University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

2 - Second Eastern Wildlife Damage Control Conference (1985) Eastern Wildlife Damage Control Conferences

September 1985

DOUBLE-CRESTED CORMORANT DAMAGE TO A COMMERCIAL FISHERY IN THE ,

Scott R. Craven University of Wisconsin-Madison

Esther Lev University of Wisconsin-Madison

Follow this and additional works at: https://digitalcommons.unl.edu/ewdcc2

Part of the Environmental Health and Protection Commons

Craven, Scott R. and Lev, Esther, "DOUBLE-CRESTED CORMORANT DAMAGE TO A COMMERCIAL FISHERY IN THE APOSTLE ISLANDS, WISCONSIN" (1985). 2 - Second Eastern Wildlife Damage Control Conference (1985). 11. https://digitalcommons.unl.edu/ewdcc2/11

This Article is brought to you for free and open access by the Eastern Wildlife Damage Control Conferences at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in 2 - Second Eastern Wildlife Damage Control Conference (1985) by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. DOUBLE-CRESTED CORMORANT DAMAGE TO A COMMERCIAL FISHERY IN THE APOSTLE ISLANDS, WISCONSIN by Scott R. Craven and Esther Lev

ABSTRACT eliminated perching and an old-fash- ioned scare-crow. National Park Service The endangered classification of policy precludes direct control of the double-crested cormorant (DCC) the increasing DCC population. in Wisconsin resulted in complete protection and significant management efforts in the 1970's., These efforts, INTRODUCTION probably coupled with reduced pesticide Double-crested cormorants loads, resulted in a resurgence of (Phalacrocorax auritus) have conflicted Wisconsin cormorant populations from with commercial fisheries along the a low of 66 pairs in 1972 to 1028 coast of Maine, in Wisconsin, in the pairs in 1982» The DCC was reclassified Great Lakes region in general (Matteson as a threatened species in 1982. 1983)> and perhaps in other areas. This apparent success story did not Such conflict was especially evident take into consideration the potential during the period 1920-45 and again negative impact of an abundant pisci- in recent years. Reports of cormorant vorous bird. In 1978 a colony of (DCC) depredations have been propor- DCC's became established on a remote tional to changes in cormorant numbers in the Apostle Islands throughout much of their North American National, Lakeshore, in . range. From 17 pairs in 1978 the colony in- Cormorants were abundant throughout creased to 289 pairs in 1985. By the Great Lakes region throughout 1982, commercial fishermen in the the 1800s (Lewis 1929). However, Apostle Islands began to complain there were no reports of DCC colonies about damage to the valuable catch in the Great Lakes in the early 1900s. of Lake Whitefish. They accused DCCs By the 1920s DCC numbers and colonies of feeding within pound nets and thus began to increase and as they did, causing sub stantial damage by gilling so did complaints from commercial and scaring captured whitefish. Annual fishermen. Persecution by fishermen loss was estimated at $5-10,000 distri- during the 1940s followed by the bio- buted amongst 3 fishermen. accumulation of DDT, DDE, DDD, PCB, The interaction of DCC's and the and other contaminants between 1950 whitefish fishery was studied from and the early 1970s combined to devas- 1983-84. Food habits data did not tate the DCC population of the Great suggest that commercial fish species Lakes. Subsequent protection and were important to the diet of DCC's management reversed these trends during in the Apostle Islands. Observations the 1970s. Vermeer and Rankin (1984) suggested that the attraction of pound present an excellent review of historic nets centered more on the use of net DCC population trends. support poles for perch sites than Prior to the 1890s DCCs inhabitated on the availability of food within the isolated and larger lakes of the the net. Nine abatement techniques northern and central parts of Wisconsin were tested. Damage was reduced for (Carr 1890). The number of colonies periods of up to 4 weeks by a combina- in Wisconsin increased substantially tion of structural modifications that between the 1920s and mid-1950s (Mat- teson 1983)o Between 1923 and 1966 ®Scott Craven, Extension Wildlife cormorant colonies were observed in Specialist and Associate Professor, 16 Wisconsin counties (Anderson and Department of Wildlife Ecology, Univer- Hamerstrom 1967, Scharf 1979). As sity of Wisconsin-Madison 53706 noted, by the mid-1960s the Wisconsin Esther Lev, Project Assistant, Depart- cormorant population was reduced by ment of Wildlife Ecology, University pesticide contamination, human perse- of Wisconsin-Madison 53706 cution and habitat loss. In 1972,

