Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A discussion of the distributional ecology of the Bufflehead (Bucephala albeola; Anatidae; Aves) based… Erskine, Anthony John 1960

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1960 A8 E7 D4.pdf [ 8.37MB ]
JSON: 831-1.0107010.json
JSON-LD: 831-1.0107010-ld.json
RDF/XML (Pretty): 831-1.0107010-rdf.xml
RDF/JSON: 831-1.0107010-rdf.json
Turtle: 831-1.0107010-turtle.txt
N-Triples: 831-1.0107010-rdf-ntriples.txt
Original Record: 831-1.0107010-source.json
Full Text

Full Text

A DISCUSSION OF THE DISTRIBUTIONAL ECOLOGY OP THE BUFFLEHEAD (Bueephala albeola; Anatidae; Aves) BASED UPON BREEDING BIOLOGY STUDIES IN BRITISH COLUMBIA by ANTHONY JOHN EBSKINE B.Sc., Acadia University, 1952. M.A., Queen's University, 1955. Ph.D., Queen's University, 1957. A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Department of Zoology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March, I960 . ABSTRACT This study discusses the breeding biology of the Bufflehead (Bucephala albeola; Anatidae, Aves), and examines factors operating upon the species during the breeding season to evaluate their effectiveness in setting limits to its breeding distribution. A definition of the breeding habitat utilized by the species is offered, with special attention being paid to the form and spatial relationships of the nest-hole. Precise details on nesting are presented and compared with those of related species. Factors which may affect the survival of young under varying conditions are examined critically for correlations with observed mortality. On the basis of these data i t is suggested that climate is of major importance in limiting the breeding distribution of Bufflehead. Climate limits the distribution of trees, and thus of nest-sites, while luxuriant growth of shrubs in regions of high rainfall may limit the fraction of young reaching the water. Adverse weather may cause mortality both by causing chilling and by preventing small young from feeding. It is felt that further studies on waterfowl which involve relationships between nest-sites and territories might profitably use this species as a vehicle. In presenting this thesis i n p a r t i a l fulfilment of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of ray Department or by his representatives. It is understood that copying or publication of this thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of The University of B r i t i s h Columbia Vancouver 8, Canada, Date > r ^ U &,#10 — i i -ACKNOWLEDGEMENTS I wish to express my thanks and appreciation to a l l who have helped with this study, and in particular to the following: to Dr. I.MoT. Cowan, Head, Zoology Department, University of British Columbia, for general supervision of the problem, and most especially for clearing financial obstacles from i t s path; to Canadian Industries Limited (Ammunition Division), for the award of Wildlife Research Fellowships for the years 1958-60; to the University of British Columbia, for the use of Library f a c i l i t i e s , and for office space and the use of equipment in the Zoology Department; to the members of my Committee, Drs. I.McT. Cowan, J.R. Adams, M.D.F. Udvardy, and J.F. Bendell, for much helpful criticism and discussion during the preparation of this manuscript; to fellow graduate students in the Zoology Department, for helpful and stimulating discussion throughout the study, and especially to M.T. Myres (who originally suggested the topic), and Miss M.F. Jackson, who contributed many valuable f i e l d observations; to Mr. and Mrs. G.G. Gibson, fellow students i n the Zoology Department, for examining the specimens of downy young Bufflehead for parasites; to A.J. Wiggs and W.D. McLaren, fellow students i n the Zoology Department, and to L.G. Sugden, then Regional Biologist with the B r i t i s h Columbia Department of Recreation and Conservation, Fish and Game Branch, for assistance in the f i e l d ; and especially to the last-named, who with his wife made me welcome at their home i n Williams Lake throughout the f i e l d studies; and to P.W. Martin, Regional Biologist with the Fish and Game Branch, and to H. Green and V. Geist, Biologists with the Parks Branch, of the above-named Department, - i i i -for assistance during the v i s i t s to the Kamloops and Wells Gray Park areas; to F. Davis, Highland Ranch, 105 Mile House, for permission to camp upon his property at Watson Lake; to the following, who most kindly answered letters i n connection with this study: H.L. Mendall, Maine Co-operative Wildlife Research Unit; Rev. S.D. Robbing, Editor, Wisconsin Society for Ornithology; L.R. Jahn, State of Wisconsin Conservation Department; A.S. Hawkins, J.H. Stoudt, and E.G. Wellein, of the United States Fish and Wildlife Service, i n Minnesota, South Dakota, and Colorado, respectively; Brina Kessel, University of Alaska; A. Murie, Moose, Wyoming; R.L. Salter, State of Idaho Department of Fish and Game; J.M. B a i l l i e , Royal Ontario Museum of Zoology; N.G. Perret, Canadian Wildlife Service, then i n Manitoba; R.W. Sutton, The Manitoba Museum; W.G. Leitch, Ducks Unlimited (Canada); C.S. Houston, Torkton, Saskatchewan; O.C. Furniss, Parksville, B.C.; R.I. Edwards and R.W. Ritcey, Br i t i s h Columbia Department of Recreation and Conservation, Parks Branch; and most especially to H.A. Hansen and A.G. Smith, United States Fish and Wildlife Service, i n Alaska and Utah, respectively; H.W. Burns, Leduc, Alberta; W.E. Griffee, Portland, Oregon; and A.E. Naylor, State of California Department of Fish and Game; who contributed particularly valuable information; last but not least, to my wife, for contributing moral support from the very-start of these studies, while forgoing my company during much of the time necessary for their completion. Anthony J. Erskine - iv — TABLE OF CONTENTS Section Page INTRODUCTION . « . . « o . . . » . « o o « . « . . . « . » . « 1 The Objects of This Study . . . . » 1 Times and Places of the Field Studies . . . . . . . . . . . 1 Sources of Data . 1 Methods . . . . . • » . • « • • . » » . • . • • » • • « • • 3 The Subject of the Investigation 3 BREEDING HABITAT, WITH PARTICULAR EMPHASIS ON THE NEST-SITE . 5 The Size of the Bird 5 The Nest—Hole . . . . . . » . • . . . . . . . . . . . . . . 6 Location and Orientation of Nest-Holes . . . . . . . . . . . 10 Height Relationships of the Nest-Hole . . . . . . . . . . . 12 Plant Communities and Phytogeographic Regions within the Breeding Range of the Bufflehead . . . . . . . . 14 Abundance of Nest-Sites in Various Areas . . . . . . . . . . 20 "Competition" for Nest-Sites . . . . . . . . . . 22 Usage of Sites by Bufflehead . . . . . . . . . . . . . . . . 23 Homing Tradition . . . . . . . . . . . . . . . < • « . • . * 27 Ter r i t o r i a l i t y 29 Lake-Types used by Bufflehead . . . . . . . . . . . . . . . 33 Recapitulation 34 NESTING BIOLOGY 36 Arrival on the Breeding Grounds and the Start of Nesting . . 36 Laying Schedules 38 Clutch-Size . . . . „ 40 Dates of Laying and Variation in Clutch-Size with Laying Date . . . . . . . . . . . . . . . . . . . . 43 Egg Measurements • 44 Incubation . • 46 Attentiveness . . . . . . « • . . . . » « • « 48 Hatching, and Brooding of Young in the Nest . . . . . . . . 49 Leaving the Nest . . . . . . . . . . . . . . . . . 50 Success of Nests and Eggs ..<>*. 52 Success in leaving the Nest, including Nest-Boxes . . . . . 56 Recapitulation . . . . . . . . . . . . . . . . . . 56 FLIGHTLESS YOUNG, AND MORTALITY FACTORS ACTING UPON THEM . . . 58 Food Supply . . . . . . . . . . . . . . . . . . . . . . . . . 58 Competition . . . . . . . . . . . . . . . . . . . . . . . . 62 Predation . . . • • • • . « • « . « . » . . » . « . « « • . 63 Parasites . . . . . . . . . . . . . . . . . . » 64 Drowning . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Climatic Environment . . . . . . . . . . . . . . . 64 Brood Home Range, and Mixing of Broods 66 Survival of Young on the Lakes 69 _ y — Section Page Brood Schedules 76 Recapitulation • 80 DISCUSSION OF THE BREEDING DISTRIBUTION, MAINLY ON THE BASIS OF FACTORS ACTING DURING THE BREEDING SEASON . . . . 81 Relationships of Breeding Distribution to Habitat Used . . . 82 Relationships of Breeding Distribution to Nesting Biology , . 85 Relationships of Breeding Distribution to Factors affecting Survival of Flightless Young . . . . . . . . . . . . 85 A Comparison of the Breeding Distribution of Bufflehead and of Related Species • • » • • • 87 Systematic Position of the Bufflehead . . . . « . . 88 Evolution of the Present Breeding Distribution . . . . . . . 89 SUMMARY . • . . . . . . . . • . • • • • • • • • • • • « • • • . 94 REFERENCES CITED . . . . . a . * . . . « « . » . « • * • « « . 97 APPENDICES . . • . . . • • « « . * • • • » • • • • • « • • • « 107 Localities at which the Present Studies were carried out . . 107 Procedures and Results of Invertebrate Sampling . . . . . . . . 110 Climatological Data for Selected Stations . . . . . . . . . . 1 1 2 LIST OF TABLES Table Page 1. Measurements of Ducks i n the Mergini . . . . . . . . . . . 5 2. Measurements of Nest-Sites . . . . . . . . . . . 9 3. Angles between Nearest Water and Hole Direction . . . . . 10 4. Numbers of Trees i n Quadrats around Nest-Trees . . . . . . 11 5. Height Relationships of Nest-Holes . . . . . . . . . . . . 13 6. Frequencies of Occurrence of Various Plant Species on Plots around Nest-Trees of Various Species . . . . . 15 7. Female Bufflehead trapped on Nests in Two or more Years . 23 8. Comparison of Numbers of Nests (including those inferred from Broods not assignable to Known Nests) with Numbers of Bufflehead counted on May Censuses . . . . 32 9. Mean Temperatures and Dates of "Breakup" and Laying . . . 38 10. Intervals between the Laying of Successive Eggs . . . . . 39 11. Comparison of Egg-Laying Intervals for Different Eggs in a Clutch . 39 12. Probable Times of Day of Laying 40 13. Clutch-Size of Bufflehead in Various Areas . 42 14. Laying Dates and Seasonal Variation in Clutch-Size . . . . 43 15. Measurements of Bufflehead Eggs 45 16. Incubation Periods in Other Waterfowl . . . . . . . . . . 48 17. Attentiveness of Females during Incubation . . . . . . . . 49 18. Nest Success . . . . . . . . . . . . . . 53 19. Egg Success . 53 20. Condition of Eggs that failed to hatch (Successful Nests only) < 54 21. History of Unsuccessful Nests . . . . . . . . . . . . . . 54 - v i i -Table Page 22. Foods of Downy Young of Bufflehead . . . . . . . . . . . . 59 23. Weekly Mean Temperatures and Total Precipitations at Watson Lake . . . . . . . . . . . . . . . . . . . . . 65 24. Mean Distances between Furthest-Separated Observations of Individual Broods within a Given Age-Class, at Watson Lake, 1958 „ 67 25. Progressive Decline in Average Size of Waterfowl Broods . 71 26. Survival of Young Bufflehead at Watson Lake . . . . . . . 74 27. Survival of Young Bufflehead, Pete Kitchen Lake, 1959 . . 75 28. Buff lehead Brood Dates in Tarious Regions » 77 29. Brood Dates for South-Central British Columbia . . . . . . 79 30. Sources of Distributional Data 81 31. Climatic Data for Selected feather Stations . . . . . . . 83 - v i i i -LIST OP FIGURES Figure to follow Page 1. Measurements of Nest-Sites . . . . . . . . . . . . . 8 2. Distance of Nests from Shore . . . . . . . . . . . . . . . . . 10 3. Height of Nest Entrance-Holes . . „ . . . . . . . . 12 4. Illustrations of Bufflehead Nests i n B r i t i s h Columbia . . . . 14 5. Aspen Groveland and Douglas F i r Woods 16 6. Spacing of Nests at Watson Lake . . . . . . . . . . . . . , 9 31 7. Change of Modal Interval between Laying of Successive Eggs with Position of Egg in Sequence 37 8. Laying Dates, and Variation i n Clutch-Size with Laying Date . 43 9. Measurements of Eggs « « . « . . . , . . . . . . 45 10. Incubation Periods obtained i n this Study . . . . . . . . . . 46 11. Hatching Dates and Survival of Young Bufflehead at Watson Lake, compared with Temperatures and Precipitation during the Same Period . . . . . . . . . . . . . . . . . . . . . . . 66 12. Progressive Decline i n Average Size of Waterfowl Broods . . . 71 13. Survival of Young Bufflehead at Watson and P h i l i l l o o Lakes, 1958 and 1959 . . . . . . . . . . . . . . . 73 14. Breeding Distribution of Bufflehead, based largely upon the literature . . . . « . . . . « . . . . . . * . . . . . . 81 15. Areas covered by the Canadian Shield . „ . ..» 83 16. Breeding Distribution of Flickers (Colaptes spp.) . . . . . . 84 17. Isopleths of Precipitation and Length of Growing Season, compared to Breeding Distribution of Bufflehead . . . . . 86 18. Bufflehead Nests i n Populus woods near Leduc, Alberta . . . . 86 19. Major Migration Directions of Bufflehead, as indicated by Banding Returns . . . . . . . . r . . . . . . 90 in Appendix I, Map of Bri t i s h Columbia, to show Locations of Study Areas li s t e d i n Appendix I . . . . . . . . . . . . . . . 109 INTRODUCTION The Objects of this Study, The major interest i n this study was to learn what factors set limits to the breeding distribution of the Buffiehead (Bucephala albeola). Such factors may act at various stages of the l i f e cycle, but time did not permit a complete l i f e history study, A detailed knowledge of the breeding biology of the species i s necessary for any further studies, and i s not available from published sources* Consequently, f i e l d work i n this study has been concentrated upon securing details of the breeding biology. The influences of various factors acting upon the species during the breeding season are discussed i n an attempt to throw light upon what actually limits the distribution. Times and Places of the Field Studies, The main study area was Watson Lake (not to be confused with the much larger lake of the same name on the Br i t i s h Columbia-Yukon border), about four miles northwest of 100 Mile House, i n south-central Br i t i s h Columbia, while other lakes i n the same general area were visited frequently. Several v i s i t s were made to the Riske Creek area and to the ranges south of Kamloops, while single v i s i t s were paid to Wells Gray Park and to the Big Creek area. Areas i n which studies were made are shown i n a table and map in the Appendices. Studies i n the f i e l d occupied the following periods: April 5-6 and May 2 - August 17 in 1958, and April 25 - August 8 and October 10-11 in 1959, In addition, a v i s i t was made to two lakes near Leduc, Alberta, on August 31, 1959. Sources of Data. The majority of the f i e l d data were obtained by the writer, in the periods li s t e d above. A.J. Wiggs, L.G. Sugden, and W.D. McLaren provided assistance i n the f i e l d , particularly by locating nests, during the same periods. Miss M.F. Jackson and Mr. M.T, Myres contributed data obtained i n the Springhouse area i n the summers of 1952-54 and 1955-56, respectively, Myres also made a very important contribution to this study by banding female Bufflehead on eight of the eleven nests of this species which he located around Watson Lake in May and June of 1957. This provided a valuable third year of data for the main study area. Field observations for 1952-54 and 1955-57 are due to Jackson and Myres, respectively, unless otherwise credited. The f i l e s of the B r i t i s h Columbia Nest Records Scheme were made available for this study, and a l l cards relating to Bufflehead were examined. Such of these data as were used are credited to this source. Banding data other than those collected i n •this study were largely from the records of banding carried out from 1948 to 1952 by the university of B r i t i s h Columbia (U.B.C.) i n collaboration with the B r i t i s h Columbia Game Commission, from 1953 to 1955 by U.B.C. students with outside assistance, and from 1957 to 1959 by the B.C. Game Commission biologists. No Bufflehead were banded i n B r i t i s h Columbia i n 1956. Records of band returns of Bufflehead banded elsewhere have been provided by the United States Fish and Wildlife Service banding office and by Ducks Unlimited. An exhaustive survey was made of the major English-language periodicals dealing with birds, the results obtained being mainly of distributional value. The major references, Bent•s "Life Histories of North American Wild Fowl (Order Anseres)" (1925), P h i l l i p s * "A Natural History of the Ducks" (1925), and Kortright's "The Ducks, Geese, and Swans of North America" (1942), were consulted freely, as was the only previous detailed study of the species, Munro's "Studies of Waterfowl i n B r i t i s h Columbia. Bufflehead" (1942), Wherever possible, insufficiently detailed publications were supplemented by correspondence with the authors, and a number of other individuals were also contacted. Methods. The study i n the f i e l d was carried out by direct observation, with the naked eye or with eight-power binoculars, special apparatus being used only for specific problems. Measurements made in feet and inches i n the f i e l d were converted to the metric system later, but