During July-December 1978-1984, Bonaparte's Gulls (Larus philadelphia) on annual migration through the Quoddy region off New Brunswick, Canada, fed on fish, euphausiids, insects, and other marine invertebrates (mainly polychaetes and amphipods) in varying proportions as the summer and autumn progressed. The seasonal variation in the diet was related directly to food availability, which, in turn, was dictated by natural cycles of prey in the region. The frequency with which plastic particles were found in the stomachs also had a statistically significant seasonal trend. Fish provided the largest energy contribution (75-91%) to the diet at all times in the region. There was no significant difference between diets of juvenile and adult birds. Received 23 April 1986, accepted 15 August 1986.

Department of Zoology, University of Guelph, Guelph, Ontario NIG 2W1, Canada BONAPARTE'S Gull (Larus philadelphia) is a widely distributed North American species. The Quoddy region of New Brunswick in the Bay of Fundy, Canada, is a major autumn staging ground for migrating Bonaparte's Gulls, result- ing in the largest known concentration of this species in eastern Canada (Canadian Wildl. Serv. 1979). The anomalous tidal regime of the Quoddy region results from interaction of large- amplitude, semidiurnal tides with the complex physiography of the area; these conditions af- fect the local concentration and distribution of zooplankton, thereby creating important feed- ing sites for fish, seabirds, and marine mam- mals (Smith et al. 1984). This localized concen- tration of food, in combination with ease of access to the area, creates a favorable situation for intensive study of Bonaparte's Gull (Braune and Gaskin 1982a). The diet of the Bonaparte's Gulls in the Quoddy region during September 1978 and August 1979 was summarized by Braune and Gaskin (1982a). The objective of the present study was to evaluate seasonal aspects of the diet of Bonaparte's Gulls during autumn mi- gration during 1978-1984, and to compare prey types with those taken by the birds in other areas of North America. METHODS Collection of birds.--As part of a study on heavy metal accumulation in the marine ecosystem, 222 Bona- parte's Gulls were collected by shotgun (under Sci- entific Kill Permits issued by Canadian Wildlife Ser- vice) from 22 July through 28 December 1978-1984 in the Quoddy region off southeastern New Bruns- wick, Canada (Fig. 1). The birds were pooled into 4 periods of approximately 6 weeks each (Table 1). Birds were collected at various times of day on flood and ebb tides. An adult gull also was obtained from the breeding grounds near Churchill, Manitoba, on 8 August 1982. The proventriculus and gizzard of each bird were removed as a unit and the contents pre- served in 10% formalin in 1978 and 70% alcohol in 1979-1984. Seven stomachs were empty. The birds were then aged as juvenile or adult by plumage (Grant 1982) and sexed by examination of gonads. Diet analysis.--Food items from collected stomach contents were categorized into four major groups: insects, euphausiids, other marine invertebrates, and small fish. Insects were identified to order (Borror and White 1970), and euphausiids and most other marine invertebrates were identified to species (Gos- her 1971). Fish were assigned to species based on morphological characteristics (Leim and Scott 1966) and otoliths (Smith and Gaskin 1974). Numbers of partially digested fish were calculated based on counts of vertebrae and otoliths. To minimize bias resulting from birds feeding op- portunistically on one food type to the exclusion of most others at the time of sampling, stomach-content data were pooled over the years to gain average rep- resentative dietary trends as well as a temporal se- quence (July-December) of dietary information. Data are presented as percentage frequency of occurrence of major food types among stomach-content samples. Food items unidentifiable to order or species were included only in the respective major prey category. The G-test of independence using Williams's cor- rection (Sokal and Rohlf 1969) was used to test for Fig. 1. The Quoddy region in the southwestern Bay of Fundy off New Brunswick, Canada. differences in diet between juvenile and adult Bo- naparte's Gulls collected from the same feeding flock. For each of six disparate samples containing both ju- veniles and adults, G-tests were