14 with a total state population of 66 the development of depredation abatement pairs (Matteson 1983)? the DCC was techniques. listed as an endangered species. We thank Bruce Swanson (WDNR) for Positive management practices such logistical support; Dennis Pratt (WDNR) as erection of artificial nesting for identification of fish; Scott structures were vigorously pursued Hulse (WDNR) for collection of fish; by the Wisconsin Department of Natural and Sumner Matteson (WDNR) for field Resources (Meier 1981). Only a decade assistance, constructive criticism, later the number of nesting pairs and a wealth of information. Commercial had increased to a minimum of 1028 fishermen Morris Boutin, Jack Erikson, and the bird's status was downgraded and Dean Halvorson were extremely to threatened. There are no documented cooperative and willing to test abate- explanations for the dramatic increase ment devices and provide catch data. in population, however, reduced pesti- James Selgeby (USFWS) assisted with cide loads in the environment, immigra- the identification of otoliths. The tion from other DCC populations, manage- NPS Apostle Islands National Lakeshore ment, protection, and other factors office, especially Park Ecologist are probably all involved. Merryll Bailey, provided valuable In 1978, 17 pairs of DCCs were logistical support. Terry Daulton found nesting on in the deserves special thanks as a hard- Apostle Islands National Lakeshore working and dedicated field assistant (AINL) in Lake Superior under National in 1983 and 1984. Donald Rusch, Leader, Park Service jurisdiction,, By 1985 WCWRU, provided the boat that made the Gull Island colony had increased field work possible. J. Philip Keillor, to 289 nesting pairs and at least Sea Grant, provided many of the initial one satellite colony had been establ- ideas and contacts that led to this ished. In 1980, only 3 years after study. Financial support for this the Gull Island colony was discovered, research was provided by the U.S. commercial pound net fishermen in National Park Service and the Department the Apostle Islands began to complain of Wildlife Ecology, University of about cormorant depredations. Fishermen Wisconsin-Madison. claimed losses of 30-40$ of the white- fish catch in 1982 due to direct consump- STUDY AREA tion, scarring or gilling caused by The principal area of research DCCs (B. Swanson, pers. commun.). was the Apostle Islands National Lake- (Note: Gilling results when fish shore, Bayfield, Wisconsin (Fig. 1). become entangled in the pound net The 22 Apostle Islands, 20 of which mesh.) Five pound net fishermen were are part of the National Lakeshore, affected in 1982 and at least 40 pound are located off the tip of the northern nets which provided 60-70$ of the Wisconsin mainland in Lake Superior. commercial fisherman's income were The twenty islands comprise 15,778 involved in the depredation problem hectares of land, ranging in size (Mary Halvorsen, pers. commun.). from Gull Island (1.2 ha) to Stockton Thus the problem was viewed as serious Island (4,021 ha). The Apostle Islands by local resource managers. Fishermen lie in the transition zone between attempted to abate depredations with northern boreal coniferous forest rubber snakes, wind wheels, brightly- and deciduous forest. Sugar maple colored flags, eagle decoys, pieces (Acer saccharum), yellow birch (Betula of metal, and covered nets with no lutea), hemlock (Tsuga canadensis), success. red pine, (Pinus resinosa), white In response to National Park Service birch (Betula papyrifera), white cedar (NPS) concern about the AINL depreda- (Thuja occidentalis), balsam fir (Abies tions problem, we initiated a study balsamea) and black spruce (Picea of the DCCs in the Apostle Islands mariana)' are common trees (USNPS 1983). in 1983. Data were collected on the The understory of several islands food habits and ecology of DCCs and is dense Canada yew (Taxus canadensis).

