Predation on Ruffe by Native Fishes of the St. Louis River Estuary, Lake Superior, 1989-1991 DEREK H. OGLE JAMES H. SELGEBY JACQ

Home , European perch, Gymnocephalus, Ruffe

North American Journal oj Fisheries Management 16:1 15-123. 19% *.'< Copyright by the American Fisheries Society 1996

Predation on Ruffe by Native Fishes of the St. Louis River Estuary, Lake Superior, 1989-1991 DERE . OGLKH E National Biological Service, Minnesota Cooperative FishWildlifeand Research Unit Department Fisheriesof Wildlife.and University Minnesota.of Paul,St. Minnesota 55/08.USA JAMES H. SELGEBY National Biological Sen'ice. Ashland Biological Station Ashland. Wisconsin 54806.USA JACQUELINE F. SAVING National Biological Sen'ice. Great Lakes Science Center 1451 Green Road. Ann Arhor. Michigan 48/05. USA RAYMOND M. NEWMAN Department of Fisheries and Wildlife. University of Minnesota St. Paul, Minnesota 55108. USA MAR . HENRYG Y National Biological Sen'ice. Minnesota Cooperative FishWildlifeand Research Unit

Abstract.—The ruffe Gymnocephalus cernuus, an exotic Eurasian percid, recently became established in the St. Louis River estuary. Lake Superior, after accidental introduction. Management actions (catch regulations and stockings) were enacted in 1989 to increase the density of top-level predators in the estuary, and thus to increase predation on ruffe. We conducted a field and laboratory stud determino t y d to an wha . eif t extent, native piscivores consume ruffe. Stomach f 3.66o s 9 predators were examined in 1989-1991. Ruffe occurred in 6.7% of burbot Lota lota. 5.8% of bullheads Ictalurus spp., 4.7 f smallmout%o h bass Micropterus dolomieu. 2.6 f norther%o n pike Esox tucius. 2.6% of black crappies Pomoxis nigromaculatus, and 1.3% of yellow perch Perca flavescens (4.5%> after 1989) captured during the 3-year study ruffo .N e were stomach7 foun96 n di s f walleyeo s Stizostedion vitreurn examined. Ruffe were 22.7 diee weighty th (b tf %o f bullhead)o s (during the only year bullheads were captured) and 0.1-17.9% of the diet of northern pike. Ruffe were 0.9-24.5% of the diet of smallmouth bass that contained fish. 1.5-6.9% of yellow perch that contained fish, and 0.0-10.9% of black crappies that contained fish. Most ruffe eaten were age-0 or small age-1 fish. In the laboratory, walleyes that were first fed soft-rayed prey or that were also offered soft-rayed prey consumed very few ruffe, whereas walleyes that were first fed spiny-rayed yellow perc r werho e also offered yellow perch consumed about equal number f ruff syelloo d an e w perch. Northern pik burbod ean t consumed about equal number f ruffsyelloo d an e w e percth n hi laboratory s unlikeli t I . y that predation will effectively contro e initiath l l expansio f ruffno n i e other area f tho s e Great Lakes because native predators initially consum ruffew efe , especiallf yi more preferred soft-rayed pre availablee yar .

Ruffe Gymnocephalus cernuus (formerly Acer- range of ruffe is thought to extend from the Great ina cernua). a percid native to Europe and Asia, Plains to the eastern seaboard and north into Can- were accidentally introduce . Loui(Busiaha dSt ad inte sth o n 1993) Greae th , t Lakes being prime River estuary of Lake Superior in the early 1980s, areas for colonization (Busiahn 1993; Edsall et al. presumably through ballast water discharge from 1993). transoceanic ships (Simo d Vondruskan n a 1991; Ruff hary ma em nativ. St e e faund fisth an hf o a Pratt et al. 1992). Ruffe were first discovered in Louis River estuary. Lake Superior, and the other the St. Louis Rive n 198 i rd positivel 6an y iden- Great Lakes (Edsal . 1993al t e l ; Mill. 1993)al t se . tified in 1987 (Pratt et al. 1992). A ruffe population Their diet in the St. Louis River estuary suggests was established in the estuary by 1988 and has that ruffe may compete with yellow perch Perca increased dramatically since then (Pratt etal. 1992; flavescens, trout-perch Percopsis omiscomaycus. . Mill1993)al e potentiat e s Th . l North America othed an rn benthic-feedin g fishes (Ogl . 1995)al t ee . 115 116 . OGLAL T EE

