Shively 20011. These observations suggest that a large portion of the aclult popula- tion of both species is not in spawning condition cluring any given spawning season. Observations of tagged fish frequenting more than one lake spawn- ing site in a year suggest multiple spawning events for incliviclual fish. Frequency of spawning is relevant to the populations' potential for recovery. Larvae Embryos remain in the grave! for 2-3 wk (USFWS 20021. The subse- quent larval stage lasts for about 40-50 clays (Markle ancI Cooperman 20021. Stream-spawnecI larvae emerge ("swim up") from the grave! ancI immecliately move downstream, mostly at night, in late March to early rune, clepencling on spawning ciate (Moyle 20021. The abundance of larvae peaked in the Williamson River system 21 clays after the peak in spawning (Coleman et al. 1988, cited in Scoppettone ancI VinyarcI 19911. Larvae spawned in the Williamson River system pass to Upper I(lamath Lake in as little as a clay. More than 99°/O of larvae enter the lake before the caucial fin has formecI ancI well before the yolk sac is absorbed, after which the fish must feecI (Cooperman ancI Markle 20001. How these movement rates are relatecI to location of spawning (lower Williamson River or Sprague River below or above Chiloquin Dam) ancI how different they wouicI be if more fish spawned above the clam are unknown. Larval mortality in the William- son River is around 93°/O per clay (L. Dunsmoor, personal communication, in Markle ancI Cooperman 20021. Mortality in fishes with planktonic lar- vae is in general very high (Houcle 1987, 19971. Larval habitat is best clescribecI as shallow, nearshore, ancI vegetated in both rivers ancI lakes (Figure 5-3) except in Clear Lake ancI Gerber Reser- voir, which lack vegetation (I(lamath Tribe 1991, Markle ancI Simon 1994, Reiser et al. 20011. Larvae are most abundant in the northeastern portion of Upper I(lamath Lake, inclucling the Williamson River estuary ancI the lower Williamson River (Markle ancI Cooperman 20021. In Upper I(lamath
FISHES OF THE UPPER KLAMATH BASIN ,, 97 ~^^A_AA~A~ Larvae Juveniles 1 ft. 2 ft. 3 ft. 4 ft. Adults FIGURE 5-3 Generalized view of habitat of young suckers in Upper I(lamath Lake. Source: USFWS 2002, p. 83. Lake, larvae first concentrate near emergent vegetation at the mouth of the Williamson River for several weeks ancI then appear in other regions of the lake where emergent vegetation is found; that this process can continue for more than 2 ma is not surprising, given the protracted spawning period of the suckers (Cooperman anti Markle 20001. Studies of the larval use of habitat have focused on the importance of clepth anti vegetation as components of habitat. Observations by Coleman et al. (1988), Buettner anti Scoppettone (1990), the I(lamath Tribes (I(la- math Tribe 1991; I(lamath Tribe, Natural Resources Department, Chilo- quin, Oregon, unpublishecI material, 1996), Cooperman anti Markle (2000), anti Reiser et al. (2001) indicate use of shallow water (less than 4.3 ft anti
198 FISHES IN THE KLAMATH RIVER BASIN often less than 20 in.) sometimes in areas clevoicI of cover but more usually near emergent vegetation, such as bullrush beds. Larvae use emergent veg- etation primarily from early May through late rune, although larvae may be founcI up to micI-luly (see Reiser et al. 2001) because spawning continues into late May. Submerged aquatic vascular plants apparently are less im- portant than emergent vegetation (Cooperman 2002), probably because macrophyte becis are selclom well clevelopecI in spring, when much of larval growth occurs. Larvae may not necessarily aggregate within clense vegeta- tion itself but rather near it or in openings in the vegetation in areas cle- scribecI as "pockets of open water surrounclecI by emergent vegetation," or "the open water/emergent vegetation interface" (Reiser et al. 2001, p.4-91. Successful spawning ancI recruitment of suckers in Clear Lake, which is largely clevoicI of emergent anti submerged vegetation, show that larvae can survive without such vegetation. Clear Lake is very turbid, however, ancI this may provide protection from visual predators. Laboratory tests show that predation on larvae by fathead minnows is highest when larvae lack cover (Dunsmoor 19931. Young and small fishes in freshwater and marine habitats woric~wicle often take refuge in clense vegetation when threatened by preciators, although larvae of some species are entirely pelagic. Clear Lake contains flooclecI annual grasses ancI herbs ancI emergent ancI submerged vegetation in tributaries that may be used by larvae, ancI it has fewer introclucecI predators, such as yellow perch anti fathead min- nows, than cloes Upper I(lamath Lake (USFWS 20021. Thus, successful recruitment in Clear Lake cloes not demonstrate that vegetation is unimpor- tant in Upper I(lamath Lake. Successful spawning apparently cloes not occur in any of the main-stem reservoirs, which have steep shorelines, lack substantial emergent vegetation, have abundant predators, anti may lack spawning areas (Desjarclins ancI Markle 20001. Juveniles (1-4 Inches) Larvae are consiclerecI juveniles at a length of 1-4 in., which the suckers generally achieve by the encI of luly (USFWS 20021. Juveniles are termed young of the year (YOY) or age O through their first winter. They spencI daytime near shorelines over clean, rocky bottoms composed of sancI, gravel, ancI small bouiclers (Simon et al. 2000; Figure 5-31. YOY use both vegetated ancI unvegetatecI portions of shoreline, generally in water less than 4.3 ft creep (USFWS 20021. I(nowledge of the extent to which vegetation is used is complicatecI by the clifficulties of sampling juveniles in clense vegetation (Reiser et al. 20011. Abundance of YOY at first is greatest in the northeast- ern portion of Upper I(lamath Lake; as summer progresses, young fish move southward in the lake anti into creeper water anti become less associ- atecI with shorelines, ancI they become more oriented toward the lake bot-
