North American Journal uf Management 16:560-569. 1996 <*•• Copyrigh e Americath y b t n Fisheries Society 1996

Effect of Electrofishing Pulse Shape and Electrofishing-Induced Spinal Injur Long-Tern yo m Growtd han Survival of Wild STEVE . DALBEYNR THOMAD AN ' . MCMAHOE S N Biology Department, FishWildlifeand Program Montana State University. Bozeman. Montana 59717. USA

WADE FREDENBERG2 Montana Fish, Wildlife, Parksand 1400 South 19th Street. Bozeman. Montana 597/5. USA

Abstract. — High rate f spinaso l injury from electrofishing have been reporte r rainbodfo w trout Oncorhynchus mykiss. but little is known about the fate and performance of injured fish. We conducted a long-term experiment to evaluate how incidence and severity of electrofishing-induced spinal injury affects growth and survival. We eleciroshocked 866 wild rainbow trout from the Gallatin River, Montana, using one of three different DC pulse shapes (smooth, half pulse, and full pulse). X-rayed the fish to determine degree of spinal injury, and compared short-term (100-d) growt long-terd han m (335-d) growt survivad han l after transplantin e fisgth h to0.6-ha a pond (stocking biomass 5 kg/ha)25 , . Rainbow trout shocked wit hsignificantld pulseha C dD y higher (P = 0.0001) incidence (40-54%) of spinal injury than those shocked with smooth DC (12%); injuries were also more severe among fish captured with pulsed DC (P < 0.01). Incidence and severity of injury were positively correlated with fish length (r = 0.79-0.83. P < 0.02). Few surviving fish (7 of 418) at the end of the study exhibited outwardly visible spinal deformities. Healin f spinago l injurie s readilwa s y apparen radiographn o t s evidencesa calcificatioy db d nan fusion of damaged vertebrae. Long-term survival of rainbow trout was not affected by pulse shape used during capture or by seventy of electrofishing-induced injury. However, fish with moderate to severe injury (spinal misalignment and fracture), representing 28% of the total number shocked, had markedly lower growth and condition after 335 d than fish with no or low spinal injury.

Electrofishing has long been viewed as a rela- phy trout waters in the state (Holmes et al. 1990). lively harmless fish-sampling method whe3 wil69 nd f userainboO d w trout shocke Montann i d a appropriately. However, recent report f signifio s - rivers (201-554 mm), 50-70% sustained spina- in l cant skeletal damage and hemorrhaging from jury resulting in 769 hemorrhages and 2,647 in- shocking with standard equipmen d settingan t s jured vertebrae (Fredenberg 1992). Pulse shape have caused biologists to question the assumption and frequency are among the variables that influ- that its effects on fish health are minimal (Snyder ence incidence and severity of injury (Sharber and 1992, 1995). Carothers 1988; McMichael 1993; Sharbe. al t e r Rainbow trout Oncorhynchus mykiss appear par- 1994). Lowest injury rates in rainbow trout have ticularly susceptible to injury during shocking been observed with smooth DC and highest rates (Reynolds and Kolz 1988; Snyder 1992). Sharber with pulsed DC at high (>60 HZ) frequency (Fre-

Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 and Caiolhers( 1988) found spinal injurie% denber44 n i s g 1992; McMichael 1993; Sharbe. al t e r f largo e% t(>30o67 form 0m k length) rainbow 1994) trout electrofished with pulsed direct current (DC). Despite reports of high injury rates -n rainbow Wild rainbow trout larger mtha0 mn40 shocke d trout and Qther salmonids, the fate of injured wild with Alask n pulsei C ad D ha d4a 1 % spinal injury fish Qver the ,ong term and ^ potentia, lation rate and 14% short-term (96 h) mortality, thus effects of shocki remain , , unknown (S prompting a moratorium on electrofish.ng in tro- def ,992) Immediale mortality after electrofishing s usualli y lowexternad an , l sign f injuro s e yar ____ often absent or undetected (Hudy 1985; Mc- P re l ad reSS M l Fis 1 Wildlif ? A ' c f " ^ D °" ri ! ; *' *"? ^ Michael 1993; Hollender and Carline 1994; Shar- 475 Fish Hatchery Road, Libhy, Montana 59923, USA. . , tnf.. . ,. . ,^^ ... 2 Present address: U.S. Fish and Wildlife Service, 780 ber et aL I994; but see Hauck 1949' Holmes et al« Creston Hatchery Road, Kalispell, Montana 59901, 1990). Some investigators have inferred tha- in t USA. juries often heal and long-term mortality is low, 560 ELECTROFISHING EFFECTS ON RAINBOW TROUT 561