European studies indicate that ruffcompety ma e e piscivores (walleye, northern pike burbod an , t Lota with European perch Perca fluviatilis. a close rel- lota) to consume ruffe under artificial conditions. ative of the yellow perch. In enclosures in small These two pieces of information provide a basis Swedish ponds, the growth of Eurasian perch de- for evaluating the effect of predation on ruffe dur- clined and the amount of zooplankton in their diet speciesine gth ' earl . LouiySt e yearsth Riven si r increased when ruffe densit s experimentallwa y y estuary, and they also provide a baseline with increased (Bergma Greenberd nan g 1994) Locn I . h which future determinations of predator diet can Lomond, Scotland, the relative abundances of ruffe be compared. and Eurasian perch were inversely related (Mait- land 1990; Adams and Tippett 1991). In addition Methods to competition for food, egg predalion by ruffe has Field Stud\ been implicated in the decline of some populations of powan Coregonus lavaretus (Balagurova 1963; Several gears were use colleco dt adequatn a t e Pavlovski Sterligovd yan a 1986; Adam Tipd san - numbe f predatoro r s fro e estuarmth y durine gth pett 1991 d vendacan ) . cilbulaC e (Pokrovskii open-water season f 1989o s , 1990 d 1991an , n I . 1961). Thus, expansion of ruffe in the Great Lakes each year e collectew , d predators regularly with may have serious ecologica economid an l c impacts maio tw n gears. Every other week during all three on existing fish communities (GLFC 1992; Edsall years, we towed a 5.2-m-lead-line bottom trawl t ale . 1993). for 5 min at 2.6 km/h at 40 locations selected ran- Local resource agencies began an intensive domly from throughout the estuary. More details management plan in an attempt to limit the density about this gear and selection of locations were of ruffe in the St. Louis estuary and to discourage give . (1995)Ogly al b n t e n week O . s whee w n the dispersa f ruffo l e fro e estuarmth y (Busiahn did not trawl, we set fyke nets (25-mm-bar mesh) 1993). As part of this plan, piscivores were stocked for approximatel t foua h r 4 locationy2 r eleco s - and restrictions were impose e harvesth n f o do t trofished t (pulsenigha ) t DC dalon ga predeter - native piscivores. In 1989 and 1990, an average mined length of shoreline in Allouez and Superior f walleyo f 63y o 1fr e Stizostedion vitreum. 13 fin- bays (Figur t fykse e ee I) netW . s onl 198n yi d 9an gerling walleye, and 6 subadult northern pike Esox 1991 and electron'shed only in 1990. To obtain Indus were stocke r hectare 4,654-hpe dth n i e a additional samples e conductew , o intensivtw d e estuary n 1989I . , daily creel limits were reduced sampling efforts in 1989 and obtained stomach from six to two walleyes and from five to two contents from fish collected by cooperating man- northern pike. The minimum length limit on wall- agement agencie r othefo s r purpose l threal n ei s alss . owa increasee mm ey 1 38 o dt fro5 m30 years. The two intensive efforts were conducted e effectivenesTh f predatoro s controllinn si g ruffe in Alloue Superiod zan r bays wit n 11.2-m-leadha - has varied in Europe. Populations of ruffe and oth- line botto km/0 md 4. t trawhan a n l mi towe 5 r dfo er forage species declined after the abundance of fyke nets (as described above). In addition, we /ander (pikeperch) Stizostedion luciopercad an obtained predators from experimental gill nets (1.8 northern pike increase somn di e Polish lakes (Bon- m deep and 15.2 m long with mesh bar measures ar 1977) and after regulations restricted the harvest of 19, 25, 32, 38, and 51 mm, hung at 50%) set of zander in an Estonian lake (Pihu and Maemets in 1989 and 1990, from an open-water creel survey 1982). Increases in ruffe abundance following de- in 1989 d froan ,m fyke nets (sam s describea e d cline f predatoro s y fis somn i h e Russian (Popova above) set at additional locations in November and Sytina 1977) and English (Duncan 1990) wa- 1989 l walleyesAl . , northern pike, smallmouth ters also indicate that predatory fish may regulate bass Micropterus dolomieu, and black crappies Po- ruffe abundance. However, in Lake Tjeukemeer, moxis nigromaculatus collected in all three years, increased /.ander abundance cause terminatioy db n all yellow perch over 150 mm total length (TL) of the zander fishery did not result in a decreased collecte n 199 i dd 1991an 0 , all bullheads (pri- abundanc f ruffeo e becaus abundance eth - Eu f eo marily Ictalurus melas. somt bu . nebulosus)e/ col- ropean smelt Osmerus eperlamts. e preferreth d lected in 1991, and all burbot collected in 1989 prey of zander, was high and unaffected by zander from these gears were use completr dfo e diet anal- abundance (Lammens et al. 1990). ysis. Altogether collectee w , d 1,925 predators with In this paper e describw , e extenth e o whict t h regular sampling, 480 predators with intensive native predators consumed exoti. St ce ruffth n i e sampling, and 400 predators with the other gears Louis River estuary and the ability of three large just described additionn I . includee w , l yellodal w PREDATIO RUFFN N O LAK A N EI E SUPERIOR ESTUARY 117