performed separate- ly for each of the four major food types (insects, eu- phausiids, fish, other invertebrates) based on raw frequency of occurrence data. Chi-square tests were used to test for seasonal changes in the frequency of occurrence of the food types in the stomach-content samples collected over the four time periods. For all statistical analyses, the level of significance was set at P --- 0.05. RESULTS With the exception of two isolated cases, there was no significant difference between diets of juvenile and adult Bonaparte's Gulls in the Quoddy region (Table 2), so data from both age groups were pooled. Euphausiids were taken more frequently than fish during July to early October, but fish were taken with significantly greater frequency during October-November than earlier in the season (Table 3). Other in- vertebrates were taken significantly more fre- quently as the autumn progressed (Table 3). The occurrence of insects decreased significantly until none were taken during December. The insects most commonly found in the stomachs of Bonaparte's Gulls were from the orders Coleoptera, Hymenoptera, and Diptera. Other than insects and euphausiids, poly- chaetes (Nereis virens) and amphipods (Gam- marus oceanicus) were found frequently. The stomach of the adult bird collected from the breeding grounds (Churchill, Manitoba) con- tained insects (Ephemeroptera, Trichoptera) exclusively. The frequency with which stones were found in the gizzards paralleled the frequency trend of the other invertebrates, and the presence of plastic particles (fragments and pellets: 1-4 mm diameter) increased significantly over the sea- son (Table 3). DISCUSSION Quoddy region.--Euphausiid surface swarms provide a highly concentrated and readily ac- cessible food source in the Quoddy region (Smith et al. 1984). Such surface swarms are particularly important for seabirds because most species, including Bonaparte's Gull, feed only at the surface (Brown et al. 1979). Surface swarming of euphausiids occurs most frequent- ly in August and September, with another peak during November-December (Fish and John- son 1937). The occurrence of euphausiids in the diet of Bonaparte's Gulls closely followed swarming trends (Table 3). Young herring (Clupea harengus) tend to con- centrate in inshore shoal areas, particularly in summer and autumn (Ridgway 1975), to feed in the upper water layers (Legare and Maclel- lan 1960), where they are readily available to feeding gulls. Species of fish found in the stomachs of gulls varied over the years; harbor pollock (Pollachius virens) replaced herring as TABLE 1. Numbers of Bonaparte's Gulls (J = juveniles, A = adults) collected from the Quoddy region during July-December 1978-1984. Year Total Period 1978 1979 1980 1981 1982 1983 1984 J A 22 July-30 Aug 7 17 30 23 10 14 73 31 Aug-9 Oct 18 25 22 21 31 55 10 Oct-18 Nov 20 5 12 2 35 19 Nov-28 Dec 11 1 12 Total 18 7 17 55 76 37 12 47 175 TABLE 2. Corrected G-values for comparisons of frequency of occurrence of major prey types in stomach contents between juvenile (j) and adult (a) Bonaparte's Gulls collected from the same flock in the Quoddy region on six occasions. a Aug 1981 Sept 1978 Sept 1981 Sept 1982 Oct 1983 Nov 1984 n i=9 n i=3 n i=6 n=11 n i= 11 n i=2 Prey type na = 10 n, = 8 na = 8 na = 11 na = 10 na = 10 Insects 4.37* 0.01 0.78 1.12 0.94 0.53 Euphausiids 0.39 2.44 0.78 1.60 0.00 0.05 Fish 0.00 0.21 0.00 3.96* 0.25 0.05 Other invertebrates 1.04 0.01 0.00 0.96 3.17 0.05 * = 0.025 < P < 0.05. the major fish prey species during 1980 and 1982. Harbor pollock were taken only during late July to early September (Table 3). The pol- lock is a larger, heavier fish than the herring (Kohler et al. 1970), and after September most harbor pollock are too large for gulls to capture and swallow easily. The highest concentrations of insects oc- curred mainly during mass emergences such as those of winged ants (Hymenoptera) during August-September. During the emergences, the birds fed on the insects mainly by "hawking" for them. At other times, insects were taken from the water surface layer or from floating weed fragments (Braune and Gaskin 1982a). As insect availability declined during the autumn, other invertebrate species, primarily poly- chaetes and amphipods that occur in large con- centrations locally inshore, were ingested more frequently (Table 3). The parallel trend in fre- quency of