15 METHODS colony, placed in plastic bags, and Field research was initiated on frozen for later identification. July 11, 1983» Prior data were col- Cormorant pellets were collected lected on the biology of the DCC in from Gull and Eagle Islands (Fig. 1). the Apostle Islands by USNPS biological Pellets are a mass of indigestible aids in cooperation with WDNR staff. material enveloped in mucus and expelled Collaboration between the Department by adult and subadult DCCs. Young of Wildlife Ecology, USNPS, and WDNR birds do not begin producing pellets personnel and 3 commercial fishermen until they are able to fly (Ainley continued throughout the study. Obser- and Kelly 1981). Cormorants typically vations of DCCs and data on their produce one pellet daily and these ecology, food habits and interaction pellets are comparable to regurgitation with the commercial whitefish industry samples as indicators of diet (Jordan were collected from May-October, 1983 1959). In many cases food remains and May-September, 1984. were digested to the point where fish Four aerial surveys of the Apostle species could not be identified by Island's cormorant population were gross examination. Thus, we used conducted between July 17 and September otoliths as a diagnostic tool (Ainley 2, 1983. Each 2-hour survey was flown and Kelly 1981). Fish otoliths (calcare- in the morning around all of the is- ous concretions in the inner ear) lands. Number of birds observed and were easily separated from pellets. location were recorded„ Biweekly Since each species of fish has distinc- cormorant counts were conducted by tive otoliths which are unaffected boat in both years of the study. by digestion, we could identify fish Boat counts concentrated on numbers species in all samples. The number of birds at pound nets and at key of otoliths of each species was divided- feeding areas. Pound nets in the by 2 to estimate the number of fish north and south sectors of the AINL eaten (Ainley and Kelly 1981). A were censused on alternate days. reference collection of otoliths of Cormorant observations reported by common Lake Superior fish was developed. USNPS and WDNR employees were collected Samples of fish were caught, identified, and mapped. and measured, then the otoliths were removed, mounted, and labeled. One hundred and eighty four DCCs were banded on 3 trips to Gull Island To further document food habits, (11 July, 8 August 1983 and 12 July a 38 m (125 ft) experimental mesh 1984). Prefledging 3-7-week-old birds gillnet was set in areas where cormor- were banded with standard size 8 USFWS ants were observed feeding. Mesh aluminum leg bands. Sixty-four of size of experimental net increases the birds were also banded with 3 in roughly 1 cm increments from 2.5- different colored aluminum leg bands 8.9 cm along the length of the net, for individual identification. Flight- allowing sampling of diverse fish less young were herded into a group, species and size classes.. Gillnets individually captured, and placed were set for one or two nights in in nests covered with burlap to protect water depths of 4.5-6 m. Species them from overheating. Nests with and size of fish caught were recorded. eggs were also protected. Birds were Nine deterrent (abatement) techniques aged (Canadian Wildlife Service 1977) were tested during the 2 field seasons but not sexed. (Table 1). Pound nets were selected Food habits of both immature and to receive abatement techniques on adult DCCs were studied by observation the basis of cormorant activity, gilling of feeding activity and collection rates, and the availability of a second of food remains. Adult and well deve- net to serve as a control. Each tech- loped immature birds regurgitate stomach nique was evaluated for a minimum contents when disturbed or frightened of 2 weeks unless the technique was (Lewis 1929)o Samples of regurgitations clearly ineffective. Bird activity were collected at the Gull Island at nets with deterrent devices was