FIGURE I.—Map of the St. Louis River estuary. Lake Superior. perc caplurem h m oveO l bullJ 1 al rn 198d i d - 9an measured for length, their weight was estimated head capturem sm ove0 199n 16 di r r ih0fo e pur- e meath s na weigh f measurablo t e specimenf o s pose f evaluatino g predatio n ruffo n e only (i.e., that species found in stomachs collected during these analyses wer t includeno e n Ihi de general e samth e month. Ruffe foun n stomachi d s were diet description). These samples broughl ihe num- classified as age-0 or age-1 and older based on TL. ber of predator siomachs examined for ruffe to Diet composition of predators caught in the field 3,669. summarizes wa percene th s da f tota t weigho t we l t Individual predators were measured (TL ± 1 ai tim f ingestioneo . Diet composition summaries mm) and their whole stomachs or stomach contents for species thought to be primarily piscivorous were removed and preserved in 10% formalin. (walleyes, northern pike bullheadd an , ) were based Stomach contents were flushed (by a method sim- n identifiablo e invertebrate d fishan s , whereas ila o tha t rf Seaburo t g 1957) from live d fisan h summaries for species thought to be less pisciv- whole stomachs were removed from all dead fish orous (smallmouth bass, yellow perch, and black (which were mostly from gile creeth l netld an s crappie) were base onln do y identifiable fish each survey). Stomach contents were identified, count- year ensuro T . e adequate sample size describr sfo - ed, and weighed wet (±0.1 g). Invertebrate prey ing the diel of each predator, all stomach samples were identified to family or genus but were pooled collected in the same year were pooled. Walleyes for summarization. Fish prey were identifieo t d capture e 11.2-th n di m botto my 198 trawMa 9 n i l specie casr famile so th f smal eo n y(i l centrarchids wer t include eno diee th t n summariedi s because and ictalurids) with the aid of reference fish spec- they were collected earlier that year than in 1990 imens, body parts (e.g., ololilh preopercles)d san , d 1991 d an thean ,y wer n theio e r spawnin- mi g and meristics (Ogle 1992). Prey fish that were gration from Lake Superior. These walleyes con- largely intact were mm)measure1 e ± Th . L d(T taine numbery d (b nearl % y)97 rainbow smel- Os t wet weight at time of ingestion for intact prey fish merus mordax, whic d alsha ho congregater fo d was estimated from length-weight equations de- spawning near Superior Entry. rived from forage fish collecte n routini d e irawl Laboratory Study catches or from literature values (Ogle 1992). The ability of walleyes, northern pike, and bur- When partially digested prey fishe b coul t no d bot to consume ruffe was tesled in trials with only 118 . OGLAL T EE

TABLE 1.—Numbers and sizes of predators sampled in the St. Louis River estuary, 1989-1991, and numbers, ages, and si/es of ruffe found in stomachs of those predators. Age and size are summarized for all measurable ruffe observed in predator stomachs collected durin years3 l gal . Total length smillimetersn i (TL e ar ) .

Predators Ruffe consumed Sample size Year Age Mean TL Species Mean TL (range) 1989 1990 1991 1989 1990 1991 0 >1 (SE) Walleye 419(180-758) 524 203 240 0 0 0 0 0 Northern pike 606(251-1.015) 262 40 275 2 1 12 2 13 114(8) Burbot 553 (333-830) 60 0 0 5 4 1 86(18) Bullheads 256(190-441) 0 383 742 11 80 II 80 81(2) Small mouth bass 300(173-468) 47 36 89 1 4 3 5 3 53 (6) Yellow perch 238(169-335) 491 75 126 0 8 4 9 3 43 (5) Black crappie 280(184-335) 20 17 39 0 1 3 4 0 28 All predators 1,404 754 1.51 1 8 25 102 35 100 79(2)

ruffe, and the preference for or against ruffe was northern pike; 93 ± 2.4 mm for burbot) and yellow teste trialn di s with ruff alternativd an e e preyr Fo . r northerfo percm m n h 8 pike(158. ± 0; 11± 5 walleye only, thes testo etw s were conducted with 3.8 mm for burbot). In addition, a choice of large predators that were first fed either soft) -mm osmald 8 r 4. spinyan ) l± ruff4 -mm (8 e6 (147. 5± rayed prey. Predators were acclimate tested dan d was offered to each northern pike. The second in circular tanks that wer 5 me0. dee eithed pan r northern pike quit feeding during weeks 9-12t I . diameten i 1.5m r (walley m burbotd 5 an e 2. r o ) s subsequentlwa y offered yellow 3 perc r fo h in diameter (northern pike). Water temperatures in weeks, during which time it fed consistently and the tanks were 11-12°C. Yellow perch, which are thes wa n reteste combinatioa n do yellof no w perch . LouicommoSt e sth Rivern ni , were selectes da an dadditionan a ruff r fo e weeks3 l e burboOn . t e alternativth e spiny-rayed prey. Lake trout Sal- died during the mixed-prey species test and data velinus namaycush d rainboan w trout Oncorhyn- from that includeno t fise ar hthin di s analysis. chus mykiss were used as soft-rayed alternative Comparisons among treatments in the experi- prey because they were readily available and easily ment r walleyfo s e (two tanks d northeran ) n pike maintaine e laboratoryth n i d . During testsa , (two individuals) were based on the mean number known numbe f preyro excesn i , f possiblso e daily of fish consumed per predator per day over each consumption rates (3-5 f %predatoo r biomass), weekly interval. Sample sixes were too small for added an wereac) o dt hemm measure1 ± L d(T meaningful statistical analysis. For the experi- test tank numbee Th . f preo r y eate recordes nwa d ments with burbot, the median number of fish con- and prey were replenished same dailth eo yt den- sumed per burbot per day over the entire 14-d sity. experimen response th s wa te tested. Four tanks containing thre r fouo e r walleyes (450-600 mm TL) per tank were provided ruffe Results and alternative prey. Walleyes in two of the tanks Consumption of Ruffe and Diet Composition in were tested first on spiny yellow perch (mean TL Fieldthe ± SE, 112 ± 4.9 mm), then on ruffe (84 ± 2.2 mm), and finally on a combination of ruffe and All predators examined from the estuary except yellow perch (the sam single-pree eth sizen i s sa y walleyes consumed some ruffe (Table I). Ruffe tests). Walleye othee tanko th n rtw si s were tested occurre 6.7n i d f %burboto , 5.8 f %bullheadso , first on soft-rayed lake trout (123 ± 5.0 mm), then 4.7 f %smallmouto h bass, 2.6 f %northero n pike, on ruffe (12 0± 12. 5 mm) finalld coma an , n o y- 2.6 %f blaco k crappies 1.3d an ,%f yello o w perch binatio f lakno e trou ruffed an t . Walley firse th tn ei (4.5% after 1989) captured during the 3-year study. f tanko t sse were give n additionaa n l choicf o e No ruffe were found in 967 walleye stomachs ex- large (137 ± 5.4 mm) and small ruffe (85 ± 3.7 amined (Table 1). Northern pike and bullheads mm) during weeks 11 and 12 of the trials. generally consumed age-1 and older ruffe, whereas Individual northern pike (N = 2; 600 and 750 the other predators consumed mostly age-0 ruffe mm) and burbot (N = 6; 380-550 mm) were tested (Table 1). However, bullheads consumed only age- similarly t onlbu ,y witr fo h ruffm m e 8 (158. 0± 0 ruffe in 1990 and small age-1 ruffe in 1991. PRKDATION ON RUFFE IN A LAKE SUPERIOR ESTUARY 1 19

