j. Field Ornithol., 59(2):110-119

THE CHANGING PATTERNS OF BRANT MIGRATION IN EASTERN

A. J. ERSKINE Canadian Wildlife Service P.O. Box 7590 Sackville,New Brunswick,EOA 3C0 Canada Abstract.--Thepattern of BrantMigration through the Maritime Provincesof Canadahas changedmarkedly since 1930, and the formerly authoritative accounts by Bent(1925) and Lewis(1937) nolonger apply. In the 1930s,use of thecoastal route through the Maritimes decreasedgreatly in favorof a directmovement between New and James Bay, leaving lessthan one-tenth of theAtlantic population following the coastal route in spring.Autumnal useof the coastalroute virtually ceased by the 1940s.Spring use of Marltimesstaging areas decreasedbetween 1960-1964 and 1974-1985, but accountedfor about 7% of the total wintering populationsin both periods.

CAMBIOS EN LOS PATRONES MIGRATORIOS DE BERNICLA EN LA PARTE ESTE DE NORTE AMERICA Resumen.--Lospatrones migratorios de Branta bernicla hah cambiado notablemente a parfir del1930, pot lo quelos trabajos de Bent (1925) y Lewis(1937) no se ajustan a losmovimientos actualesde estosgansos. En la d&ada de 1930 el usode la ruta costaheraa travezde las ProvinciasMaritime (Cfinada)disminuy6 considerablemente en favor de un movimiento directoentre New Jerseyy la BahiaJames. Menos de un 10%de la poblaci6ndel Atlfintico tom6la ruta costaneradurante la primavera.Para la d&ada del 1940 el usode esteruta cay6en casi total deshuso durante el otofio. E1 uso en la primaverade la rutade las Maritimes disminuy6durante los aftos1960-1964 y 1974-1985, con tan soloun 7% de las aves utilizandodurante ambos periodos la ruta de Bahiade Fundy/Golfode St. Lawrence. The westernAtlantic population of Brant (Brantaberr•icla) is known to have changedin numbersdramatically during the last 60 yr (e.g., Rogers1979). Observationsin the Maritime Provincesof Canadain 1960-1984 suggestthat Brant migrationpatterns have changedfrom earlierdescriptions (Bent 1925, Lewis 1937). The accountsby Bellrose (1976) and Palmer(1976) seemto underestimatethat change.The errors probablyarise from the enormously increased volume of data accumulated since1930, muchof it still unsummarizedand unpublished.No handbook compilerdealing with dozensof speciescan examine all sourcesof un- publisheddata, and most major journals discourage descriptive papers summarizingsuch information. I haveassembled Brant data from the Maritime Provinces(New Brunswick,,Prince Edward Is- land), and review the changesin Brant numbersand movementsthat becameapparent from a comparisonof Maritimesrecords with the overall picture.

SOURCES The analysisis basedon several,non-overlapping sources. A primary sourceis my repeatedobservations of Brant along the north shore of Nova Scotia(NS) during 1960-1969 and 1974-1985. An unpublishedcom- pilationfor theMaritime Provinces (Aug. 1974) by BarryHughson, then

110 vol.29, No. 2 BrantMigration [ 1 1 1 of the Canadian Wildlife Service (CWS) drew on unpublishedsources as well as published data. Very few of my recordswere included in Hughson'scompilation. The publishedrecords of the Nova Scotia Bird Society(NSBS Newsletter 1957-1980, Nova ScotiaBirds 1981-present) were abstracted,extending Hughson's coverage by 15 yr. The speciesfiles in the New Brunswick(NB) Museum (not consultedby Hughson)were abstracted,as the published summaries based on them (Squires 1952, 1976) were extremelycondensed. Personal records of David Christiewere also consulted. An aerial survey,by United StatesFish and Wildlife Serviceand CWS personnel,in May 1977 (A. Reed, unpubl. rpt. in CWS files) covered most Brant stagingareas in Canada and the U.S.A. within a few days.