geographic distances are given i n miles. Nest-holes were opened by a procedure developed during the study (Erskine, 1959b), and capture of nesting females was effected by the procedure described i n the same paper. Flightless young were trapped by the technique described by Cowan and Hatter (1952). Procedures for sampling aquatic invertebrates are described i n the Appendix on that topic. It must be pointed out that no reliable method exists for determining the age of female Bufflehead while the bird i s alive, whether in the hand or at a distance. At longer ranges, particularly after the start of the moult i n early June, yearling drakes may also be easily mistaken for females, so census data except for adult males must be used with caution. The Subject of the Investigation. The Bufflehead i s a small, black-and-white duck. More detailed descriptions may be obtained from most standard ornithological works (e.g. Kortright, 1942; Peterson, 1947). Although the most recent Checklist of North American Birds (A.O.U,, 1957) s t i l l places a l l diving ducks except the "Ruddy Duck (Oxyura jamaicensis) and the Mergansers i n one subfamily, the treatment of Delacour and Mayr (1945 and - 4 -1946) seems more accurate. In their work the Bufflehead i s placed i n the Tribe Mergini, between the Goldeneyes and the Mergansers, with which i t shares similarities i n plumage, particularly that of the downy young, and i n habits, particularly that of nesting in tree-holes. One other group of waterfowl, the Cairinini of Belacour and Mayr, also has members which nest i n tree-holes, the Wood Duck (Aix sponsa) being the only North American representative of this group. In this study, comparisons w i l l be made with other members of the Mergini, for the most part, and particularly with the Barrow's Goldeneye (Bucephala islandica), which also nests in tree-holes and which has been carefully studied in the same general area. Authoritative works,dealing with other tribes of waterfowl, which have been used extensively i n comparisons are those of Hochbaum (1944), Sowis (1955), and Mendall (1958). With one exception (Munro, 1942), previous studies of Bufflehead have been superficial, based largely upon observations obtained i n the course of studies on other species, or else highly specialized. In the latter category are two recent studies (Myres, 1959a and b; Dane et a l . i n preparation) dealing with the behaviour of the Bufflehead. Courtship and display have been largely ignored i n the present study, as have a l l questions regarding the moult of adults and yearlings. In the present dissertation, the breeding biology of Bufflehead w i l l be discussed under three main headings. I. Breeding Habitat, with particular emphasis on the Nest-Site. I I . Nesting Biology. I I I . Flightless Young, and mortality factors acting on them. BREEDING HABITAT,.WITH PARTICULAR EMPHASIS ON THE NEST-SITE The Bufflehead's habit of nesting in tree-holes w i l l be discussed in considerable detail. The terms "holes" and "hole-nesting" w i l l be used throughout as synonymous with "tree-holes" and "tree-hole-nesting". The Size of the Bird, Most of the common hole-nesting ducks i n the Mergini are similar in size. The Common and Red-breasted Mergansers (Mergus merganser and M, serrator) are slightly larger and the Hooded Merganser (Lophodytes cucullatus) slightly smaller than the other three Mergansers and the Common (Bucephala clangula) and Barrow's Goldeneyes, Even the Smew (Mergus albellus) is very much larger than the Bufflehead, The following table illustrates this point. Table 1, Measurements of Ducks in the Mergini, v Species Sex Common Merganser Male Female Red-br, Merganser Male Female Hooded Merganser Male •Female Smew Male Female Common Goldeneye Male Female Barrow's Goldeneye Male Female Bufflehead • Male Female Length (cm.) Weight (gm,) (No, birds measured i n parentheses) 65 (42) 1560 (35) 58 (39) 1075 (20) 59 (27) 1190 (15) 53 (21) 820 (11) 46 (31) 680 (19) 44 (23) 550 (12) .41 (-) 480-740 (-) 49 (43) 980 (36) 43 (40) 780 (33) 49 (2) 1300 (1) 43 (2) 740 (1) 38 (34) 450 (17) 34 (16) 340 (14) Not taken 290 (39) These data, are taken from Kortright (1942), except those for the Smew, which follows Fisher (1951), and those for female Bufflehead marked *, which were incubating birds weighed in this study). - 6 -The size of these birds i s dwelt upon at some length here, since i t i s of fundamental importance to the nesting habits. The Bufflehead i s the only speeies small enough to enter and use the holes of Flickers (Colaptes), Other North American hole-nesting ducks depend upon nest-holes of the Pileated Woodpecker (Dryocopus pileatus), which i s much rarer i n the north than are the Flickers, or else upon Flicker holes enlarged, by rot or other agency, or upon natural cavities, as do those species found i n Eurasia, It might, however, be possible for the Smew to use holes of the Green Woodpecker (Picus v i r i d i s ) . the Eurasian ecological counterpart of the Flickers, where their ranges overlap. Alone among the species discussed so far, the Red-breasted Merganser habitually nests upon the ground, as do the other members of the Mergini, the Scoters, the Oldsquaw, and the Harlequin Duck, Some other species, particularly the Common Merganser and the Barrow1s Goldeneye, sometimes nest upon the ground or i n holes or burrows. It has been stated that in areas where trees are scarce the Bufflehead w i l l also nest i n holes i n the ground, but such statements are a l l traceable to one record (Raine, 1892) of a nest i n a "gopher burrow" (presumably Citellus sp.?) i n southern Saskatchewan. This report seems plausible, but the fact that no similar occurrence has come to light either before or since suggests that adaptability to "substitute nest-sites" is not strong in Bufflehead, The Nest-Hole. Most oologists considered that nests of the Bufflehead were rare and d i f f i c u l t to find. The most probable cause for this fallacy was that such people were not working i n the right areas, the breeding range of Bufflehead being di s t i n c t l y less extensive than has been indicated by many workers. As with most other species, the nests of Buffle-head are readily found, i n those areas where the species i s a common breeder, by the use of techniques adapted to the particular nesting habit involved. Careful searching for Flicker holes around water bodies where Bufflehead are present during breeding season usually reveals some nests, a combination of lamp and mirror being convenient for inspecting the contents of holes too small to admit the hand. This last point may also have contributed to the alleged scarcity of Bufflehead nests. Most workers have fa i l e d to appreciate the smallness of entrance through which the female Bufflehead can pass, and Kortright (1942) has perpetuated this misconception with the statement that entrance-holes "as small as ~h\ inches i n diameter" are used. This appears to be a misquotation from Bent (1925), who correctly quoted Brooks (1903) to the effect that entrance—holes "not more than 3 i inches i n diameter" were normally used. Taverner (1928) also noted that Flicker holes with entrances "no larger than normal for that species" were used by Bufflehead. Although holes of the Pileated Woodpecker are also used, particularly i n Alberta (Henderson, 1927 ), there can be l i t t l e doubt that Flickers provide most of the nesting sites. Myres started measuring nests of Bufflehead in 1956, and a catalogue of nests measured in Br i t i s h Columbia (deposited with the Zoology Department, University of Br i t i s h Columbia) now includes 107 sites. A l l but five of the nests measured by Myres have been remeasured in this study, as were a l l new sites found. The smallest entrances which have been measured are 5.7 by 5.7 cm. and 6.2 by 5.1 cm., the horizontal measurement always being given f i r s t . These values are near the lower li m i t of entrance size for Flicker sites measured by McLaren and the writer i n the same area i n 1958 and 1959. As noted by Brooks (1903), the majority of holes are smaller than 8.2 cm. (3^ inches) i n diameter, and these smaller holes may have been ignored by oologists since most other species using such sites, v i z . Tree Swallows (Iridoprocne bicolor), Bluebirds (Sialia spp.), and Starlings (Sturnus vulgaris), are well represented i n egg collections from points beyond the breeding range of Buffiehead« Minimum entrance size i s not the only li m i t upon the usage of holes by Bufflehead. The measurements i n the catalogue suggest that shallow nest-holes are avoided, none of those li s t e d being less than 19 cm. from s i l l to bottom of nest. At the other extreme, young Bufflehead have successfully l e f t a nest-hole 134 cm, deep (Erskine, 1959a, referring to nest # 54 i n the catalogue), although in this case a knot-hole about halfway up may have been used rather than the main entrance. In the one case known of a Bufflehead occupying a nest-site i n a stump with an open top (#45), a side entrance was used rather than the larger top entrance. Such sites are often used by Goldeneye, which almost always use the top entrance even when a side entrance of suitable size i s present (Jackson, verbal). In most cases where Bufflehead eggs have been found i n nests with entrances large enough to admit Goldeneye these are incubated by the latter species, unless the cavity i t s e l f is too small to accomodate the larger bird i n a brooding position. The largest entrance measured was 15.2 by 45.5 cm., but the cavity (in a rotted-out Spruce stumpj # 49) was no more than large enough for Bufflehead. Nest-site measurements i n the catalogue are presented graphically (Pig. l ) . These data are grouped to show the modal and less-used values in Table 2, with some measurements of nest-sites of Barrow * s Goldeneye (measured by Jackson, McLaren and the writer) given for comparison. It i s quite apparent that the nest-sites used by Goldeneye are markedly larger, on the average, in entrance and "cavity"-size, as might be expected for a larger bird. The much greater v a r i a b i l i t y of the depths suggests a greater tolerance for this factor i n the Goldeneye, probably brought about by a scarcity of sites meeting the other size requirements. Species No, of Nests 5.1-5.7 5 Entrance Size Ranges (cm,) .7-7.6 7.6-10,0 10.1-15,2 15.2-45.5 v Bufflehead Barrow's Goldeneye* 201 30 2 0 154 34 8 5 16 7 3 2 Species No, of Nests Depth Ranges (cm,) 0-18 18-25 25-37 37-53 53-134 Over : 134 Bufflehead Barrow's Goldeneye 105 19 0 9 4 2 76 14 6 0 5 1 5 2 Species No. of Nests under 8,8 "Cavity" Ranges (cm,) 8.8-11.4 11,4-16.5 16.5-22.8 22.8-24.0 Over 24 Bufflehead Barrow's Goldeneye 99 14 0 0 10 56 32 0 1 8 1 0 0 5 1-3 <t> o H j 52! ro 01 I ® I I * Seven other nests were in "open-top stumps". (Height and width of entrances are grouped together; depth refers to the distance from entrance s i l l to bottom of nest; "Cavity" refers to the distance from inside of s i l l to back of excavation at same level, this being useful only for comparison with other nests so measured). - 10 -Location and Orientation of Nest-Holes. If Flicker holes are numerous i n the study areas i t might be expected that a hole-nesting duck would preferentially occupy those sites situated near to water. Distances of occupied nests from shore are plotted i n Figure 2, and i t appears that this i s actually the case. Trees or stubs actually standing in the water were sometimes u t i l i z e d , but no particular preponderance of nests i n such sites was noted except at Gummings and Soda Lakes, where the majority of the woodland fronting on the lake i s quite dense. There seems to be no clear tendency for the nest entrance to face the water, at least for nests i n deciduous trees, although every site i n trees standing i n the water near the shore faced towards the open lake. The orientation of the entrances with respect to the shore i s shown i n Table 3. Table 3. Angles between nearest water and hole direction. No. nests Tree i n Water Angle Under 90° 90° Over 90° In deciduous In conifer 2 8 27 24 8 7 22 5 Field observations give the impression that the entrances to Bufflehead. nests tend to face towards the open, whether this be the edge of the lake, the open range, or a clearing or opening i n woodland. A quantitative measure of this tendency was sought, and the following method was used. The trees were counted on four quadrats (10 metres square) each having one corner at the nest tree, but when counts for a l l nest-trees were combined there was no significant difference between the number of trees i n the quadrat in front of the nest and those beside or behind i t . to follow p.10 However, when the totals for the front quadrat were combined with whichever of the side quadrats had the lower value, a significant tendency (p=0.05) to face into that half of the f u l l circle having the fewer trees was demonstrated. Two further assumptions had to be made, namely: a l l trees which did not reach up to within five feet of the entrance hole were eliminated, this being only of significance for high nests, and quadrats having fewer than ten stems were considered as neutral i n effect, since no serious obstruction to f l i g h t would be imposed by this number. The results are shown i n Table 4. Table 4. Numbers of trees i n quadrats around nest-trees. Group A+B vs C+D A+C vs B+D Total 1) A+B = C+D 4 A+C = B+D 7J 11 2) A+B> C+D 50; A+C > B+D 38; 88 3) C+D < A+B < 10 9} B+D < A+C < 10 13; 22 4) C+D < A+B > 10 27; B+D < A+C > 10 32; 59 ("A" i s the quadrat i n front of the entrance, "D" that behind, "B" and "C" those on either side). Groups l ) and 3) are considered neutral i n effect, leaving scores of 88 "facing the open" against 59 "facing away from the open". It i s probable that a more satisfactory measure could be evolved for this tendency. It seems l i k e l y that the undulating f l i g h t of Flickers must demand that nests of that species (and thus of Bufflehead) face towards the open. On the downward swoops the wings are closed and l i t t l e manoeuvering i s possible, and leaving the -nest i s usually achieved by such a swoop, of perhaps one or two metres i n depth and three to five metres long (cf. Burns, 1900). The entrance to the nest must face towards - 12 -a f l i g h t path of these dimensions. The Bufflehead female appears more manoeuvrable i n the woods than i s a Flicker, and i t i s probable that Bufflehead have no d i f f i c u l t y i n flying to and from almost any hole made by a Flicker or larger woodpecker. Such analysis of f l i g h t mechanics i s better done from motion picture sequences, but i t seems l i k e l y that these generalizations, based upon numerous observations, would be supported by the more refined technique. Height Relationships of the Nest-Hole. It seems doubtful that Bufflehead have any preference as to the height of their nests above ground or water. The catalogue shows a majority of nests at heights between one and three metres, but t h i s , besides being the range i n which nests are most easily found, probably reflects the heights at which Flickers are obliged to nest more than i t does a preference on the part of the duck. The holes must be made at levels such that the branches of the canopy w i l l not obstruct f l i g h t , and where the trunk i s of sufficient diameter that a cavity of suitable size may be excavated within i t . A considerable proportion of nests are made in stubs, many of which are only two or three metres i n height. However, holes tend to be made near the tops of stubs, without reference to the height of the l a t t e r . Although many holes i n t a l l stubs were not reached, enough nests were located at heights above six metres to suggest that these represent a significant fraction of available sites i n some regions. This i s particularly the case in logged—over or burned areas, where only t a l l stubs project above the regenerating forest. The heights of the entrance-holes are shown graphically (Fig. 3), while other pertinent data are tabulated below. to follow p.12 Figure 3. Height of Nest Entrance-Holes» (this sample is biassed towards heights of less than 6 metres, since higher sites are d i f f i c u l t both to locate and to inspect.) - 13 -Table 5. Height relationships of nest-holes. Number of Nests Hole Height (m.) under 0.5 0.5-1.1 1.1-3.0 3.0-6.7 6.7-15.2 109 0 7 56 36 10 Number of Nests Condition of "Trees" Live Trees Dead Trees Stubs 108 31 20 57 Number of Nests Height of "Trees" (m.) under 3 3-6 6-9 9-12 Over 12 In trees 51 In stubs 57 0 1 13 24 13 15 20 14 4 4 Number of Nests Distance from top of stubs, range (m.) under 0.6 0.6-1.2 1.2-1.8 1.8-2.4 Over 2.4 57 21 14 7 4 11 One other factor likely to influence the height of Flicker holes, and thus also Bufflehead nests, is the height of the shrub layer. In the main study areas the shrub layer is quite low and often, especially in grazed areas, almost completely lacking. Prostrate shrubs