Notes on the South Atlantic Canada Goose Population

BIRD-BANDING A JOURNALOF ORNITHOLOGICAL]NYESTIGATION

VOL. XXIII JANUARY,1952 No. 1

NOTES ON THE SOUTH ATLANTI.C CANADA GOOSE POPULATION BY HAROLd) C. HANSON AND RICHARD E. GRIFFITH Recent studies (Williams, 1945; Hanson and Smith, 1950) have shownthat each of the four main flywaysin North America contains two or more distinctpopulations of Canadageese, Branta canadensis (Linnaeus). In a comprehensivereport on the Canada geeseof the Mississippiflyway population, Hanson and Smith (1950) describedthe rangeof a distinct,heretofore unrecognized population, the Southeast population,that wintersin the inland regi.onsof the southe.astern states. It also outlined in brief the rangesof two other populationswhich winter on the Atlantic seaboard, the North Atlantic and the South Atlantic populations. BecauseCanada geese tend to be segregatedby distinct population groupingswith strict adherenceto individual breedingand wintering ranges,these birds are particularlyadapted to effectivemanagement by populationgroups. For this reason,there is specialmerit in careful identificationand study of thesepopulations. Thi.sreport dealsprimarily with the geeseof the South Atlantic population,but data and mapsshowing recent unpublished recoveries from the Mississippiand Southeastpopulations are also includedfor comparative purposes. MATERIALS AND METHODS The nucleusof this report is providedby the .bandingand recovery records of the Jack Miner Migratory Bird Sanctuary located near Kingsville,Ontario. The recoveryrecords consist chiefly of the original letters received from hunters in the United States and Canada and from missionaries and fur traders in the far North. The records of Canada geesebanded at the Jack Miner Sanctuary in autumn, the majority of which were of geesebelonging to the MississippiFlyway and the Southeastpopulation, were analyzedfor the MississippiFly- way report. The recoveriesof Canada geesebanded at the Miner Sanctuaryin spring, chiefly of the South Atlantic population,are treatedhere. Repliesto questionnairessent by Hansont,o the principal fur trade posts around Hudson and James Bays in 1947, provided additional valuable information about conditions on the breeding grounds. Information on wintering groundsand populationsis •based mainly on personalknowledge and data obtainedfrom the files of the Branch of Refuges,U.S. Fish and Wildlife Service. ACKNOWLEDGMENTS For permissionto study the Miner records,we are indebtedto Manley Miner, Presidentof the JackMiner Migratory Bird Foundation, Inc., acting in behalf of the Miner family and the Miner Foundation. 2] HANSON,GalrrlT•, Canada Goose Population Bird-Banding January Vol. XXIII 1952 HANSON,GRIFFITH, Canada Goose Population [3

H U D S 0 N

B A Y

BELCHER

Whole R.

L E G E N D JAMES g TWIt I-Eskimo Point 12-Eastmain o] is. 2- Churchill 13-Old Factory 3-York Factory 14-FI. George 4-Ft Severn 15-Great Whale River 5-Weenusk IG-Port Harrison G-Lake River 17-Povungnituk 7-Attawopiskot 18-Cope Smith 8-Akimiski Island 19-Walstenholme 9-Ft. Albany 20-CopeHenrietta Mori, 10-Moose Factory 21-Cape Jones II-Rupert House 22-Ogoki 23-R•chmond Gulf

Fig. l.--Map showinglocation of principal fur trade posts in the H.dson-James Bays area.

We are appreciativeof the official co-operationextended by Dr. Harrison F. Lewis and T. S. Hennesseyof the Canadian Wildlife Serviceof the Canada Department,of Resourcesand Development; and by Dr. ClarenceCottam and Mr. Frederick C. Lincoln, of the United States Fish and Wildlife Service. For supplementarydata on populationsand kills we are indebted as follows: for New York, A. W. Holweg; for Pennsylvania,Thomas D. Frye; for New Jersey,L. G. MacNamara; and for North .Carolina, T. Stuart Critcher. HANSON,GRIFFITH, Canada Goose Population Bird-BandingJanuary