FISHES OF THE UPPER KLAMATH BASIN 199 tom (Gutermuth et al. 20001. Simon ancI Markle (2001) suggest that over- winter mortality of first-year juveniles approaches 90°/O. After their first year, juveniles are founcI throughout the lake but are most abundant in the northern one-thircI of the lake, as are aclults, although it may be important that sampling has been concentrated on this area (Reiser et al. 20011. luvenile Lost River suckers appear to clepencI less on shallow-water habitats than juvenile shortnose suckers, as shown by sampling with beach seines (Simon et al. 2000), ancI juvenile shortnose suckers are apparently more strongly oriented toward the lake bottom than juvenile Lost River suckers (Scoppettone et al. 19951. Subaclults (4-10 Inches) ant! Aclults Subaclults are the least-stucliecI age group. It is assumed that their re- quirements ancI habits are most like those of nonspawning aclults but their behavior is obscure because they are too fast to catch in seines or trawls, too creep to catch in cast nets, ancI often too small to gilinet. Given that suckers may spencI the first 3-8 yr of their lives as subaclults, aciclitional information on this stage couicI be important. Lost River suckers grow rapicIly for their first 5 or 6 yr to a length of 14-20 in. (Scoppettone 19881. Some males reach maturity (i.e., are capable of spawning) at 4+ yr ancI 15 in. ancI some females clo so at 7+ yr ancI 21 in., but most fish mature at 8 or 9 yr; males often mature earlier than females. At maturity, growth slows (Scoppettone 1988, Buettner ancI Scoppettone 1990, Scoppettone et al. 1995, Perkins et al. 2000a). The largest ancI oiclest fish are females. The oiclest known Lost River sucker (43 yr) was obtained in Upper I(lamath Lake cluring a fish kill in 1986 (Scoppettone 19881. Female shortnose suckers apparently grow faster ancI larger than males. Both male ancI female shortnose suckers mature as early as 4+ yr. Males can be mature at 11 in. ancI females at 13 in., although maturation at 5-7 yr is more usual. The oiclest known shortnose sucker (33 yr) was taken from Copco Reservoir in 1987 ancI was 19 in. long (Scoppettone 19881. Aclult Lost River suckers forage primarily on zooplankton ancI benthic (bottom-c~welling) macroinvertebrates (Coleman et al. 1988, Scoppettone ancI VinyarcI 19911. The shortnose sucker, as couicI be preclictecI from the more terminal position of its mouth, feecis preclominantly on clacloceran zooplankters (water fleas), although the guts of only a few aclults have been examined (Coleman et al. 19881. The presence of detritus in the guts of shortnose suckers from Clear Lake indicates that shortnose suckers may also feecI close to the bottom (Moyle 20021. Aclult suckers select water depths of 3-15 ft. as shown by ciaylight spring ancI summer observations; their strongest preference appears to be for 5-11 ft (Reiser et al. 2001, USFWS 20021. Their minimal use (1% of
200 FISHES IN THE KLAMATH RIVER BASIN daytime observations) of shallower water couicI reflect avoidance of high light intensities ancI thus of aerial predators; limitecI use of the deepest water (about 4°/O of daytime observations), particularly in summer, may reflect avoidance of low concentrations of clissolvecI oxygen (Chapter 31. Although aclults of the Lost River suckers ancI shortnose suckers are captured together in many places in Upper I(lamath Lake, some differences in their distribution suggest different habitat preferences. For example, in 2001, Lost River suckers were 2-3 times more abundant in trammel net samples from the western shoreline of Upper I(lamath Lake, whereas short- nose suckers were 2-3 times more abundant in samples from the eastern shore (Coen et al. 20021. Possible habitat differences in these regions might be worthy of further investigation, although the differences couicI reflect chance encounters with aggregations of the two species. Physiological Tolerances Lake suckers in general are relatively tolerant of water-quality concli- tions that are unfavorable or even lethal for many other fishes. For ex- ample, suckers in goocI condition occur in Tule Lake, which perioclically experiences extremes of clissolvecI oxygen, pH, ancI ammonia that are toxic to fathead minnows, a tolerant species (Dileanis et al. 1996, cited in USFWS 20021. Other lake sucker species are similarly tolerant. Endangered cui-ui evolvecI in the very alkaline (pH, 9.0-9.5) ancI saline (5 ppt) waters of Pyramid Lake, Nevada, where only five or six other native fish species persist. The only nonincligenous fish species to have successfully colonizecI Pyramid Lake is the Sacramento perch (G.G. Scoppettone, U. S. Geological Survey, Reno, Nevada, personal communication, 20021. Most fishes cannot tolerate sustained pH in excess of 9 (Falter ancI Cech 19911. Upper I(lamath Lake suckers can tolerate pH approaching 10, temperatures up to 31-33°C, concentrations of unionized ammonia up to 0.4-0.5 mg/L, ancI clissolvecI oxygen concentrations clown to 1.5 mg/L. Beyond these threshoicis, the suckers clie in laboratory tests (typically con- cluctecI on juvenile fish); larvae are more sensitive than larger fish (Falter ancI Cech 1991, Martin ancI Saiki 1999, Saiki et al. 1999, Moyle 20021. Mortality is high in aclult suckers below oxygen concentrations of about 1 mg/L (Chapter 61. Falter ancI Cech (1991) found that shortnose suckers hacI much lower tolerance of high pH Measured as pH at which swimming equilibrium was lost) than two other endemic fishes, the I(lamath tui chub ancI the I(lamath largescale sucker. Shortnose suckers lost equilibrium at a mean pH of 9.55, tui chub at 10.75, ancI I(lamath largescale suckers at 10.73. Maximum pH in Upper I(lamath Lake cluring summer phytoplank- ton blooms frequently exceeds 9.5 at the surface cluring ciaylight hours, but pH cluring episodes of mass mortality generally is about 7.5-8.5 (Perkins et