but pronounced spinal deformities after electro- system (Vincent 1971) anode 25.4-ca s Th . ewa m have been reported in other studies (see aluminum triangle attached to electrical cable, and Snyder 1992). To date, most studies that have ex- the cathode was a 1.2-m2 stainless steel plate at- amined effect f electrofishino s growtn go surd han - voltageC tacheD boate e th o Th .sdt (rectifie) dAC vival in trout (Hudy 1985; Gatz et al. 1986; Taube were adjuste provido f currentt do A 3 ea leve , l 1992; McMichael 1993; Dwyer and White 1995) use norman di l electrofishing operation- re e th n si have been conducted over short time intervals gio produco nt e electrotaxi higd an sh captur- eef (<35 d) with hatchery fish (e.g., Hudy 1985; ficienc f trouo y t (Fredenberg 1992) portablA . e McMichael 1993; Dwyer and White 1995; but see oscilloscop s use measuro ewa dt e output voltage Gat/ et al. 1986), and, except for the study by ane voltagth d e gradient (V/cm) surroundine gth Taube (1992), they hav distinguishet eno d between anodfielde th n .ei Voltages varie0 40 do t fro0 m35 injure uninjured dan d fishr stud designes Ou . ywa d r half-pulsfo 0 40 r smoot o eVfo t , fro0 hDC m20 (1o t ) characterize short-term (100-d) growtd han DC, and from 0 to 400 V for full-pulse DC (Figure long-term (335-d) growt d survivaan h f wilo l d I) s delivere. Pulsewa half-sina C dD s da e wave rainbow trout shocke df thre o puls witC e eD hon e . Voltag aa frequenctHz 0 6 e f gradiento y e ar s shapes (smooth, half pulse d fulan l, pulse)) (2 ; shown in Figure 2. compare growth and survival of injured and un- e thre pulsC Th D e e shapes were used sequen- injured shocke ) evaluat (3 e fieldd d th fis n an ;i h e tially (smooth, half pulse d fulan ,l pulse) until healing of electrofishing-induced spinal injuries. about 330 fish had been collected for each pulse shape test group. Only one pulse shape was used Methods sectioa timea t d an , n 100-1,00 lon0m g sepa- Experimental site.—Growth and survival of rated reaches shocked with different pulse shapes electroshocked rainbow trout were evaluaten a n di to ensure that fish were shocked with only one irrigation storage pond on the Montana State Uni- pulse shape. Fish were hel livn di ee wellth n o s versity campus e 0.6-hTh . a maximua pon d dha m electrofishing boat, then transported in oxygenated unifora mdept5 d ,4. meaf an h o , m m n dept5 3. f ho tank separato st e outdoor hatchery racewaye th t sa mugraved an d l bottom. Summer water tempera- Bozeman Fish Technology Center. , fisAfteh h4 2 r tures were typically less than 20°C (range- 10 , from each test group were move indooo dt r tanks 26°C ranged wintern i an ) C d5° . betweed an 2 1. n for processing. Fish were anesthetized with Dissolved oxygen concentration ranged fro0 m8. MS-222 (tricaine methanesulfonate), measured for 11.o t 5 mg/L diversioy . b Inflopone s th dwa o wt n fork length, weighed, and tagged with both an un- from a nearby creek and outflow was through an coded wire tag in the snout or base of left pelvic irrigation pump. Fish movement in and out of the r fi dorsaidentifo o nt n fi l y pulse shape test group pons preventewa d screeney db d inflo outd wan - and an uniquely coded visible implant (VI) tag in flow pipes. Quafitative plankton pone towth dn si postorbitae th l adipose tissu r individuaefo l iden- showed that abundant trout food—Gammarus, tification. Fish were X-rayed (right lateral view) Daphnia, Diaptomus—d an was present. Before fish in groups of 10 with a Minxray X750G portable introduction, a few (N = 16) brook trout Salvelinus X-ray machine with Konica Blue film exposed for fontinalis were removed with gill nets. No source-to-imaga othe d an r V k 0.6-0. 0 5 t a 8es distance fish species were present 3-m-higA . h fence sur- of 760 mm. Trout ranged from 153 to 388 mm fork rounding the pond prevented angler access. length. Length distributions (mean ± SE) of fish Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 Fish collection.—Wild rainbow trout were col- were similar among the three test groups: smooth lected fro ma 14-k m e Gallatireacth f ho n River = 252. ; hal± 616. f3 mm puls 4 244.= e 2. ± 5 downstream from Big Sky, Montana, on July 27 mm; and full pulse = 245.4 ± 16.3 mm (Koi- and 28, 1992, at water temperatures between 13 mogorov-Smirnov testr 1984 0.22-0.33= Za , ;D , and 16°C and conductivity of 260 u,S/cm. Fish in P > 0.6). thi likelihoow s lo sectio a d f previoudno ha - ex s Survival and growth.—After processing, 866 posure to electroshock. Limited shocking had oc- fish (smooth DC, N = 241; half-pulse DC, N = e sectioth f o n 1989 ni curreuppee m th k ,n 6 i dr 309; 316= full-puls N ) , werDC ee transported and the last extensive population estimate was in fro hatchere mth pone releasedd th dan o yt 3 1- , 1984. Fish were captured by single-pass electro- d after collection, on July 29-30, 1992, at an initial fishing from a 3.8-m fiberglass drift boat equipped stocking biomass of 255 kg/ha. Fewer fish from wit a 3,000-Wh , 220- generatorC VA a locall, y the smooth DC test group were released because manufactured electrofisher, and a mobile electrode some had escaped from the outdoor raceway. The 562 DALBEY KT AL.