16 monitored for two weeks prior to place- Sexually immature DCCs concentrated ment of the device and continued after at several other points in the AINL placement. Each experimental net away from Gull Island. , was paired with an adjacent unprotected 55 km from Gull Island, was the primary net. Location, time, number of birds, roosting area for nonbreeders. The activity of the birds, and other avian interior of Eagle Island (9.7 ha) species in the area were recorded. is dense white and yellow birch and When possible, data on the number Canada yew. Rocky ledges around the and species of fish caught and gilled perimeter of the island provide excel- were collected by observing the commer- lent roosting sites. Up to 170 DCCs cial fishermen as the nets were lifted. were observed at Eagle Island at any Additional data were provided by WDNR one time with smaller groups of 20- staff and the commercial fishermen. 40 frequently observed at Little Manitou and Hermit Island Rock. In June of RESULTS 1984, 3 nests, one with 2 eggs and 2 empty, were found on Eagle Island. Colony Size and Production These nests constituted the first In 1984, Gull Island, the smallest known nesting attempt on Eagle Island. of the Apostle Islands (Fig. 1) was Neither of the 2 eggs hatched. In the only known double-crested cormorant 1985, at least 10 nests with young colony in the area. Gull Island lies were found in trees on Eagle Island, only 30-70 cm above lake level and in addition to several ground nests. consists of a pebble and stone substrate with scattered growth of bluejoint Food Habits grass (Calamagrostis inexpansa), red Double-crested cormorants appear elderberry (Sambucus pubens) and moun- to be opportunistic feeders, feeding tain maple (Acer spicatum). upon the most available and abundant From the first discovery of DCCs fish source at a given time. One on Gull Island in 1978, the colony hundred and fifty regurgitation and increased dramatically to 254 nesting pellet samples were collected and pairs in 1984. The slight increase analyzed during the 1984 field season. of 11 pairs over the 243 pairs present Thirteen species of fish were identified in 1983 suggests that the size of in food remains but no single fish the colony and its growth rate may species dominated DCC diet in the be leveling off. The 1984 production Apostle Islands (Table 2). Small rate of 1.67 young per nest (fledged) forage fish, ninespine sticklebacks was 2.3 times greater than the 1983 (Pungitius pungitius), slimy sculpin rate of 0.73 young per nest. Based (Cottus cognatus), spoonhead sculpin on weekly counts, the estimated 1984 (Cottus ricei), and burbot (Lota lota) population of adult DCCs (including are the most frequently taken species. nonbreeders) in the AINL was 700, We found lake whitefish and lake trout up 100 from 1983. (Salvelinus namaycush), the two impor- The chronology of the nesting season tant commercial species of the area, varied between years. In 1984 DCCs in only 2% of the samples. DCCs gener- arrived in Chequamegon Bay on 15 April, ally eat fish 12-15 cm long (Bartholomew all young had fledged from Gull Island 1942) and this was true for most samples by 28 August, and DCCs began to move we could distinguish. to staging areas in Chequamegon Bay Flock feeding behavior described by 28 August. In 1983, DCCs arrived by Bartholomew (1942) was not observed on Gull Island in mid-April, flightless until October 1983 and August 30, young were still present in the colony 1984 when birds moved to a staging in mid-August, and birds were not area in Chequamegon Bay. Lack of observed on staging areas until 3 this behavior may be related to the October. Overall, the nesting season fact that the commonly eaten fish was about 3 weeks earlier in 1984 species in the Apostle Islands do than 1983= not school. Preferred feeding sites

17 in the AINL tended to be shallow (5- pers. commun.). 18 m deep) sandy or shoal areas close Gilling caused an estimated economic to island shorelines (Fig, 4). DCCs loss of about $7000 for the period have been observed diving to depths June-August 1983 and $4500 for the of 22 m (Lewis 1929), however, they period July-August 1984 (Table 4). tend to prefer feeding in shallow These estimates were based on a dockside water. value of $0.50 per pound for whitefish. Comparison of food samples between Deduction of 5% of the total for normal Gull and Eagle Island suggest that gilling losses yielded an estimate breeding adults and nonbreeders had of the economic impact of DCCs. How- a similar diet (Table 3). The four ever, some of the gilled fish are species most frequently eaten in both sold as second quality fish or smoked areas are small forage species rather and sold. Thus the fishermen recover than commercial species. some of the loss. These data suggest that the total loss for both 1983 Abatement Techniques and 1984 would likely not exceed $15,000 Nine abatement techniques were distributed amongst 3 fishermen. tested. Each technique, the length of the test, and the generalized DISCUSSION success/failure of the technique are summarized for ease of reference (Table DCC Population and Interactions With 1). In general, DCCs adjusted to Other Species all abatement devices within 4 weeks The Gull Island colony increased or less of installation. DCC activity in 1984 for the 7th consecutive year. at nets dropped or was eliminated However, the rate of increase began following installation of electric to decline in 1983 and 1984 and the shockers, nails, cones, and scarecrows pattern of increase suggests that (dummies), with or without boats. the colony followed a typical pattern Over time, bird activity approached of logistic growth (Fig, 2). The pre-abatement levels. Combinations 1985 colony size of 289 pairs, deter- of techniques designed to make certain mined after the previous analysis parts of the pound nets inaccessible was completed, further supports a for DCCs (cones, nails, electric leveling off of the Gull Island colony. shocker) and scare devices (scarecrow/ Extrapolation of the observed trend dummy) were the most successful suggests that colony size will stabilize approach. The Av-Alarm device and at about 325 nesting pairs (Fig. 2). models of predatory birds were inef- The 1984 production rate of 1.67 young fective. The Av-Alarm proved to be per nest was 2.3 times greater than incompatible with the NPS wilderness the 1983 rate of 0.73° Since cormorants objectives for the AINL. do not nest when 1 year old, the low production (we assume this rate is Damage to Commercial Fishery low in the absence of data from other Throughout the summer commercial years at Gull Island) cannot be respon- pound nets were monitored to determine sible for the small increase in colony the percentage of gilled whitefish. size between 1983 and 1984. We believe Fishing records before cormorant prob- disturbance on Gull Island was the lems began suggest a natural gilling primary explanation for the depressed rate of 5% in the absence of cormorants production in 1983. (B. Swanson, pers. commun.). The Gull Island supported both a cor- average gilling rates for July and morant and a herring gull colony and August, in excess of the baseline the herring gull colony increased 5%, were essentially identical for over the same period the DCC enlarged 1983 (33.9?) and 1984 (31.8%). Gilling (S. Matteson, pers. commun.). When rates were also comparable to the disturbed, the adult cormorants flushed, 35$ loss figures estimated for the circled overhead and then landed on 1981 and 1982 seasons (B. Swanson, the water several hundred meters away,