FISHES OF THE UPPER KLAMATH BASIN 20 al. 2000b), indicating that high pH cloes not cause mass mortality (Chapter 31. In Upper I(lamath Lake in late summer, cluring times of physiological stress, suckers may seek higher water quality, such as that of springs ancI river mouths, even though such areas are otherwise avoiclecI, probably because they are too shallow or too clear (USFWS 2002, Appendix D; Chapter 61. Physiological tolerance tests generally are performed in a laboratory on single factors helcI at constant values, whereas factors in nature often vary over time ancI space, co-occur, ancI can operate synergistically. Summer conditions in Upper I(lamath Lake typically involve episodes of high pH, high unionized ammonia, ancI low clissolvecI oxygen in combination with high temperatures that increase the oxygen clemancI of the fish. High con- centrations of unionized ammonia can cause structural ciamage to gills, which can increase the susceptibility of fish to low concentrations of clis- solvecI oxygen. High pH (over 9) inhibits ammonia excretion, thus creating stress (Lease 2000, cited in USFWS 20021. Susceptibility to columnaris disease, which is caused by the bacterium Flavobacterium columnare, in- creases with increasing temperature but decreases with increasing ammo- nia concentrations (Morris et al. 2000, Snycler-Conn et al. unpublishecI in USFWS 2002). As an adjunct to laboratory studies, Martin ancI Saiki (1999) placecI cages containing juvenile Lost River suckers in Upper I(lamath Lake for 4- ciay periods. High mortality occurred at high pH, high concentrations of unionized ammonia, ancI low concentrations of clissolvecI oxygen; low clis- solvecI oxygen was the strongest correlate with mortality. At sublethal tem- peratures ancI concentrations of unionized ammonia, fish were tolerant of higher pH than expected from the laboratory studies (fish toleratecI pH as high as 10.81. The stucly suggests that laboratory tests of single factors shouicI be viewed as being only indicative of the extremes that can be toleratecI; they are not strictly predictive of responses in the fielcI. From the viewpoint of physiological stress on fishes generally, ancI especially for coicI-water fishes, water-quality conditions are poor through- out much of the I(lamath basin, as explainecI in Chapters 3 ancI 4. Physi- ological threshoicis for suckers, however, are reachecI or exceeclecI less ex- tensively than for most fishes because of the high tolerance of suckers. Harm to suckers caused by poor water quality is known for Upper I(lamath Lake ancI may also occur in the Lost River ancI upper I(eno Reservoir (Lake Ewauna). In other lacustrine or flowing-water environments of the basin, however, poor water quality may be much less important than other factors for suckers, although it may strongly affect some other fishes. In Upper I(lamath Lake, suckers are aciversely affected by poor water quality, which is a byproduct of very high abundances of Aphanizomenon flos-aquae, a planktonic bluegreen (cyanobacterial) alga. Peak abundances
202 FISHES IN THE KLAMATH RIVER BASIN of Aphanizomenon occurring in late summer or early fall cause very high pH. Uncler certain meteorological conditions overturn of a stratified water column ancI collapse of the Aphanizomenon population combine to cause clepletion of oxygen throughout the water column ancI distribution of high concentrations of unionized ammonia (Chapter 31. The adverse water-quality conditions in Upper I(lamath Lake poten- tially have three types of effects on enciangerecI suckers in Upper I(lamath Lake: (1) mass mortality of large fish, (2) mortality, either episodic or continuous, of small fish or larvae, ancI (3) physiological stress on one or more age classes, which leacis to physiological impairment but not necessar- ily cleath. Poor water quality in Upper I(lamath Lake is a clocumentecI cause of the episodic mass mortality of large suckers in the lake. The recent history of these episodes is given in this chapter, ancI the factors producing cleath are cliscussecI in Chapter 3. Extensive research on the clirect cause of mortal- ity cluring episodes of mass mortality has lecI to the reasonably firm conclu- sion, supported by scientific evidence, that mortality is caused by inacI- equate amounts of clissolvecI oxygen. The two other potential clirect causes of mortality, pH ancI unionized ammonia, appear not to control mass mortality. DissolvecI oxygen, unlike pH ancI unionized ammonia, remains adverse continuously for many clays cluring episodes of mass mortality, whereas pH ancI unionized ammonia clo not. Thus, although aciclitional studies of mechanisms leacling up to mass mortality are warranted, the clirect cause in large fish seems to be unclerstoocI reasonably well. There is insufficient evidence to show whether extreme water-quality conditions also cause mortality of juveniles ancI larvae. Laboratory experi- ments indicate such potential, but it has not been clocumentecI in the fielcI. FielcI documentation, especially if mortality were steady rather than epi- soclic, wouicI be clifficult for the smaller life stages of fish because of their quick deterioration ancI clispersal after cleath. The possibility that graclual or episodic mass mortality of small fish occurs shouicI be stucliecI. Adverse water-quality conditions can affect fish inclirectly, as explainecI above. Laboratory studies are useful, but fielcI indicators of stress also are important in that sublethal responses to stress cannot always be proclucecI in an interpretable way in the laboratory. Indicators of physiological stress inclucle unusual or recurrent epizootics, poor bocly-conclition factors, physi- cal anomalies, ancI low growth rates compared with those in populations that are not exposed to adverse water-quality conditions, abnormally low fecundity or fertility of mature fish, ancI behavioral aberrations. Some at- tention has been given to the indicators for example, physical anomalies in suckers of Upper I(lamath Lake are common (USFWS 2002) but a more comprehensive effort at evaluating indicators of stress probably is warranted.