SmootC hD remove l remainindal g fish fro pone m th treaty db - ing it with 5% liquid rotenone on July 1, 1993 (33 5postrelease)d . Rotenon s uniformlewa y dis- tribute a dri a pvi dhos e placee propelleth n i d r n outboarwasa f o h d motor. Most mortalit- oc y curre e continuew t d bu withi . dd monitorin3 n g afted pone 7 r th applicationdfo believe W . e total removal was achieved because surface feeding by trout, commonly seen before application, was not observe wee1 r kfo d after application e colTh .- 0 5 0 4 0 3 0 2 0 1 O lected fish were weighed, measured identified an , d by tag. Long-term survival (S) was determined by 8001 Half-pufse DC numbee th wher . s G rwa + eintroduced/ R - I S= , /? the number removed from the pond at the end 4OO 0) of the study, and G the number removed to assess O) short-term growth. CO 90° We use da log-likelihoo d contingency analysis (G-test, Zar 1984) to compare long-term survival amon e threth g e pulse-shape groups. Short-term d long-teran m change growtn i s d conditioan h n were compared among pulse-shape groups by anal-

10 20 90 40 SO ysi f varianco s e (ANOVA Tukey'd an ) s multiple comparison test (Zar 1984). Following Tripped an l Full-pulse DC Hubert (1990) usee w , d analysi f covarianco s o et adjus r effectfo t f fiso s h lengt growthn ho . Spinal injury.—We assessed spinal injur raty yb - ing incidenc severitd ean f injuryo y accordino gt criteria propose . ReynoldJ y b d s (Alaska Coop- erative Fish and Wildlife Research Unit, personal communication e alsse ;o Hollende d Carlinan r e 1994) (Table r eacI).Fo h e fish,e ratingth th r fo most serious injury observed was designated as e overalth l injury rating. X-ray plates from dif- 4O 9O ferent treatments were shuffled and read randomly eliminato t e bias. Spinal abnormalities presen- be t Time (ms) fore electroshocking were readily distinguished from recent injury by the evidence of calcification FIGURE 1.—Electrical pulse shapes (measure- os y db (Sharber and Carothers 1988; Fredenberg 1992), cilloscope) generated by electrofisher at smooth, half- pulse full-pulsd an , settingsC D e . and these were eliminated from our injury rating; such abnormalities were present in only 1% of fish

Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 X-rayed e assesseW . d healin f spinago l injurt a y pons visuallwa d y inspecte r deafo d d fish twice the end of the study by X-raying 38 randomly se- daily for 3 months postrelease and three to five lected fish having various degree f initiaso l spinal time weea s k thereafter during ice-free periods. injur ratind an ye X-ray gth f thesso e fish without To measure short-term growth among test knowledge of initial injury. We then determined groups, we removed about 10% of the fish stocked degre f healineo y comparinb g g initia d finaan l l from each group at 100 d postrelease (7 November) X-rays. using fyke nets and short-term gill-net sets. Sam- We compared incidenc severitd an e f spinayo l pled fish were weighed, measured d identifiean , d injury among pulse-shape treatment groups with a by tag. Change lengthn i s , weight conditiod an , n C-test. Effect f injuro s y severit n growto y d an h (K = 105 x [weight. g]/[length, mm]3) were de- condition were analyzed by ANOVA. Associations termined for fish that retained the VI tag. between fish length and injury severity were as- To assess long-term survival and growth, we sesse simply db e linear correlation (Zar 1984). ELECTROFISHING EFFECTS ON RAINBOW TROUT 563 10-1 -°- SmootC hD -*- Half-pulse DC -*• Full-pulsC eD

O 6O O 5O O 4O O 3O O 2O O 1O O Distance from Electrode (cm) FIGURE 2.—Peak voltage gradients (V/cm; measure oscilloscopey db ) surroundin three anode th C gth r D e efo pulse shape sstudye useth n di .