BRANT MIGRATION THROUGH THE MARITIMES Table 1 summarizesthe accumulatedrecords of Brant migration from 1960 to 1984 for comparisonwith earlier periods.One striking feature of the movementthrough our area is the narrow width of the migration corridor (Fig. 1). A few Brant reach the Maritimes in February, or late January in some years, but reportednumbers seldom exceeded 200 in thosemonths. The main movementevidently begins in March, when severalthousands ap- peared at Grant Manan and at Brier Island, at the mouth of the . The last stagingarea to the southwestis apparentlyat Cape Cod (cf., Bent 1925, Palmer 1949, Phillips 1932), and the 400 km acrossthe Gulf of would require less than 7 h flight; leaving Cape Cod in the morning would reach the Maritimes well before dark. The numbers reported from Grand Manan and Brier Island are far larger than anywhere else around the , which suggestsdirected movementin daylight.The third major concentrationarea in the Bay of Fundy, at Maces Bay, New Brunswick,is about 1 h flight farther from Cape Cod, but the numbersthere peak later, usuallyin April. Numbers farther south(in Nova Scotia)and northeast(up the Bay of Fundy) are much smaller, perhaps reflectingonly dispersalto ice-free areas after arrival. Beachesand intertidal areasin the lower Bay of Fundy are partly ice-coveredin winter and early spring, and the March arrival of Brant at Grand Manan and Brier Island probably evolvedto match the time when suitable feeding areas there becomeice-free. Suitableareas in the upper Bay of Fundy probablycould accommodate only a small part of the population,and the next important stagingareas are in Prince Edward Island, northern Nova Scotia and eastern New Brunswick,which usuallydo not becomeice-free until mid-April, a month or more later than at Grand Manan. My earliest record for the north shore of Nova Scotia was 4 Apr. 1968, at Linden. In the 1960s more Brant were seenat Linden than at more easterlylocations during April, but in May larger numbers were found at Port Philip and Bayhead (Table 2). After 1970 I rarely found any Brant in the more easternareas, and few in May. The milder, moreopen springs recently may allow Brant 112] .4.J. Erskine J.Field Ornithol. Spring 1988

(TO UNGAVABAY)

QUEBEC

BRETON ISLAND

NEW BRUNSHIC

L__•GE •, LONG ISLAND, NEN YORK 2, CAPE COD, MASSACHUSETTS •. GRANDMANAN ISLAND, NEW BRUNSWICK q. BRIER ISLAND, NOVA SCOTIA 5. MACESBAY, NEW BRUNSWICK 6. LiNDEN-RIVER JOHN, NOVA SCOTIA 7. CENTRAL PRINCE EDWARD ISLAND 8. TABUSINTAC, NEW BRUNSWICK 9. SEPT-•LES, QUEBEC •0. MONTMAGNY-MATANE,QUEBEC •1. P•CTON TO KINGSTON, ONTARIO FIGUREl. Springmigration routes of Brantin easternCanada, with known(past and present)major stagingareas. to movefarther north earlier than in the past.In the 1960s,from mid- Aprilto lateMay Brantwere found regularly in northernNova Scotia and on PrinceEdward Island.Counts on PrinceEdward Islandexceeded 2000in theearly 1960s, whereas the north shore of Nova Scotia harbored fewer than 1000 Brant. Traditional (illegal) huntingfor sea ducksin easternNew Brunswickduring April andMay disturbedother waterfowl sothat few peoplehave made Brant observations there. Jack Wishart, long-timeresident of Tabusintac,reported estimates of 6000Brant as lateas 1956, but Hughson and I wereunsure whether those were peak counts,or cumulative totals, including repeated sightings, ofstaging flocks. CWS aerialsurveys in 1974and 1977found only 1000 and 1400Brant on thoseshores. The finaldeparture from the southern Gulf of St. Law- rencevaried from about 25 May to 5 Jun. in the 1960s;only in 1961,a verylate spring, did I seeover 100 Branton a Junesurvey in Nova Scotia.I sawno Branteast of RiverJohn (long.63ø03'W), although E. Holdway(reports to NovaScotia Bird Society[NSBS]) saw a few at Pictou(62ø40'W)--which I did notvisit--in mostsprings in the 1960s; vol.59, No. 2 BrantMigration [ 1 13 1 14] A.J. Erskine J. Field Ornithol. Spring 1988 TABLE2. Summaryand comparisonof Brant migrationrecords by A. J. Erskine,northern Nova Scotia, 1960-1984.