FISHES OF THE UPPER KLAMATH BASIN 203 Overall, there is no cloubt that poor water-quality conditions are suppressing the enciangerecI suckers of Upper I(lamath Lake through mass mortality of large fish. Less clear is the role of potential aciclitional sup- pression through mortality of smaller fish or sublethal effects of physi- ological stress caused by poor water-quality conditions on any or all life stages. Population Size Abundances of larval anti juvenile suckers have been estimated from fielcI samples over the last several years (e.g., Simon et al. 20001. Calcu- latecI population sizes of aclults have been basecI on recapture of tagged fish cluring fish kills. The confidence intervals around the numbers are very large anti, because many of the assumptions of mark anti recapture methods are not met by these estimates, the estimates are of limitecI value (R. S. Shively, USGS, unpublishecI memo, 5 March 2002; USFWS 2002). Newspaper reports, eyewitness accounts, anti ciata on catch per unit effort leave little cloubt that the sucker population exploitecI by the snag fishery in the 1960s anti earlier was much larger than it was by the 1980s. Relative estimates of the size of the spawning run of suckers in the William- son River were first basecI on estimated catch rates anti later on stanciarcI- izecI recapture anti electrofishing methods. The estimates showed a marked decrease in abundance of fish cluring the micicIle 1980s. In 1984' the run of spawning Lost River suckers was estimated at 23~000' but it fell to 12~000 in 1985. Catch per unit effort of electrofishing fell by 57% for Lost River suckers anti by 83% for shortnose suckers from 1984 to 1986 before the major fish kill of 1986 (Scoppettone 1986' Bienz and Ziller 1987' SCOP- pettone anti VinyarcI 19911. The fishery was closecI in 1987. More recent estimates of abundance clepencI on catch per unit effort in stanciarclizecI trammel-net samples anti can be compared only among collections for the years 1995-2001. No universal or absolute estimates of the size of any age class of sucker are available. Estimates are relative, limitecI to specific sites (e.g., spawning areas), or are otherwise qualifiecI from the viewpoint of making an overall numerical assessment of the population. While the use of qualifiecI or rela- tive estimates is beneficial, efforts to make more comprehensive population size estimates in the future wouicI be clesirable (see Chapter 61. For purposes of ESA actions, the critical facts, which are known with a high clegree of certainty, are that the fish are much less abundant than they originally were anti that they are not showing an increase in overall abundance. Thus, the point of departure for research anti remecliation in the future is the neecI to restore abundance of the listecI suckers.
204 FISHES IN THE KLAMATH RIVER BASIN Age-Class Structure Most adult suckers in Upper I(lamath Lake are large and old. The uneven age distribution has characterized the populations for several de- cades. Through the 1980s, the age distribution of Lost River suckers was heavily skewed to fish 19-28 yr old. In 1986, the year before fishing was banned, recruitment had apparently been poor for about 18 yr; 95°/O of adult Lost River suckers were 19-30 yr old (Figure 5-4; Scoppettone 19881. The data for Lost River suckers shown in Figure 5-4 are based on fish obtained during fish kilos a sampling method with unknown but multiple v , ~ v ~ 1 · · 1 1 ~ · 1 .1 . 1 1 1 r. 1 rr 1 ~ . ~ biases, Including some evidence that older, larger fish sutter d~sproport~on- ately high mortality (Chapter 61. Assuming that the fish collected during fish kills are representative of the adult population as a whole, it can be concluded that many age classes were essentially missing from the lake before 1988, when the fishery was active. Closure of the fishery in 1987 greatly reduced mortality of spawners, after which additional mature fish began entering the spawning population (Figure 5-4B). Cessation of fishing apparently contributed to the produc- tion of a strong year class of both endangered sucker species in 1991, and to smaller but notable year classes also produced in 1990, 1992, and 1993 (Figure 5-4B; see Markle and Simon 1994, Cunningham and Shively 20011. These fish would have been expected to mature in the late 1990s, but the major fish kills that occurred in 1995,1996, and 1997 affected not only old spawners but also probably young spawners. Spawning runs declined in the late 1990s, with little evidence of substantial recovery until 2000 (Figure 5- 51. The upsurge in spawning numbers in that year and again in 2001 may represent maturation of fish from the 1991 and later year classes. It is possible that fish that lived through the fish kills of the middle 1990s were stressed by poor water quality and as a result experienced delayed matura- tion (e.g., Trippe! 1995, Baltz et al.1998), although Terwilliger et al. (M.R. Terwilliger et al., Oregon State University, Corvallis, OR, unpublished material, 2000 ~ found no evidence of impaired growth associated with periods of poor water quality in juvenile suckers of Upper I(lamath Lake. That spawning runs apparently increased in 1999-2001 shows that the species have substantial resilience, but this is no guarantee of recovery. Comparisons between 2000 and 2001 data indicate a weak but signifi- cant trend toward increasing average size among all spawning shortnose suckers and female Lost River suckers in the Williamson River (Cunningham et al. 20021. A similar significant trend toward increased median size at a variety of nonspawning sites in Upper I(lamath Lake was also found (Coen et al. 20021. When combined with evidence of low numbers of small river- spawning fish in recent years (Cunningham et al. 2002), the data could indicate year-class failure among fish that hatched in the middle 1990s and
