|
07Z7/
<br />9 TheSoulhteestem Naturalist
<br />- vol. 30, no. I
<br />length) 11 the nearest 0.1 nom and assigned to a developmental phase according to Snyder (1976).
<br />During the larval period (54 days), daily growth ranged from 0.0 to 0.7 nun. ,%Iran daily' growth
<br />w•as 0.3 mm (ranging from a mean of 0.2 for mesolaivae to 0.4 mm for protolarvae), compared to
<br />0.5 to 0.6 mail for hairlicry-reared larval humpback chub :laid bonyctil. respectively, culturred al
<br />lemper:uures ranging from 119 to 25.11°C (Ilanunan. 1982x; 1982b). Touil length of larvae at
<br />selected developmental events (Fig. 1) correlated closely with data presented by Snydcr (United
<br />States Dep. flier. Bur. Land Mgt. Bio. S(i. Set. No. 3, Denver, Colorado, 1981) for fie•Id-collecutl
<br />roundlail (.hul) inesohilvae and Invialalvae.
<br />Out- of the- major problems confronting aquatic biologists working in the Colorado River Basin
<br />is identifying young-o1'-the-year ioundmil chub, humpback chub and bonytail (Valdez and
<br />Clemmer, pp. 109-119 in Fishes of the Upper Colorado River System: present and future (W. H.
<br />Miller, H. M. Tyus, and C. A. Carlson, eds.), Am. Fish. Soc., Bethesda, Maryland, 1982). Reliable
<br />discriminating rharacu'rs must be dr(ermined before collections of larval and early juvenile Gila
<br />can be used to monitor these fish populations and :answer many questions regarding life history
<br />and habitat requlrCluC111S. We anticipate that the taxonomic research being conducted by the Colo-
<br />rado Division of Wildlife and Colorado St:rte University will contribute significantly to solving
<br />this problem. The roundtail chub developmental series produced from the present study is a vital
<br />component in this research.
<br />This research was funded under Federal-Aid matching grant 5E-3.5 administered through the
<br />Nongame Research Group, Colorado Division of Wildlife. G. Dean, M. Grode, R. Hul'zigei, M.
<br />Muth, S. Plaiania, G. Skiba, D. Snyder, R. Van Buren and E. Wick assisted in the field of
<br />laboratory.-ROBERT T. Murt, Larval Fish Lob., Dept. of Fishery and Wildlife Biol., Colorado
<br />State Univ., CHARLES M. HAYNES. Nongame Research Group, Colorado Div. Of Wildlife, Fort Col-
<br />lins, CO 80526 and CLARENCE A. CARLSON, Dept. of Fishery and Wildlife Biol., Colorado State
<br />Unit., Fort Collins, CO 80523.
<br />CHOKING OF COLORADO SQUAWFISH, PTYCHOCHE/LUS LUC/US (CYPRINIDAE),
<br />ON CHANNEL CATFISH, ICTALURUS PL'NCTATC/S (1CTALURIDAE), AS A CAUSE OF
<br />MORTALITY.-The Colorado squawfish (Plpchocheilus lu(ius), the largest cyprinid native to
<br />North America, is endemic to the Colorado River basin and historically was widely distributed in
<br />the Colorado River and its major tributaries. Following major water developments on the Colo-
<br />rado River system, this species declined and is now listed by the U.S. Fish and Wildlife Service as
<br />endangered (U.S. Fish Wildlife Service, Federal Register 39:1175. 1974).
<br />Historically the Colorado squawfish has pteyetl un soft-rayed fishes. Early in this century the
<br />channel catfish (ittalurus pun(talus) was established in the Colorado River system as a sport fish.