Mean number (frequencyof records)a 1960-1969 1974-1984 Area (westlongitude) Apr. May Apr. May Linden (63ø49') 167 (5/6) b 81 (5/8) 38 (5/6) 27 (1/4) Port Philip (63043') 11 (2/7) 145 (9/10) 0 (0/5) 38 (1/3) Bayhead-Tatamagouche(63022 ') 101 (1/2) 137 (5/8) 0 (0/3) 0 (0/2) River John (63ø03') 0 (0/2) 94 (5/6) no data no data

Omittingzero counts before earliest date or after latestdate of recordin any year. "5/6" meansBrant seenon five of six visitsduring migration period. one spring record at Merigomish Island (62ø25'W) in 1977 (NSBS) was exceptional.Hughson noted small numbersseen in Prince Edward Island eastto Souris (62ø15'W) fairly regularly. Recent data give few clues as to the destinationof Brant that pass throughthe Maritimes.The formerconcentration at Sept-lies,Quebec, (Lewis 1937) seemsnot to be important now (Bellrose 1976), but Brant using the coastalroute may stage farther west. Lehoux et al. (1985) showedBrant using various locationsin the St. Lawrence estuary, with peak numbers in the first half of May. These dates seemearly for the coastalroute, aspeak Brant numbersin the southernGulf of St. Lawrence in the 1960spersisted until late May. The 1977 aerial survey(A. Reed, unpubl.) found 12,000 Brant alongthe St. Lawrenceestuary on 16 May. The 1700 farther eastin Quebec,with 5100 in the Maritimes and 1300 in New England eastof Long Island, New York, in the sameweek, added up to over 8000 Brant using the coastalroute, which agreedwith other estimatesin the 1970s (seeTable 3). If the 12,000 on the upper estuary of the St. Lawrence arrived by the coastalroute, passagethrough the Maritimes in April and early May would have to be more rapid than is suggestedby countsin the larger stagingareas of the Maritimes. One alternative is a direct overland flight bypassingthe Maritimes, probably from the Long Island area to the St. Lawrence estuary,as was assumedby Vangilderet al. (1986). No reportsof largenumbers of Brant migrating overlandalong sucha route havebeen received (A. Reed, pets. comm.);in view of the high lands along that route, the birds might fly too high to be detectedfrom the ground. Such a flight is energetically possiblefor Brant, as describedby Vangilder et al. (1986), but it hasnot beenverified. In the estuaryand northernGulf of St. Lawrence,eelgrass (Zosteramarina), a preferredfood of Brant, neverrecovered after the die- off around 1930 (A. Reed, CWS, in litt.), althoughrecovery in the Mar- itimes was complete.Without the abundant food in the northern Gulf that formerly allowedreplenishment of fat reservesby large numbersof Brant, the main benefit to the birds following the longer coastalroute is that the populationcould spread out during March and April, as coastal Vol.59, No. 2 BrantMigration [ 1 15

T^BLE 3. (a) Estimationof spring Brant numbersin the Maritimes and (b) Comparison with winter populationson the Atlantic coast, 1960-1964 and 1974-1984 (figures rounded to nearest50, except 1977 survey).

Maximum spring population estimatesa Mar. Apr. May 1960- 1974- 1960- 1974- 1960- 1974- (a) Area 1964 1984 1964 1984 1964 1984 1977 i) In Bay of Fundy Brier Island 3000 1000 1000 1000 500 450 Grand Manan 6500 3700 7000 5300 8000 550 Maces Bay 500 400 2000 1650 150 300 10,000 5100 10,000 7950 8650 1300 1513 ii) In Gulf of St. Lawrence PEI 2300 1250 2200 850 1093 North NS East NB 1000700 700350 1300900 1250 500 } 2534 4000 2300 4400 2600 3627 Overall (rounded to 1000s) 10,000 5000 14,000 10,000 13,000 4000 5140

(b) 1960-1964 1974-1984 Winter estimates(from USF & WS unpubl.) 124,600-265,600 35,000-127,000 mean 197,000 mean 89,000 Spring estimatesin Maritimesb (from above) 12,000 6000 (6%) (7%)