figure prominently in the ground cover, and very few plants other than tree regeneration exceed one metre in height, thus presenting l i t t l e obstacle to nests in the one to three metre range, and none to higher nests. By contrast, of eight sites found by H.W. Burns near Leduc, Alberta (six of which were visited in this study), only one (2.3 m.) was less than 3.7 metres from the ground, the others ranging up to 7.5 metres, and most of the nests reported in the literature from Alberta, Saskatchewan, and Manitoba, are also in this range. The shrub layer at Leduc averaged 2.5 to 3.0 metres in height in many places, almost always being at - 14 -least 1.5 metres t a l l and often very dense, which set obvious lower limits to the usable height of Flicker holes. Furthermore, the trees there averaged 15 metres i n height (compared to about 10 metres i n the main study areas; Table 5) with a corresponding rise i n the level of the bottom of the canopy, so a greater average length of trunk was available for Flicker use. Plant Communities and Phytogeographic Regions within the Breeding Range of the Bufflehead. Differences i n the height of the shrub layer around Bufflehead nest-trees i n different areas are probably associated with actual differences i n the plant communities involved. The species of trees u t i l i z e d vary with a v a i l a b i l i t y , but Flickers show a marked preference for species with soft wood. These include particularly various poplars (Populus spp.), as well as dead trees and stubs of almost any species that by rot becomes soft enough to be readily excavated, notably Douglas F i r (Pseudotsuga menziesii) i n Br i t i s h Columbia. Even when dead, pines (Pinus) and spruces (Picea) are rarely rotten enough to be suitable, the more so since both usually grow i n dense stands and often (especially Pinus contorta) with very slender stems. Bufflehead nests have been found i n the following tree species in B r i t i s h Columbia (mostly during the present study): Aspen (Populus tremuloides) 63, Douglas F i r 35, Lodgepole Pine (Pinus contorta) 4, Ponderosa Pine (P. ponderosa) 5, Spruce (sp.?) 3; total 110. Typical nests i n Aspen and Douglas F i r are illustrated i n Figure 4. The vegetation was li s t e d on 20-metre square quadrats around a l l nests visited i n the present study, except those in Alberta, namely: 57 i n Aspen, 33 i n Douglas F i r , 4 i n Lodgepole Pine, 2 each in Ponderosa Pine and i n Spruce. Frequencies of occurrence of the various plants found were computed. Much the same species, both of trees and of shrubs, were found around a l l nests, with two to follow p.14. Figure 4. Illustrations of Bufflehead Nests in British Columbia (note general lack of shrub layer in both pictures). ( a) Typical nest (# 51) in Aspen groveland, Riske Creek, B.C. (b) Typical TIest (§= 41) in climax Douglas Fir woods (Aspen regeneration in background), Phililloo Lake, B.C. - 15 -exceptional Ponderosa Pine was found only around the two nests in trees of that species (# 75 and # 87), while Betula glandulosa was found only around the one nest near Big Creek (# 107). This suggested that only one "bioclimatic zone" was involved for the other 95 nests, but the relative frequencies of occurrence of certain trees and shrubs suggest that two different "associations" within that zone are u t i l i z e d . This may be seen from the following Table, Table 6. Frequencies of Occurence of Various Plant Species on Plots Around Nest-trees of Various Species. Per cent of plots on which species found"*" Species Nests i n As DP LP * Sp * pp Betula occidentalis 11 45 0 0 0 Juniperus scopulorura 2 12 0 0 0 Picea sp.? 7 12 25 50 0 Pinus contorta 70 45 50 0 50 P. ponderosa 0 0 0 0 50 Populus tremuloides 98 55 100 0 100 Pseudotsuga menziesii 9 94 0 0 50 Salix sp.? 33 15 0 0 0 (Regeneration of trees) 64 25 0 0 Betula occidentalis 9 Juniperus scopulorum 11 24 25 0 0 Picea sp.? 5 12 50 0 0 Pinus contorta 60 21 75 0 0 Populus tremuloides 67 73 100 0 50 Pseudotsuga menziesii 5 79 0 0 0 Salix sp.? 33 33 75 0 0 (Shrubs) 0 0 Amelanchier a l n i f o l i a 58 52 0 Arctostaphylos uva-ursi 44 70 50 0 0 Berberis repens 2 3 0 0 0 Cornus stolonifera 0 6 0 0 0 Juniperus communis 9 12 25 0 0 Linnaea borealis 7 21 0 0 0 Ribes lacustre 21 21 0 0 0 Rosa sp.? 90 100 100 0 50 Sheperdia canadensis 49 70 25 0 50 Symphoricarpus sp.? 37 21 25 0 0 (No shrubs at a l l 5 - 0 0 100 50) (Nest-tree species are abbreviated as follows: As - Aspen, DF - Douglas F i r , LP - Lodgepole Pine, Sp - Spruce, PP - Ponderosa Pine). * Both nests i n Spruce were i n stubs standing in water, hence the lack of trees and shrubs around themj one nest in Ponderosa Pine was i n a tree standing alone i n grassland. + Not including nest-tree i t s e l f . - 16 -According to Tisdale and McLean (1957) the study areas lie within the Interior Douglas Fir Zone, Their studies were based largely upon the Tranquille Forest some 80 miles to the southeast, and almost no systematic botanical work has been done on the areas covered in this study. However, although the criterion for selection of the plots examined here (i.e. presence of a Bufflehead nest) would hardly be acceptable to a botanist, a close correspondence was observed between the frequencies of occurrence of the trees and shrubs around the nests in Aspen and Douglas Fir and those in the "Populus - Calamagrostis" and "Pseudotsuga - Calamagrostis" associations, respectively, of Tisdale and McLean (1957, supplementary table). These types are illustrated in Figure 5. The following qualifications should be made, however; l) frequencies of Salix spp, and of Betula occidental is do not correspond well, probably because in this study they occurred principally as water-edge species; 2) frequencies for Linnaea borealis and for Berberis repens were much lower in this study than in the closed forest stands studied by Tisdale and McLean; these plants were observed in closed stands, where nests did not occur, in the areas studied here also; 3) Vaccinium membranaceous and V. caespitosium, found abundantly by Tisdale and McLean, were not noted he re. The Douglas Fir association has that species as climax tree and as a major element of regeneration, while Aspen and Lodgepole Pine occur mainly in successional stages; Betula occidentalis grows commonly at the water's edge, while Arc to staphylo s and Sheperdia are the most diagnostic shrubs. Aspen dominates the other association, with Lodgepole Pine as a frequent companion, both also appearing in regeneration; Salix spp, at least partly replace Betula along the shores, while the most diagnostic shrub is Symphoricarpus. The Aspen-Pine community occurs on the presumably more alkaline soils bordering to follow p.16. Figure 5. Aspen Groveland and Douglas F i r Woods. (a) Aspen groveland and open range around Pete Kitchen Lake, B.C.; Douglas F i r woods on ridges beyond. (b) Climax stand of Douglas Pir (18-2S m.tall), with Aspen regeneration at right; nest jf= 41 is in stub (5 m.tall) at l e f t centre; Phililloo Lake, B.C. - 17 -either alkaline lakes, or grasslands, which in many cases fringe white alkali flats devoid of plant cover. Fire is probably also instrumental in extending the areas of Aspen - Pine forests, and in some areas, particularly at higher elevations, repeated burning on poor soils has led to pure Lodgepole Pine stands of vast extent, the "Pinus — Calamagrostis" association of Tisdale and McLean (1957). These last areas are of l i t t l e significance in the distribution of Bufflehead, since Flickers prefer other tree species for nesting sites, but data from this study and from other published and unpublished reports suggest that the Bufflehead breeds throughout the other associations of the Interior Douglas Fir Zone of British Columbia, Although Douglas Fir also occurs extensively through Washington and Oregon to northern California, the only other record of Bufflehead breeding which can be referred to this zone is from central Oregon; W.E. Grriffee (in l i t t . , 1959) reported a nest found in 1959 on the eastern slope of the Cascade Mountains, at about 1000 metres elevation. This nest was situated in a "cottonwood" (presumably Populus trichocarpa) close to a lake, and is the only record to date for a Bufflehead nest in this tree species, although such doubtless occur frequently in some areas. Below the Douglas Fir Zone lies a zone with Ponderosa Pine as dominant tree (cf. Halliday, 1937; Krajina, 1959). The only two nests found in Ponderosa Pines in this study were situated at 1050 metres elevation and were felt to be probably in the lower part of the Douglas Fir Zone, but i t is possible that some of the nests found in the Okanagan Valley by Munro (e.g. 1918) may have been in the lower zone. This zone extends south to California, and the small area where Bufflehead breed in the northeastern part of that state may be referable to that zone (cf. Grinnell et al, 1930; Miller, 1951). No records are known for the extensive Ponderosa Pine "parklands" in Oregon, Washington, Idaho, Montana, Colorado, and South Dakota, _ 18 -On the wetter, west-facing slopes of the Pacific coast and Rocky Mountains are two zones characterized by forests of Western Hemlock (Tsuga heterophylla) and Western Red Cedar (Thuja plicata). A few Bufflehead breeding records are referable to areas close to or just within these zones. These include two records from Wells Gray Park (R.I. Edwards, in l i t t . . 1959; R. Helsitt, verbal), one from Hazelton (Taverner, 1919b), and one from Sumas Lake (Brooks, 1917). Of these, only that of Helsitt from Murtle River suggests other than accidental breeding, and that area is referred by Halliday (1937, map) to the "Central Douglas Fir Section" rather than to the Hemlock-dominated "Columbia Forest", Above the Douglas Fir belt l ie "the subalpine forests of Englemann Spruce (Picea englemanni) and Subalpine Fir (Abies lasiocarpa). in -which Lodgepole Pine covers extensive areas in succession to fire or on poor soil (Krajina, 1959). One nest found in this study (# 107) appears referable to this zone; i t was situated at about 1450 metres elevation (400 metres higher than any other nest in the catalogue) at the edge of an extensive forest of Lodgepole Pine, in which were scattered Spruces and Aspens, Betula glandulosa was found on this plot and on no other, as was a distinctive Salix species. Two other breeding records from British Columbia (Cowan, 1948; Ritcey, in l i t t . . 1959) also represent this zone, as far as one can judge from the elevations and vegetations of the areas, while two records from high elevations in Oregon (Evenden, 1947; Griffee, 1958) probably also belong here. There are two records from lellowstone Park in Wyoming (Rosche, 1954), as well as two less certain ones from adjacent parts of Idaho (Salter, 1953, in l i t t . . 1959; Steel et al, 1956), which were also at high elevations and were probably also in the subalpine forest zone0 The subalpine spruce-fir forests of the western mountains blend almost - 19 -imperceptibly with the boreal and sub-boreal spruce-fir forests (Halliday, 1937; Krajina, 1959), which stretch across North America from Alaska to Newfoundland. Although the literature contains many references to Bufflehead breeding in this area i t may be questioned whether the climax spruce forests are utilized in most cases. More probably i t is the Aspen groves of fire succession and the river-side communities of Black Cottonwood (Populus trichocarpa) and Balsam Poplar (P. balsamifera) that are the centres of breeding, with only scattered nesting along water bodies bordered by spruce itself. Hansen (in l i t t . , 1959) states that in Alaska "the Bufflehead is further restricted to the major river valleys where adequate stands of cottonwood are available for nest sites". Baird et a l (1884) also mention nests found in poplar trees along the Black River in Alaska. Rand (1946) and Godfrey (1951) noted many broods along the river valleys of Yukon Territory, in which Populus species were described as abundant. There are two records for western Ontario ( B a i l l i e , i n l i t t . , 1959) in boreal forest areas, but precise information on the presence or absence of poplars is lacking. The second major breeding area of Bufflehead is the "poplar region" and parkland belt along the southern border of the boreal spruce forests. Numerous references i n the literature deal with the Bufflehead in both the isolated parklands of the Peace River area (e.g. Cowan, 1939) and in the main parkland - poplar belt (e.g. Taverner, 1928), while f l o r a l discussions of various parts of this region are given by Bird (1930), Moss (1932), and Lynch (1955). Bufflehead breeding records are numerous in this region in central Alberta, but rapidly dwindle as the parkland belt narrows to the south along the eastern slopes of the Rockies and to the east in Manitoba; in both cases almost no records are known south of the 49th parallel, A few records, mostly from early work and thus impossible of verification, come from valleys in the - 20 -prairies and the Great Plains, perhaps from breeding in riparian communities, in which cottonwoods are prominent. The question of whether the Bufflehead ever bred in Maine, where climax forests include Hemlock (Tsuga canadensis), Red Spruce (Picea rubens), and the so-called "Northern Hardwoods", has been threshed out without positive conclusions being reached. Possibly accidental breeding may have been reported there without details, so that later workers assumed breeding to be regular. Another extra—limital record, from Ontario (Morden and Saunders, 1882), has been discredited (Baillie and Harrington, 1936) with apparent consent of the authors, while other records from the eastern part of "that province (Alberger, 1890} Forster, 1772) lack details without appearing improbable. Abundance of Nest-Sites in Various Areas. Quantitative data upon the abundance of Flickers in various parts of North America, both within the breeding range of Bufflehead and outside i t , would be desirable, since the abundance of Flicker holes may be expected to roughly parallel that of Flickers. This type of data is best obtained in North America from the breeding-bird censuses published annually in "Audubon Field Notes" and its predecessors, "Audubon Magazine", and "Bird-Lore". This source of data is, however, only of questionable value for Flickers. In theory, such censuses cover uniform tracts, of woodland, scrub, grassland, marsh, etc., whereas Flickers are a "forest—edge species". Censuses covering pure stands may thus not represent its populations with any degree of accuracy. However, there are enough censuses which have failed to restrict themselves to uniform habitats that, by careful selection, populations of edge-areas may perhaps be approximated. Almost no counts have been made in any regions where the Bufflehead is known to breed, except for a single count from the Mackenzie Valley (/_ 3.3 pairs per 100 acres; Stewart, 1955), so the best basis for comparison is some rough - 21 -estimates of Flicker density in interior British Columbia which were made in the course of this study from counts of known Flicker nests. These estimates approximated 4-5 pairs per 100 acres of Aspen groveland, which is not dissimilar to those from generally coniferous forest in Nova. Scotia, Maine, and Vermont (5 pairs/100 acres), for deciduous forest in Ohio, Indiana, Illinois, and Michigan (3 pairs/100 acres), and for coastal deciduous forest in New Jersey, Maryland, Virginia, North and South Carolinas (6 pairs/100 acres), a l l of these being based upon averages of a number of counts by various workers (Audubon Field Notes, 1936-1958). Much lower values (J_ 1 pair per 100 acres) were obtained from a smaller number of censuses in Ponderosa Pine and other dry forests in South Dakota and Colorado, whereas much higher values were obtained upon three censuses in valley bottomlands in Colorado and Wyoming. Extreme southern areas (Georgia, Alabama, Texas, Mexico) were not considered, nor were the very few censuses from lowland areas in California and Oregon. If these data present even an approximation of Flicker abundance in various areas, i t seems 1ikely that Flickers provide an adequate supply of holes both in the range of Bufflehead and beyond i t . However, it is certain that Flickers do not occur with equal abundance in a l l localities within a region. As mentioned before (pp. 11-12), i t is unlikely that Flickers will nest in areas which are densely forested, and i t was quite obvious in the main study areas that Aspen groveland supported far higher populations of Flickers than did the Douglas Fir woods. For example, around Watson Lake in groveland a total of 79 nest-holes apparently suitable for Bufflehead were examined in 1958 and 1959. 