<br />Since that time, there were undocumented reports of Cblol'ado squawfish found dead with channel
<br />catfish lodged in their esophagus (Seeth:ler, M.S. Thesis, Utah State Universi(y. Logan, 1978;
<br />Vani(ek :laid Kramer, -Trans. Arner. Fish. So(., 98:193-208, 1969). Until now, the onlydot'unu•nted
<br />evidenrt• of this phenomenon was re•pnrted by M, Ada (Soulhwesic•rn Nat. 29:119-120. 1983) who
<br />collected :la1 adult Colorado squawfish (550 min 'I-L) with a channel catfish (120 min 'I'L) lodged
<br />head first in its esophagus, posterior to the gill ;aches.
<br />Some concern exists that the lodging of juvenile (flannel catfish in throats of Colorado squaw-
<br />fish could be one of the causes of the decline in numbers of this endangered species. Hence, a
<br />laboratory experiment was designed to study the phenomenon in more detail. Objectives of the
<br />study were to 1) determine if adult Colorado squawfish would prey on channel catfish without
<br />prior conditioning, 2) determine if channel catfish could be swallowed or tend to lodge in the
<br />esophagus, and 3) determine if Colorado squawfish are injured in the process of trying to eat
<br />channel catfish. Since the completion of this study, three more cases of Colorado squawfish chok-
<br />ing on channel catfish were reported from the Green River in Utah. This previously unpublished
<br />information is included here and contrasted with the laboratory study.
<br />Colorado squawfish used in the laboratory experiment were hatched in 1974 and reared at the
<br />Willow Beach National Fish Hatchery. Sixty individuals ranging in length from 360 to 512 min
<br />TL (x = 994 mm, SD=29 mm) were used in these tests. Fish were fed pelleted commercial trout
<br />feed or rainbow trout (Salmo gairdneri) on alternating days. Channel catfish were hatched in Jun(-
<br />1982 and reared at a commercial hatchery in southern Idaho. Lengths ranged from 51 to 125 mm
<br />TL (K = 77, SD = 10.5, n=600) at the time of testing. Both species were held at 15 t 1 C and it
<br />photoperiod of 14L:IOD in Logan City dechlorinated water.
<br />March 1985 Notes 155
<br />The chamber where feediung trials took place was a covered flow-through circular tank 1.5 in
<br />in diameter and 30.5 cm deep. Water temperature was 15 t IC, photopoeriod was HLAOD. Colo-
<br />rado squawfish seemed to lack sufficient motivation to feed when tested singularly and were there-
<br />fore tested in groups of three. In preliminary tests. they also appeared to lack inivicsi in channel
<br />catfish as prey items; hence, only those- fish that would eat catfish with their spines removed
<br />(CCSPL.) were exposed to catfish with their spines inmrt (CCSP). The experimental protocol was
<br />as follows: 'I'hc Colorado sgtlawlish to be used in it particular feeding trial was suu'vcd fur 5 days
<br />before ilia- trial. On the morning of the trial they were ransfe•rred to the experimental aquarium
<br />and allowed to recover from handling stress and acclimate for 6-7 hr. The 40 catfish to be used in
<br />the feeding trial were tranquilized with tricane methane sulfonate and quinalidine and measured.
<br />On the morning of the vial, dorsal and pectoral spines were clipped and fish were placed in a
<br />recovery aquarium for 5.6 hr. This method of spine removal evidently has no apparent effect on
<br />the prey (Reis(, Can. J. Zool., 58:1245.1252, 1980), The CCSPL were released in the experimental
<br />aquarium with the Colorado squawfish in the later afternoon at a time when squawfish are nor-
<br />mally fed. They were observed for the first 0.5 hr and then left with the squawfish for 90 hr. The
<br />catfish here fell during this period, but the Colorado squawfish were not. On the morning of the
<br />fifth day the Colorado squawfish were removed from the aquarium, after which the catfish were
<br />tranquilized, renleasured, and then returned to the recovery aquarium. Colorado squawfish used in
<br />the trial were placed back in the experimental aquarium and left alone until the afternoon. If any
<br />CCSPL hall been eaten, then 40 catfish with spines intact (CCSP) were placed in the aquarium.