BREEDING RANGE Our knowledgeof the breedingrange ,of the SouthAtlantic population is basedon a comparativestudy of the recoveriesfrom bandingsat HorseshoeLake, Illinois; the Miner Sanctuary,Kingsville, Ontario (Hanson and Smith, 1950); Earleville, Maryland; and Lake Matta- muskeet,North Carolina, Ta.ble 1. These rec,overyrecords indicate that the range of the South Atlantic populationincludes suitable areas inland from the east coastof Jamesand HudsonBays, from perhaps as far southas the Rupert River north to W,olstenholme. Fig. 1. The BelcherIslands in HudsonBay and the Twin Islandsin central James Bay are believedto lie within the breedingrange of this flock as possiblyalso do portionsof southernBarfin Island. The main rangeeast of JamesBay may be confinedto the relatively low coastalplain that extendsinland for a distanceof about 60 miles. Low (1896 p. 324) writesin regardto the Canadagoose: "... abundant on the East Main River . . . especiallyon lower part, where the river is cut out of clays, with good bottomlands;breeds in large numbers on the islandsof JamesBay." Few band recoverieshave been obtainedfrom the countrylying betweenCape Jonesand the RichmondGulf area, fig. 1, the reason being that this region apparently does not afford either attractive feedingor nestinggrounds. Consequently, few geeseare killed in this sector. Whereasthe coastof JamesBay from Rupert Bay to Cape Jonesis low and undulatingand offers a number of feeding areas attractive to migrating geese, the coast of Hudson Bay from the neighborhoodof CapeJones to CapeDufferin is of a differentcharacter, the land being higher and more unevenand rising gradually as it extendsnorthward, becomingrugged and precipitousat the head of ManitounukSound (Bell, 1879). This coastline,as w,ould be expected, affordsno feedingareas for geese.In a letter to Jack Miner (Dec. 20, 1918) L. G. Mayer, then a r•ostmanager at Great Whale, related that few geesewere killed at Great Whale becausethere were no feeding placesalong that sectorof the coast and that the geesemade their autumn flights in his area over water. From the descriptionsgiven by geologists(Bell, 1879; Low, 1888, 1902), largesectors of the country lying inland from the lower half of the east coast of Hudson Bay are apparentlytoo high, rugged, and barren to afford ideal nesting habitat for Canada geese. Reportsof residentsat Povungnitukand Port Harrison.the accounts of explorers (Low, 1902; Rousseau,1948), and band recoveries,all point t,o the presenceof large numbersof Canada geesebreeding throughoutthe UngavaPeninsula. This beliefwas recently substantiated by Eklund and Cool (1949). In an aerial surveyof waterfowlpopula- tions in the central sectionsof the Ungava Peninsula,they found the highest populationsof Canada geese between the Koguluk and PovungnitukRivers. There, they obtaineda count of 2.1 geeseper mile as comparedwith over-all averagesfor tundra of 0.21 and 0.17 geeseper mile of rivers and lakes, respectively.Their finding of a high populationof Canadageese between these two adjacentrivers in Ungava is at leastreminiscent of the relation of CanadaGoose produe- Vol. XXIiI 1952 HAh•ON,GRIFFITH, Canada Goose Population [5 tion centersto rivers in Northern Ontario (Hanson and Smith 1950). Adequatedescriptions of the exactnature of the interior of the Un- gava Peninsulaare scarcebecause few peoplehave traversedthe interior. As Flaherty t1918, p. 124) stated,knowledge of this country "is derivedsolely from the informationof the Eskimo. Accordingto them it is, generallyspeaking, a rolling plateau of low, long-sloping hills, everywhereintersected by countlessthousands of lakes and con- nectingstreams." On the BelcherIslands, large numbers of Canadageese nest, particularly along a V-shapedlake, 65 miles long, on Flaherty Island (letter from Jack Tryer to Jack Miner, date unknown). A portion of the breedingrange of Canadageese on southernBarfin Island may be considereda part of the range of the South Atlantic population,particularly the westerlyportions of the southerncoast, althoughthe speciesis known to nest "more or lessuniformly all the way along the southerncoast from Gabriel Strait to Cape Dorset. re- sortingto islandsas well as mainland" t Soper1946, p. 17). The Net- tilling Lake Eskimoshave reportedthat many Canadaand other kinds of geeseare to be found in the regionof AmadjuakLake and along the river of the samename (letter from Arnold C. Herbert to JackMiner, April 3, 1939,}. A goosebanded by the Miners in the springof 1927 and shotat AmadjuakLake lendscredence to the reportby thesenatives, but it shouldbe pointedout that it may be principallythe non-breeding geesewhich frequent the areaslying north of the main breedingrange on the mainland. Canada geeseare also reportedto nest 15 miles inland from Lake Harbour along the Soper River (ConstableDaoust, R.C.M.P., Hansonquestionnaire, 1947, Soper 1946). The Canada geeseof the South Atlantic populationwhich nest on the BelcherIslands and inlandfrom the eastcoast of HudsonBay as far as the heightof land are a part of a singlerecognizable population, the race B.c. interior, Todd (1938).

MIGRATION Prior to the timesouthward migration occurs, there are localflights to inlandlakes for the purposeof .molting.A. Lunan,former post managerat Port Harrison, recentlystated t personalcommunication. 1949• that there were flights of Canadageese into the tundra lakes of thePort Harrisonregion from the southin earlyJune, reportedly for thepurpose of molting.Flights of Canadageese into this regionfrom areasnorth of Port Harrisonhave also been observed. According to the Rev. H. S. Shepherd,Canada geese have appearedat favoredlakes in the Port Harrisondistrict about the end of July, with the movement extendinginto August. On theirautumn migration, the geese of the SouthAtlantic population follow the eastcoast of Hudsonand JamesBays south as far as RupertBay. There,a portion,if notthe main body, of theflight strikes inland,follo;• ing the generalcourse of theNottaxvay River to the southeast. Otherscontinue on to Hannah Bay at the southend of James Bay,from which they fly southto theMiner Sanctuary, their migration routesmerging with those of the Canadageese of the Mississippiand Bird-Ba•din,g 6] HANSON,GRIFFITH, Canada Goose Population January