FISHES OF THE UPPER KLAMATH BASIN A. 3o- 25 - 20 - cow ~ 15- as B- 40- ~ 25- as 35 - 30 - 20 - 1 ~ 0 - 20s Lost River Sucker N=l9O 0 - 5- o- till I,,,,ll,,,, In,,, I 0 5 10 15 20 25 Upper Klamath Lake 1986 30 35 40 45 Age (Years) . T ~ . 1995 1 1990  Lost River Suckers 1~ Shortnose Suckers au us o u 5- au . - row 1R ~ ~ ~ ~ n ~ ~ ~ n rem ~~ n n n not n n ~ n ~ ~ 7 1 1 I T I I T I I T I T T I T T T T I T T I T T I T 985 1980 1975 1970 1965 1960 Year Class FIGURE 5-4 Age distributions of suckers in Upper I(lamath Lake based on fish kills. (A) Age distribution of Lost River suckers in Upper I(lamath Lake based on the 1986 fish kill. Multiple peaks indicate strong year classes estimated as 1958, 1961, 1964, 1967. Source: Scoppettone and Vinyard 1991. Pp. 359-377 in Battle Against Extinc- tion: Native Fish Management in the American West, W.L. Minckley and Tames E. Deacon, eds. Copyright 1991 The Arizona Board of Regents. Reprinted by permis- sion of the University of Arizona Press. (~) Age frequency distributions of Lost River suckers and shortnose suckers in Upper I(lamath Lake based on fish collected from the 1997 fish kill. Effects of fishery closure in 1987 and of entry of successful 1991 year class are evident. Fish as old as 35 yr (spawned in 1962) were present. Source: Markle and Cooperman 2002, based on data from R. Shively, USGS.
206 300 - 250 - 200 - 50 - o- FISHES IN THE KLAMATH RIVER BASIN Spawner Abundance, Lower Williamson River, 1995-2001 ~ Lost River Suckers 1~ Shortnose Suckers 1 1, 1995 1996 1997 1998 1999 2000 2001 FIGURE 5-5 Spawning-run abundances of lake suckers, lower Williamson Riv- er, 1995-2001. Decline in spawners consistent with expected changes given fish kills of 1995-1997 is evident (1995 data were obtained before the fish kill that year). CPUE is a measure of catch per unit effort based on fish caught per unit of time spent fishing with trammel nets. Source: Modified from Cunningham et al. 2002, p. 30. that wouicI mature in the early 2000s. Concern over lost year classes might be tempered by an apparent trencI in increased overall abundance among river spawners in 1999-2001 (Figure 5-51. Catches of both species from the Williamson River in spring 2002 clecreasecI, however, by about 50°/O com- parecI with 2001 (R. S. Shively, U.S. Geological Survey, I(lamath Falls, Oregon, personal communication, October 8, 20021. Abundance inclex (catch per unit effort) for lake-spawning Lost River suckers clo not indicate an increase in numbers of spawners (1999, 3.0 fish/h; 2000, 2.0 fish/h; 2001, 2.4 fish/in), ancI the average size of lake-spawning fish increased significantly between 2000 ancI 2001, suggesting lack of recent recruitment into the spawning population (Hayes et al. 20021. Catches at the shoreline areas in 2002 also clecreasecI by about 15-20%. In fact, sampling in 2002 indicates that there has been no substantial recruitment into the aclult popu- lation since 1999 (R. S. Shively, U.S. Geological Survey, I(lamath Falls, Oregon, personal communication, October 8, 20021. Observations on size of spawners since 1984 (Perkins et al. 2000b) indicates that very large Lost River suckers (over 25 in. for males, ancI
FISHES OF THE UPPER KLAMATH BASIN 207 over 28 in. for females) have been lost progressively from the population, that recent spawning aggregations are macle up largely of meclium-size fish (18-24 in.), ancI that the median age of spawners for Lost River suckers is 12 yr ancI for shortnose suckers is 9 yr (as jucigecI from age- length relationships; Markle ancI Cooperman 20021. These findings sug- gest that successful year classes after 1991-1993 are largely absent, that is, that little recruitment of young spawners has occurred at the same time that the largest fish have been progressively removed by the fish kills; this raises a concern over future numbers of spawners ancI total reproductive output of the population. As with Lost River suckers, knowlecige of age distributions of shortnose suckers in Upper I(lamath Lake comes chiefly from three fish kills in the l990s, except that the ciata are even less complete ancI earlier ciata are lacking (Figure 5-4B). Indications from age distributions of fish collectecI after fish kills have indications similar to those for the Lost River suckers. One other trencI of note is that larger fish appear to spawn earlier in the season (Perkins et al. 2000b), but this trencI may have been obscured in recent years by a relative lack of small spawners (Hayes et al. 20021. RegarcIless of cause, multiple strong year classes with temporal separation in spawning between year classes is potentially advantageous because it decreases the likelihoocI of failure of all the year's larvae if environmen- tal factors vary for year to year cluring the breeding season (e.g., Trippe! 1995). Information on age distribution is a funciamental indicator of the status of a population, ancI it sometimes suggests reasons for failure of a species to recover. Although the 1990s, in apparent contrast with earlier years when the fishery was in place, have proclucecI recruitment into the subaclult ancI aclult stages, the fish entering these stages have been killecI in large numbers cluring episodes of mass mortality in Upper I(lamath Lake. Thus, one rea- son for failure of the populations to recover is probably suppression of reproductive capacity of the population clue to selective mortality of aclult fish. This cloes not, however, rule out the possibility that part of the expla- nation for lack of recovery lies in suppression of the number of fish entering the subaclult ancI aclult phases. The fish collectecI cluring fish kills indicate recruitment into the subaclult ancI aclult stages in all years, ancI especially in some years with notably abundant year classes (such as 1991), but the amount of this recruitment may be insufficient to support overall growth of the population. Thus, one bottleneck almost certainly involves the mass mortality of large fish, ancI a second bottleneck couicI be at one or more places in the life cycle between laying of eggs ancI the entry of fish into the subaclult ancI aclult categories. As cited above, numerous efforts are uncler way to identify unusual mortality or suppression of vigor in young fish, but no conclusions are yet available on this important matter.