22 were used by Bufflehead in these years, and three more were known to have been used in 1957. Around Phililloo Lake in climax and logged-over Douglas Fir woods only 25 suitable holes were found, of which 14 were used by Bufflehead in 1958 and 1959, three - 22 -others containing evidence of former use. While the coverage at Phililloo Lake was less complete, i t is felt that this discrepancy would by no means account for the three-fold difference in abundance of holes around these two lakes of roughly comparable size. Thus scarcity of nest—sites may be of importance on a local scale, "Competition" for Nest-Sites. "Competition" is used here in the broadest sense; e.g. two species or individuals which can use the same nest-hole are competing. The Bufflehead and its relatives are not the only birds which use old woodpecker holes for nesting, A detailed study of interactions between various hole-nesting species is presently underway (McLaren, in preparation). The following conclusions are in part based upon that study and in part upon observations made in this and previous work. No competition is provided by members of the hole-nesting families Paridae and Sittidae, which use much smaller holes, but a number of species besides Flickers have demands which overlap those of Bufflehead. Chief among these are the Common Starling, Mountain Bluebird (Sialia currueoides). and Tree Swallow, while rarer species such as the Sparrow Hawk (Falco sparverius). Hawk Owl (Surnia ulula), and Saw-whet Owl (Cryptoglaux acadica), may be of some influence (e.g. Henderson, 1927). The Bufflehead has a distinct advantage over many of these species in that i t starts nesting earlier. On the other hand, the Bufflehead is unable to remove from the nest the closely-compacted mass of straw and debris resulting from the nesting of most of the other species listed, A nest may thus be f i l l e d up, by consecutive nestings of other species, beyond the minimum depth tolerated by Bufflehead. The Bufflehead has been known to dispossess other species. One female was caught on the same nest in both 1958 and 1959 (# 51), a Starling nest and eggs found on an earlier visit in 1959 having been destroyed (by the Bufflehead?). In at least two cases in 1959 (# 3 and # 34) - 23 -Bufflehead commenced laying, at rather late dates, upon the almost complete nests of Tree Swallows, one egg of that species being l a i d while the Bufflehead was laying. The reverse situation i s more common, that a Bufflehead nest i s deserted after the eggs have been buried in straw. This can -only occur during the laying period, since later the female incube/tes very closely, A few specific•instances of nest-site competition are cited he re B Attempts by Bufflehead to u t i l i z e a nest-cavity (# 54) suitable for, and occupied by, Barrow's Goldeneye have terminated fatally for the smaller species on at least three occasions (Erskine, 1959a, I960), while joint clutches have also been found in other nests (Henderson, 1927; Burns, in l i t t , , 1959), One nest (# 10) was used successfully by Bufflehead in 1957, but in both 1958 and 1959 the laying schedule was interrupted by the addition of straw to the nest, probably by Mountain Bluebirds. In 1958 several eggs were laid after the f i r s t addition of straw, probably by at least two different females (see p,45), but the nest was finally deserted; in 1959 the nest was deserted shortly after the f i r s t addition of straw. In another nest (# 69) Starlings laid one egg; later four Bufflehead eggs were laid, and when finally visited the nest contained a Tree Swallow nest and eggs above the Bufflehead and Starling eggs, a l l having been deserted. One nest (# 103) when found in 1959 contained the accumulated material from two or more Tree Swallow nests, reaching to within 8 cm, of the s i l l , above a thick layer of shell fragments from earlier Bufflehead nestings. In general, relatively few nests are affected by activities of other species while the sites are actually occupied by Bufflehead,- but the use by other species may ultimately render a site unsuitable for Bufflehead, Usage of Sites by Bufflehead.. Nests may become unavailable through - 24 -the activities of other Bufflehead as well as of other species. When a clutch of Bufflehead eggs is deserted during the nesting season, that nest may not be accepted for use i n the following year, the presence of eggs of its own species perhaps deterring a female from starting to lay in a nest. It appears that egg membranes probably disintegrate during their second winter, allowing the eggs to collapse to shell fragments, so such nests w i l l again become available for use. Two nests contained deserted clutches from 1957 when found in 1958; the eggs were removed from one nest (# 47), which was used successfully in 1959, the other (# 48) remaining empty although the eggs had crumbled to fragments by 1959. However, i t seems unlikely that a given bird will return to a nest in which i t has once deserted a clutch, so such re-occupations probably involve shifts of occupants. Banding has provided a picture of some of the local movements of female Bufflehead, as well as showing that birds will often occupy the same nest in consecutive years. Returns having significance in these connections are listed below. Prom the table i t can be seen that birds reoccupied the same site in 10 cases out of 19 pairs of consecutive years, while movements occurred i n nine cases. Nest—site tenacity has been demonstrated by banding i n Barrow's Goldeneye (Jackson), while H,W. Burns (verbal) states that Bufflehead sites may be used for many years in succession. The data given here suggest that such unbroken series in Bufflehead often result from several different females using the nest in succession, - 25 Table 7, Female Bufflehead trapped on nests in two or more years. Band Number Known History of Bird 38-520461 Banded as young 1949, rebanded 515-42805 on nest at same lake 1957, on same nest 1958, 38-520622 Banded as on same young 1950, on nest at lake \ mile away 1952, nest 1954, but not 1953. 505-50120 Banded as 1952, on nest at lake 96 miles away 1957, moved 265 m, W to 1958 (re-) nest. 505-50254 Banded as on subadult (?) 1955, on nest at same lake 1957, nest 1958, moved 67 m, E to 1959 (re-) nest. 505-50471 Banded on nest 1958, on same nest 1959. 505-50472 Banded on nest 1958, on same nest 1959, 505-50473 Banded on nest 1958, on. same nest 1959. 505-50474 Banded on- nest 1958, on same nest 1959. 505-50476 Banded on nest 1958, on same nest 1959. 505-50477 Banded on nest 1958, moved 82 m, SE to 1959 nest. 505-50478 Banded on nest 1958, moved 450 m« SSE to next lake to 1959 (re-) nest, 505-50483 Banded on nest 1958, moved 1100 m. ESE to 1959 nest. 515-13721 Banded as bird on nest. young 1955, on nest at same lake 1957 (1958?, same nest not caught), moved 70 m. SE to 1959 515-42801 Banded on nest 1957, moved 900 m. E to 1958 nest, moved 180 m, S to 1959 nest. 515-42803 Banded on nest 1957, on same nest 1958 and 1959. 515-42804 Banded on nest 1957, moved 134 m. NW to 1958 (re-) nest, moved (from 1958) 255 m. ESE to 1959 (re-) nest. 525-19725 Banded on nest 1958, on same nest 1959. (m, — metres; miles spelled out). Cases are known of nine nests which were used by one female Bufflehead in one breeding season and by a different bird in the next year. In two of - 26 -these cases the former tenant had been shot in the intervening hunting season, but in four cases the original tenant was found upon a different nest in the second year; in three cases the second tenant had been found upon a different nest in the f i r s t year. Although disturbance may well have been responsible for some or a l l of the moves, this does not affect the fact that the sites were reoccupied. Two of these cases probably involve renesting, the f i r s t tenant having deserted its old nest early in the season, while at least one of the new tenants was a two year-old, breeding late and probably for the f i r s t time. Tradition can hardly be invoked to explain usage of one site by more than one bird, unless a l l of the birds using a given nest had resulted from earlier nestings in that site. No proof of such precise homing has been obtained. The desirability of such sites remains an enigma, however; other apparently suitable sites remain unoccupied year after year. Precise data are lacking on the duration of occupancy of a site by Bufflehead, Nine nests on the main study areas are known to have been used in three consecutive years, while seventeen other nests were used in two consecutive years; three other cases are known of nests used in more than one year in broken sequence. Many nest—sites soon become unavailable due to the trees falling down or breaking off at the level of the cavity, Five of 13 nests found in 1952-1955 had fallen down before 1959, while four others had become untenable due to human or insect action, vandals and oologists having chopped open entrances while carpenter ants ate out the floors of the nests. Tree Swallows and Flickers occupied three of the other four nests in 1958 and 1959. Regeneration of tree cover may have been responsible for the final abandonment of one site (# 15), since another site (# 101) only eight metres away but three metres higher up was used in 1959, - 27 -while # 15 has not been used since 1952. Stubs standing in water avoid this contingency, and one of these has provided the only important longevity record for a nest-tree. A Douglas Fir stub (# 106) at Cummings Lake was occupied by Bufflehead in 1941 (Munro, 1942); i t was also or again occupied by Bufflehead in 1954, and was s t i l l usable in 1959, Flickers having been the only recent tenants. In a l l cases where movements are known to have occurred from one year to the next the female Bufflehead had been handled for banding, which of necessity occasions some disturbance. In six cases of the nine movements documented in Table 7 severe disturbance was known to have taken place at the nest occupied prior to the move, and i n four of these desertion resulted (only one desertion resulting from banding operations). Proven movements are relatively short, six of nine being less than 300 metres, and none over 1100 metres. These are of the same order of magnitude as the movements listed by Sowls (1955) for dabbling ducks. In a species with as stringent nest-site requirements as the Bufflehead such a lack of tenacity to the precise site might seem to be an undesirable t r a i t . However, i t is possible that this is a local condition induced by a, ready availability of alternative nest-sites, and the habit of homing to the natal lake in combination with the short distance of most moves probably balances any adverse effects resulting from lack of nest-site tenacity. Homing Tradition. Studies of many bird species (see Tinbergen, 1953) have indicated that either or both sexes may tend to return to the area where they were born or else to that in which they f i r s t bred. The most usually held concept of "homing tradition" in ducks i s that the females mate on the wintering grounds and then home to their natal lake, the males following their mate regardless of their own place of origin. The work of the Delta Waterfowl - 28 -Research Station (Hochbaum, 1955; Sowls, 1955) is recognized as authoritative on this subject, as far as waterfowl are concerned. For female Bufflehead, banding evidence supports homing both to the natal lake and to a previous breeding ground, and also to the traditional "moulting lake" after breeding is finished. Adult and yearling males have been captured in such small numbers that conclusions for that sex are unreliable. In British Columbia, there are now 17 records of adult female Bufflehead banded either on the nest in June or with their broods in July, and retaken in subsequent years under similar circumstances. A l l of these birds, recaptured one or two years after banding, were on the lake where they were banded, except one which had moved to another lake one-half mile away. Eight other females were banded as flightless young and were sub-sequently caught on nests two or more years later. Six of these were on the lakes where they were banded, while two were caught on lakes about one—half mile away. One of these birds (38—520461) is the oldest Bufflehead known, having been almost nine years old when last seen. Although none of the birds banded as young and later caught on nests had moved more than a mile, birds banded as young and later caught on "moulting lakes" are more scattered, two individuals having been caught 17 and 36 miles from the point of banding. Likewise, one adult banded while moulting was later retrapped on the nest at a lake 96 miles away. In general, movements for moulting are short, the other six records involving shifts of less than three miles, as did 12 of 14 records of birds banded as well as retrapped upon "moulting lakes". As a result of movements for moulting, band returns from birds shot in hunting season are of l i t t l e significance in establishing homing. 42 out - 29 -of 51 returns of B.C. banded Bufflehead from the interior of that province are from the f a l l following banding. Seven returns of birds shot prior to mid-October, in years after that in which they were banded, are a l l within 30 miles of the point of banding, but this is not of itself very meaningful. Territoriality. Tolerance of the Bufflehead for others of its own kind must now be considered. As already stated, the presence of deserted eggs in a nest may lead to its disuse in the succeeding season. This fact, combined with the scarcity of extremely large clutch-sizes (over 15 eggs) (see p. 42), suggests that relatively l i t t l e intra-specific competition is centred upon the nest-site itself, at least in areas where nest-sites are numerous. However, intra-specific rivalry in Bufflehead is often noted (e.g. Dice, 1920; Stansell, 1909a). Although displays involved in such rivalry are not discussed in this study, some of the limits set by territorial competition must be considered, even though colour-marking of adults, the surest method for precise delimitation of territories, was not employed. Various workers, from Altum and Howard to the present, have discussed the subject of territory in bird l i f e , Nice (1941) summarized the available evidence and distinguished several types of territory, differing in the combinations of various activities carried on in them, viz, feeding, mating, but a l l characterized by being "defended areas", Hochbaum (1944) discussed territory in several species of ducks, pointing out the need for water, loafing spots, food, and nearby nesting cover, to occur together on the territory, Sowls (1955) favoured the term "home range", which included besides those areas defined by Hochbaum al l other areas habitually frequented by birds of a pair during their summer period of - 30 -isolation. He also pointed out that wide variations occur in the frequency and intensity of t e r r i t o r i a l defence, both between species and within a species, Mendall (1958) f e l t that the concepts of Hochbaum (1944) and Sowls (1955) applied to "dabbling" ducks, but that in. the Ring-necked Duck (Aythya cpilaris) "mutual respect" between pairs was a major factor in maintaining terr i t o r i e s , once nest-sites had been selected, actual f r i c t i o n being very rare a,nd defence of the territory, as distinct from that of the female, seldom noted, Munro (1942) stated that "a breeding pair (of Bufflehead) establishes a definite territory which the male vigorously defends from encroachment by other males". In this study, although, males were seen to chase yearlings (males?) on or near the area where they were feeding alone, no conflicts were witnessed between adult male Bufflehead except when one was accompanied by i t s mate. This is similar to Carter's experiences (1958) with the Common Goldeneye, Thus, defence of the female seems more probable than defence of a territory. However, there i s no doubt that breeding pairs are well spaced along the shores of larger lakes, and as Mendall (1958) noted for the Ring—necked Buck an acceptable degree of isolation i s maintained, with l i t t l e conflict after the i n i t i a l shuffling associated with nest-site selection. For convenience, the term territory w i l l be used in this discussion, but with the understanding that only a concept similar to that of Mendall i s implied by the term. Territory in Bufflehead includes a pond or part of a larger body of water, loafing spots on shore being less important than for dabbling ducks since Bufflehead usually loaf on the water. The nest is distinct from the territory, the female flying thence from the water and vice versa, except under special circumstances. As shown earlier (Fig. 2), the nests are not always close to the water, and the territory does not always include - 31 -that part of the shoreline closest to the nest. This is most noticeable when nests are unusually close together; for example, in 1959 three trees or stubs were each known to contain two simultaneously occupied nests of Bufflehead (# 11 and # 26, # 33 and # 74, # 81 and # 82), a l l of which were successful, More usually, however, nests