Southeastpopulations. For reasonsdi.scussed more fully by Hansonand Smith (1950), the numbersof SouthAtlantic geese migrating through the regionof Kingsville,Ontario, in the autumnare believedto be but a small part of the total flight. Ban.d recoveries,fig. 2, exaggeratethe importanceof this latter migration route for severalreasons: (1) a greater hunting pressure is associatedwith the •nore dense human populationin the regionsdirectly south of JamesBay as compared with the countrylying southeastfrom JamesBay; {2) mostof the area

Fig. 2.--Location of band recoveriesfrom Canada geesebanded at the Jack Miner Bird Sanctuary,Kingsville, Ontario, in the spring, 1915-49,and reported recoveredsouth of James Bay in Canada. Recoveriesreported from fur trade postson the east coastsof Hudson and James Bays are given in table 1. Vol.1952XXIII HANSON,GRIFFITH, Canada Goose Population [7 to the southeast,between James Bay and the northernborder of western New York is a forested wilderness whereas between the south end of JainesBay and Kingsvillethere are considerableareas of farmland attractiveto Canadageese for feeding; (3) overlapsoccur in banding operationsat the Miner Sanctuaryowing to the variablemigration behaviorof the geesethemselves, some geese from boththe Mississippi Valleyand Southeastpopulations stopping at the Miner Sanctuaryin the springand receiving"S" markedbands from the seriesused in springmainly on the SouthAtlantic geese. When these spring-banded geeseare shotin autumnalong their normalmigration route between

YEAR OF RECOVERY ß 1940

•ß 19421941 HOLLOWSYMBOLS SHOW .... ß 1945 LOCALITY IS DOUBTFUL • ...... ß 1944

Fig. &--Location of band recoveriesfrom Canada geesebanded at the Jack Miner Bird Sanctuary, Kingsville, Ontario, in the spring of 1944 or earlier springs and reported recoveredin the United States.during 1940-44. Bird-Banding 8] HANSON,GRIffITH, Canada Goose Population January 'Vol.1952XXIII HANSON,GRIFFITH, Canada Goose Population [9

JainesBay and the Miner Sanctuary, the recovery records are presumed to be of SouthAtlantic geese,although for reasonsstated this is not true in many cases. The flightsof Canadageese which fly southeastwardfrom Rupert Bayon JamesBay are believedto be thesame as thosethat crossinto the United Statesin the region of westernNew York, fig. 3. These flightscontinue on acrosseastern Pennsylvania, attaining the middle Atlantic coastin the region of ChesapeakeBay. I,ittle is yet known regardingthe patternof the "distributaryflights" whereby the various componentsof the populationsegregate out on the differentwintering areas. Perhapsthe geesewintering at Lake Mattamuskeettend to use somewhatmore westerlyroutes than the geesewintering further north alongthe co,ast.In all likelihood,the geeseusing any one sectorof the winteringgrounds tend to return to the samesector in subsequent years. The spring migration routesextend farther west than do the autumn migrationroutes, perhaps explained in part by the northwesterlypro- gressionsof the spring isotherms.En route, large numbersof these geesevisit the Miner Sanctuary,particularly during the first two weeks in April. Their final flightcarries them directly to the breedinggrounds, the coastsof Jamesand HudsonBays not being followedto the extent that they were in the autumn. The more moderatecli•nate of the interior may be one reasonfor their avoidanceof the coastsof the bays in earlyspring. According to GeraldParsons, post manager at ObJjuan, locatedjust southof Gouin Reservoirin Quebec,few flocksare seenin his area in the autumn,but many are seenin the springflying directly north. At Great Whale River, fig. 1, few geeseare seenin spring,but many are observedin the autumn (letter from L. G. Mayer to Jack Miner, 1918). Thus, it is in spring,while they are still on their inland trappingterritories, that the Indians who summerat the postson the east coast of JamesBay make their principal kill of Canada geese, table 2.

WINTERING GaOUNDS When the recoveriesfrom the Miner bandingswere plottedby exact localityrather than simplyby states,as wasthe practiceof Jack Miner, they revealedthat the autumnflight stoppingat the Sanctuaryactually consistedof two populations,the MississippiValley populationand the Southeastpopulation, fig. 4 and table 3. Band recoveriesindicated that the Southeastpopulation wintered chiefly in the inland pied•nontareas of Virginia, North Carolina, South Carolina, Georgia, Alabama, and on the gulf coastof Florida in the vicinity of St. Marks. Within this range,these geese used nearly every major river and reservoirof im- portance.For a more detaileddiscussion of the breedingand wintering range of the geeseof the Southeastpopulation the reader is referred to Hansonand Smith (1950). The Southeastpopulation will be dis- cussedhere only insofaras new data are availableor whenit is desir- ableto reviewpublished data to clarify the statusof the SouthAtlantic population.In contrastto the inland range of the Southeastgeese, the SouthAtlantic geese frequent almost exclusively the coastalareas, from southernNew Jerseyto Lake Mattamuskeet. ]10] HANSON,GRIFFITH, Canada Goose Population Bird-BandingJanuary Vol. XXIII 1952 ]:lANsON,GRIFFITH, Canada Goose Population [] 1