18 where they awaited cessation of the Comparison of food samples between disturbance. They remained wary even Gull and Eagle Island suggest that after the departure of human intruders breeding adults and nonbreeders had and were slow to return to the colony, a similar diet. One obvious difference, leaving their young unattended for the prevalence of nine-spined stickle- 20-60 minutes. Herring gulls are backs at Gull Island is an artifact known to prey on unattended cormorant of our technique. Samples from Gull eggs and young (Ellison and Cleary Island were both regurgitations and 1975)o Ellison and Cleary (1975) pellets; only pellets were collected reported that gull predation in cor- at Eagle Island„ Nine-spined stickle- morant colonies often resulted in backs could only be identified in nest abandonment and failure. Other regurgitations; their microscopic studies have linked human disturbance otoliths could not be detected in with increased gull and crow predation pellets. This species was abundant on DCC eggs and young (Mendall 1936, in the waters around both sites (USFWS, Drent et al. 1964, Vermeer 1970, Lock WDNR) and we suggest sticklebacks and Ross 1973, Kury and Gochfeld 1975). are, in fact, eaten in both areas. Ten visits to Gull Island by biologists The top four species consumed in both between May and September and the areas are small forage species rather presence of a red fox (Vulpes fulva) than commercial species. Cormorants which was removed added to the distur- do not appear to use fish of commercial bance on Gull Island in 1983« species and size as a food source. Eagle Island (9.7 ha, Fig. 1) was We suggest that the attractiveness the primary roosting area for non- of the nets as perch sites results breeding DCCs and it also supported in gilling or damage to valuable fish the only Great Blue Heron rookery simply due to birds presence. Natural in the Apostle Islands. Heron nests perch sites are virtually non-existent were located in white and yellow birch in the waters around the islands. and a few spruce and fir over a dense We observed cormorants diving in the understory of Canada yew; about half pots of the nets but they rarely sur- of the trees were dead. As noted faced with fish in their mouths. previously, 3 DCC nests were found Generally, the fish in the nets were during surveys of Eagle Island in too large for the cormorants to eat. 1984. No DCC nests were found during A DCC may chase and subsequently gill heron surveys on Eagle Island in 1982 fish as a simple stimulus response or 1983. The presence of 10 DCC nests to movement below it while it is per- in trees on Eagle Island in 1985 sug- ched. In early September 1983 and gests that DCCs produced at Gull Island 1984 DCCs were observed perched at may begin pioneering new colonies nets where the pots had been pulled. when they reach breeding age. Installation of alternative perch sites near active fishing nets with Herons and cormorants are known deterrents would provide more infor- to coexist in the same colony in other mation on this theory and could serve areas with herons nesting in the upper as an additional abatement technique. parts of the trees and cormorants However, due to cost and potential nesting in lower limbs. However, navigational hazard alternative perches there are also examples, notably from were not installed. New York (S. Matteson, pers. commun-), where DCCs have taken over great blue The incidence of gilling increased heron rookeries and displaced the as young DCCs fledged and more birds herons. Eagle Island will require were observed around pound nets in annual surveillance for interspecific late summero Their inexperience may competition. Such new colonies suggest also contribute to the problem as increased depredations problems for they try to capture fish seen as they commercial fishermen. perch on pound nets. Most of the fish which appeared frequently in Food Habits the DCC diet are small, shallow water,