are spaced out, and some idea.of the size of territory or area of isolation may be obtained from distances between nests. For example, at Watson Lake pairs of simultaneously occupied nests were separated by 114, 82, 112, 165, 67, 250, 80, 75, 100, 139, 140, 134, 122, and 137 metres, respectively, as shown in Figure 6, while some data from other lakes include 38, 215, 380 and 205 metres between simultaneously occupied nests. The minimum distance between nests thus approximates 100 metres, nests separated by less than this distance from a nest on one side usually being separated from nests on the other side by a considerably greater distance. On the larger bodies of water where most of these studies were made (Watson and Phililloo Lakes) not a l l of the area was utilized as territory, the central parts appearing to be used as communal feeding grounds by breeding birds, migrants, and parties of moulting birds. The minimum area may be estimated from the smallest size of pond occupied, which approximates 0,4 hectares (l acre). On such small areas the brood rarely remains on the pond long, and the adults may f l y elsewhere to feed at times, so perhaps 0.8 hectares (2 acres) better approximates the minimum, at least for isolated ponds. In cases where a female was observed to f l y from a nest to join a territorial male connection could be established between the nest and territory, but usually such connections could only be inferred. Censuses of Bufflehead on Watson Lake indicated that the numbers present during May bear no clear relationship to the number of nests occupied. It is assumed that nearly a l l nests there were actually found in 1958 and .1959, to follow p.31 Figure 6. Spacing of Nests at Watson Lake. - 32 -since nearly a l l broods could be accounted for from known nests. This may be seen from Table 8. Table 8. Comparison of numbers of nests (including those inferred from broods not assignable to known nests) with numbers of Buffle-head counted on May censuses. lear Ho. Nests Used No. Nesting Females Dates Censuses M F S 1958 18 14-16 May 8 31 17 5 May 13 27 20 31 May 18 32 19 27 May 24 34 21 12 1959 19 18 May 17 32 11 May 23 29 10 M — males. F - females, S - subadults; see p. 3 for distinction of females and subadults, but parties of 3 or more "female-like" birds were listed as subadults). In both years the number of males censused was approximately double the number of "nesting females", whereas the number of "females" censused differed much less from the actual number nesting. Since counts were a l l made in the early forenoon, i t is quite possible that some females were absent laying or incubating at tne time of the count. It was suggested by (the late) J.A. Munro (verbal) that such May counts probably include late migrants; however, a census by Munro on June 7, 1958, only showed ten adult male Bufflehead, so that by then some of the breeding males had probably left for the moulting grounds, especially since the f i r s t brood appeared on Watson Lake on June 8. Although the June 7 count actually equalled the number of nests from which young were hatched that year, i t is f e l t that this was merely a coincidence. Miss Jackson secured somewhat comparable figures by comparing censuses with numbers of known broods in Barrow's Goldeneye (nests not being included in the comparison), but felt that the discrepancy there might best be explained by a shortage of suitable nest-sites limiting the numbers of pairs actually breeding. It seems unlikely - 33 -(cf . p.2l) that this could be the case for Bufflehead. at least at Watson Lake. In census counts i t is readily apparent that not a l l parts of the shoreline of a given lake are utilized for territories by Bufflehead, and many water bodies lack this species altogether. Evidently certain physical or biological features of lakes are of significance in the selection by Bufflehead of a breeding area. Lake-Types used by Bufflehead, Students of limnology and lake morphometry have made many studies of "lake-types", but classifications based upon vertebrate faunas other than those of fishes are few, A paper by Palmgren (1936) classified the water bodies of the Aland Islands according to their avifaunas, and compared the result with classifications based upon limnological and botanical criteria, Palmgren*s "Nyroca lake" and "Podiceps lake", characterized by the presence of breeding birds of Aythya (Nyroca) ferina and of Podiceps cristatus. respectively, were eutrophic (Potamogeton) types, distinguished mainly by the presence of a hypolimnion in the latter and it s absence in the former, 3 m, depth being roughly the dividing line. The third of Palmgren*s types, the "Colymbus lake" characterized by breeding birds of Gavia (Colymbus)arctica. was an oligotrophia lake. The avian "indicator" species used by Palmgren are lacking in North America; Aythya americana or A. valisineria. Podiceps grisegena. and Gavia immer appear to be the nearest ecological counterparts in the study areas, although these two species of Aythya are by no means restricted for breeding to lakes of less than 2 m. depth, . On the basis of Palmgren*s descriptions of the lake-types and the occurrence of the "indicator" species, the Bufflehead usually frequents lakes of "Podiceps-type", although also - 34 -occurring upon small ponds of similar nature where J?. grisegena i s never found. Generally speaking, favoured breeding lakes for Bufflehead have rela t i v e l y open shores, although woods near the shores are necessary to provide nest-sites, and lakes or bays with extensive reedbeds are avoided. The species is generally lacking both from the very shallow, reedy sloughs of the "Nyroca type" and from the large deep, gravel-shored lakes of the "Colymbus type". Recapitulation. The Bufflehead deviates markedly from other ducks of the Mergini i n size, this adaptation permitting the use of the more numerous cavities of sizes too small for the other species. Probably related to this greater supply of nest-holes i s a decreased adaptability to sites deviating from the form of a typical Flicker hole. Nest-sites are selected near the water, but other spatial relationships are less well defined, except for a need for re l a t i v e l y open areas in front of the nest. In B r i t i s h Columbia, the Interior Douglas F i r Zone i s the main breeding area, few records coming from vegetational areas with other climax species; the main nesting trees are Aspen and Douglas F i r . In other regions, the poplar parkland belt along the northern edge of the Prairies i s a favoured breeding area, while most other breeding probably takes place i n riparian poplar communities. Abundance of nest-sites i s f e l t to be limiting only on a local scale. Several other species occupy nest-sites suitable for Bufflehead, and their occupancy may ultimately make the site undesirable. Such a c t i v i t i e s rarely interfere with use of a nest i n which Bufflehead have commenced laying in a given year. Some sites are particularly favoured and may be used over a period of years, but there may be a rapid turnover of tenants, "Live returns" of banded female Bufflehead indicate that homing to the natal lake i s well-developed in this species, but tenacity - 35 -to a given nest-site i s probably less than for Goldeneye. Even where the species has stringent requirements for its nest-site no ultimate disadvantage should accrue from such movements, in areas where nest-sites are numerous. Territory consists of an area within which the female is defended by the male, and in which the pair are relatively l i t t l e disturbed by other pairs. Territories upon the lakes are related to spacing of nests along the shore-line, but census figures are inadequate to accurately evaluate nesting populations of sizeable lakes. Eutrophic lakes of moderate depth, in which reedbeds are relatively sparse, are favoured for breeding areas. NESTING BIOLOGY This section takes up that stage of "the breeding cycle which involves the use of nests. Arrival on the Breeding Grounds and the Start of Nesting. I t was not possible to be on the study area from the time that the lakes became clear of ice ("breakup") to study arrival and dispersal on the breeding grounds* I t i s usually stated (e.g. Hochbaum, 1955) that ducks pair on the wintering grounds, and that the males then follow their mates to the previous breeding area or natal lake of the latter. However, Munro (1942) has shown that early migrating flocks of Bufflehead include mostly males, and recent data confirm t h i s . The buildup of numbers of each sex is well shown by figures collected i n the San Jose Valley of British Columbia i n 1959 (Sugden, i n l i t t . . 1959). The f i r s t male was seen March 26, the f i r s t female April 9; not more than 11 males were seen on one day u n t i l April 10 when the count was 149 males, 22 females (ratio 677:100). From then on the numbers of both sexes increased, the April 19 count giving 235 males, 104 females (ratio 226:100), which i s comparable to the ratio of 258:100 obtained by Munro (1942) from 441 birds seen i n the period April 16-24, 1941. Counts made i n this study on Ap r i l 5-6, 1958, gave 132 males, 9 females (ratio 1466:100), which approximates the stage reached on April 10 i n 1959. The evidence of these counts indicates that by no means a l l pairing of Bufflehead can occur prior to migration, unless most of the early migrants are birds that ultimately remain unmated, which seems unlikely. Laying was already i n progress when the writer arrived on the study area i n 1958, and i t commenced very soon after his arrival i n 1959. Consequently, much of the process of pair formation would have been missed - 37 -i n any case, and i n fact no study was made of this phenomenon. I t i s assumed throughout this discussion that Bufflehead arrive on the "migration lakes" very shortly after "breakup" on the latter, and that dispersal to the breeding lakes follows as soon as these are clear. Brake Bufflehead were seen on partly ice-covered lakes on April 5-6, 1958, and Jackson (in preparation) found that Barrow's Goldeneye also appeared on the breeding lakes almost as soon as they were open. The f i r s t heavy movement of Bufflehead into the general breeding area occurred about April 10 i n 1959, about four days later -than i n 1958, and the f i r s t appearances of eggs and broods i n the two years were separated by similar intervals. Sates of spring "breakup" are presumably related to temperatures during March and A p r i l . The latter may be obtained from weather station reports, but "breakup" dates are rarely known with great accuracy. Such figures as were obtainable on this subject are given i n Table 9. From general impressions i t appears that 1951, 1954, and 1955, were late i n phenology, 1953, 1956, and 1957 about average, and 1952, 1958, and 1959 early, 1958 i n particular being almost unprecedentedly early. Although the data are much less precise than could be wished, certain trends are discernable. Arrival on the "migration" lakes precedes the "breakup" on the breeding lakes by 10 to 15 days, while the f i r s t eggs are l a i d usually within 10 days of the arrival on the nesting lake. I t may be anticipated that further north the interval between arrival i n the area and start of laying w i l l be shorter, probably within a week of "breakup", while i n more southern regions a longer interval may perhaps occur. - 38 -Table 9. Mean Temperatures and Dates ef "Breakup" and Laying, Year Daily Mean Temperature(°C) March April "Breakup" "Migration" lakes dates for "Breeding" lakes F i r s t Layings 1951 -6 5 1952 -2 6 ca.Apr, 25 by May 5 1953 1 4 ca.Apr. 21 1954 -3 0 May 12 ca.May 15 1955 -7 2 May 6 1956 -3 5 Apr. 28 ca.May 14 1957 -1 5 ca.Apr, 18 ca.Apr, 28 ca.May 1 1958 -1 5 ca.Apr, 5 ca.Apr. 18 ca.Apr. 25 1959 2 5 ca.Apr, 9 ca.Apr, 20 Apr. 27 (Temperatures are the averages of those for Dog Creek Airport and 150 Mile House, taken from "Climate of Bri t i s h Columbia" reports for the years concerned; dates of "breakup" were obtained from Jackson, Myres, and notes of banding crews; "Migration lakes" include E l l i o t , Murphy (142 Mile), Cummings, 149 Mile, and Williams Lake; breeding lakes are Westwick Lake for 1952-1956, while the values for 1957-1959 are obtained by interpolation between April v i s i t s and the start of f i e l d studies i n May). Laying Schedules. Without daily v i s i t s to nests i t i s impossible to pinpoint the start of laying exactly. Sowls (1955) found i t possible to back-date nests of many ground-nesting ducks quite precisely, even after the onset of incubation (by using aged embryos), since the laying schedules almost invariably added one egg per day u n t i l the clutch was complete, Jackson found that this was not the case with Barrow's Goldeneye, and i n this study laying schedules of Bufflehead were also found to be irregular. The times of v i s i t s to a l l nests were carefully recorded, and where more than one egg had appeared i n the interval between v i s i t s the time i n hours was divided by the number of eggs added to obtain an average interval. Figures were obtained i n this manner for the laying intervals of 138 eggs, which are grouped below. Table 10. Intervals between the laying of successive eggs. Range (hr.) under 20 20-30 30-40 40-50 50-60 60-70 over 70 No. of eggs 0 37 57 29 10 2 3 Presumably, since v i s i t s were not made more than once a day and not always at the same time of day, the 20-30 hour range may be taken to represent eggs la i d at one-day intervals, and the 40-50 and 50-60 hour ranges eggs l a i d at two—day intervals. The 30-40 hour group may involve cases where some eggs were l a i d on successive days and some on alternate days. The mean value is 37.7 hours, or approximately 1.5 days between successive eggs. It i s also enlightening to examine whether changes i n interval occur as the clutch is l a i d . A few eggs of which the sequence was uncertain were eliminated from the 138 lis t e d above, leaving 127. These are shown i n Table 11 and i n Figure 7. Table 11. Comparison of egg-laying intervals for different eggs i n a clutch. Range Order of Egg in Sequence (hours) 1 2 3 4 5 6 7 8 9 10 11 12 Total 20-30 2 3 3 2 4 2 4 5 5 3 1 34 30-40 4 6 8 8 7 7 6 2 2 1 1 1 53 40-50 2 2 5 4 4 4 1 1 1 1 25 50-60 1 1 2 2 1 1 1 1 10 over 60 1 1 1 1 1 5 As Figure 7 clearly shows, there is a pronounced shift in the modal interval after the sixth egg has been l a i d in a nest, from two or one and one-half days to one day. In the only nest visited daily during most of the laying period an average interval of at least two days was noted. This, however, is hardly typical, and at the other extreme 12 eggs were l a i d in another nest within 14 days. to follow p.39. Figure 7. Change of Modal Interval between Laying of Successive Eggs with Position of Egg in Sequence. - 40 -Compilation of the times of day at which females were found on nests during the laying period gives a suggestion of the times of day at which eggs were l a i d . Table 12. Probable times of day of laying. Time of V i s i t (PST) Total No. T i s i t s During Laying No. Vi s i t s Female Found on Nest (%) before 0800 2 1 (50) 0800-1000 18 7 (39) 1000-1200 20 5 (25) 1200-1400 18 3 (17) 1400-1600 20 0 (0) 1600-1800 15 0 (0) 1800-2000 10 2 (20) This circumstantial evidence suggests that laying takes place mainly i n the mornings. The two birds found on the nest i n the evening may also have been laying, however; Mendall (1958) l i s t e d one authenticated case of a Ring-necked Duck laying i n the evening, although that species also more often l a i d i n the morning. Clutch-Size. Davis (1955) has c r i t i c i z e d the use of "eggs found i n the nest" as indication of clutch-size, and with considerable justification. No other method i s yet available for the Bufflehead, and the errors involved i n i t s use seem less than for most species discussed by Davis. Unexplained losses of eggs, as distinct from loss of the whole clutch, are rare i n hole-nesting species, and only one such loss was noted at any stage i n a l l of the nests studied. In that case one of the seven eggs previously present was broken, and a Flying Squirrel (Glaucomys sabrinus) was seen leaving the nest. - 41 -The broken egg was removed, and two days later two more eggs had disappeared, but after eleven days laying was found to have been resumed; the female which ultimately incubated the clutch was that banded upon the same nest in the previous year. On rare occasions several females may lay i n the same nest (see under # 10, p.45); however, were this of frequent occurrence the proportion of large clutches (over 12) would be much greater, as i s the case with Wood Ducks (McLaughlin and Grice, 1952) or Goldeneye, for which clutches of 15 or more eggs have been noted not uncommonly. The greatest number of eggs recorded i n one Bufflehead nest i s 16; no other nests with over 14 are known, and even clutches of 13 are rare. The distribution of clutch—sizes i s l i t t l e skewed with respect to the mean, which would probably not be the case were joint layings frequent. Only one case i s known of an egg being dropped elsewhere than i n a nest, this being within five metres of a nest (# 54) for which Bufflehead and Barrow*s Goldeneye were competing (Erskine, 1959a; 1960), On the whole, the number of eggs found i n the nest i s probably a good approximation of the actual clutch-size i n Bufflehead. Some data on clutch-size i n Bufflehead are given i n the literature, but often no indication was given that laying had been completed. I t seems better to tabulate a l l data from the literature, as well as values quoted by other observers, separately from those obtained i n this study. Nine clutches from this study to which doubt was attached were omitted, as were a l l counts based upon numbers of egg membranes found i n the nest after hatching, since this method has been found to deviate from known clutch-sizes by one or more eggs. The data are given below, grouped by p o l i t i c a l divisions. - 42 -Table 13. Clutch-Size of Bufflehead in Various Areas. Area References Number of Eggs 8 9 10 11 12 13 14 15 16 B.C. This study Jackson (unpubl,) Myres (unpubl.) Munro (1918, 1942) Maguire coll'n* Wynne coll'n* Williams (1933) Alta. Henderson (1927) Raine (1927) Stansell (in Bent, 1925) Sask. Arnold (1895) Furniss (1938) Massey (in Bent, 1925) Raine (1892) Man. Raine (1892) M (in Maeoun, 1900) Alaska Lockhart (in Baird et a l , 1884) Oregon Griffee (in l i t t . . 