Results groupP F M d S san e thath P H n i n curre e th n i d Spinal Injury group. Similarly, fish in the FP group had a sig- nificantly higher proportion of class 1 and 2 in- X-rays were of good quality with greater than jurie groupP 0.05< H P s( e ). th tha Howeverd ndi , 6 readable86 e th f . o Initia % 99 l spinal injury dif- e proportioth f fisno h wit e moshth t severe injury fered significantly among the three pulse-shape (class 3) did not differ significantly among pulse- group botn i s h incidence = 115.7(Tabl C ; P ,2 e shape groups (P = 0.13); injury ranged from 6% 0.0001)= severitd an y ( G 49.8= 0.0001) = P , . group P grouP H F d e 10po st th an (Tabl %n i M inS e Rainbow trout shocked with half-pulse (HP) and . Overall 2) f rainboo % 9 , w trout shocked sustained full-pulse (FP) DC waveforms had about three to class 1 injury, 20% class 2, and 8% class 3. four times higher proportio f spinano l injuries than Incidence and severity of injury varied with fish those shocked with smooth (SM (HPC D ) , 40%; length (Figure overalTh . e3) l incidenc f injureo y Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 FP, 54%, SM, 12%). A significantly higher pro -s positivelwa y correlated with fish lengt r ( =h portion of class I and 2 injuries (P < 0.01) oc- 0.83 0.01)= proportioe P , Th . f fisno h with class 3 injury also was positively correlated with length TABLE I.—Criteria used to rate severity of electrofish- 0.79(= r0.02)= P , , wherea proportioe th s f no ing injur rainboo yt w trout. Injuries were determined from fish with clas injur1 s s negativelwa y y correlated examination of radiographs. with length (r = -0.92, P < 0.001). The propor- tio f rainbono w trout with clas injur2 s y showed Injury class Spinal damage no significant association with length across all length classe 0.46 = 0.9) = r s( P , ; however, pro- 0 None apparent 1 Compressio f vertebrano e portio f clasno injurs2 positivels wa y y correlated 2 Misalignment and compression of vertebrae with lengt r 0.98= fis fo hr h( ,lesm sm tha0 25 n 3 Fracture of one or more vertebrae or complete sep- 0.02)P= . aration of two or more vertebrae f 41 studye O 8th e fisf o w , h d captureen e th t da 564 DALBEY ET AL.

TABLE 2.—Incidence and severity of spinal injury in rainbow trout collected by electrofisher from the Gallatin River, Montana. Three DC pulse shapes were used. Number and percentage (in parentheses) of fish in each injury class (Table I) for each pulse-shape group are shown for all fish at the beginning of the study and for all fish individually identifiable (i.e.. retained VI tag) at the end of the study (335 d).

Beginning End Number (%) by injury class: Number (% injury b ) y class: Waveform N 0 1 2 3 N 0 1 2 3 SmootC hD 241 212 5 9 15 54 47 1 3 3 (88) (2) (4) (6) (87) (2) (5.5) (5.5) Half-pulse DC 309 186 31 74 18 57 35 7 14 1 (60) (10) (24) (6) (61) (12) (25) (2) Full-pulsC eD 316 145 44 95 32 103 48 17 28 10 (46) (14) (30) (10) (47) 06) (27) (10)

observed outwardly visible spinal deformities in from 16 to 68% (Table 3). Proportions of fish clas- 0 SM, 2 HP, and 5 FP fish. Healing (calcification) sifie eacn di h injury beginnine clasth d t sa en d gan f injureo d vertebra readils ewa y apparent fro- mra of the study were significantly different (P < diographs take ninjure8 3 fro e mth d fish choset na 0.001). For the 38 fish examined, the spinal injury random and X-rayed at the end of the study. Cal- rating increase severitn di r 60%yfo , remaine- dun cification made judging severit f initiayo l injury change r 34% decreased fo . an , 5% r dfo difficult but also made spinal damage from elec- troshocking more visible. Damaged vertebrae, par- Survival ticularly those with hairline fractures (class 3), During the period of electroshocking, transfer- wer t alwayno e s detecte n initiai d l radiographs. hatchere rinth o gt y holding facility. X-rayingd an , proportioe Th f fisno h with clas injurs3 y increased transporting to the pond, only 1 fish (in the HP

D Class 1

50- £2 Class2 Clas• s 3

40- (26)

(61) — 3OH "c (186) (180) CD (22)

Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 y CD 20 (147) (131)

(113)