19 bottom species. Nine-spine stickle- and habitat improvement (e.g., reduced backs, sculpins, menorainee whitefish, pesticide loads, etc.) have enhanced burbot, and longnose suckers are all cormorant populations throughout the common in waters less than 10 m. Great Lakes (Scharf and Shugart 1981). Smelt and herring are also very abundant Thus we cannot conclude that the cor- but may swim too fast for DCCs to morant population in the Apostle Islands utilize regularly. Whitefish frequent is "unnatural"; a classification which deeper water and are very sensitive would allow greater management flexi- to warm water. Thus they may be un- bility under National Park Service available to DCCs except around pound policy guidelines. nets. An attempt to set an optimum popu- lation size is not warranted if a MANAGEMENT RECOMMENDATIONS "hands-off" management approach is to be taken. However, if management Cormorant Population is undertaken in the future because As described, the double-crested of adverse impacts on other species, cormorant population on Gull Island such as great blue herons on Eagle increased rapidly from 1978 through Island, then we suggest that a popu- 1982. From 1982-1985 population growth lation of about 300-350 would be suf- slowed and began to level off. The ficient to ensure the safety of the curve produced by plotting the number DCCs in the AINL and minimize necessary of nesting pairs over time closely control (management). The population approximates a typical logistic growth data collected from 1978-1984, suggest curve. Unpredictable factors such that the colony is most productive as food supply, weather, disease, at a level of about 200 and would human intervention, and perhaps others, thus require the maximum level of could result in substantial annual control (removal of birds). This fluctuations in the nesting colony. would become an annual burden. If However, in the absence of such factors in fact the colony does begin to level these data suggest that the Gull Island off in the next 4-6 years, as the Colony will stabilize at about 325 data suggest, then control in terms nesting pairs. of bird removal would be kept at a To the human eye, it appeared that minimum. space for nests is not limiting colony If the new colony identified on size on Gull Island. The colony of Gull Island is indicative of coloni- herring gulls on Gull Island also zation of additional islands, then increased by about 50$ during the population management would have to period of growth in the cormorant be expanded. colony (S. Matteson, pers. commun.). Since the gulls nest several weeks Conflict with Fishery earlier than the cormorants, their The key resource management problem nesting territories may exclude the involving the cormorant population cormorants from parts of Gull Island. in the AINL is the conflict with the Herring gulls may also exert more commercial fishery. Strong emotions, predatory impact on cormorant eggs limited abatement techniques, legal and young birds as their numbers in- restrictions, and other considerations crease. Thus, herring gulls may be make the fishery/cormorant issue very partially responsible for the ultimate difficult to deal with. Recognizing size of the cormorant colony. However, these facts we offer the following it is impossible to predict exactly recommendation/analysis for a variety what factors are causing the cormorant of solutions: population to stabilize. Human activity 1. Population reduction: The cor- cannot be implicated in either the morant colony on Gull Island establishment or stabilization of (or any newly identified nesting the Gull Island colony, except to islands) is vulnerable to a the extent that protection, management, variety of techniques to reduce