1959) Calif. Davis (1941) Naylor (in l i t t . . 1959) 8 15 15 15 1 2 4 7 2 2 1 6 2 in collection of Zoology Department, University of British Columbia. In view of the small samples from the various states and provinces, means were calculated only for data obtained in this study and for a l l other data together. These values are: present study - M = 8.51, range 5 to 16, s = 2.22} other studies - M = 8.70, range 5 to 14, s = 1.97. Five clutches in the present study were considered to be the results of renesting (M =o 6.80, range 5 to 9, s = 1.64); with these removed the values.for f i r s t nestings in the present study are M = 8.62, range 5 to 16, s = 2.22, - 43 -Dates of Laying and Variation in Clutch-Size with Laying Date, It was shown earlier (Table 9) that the start of laying i s related to the time of "breakup" of ice on the nesting lakes. The dates cited there referred to the earliest known nestings in each year. In the following Table the numbers of known nests initiated during each weekly interval during the nesting seasons of 1958 and 1959 are given, together with the mean clutch-size for nests begun in each week. Since the schedules in the two years were very nearly synchronized the data from both years are combined. Table 14. Laying Dates and Seasonal Variation in Clutch-Size. Period When First Egg Was Laid No. Clutches Begun Mean Size (No. dutches Standard Used Deviation Apr. 24-30 May 1-7 May 8-14 May 15-21 May 22-28 May 29-June 4 17 (+6) 28 (+3) 15, + 1 renest (+3) 12, + 1 renest (+3) 4, + 2 renests 1, + 1 renest 10.2 (15) 9.4 (28.5) 8.2 (10), renest 8.0 6.6 (10), renest 9.0 6.9 (5.5), renest 5.5 6.0 ( l ) , renest 6.0 2.41 2.08 1.85 1.07 1.88 (Nests for which the date of initiation was less precisely known are included between the groups, and are averaged between the weeks involved in calculations of clutch-sizes (hence, "half-clutches"); only complete clutches are included in the clutch-size calculations, but a l l nests in which eggs were known to have been laid are listed in the second column). The sample sizes could with advantage have been larger, but there seems l i t t l e doubt that the clutch-size actually does decline through the course of the nesting season. This is probably in part due to the presence of renesting undetected by the present study, but i t is possible that this is partly an adaptation towards the synchronous hatching of both early and late nests, as 32 No. of Clutches Begun 24 16 Mean Size of Clutches * Started in 7 Week 24-30 April 24-30 April \ Original netts 1-7 «-14 15-21 May 22-2% Date at which laying commenced 2S-4 June 1-7 • 15-21 May 22-21 2* - 4 June Date at which laying commenced d CD ^ t) M » & ct" «< CD p. OS ts » (TO SO Q, c+ © <d © c f H " O y o Ct O N CD cf -O O - 44 -was suggested by Koskimies fox the Telvet Scoter (Melanitta fusca)(1957a), It i s probable that the smaller number of nests initiated i n later weeks i s an actual decline, but i t may also be associated with a smaller intensity of nest-hunting effort i n that period, particularly i n 1958, No striking uniformity of clutch—size for individual birds was noted, whereas Koskimies (1957a) found i n the Velvet Scoter a maximum spread of two eggs for any one bird. I t i s possible that the extreme cases found i n this study (e.g, 1 3 - 7 ; 7 - 12) may be due to two birds laying i n the nest i n one year and only one i n the other, but this seems unlikely (cf, pp. 41, 45-46). The influence of age upon clutch-size has been worked out for a few species (e.g. Richdale, 1949); generally, birds breeding for the f i r s t time tend to lay smaller clutches than do older birds, i f the clutch-size shows a range of values. No yearling Bufflehead have been found nesting, but five two year-olds had clutches averaging 8.2 eggs. 21 clutches l a i d by older birds (or birds known to be breeding for at least the second time) averaged 9,2 eggs, the difference not being significant. More data, particularly for two year-olds, are needed to c l a r i f y this point for Bufflehead. Egg Measurements. A number of sources quote measurements of Bufflehead eggs, but some disagreement exists between the values given. For comparison, these are tabulated, together with the measurements of 143 eggs, from 44 separate nestings, collected i n this study. These were largely deserted clutches or else eggs l e f t unhatched i n the nest, and included eight apparently f u l l clutches. The fact that these were a l l eggs that failed to hatch should not affect the v a l i d i t y of the sample, since i t remains to be shown that the external measurements of eggs (runt eggs aside) in any way - 45 relate to their success* fable 15* Measurements of Bufflehead Eggs* Source No. of Eggs Length (mm.) Mean Bange Width (mm.) Mean Range Bent (1925) 86 48.5 55-40 34.7 38-26 Phil l i p s (1925) Not given 50.5 55-44.5 36.5 38-33.5 Arnold (1895) 9 52.1 54.5-49.0 36.7 37.1-36.1 Baine (1892) 10 50.7 Not given 36.9 Not given This study 143 50.7 56.9-46.7 36.6 38.8-33.3 (s - 1.62) (s * 1.23) Both of Bent's values seem low, and his data include one egg measuring 40 by 26 mm. There are two references to "robin-sized" runt eggs of Bufflehead (Cook, 1930; Henderson, 1936). The other mean values agree well with one another. Variation of egg size within individual clutches was only determined i n a few cases, and s t a t i s t i c a l treatment such as that given for the Velvet Scoter by Koskimies (1957b) i s not possible here. By plotting such data as were obtained i n this study (Figure 9), i t may be seen that lengths are relatively uniform, a l l of the means for single clutches lying within one "group standard deviation" of the group mean. The widths show considerable scatter between clutches but relatively l i t t l e within a clutch. Since the width of eggs l a i d by a given bird w i l l i n part be governed by the diameter of the oviduct, the small scatter within most clutches supports the contention that laying by more than one bird i n a single nest is rare. The data for nest # 10 show that this does occur sometimes; the f i r s t eight eggs i n 1958 and the four i n 1959 were thought, on other grounds, to have been l a i d by the same bird, while the other groups of two and three eggs were probably l a i d by two other birds. Furthermore, the two unhatched eggs i n nest # 73 had measurements suggesting that t h i s , the largest clutch on record (16 eggs), to follow p.45 Figure 9, Measurements of Eggs. Nest No. No. Eggs #4-11 #7- 11 195* •#10-8 2 3 J 4 1959 « 3 - 8 #34-7 #4 6-13 #56-1' #65-6 #73-2 #90-7 TOTAL ^ S A M P L E Length (mm.) 46 47 49 49 5 0 51 52 53 54 M( * * -j f - t e — w ) > » X M—K tt M—m tt 5fc -M *-5.5 #7—11 * 1 0 - » ] 2 > 19S« 4 1959 #4 6-73 #56-11 # « - 6 #73-2 #90-7 TOTAL SAMPIE 1 43 Width (mm.) 3 3 3 4 • 3 5 l 3t6 3 7 • 3 • 39 »4= Legend: individual measurements marked x j clutches of under 5 eggs omitted, except for nests # 10 and # 73. 4 0 Mean ' 1 range ' M-s M*s * - 46 -was actually the result of two birds laying i n the same nest. On the other hand, the 13-egg clutch from nest # 46 has one of the smallest standard deviations l i s t e d , so that a clutch of 16 may be within the limits of possibility for one bird. Data on weights of eggs are lacking, and adequate descriptions of shapes and colours of eggs are given by Bent (1925) and other standard reference works0 Incubation. Various authors i n recent years have levelled criticism at careless use of the term "incubation period". Although the start of incubation could never be established very precisely i n this study, "incubation period" as used here means the length of time from the laying of the last egg u n t i l the hatching of the clutch, a l l eggs having hatched out overnight i n nearly a l l cases observed. This follows the definitions of Nice (1954) and Davis (1955) i n using the viewpoint of the egg, rather than that of the parent(s). The duration of incubation was not previously known for Bufflehead. Ph i l l i p s (1925) reported that wild-taken eggs had been hatched under a hen i n 21-22 days, but this i s very much out of line with a l l values for related species. In the present study the modal value was 30 days, but the values 29-31 days f i t the observed data equally well and better allow for the probable v a r i a b i l i t y between individual cases. The data are presented graphically (Figure 10) for a l l clutches where the incubation time was known to within a range of five days, but the picture i s unchanged by the addition of a further 16 clutches for which only upper or lower limits to the incubation time were ascertained. I t might be noted that the two clutches with the longest incubation times (# 58 and # 59) were both early, and at the time incubation was begun temperatures were below freezing nearly every night, daytime maxima frequently being only of the order of 10-12°C, which is well to follow p.46. Figure 10. Incubation Periods obtained in Ihis Study. <D • . • H CO •» o> fl rH 5 5 ^ -p © (0 -P <D © o -p g >> • H a5 © u co <D to fl ® o 0 © § (0 § • H -P OS >» 05 fl O • H • P 03 CT> r£> CM o fl © 3 99 OS T3 I EH fl O • r t - P «3 . Q O 1 1 1 *d (to •I en i n en <5 m ui S •> en en n I— • <"> • » <^  <N r» <"> * * % *v 5 5 ffi S % % * ^ - 47 -below the usual level for that time of year there. In nest # 23 i n 1958 even the assumption of the very short incubation period shown requires that incubation have begun before the last two eggs were l a i d , and the two eggs that failed to hatch (the last laid?) contained nearly fully-formed young. In two other nests the last egg may have been l a i d after incubation was commenced; i n one case the young bird i n one egg died after the shell had been "pipped", although this may have been due to weakness rather than to a retarded schedule. Another nest contained an apparently normal young alive after the remainder of the brood had been gone for several hours; this bird was s t i l l alive and vigorous on the following morning and was then released. Although Nice (1954) has shown that incubation periods given i n the literature must be viewed with suspicion, i t may be worth l i s t i n g the values given for various other waterfowl for comparison with that found here for the Bufflehead. The value found here i s i n line with those found for other hole-nesting ducks, i f the 20-day values for Common Goldeneye are discarded. Ineubation times for these species certainly appear to average longer than those for ground-nesting species, as already noted by Nice (1954). - 48 -Table 16. Incubation Periods i n Other Waterfowl. Species Bent Values (in days) Given by Phi l l i p s Pough Jackson (Mergini) Common Merganser Bed-breasted Merganser Hooded Merganser Common Goldeneye Barrow's Goldeneye Harlequin Duck Oldsquaw White-winged Scoter Common Scoter 28 26-28 31 20 (?) 28 24 24 28,32 26-28,29 31 20(?),28,30 28,31 31(?) 24,26 27-28 32-33 34 27 31 26 — 31-34 24 24 27 30 (Aythyini) Eing-necked Duck Bedhead and Canvasbaek Lesser Scaup 26-28 24 22-23 (Anatini) Mallard. Gadwall, P i n t a i l , Shoveler, Bluewinged Teal, A l l i n range and American Widgeon 21-24 (Cairinini) Wood Duck 30 (Oxyurini) Buddy Duck 21 (The values for the Mergini are taken from Bent (1925), Ph i l l i p s (1925), and Pough (1951), except for Miss Jackson's value for Barrow's Goldeneye; the other values are quoted from Hochbaum (1944), Mendall (1958), and Nice (1954)). Attentiveness« No direct study of attentiveness was made. However, tabulation of times of day at which females were found on or off the nests gives some information on this point. Comparison of v i s i t s during the f i r s t and second halves of the incubation period showed no significant differences, so a l l data are pooled below. 49 -Table 17. Attentiveness of Females During Incubation. Time of Vi s i t s Total No. V i s i t s No. V i s i t s Female (PST) Made to Nest Found on Nest (jt) 0500-0600 3 3 (lOO) 0600-0800 26 24 (92) 0800-1000 85 75 (88) 1000-1200 48 40 (83) 1200-1400 42 33 (79) 1400-1600 49 39 (80) 1600-1800 35 21 (60) 1800-2000 32 20 (63) 2000-2200 4 1 (25) (No v i s i t s were made prior to 0500, and very few were made prior to 0800 during the f i r s t half of incubation to avoid desertion). Attentiveness thus seems to decline during the day, but does not f a l l below 80 percent u n t i l after 1600. A tendency to leave the nest later i n the day seems probable, and i n one nest kept under quite close observation (# 33) the female was only found on the nest three times i n ten v i s i t s after 1800. these three a l l being i n the f i r s t week of incubation when night temperatures were cool. Leaving the nest i n the evening was also noted i n the Ring-necked Duck (Mendall, 1958) and i n the Barrow's Goldeneye (Jackson, i n preparation), but no evidence for a morning break i n incubation such as found for the Ring-neeked Duck was noted i n Bufflehead. Further evidence that such does not usually occur i s given by the fact that females trapped on the nest for banding i n the forenoon usually defecated profusely, much more so than might be expected i f the birds had recently l e f t the nest. Birds breeding for the f i r s t time may be more attentive than older birds. Two year-olds were found on their nests 14 times on 16 v i s i t s or 88 percent, whereas older birds were found bn the nest 103 times on 134 v i s i t s or 77 percent; this difference i s not of s t a t i s t i c a l significance at present. Hatching, and Brooding of Young i n the Nest. The hatching process - 50 -in Bufflehead is comparable to those described for other species of ducks. Tapping and "peep"-ing noises may be heard inside the eggs on the second afternoon before hatching takes place, 24 hours after this a l l or most of the young hare "pipped" the shell, but the membranes are usually s t i l l intact. Hatching seems to take place during the night. Seven cases in which more than one v i s i t was made to a nest during the hatching process are compatible with the schedule proposed, but two exceptions were noted. In one case one of five eggs had "pipped" as early as the second afternoon before hatching, although none of the membranes were yet broken at 1730 on the next day. In another nest four of nine eggs were noted as having tapping and "peep"-ing sounds one afternoon, none having pipped, yet the young had hatched and left the nest by the morning of the second day following. Tapping sounds and "pipping" of the shell were noted mainly in the afternoon, most probably because the females are less attentive at that time of day. During the last three days before hatching, females were found on nests on 40 visits and off on 13 visits (75 percent attentiveness). Eggs were removed from under sitting birds and found to have tapping sounds inside on two occasions, while tapping or"pipping" was noted on eight of 13 visits when the females were absent. The young remain in the nest for 24 to 36 hours after hatching and are brooded assiduously by the female during this period. The females were found on the nests in 25, or 83 percent, of 30 visits. Leaving the Nest loung Bufflehead usually leave the nest during the forenoon. In eight nests young were seen on one day and were gone before noon on the next day, while in seven cases, including four of those given above, young left between the f i r s t and second visits on a given day, five of these departures certainly being before noon. The actual departure was only - 51 -watched once, but since this process has not previously been described for Bufflehead and presents some interesting features a detailed account is warranted. On June 19, 1958, at Watson Lake, the observer was concealed 25 metres from the nest (# 33), which is situated 0.8 metres from the ground in a dead Aspen 31 metres from the shore. The observer was almost directly in front of the entrance, which faced away from the shore. The edges of the grove were about 15 metres away on either side, and numerous shrubs 0.3-1.0 metres in height were present in the shaded area; birds on the ground could not be seen from the"blind". Observation was commenced at 0815 PST, the female then being on the nest with the young around her. The data following are transcribed from f i e l d notes, without change except by giving in f u l l words then abbreviated. 