The over-all wintering range of Canada geeseof the South Atlantic populationis less extensivethan that of either the MississippiValley or Southeastpopulations, figs. 3 and 4, but the total refugeareas available to these geeseare considerable,national refuges alone totalling about 69,000 acres. In addition, there are enormous water areas that oiler thesegeese safe retreat from gunning. The main winteringrange extendsfrom the upper reachesof ChesapeakeBay and coastalareas of southernNew Jerseysouth to CurrituckSound and Lake Mattamuskeet.

19461945 HOLLOW SYMBOLS SHOW • 1947 LOCALITYIS DOUBTFUL ß 1948 ß 1949 . ß YEAROFRECOVERY •'•,,•.,,.

Fig. 4.--Location of band recoveriesfrom Canada geesebanded at the Jack Miner Bird Sanctuary,Kingsville, Ontario, during the autumn of 1948 or earlier autumns and reported recovered in the United States during 1945-49. Bird-Banding 12] HANSON,GRIFFITH, Canada Goose Population January

Major concentrationsoccur at the BlackwaterNational Wildlife Refuge and the .shoalsaround Fox, Smith, and Tangier Islands in Chesapeake Bay; at Back Bay National Wildlife Refuge,Virginia; at Currituck Soundand at Pea Island and MattamuskeetNational Waterfowl Refuges in North Carolina. Smaller concentrationsnumbering from a few hun- dred to 1,000 or more are found scatteredthroughout Pa•nlico and Albe•narleSounds and alongthe C,arolinaBanks. W,ithinthis general region,about 200,000 Canada geese winter annually. The sizeand range of any one winteringflock varies frownyear to year, dependingupon the quality of natural feedinggrounds and the availabilityof supple- •nental feed in nearby agriculturaldistricts. The wintering groundsare essentiallymariti•ne in character,and in keepingwith this habitat,the principalnatural feedinggrounds are beds of submergedaquatic vegetationand the •narshlandsadjoining the coast. Someflocks feed extensively on cultivatedfields, but in •nost cases,the feedingareas lie at no great distancefrom salt water. Only a very smallpercentage of the tidal •narshwithin the wintering range can be classifiedas productivefeeding grounds. The extensive Juncusmarshes of the North Carolina and Maryland coastslack the succulentbrowse and rootssought by the geese.Where available,tidal flatsof Spartinaalternifiora supplement the submergedvegetation. Ex- tensive beds of aquaticsin the waters of Back Bay, Virginia, and Currituck Sound,North Carolina, are an important sourceof foods for the geeseusing these areas in late fall and early winter. When these aquaticbeds are depleted,the geesefeed upon the bedsof American bulrush(Scirpus americanus) found along the inner beachand on the fresh-water•narshes. Where present, extensive use is made of eel grass (Zosteramarina) in ChesapeakeBay and the soundwaters of North Carolina whenevertide levels and weather conditionspermit. Shoal grass(Halodule wrightii) is takenin considerablequantity in Pa•nlico Soundand Widgeon grass (Ruppia maritima) is alsoimportant locally. Intensiveseining in recentyears by com•nercialfishermen has elimi- natedmany .beds of submergedaquatics which formerlyhad attracted large flocksof Canadageese as well as other waterfowl. The practiceof burning coastalmarshes, for•nerly done to afford pasturagefor livestock(and still carriedout by trappers),is followed by wildlifeagencies to provideadditional green food for geeseduring late winter and early spring. To someextent, this activity affectsthe distributionof winteringflocks. The nmnbersof Canadageese that winteredalong the southernJersey coastprior to the disappearanceof the eel grassin 1931-32 are not known,but judging from Urner's recordsfrom the Cape May area (Stone1937) they wereconsiderable. There is no question,however, that the formereel grassbeds and unditchedcoastal •narshes afforded a moreextensive wintering range in that statethan now exists. Some eel grassbeds survived in portionsof ChesapeakeBay and in Pamlico and AlbemarleSounds even after .theplant ceasedto exist throughout much of its for•nerrange, and it is in theseareas where eel grassis •nakinga goodrecovery that Canadageese have now concentratedin n um•bers. Vol. XXIII 1952 HANSON,GRIFFITH, Canada Goose Population [13