20 the number of cormorants including the poles and rigging of shooting, nest destruction, pound nets is as an ideal and excessive disturbance,, perch rather than as a feeding Additionally, non-breeding cor- site. Installation of poles morants or free-flying adults in suitable cormorant feeding from the colony could be shot areas away from active pound or trapped at or around roost nets could occupy some birds sites or pound nets. that might otherwise be a Recommendation; While this problem at net sites. The may be a popular solution with navigational hazard posed the fishermen, direct population by additional poles or floating reduction is incompatible with perch sites would have to NPS policy and the legal status be resolved. ("threatened") of the cormorant d. A modified acoustic scare in Wisconsin. It would also device is being developed be disasterous for public rela- at the University of Wisconsin tions. It should not be consi- for Canada goose damage con- dered. trol. The device senses 2. Damage abatement: A variety the presence of birds by of techniques to reduce fish "hearing" their calls or losses to cormorants were tested detecting motion. Then and and reviewed in the text. No only then does it activate single technique was 100? effec- either a propane exploder tive, but rarely is any abatement or an amplified distress technique 100? effective in call. This would maintain animal damage control work. the wilderness tranquility Success must be measured in of the AINL but provide a the reduction of loss rather scare device when it is needed. than in terms of elimination. 3° Compensation: Wisconsin adopted Several techniques did reduce new wildlife damage control loss in test nets. legislation in 1983° The program Recommendation: provides abatement assistance a. Combine the techniques that and direct compensation if abate- showed promise; e.g., a dummy ment fails for damage caused or dummy/boat combination by deer, bear, and Canada geese. in conjunction with metal A segment (3%) of the Wisconsin cones on poles and an elec- Endangered Resources Fund (tax trified wire or porcupine contribution program) is also wire on the horizontal supports earmarked for endangered wildlife should reduce loss to tolerable damage compensation. The prece- levels. dents, guidelines, and financial b. Concentrate use of abatement resources of these 2 programs techniques during periods suggest that some form of compen- sation for cormorant damage of peak gilling (July/August). be examined as a possible solu- Cormorants demonstrated substan- tion. tial adaptability and tended Recommendation; to ignore scare devices after repeated exposure. Thus a. Have all parties (WDNR, NPS, they should be used only fishermen) agree on an accep- when they can provide maximum table baseline level of gilling benefit. and financial loss. c. Investigate the use of alter- b. Consider the possibility nate perch sites as a means of direct compensation for of "diluting" the problem. loss (total or prorated de- Observations suggest that pending on available funds) the primary attraction of or in-kind compensation in 21 the form of longer seasons (Phalacrocorax bougainvillii L.) or a relaxation of other y un metodo para estimar la ingestion regulations compatible with diaria, Bol, Cia. Adm. Guano 35 fishery management objectives„ (4):23-4O, Abatement devices could be KURY, C. R,, AND M, GOCHFELD. 1975, purchased for the fishermen Human interference and gull predation as a cost-effective use of in cormorant colonies. Biol. limited funds in keeping Conserv, 8:23-33, with the "try abatement first" LEWIS, H. F. 1929, The natural history philosophy of animal damage of the double-crested cormorant» control in Wisconsin,. Ru-mi-Lou Books, . 94pp. Co Compensation or incentives LOCK, A. R,, AND R. K, ROSS, 1973° could also be offered to The nesting of the great cormorant encourage fishermen to use (Phalacrocorax carbo) and the double- alternate fishing techniques crested cormorant (Phalacrocorax less susceptible to cormorant auritus) in Nova Scotia in 1971° damage or to avoid fishing Can, Field-Nat, 87:43-49, in areas of cormorant activity. MATTESON, S. Wo 1983, A preliminary In conclusion, an integrated approach review of fishery complaints asso- of abatement, possible compensation ciated with changes in double- programs or incentives, and long-term crested cormorant populations in monitoring of the APIS cormorant popula- Maine, Wisconsin, and the Great tion should allow the commercial white- Lakes Region. Rept. to Bur, End. fish fishery and the double-crested Res, Wisconsin Dep. Nat, Res,, cormorant population to coexist, Wisconsin, 16pp. MEIER, Thomas I. 1981. Artificial LITERATURE CITED nesting structures for the double- AINLEY, Do Go, AND P. R. KELLY, 1981, crested cormorant. Technical Bul- Feeding ecology of marine cormorants letin No. 126, Dept. of Nat, in southwestern North America, Resources, Madison, WI. Condor 83:120-131, MENDALL, H, L. 1936, The home life ANDERSON, D, W., AND F, HAMERSTROM. and economic status of the double- 1967, The recent status of Wisconsin crested cormorant (Phalacrocorax cormorants. Pass, Pigeon 29(1):3- auritus auritus) (Lesson). Maine 15, Bull, 39:1-159, BAILLE, J. L, 1947, The double- SCHARF, W, C. 1979« Nesting and crested cormorant nesting in Ontario, migration areas of birds of the Can, Field-Nat, 61:119-126, U.S. Great Lakes (30 April to BARTHOLOMEW, G. A,, JR, 1942. The 25 August 1976). U,S. Fish and fishing activities of double-crested Wildlife Service Off, Biol. Serv. cormorants on San Francisco Bay, FWS/OBS-77/2- 113pp° Condor 44(1):13-21, , AND G, W, SHUGART. 1981. CARR, C. F, 1890, A list of the Recent increases in double-crested birds known to nest within the cormorants in the United States boundaries of Wisconsin, with a Great Lakes, American Birds 35(6): few notes thereon, mimeo, 910-911o DRENT, R,, G, F. VAN TITS, F, TOMPA, VERMEER, K, 1970. Some aspects of AND K, VERMEER, 1964, The breeding the nesting of double-crested cor- birds of Mandarte Island, Can. morants at Cypress Lake, Saskatch- Field-Nat, 78:208-263. ewan, in 1969; a plea for protection. ELLISON, Lo No, AND L, CLEARY, 1978, Blue Jay 28:11-13, Effects of human disturbance on , AND L. Rankin, 1984, Popula- breeding of double-crested cormor- tion trends in nesting double- ants. Auk 95:510-517o crested and pelagic cormorants JORDAN, R, 1959» El fenomeno de in Canada, Murrelet 65(1):1-9= las regurgitaciones en el Guanay

22 Table 1. Results of abatement techniques tested in the Apostle Islands, 1983-8*4.

Technique Location Trial Period Results

Av-Alarm Madeleine Island 1 week Not successful. Cormorants observed (audio scare device) perching within 7 feet of speaker. Also poor public acceptance in populated areas.