TABLE 2.—Percentage f totaso l prey weigh stomachn i t f walleyeso s , northern pike d bullheadan , s contributey db identifiable prey items. This summary is restricted to stomachs with identifiable food. The total numbers of stomachs examined arc reported in Table 1.

Walleye Northern pike Bullheads 1989 1990 1991 1989 1990 1991 1991 Prey items (N = 142) (/) V -89 ) 35 (= N (N= 112) (= N16 ) (N = 248) Ruffe 0.0 0.0 0.0 O.I 17.9 11.7 22.7 Yellow perch 19.1 2.3 32.1 21.3 53.3 38.7 13.6 Black crappie 6.9 2.6 2.9 6.2 0.0 17.5 7.0 Walleye 0.0 0.0 0.0 6.8 0.0 2.6 0.0 Btheostomatinae 0.0 6.7 0.0 0.0 0.0 0.0 0.3 Trout-perch 0.1 0.0 0.0 0.8 16.9 0.7 8.9 Other spiny-rayed fish" 1.8 1.3 0.0 4.5 0.0 0.0 0.3 Ictaluridae 17.2 24.7 11. 1 0.2 0.0 2.7 0.6 Kmcrald shiner 11.8 30.9 15.1 5.2 9.2 4.3 1.9 Spottail shiner 13.4 9.2 23.4 2.1 0.0 11.2 37.1 Rainbow smelt 25.0 11.3 3.8 4.1 0.6 0.0 0.0 Catostomidae 0.2 0.3 0.0 47.5 0.0 10.2 O.I Other soft-rayed lishb 2.8 0.0 2.8 0.9 0.0 0.0 3.5 Invertebrates 1.6 10.9 8.8 0.3 2.1 0.4 4.0 a Rock bass Amhtoplites ruprstris, smallmouth bass, white perch, and unidentified Centrarchidae. h Ale wife Alostt pseudohurengus. common shiner iMxilus cttrnutus. and golden shiner Notemigonus crysttleucas.

Northern pike consumed larger ruffe than burbot shiners Notropis atherinoides d Notropisan hud- and bullheads, and these three species consumed sonius in all three years, but varied their con- larger ruffe tha othee nth r predators (Table 1). Ruffe sumptio f otheno r prey among years e (TablTh . e2) were more than 10weighty %complete (b th f o ) e other major prey eate walleyey nb s were ictalurids die f northero t aln i nl threpik f n 199o i ed 199 d years0an an 1 , yellow perc 198n h i d 19919an , bullhead n 1991i s e onl, th whic ys yeawa h rd rainbo thaan t w smel n 198i d t 1990an 9 . Inverte- the diet compositio f bullheadno determines swa d brates, mostly burrowing mayflies Hexagenia spp., (Tabladditionn I . e2) , ruffe were more tha% wer10 n e about 10% of walleye diet 199n s i 1991d 0an , piscine th f o e die f smallmouto t n 1989i t onl h % basbu .2 y Walley199n si d 0an e consumed mostly f blaco 199d k1 an crappie 199n si 0 e (TablsmalTh . e3) l ictalurid s, 19.5)(meanSD ; , ,TL 55.m 3m relatively high percentage of ruffe in the diet of which have hardened soft rays (Hubbs and Lagler northern pike and black crappies in 1990 was due 1958), but a wide range of sizes of spiny yellow to a single ruffe found in relatively few samples perch (mean, 111.8 mm; SD, 33.9). of both predators. Northern pik primarild efe catostomidn yo d san Walleyes fed mostly on emerald and spottail yellow perch in 1989, and yellow perch and a va-

TABLE 3.—Percentage f totao s l prey weigh stomachn i t f smallmouto s h bass, yellow perch d blacan , k crappies contribute identifiably db e prey items. This summar restrictes yi stomacho dt s with identifiable fishe totaTh . l numbers of stomachs examine e reportedar Tabln di e 1.