0829 - female appeared in entrance, then backed in until only head showed; 0832 — female disappeared inside; 0850 — female appeared in entrance, looked around, backed in until only head showed, moved into entrance again; 0851 - female dropped to the ground (out of view of the observer); 0908 — female flew in from the west side of the grove (from the lake?) and entered nest; 0915-0916 - female left nest exactly as before; 0927 — female flew back and entered nest exactly as before; 0929 — female looked out, then dropped to the ground without f i r s t backing in again; 0932 - female flew to the lake from the ground near the nest; 0933 - one young appeared in the entrance and dropped to the ground; 0934 - female flew back and entered nest as before; 0935-0936 - female left the nest as at f i r s t , looking around and backing in before dropping to the ground; 0937 — two young left the nest in rapid succession; then a pause; 0938 - two more young left in quick succession; 0941 - one young left the nest; then a pause; 0942 - one young left the nest; 0945 - one young left the nest; 0957 - observer left "blind" and hastened to shore; 1000 - female was leading seven young out on the lake. At 1005 the nest was seen to contain one unhatched egg (later found to be undeveloped), but a search for the eighth young was unsuccessful. Whether - 52 -i t was the f i r s t to leave the nest, three minutes before the female dropped to the ground, or the last , three minutes after that preceding i t , i s not known, but the latter seems more l i k e l y . In any case one young was lost during the trek of less than 40 metres to the water. This sequence of departure from the nest differs from those described for Goldeneye (summarized by Myres, 1957) i n that the period of departure was here prolonged over 12 minutes. Most references to Goldeneye leaving the nest also refer to an auditory signal, but Jackson (verbal) noted one departure i n which no clearly audible signal was given, and this one example for Bufflehead may also be atypical i n this respect. One other observation suggests that departure from the nest may be s t i l l more prolonged on occasion. On June 8, 1958, a nest (# 23) contained at least two young, one of them s t i l l damp, and two eggs, the original clutch having numbered ten, while a female was seen on the lake nearby accompanied by five very small young. This was at 1345, but at 1925 on the same day a female with seven small young was seen i n the same area; at 1305 en the following day only the two unhatched eggs remained i n the nest, and a female with eight young was seen nearby shortly afterwards. No other broods appeared upon the lake for two days, so i t seemed probable that a l l these young came from the same nest, having departed independently and been gathered into the only brood then present. Success of Nests and Eggs. About 80 percent of occupied nests found in this study were successful, i f a l l areas are considered together, in both 1958 and 1959. Marked differences were noted between lakes or areas when these smaller samples were considered separately. - 53 -Table 18. Nest Success. lear Total Nestings for Which Fate Known No. Failed No. Successful * 1958 1959 36 70 8 14 28* (78) 56' (80) * includes one joint clutch incubated by Barrow's Goldeneye (Erskine, 1959a), and two nests for which clutch-size i s unknown, one of which was found after the young had l e f t . * includes one joint clutch incubated by Buffle-head (Erskine, I960), and four nests found after hatching for which clutch-size i s unknown, two unlisted nests were not revisited and their nesting success i s unknown. In successful nests 92-93 percent of eggs l a i d were hatched successfully. Egg success varied between areas, although less markedly than did nest success. Table 19. Egg Success. lear Successful Nests for Which Clutch Known* Eggs Laid Eggs Hatched ($) Unhatehed Eggs 1958 1959 25 51 211 435 194 (92) 407 (93) 17 28 * successful joint clutches omitted. These figures might be further broken down into original nests and renestings, thus: i n 1958 original nests had 180 of 197 eggs hatched and renests had a l l of 14 hatched, while i n 1959 the figures were 388 of 415 eggs hatched i n original nests, 19 of 20 eggs in renests. The percentages are thus 91 and 93 percent for original nests, 100 and 95 percent for renests, these differences not being significant. The status of eggs f a i l i n g to hatch varied, the majority being either i n f e r t i l e (or dying before detectable development had taken place) or else dying at an advanced stage of incubation. These data follow. Table 20. Condition of Eggs That Failed to Hatch. (Successful Nests Only) Content When Opened, or Other Fate 1958 1959 Dead embryo, one-half term or older 5 8 Dead embryo, less than one-half terra — 2 Infertile or undeveloped 6 13 Rotten (cracked early i n incubation) — 3 Broken i n opening nest during laying 1 2 Other 5* -Total 17 28 * Includes two which disappeared after one broken by Flying Squirrel (see pp. 40—41), one not collected for examination, one buried i n straw below floor of nest and thus not subjected to incubation. In unsuccessful nests the cause of failure also varied. Most were deserted before incubation was begun, either with complete or partial clutch. The data are tabulated below. Table 21. History of Unsuccessful Nests. Fate of Nesting 1958 Nests Eggs 1959 Nests Eggs Incubated 60 days (did not hatch; Deserted: After 2-10 days incubation Before incubation ( f u l l clutch?) During laying (partial clutch) Nest used by Flying Squirrel + Nest used by Tree Swallow + Female k i l l e d on nest Eggs destroyed g 13,11,7 13 4,4,3,1 4 7(1 broken) 5,1 1 3 3 1 11 13,8,14* 6,2,2 4 3 1 4 1 1 2 Total 60 13 81 * Multiple clutch, (see p.45). + Eggs unharmed. - 55 -The causes of desertion were not known i n a l l cases; the ultimate desertion of the multiple clutch i n 1958 was probably due to the capture of the (third?) female for banding, and two of the clutches deserted early i n incubation i n 1959 were probably also from this cause. In one nest deserted with a partial clutch three of the four eggs bore small punctures (as of claw-marks), and the presence of a fresh feather indicated that a Flicker had been i n the nest, Down plastered on the entrance suggested that the Bufflehead had been repelled upon several attempts to enter the nest, and the bird that had occupied that nest i n the two previous years renested i n another site 67 metres away. The predation of the single egg was by a bird (Starling?), but the other predations were more l i k e l y by mammals. Other studies have given comparable data upon nest and egg success. Jackson provided data on six nests; two had a l l eggs hatched, two clutches were collected by oologists or others, one was deserted after 5-6 days of incubation, and two of the 10 fresh eggs i n another deserted nest bore claw-mark punctures* Myres found a l l eggs hatched i n three nests, while the female deserted after banding i n a fourth. Burns (in l i t t . , 1959) found a l l eggs hatched i n eight nests, the nest-tree having fall e n down during the laying period i n a ninth case; one of the successful nests was -thought to have been a renest of the one that f e l l , Naylor (in l i t t . . 1959) reported on nine nestings, mostly i n boxes; a l l eggs hatched i n four cases, while three nests l e f t one, -three, and four eggs with dead embryos. One nest was deserted early i n incubation, and the female was found dead but apparently uninjured upon the other nest* Furniss (1938) reported one nest i n which a l l eggs hatched. Most of the references i n the literature to Bufflehead nests were by oologists; a l l of those nests were unsuccessful, since the eggs were collected. - 56 -Success i n leaving the nest, including nest-boxes. Failure to leave the nest does not appear to be a significant mortality factor for young Bufflehead. In the 84 successful nestings l i s t e d earlier (Table 18) young were found dead i n the nest i n only two cases, while a l i v e young which might have died i f l e f t there was found alone i n another nest. Other reports are similar. Myres i n 1957 found one dead young i n a nest where a l l eggs had hatched. Naylor (in l i t t . . 1959) found that Bufflehead nesting i n deep boxes (60 cm.) erected for Wood Bucks lost nearly a l l young through failure to leave the nest; a l l young (6 and 9) hatched i n two nests died there, while only one of four young hatched i n a third nest succeeded i n leaving the box. Naylor also suggested that the female found dead i n one box had f a i l e d to climb out. Burns (in l i t t . . 1959) quoted the experience of a nearby worker on the need for a "ladder" to help downy Bufflehead climb out of nest-boxes made from nail-kegs, while Naylor solved the same problem by making boxes with the bottom slanting up to within 10 cm. of the entrance. Four nest-boxes set up by the Br i t i s h Columbia Game Commission were used by Bufflehead i n 1959, two of these having one or more young hatched. A l l five hatched young l e f t one box successfully, this box being made of rough Douglas F i r boards, while one of three hatched i n another box made of smooth cedar boards f a i l e d to leave the nest. A young Barrow's Goldeneye hatched from the joint clutch i n the cedar box also failed to leave the nest (Erskine, I960). The depth of both of these boxes was about 20 cm., but the depth seems less c r i t i c a l (of. Table 2) than a rough surface i n enabling young to leave a nest-site. Recapitulation. Egg-laying by Bufflehead commences within ten days of the "breakup" on the breeding lakes, although the birds may have been present upon open water elsewhere i n the area for as much as two weeks more. - 57 -Eggs are l a i d at intervals of about 1.5 days, but this may approach one day towards the completion of a large clutch. Laying probably occurs most often i n the forenoon. The mean clutch-size i s approximately 8.6 eggs for original nests, and about two eggs less for renests. Clutch-size appears to be less for clutches started later i n the season. Measurements of eggs, particularly the width, may give indications of whether more than one female has l a i d i n a given nest. The incubation period i s most commonly 29 to 31 days, and i s thus similar to most other hole-nesting ducks of the tribe Mergini. The eggs are brooded quite assiduously, attentiveness averaging over 80 percent except i n the early evening. The hatching process is similar to that of most other waterfowl, but the one departure from the nest that was witnessed deviated markedly from those described for the closely related Goldeneyes. Nest success amounted to about 80 percent, while about 92 percent of eggs i n successful nests hatched. Desertion during laying, from unexplained causes, and the death of embryos either very early or after the middle of the incubation period, were the main causes of failure, predation being of minor importance. Failure to leave the nest is not a significant mortality factor under normal circumstances, but may have importance i n management programs i f nest-boxes made of other than rough lumber or of great depth are employed. Once the Bufflehead has located a nest-site i t w i l l be able i n a majority of cases to succeed i n hatching young. Regulation of numbers most probably takes place at other stages i n the breeding cycle, since there appears to be l i t t l e during the "nest stages" to produce this effect. - 58 -FLIGHTLESS YOUNG, AND MORTALITY FACTORS ACTING UPON THEM Factors acting upon the survival of flightless young Bufflehead may be expected to vary with l o c a l i t y , some factors being more important i n one part of the range, and others elsewhere. The most important of such factors are, not necessarily i n this order, food supply, competition, predation, parasites, drowning, and climate, and these w i l l be discussed i n turn before survival as a whole is taken up. Food Supply. Lack (1954) has considered food as the ultimate regulating factor for populations of most nidicolous birds. In that study, however, waterfowl were hardly mentioned, and the nidifugous habit of this group i n combination with the independent feeding exhibited by their young is incompatible with a limitation imposed by the a b i l i t y of the parent(s) to provide food. Even very young ducks appear to secure a l l of their own food, and the parent seems to exercise l i t t l e direct control of where such feeding takes place once the brood "home range" i s established. Few studies have been made of the foods eaten by diving ducks, and only those of Cottam (1939) and Munro (1942) contribute significantly to the picture for Bufflehead* Munro (1942) quoted figures for 20 downy young and 13 adults taken on the breeding grounds during June, July, and August, while Cottam (1939) had examined three young and 16 adults taken at the same time of year. In addition, Dixon (1926) reported the stomach contents of four young taken i n June. A l l sources agree that insects provide the major amount of the food of Bufflehead, Crustacea and Molluscs also contributing significant proportions, especially i n birds wintering on salt water. In a comparison with other species, Cottam (1939) showed that the Goldeneyes also depended - 59 -upon insects to a major degree, but Molluscs form a more important source of food for these species, particularly i n winter, than for Bufflehead. Oldsquaw, Harlequin Buck, and Steller's Eider had fed mainly upon Crustacea, while the Scoters and the larger Eiders ate predominantly Molluscs. However, these results must be viewed with caution, since the majority of the birds examined were taken during the f a l l hunting season. Insects are probably of major importance for the young of a l l species frequenting fresh water areas for breeding. The foods of young Bufflehead are shown below. Table 22, Foods of Downy Young of Bufflehead (in percent by volume). Percentages for Various Areas Materials Found i n B.C. (Munro ) Alta. (Cottam) Cal i f . (Davis) Birds Examined 20 Birds 3 Birds 4 Birds Annelida mm mm 1.25 Arthropoda Crustacea Amphipoda 17.85 - — Insecta Odonata 26.05 58.66 43.75 Coleoptera 32.10 24.66 -Hemiptera Notonectidae — 1.25 Corixidae 5.95 *? 