208 FISHES IN THE KLAMATH RIVER BASIN Perspective on Age-Class Structure ant! Strength, Mortality, ant! Reproductive Output Most fishes experience astronomically high mortality in their early life- history stages. The millions or even billions of inclivicluals that hatch in a population are reclucecI by many orclers of magnitude at the time of matu- ration. On the average, a male ancI female just replace themselves over a lifetime of spawning, even though they may produce millions of fertile eggs. These facts are relevant to sucker recovery in several ways. High mortality among larvae ancI small juveniles is to be expected, but the rates shouicI plummet in later years, ancI oicI fish shouicI show low mortality. Small percentage changes in mortality of vouch fish can translate into large nonu- 1 . 1 rr 1 . 1 r .1 1 1 --if ~ - J ~ ----I ----- ----- -- --------- - ---- ~ ----I- r ~ r -- ~arlon alrrerences later Because or tne nlgn numbers of young inclivicluals. Thus, any steps that can be taken to increase larval anti juvenile survival in I(lamath Lake suckers couicI produce great benefits. The high mortality experienced by very oicI fish cluring the fish kills of the micicIle 1990s is especially alarming given the reproductive potential of these fish (e.g., Conover ancI Munch 20021. Large, oicI fish of most species produce clisproportionately more eggs than smaller fish. For example, in recI snapper (L?vtjan?vs campechan?vs), which is heavily fished ancI clepletecI throughout its North American range, a single 10-yr oicI female (26 Ib, 24 in.) can contain 9 million eggs, which is equivalent to the total egg output of 212 aclult females that are 3-4 yr oicI, weigh 2.2 Ib each, anti are 17 in. long. One 26 Ib oicI fish produces more eggs than 250 Ib of younger fish. Thus, loss of larger size classes in a population can have a disproportionate effect on egg production ancI future recruitment (Bohnsack 19941. The value of large fish, even in small numbers, is evident in the listecI suckers. ~1 1 ~ 1 1 1 11 · 1 · 1 1 1 1 the number or young produced and eventually recruited Into adulthood increased greatly just after the snag fishery was closecI (see Figure 5-4B), demonstrating that even low numbers of large fish can produce large num- bers of recruits (Markle anti Cooperman 20021. The clisproportionately high contribution of oicI fish is even greater than fecundity wouicI indicate. Because the quality of eggs (size ancI amount of yolk) proclucecI by oicI females may be greatest, larvae hatching from these eggs may be larger and more likely to survive the early periods of high mortality (e.g., Trippe! 19951. Although numbers of spawning fish in the Williamson River appear to have climbed in recent years, the reproductive potential of the population is lower than it was before the fish kills because the fish are smaller (Markle and Cooperman 20021. Reproductive output of a population is determined jointly by the number of spawners and the ace .. .. . , ~ . . . . . . . .... dlstrlbutlon ot spawners. 1 wo populations ot equal size that contain dltter- ent size distributions of fish will not be equal in reproductive value; the population with more old, large fish will have much higher reproductive
FISHES OF THE UPPER KLAMATH BASIN 209 potential. Any alterations that can be macle in the environmental conditions that clirectly affect the probability or severity of fish kills shouicI receive especially careful consideration (Chapter 31. Species that are long livecI ancI late to mature, such as the enciangerecI suckers of the Klamath basin, may responcI slowly both to clegraciation ancI to restoration of habitat requirements, in contrast to other species that mature more quickly. Thus, the presence of oicI fish is not in itself evidence of a sound population. In fact, even if oicI fish are numerous, their failure to propagate wouicI render them implicitly extinct until a reversal of the situ- ation occurs. Similarly, improvement of environmental conditions may leacI to beneficial changes in the population through recruitment of young age classes, but the final evidence of progress toward recovery, which is survival of these younger classes to maturity ancI oicI age, will not be evident for a clecacle or more. This special perspective on the long livecI, slow maturing suckers must be maintained in any evaluation of prospects for extinction ancI response to remecliation. Enciangerec! Suckers in Other Klamath Basin Waters Suckers occurred naturally in Tule Lake, Sheepy Lake, ancI Lower Kla math Lake, from which spawning fish ran up the Lost River (Table 5-31. All three of the lake populations apparently were extirpated when their waters were cirainecI for agricultural purposes around 1920 (Chapter 21. During the 1930s, after farming failecI in the former lake becI, the lakes were to some extent reinunciatecI, but not to their former depths. Suckers recolo- nizecI Tule Lake but not the other two lakes. There has been no evidence of successful spawning in Tule Lake, although fish from the lake eviclently spawn in the lower Lost River. Fish of both species, but mostly shortnose suckers, have been founcI regularly in the reservoirs between Keno ancI Iron Gate Dam (e.g., I. C. Boyle, Copco, Iron Gate). Apparently, they clo not spawn. Fish in these impoundments probably consist of inclivicluals that enter the Link River from Upper Klamath Lake anti survive passage at Link River Dam; they tencI to be oicI ancI large (Figure 5-61. The trip out of Upper Klamath Lake is one-way, inasmuch as no fish laciclers suitable for suckers are locatecI at Link River Dam or at any of the other clams along the Klamath River (Chapter 61. The great size ancI age of female fish as suggested by Figure 5- 6 couicI make such fish valuable as transplants to more favorable habitats. Reproducing populations of enciangerecI suckers exist in Clear Lake, in Gerber Reservoir, ancI in portions of the Lost River downstream (the Lost River couicI receive fish from Gerber Reservoir in its upper portion ancI from Tule Lake in its lower 7 mi, below Anderson Rose Dam). Clear Lake,