Smallmouth bass Yellow perch Black crappie 1989 1990 1991 1990 1991 1989 1990 1991 Prey items (N= II) (N = 20)) 27 (= N ) 20 (N= (N = 37) Ruffe 0.9 21.4 24.5 6.9 1.5 0.0 10.9 3.0 Yellow perch 35.9 10.1 0.0 0.0 5.4 11.3 0.0 0.0 Black crappie 16.3 2.7 18.4 0.0 13.8 7.3 0.0 93.9 White perch 0.0 0.0 0.0 0.0 0.0 0.0 28.3 0.0 Kiheostomatinac 10.6 18.2 6.3 10.0 1.7 0.0 0.0 0.0 Trout-perch 1.4 17.5 4.6 13.6 7.2 0.0 0.0 0.0 Other Centrarchidae 4.9 0.4 2.9 0.0 0.0 25.4 0.0 0.0 Walleye 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.1 Ictaluridae 2.1 0.0 0.0 0.0 0.0 10.0 0.0 0.0 Hmcrald shiner 0.0 11.0 11.2 55.7 20.5 10.7 60.7 0.0 Spottail shiner 14.0 18.7 24.3 13.8 51.0 25.9 0.0 0.0 Other soft-rayed fish0 13.8 0.0 7.9 0.0 0.0 9.4 0.0 0.0 a Alewife. golden shiner rainbod an , w smelt. 120 . OGLAL T EE

riet f otheyo r item n 199 i sd 1990an 1 (Tabl. 2) e The other major prey eaten by northern pike were ruffe and trout-perch in 1990 and black crappies, ruffe, spottail shiners, and catostomids in 1991. Amon majoe gth r spiny fish besides ruffe, northern pike consumed medium-size to large yellow perch (mean, 133.4 mm; SD, 41.0) and small black crappies (mean , 31.3), SD 105.; . 4mm Bullheads fed primarily on spottail shiners, ruffe, and yellow perch (Table 2). In August-September 9 11 13 of 1991, many age-0 black crappies were captured I Trout Ruffe Trc>ut Kilix in the fyke nets along with many bullheads that (0 o 2.0 —— contained only age-0 black crappies. This suggests O that bullhead havy ma se consume e abundandth t small black crappies in the fyke nets and that the consumptio f blacno k crappie bullheady b s s wa s 3 1.0 overestimated addition I . age-e th o 0nt black crappies, bullheads consumed intermediate-size spiny 0 yellow perch (mean, 95.6; SD 33.0). 0.0 Smallmouth bas e mostlat s y spiny-rayed fish, 2345 6 and ruffe was a prominent noninvertebrate food Wee Tesf ko t ite n m199 i d 1990an 1 (Tabl . Yello3) e w perch FIGURE 2.—Mean consumption ratr individuaepe r pe l consumed mostly emerald shiners in 1990 an walleyey db y da s during each wee predatiof ko n tests with spottail shiners in 1991 (Table 3). Large black yellow perch, ruffe, or a combination of yellow perch and ruffe (top r wito ) h trout, ruffe combinatioa r o , f no crappie mostle sat y centrarchids Lepomis sppd an . trou d ruffan t e (bottom) choicA . f larg o ed smal an e l spottail shiner 1989n si , emerald shiner 1990n si , ruffe was offered in weeks 11 and 12 during the yellow and age-0 black crappies in 1991 (Table 3). More perch-ruffe series. Filled bar consumptioe sar n ratf eo than 90% of the fish eaten by yellow perch and ruffoped an en bar consumptioe sar n rate f alternativso e smallmouth bass and 100% of the fish eaten by prey. black crappies were less than 100 mm TL. timed on . When give choicna f preeo y size, both Consumption of Ruffe in the Laboratory northern pike ate only large ruffe. Walleye consumptio f ruffo n n laboratori e y Burbot ate similar numbers of ruffe and yellow tanks was affected by the walleye's previous feed- perc singlen i h d mixed-prean - y species testP ( s ing experienc e typ th f alternativo ed an e e prey 0.10)> single-pren I . y species tests, burboa e at t offered. Walleyes that were first fed soft-rayed fish median (95% confidence interval) of 0.32 (0.14- (trout) consumed no ruffe when only ruffe were 0.50) yellow perc 0.3d han 6 (0.07-0.71) ruffr pe e offere d verruffw an dfe y e when both ruffd an e day. In mixed-prey species tests, burbot ate 0.43 trout were offered (Figure 2). Walleyes that were (0.28-0.64) yellow perc 0.1d han 4 (0.00-0.50) ruffe first fed yellow perch consumed ruffe offered alone r daype . aa ratt e similae ratth e o observet r r yellofo d w perch alon consumed ean d ruffyellod an e w perch Discussion at similar rates when both prey were offered (Fig- Bullhead d northeran s n pike probably exerted . Whe2) e nur presented equal number f smalo s l the greatest overall predation pressure on the ruffe and large ruffe, 0.33 small and 0.03 large ruffe populatio . LouiSt e sth Riven ni r estuary fiele .Th d were consumed per walleye per day. study indicated that bullheads, burbot, piscivorous Northern pike testelaboratore th n di y gave var- smallmouth bass northerd an , n pike wer prie eth - resultsd ie . Both pike consumed about equal num- mary consumers of ruffe. Overall predation by ber f ruff syelloo d ean w perch when each pres ywa bullhead s likelwa s y greatest because thee ar y offered alone (Figure 3). The first pike consumed abundant in the estuary (Ashland Biological Sta- no ruffe when yellow perch were simultaneously tion, unpublished data). Overall predatio bury nb - offered secone Th . d pike consume fooo dn d when bot and smallmouth bass was likely low because both ruff yellod an e w perch were first offeredt bu , burbot inhabit the estuary for only a short time consumed both prey when they were offere secda - (Schram 1983 d smallmoutan ) h bass havw lo e PREDATIO RUFFN N O LAKA N EI E SUPERIOR ESTUARY 121