1O m. I 15O 175 2OO 225 25O 275 3OO 325 Length Group (mm) FIGURE 3.—Severit f spinayo l injur rainbon i y w trou length-grouy b t 866)= N p( . Values for lowee mth r bounds of each 25-mm length-group. The sample size for each group is shown in parentheses. ELECTROFISHING EFFECT RAINBON SO W TROUT 565

TABLE 3.—Spinal injury ratin8 injure3 f go d rainbow 60% = 0.3 0.8)= = P F P ,;G . Injury severity also trout X-raye t tha e d eth beginnin f d (33o en ) 5d d an g appareno n d ha t influenc n long-tero e m survival. study numbee Th . proportiod ran parenthesesn n(i f fis)o h Among the survivors that retained the VI tag (215 are shown for each injury class. or 51%), the proportions in each injury class were Number (% injury b ) y class similar within each pulse-shap- be e e grouth t pa Study time 1 2 3 ginning and end of the study (Table 2; G = 0.3- 0.5, P = 0.5-0.9). The overall proportion of in- Beginning 8 24 6 (21) (63) (16) jured and uninjured fish within each pulse-shape End 0 12 26 e studs th f wa y o groue beginnin th d t en pa d gan (0) (32) (68) also similar (G = 0.02-0.5, P = 0.8-0.9).

Growth group) died. Of the 866 fish stocked in the pond, 115 were remove e short-terth r dfo m growtd han Measuremen f short-tero t m growt condid han - condition assessment, 21 were recovere s deada d e tioth f no chang % s limite 54 fis2 ewa r 6 o h o dt fish, and 418 were recovered after the rotenone sample (N = 115) becaus remaindee eth losd ha rt treatment unaccountee th ; fis2 h31 werr dfo e con- the VI tag. Average changes in length, weight, and sidere n siti d u mortalities because screeninn go condition over this period were not significantly inflo outflod wan w pipes prevented fish emigration different among the three pulse-shape groups (Ta- from the pond. . Fis4) he frobl l thremal e waveform groups lost Short-term mortalit postreleased 7 y( w lo s wa ) condition contrastn I . , chang lengthn i e , weight, among all test groups (FP = 1, HP = 2, and SM d conditioan n showed significant differences = 12). The higher short-term mortality in SM among injury classes, although comparisons were group fish was probably due to the increased stress based on small sample sizes for injury classes 1 incurred when escaped fish were recaptured from and 3. Uninjured (class 0) fish increased signifi- the outdoor raceway. Of the 21 dead fish recovered cantly mor lengtn ei weighd h an (19.) t 5(28.mm 8 fro e ponmth d durin e studyth g (71%5 1 , ) were g) than injured fish (means: 6.1 to 11.5 mm, - 19.5 recovered within the first week following intro- o 12.t . Conditio7g) n decline fisn di h- froin l mal duction. jury classes after 100 d, but the decline was sig- Long-term surviva s similal (33wa ) 5d r among nificantly greater for injured fish. all pulse-shape groups (SM = 54%, HP = 58%, Unlik e 100-th e d comparison, average weight

TABLE 4.—Mean change lengthn i s , weight d conditionan , f electroshockeo 8 d rainbod 5 w 33 trou d an t afte0 10 r classified by DC pulse shape (SM = smooth, HP = half pulse, and FP = full pulse) or severity of spinal injury (Table 1). Standard deviation f meanparenthesessn o i e totae sar th ls i numbe N ; samplee indicateth f P fis ro n d i h an , s results of analysis of variance tests. Within pulse-shape and injury class groups, values along a row without a letter in common are significantly different (P ^ 0.10).

Pulse shape Injury class Variable SM HP FP P 0 1 2 3 P Time interval: 100 d

Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 Length (mm) 17.0 / 7.8 z 14.2 z 0.14 19.5 z II.5y 6.1 y 9.3 y 0.003 (14.2) (8.0) (13.2) (13.4) (6.5) (7.4) (18.3) Weight (g) 22.6 z 8.8 z I3.4z 0.41 28.8 z I2.7y 1.5 y -19.5y

60-

50-

40

160 180 200 220 24O 26O 28O 3OO Initial Length (mm) FIGURE 4.—Length-growth regressions for rainbow trout having low (classes 0 and I) and high (classes 2 and 3) electron"shing injury.