210 FISHES IN THE KLAMATH RIVER BASIN 4 Shortnose Sucker N = 19 3 2 1- o- Copco Reservoir 1987 . ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 0 5 10 15 20 25 30 35 40 45 Age (Years) FIGURE 5-6 Age structure of a small sample of shortnose suckers taken from Copco Reservoir, 1987. Source: Scoppettone and Vinyard 1991. Pp. 359-377 in Battle Against Extinction: Native Fish Management in the American West, W.L. Minckley and Tames E. Deacon, eds. Copyright 1991 The Arizona Board of Re- gents. Reprinted by permission of the University of Arizona Press. which was establishecI in 1910, contains populations of both species (Scop- pettone et al. 1995 estimated that 73,000 suckers occupied the lake), anti both show recent evidence of cliverse age structure anti continued successful reproduction anti recruitment. The reservoir is a source of irrigation water anti can be drawn clown cluring drought, which exposes the fish to multiple threats. Clear Lake was drawn clown to as low as 5°/O of capacity cluring 1992, anti fish collectecI after the cirawclown anti in the next spring were in poor condition, although their condition rebounclecI by the encI of the next summer (USFWS 20021. Success of shortnose suckers anti Lost River suck- ers in Clear Lake is encouraging in its own right anti as a potential rescue population that couicI be used for restoring populations in other water bodies. Extreme cirawclown, although prohibited by the USFWS biological opinion of 2002, is a threat if it shouicI occur inacivertently, anti the lake anti its suckers presumably are vulnerable to major environmental clisas- ters, such as a break in the clam (Moyle 20021. Unexpected changes in the spawning anti rearing habitats in Willow anti Boles Creeks above the reser- voir also couicI affect sucker abundances.
FISHES OF THE UPPER KLAMATH BASIN 2 Gerber Reservoir, which was created in 1925' contains shortnose suck- ers but not Lost River suckers. Shortnose suckers in Gerber Reservoir ex- hibit a wicle range of size classes, indicating successful reproduction anti recruitment. Gerber Reservoir is not connected to any other sucker popula- tion, so there is no possibility of genetic exchange. Condition of fish in Gerber Reservoir is known to vary from poor to goocI; poor condition was associated with lowest water levels in 1992 (the lake was drawn clown to 1°/0 of capacity). The population has not received a great clear of attention. Gerber Reservoir flows into the Lost River, which flows into Tule Lake (Figure 1-2~. Historical sucker runs out of Tule Lake anti up the Lost River were substantial; these runs supported commercial fisheries anti canneries (USFWS 2002~. Today, after the construction of multiple clams, only small numbers of the two enciangerecI species occur in the Lost River; shortnose suckers are more common than Lost River suckers. It is not known whether these populations are self-sustaining (USFWS 2002~. Spawning habitat is limitecI, anti spawning has been observed at only about three locations, although several other sites appear to provide appropriate spawning habi- tat. Small numbers of larvae anti juveniles have been collectecI in the river, but these fish couicI originate in Gerber Reservoir. Upstream movement from Tule Lake encis at Anderson Rose Dam, 7 mi above the lake. Spawn- ing habitat in the 7-ml reach is scarce, anti rearing habitat is compromised by poor water quality from water connected with Tule Lake sumps anti agricultural return flows. Water quality in the Lost River is generally poor; the river fails to meet several Oregon state-specifiecI water-quality thresh- oicis. Gradients in portions of the river are unfavorably steep for suckers? anti seasonal clewatering is common, as are clense plant growth anti algal blooms associated with poor water quality. Both summer anti winter fish kills were clocumentecI for the Lost River Diversion Canal region in the late l990s. Brown buliheacI (Amei?vr?vs neb?~Ios?vs) anti pumpkinseecI (Lepomis gibbos?vs) are abundant anti nine of the 16 fishes in the river are warm- water nonnatives. USFWS (2002' Appendix E, p. 31) conclucles that the Lost River is highly clegraclecI anti "can perhaps be best characterized as an irrigation water conveyance, rather than a river." Tule Lake, once larger than Upper I(lamath Lake but now less than 15% of its original size, contains populations of both enciangerecI species amounting to perhaps a few huncirecI fish represented by a few size classes of oicI fish (for example, 16-24 in.; Scoppettone et al. 19951. Suckers in Tule Lake typically have higher condition factors anti lower incidence of exter- nal parasites than suckers in other parts of the basin (USFWS 2002~. The Tule Lake populations historically were maintained by spawning runs up the Lost River, which for reasons listecI above now are extremely limitecI. Conditions within Tule Lake are deteriorating because of accumulation of sediment from agricultural sources. Alterations in water-management prac-