occurrence of stones probably can be attributed to their ingestion with polychaetes and amphipods along the shoreline. The number of prey individuals equal to the energy content of one herring (4 g) varied dra- matically among the different prey types (Table 4). Insects, euphausiids, and other inverte- brates may be exploited most effectively only when they occur in large, dense concentra- tions. When high prey densities occur, the gulls may settle on the water and seize surface prey (Braune and Gaskin 1982b). Fish contributed the most energy to the diet (Table 5). Euphausiids contributed more ener- gy to the diet during July to early October than from October to December. Insects were taken more frequently than their dietary energy in- put would suggest (cf. Tables 3 and 5), but overall, insects and other invertebrates contrib- uted little energy to the diet. Prolonged retention of fish otoliths and polychaete jaws presented a potential bias to TABLE 3. Percentage frequency of occurrence of prey types and nonfood items collected from the stomachs of Bonaparte's Gulls, July-December 1978-1984. Chi-square values are given for seasonal changes in fre- quency of occurrence of prey types and nonfood items found in stomach contents. a 22 July- 31 Aug- 10 Oct- 19 Nov- 30 Aug 9 Oct 18 Nov 28 Dec Stomach contents (n = 85) (n = 85) (n = 35) (n = 10) Insects b 49 41 37 0 9.58'* Euphausiids 65 68 43 70 7.40 Meganyctiphanes norvegica (51) (64) (31) (50) 10.53'* Thysanoessa inerrnis (34) (44) (20) (30) 6.33 Fish 58 32 71 60 20.13 * * * Clupea harengus (27) (16) (34) (30) 5.26 Pollachius virens (22) (1) (0) (0) 28.47*** Other invertebrates c 8 15 43 20 21.35'** Stones 4 16 23 I0 11.34'* Plastic particles 0 2 9 20 16.45'** ** = 0.01 < P < 0.025, *** = P < 0.001. b Ephemeroptera, Orthoptera, Hemiptera, Homoptera, Coleoptera, Trichoptera (larvae), Lepidoptera (eggs), Diptera, Hymenoptera, and un- identified pupae. c Polychaeta, Arnphipoda, Mysidacea, Gastropoda, and Bivalvia. TABLE 4. Mean wet mass (g/individual), energy content (kJ/g wet mass), and number equivalents based on energy values of major prey types found in the stomachs of Bonaparte's Gulls collected from the Quoddy region. Prey Mean bet wet Energy equiv- Prey mass content alents Fish Clupeidae Gadidae Other invertebrates Polychaeta Amphipoda Euphausiids Meganyctiphanes norvegica Thysanoessa inermis Insects 4.0 a 10.925 e 1 6.0 a 5.651 f 1.3 2.24 b 2.675g 7 0.138 b 3.910g 81 0.127 c 4.098 c 84 0.032 c 4.098 c 334 0.032 a 3.177g 430 ' Kohler et al. (1970); based on fish of 8 cm total length such as those found intact in the digestive tracts of gulls. b Based on individuals collected from areas with feeding gulls (Poly- chaeta: n - t0, Amphipoda: n = t0). c Kulka and Corey (1982). d Based on intact individuals found in proventriculi (n = 50). e Keiver (1982). f Wiens and Scott (1975). Cummins and Wuycheck (1971). the dietary information because counts of in- dividual prey often were based on numbers of these structures present. In a variety of seabirds, fish otoliths are retained for no more than 24 h (Furness et al. 1984). However, Bar-tailed Godwits (Limosa lapponica) retained polychaete jaws for 6-7 months (Smith 1975), and a Shy Albatross (Diomedea cauta) retained squid beaks (similar in composition to polychaete jaws) for 1-2 months (Furness et al. 1984). Retention time of such structures is likely to be species specif- ic, depending on grinding efficiency of the giz- zard and the presence of grit. Small stones and grit were found regularly in the muscular giz- zards of Bonaparte's Gulls, suggesting a shorter retention time for the jaws in gulls than in shorebirds (see Table 3). The dietary energy contribution of polychaetes probably was over- estimated because of the inclusion of retained jaws in counts of polychaete numbers (Table 5: Other invertebrates). The frequency with which plastic particles occurred in the stomach contents of Bona- parte's Gulls in the Quoddy region increased significantly throughout the summer and au- tumn (Table 3). As the availability of insects decreased and surface swarms of euphausiids and shoals of small fish became less frequent than in summer and early autumn, the birds began to exploit other food resources. This was apparent in the increased consumption of oth- er invertebrates, and probably also accounts for the increased occurrence of plastic particles in the diet. With the increased patchiness and de- creased availability of fish and euphausiids in the area, it is likely that the birds more fre- quently picked up items that appeared edible. Other areas.