Few Canadageese winter alongthe Delawarecoast, possibly because the coastalmarshes there are relatively unproductiveof natural :[oods, no large beds of submergedvegetation being presentexcept at the mouth of the DelawareRiver (SusquehannaFlats). Large flocks of geeseoccur on the marshesof upperDelaware Bay and alongthose of southernNew Jersey,but only after the winteringconcentrations break up. Agricultural trendsare reflectedby local distributionof the wintering flocks. Fields along the ChesapeakeBay shore, formerly planted to grain, attractedand held small winteringflocks. Later, when such areaswere permittedto revert back to the natural vegetation,even the nearbyshoal-water feedin,g grounds were abandoned.Some such areas wherefarming was discontinuedat the turn of the centuryare again beingput undercultivation and againattracting Canada geese. Demand for increasedfood suppliesduring World War I! resultedin the clearing of additional coastallands and the growing o:[ such crops as peanuts, corn, soybeans,and winter grains-- all attractiveto wintering geese. As a result,local concentrations have increased .beyond what wouldhave been the capacityof the range in its original primitive condition. There has been a closecorrelation between agricultural development and winteringgoose populations in the Lake Mattamuskeetarea. Avail- able data indicatethat few birds usedthe area prior to the draining of the lake and the farming of the bottompeat lands. Inability to dry out the lake area and intermittent cultivation of the soil contributed to the developmentof extensivebeds of American bulrush. When floodedshallowly during the autumn,these bulrush beds provided ideal feedinggrounds and attractedthousands of Canadageese. Since the restorationof the lake in 1934 the marshlandsbordering the shoreline have continuedto hold the winteringpopulation. The almostcomplete disappearanceof the eel grass from Pamlico and Albermarle Sounds and the deteriorationof other aquaticfeeding grounds,caused in large measure by commercial fishing activities, are also believed to have influencedthe gradual build-up of the wintering flock at Lake Matta- muskeet that has occurred since 1920. As at HorseshoeLake and other areas in the MississippiFlyway, whereverSouth Atlantic Canadageese have been encouraged to become dependentupon agricultural crops, they seem to lose much of their warinessand are taken by hunters in larger numbersthan are self- reliant flocks,accustomed to foraging on tidal marshesand in shallow bays. In the South Atlantic Flyway wintering grounds,about 30 per centmore geese are baggedin agriculturaldistricts than on the marshes. Althoughthe increasedtake of geesein agriculturalareas is partly a resultof greateravailability to the hunters,the geeseon thesewintering groundsare partially safeguardedby the numerousbays availableto them where they can rest well beyondthe reach of gunners. POPULATIONS Completepopulation figures over a period of years are not available for the SouthAtlantic geese,partly becausethe identity of the various winteringflocks in the middleAtlantic stateswas not fully knownuntil recently. Also, inventory data from earlier years do not match the Bird-Banding ]4] HANSON,GRIFFITH, Canada Goose Population January standardsof recentinventories. For example,data from extensiveareas like ChesapeakeBay for yearsbefore the useof planescan be expected to show, at best, only trends. In someyears, the January inventory indicatedas many as 200,000geese for the entire ChesapeakeBay area, but it is doubtful that 100,000 could now be found there. The only large concentrationsfrequenting the adjoining mainlandoccur at the BlackwaterNational Wildlife Refuge, fig. 5. Large numbersalso use the shoalsaround Fox, Smith, and Tangier Islandsin the lower bay region. The MattamuskeetNational Wildlife Refuge holds the greatestnum- ber of geese,the winteringflock numbering between 40,000 and 70,000 since 1943, fig. 5. The populationat the BackBay NationalWildlife Refugehas peaked at 30,000 or more for six of the last eight years,fig. 5. Someinventory recordsfor Virginia have pointed to a total state-wi.de population of between35,000 and 40,000 Canada geese. The evidenceavailable at this time, however,strongly indicates that the geesewhich inhabit the inland areas, i.e., the total Virginia state-widepopulation minus the Back Bay population,belong to the Southeastpopulation. The smallestmajor concentrationfrequents the Pea Island National Wildlife Refugeand the adjoiningareas of PamlicoSound. It would appear from migration dates and flock movementsthat the geeseat Back Bay and Pea Island probablybelong to the samewintering population. On the other hand, the Mattamuskeetflock appearsto be a distinctentity during the fall-winter period. The trend of the combinedpopulations using thesefour major wintering groundshas been sharplyupwards in recentyears, fig. 5. As the annual total numbersat the four federal refugesrepresent at least 60 per centof the total flywaypopulation, it seemsprobable that total flywaynumbers have al, so increased, but becauseof thetendency of geese to concentratein the vicinity of refuges,it is debatablewhether the over-all increasehas been proportionalto the increaseat the refuges.