212 FISHES IN THE KLAMATH RIVER BASIN tices, however, could arrest deterioration. Some changes might even restore spawning runs. In 1999, the U.S. Bureau of Reclamation began releasing 30 cfs during the spawning and incubation period (April-lune), which led to detectable spawning activity below Anderson Rose Dam within 2 days (USFWS 20021. Such spawning could presumably lead to juvenile recruit- ment, but monitoring for presence of juveniles is needed. Collection of larvae reported by Shively et al. (2000a) is additional evidence of reproduc- tion. The relatively good condition of suckers in Tule Lake makes these populations valuable for the long-term survival of both species of suckers, especially given the continuation of fish kills in Upper I(lamath Lake. Conservation Status Lost River suckers and shortnose suckers were declared endangered by California in 1974 (Moyle 20021; Oregon placed both Lost River suckers and shortnose suckers on its protected list in 1987. USFWS first listed both sucker species as candidate (Category 2) species in 1982. They were pro- posed for listing as endangered in 1987 and were designated as endangered species in 1988 (53 Fed. Reg. 27130 F19881). Despite the controversy sur- rounding the species in recent years, only 13 written comments were re- ceived by USFWS during the comment period before listing; 12 of the comments favored listing, one expressed no opinion, and there were no comments opposing the listing. Reasons for listing are given in Chapter 6. A federal recovery plan has been developed (Stubbs and White 19931. Critical habitat was proposed in 1994 (59 Fed. Reg. 61744 F19941) but has not yet been formalized, nor has a recovery team been designated. CONCLUSIONS Human activities in the upper basin have affected not only the listed suckers, but virtually all the native species, several of which are greatly diminished in distribution and abundance. In particular, bull trout and slender scuipin have become rare in the basin in recent years. The Lost River system, which appears to have changed the most in the last 30 yr was dominated by blue chub, tui chub, and the three native sucker species, but it is now dominated by nonnative species. Upper I(lamath Lake also has a high abundance of nonnative species, and most of its native species appear to be declining. A downward trend may be common, in fact, to native fishes in most aquatic habitats in the upper I(lamath basin, although documenta- tion is weak. The overall status and biology of the fishes of the basin, except for the two endangered suckers, is poorly known or at least poorly re- corded. Research over the last 15 yr has produced many unpublished re- ports and extensive data but very few peer-reviewed papers. Thus, the
FISHES OF THE UPPER KLAMATH BASIN 213 utility of the available information is harcI to jucige. One possible remedy wouicI be to provide funding for postcloctoral scholars to compile informa- tion ancI write papers by working with university ancI agency scientists who have collectecI ciata. Future status of the suckers ancI other native fishes ancI the spreacI of nonnative species cannot be jucigecI without periodic basin-wicle survey of fishes. Monitoring is a key feature of aciaptive management (see Chapter 101. Also, most information on the biology ancI status of the suckers ancI other native fishes has not been publishecI in peer-reviewecI journals or books. Also, further studies on the systematics of I(lamath basin fishes are neeclecI so that managers can avoid being surprised by the discovery of new enciangerecI species, as are studies of the effects of nonnative species on the listecI suckers ancI other native fishes. Introductions or spreacI of nonnative species aireacly in parts of the basin are major threats to native species. The Sacramento perch in particular has the potential to spreacI through the canal system from the Lost River to Upper I(lamath Lake, where it couicI become a preciator of juvenile suckers ancI other native fishes. Populations of the two listecI sucker species in the upper I(lamath basin have cleclinecI greatly in overall abundance ancI breadth of distribution. Stable reproducing populations of the two species occur now only in Clear Lake ancI Gerber Reservoir (Gerber Reservoir has only shortnose suckers). The formerly large populations of the two suckers in Upper I(lamath Lake are cirastically reclucecI, although no quantitative estimates are available for former or present population sizes. The sucker populations showed a sub- stantial increase in recruitment, as inclicatecI by year class strength, follow- ing the encI of fishing in 1987. While the populations of Upper I(lamath Lake are reproducing ancI all age classes are present, they are not rebouncI- ing in abundance. Episodic mass mortality of large enciangerecI suckers is one explanation for failure of the populations of Upper I(lamath Lake to rebound. Other age classes may be aciversely affected in other ways, but these mechanisms are not as well clocumentecI. ProlongecI low concentra- tion of clissolvecI oxygen cluring the late summer of some years is probably the clirect cause of mass mortality in Upper I(lamath Lake. The two enciangerecI sucker species are present at other locations, but at none of these locations are substantial numbers of all age classes present. Large suckers are present in the five main-stem reservoirs of the upper I(la- math basin ancI in the upper ancI lower portions of the Lost River main stem, as well as Tule Lake, but there is no recruitment. Spawning occurs in the Lost River but cloes not sustain a population of juveniles in Tule Lake, as once was the case. Dewatering of Tule Lake ancI Lower I(lamath Lake ancI large physi- cal ancI chemical changes in the Lost River almost certainly are the cause for failure of enciangerecI suckers in the Lost River below Clear Lake ancI Gerber Reservoir to show recruitment or increase in abundance.