and length changes differed among pulse-shape regressio groupo tw ne sline th alsr osfo were sig- groups at the end of the study (335 d; Table 4). nificantly different (Figure 4; analysis of covari- Average weight gain of HP group fish (47.7 g) was ance; / = 2.7, P = 0.01 for slope). The adjusted significantly greate (30.P rF thad 2(26.M an n S ) 6g mean length gaiinjur w r fislo fo nh r yo wit o n h g) fish. The HP fish also had significantly greater (17.1 mm) was nearly 2.5 times greater than the mean length increase (19.9 mm) than SM fish (9.4 adjusted mean length gain of fish with moderate P fisF h mm d (13.an ) 2 mm). Condition factor higo t h injury (7.1 mm). changes were similar among pulse-shape groups, wit l grouphal s showin increasn ga conditionn ei , Discussion unlik decrease eth e observe t 10da . 0d Spinal Injury Long-term growth and condition differed mark- edly with degre f injuryeo . Classe d I an fis 0 sh Overall (32% 866f 237 ,o rainbof )o w trout cap- gained an average of 36.9 to 44.3 g over the 335-d tured in our study sustained spinal injury during period, compare r averagn a fo o dt g e5 4. los f o s electroshocking. This proportion is comparable to Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 class 3 fish (Table 4). The mean weight gain of the incidences of electrofishing-induced injury classes 0 and I fish was more than twice that of among wild rainbow trout reported by Sharber and fish with class 2 injuries (15.4 g). Mean condition Carothers (1988), Holme t ale s . (1990), Freden- change of class 0 (uninjured) fish was significantly berg (1992) d Sharbe . an (1994), al t e rthein I . r greater than tha f injureo t d fish (classed san , I2 , summar f electrofishinyo g injury studies, Hollen- 3). Condition increased in class 0, I, and 2 fish der and Carline (1994) reported that an average of but declined in class 3 fish. 33% of wild trout sustained spinal injury during Fish with no or low injury grew 5-10 times more shocking. These studies encompasse widda e range in length (means, clas = 17.1 s0 , clas s = 22.I 3 of electrofishing equipment, settings, and environ- mm) than fish with moderate to severe injury mental conditions. (means, clas =s2 4.8, clas = s3 (Tabl) 1.5mm e Pulsed-DC waveforms caused three to four 4). Although growth was negatively correlated times higher incidence of injury to rainbow trout with initial length for both groups, length-growth than did smooth DC in our study. A strong positive ELECTROFISHING EFFECT RAINBON SO W TROUT 567

relation between injury rate and pulse frequency prevalent than damage to the spinal column (Fre- bees ha n observed consistentl studien yi f elecso - denberg 1992; McMichael 1993). trofishing injury (Sharber and Carothers 1988; Fre- denberg 1992; McMichael 1993; Sharber et al. Effects Survivalon Growthand 1994). Type of pulse shape used in our study also There was no evidence that pulse shape or de- affected severity of injury. Rainbow trout shocked gree of spinal injury adversely affected long-term with half-puls a signifid full-puls d an e ha - C D e surviva f wilo l d rainbow trout, despite significant cantly greater proportion of spinal compression difference initian i s l injury. Indeed proportioe th , n (clas injury1 s d spinaan ) l misalignment (clas2 s f injureo d fish within each injury categors wa y injury) than those shocked with smooth DC (Table remarkably similar at the start and end of the study 2). In contrast, the proportion of spinal fracture range (Tabl Th f mortalit. eo e2) f rainboyo w trout (clas injury3 s s relativelwa ) y constant (6-10%) among pulse-shape groups in our study (40-46%) amon three gth e pulse-shape groups. Similar levels was comparable to annual mortality of age-2 and of clas injur3 s y (5-10%) among different pulse older wild rainbow trou Montann i t a rivers (about shapes were also note Fredenbery db g (1992d an ) 50%; Vincent 1987). Taube (1992) also founo n d Hollender and Carline (1994). The causes for this differenc n long-teri e m survival (20 ) amon3d g consistent level of class 3 injury during electro- injury classe n hatcheri s y rainbow trout after shocking need further investigation. Reynoldd san shocking. Kolz (1988) suggested that rainbow trout are se- Electrofishing injury, however d havdi ,a e verely injured when fish are shocked within the marked influence on growth. Rainbow trout with area of high voltage that occurs near the anode at moderat severd ean e spinal injury (classed an 2 s all waveforms (Figure 2). Alternatively, Sharber showe) 3 d increaslittlo n r eo lengtn ei weighd han t et al. (1994) postulated that "injury-causing sei- compared ove5 33 r uninjureo dt r slightlo d - yin zures are relatively independent of the strength of jured fish which average dgaia f 17.no 22.o t 1 3 the electrical stimulus" and are more a function m lengtmn i 36.d h44.an 9o weight t n i 3g . Effects were especially pronounce mose th r t dseverelfo y of duratio f exposureno . injured fish (class 3), which lost weight and con- findingr Ou s support earlier work demonstrating dition during the study. Thus, negative effects of a positive relation between fish size and frequency electrofishing injury on growth and condition re- of electrofishing injury (Lamarque 1990; Hollen- maine t leasa r yea1 tdfo r after injury. This con- der and Carline 1994). For example, we found that trasts wit rapie hth d recovery, measure hourn di s about 40% of trout longer than 200 mm sustained or days, from physiological and behavioral effects injury compared to about 27% in smaller fish (Fig- of electroshock (Horak and Klein 1967; Mesa and ure 3). We also found that longer fish sustained Schreck 1989; Mitto McDonald an n d 1994)e W . more severe injury. Previous studies of electro- found no evidence for negative effects of electro- fishing injur wiln i y d rainbow trout have focused fishing pulse shape on long-term growth and con- on injury to fish exceeding 300 mm (Sharber and dition. These findings underscor- in neee r eth dfo Carothers 1988; Holme . 1990al t se ; Sharbe. al t re vestigator consideo st r severit f injuryo y whe- nas 1994). Our study demonstrated that spinal injuries sessing incidence and effects of electrofishing in- are detectable in wild rainbow trout at least as jury cause differeny db t pulse shapes. smal , althougs 15a l 0mm h damag smalleo et r fish shoule On d interpre resultr ou t s with severa- ca l Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 is more difficult to detect (Fredenberg 1992). veat mindn si . First suspece w , t dynamithaa n i t c Healing of spinal injury was readily apparent in stream environment, skeletal damage could pos- radiographs of injured fish about 1 year after sibly have greater negative effects on growth and shocking. Injury sites showed significant calcifi- survival than it did in our study conducted in a cation, including apparent fusion of vertebrae. relatively benign pond environment, becausf o e f 418Ver o fis w 7 ( hy)fe exhibited outwardly vis- higher physiological demands require r capfo d - ible spinal deformities. Our comparison of initial turing food and maintaining position in flowing and final injury status, though based on a limited water (Horak and Klein 1967). Second, because number of fish (N = 38), further suggests that in- our study lacked an unshocked control group, we cidence and severity of spinal injury from elec- were unabl o determint e e overalth e l effectf o s e troshocunderestimateb y ma k d with standard electrofishin survivan o g growthd an l hencd an , e X-ray equipment. In addition, radiographs do not coul t definitiveldno y determine whether potential detect soft tissue damage, which may be more negative effects from electrofishing were additive 568 DALBEY ET AL.