ANNUAL BAC The SouthAtlantic geese are huntedon the breedinggrounds by two groupsof natives-- the Eskimosand the CreeIndians. The Eski•nos concernedare distributedalong the eastcoast of HudsonBay; the In- dians that inhabit the breedinggrounds of thesegeese hunt over the areaslying inland from the east coast of JamesBay, fig. 6. Their com- binedpopulation in the early 1940'swas approximately 2,800 (Anon, 1945, Robinson1944). The seasonalaspect of the kill by thesenatives is evident in table 2. Information regardingthe number of Canada geesetaken by these nativeswho hunt out from the various postsis limited. Accordingto studiesmade by A. J. Kerr, an anthropologistin resi.denceat Rupert House in 1947-48, the kill of Canada geeseby Rupert House Indians from July, 1947, to June,1948, was417 birds (letter from J. G. Honig- man to H. C. Hanson, Sept. 10, 1948). The kill made betweenNastapoka Sound and Povungnitukis estimated by Rev. H. S. Shepherdto be between1,500 and 2.000 geese 'Vol. 1952XXlII HANSON,GRIFFITH, Canada Goose Population [15 i Hansonquestionnaire, 1947). D'Arcy Monroe of the Hudson'sBay Companyrecently informed Hanson (August,1949) that the native kill at Povungnitukfor the yearswhen he was stationedat this post, around1946 and 1947, was about2,000. The Royal CanadianMounted Police at Port Harrison estimatedthe Canadagoose kill in their district

POPULATION (THOUSANDS) -•o 5 • • o• • o 8

I I '•. • POPULATION(THOUSANDS)

I I . •. •o o ' [ •- • ?r ' '/ ' • ....

•/ m ,• • X -

Fig. 5.•Populations of Canada geeseat four principal wintering areas on the middle Atlantic coast, 1940-41 • 1950-51. Bird-Bandin g 16] HANSON,Ga,rr•T., Canada Goose Population January Vol. Hill 1952 HANSON,Gaz•'rn',, Canada Goose Population [17 at about 1,000. In the Port Harrison and Povungn,ituk districts the principalkill is made during the summer,table 2, when the geeseare in flightlesscondition. The ReverendMr. Shepherdknew of oneinstance of threehunters killing 130 geeseon a singletrip. A. Lunan of the

A -- WOLSTENHOLME B -- MANSEL ISLAND C -- CAPE SMITH D - POVUNGNITUK E -- PORT HARRISON F -- RICHMOND GULF G - BELCHER ISLANDS H -- GREAT WHALE RTVER I -- FORT GEORGE J -- EASTMAIIq K -- RUPERT HOUSE L -- NEMISCAU M -- NEOSKWESKAU

662

JAMES

BAY

Fig. 6.--Map showinglimits of the trappingand huntinggrounds and sizeof the various bands of Indians and Eskimoseast and south of Hudson and James Ba2s. Solid lines delineate Eskimo territories i from Robinson 194•1); interrupted lines delineate Indian trapping territories. (After Cooper, 1933.) ].8] HANSON,GRIFFITH, Canada Goose Population Btrd-l•ndtag$anuary Vol. XXIII 1952 HANSON',GRIFFITH, Canada Goose Population [19