a Highest single day figures coveringmost concentrationsin area shown. bMean used(rather than maximum) to reducepossibility of duplicationbetween Bay of Fundy and Gulf of St. Lawrence areaswithin a month. feedingareas becameice-free, long before stagingareas in James Bay and northward were accessible.Dispersal along the coastwould reduce pressureon foodresources during the periodwhen the birds are building up fat reservesprior to the main movementnorth to the breedingareas. There are no recent observationsof migration northward beyondthe gulf and estuaryof the St. Lawrence.The former Sept-Islesstaging area wouldbe preferable to the St. Lawrenceestuary (Quebec-Matane) mainly for departuresnorthward to Ungava Bay--880 km from Sept-Iles vs. 1050-1200 km from the estuary.The distanceto JamesBay is comparable from both stagingareas. The scarcityof Brant in the northern Gulf now suggeststhat migration via Ungava may no longer occur (A. Reed, in litt.). Lewis (1937) indicatedthat Brant did not stopin Ungava Bay, but merelypassed there enroutenorthward. Vangilder et al. (1986) suggested that sucha migration would be possibleonly with goodfeeding oppor- tunities for replenishingfat reservesbefore the flight. I concludethat Brant migratingthrough the Maritimes now continuenorthward largely or exclusivelyby way of James Bay. 116] A.J. Erskine J.ridd Ormthol. Spring 1988 MacLaren Marex Inc. (1979) reportedsome Brant aroundthe northern tip of Labrador, and near FrobisherBay (Ikaluit), NorthwestTerritories, in May-Jul. 1978, with most birds in 9-10 Jun. Those birds were 300- 500 km east of any migration route along the east side of Hudson Bay; possiblythey representedremnants (< 1000 birds in all) of the former migration by way of Ungava Bay, but drift during migration seemsequally plausible as an explanationof their presence. Fall migration of Brant through the Maritimes now is almost non- existent. I saw them only once in fall. Similarly, the NSBS recordsin- cluded fall dates in only a few years, notably in 1977 from 15 Nov. onward, with a remarkable estimateof 2000 birds in the Upper Bay of Fundy on 9 Dec. The only recent Brant wings from the Maritimes in the CWS speciescomposition survey also were receivedin 1977. A few Brant stayedaround Brier Island all that winter, the only recent year with recordsoutside of Grand Manan, although Christmas Bird Count reports occur every few years. The earliest fall recordswere in late October,but one Cape BretonIsland record(15 Sep. 1968, 15 birds) and one at Pictou (10 Sep. 1963, nine birds) seemedso aberrantlyearly that birds of a different (High Arctic?)population might have beeninvolved.

BRANT MIGRATION FARTHER WEST The major spring movementof Brant in easternNorth America now leavesthe Atlantic coastnear Long Island and runs overland to James and Hudson bays (Bellrose 1976), with a relatively small proportion stoppingenroute on Lake Ontario (cf., Quilliam 1965). I saw what were apparently Brant flying northwestin westernQuebec almost on a direct line from Long Island to James Bay, on the eveningof 22 May 1970. If Brant leave the Long Island coastin the morning, a continuousflight at 60 km/h would bring them to Lake Ontario by afternoon,into western Quebecby evening,and to JamesBay by dawnthe next day; at 100km/h (Palmer 1976), Long Island to James Bay would be only a day's flight (14 h). In fall, the main movementof Brant passessouthward through Hudson and James bays and thenceoverland to the Long Island area (Bellrose 1976). Stopoverson Lake Ontario are lessregular than in spring (Quil- ljam 1965; Spragueand Weir 1984); possiblyfall flights passover that area during darkness,although that would call for evening departures from James Bay, in contrastto the daytimemigration suggested in spring.