Yellow j Ruffe ' Mix No Tests ical Station, unpublished data). This increase in 0.6 Perch ran bottoe f ruffth ko n i em trawl e catchedu s swa to both an increase in ruffe abundance and a de- crease in the abundance of yellow perch and small 0.4 forage fish (GLFC 1992). Furthermore, more small ruffe, whic e fielhth d stud laboratord yan y exper- 8. iments suggest are preferred by most predators 0.2 (except northern pike), became available as large year-classes of ruffe were produced and body growth of ruffe declined (Ashland Biological Sta- Yellow , Ruffe Mix Yellox Mi w tion, unpublished data). However, ove e samth r e H Perch 3 Perch time e percentagth , e die f f ruffth moseo o t n i et predators remained low or was lower than per- S centage f lesso s abundant prey r exampleFo . , wall- 5 1.0 eyes did not consume ruffe during the study, and northern pike consumed much less ruffe than they 0.5 did yellow perch in the last two years of the study. e ruffe'Th s large dorsal, anal, pelvic pred an , - I opercular spine detey sma r predatio walleyesy nb , 0.0 but may have less of an effect on northern pike. '5' '?• '9' IV '13' '15' \T Week of Test Spineeffectivn a e sar e antipredator adaptatio- nbe cause they increas apparene eth preye t th siz f ,eo FIGURE 3.—Mean consumption rate per individual per limit the attack to the prey's head, and may punc- firse th t y (top b secon d y )an da d (bottom) northern pike during each week of predation tests with yellow perch, ture the predator's throa r stomaco t h lining (Eklov ruffea combinatio r o , f yellono w perc d ruffean h A . Hamrid an n 1989) fielde th n ,I . walleyes consumed choic f larg eo smald ean l ruff s offeree firsth wa e t o t d both soft-raye spind dan y fish, althoug spine hth y northern pike in weeks 5-7 and to the second northern fish were generally small. In the laboratory, wall- . pikFille9 weekn ed i dan bars 8 consumptioe sar n rate eyes consumed fewer yellow perch and ruffe than of ruffe and open bars are consumption rates of alter- soft-rayed trout, very few ruffe when trout were native prey. present mord an , e smaller than larger ruffe, results consistent with an aversion to spines. However, abundance (Ashland Biological Station, unpubli- walleye laboratore th n si y readily consumed ruffe shed data). Overall predatio ruffn n o northery b e n when they were previousl yellod yfe w perch. Thus, pike was probably high because the resident pop- walleye conditionee b n sca t ruffe t ea the o bu dt ,y ulation of northern pike is large, even though some seem to prefer soft-rayed prey to ruffe. When ruffe fish may leave the estuary for short periods (J. were first introduced into the St. Louis River es- Spurrier, Minnesota Departmen f Naturao t - Re l tuary, soft-rayed emerald and spottail shiners were sources, personal communications). Predation by abundant there (Ashland Biological Station, un- . Loui walleyet detecteSt no e ss th Riven wa sdi r published data). The availability of these prey may estuary, even though we examined 967 walley walleyo reasoe n ey on e nwh b e stomachs froe mth stomachs. Furthermore, most adult walleyes only estuary contained ruffe. Northern pike also con- monthw residestuare fe th a n r sei ydurinfo d gan sumed both spiny soft-rayed -an d field e preth n y,i following the spawning season (Schram et al. spinbue th t y prey were larger than those consumed 1992). Yellow perc blacd han k crappies consumed by walleyes, which suggests that spines deter some ruffe, but overall predation was likely low northern pike less. Furthermore, both laboratory because piscivorous size-classes of these predators and field evidence suggest that northern pike prefer wer t abundanno e t (Ashland Biological Station, larger ruffe. unpublished data). In addition to the ruffe's spiny armor, its habitat The field results suggest that some predators, preferences may reduce the probability of en- especially walleyes, may have selected other prey counter between it and predators. Age-1 and older over ruffe. Ruffe increased fro sevente mth h most ruffe stay nea bottoe rth m (Ahlbert 1970; Bergman abundant species captured in the routine bottom 1988 f deepero ) , darker waterd an s y durinda e gth trawl sample 198n si 9 secon e toth d most abundant of shallower littoral areas at night (Ogle et al. fish captured in 1990 and 1991 (Ashland Biolog- 1995). This behavior would reduce encounters 122 OGL ALT EE .