Hudson'sBay Companyrecently stated (personalcommunication to H. C. Hanson,1949) that 75 per cent of the kill around Port Harrison is of nonbreeding,molting geese. In a letter of August8, 1943, to JackMiner, BrotherG. Lavoiewrote regardingthe goosekill by the Indians at Eastmain:"The best shots here kill about 60, averageis about 20." Inquiry of the Indians at Rupert House revealedthat the largest individual spring kill of Canada geesein that area, usually 25 to 30, was 40 in the spring of 1946. At a point betweenRupert House and Eastmain6 or 7 hunterskilled about 150 geesein the spring of 1946. The averagekill by the Rupert HouseIndians is not over 7 or 8 i letter from Father Damase Couture to H. C. Hanson, Nov. 12, 1946). If Kerr'skill figurefor RupertHouse, 418, is dividedby 111, the approxi- mate number of hunters (men, and boys 17 or more years of age) at this postin 1944 (Anon. 1945), an averagekill of 3.8 geeseper native hunteris indicated.The principalkill by the Indianstakes place during the migration periods. Very few Canada geeseare shot during the nestingperiod asthese Indians spend their summersat the coastalposts. There is ample justificationfor the kills made by the Eskimosand Indians, as often their very survival dependson the availability of geese.The ReverendH. A. Turner (letter to Jack Miner, August 23, 1938) writesfrom Port Harrison: "The peoplehere would often be in a very bad way exceptfor the birds as there are no sealsand fish are not very abundant." An explanationfor the high band recoveryrate from the Port Harrison-Povungnitukregion is furnished by W. A. Tolboom,a postmanager at Povungnituk:"Later on in the seasonthey moult and are then unableto fly. Great num.bersare thus killed by the Eskimoswho hunt inland and run them down. However•being very fast of foot thesebirds have no trouble outrunning a human, and it's only by careful stalkingthat nativescan get them." (Hansonquestion- naire, 1947). The kill made in the inland areas over which the geesepass in migration south of JamesBay is slight, as might be expected. Gerald Parsons,post managerat Obijuan, Quebec,estimated that not over 10 or 15 Canadageese are shot annuallyover an area of 50 squaremiles about that post. The paucity of recoveriesbetween James Bay and the United Statesis further evidenceof a low kill throughoutthis area, fig. 2. One reasonfor so few recoveriesis that the geesemigrate over this forestedportion of the rangewith too few stopsof sufficientdura- tion to afford much shootingto the natives. It wasestimated by Hansonand Smith (1950) that only 1 to 2 per cent of the MississippiFlyway populationis baggedby hunterslocated betweenHudson and JamesBays and the United Statesborder. Native and white huntersresiding in the compara'blearea in Canadathat lies betweenJames Bay and easternLake Ontario are believedto take an evensmaller portion of the SouthAtlantic Flyway populationwhich is available to them in the autumn. Information on the numbersof Canadageese killed on the wintering groundsis very unsatisfactory,few reliable data being available. Al- thoughkill card data are includedin table 5, they can at best be ex- Bird-Ba,ndin• 20] HANSON,Gmr•'•Ta, Canada Goose Population sanuary peeredto showonly trends. Figuresfor the kill in New York and New Jerseyare basedon kill card data, and in all probability exaggerate the actualnumber of geesebagged by severaltimes. The mostdepend- able data are those for the coastal areas of North Carolina where close check is kept of the kill by trained observers.An indication of the shareof the,total kill of SouthAtlantic geesetaken by the variousstates can be gainedfrom compilationsof band recoveries,table 4. If we can assumethat the geesebanded at the Miner Sanctuaryin the spring representan adequatecross section of the entire population of South Arian.ticgeese, a not unwarrantedassumption, a rough appraisalof kill of thesegeese in the United Statesin recentyears can be obtained from the use of band recoveries and kill records, tables 3 and 5. Reliablekill recordsare availableonly for the 1947-48 to 1949-50 seasonand then only for the state of North Carolina. Somegeese are killed elsewherein North Carolina besidesthe localitiesgiven in table 5, but the number would be small relative to the total coastal kill. Band recoveryrecords indicate that the kill in North Carolina,table 5, chiefly the coastalareas, fig. 3, for the period1945-49 as well as earlierperiods, amountsto at least 70 per cent of the total kill of SouthAtlantic geese in easternUnited States. The averagekill in North Carolinafor the period 1947-48plus 1949-50was in the neighborhoodof 7,000 birds. Assumingthen the 70 per centrecovery level from North Carolina,table 4, is representativeof the 7,000 bird kill, the total kill for all eastern statesfrequented by the SouthAtlantic geesehas in recentyears aver- agedroughly 10,000. Admittedlythe trendof thekill in the Lake Matta- mu.skeetarea has beensharply upward for the yearscited; perusalof fig. 5 will revealthat the populationfrequenting that areaalso increased considerably.The kill of SouthAtlantic geese in individualstates other than North Carolinacan be roughlyestimated from bandrecoveries by the methodoutlined above. While this appraisalis opento many criti- cisms,it is the only meansavailable thus far for estimatingthe kill. A similar appraisalof the over-allkill on the breedinggrounds can als9be obtainedby the useof bandrecoveries. Assuming that the kill of 1,500 to 2,000 geesein the Port Harrison district, as reportedby two residentobservers, is approximatelycorrect, and that the number of band recoveriesfrom the varioussectors of the breedinggrounds is approximatelyproportional to the local kills, it would appearthat the kill in the Port Harrisonregion is about36 per cent of a total kill of about6,000 'birds. T•hescarcRy of band recoveriesbetween James Bay and the border is fairly reliable evidencethat the kill of thesegeese in Canadaaway from the breedinggrounds is negligible.Thus, the best estimate we can make of the Canadian kill at this time is that it is roughly one-halfthat of the kill made in the United States.

SUMMARY The Canadageese of the SouthAtlantic population breed inland from the eastcoasts of Hudsonand JamesBays, probably on The Twin Is- landsin JamesBay, on the Belcher Islands and perhaps to a verylimited extent on southern Bail:in Island. In autumn,most of thesegeese migrate south along the eastcoasts of Hudsonand JamesBay to the regionof RupertBay. From there they Vol. XXlII 1952 HANSON,GRIFFITH, Canada Goose Population [21 strike south-eastwardcrossing over into United Statesin the region of westernNew York. On their final flights to their wintering grounds, they mi.grate acrosseastern Pennsylvania, some attaining the Atlantic seaboardin the regionof ChesapeakeBay, othersapparently continuing southwardalong a more inland coursebefore reachingtheir coastal winteringgrounds. Their winteringrange extends from the upperreaches of Chesapeake Bay southto Lake Mattamuskeetarea (chieflyHyde and Dare Counties) of North Carolina. It includesfour national refugestotalling over 69,000 acres. The annualbag of SouthAtlantic •geese during the 1940'sis estima,ted asbeing somewhere in the neighborhoodof 16,000'birds. Roughlyone- third of the annualkill is made by the Indians and Eskimosliving on the breedinggrounds. The relative importanceof the annual kill by thesenatives has probably not changedappreciably in recent decades and is often necessaryfor their survival.