Electric-shocker South Twin Island 2 months Successful at keeping cormorants (electrified wires) Rocky Island from perching. Little Squaw Bay Raspberry Bay

Metal cones Raspberry Bay 1 month Successful at keeping birds off (on tops of poles) poles. Best used in combination with another technique to keep birds off the rest of the net.

Nails South Twin Island 2 months Successful at keeping birds off (same purpose as Rocky Island poles. Best used in combination cones) Sand Island with another technique to keep birds off the rest of the net.

Owl decoy Raspberry Bay 2 days Unsuccessful. Birds observed (scare device) Raspberry Island perching next to decoy within 2 days.

Mylar helium balloons Frog Bay ' 2 weeks Unsuccessful alone. Best used (scare device) Hermit Island in conjunction with scarecrow.

Hanging scarecrow Roys Point Successful for 1 month. After Raspberry Bay t weeks birds were observed South Twin Island perched on poles.

Boat Hermit Island 3 weeks Successful for 2 1/2 weeks. (floating in pot of Best used in conjunction with net) scarecrow.

Scarecrow/Boat 6 weeks Successful. No birds observed Kapunky Bay at net for five weeks. Rocky Island Reduced gilling rate. Best used in combination with metal cones and mylar helium balloons.

K • 324

300 _ (286)-^ ©

O254 OBSERVED POPULATION SIZE

LOGISTIC PREDICTION 200 K 1 + Ae-'ml

1 K = 324 A = 18.041

r .- 0.818 APDSTLE ISLANDS

1 I I I ! L 1978 Fig. 1. Apostle Islands area, Lake Fig. 2. Logistic growth curve fitted Superior, Wisconsin to observed cormorant numbers.

23 Table 2. Frequency of occurrence of fish found in regurgltatlon Table 3. Frequency of occurrence of fish found In 150 coraorant and pellet samplos of doublo-erosted cormorants, Apostlo Islands, regurgltatlon and pellet oaaploa oollected from cull Island and Wisconsin, 1984. Eagle Island, 1984.

Species Percent Eagle Island Species Species Hlneaplne Stickleback 16 .6 Slimy Sculpin Nlnesplne Stickleback 20.2 15 .5 Slimy Sculpin 34 .0 Burbot Spoonhead Sculpin 13 .It 15.5 Spoonhead SouLptn 24 .0 Sllny Sculpin 11.4 Burbot 13 • 3 Henomlnee tfhltefish 8.3 Uke Northern Chub Lake Northern Chub 10 11.2 Uke Northern Chub 6 .4 Spoonhead Sculpin Longnose Sucker e.6 10.4 Longnose Sucker 6.4 Trout Perch Trout Perch s.3 9.5 Unknown 5.5 Longnose Sucker Unidentified 5 9.5 done 3.7 Unknown Henomlnee Whlteflsh 2 .11 5.3 Burbot 3.7 Nothing Hone 1 .7 1.4 Common tfhltefish 2.7 Smelt Smelt 1 1.0 Trout Perch Lake Trout Lake Trout 1 1.0 Smelt

Cosmon Whlteflsh Henonlnee tfhltefiah 1.0 Lake Trout Common Uhitefish Herring .5 .6 Unknown Invertebrates Herring Worms .5 .6

1 Horn Sculpin Unknown Invertebrates .4 4 Horn Sculpin 98.7

Table 4, Estimated damage to the cocnerclal whitefish catch caused by ooraoranta In the Apostle Islands, 1983-60.

Date Total catch t allied Lbs lost 0 lost

June 1983 16,376 31.4 5137 2569 July 1963 26,123 32.0 6364 4182 August 1983 2.720 3SA 1049 5?5 Totals 45,219 14,590 7276

June 1984 31,142 36.3 6845 3»J2 August 1984 7.800 37^6 2074 1037 Totals 38,942 8919 4460

24