or compensatory. This may have been an important trout and juvenile Arctic grayling and cutthroat limitation because studies have detected signifi- trout. North American Journa f Fisherieo l s Man- cant difference growtn si h among shocke- un d dan agement 15:148-151. Fredenberg 1992W. , . Electrofishing-induced spinain- l shocked groups of hatchery rainbow trout (Taube juries caused by field electrofishing surveys in Mon- 1992: Dwye d Whitan r e 1995 d wilan )d brown tana. Montana Fish, Wildlife d Parksan , , Report. trout Salmo trutta rainbod an w trout (Gat t alze . Helena. 1986). Obtaining an unshocked control group of Cada. GatzF . J.. . LoarG ,A . , Jr.M d 1986. J ,,an . Effects wild rainbow trousiza f eo t comparabl thao et f o t of repeated electroshockin n o instantaneoug s the shocked test group s beyonwa sr logistidou c growt f troutho . North American Journa f Fisho l - eries Management 6:176-182. capabilities. Nonetheless, by classifying fish by Hauck . 1949R . F , . Some harmful effectelece th f -o s injury severity e coulw , d compare survivad an l troshocker on large rainbow trout. Transactions of growt f uninjureho d injurean d d fish. the American Fisheries Society 77:61-64. Hollender Carline. F . R . A.d .B , an , 1994. Injur wilo yt d Management Implications brook trou y backpacb t k electrofishing. North Our results from the first long-term study of American Journal of 14: effects of electrofishing-induced injury to wild 643-649. rainbow trout lend additional suppor manageo t t - Holmes . McBride.N R.. D , . . ReynViavantT B , . J -d an . ment recommendation r usinfo s g smootr o C D h olds. 1990. Electrofishing induced mortalitd yan injur rainboo t y w trout, Arctic grayling, humpback low pulse rate o reduct s e injury during electro- whitefish, least cisc northerd oan n pike. Alask- aDe fishing, especially among larger fish (Fredenberg partment of Fish and Game, Division of Sport Fish, 1992; Snyder 1992, 1995: Sharber et al. 1994). Manuscript 90-3, Anchorage. Although we found no evidence that severity of Horak, D. L., and W. D. Klein. 1967. Influence of cap- injur electrofishinr yo g pulse shapes affected long- ture method fishinn so g success, stamina mord ,an - term survival d findi d e significantlw , y impaired talit f rainboyo w trout (Salmo gairdneri) Colon i - growth in more severely injured fish (class 2 and rado. Transaction e Americath f o s n Fisherie- So s ciety 96:220-222. 3 spinal injuries) e neaTh .r absenc f outwaro e d Hudy, M. 1985. Rainbow trout and brook trout mor- signs of spinal injury reinforces the need for X-ray tality from high voltage AC electrofishing in a con- analysi properlo t s y assess incidenc severitd ean y trolled environment. North American Journaf o l of spinal injury from electrofishing. Even with ra- Fisheries Management 5:475-479. diographs, initial estimate f severso e injury should Lamarque, P. 1990. Electrophysiology offish in electric be considered conservative. Additional researcs hi fields. Pages 4-3 Lamarque. 3. CowP G d i. nI xan , needed to characterize the causes of severe injury editors. Fishing with electricity: applications in freshwater fisheries management. Fishing News during electroshocking, develop ways to reduce it, Books, Oxford, UK. and examine furthe e fatperformancd th rean f o e McMichael . 1993A . G ,. Examinatio f electrofishinno g injured fis thn i h e field. injury and short-term mortality in hatchery rainbow trout. North American Journa f Fisherieo l s Man- Acknowledgments agement 13:229-233. Fundin stude provides th r ywa gfo Montany db a Mesa, M. G., and C. B. Schreck. 1989. Electrofishing Fish, Wildlife, and Parks. We thank H. Johnson mark-recaptur depletiod ean n methodologies evoke and K. Aune for facilitating purchase of the X-ray behavioral and physiological changes in cutthroat . DwyeP machin . r helW fo rvoltag d n pi ean e mea- trout. Transactions of the American Fisheries So- surements Bozemae Th . n Fish Technology Center ciety 118:644-658. Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012 . McDonaldG . D Milton . d A.J an . , .C , 1994. Effects generously allowed us use of facilities. J. Bag- of electroshock, air exposure, and forced exercise danov . Black W ,e Dalbey Su , . GustafsonD , . J , swin o m performanc rainbon i e w trout (Oncorhyn- Hupka. M. Vaughan, and D. Venditti provided in- chus mykiss). Canadian Journal of Fisheries and valuable assistance. K. Beland, D. Bennett, W. P. Aquatic Sciences 51:1799-1803. Dwyer . HuntJ , . ReynoldsJ , . ZalA ed providean , d Reynolds . KolzL . A . d . 1988B.J , an , . Electrofishing helpful comments on the manuscript. We acknowl- injury to large rainbow trout. North American Jour- naf Fisherieo l s Management 8:516-517. edge J. Reynolds, N. Sharber, and D. Snyder for Sharber, N. G., and S. W. Carothers. 1988. Influence of their pioneering effort brino t s g attentioe th o nt electrofishing pulse shape on spinal injuries in adult issu f electrofishineo g injury. rainbow trout. North American Journa f Fisherieo l s Management 8:117-122. References Sharber, N. G., S. W. Carothers, J. P. Sharber, J. C. de Dwyer. W. P., and R. G. White. 1995. Influence of elec- . HouseA Vos . D ,. d Jr.1994an , . Reducing elec- troshock on short-term growth of adult rainbow trofishing-induced injur f rainboo y w trout. North ELECTROFISHING EFFECTS ON RAINBOW TROUT 569