--Other studies on the diet of Bo- naparte's Gulls indicate that insects, shrimplike crustaceans, and small fish are common prey across North America (Table 6). In inland areas, such as nesting grounds near bogs and mus- kegs, the birds are mainly insectivorous, as suggested by Bent (1921) and Plough (1951) and by the stomach contents of the gull collected at Churchill, Manitoba. Bonaparte's Gull is less of a scavenger than other gulls, and although it shows little interest in chum (Rowlett 1980), bits of refuse and food scraps floating on the surface are sometimes ingested (Sprunt 1954, Burleigh 1958). Little, if any, plant material is taken (Bent 1921). A description of diet cannot be based solely TABLE 5. Average percentage energy contribution to the diet of prey types collected from the stomachs of Bonaparte's Gulls, July-December 1978-1984. 22 July-30 Aug 31 Aug-9 Oct 10 Oct-18 Nov 19 Nov-28 Dec Prey types a (n = 85) (n = 85) (n = 35) (n = 10) Insects 1 1 < 1 0 Euphausiids 15 18 5 8 Fish 82 75 85 91 Other invertebrates 3 7 10 1 See Table 3 for detailed description of prey types. For calculations of energy content, prey unidentifiable to order or species were pooled with the order or species containing the highest number of identified individuals in each of the 4 major prey categories found in a given stomach or over a specific sampling period. TABLE 6. Diet of Bonaparte's Gull across North America. Location Food Source Nelson Lagoon, Alaska Frederick Sound, Alaska Washington state Monterey Bay, California Florida Appalachicola R. mouth, Florida Georgia South Carolina Mahomet, Massachu- setts Nantucket Is., Massachu- setts New England Quoddy re- gion, New Brunswick Large shrimp Gill and Hall (1983) Insects, eu- L.H. MacIvor phausiids, (unpubl. data) sand launce Insects Bowles (1906) Insects, eu- Baltz and More- phausiids, john (1977) marine worms, fish Insects, crusta- Sprunt (1954) ceans, snails, fish Small fairy L. Atherton shrimp (pers. comm.) Insects, fish, Burleigh (1958) other ma- rine life Insects, crusta- Sprunt and ceans, Chamberlain snails, fish (1949) Amphipods L.H. MacIvor (pers. comm.) Sand launce Heil (1984) Insects Nuttall (1974) Insects, eu- This study phausiids, fish, other inverte- brates Fish Todd (1940) Emerald shin- Axtell (1959) ers Pennsylvania Upper Niaga- ra River, Ontario- New York Coast Fish, shrimp, Bent (1921) other crus- taceans, ma- rine worms Inland Insects Bent (1921), Plough (1951) on one or two sampling sessions, even from a localized area such as the Quoddy region. Food types may remain relatively consistent be- tween years, but prey species may vary from year to year, as illustrated by the consumption of herring and harbor pollock. The diet is sub- ject to seasonal variation in prey availability, which, in turn, is related to annual cycles of prey behavior, migration, and growth. ACKNOWLEDGMENTS I thank all the members of the University of Guelph Cetacean and Seabird Research Group for assistance in the field and Dr. D. E. Gaskin, who gave his time in support of the late-autumn field sessions. Mr. R. D. McRae provided the bird from Churchill, Mani- toba. The Atlantic Reference Centre, St. Andrew's, New Brunswick, provided helpful assistance in iden- tification of stomach contents of birds. Drs. D. E. Gas- kin, V. G. Thomas, and J. B. Sprague (Department of Zoology, University of Guelph), Dr. R. Frank (The Ontario Mirdstry of Agriculture and Food Pesticide Residues Laboratory, Guelph, Ontario), and Dr. M. Gochfeld (Rutgers University, Piscataway, New Jer- sey) kindly read and criticized an earlier draft of the manuscript. The work was supported by grants to Dr. D. E. Gaskin from the Natural Sciences and Engi- neering Research Council of Canada (operating grant no. A5863) and from the Canadian National Sports- men's Fund and the Department of Fisheries and Oceans. LITERATURE CITED AXTELL, H. H. 1959. The month. Prothonotary 25: 74-75. BALTZ, D. M., & G. V. MOREJOHN. 1977. Food habits and niche overlap of seabirds wintering on Monterey Bay, California. Auk 94: 526-543. BENT, A. C. 1921. Life histories of North American gulls and terns. U.S. Natl. Mus. Bull. 113: 1-345. BORROR, D. J., & g. E. WHITE. 