LITERATURE CITED ANONYMOUS 1945 Censusof Indians in Canada, 19•4. Canada Department of Mines and Resources. Indian Affairs Branch. 1-32. BELL, ROBERT 1879 Report of an exploration of the east coast of Hudson's Bay. Report of Progress for 1877-78 Geol. Surv. of Canada ,Part C: 1-37. COOPER, JOHN M. 1933 Aboriginal land holding systems. Ms. memo to Dr. Harold W. McGill, Department Superintendent General for Indian Affairs, Canada. EKLUND, CARL R., and LEON D. COOL 1949 Waterfowl breeding ground survey in the Ungava Peninsula, Quebec. Waterfowl populations and breeding conditions summer 1949 with notes on woodcock and Wilson snipe. Special scientific report: Wild- life No. 2: 19-23. FLAHERTY,ROBERT J. 1918 Two traverses across Ungava Peninsula, Labrador. Geog. Rev. 6(2): 116-132. HANSON, HAROLDC., and ROBERTH. SMITH 1950 Canada geese of the Mississippi flyway with special reference to an Illinois flock. Ill. Nat. Hist. Surv. Bull. 25(3): 67-210. Low, A. P. 1888 Report on explorations in James Bay and country east of Hudson's Bay drained by the Big, Great Whale, and Clearwater rivers. Ann. Rep. Geol. and Nat. Hist. Surv. of Canada. 3 (Part J) 1-94. 1896 Report on the explorationsin the Labrador Peninsula along East Mai•, Koksoak, Hamilton, Manicuagan, and portions of other rivers in 1892- 93-94-95. Canadian Geol. Surv. Ann. Rep. New Ser. 8: rep. L, appendix II (List of birds of the Interior of the Labrador Peninsula), 323-8. 1902 Report on an exploration of the east coast of Hudson Bay from Cape Wolstenholme to the south end of James Bay. Geol. Surv. Canada Ann. Rep., new ser. 8: report D, no. 778, 1-84. ROBINSON,J. LEWIS 1944 Eskimo population in the Canadian Eastern Arctic. Canadian Geog. Jour. 29(3): 128-142. ROUSSEAU,JACQUES 1948 By canoe acrossthe Ungava Peninsulavia the Kogaluk and Payne rivers. Arctic 1(2): 133-5. 22 ] _•,OLETON,Banding at Norristown,Pa. Bird-BandingJanuary

'STONE?WITMER 1937 Bird studies at Old Cape May. An ornithology of coastal New Jersey. Delaware Valley Ornithological Club at Academy Natural Sciences. Piadadelphia. Vol. I. 520 pp. SOPER,J. DEWEY 19'•6 Ornithological results of the Barfin Island expeditions of 1928-29 and 1930-31, together with more recent records. Auk, 63(1): 1-24. Tot)D, W. E. CLYt)E 1938 A new race of the Canada goose. Auk, 55(4): 661-2. WILLIAMS, CECIL 1945 What about Canada geese? Unpublished MS. U.S. Fish and Wildlife Service.

Illinois Natural History Survey' and I'. S. Fish and Wildli/e Service.

THIRTY YEARS OF BANDING AT NORRISTOWN, PENNSYLVANIA BY RAYMOND J. MIDDLETON Our banding station was establishedjust thirty years ago this spring when a single governmentsparrow trap was set out. During the next seven years with one trap we banded over 800 birds. In August of 1928 five sparrow,eight chardonneretand five two-section potter traps were constructedand placed in use; with this addition the fall lnigration broughtus over 1,000 birds in the traps. Severalyears later water drip traps came into use and four char- donnerettraps were made to use with water as bait. Noting that robinsand otherthrushes would walk aroundthis type of trap lookin•g for a groundentrance we deviseda new trap circularin shapeand with two ground funnel-shapedentrances to meet thi,s need, thesebeing made of 3/•-inchpoultry netting. Thesetraps were instantlya tremendoussuccess but as somewarblers and kingletswould enter and push right thru the meshand escape as we reached in to remove them, we now made some with •/•-inch hardwarecloth and thus no small birds could go thru. Later on the multiple sectionchardonneret traps came out and two ten-cell Brenkle traps were made and have been in use since. Three years ago we saw a Modesto trap at one of our annual conventionsand four were made; they have provedexcellent for ground-feedingspecies. We now have all of our thrushtraps, which still carry their original name even though they now catch nearly every specieswe 'band (Fig. 1). These are made with •-inch mesh hardware cloth. Each has a door in the top to removebirds, or a side door if the use of a gatheringcage is preferred. The side door may be at ground level, or (particularly if many warblers are handled] near the top; the trap is sometimesmade with doors at both levels. We use six of the traps all summerwith a water drip, from a buckethanging overhead into a six- or eight-inchflower pot saucer. We use five others in the summerwith bread, crackersand small grain as bait; at other seasons all of thesetraps are usedwith grain as bait. Surprisingly,over half the warblerswe take are caughtin this, a groundentrance trap. It is definitelythe besttrap we have. Sinceusing these and the four Modesto