American Journal of Fisheries Managemen: 14 t tical error fishern si y research. Pages 93-10. J n 2i 340-346. Hunter, editor. Writing for fishery journals. Amer- Snyder. D. E. 1992. Impacts of electrofishing on fish. ican Fisheries Society. Bethesda, Maryland. Repor f Colorado t o State University, Larval Fish Vincent . 1971R . E , . River electrofishin fisd ghan pop- Laborator o U.St y . Burea f Reclamationo u , Salt ulation estimates. Progressive Fish-Culturist 33: Lake City, Utah Gled an , n Canyon Environmental 163-169. Studies Team, Flagstaff, Arizona. Vincent, E. R. 1987. Effects of stocking catchable-size Snyder, D. E. 1995. Impacts of electrofishing on fish. hatchery rainbo wwilo troutw d n trouo t t specien si Fisheries 20(0:26-27. e Madisoth n Rive d O'Delan r l Creek, Montana. Taube, T. T. 1992. Injury, survival, and growth of rain- North American Journal of Fisheries Management bow trout captured by electrofishing. Master's the- 7:91-105. sis. University of Alaska. Fairbanks. . 1984R Zar . J , . Biostatistical analysis editiond 2n , . Trippel. HubertJ . J . d A.E .. an , 1990. Common statis- Prentice-Hall, Englewood Cliffs, New Jersey. Downloaded by [Montana State University Bozeman] at 14:59 20 January 2012