CHANGES IN MIGRATION PATTERNS Bent (1925) reportedthat the main Brant movementin springfollowed the coastnortheastward to Cape Cod, crossedthe Gulf of Maine to the Bay of Fundy, continuedacross the isthmusto the Gulf of St. Lawrence, and stagedat Sept-Iles before departing northward or northwestward overland.Overland migration to the St. Lawrence estuarywas considered a minor variant. Lewis (quotedby Bent 1925, Lewis 1937) believedthat Vol.39, No. • BrantM•rat•on [ 1 1• almostthe entire population stoppedat Sept-Iles;however, his estimate of total numbers(about 60,000) seemsimprobably low, as Phillips (1932), for the sameperiod, cited single-daylocal countsas high as 80,000 birds in New Jersey. Phillips' estimateof one-third of a million birds as the maximum populationin the 1920s is consideredrealistic (Rogers 1979). Rogersconcluded that, after the widespreadeelgrass die-off, total Brant numberswere unlikely to have decreasedby as much as 90%, as asserted by Cottam (1934); poor reproductioncoupled with continuedhunting were a more plausibleexplanation of the declinethan massstarvation. Neither Cottam or Rogersseem to have consideredthat a major shift in Brant migration routes occurredcoincident with the disappearanceof eelgrassfrom the coast. Lewis (1937), however, closedhis accountby asking what route Brant could be using, if most of the populationno longer visited Sept-Iles. Brant began to appear on Lake Ontario during migration sometime duringthe 1930s(Sheppard 1949). Quilliam (1965) notedBrant regularly at Kingston, Ontario, with single-daycounts up to 2000 birds, where earlier observersin 1890 to 1930 had reportedno Brant. Thus, the present migration route, passingdirectly northward from Long Island to James Bay in spring, and in the reversedirection in fall, becameestablished only in the 1930s. Palmer (1976) suggestedthat abouthalf the populationused the coastal route in spring, but Bellrose (1976) inferred that the majority now use the inland route.Hughson in 1974 (unpublished)concluded that not more than one-tenth of the population use the coastalroute in spring. I heard similar views from CWS biologistsin the Maritimes as early as 1960, but nothingwas published,and Lewis' (1937) accounthas remainedthe acceptedpattern. My estimatesof the numbers passingin spring in comparisonwith the total winter figures (Table 3) confirm Hughson's assessment,with about 7% of the populationusing the Bay of Fundy/ Gulf of St. Lawrence route. The high numbers on the St. Lawrence estuary in May 1977 remain an anomaly. The fall migrationpattern describedby Bent and Lewis was the reverse of that in spring.Palmer (1976) indicatedthat fewer birdsused the coastal route in fall than formerly, and Bellrose(1976) statedthat the bulk of the populationmigrated overland. The late G. F. Boyer (CWS, unpubl.) noted that the fall flight of Brant through the Maritimes had virtually vanishedby 1945, and recent evidencesupports that view. The second- or third-hand reportsquoted by Bellrose(1976) may have reflectedonly memories of the former situation.

DISCUSSION As remarked by Hughson, "a crucial unknown, especiallyof early estimates,is the credibility which can be assigned(to) particular large estimates.... "Hughson believedthat local ground counts"provide ... little insightinto the magnitudeof the flight," in contrastto wide-ranging aerial counts.However, I suggestthat, giventhe few areasat which Brant 118] A.J. Erskine J.Field Ornithol. Spring 1988 stop off in numberswith any regularity, ground countsof known con- centrationareas provide important insightson populationvariations. The main caveatis that singlecounts or estimatesmust always be testedagainst the overall pattern. Some reported figures are obviouslyimplausible. Phillips (1932) discountedan alleged4 million Brant passingCape Cod in the spring of 1887. A record is in CWS files for Sept-lies of 492,700 Brant in the spring of 1929, but no other countbetween 1922 and 1934 exceeded92,000; probablythe 1929 figure involveda clericalerror. Most peak countsfrom the Bay of Fundy in 1935-1959 were round figuresof 5000 to 20,000, whereas recent figures from the same areas are a few hundredsto a few thousands;the numbersusing the coastalroute clearly havedeclined (cf., Table 2). There may be no valid groundsfor dismissing all the earlier large countsas simply guesswork. Winter estimatesof the west Atlantic populationof Brant varied from 35,000 to 266,000 (Rogers1979) between1935 and 1979. As the total population declined in the early 1930s, the North Carolina wintering stockwas depletedwhile the New Jerseyone held its own, thus becoming relatively more important. Following a die-off of Brant during severeice conditionsin New Jersey in 1976-1977, the proportionof the population wintering further south becamemore important again (e.g., Kirby and Obrecht 1982). We lack data on trendsin the breedingareas of this Brant population. Brant are Arctic nesters, and subject to "boom or bust" variations in reproductivesuccess, (cf., Barry 1962, Phillips 1932). Individual stocks in the north probablyhave fluctuatedgreatly in size, and somemay have disappeared(Boyd 1979; cf., Palmer 1976). The changein numbersof Brant following the coastalmigration route may reflect changesin sizes of particular breeding stocksas well as shifts in migration route. The map by Bellrose(1976, p. 171) showingthe coastalroute connectingvia Ungava Bay to and Ellesmere Island breeding areas seems lesslikely than a connectionwith the north Hudson Bay stocks,and later versionsof the same book do not show the Ungava Bay route at all. Phillips (1932) implied that a few hundredthousand Brant formerly used the coastalroute. Such large numberscould not have bred in the High Arctic at any recentperiod; even if someof them formerlytravelled north via Ungava Bay, they presumably dispersedwestward from southern . Migration via James and Hudson bays has been linked only with Low Arctic breedingareas around Foxe Basin and westward. Overlandmigrations must be traditional;marine birds are likely to initiate suchflights only in the companyof othersthat have usedthem in earlier years.At times of major populationreduction, it would be possiblefor a particular traditional route to be abandoned(cf., Hochbaum 1955), as the Ungava Bay route seemsto have been lost. Recent concernover a return of the eelgrassblight in New England (Kelley 1986) makes it desirableto documentwhat happenedto Brant migrationsin the northeast over the past 60 yr. vol.•9, No. 2 BrantMigration [ 1 19