<br />Fish were ubscrved for 0.5 hr afar introduction and left together for another 90 hr. After feeding
<br />trials, the Colorado squawfish were tranquilized, measured, and then segregated in three groups
<br />for long-term observation: I) fish chat had not eaten CCSPL and were not exposed to CCSP, 2)
<br />fish that had eaten CGSPL but not CCSP, and 3) squawfish that had eaten both CCSPL and
<br />CCSP. Tile Colorado squawfish were then monitored for 2 months following their exposure to
<br />catfish to check for problems arising from feeding on CCSP. Due to the extreme values of these
<br />endangered fish, none here sacrificed to check for problems arising from feeding CCSP.
<br />Fifteen trials were performed with Colorado squawfish and channel catfish ((Table 1). Two
<br />trials (9 and 10) were conducted with rainbow- trout or cutthroat trout (.Salmo Clarki) to establish
<br />a baseline. Squawfish here only used once on trials except for 3 individuals used in a preliminary
<br />and then in a liner vial.
<br />Colorado squawfish ate considerably fewer spineless channel catfish CCSPL than trout. During
<br />the 0.5-hr after inroduction Colorado squawfish attempted to eat CCSPL 60 times. Each CCSPL.
<br />was inlmedi:lady rajccutl. while rout were consumed on the first try. CCSPL eventually were
<br />eaten, but old) after several days of exposure. Six attempts to eat CCSP resulted in violent thrash-
<br />ing of the head by the squawfish. One squawfish jumped into the air and thrashed its head in its
<br />attempt to expel the catfish. 1'hc longest struggle lasted 5 nninmes anti resulted in the squawfish
<br />.'mouthing., the watt-[' for seve•rnl nninuu•s after releasing the CCSP. No evident injuries resulted
<br />and the squawfish fed on minbutw unlit upon return to tic holding aquarium.
<br />In 15 individual feeding trials, 29 Colorado squawfish were exposed to 320 CCSPL for the ruin
<br />of 720 hr which rrathed in 6% of the CCSPL being eaten (Table 2). Of the 24 Colorado squawfish
<br />exposed co CCSPL, 21 were exposed to 280 CCSP: only 1% (2 individuals) of the CCSP were eaten
<br />in trials (630 hr). Mean lengths of rainbow (84mm) and cutthroat trout (90 mm) used in this
<br />experiment were signifi(antly longer (Table 1) than the channel catfish (77 mm, P = 0.01); how-
<br />ever, ranges in hod)- length were similar. There were no significant differences in the size of fish
<br />eaten versus fish not eaten for either trout, CCSPL, or CCSP.
<br />During the 2-month observation period after testing, Colorado squawfish held in all three cate-
<br />gories fed normally on rainbow trout and pellets. There was no noticeable difference in behavior
<br />or appetite between squawfish that ate CCSP and those that did not, and no disease was evident in
<br />any fish.
<br />The phenomenon of fish predator's attempting to feed on spined prey is well studied but pri-
<br />marily from the perspective of the prey (Hoogland, Morris and Tinbergen, Behavior. 10:205.237,
<br />1957; Beyerle and Williams, Trans. Amer. Fish. Soc., 97:28.31, 1968; Mauck and Coble, J. Fish.
<br />Res. Board Can., 28:957.969, 1971; Moodie, McPhail and Hagen, Behavior, 47:95.105, 1973; Reis[,
<br />Can. J. Zool., 58:1245.1252, 1980). Ivlev (Experimental ecology of the feeding of fishes, Yale Univ.
<br />Press, 1961) defined prey vulnerability to predation as having two components, prey availability
<br />and predator preference. Prey availability is influenced by several factors such as schooling behav-
<br />ior, cryptic appearance, distribution, and armament. Channel catfish certainly have formidable
<br />armament and when this was reduced by clipping spines in the laboratory study, they were con-
|