with predators that feed by sight in good light, the Great Lakes. Native predators will likely con- such as northern pike, smallmouth bass, black sum ruffw fe e e initially, especiall f mori y e pre- crappies, and yellow perch (Scott and Crossman ferred alternative prey (e.g., soft-rayed fish or yel- 1973). Furthermore, northern pike generally feed low perch) are available. on pelagic prey in shallow littoral areas during the day, as indicated by the large proportion of yellow Acknowledgments perch and shiners in their diet and by reports in We thank Denni se Wisconsi th Prat f o t- De n the literature (Hart and Hamrin 1988; Savino and partmen f Naturao t l Resources (DNR), John Spur- Stein 1989). In contrast, there should be a high rieDennid an r s Anderso Minnesote th f no a DNR, probabilit f encounteyo r between age- olded 1an r Reed Glesne U.Sth f . o eFis d Wildlifan h e Ser- ruffe and predators that are active at twilight or vice's Office of Fisheries Assistance in Ashland, night, suc walleyes ha s (Scot Crossmad an t n 1973) Busiahm To Neid nan l Kmiece Greae th f kto Lakes and bullheads (Damel Meierottd an l o 1965). Indian Fisd Wildlifan h e Commissione th d an , e spatia Th temporad an l l overlap between age- SiscowetCoasterV V - R R as crewe d r th an fo f so ruff0 piscivoroud an e s . predatorLouiSt e sth n si sistance in sample collections; D. Zilker and T. River estuary is difficult to predict because the diel Darland for assistance in stomach analysis; and J. distribution of age-0 ruffe is largely unknown. E. Miller and G. Yearout for assistance with lab- Contact between age-0 ruffd crepusculaan e r o r oratory studies. Comment earlien o s r versionf o s nocturnal predators, suc s walleyea h d bullan s - this manuscript by J. Spurrier, P. J. Schneeberger, heads, should be high because age-0 ruffe are also an anonymoun da s reviewer were- helpfuap d an l nea e bottoth r t mnigha botn i t h shallow littoral preciated. Support for this project came from the deepen i area d an sr channels (Ogl t ale . 1995). Minnesot a GranSe a t College Program e U.Sth , . e day-timTh e distributio f age-o n t 0no ruffs ha e Department of Commerce under grant USDOC- been adequately determined (Ogle et al. 1995). NA86AA-D-SG112, the Minnesota Agricultural encountee Eveth f ni r rate betweena ruffd an e Experiment Station under project 72, the National predator is high in low-light conditions, predation Biological Service's Great Lakes Science Center, becausw lo ma e sensore yb eth y syste f ruffmo e the Minnesota DNR e Wisconsith , n DNRe th , is exceptionally sensitive. Their tapetum lucidum Great Lakes Indian FisWildlifd han e Commission, (Ahlbert 1970; Craig 1987) should help ruffe see anGreae dth t Lakes Fishery Commission. This si predator twilightn si , when predators suc walls ha - journal reprint JR324 of the Minnesota Sea Grant eye typically have an advantage over their prey. College Program, paper 20,74Minnesote th f 0o a Their sensitive lateral line system (Disled an r Agricultural Experiment Station contributio- se n Smirnov 1977; Dento Grad nan y 1989) should help ries, and contribution 916 of the Great Lakes Sci- them detect movements of active predators in low ence Center. ligho n r t o (Collett . 1977)al t ee . References Evidence from this study suggests that predation on ruffe in the St. Louis River estuary by native Adams, C. E., and R. Tippett. 1991. Powan, Coregonus predator increasy futuree sma th n ei , even though lava re l us (L.), ova predation by newly introduced ruffe, Cymnocephalus cernuus (L.) n Loci - , Lo h predation on ruffe remained low or increased only mond, Scotland. Aquaculture and Fisheries Man- slightl somr yfo e predators during this study. First, agement 22:239-246. if ruffe continue to increase in density, their Ahlbert, I. 1970. The organization of the cone cells in growth will be reduced (Hansson 1985; Bergman e retinath f foueo r teleosts with different feeding and Greenberg 1994 thed an )y will remain small habits (Perca fluviatilis L., Lucioperca lucioperca longer; most predators apparently prefer small ruffe. L., Acerina cernua L., and Coregonus albula L.). Arkiv for Zoologi 22:445-480. Second, continued predation by native predators Balagurova . 1963V . M ., The biological- basira r fo s mostln o y native prereducy yma abundance eth e tional regulation of fishery on Syam group lakes. of the preferred native prey, leading to an increase Karelya. Academy of Science Press, Moscow. in the number of ruffe consumed. Both of these Bergman, E. 1988. Foraging abilities and niche scenarios would be accompanied by undesirable breadths of two percids, Perca fluviatHis and Gym- changes in the estuary's community. nocephalus cernua. under different environmental conditions. Journal of Animal Ecology 57:443-453. o matte N predatiow ho r n ruffo n e changen i s Bergman . GreenbergA . L . d E.an , . 1994. Competition e futureth e resultth , f thio s s study indicate that between a planktivorc, a benthivore, and a species predatio s unlikeli n o effectivelt y y control ruffe with ontogenetic diet shifts. Ecology 75:1233- during early stages of colonization elsewhere in 1245. PREDATIO RUFFN NO LAKA N BI E SUPERIOR ESTUARY 123