LITERATURE CITED

BARRY,T. W. 1962. Effect of late seasonson Atlantic brant reproduction.J. Wildl. Manage. 26:19-26. BELLROSE,F.C. 1976. Ducks, geese& swansof North America. 2nd ed. StackpoleBooks, Harrisburg, Pennsylvania. BENT, A. C., ed. 1925. Life histories of North American wild fowl. Order Anseres (Part II). U.S. Nat. Mus. Bull. 130, Washington, D.C. BOYD,H. 1979. Management of Branta bernida in Canada. Pp. 198-207, in M. Smart, ed.Proceedings/Compte Rendu First technicalmeeting on WesternPalearctic migratory bird management.Int. Waterfowl Res. Bur., Slimbridge,England. COTTAM,C. 1934. The eel-grassshortage in relation to waterfowl. Trans. Amer. Game Conf. 20:272-279. HOCHBAUM, H. A. 1955. Travels and traditions of waterfowl. Univ. Minnesota Press, Minneapolis. KELLEY,K. 1986. Wildlife threatened.Scientists fear return of 1930seelgrass blight. Rural Delivery 11(3):11-12. KIRBY, R. E., AND H. H. OBRECHTIII. 1982. Recent changesin the North American distributionand abundanceof wintering Atlantic Brant. J. Field Ornithol. 53:333-341. LEHOUX,D., A. BOURGET,P. DuPuIS, ANDJ. ROSA. 1985. La sauvaginedans le syst•me du Saint-Laurent. EnvironnementCanada, Servicecanadien de la faune, Rggion du Quebec. LEWIS,H. F. 1937. Migrations of the American Brant. Auk 54:73-95. MACLARENMAREX INC. 1979. Report on Aerial Surveysof Birds and Marine Mammals in the Southern Davis Strait between April & December, 1978 for Esso Resources Canada Ltd., Aquitaine Co. of Canada Ltd. and Canada Cities ServicesLtd. Arctic PetroleumOperators Association, Project No. 146, volume I Birds, Marine Scientists & EngineersSt. John's-Dartmouth-Fredericton. PALMER,R. $. 1949. Maine birds. Bull. Mus. Comp. Zool. 102:1-636. , ed. 1976. Handbook of North American birds, vol. 2, Waterfowl (Part 1). Yale University Press,New Haven and London. PHILLIPS,J. C. 1932. Fluctuation in numbers of the eastern Brant .Auk 49:445- 453. QUILI•IAM,H. R. 1965. The historyof the birds of Kingston,Ontario. Privately printed. ROGERS,J.P. 1979. Branta bernidahrota in the USA--a managementreview. Pp. 198- 207, in M. Smart, ed. Proceedings/CompteRendu First technicalmeeting on Western Palearcticmigratory bird management.Int. Waterfowl Res. Bur., Slimbridge,England. SHE??ARI•,R. W. 1949. The American Brant on the lower Great Lakes. Can. Field-Nat. 63:99-100. S?RACUE,R. T., ANn R. D. WEIR. 1984. The birds of Prince Edward County. 2nd ed. Kingston Field Naturalists, Kingston, Ontario. SQUIRES,W. A. 1952. The birds of New Brunswick. Monograph No. 4. The New Brunswick Museum, Saint John, New Brunswick. ß 1976. The birds oœNew Brunswick. 2nd ed. Monograph No. 7. The New Brunswick Museum, Saint John, New Brunswick. VANGILDER, L. D., L. M. SMITH, AND R. K. LAWRENCEß 1986ß Nutrient reserves of premigratory Brant during spring. Auk 103:237-241. Received 17 Dec. 1986; accepted19 Sep. 1987.