1970. A field guide to the insects of America north of Mexico. Boston, Houghton Mifflin. BOWLES, J. H. 1906. A list of the birds of Tacoma, Washington, and vicinity. Auk 23: 138-148. BRAUNE, B. M., & D. E. GASKIN. 1982a. Feeding ecol- Ogy of nonbreeding populations of larids off Deer Island, New Brunswick. Auk 99: 67-76. --, & . 1982b. Feeding methods and div- ing rates of migrating larids off Deer Island, New Brunswick. Can. J. Zool. 60: 2190-2197. BROWN, R. G. B., S. P. BARKER, & D. E. GASKIN. 1979. Daytime surface swarming by Meganyctiphanes norvegica (M. Sars) (Crustacea, Euphausiacea) off Brier Island, Bay of Fundy. Can. J. Zool. 57: 2285- 2291. BURLEIGH, T.D. 1958. Georgia birds. Norman, Univ. Oklahoma Press. CANADIAN WILDLIFE SERVICE, ATLANTIC REGION. 1979. Summary of unpublished surveys of waterfowl and seabirds in Passamaquoddy region, 1973-79. Pp. 61-67 in Evaluation of recent data relative to potential oil spills in the Passamaquoddy area (D. J. Scarratt, Ed.). Fish. Mar. Serv. Tech. Rept. 901. CUMMINS, K. W., & J. C. WUYCHECK. 1971. Caloric equivalents for investigations in ecological en- ergetics. Intern. Vet. Theor. Angew. Limnol. No. 18. FISH, C. J., & M. W. JOHNSON. 1937. The biology of the zooplankton population in the Bay of Fundy and Gulf of Maine with special reference to pro- duction and distribution. J. BioL Board Can. 3: 189-322. FURNESS, B. L., R. C. LAUGKSCH, & D.C. DUFFY. 1984. Cephalopod beaks and studies of seabird diets. Auk 101: 619-620. GIII,, R. E., & J. D. HAt,I,. 1983. Use of nearshore and estuarine areas of the southeastern Bering Sea by gray whales (Eschrichtius robustus). Arctic 36: 275- 281. GOSNER, K. L. 1971. Guide to identification of ma- rine and estuarine invertebrates--Cape Hatteras to the Bay of Fundy. New York, Wiley-Intersci- ence. GRANT, P. J. 1982. Gulls: a guide to identification. Calton, England, T & A D Poyser. HEII, R. S. 1984. The winter season, December 1, 1983-February 29, 1984: northeastern maritime region. Amer. Birds 38: 292-296. KEIVER, K.M. 1982. Apparent digestible and metab- olizable energy in harp seals (Phoca groenlandica). M.Sc. thesis, Guelph, Ontario, Univ. Guelph. KOHLER, A. C., D. N. FITZGERALD, R. G. HALLIDAY, J. S. SCOTT, r A. V. TYLER. 1970. Length-weight relationships of marine fishes of the Canadian Atlantic region. Fish. Res. Board Tech. Rept. No. 164. KULKA, D. W., & S. COREY. 1982. Length and weight relationships of euphausiids and caloric values of Meganyctiphanes norvegica (M. Sars) in the Bay of Fundy. J. Crustacean Biol. 2: 239-247. LEGARE, J. E. H., & D.C. MACLELLAN. 1960. A qual- itative and quantitative study of the plankton of the Quoddy region in 1957 and 1958 with special reference to the food of the herring. J. Fish. Res. Board Can. 17: 409-448. LEIM, A. H., & W. B. SCOTT. 1966. Fishes of the At- lantic coast of Canada. Fish. Res. Board Can. Bull. No. 155. NUTTALL, T. 1974. A manual of the ornithology of the United States and of Canada, vol. II. New York, Arno Press. PLOUGH, R. H. 1951. Audubon water bird guide. Garden City, New York, Doubleday and Co., Inc. RIDGWAY, G.J. 1975. A conceptual model of stocks of herring (Clupea harengus) in the Gulf of Maine. ICNAF Ann. Meeting Res. Doc. 75/100. ROWLETT, R.A. 1980. Observations of marine birds and mammals in the northern Chesapeake Bight. U.S. Fish Wildl. Serv., Biol. Serv. Prog., FWS/ OBS-80/04. SMITH, G. J. D., & D. E. GASKIN. 1974. The diet of harbour porpoises (Phocoena phocoena (L.)) in coastal waters of eastern Canada, with special reference to the Bay of Fundy. Can. J. Zool. 52: 777-782. --, C. L. JOVELLANOS, & D. E. GASKIN. 1984. Near- surface bio-oceanographic phenomena in the Quoddy region, Bay of Fundy. Can. Tech. Rept. Fish. Aquat. Sci. 1280. SMITH, P. C. 1975. A study of the winter feeding ecology and behaviour of the Bar-tailed Godwit (Limosa lapponica). Ph.D. dissertation, Durham, England, Univ. Durham. Sor,L, R. R., & F. J. ROHLF. 1969. Biometry. New York, W. H. Freeman and Co. SPRUNT, A., JR. 1954. Florida bird life. New York, Coward-McCann, Inc., and Natl. Audubon Soc. --, & E. B. CHAMBERLAIN. 1949. South Carolina bird life. Columbia, Univ. South Carolina Press. TODD, W. E.C. 1940. Birds of western Pennsylvania. Pittsburgh, Pennsylvania, Un, iv. Pittsburgh Press. WIENS, J. A., & J. M. SCOTT. 1975. Model estimation of energy flow in Oregon coastal seabird popu- lations. Condor 77: 439-452.