Bonar . 1977A . . Relations between exploitation, yield, Maitland . 1990S . P , . Biolog f fresyo h waters edid 2n ,- and community structur Polisn ei h pikeperch (Stizo- tion. Chapman and Hall, New York. siedion lucioperca) lakes, 1966-1971. Journae th f o l Mills . Leach . Secor H . L.. L Carlton . E ,T . J . . C J , . d an , Fisheries Research Board of Canada 34:1586-1591. 1993. Exotic species in the Great Lakes: a history Busiahn. T. R. 1993. Can the ruffe be contained before f biotio c crise d anthropogenian s c introductions. it becomes your problem. Fisheries !8(8):22-23. Journa f Greao l t Lakes Research 19:1-54. Colletle. B. B., and six coauthors. 1977. Biology of the Ogle, D. H. 1992. Trophic relations of ruffe (Gymno- percids. Journa Fisheriee th f o l s Research Boarf do cephalus cernuus (L.)) in the Si. Louis River harbor. Canada 34:1890-1899. Lake Superior. Master's thesis. University of Min- Craig . 1987F . J e .biolog .Th f perco yd relatean h d nesota . PaulSt , . fishes. Timber Press, Portland, Oregon. Ogle . Selgeby H . . H.. Newman. D G .J , M . . M R , d an . Henry. 1995. Diefeedind an t g periodicit f ruffyo e . MeierottoR . R d an .Darnel) , 1965M. . R . Diurnal in the St. Louis River estuary. Lake Superior. Trans- periodicity in the black bullhead Ictalurus me/as actions of the American Fisheries Society 124:356— (Ralinesque). Transactions of the American Fish- 369. eries Society 94:1-8. Pavlovskiy Stcrligova. P . O d . L.S , an , . 1986. Predation Denton. E. J.. and J. A. B. Gray. 1989. Some obser- of ruffe, Gymnocephalus cernuus, and bcnlhic in- vations on the forces acting on neuromasts in fish vertebrates on the eggs of Lake Syam whitefish. lateral line canals. Pages 229-246 in S. Coombs, P. Coregonus lavaretus palassi. Journa f Ichthyologo l y Corner, and H. Mun/. editors. The mechanosensory 26(6):80-86. lateral line: neurobiology and evolution. Springer Pihu . MaemetsA . d E.an , . 1982. The managemenf o t Verlag Yorkw Ne , . fisheries in Lake Vortsjarv. Hydrobiologia 86:207- . SmirnovA . S d an . Dislcr. 1977N. . N , . Sensory organs 210. of the lateral-line canal system in two percids and Pokrovskii, V. V. 1961. Basic environmental factors de- their importance in behavior. Journal of the Fish- termining the abundance of whitefish. Trudy So- cries Research Boar f Canaddo a 34:1492-1503. veshchanii 13:228-234. (In Russian.) Duncan . 1990A , reviewA . : limnological management . SytinaPopovaA . L . .d A. O ,1977 an , . Foo feedd dan - biomanipulatiod an thn ni e London reservoirs- Hy . ing relations of Eurasian perch (Perca fluviatilis) drobiologia 200-201:541-548. d pikepercan h (Stizostedion lucioperca) varioun i s Edsall . Selgeby . H A. . T ,. P J . DeSorcie J . . R T ,. J d an , waters of the USSR. Journal of the Fisheries Re- French III. 1993. Growth-temperature relatior nfo search Board of Canada 34:1559-1570. young-of-the year ruffe. Journa Greaf o l t Lakes- Re Pratt, D. M., W. H. Blust, and J. H. Selgeby. 1992. Ruffe, Gymnocephalus cernuus: newly introduced search 19:630-633. in North America. Canadian Journa f Fisherieo l s Eklov, P., and S. F. Hamrin. 1989. Predatory efficiency and Aquatic Sciences 49:1616-1618. and prey selection: interactions between pike Esox . SteinA Savino . R . d . 1989F.J ,an , . Behavio f fiso r h Indus, perch Perca fluviatilis, rudd an d Scardinus predator d theian s r prey: habitat choice between erythrophthalmus. Oikos 56:149-156. open wate densd ran e vegetation. Environmenta- Bi l GLFC (Great Lakes Fishery Commission). 1992. Ruffe ology of Fishes 24:287-293. in the Great Lakes threaa : Norto t t h American fish- Schram . 1983T . S , . Seasonal movement mortalitd san y eries. Great Lakes Fishery Commission, Ann Arbor, estimate f burboso western i t n Lake Superior. Wis- Michigan. consin Departmen f Naturato l Resources, Fish Man- Hansson. S. 1985. Effects of eutrophication on fish agement Report 119, Madison. communities with special referenc o tht e e Baltic Schram. S. T., T. L. Margenau. and W. H. Blust. 1992. Sea—a literature review. Institute of Freshwater Re- Population biolog managemend yan walleye th f o t e search Drottningholm Report 62:36-56. in western Lake Superior. Wisconsin Departmenf o t . HamrinF . S d . Hartan 1988. P. , . a Pikselectiv s a e e Natural Resources, Technical Bulletin 177. predator: effects of prey size, availability, cover, an d. CrossmanJ . E Scott d . B.an W ,. . 1973. Freshwater pikdimensionsw eja . Oikos 31:220 226. fishe f Canadao s . Fisheries Research Boar f Cando - ada Bulletin 184. Hubbs. C. L.. and K. F. Lagler. 1958. Fishes of the Great Seaburg, K. G. 1957. A stomach sampler for live fish. Lakes region. University of Michigan Press. Ann Progressive Fish-Culturisl 19:137-139. Arbor. Simon, T. P.. and J. T. Vondruska. 1991. Larval iden- n Densen. R va . d T an . , L Lammens . . R.R W ,. H . E , tification of the ruffe, Gymnocephalus cernuus (Lin- Knijn. 1990. The fish community structure in Tjeu- naeus) (Percidae: Percini) in the St. Louis River kemeer in relation to fishery and habitat utilization. estuary. Lake Superior drainage basin, Minnesota. Journa f Fiso l h Biology 36:933-945. Canadian Journa f Zoologo l y 69:436—442.