43.75 Trichoptera * 12.67 -Misc. Insects 9.50 2.34 — Misc. Plant Material 8.55 1.67 — Sand — — 10.00 * Included under miscellaneous insects. Odonata are the only group having major importance i n a l l three samples, but the number of birds examined i s too small for the absence of certain groups i n individual samples to be significant. However, even i n the B.C. sample Amphipods were only taken locally, forming a major part of the food on a few - 60 -lakes but not appearing i n birds taken elsewhere. It was not possible to construct a similar table for the foods of adults on the breeding grounds. Munro*s table for adults did not break down the total percentage of insects into i t s component groups, while Cottam combined a l l samples regardless of place or month of origin. Insects clearly predominated, with Odonata of major importance i n Munro's sample and Coleoptera i n Cottam's, while Amphipods were important on a local scale. The most significant differences between the foods of adult and young Bufflehead are the higher proportions of Molluscs and of plant food taken by the older birds. Plant food was mostly seeds of Potamogeton species, which form the major food of many ducks of the tribes Anatini and Aythyini. The Bufflehead and the Goldeneyes have the most vegetarian diet of any ducks i n the Mergini. In the present study investigations into the food of Bufflehead were concerned only with sampling the invertebrates present i n various lakes. Three lakes were selected for comparison, of which Watson and P h i l i l l o o Lakes had relatively large populations of breeding Bufflehead while Lilypad Lake had no Bufflehead and few ducks of any species. After the invertebrate sampling was completed i t was learned (J.A. Munro, verbal) that twenty years previously Lilypad Lake had supported f a i r populations of ducks, including Bufflehead, while Watson Lake was barren; P h i l i l l o o Lake was not examined at that time. This at once cast suspicion upon the hypothesis that present differences between the food supplies of these lakes had any bearing upon their populations of Bufflehead, and i n fact the results of the study were negative. The three lakes are comparable i n size, but Watson is more alkaline and supports less aquatic vegetation than the other two. Probably a l l correspond - 61 -to the "Podiceps-type" (cf.pp.33-34) of Palmgren (1936), although parts of Lilypad Lake may approach the "Nyroca-type". Aquatic invertebrates were secured by plankton and bottom fauna sampling (see Appendix), but the differences between the lakes were minor. Molluscs were scarce i n Watson Lake, perhaps because of i t s al k a l i n i t y , while Amphipods were notably abundant i n the dense aquatic vegetation at one end of Lilypad Lake, The vastly larger volumes of Cladocera secured at P h i l i l l o o and Watson Lakes, especially the latter, may be a simple consequence of the later dates of sampling at these l o c a l i t i e s . The hypothesis of food as a limiting factor on abundance of Bufflehead is probably better discarded as far as these lakes are concerned, but i t may well have some bearing elsewhere. The B r i t i s h Columbia Game Commission has carried out a program of evaluating f i s h food resources i n various lakes i n this province (Northcote and Larkin, 1956). Total dissolved solids was the best indirect indication of abundant f i s h food, but even this property was not satisfactory on a local scale. Very few of the lakes then surveyed were of any significance as breeding areas for waterfowl. I t was noted i n this study that lakes where fishing was popular, such as Lac l a Hache and Bridge Lake, had few ducks breeding. Whether the lakes are of themselves unsuitable or whether the disturbance occasioned by the fishing was chiefly responsible for this i s unknown; both are presumably of influence. Furthermore, lakes favoured by ducks held few f i s h ; considerable numbers of Shiners were noted i n P h i l i l l o o Lake, but Watson Lake was apparently devoid of f i s h . Competition between breeding waterfowl and f i s h for a food supply might lead to the elimination of the former, while waterfowl can and do breed successfully around water bodies apparently too alkaline to be tolerated by f i s h . This, however, i s largely speculation. - 62 -Competition. Competition for nest-sites between Bufflehead and other speeies has already been discussed (pp, 22-23). Rivalries exhibited between species at that period are carried over into the period of flightless young by Barrow's Goldeneye, Territorial conflicts within these species occur, but must be rare i n Bufflehead, Aggressive behaviour by a female Bufflehead accompanying a brood was observed once only; the "threat" posture and attack were directed at an adult male and two female (or yearling) Bufflehead nearby. No conflicts between Bufflehead females escorting broods were witnessed, nor were such seen with Goldeneye, but such conflicts have been noted on several occasions i n the latter species by Jackson, Myres, and McLaren (verbal). Two detailed observations were made of attack on Bufflehead broods or their escorting females by females of Barrow's Goldeneye. In one case the attack was directed solely at the adult Bufflehead, which was pursued, by diving or rushing over the surface, for at least five minutes, but both females returned to their broods afterwards. In the other case the attack was directed both at the young and at the female, by turns. The young Bufflehead here were less than three days old, and this brood was not surely seen again, while the Goldeneye was not apparently accompanying a brood. It i s quite probable that the Goldeneye actually k i l l e d the young Bufflehead, or so scattered them that they subsequently died of exposure, in this case. Another case has been reported (Sugden, verbal) of an unprovoked and fatal attack by a female Barrow's Goldeneye upon a young Scaup, while Jackson twice saw Goldeneyes k i l l their own young when the latter had been colour-marked for identification. Briefer chases of Bufflehead by Goldeneye were witnessed upon other occasions, and such chases are probably of frequent occurrence while home ranges of the broods are being delimited. - 63 -Predation, No sure evidence of predation, as distinct from "assault" (described above), upon Bufflehead broods has been given i n the past, and none was obtained i n the present study. Large f i s h and turtles are lacking from most of the breeding lakes i n interior B r i t i s h Columbia, although they may occur i n breeding areas elsewhere. Horned Owls (Bubo virginianus) and Red—tailed Hawks (Buteo jamaicensis) are common in the study area, while the three Accipter species, Bald Eagles (Haliaetus leucocephalus), and Peregrine Falcons (Falco peregrinus) a l l occur there i n small numbers. Mink (Mustela vison) and Short-tailed Weasel (M, erminea), Red Fox (Yulpes fulva), and Coyote (Canis latrans) are also present i n the area. Few avian or mammalian predators are known to attack broods upon open water i n this region, so losses from predation most probably occur when broods are out on shore at night. Although broods have never been actually found on land except when travelling, they are frequently startled out of reedbeds along the shore i n the early morning; i t seems l i k e l y that such broods had spent the night either on shore or very close to i t . In 1958 losses of young were such that there was no need t© invoke predation as a causal factor; the much more severe mortality observed or assumed i n 1959 i s also well correlated with other potential mortality factors. When other factors lead to considerable numbers of dead or dying young i n an area some species may well u t i l i z e them as food, but predation upon normal young Bufflehead seems l i k e l y to be slight. Jackson (verbal) remarked on visible evidence of predation by raptorial birds upon adult Bufflehead i n the spring of 1954, when the stripped carcasses were l e f t upon the shore, and commented that young taken i n the breeding season would probably be carried to the nest or l a i r of the predator. Thus, the failure to locate remains of predated young i s not surprising. - 64 -Parasites. No studies of parasitism in Bufflehead on their breeding grounds have been carried out. Two young birds secured in this study were found to carry moderate parasite loads (13 cestodes and 6 acanthocephalans in one, a number of Spirurid nematodes and 3 cestodes in the other), but no signs of parasite damage to the hosts were noted, and i t was not felt that these parasites would have been a cause for death in the near future (D. & G. Gibson, verbal). These two young birds were estimated at 10 to 14 days old, so i t seems unlikely that most parasites would develop soon enough to affect mortality of small young. On the other hand, heavy parasite loads might render birds more susceptible to predation or to adverse weather conditions at a later stage, before flight was attained (see e.g. Jennings and Soulsby, 1958). Drowning, Young ducks of many species have been known to drown when entangled in aquatic vegetation (Hochbaum, 1944). Munro (1942) made no mention of this in Bufflehead, and none was actually observed in this study, Bufflehead broods usually frequenting open water. However, one brood in 1958 took up a territory at the shallow east end of Watson Lake where thick mats of algae covered the surface for upwards of f i f t y yards out from shore, and in the course of one week this brood was reduced from six young to two. Drowning under algal mats or in submerged aquatic vegetation was thought a possible cause. Climatic Environment. Climate acts upon the breeding of Bufflehead both directly and indirectly. Generalizations may be based upon the influences of climate, although laboratory studies on the species (cf. Kendeigh, 1934) are also desirable wherever possible. Climatic data from the study area itself are preferable, in order to reduce the effect of local variations, and temperature and precipitation data secured at Watson Lake are given in Table 23 and Pigure 11. - 65 -Table 23, Weekly Mean Temperatures and Total Precipitations at Watson Lake. Year Week Mean Temp.( C) Max. Min. Rain Days Intensity T & P Combined June 2-8 24 9 2 2 " 9-15 22 9 2 3 *» 16-22 27 9 0 0 " 23-29 21 9 5 6 Jn.30-J1.6 24 11 3 3 July 7-13 28 12 2 2 " 14-20 30 10 0 0 » 21-27 24 9 0 0 J1.28-Au.3 24 9 2 3 Aug. 4-10 23 9 3 3 June 2-8 20 4 2 4 " 9-15 20 4 4 6 " 16-22 23 8 2 2 " 23-29 18 6 7 12 Jn.30-J1.6 20 7 4 6 July 7-13 25 8 0 0 " 14-20 26 8 1 2 " 21-27 27 9 1 1 J1.28-Au.3 26 8 1 2 Aug. 4—10 22 6 0 0 1958 1959 Min.7 ; no night pptn. Min.6°j 1 night pptn. (9 ). Min.6°} no night pptn. No low temp,; 2 or 3 night pptn, Min.10°; 1 night pptn, Min.10°j no night pptn. Min.6G; no night pptn. Min.7°; no night pptn, Min.6°; 1 night pptn, Min.6°; 1 night pptn. o Min.2 ; also 2 under 5°; 2 nt. pptn. Min.3°, also 2 under 5°; 3 nt, pptn. Min.2°; no night pptn. Min.4°; 4 nt. pptn. at 4—5° Min.5°; 3 nt. pptn. ( l at 5°). Min.5°; no night pptn. Min.5°; 1 nt. pptn. (140). Min.6°; no nt. pptn,; 1 max. 12°, Min.3°; 1 nt. pptn, at 3°. Min.l 0; no nt, pptn. (Intensity of precipitation was estimated on a scale of 4 points/dayj data for weeks June 23-29 and July 21-August 3, 1958, and July 28-August 3, 1959, are less accurate than remainder, due to absences during parts of these weeks). - 66 -To sum up Table 23, rain f e l l upon not more than seven nights i n the ten-week period considered, i n 1958, while not more than six "points" f e l l i n any-one week. Although rain f e l l frequently i n the week of June 23-29 no long periods of continuous rain occurred. No temperatures of below 7°C were recorded on rainy nights, and none of below 6° were noted during the whole period of small young. In 1959, temperatures f e l l to 6° or lower i n every week, while precipitation occurred on at least 14 nights, 10 of which had temperatures of below 6 . The weather up u n t i l July 6, with the exception of the week of June 16—22, was both cold and wet, rain f a l l i n g continuously for most of the day on several occasions. In particular, the very adverse weather from June 23 to July 6 came just at the time when the greatest numbers of small young were present upon the lakes (see Figure 11). Mortality of young i n 1959 was even greater i n other l o c a l i t i e s than at Watson Lake, and local weather conditions may also have been more severe elsewhere; for example, on June 24 a severe hailstorm was experienced i n the breeding area near Eiske Creek, 70 miles to the northwest. Only one young bird was found dead during this period, on July 7. Although this bird was estimated at only ten days old, the yolk stalk had already been completely resorbed (D. Gibson, verbal). Inability to feed effectively during the frequent and heavy rains of this period i s thought l i k e l y to have been a major mortality factor i n 1959. Before continuing to analyse the survival of young Bufflehead i t may be well to look at the organization of broods, since estimates of survival must be based upon counts of broods. Brood Home Range, and Mixing of Broods. Almost alone among diving ducks the Bufflehead and Goldeneyes maintain f a i r l y distinct brood home ranges, although data on other species of the Mergini are less well known. In the Velvet Scoter, Koskimies (1955) found loose ties between parents and small young, and to follow p.66 Figure 11. Hatching Dates and Survival of Young Bufflehead at Watson Lake, compared with Temperatures and Precipitation during the Same Period. 12 18 Juy>e 6 2 4 30 6 1 2 1 8 24 3 0 5 (a) Hatching Dates - each "x" represents one nest. (b) Percentage of Hatched Young surviving at Dates shown. (c) Precipitation - each bar represents one day's precipitation, the height of the bar indicating intensity on 4-pt. scale. (d) Temperatures -.weekly means of daily maxima and minima. - 67 -general lack of brood home range. Canvasback maintain their broods as units, but wander freely, while in Scaup (Aythya affinis) "rafts" of young escorted by one to several females are formed soon after the young appear (Hochbaum, 1944). Mendall (1958) indicated that the Ring-necked Duck exhibits close adherence to brood home range and maintains brood t i e s , but this species usually frequents a very different habitat from that occupied by Bufflehead, The brood home ranges of Bufflehead often coincide i n i t i a l l y with the territories occupied by the pairs concerned, but changes are frequent. This was easily seen at Watson Lake in 1958, when a high percentage of nest failure led to a relatively sparse brood population, with broods appearing at well-spaced intervals. The average distances moved by broods i n their home range during various age classes are tabulated below. Table 24, Mean Distance Between Furthest-Separated Observations of Individual Broods Within a Given Age Class, at Watson Lake, 1958, Age Classes (No. Broods Involved i n Parentheses) 1st Week 2nd Week 3rd Week 4th Week 5th Week 6th Week Distance moved (m.) Mean 730 (9) 360 (10) 260 (10) 240 (7) 220 (5) 230 (2) Range 100-1600 100-650 100-550 100-550 180-270 180-270 The brood home range was established usually by the second week, but the rapidity of establishment was probably influenced by the presence of other broods nearby. The size of the home range was also influenced by the number of young i n a brood, larger broods tending to have wider ranges. After the establishment of the home range broods remained i n these areas almost continuously un t i l the age of five to six weeks. The groups were sometimes recognizable for s t i l l longer periods, and at one small, isolated lake (Pete Kitchen Lake) the broods were followed through to flying in 1959. On larger lakes, the gradual loosening of brood ties with age, and the possibility of mistaking well-grown broods at a - 68 -distance for parties of moulting birds, usually prevented following broods beyond five weeks of age. Munro (1942) stated that broods were often abandoned by the female, but the dates cited as examples suggest that such broods were already well over half-grown. No evidence of loose ties between parents and broods of small young has been noted for Bufflehead, although individual young may stray away from the brood; such strays may join other broods, or else may perish. Mixing of broods with those of other species or with moulting adults likewise was not observed except with nearly full-grown young. One form of brood mixing within the species which takes place with some regularity i s "pirating" of broods. In 1958 only transfer of complete broods (leaving one female alone) was observed, but i n 1959 this was seen to be a quite flexible rule. In 1958, for example, one brood of 11 four day-old young, referred to as "A", was increased to 23 of two sizes on the day that 12 others l e f t a nearby nest; the latter brood ("B") was never seen on the lake independently, although a female was seen near nest "B", calling as i f to young. A week later brood "A" increased to 34 on the next count after the last appearance of another adjacent brood of 11 young. This particular female ("A") had apparently a much greater "parental drive" than most Bufflehead, since she was the only adult female to accompany her brood into the trap during banding operations i n both 1957 and 1958. In 1959 female "A" again increased her brood, from 6 to 13 to 18 to 24, although the sources of the increases were not surely known i n that year. In 1958 another brood of two young gained four smaller companions on the day after five young l e f t a nest half a mile away along the lake, this being the last nest known to hatch that year and no other brood appearing at that time. One brood in 1959 increased from 11 to 18 shortly after seven young hatched by female "B" (whose brood was also "pirated" in 1958) l e f t a nearby nest (and never appeared elsewhere), and to 20 shortly - 69 -after the last sighting of the three survivors of yet a third brood. S t i l l another brood increased from 2 to 4 to 5 to 6 to 8 to 11 to 13 on the same home range within two weeks, while another added three young after the hatching of four eggs i n another nest and four more young after being mingled with the five from another brood during drive-trapping operations on that lake. In 1958 brood survival was high, so i t was much easier then to surely account for a l l young during their various transfers than under the severe mortality in 1959. The above evidence i s circumstantial, but i t i s nonetheless a good indication of brood transfer; colour-marking of broods would be the logical step for further studies of a more precise nature. Jackson's observations upon brood territories i n Barrow's Goldeneye corroborate these on Bufflehead to a considerable extent, although much more conflict occurs between the adult females i n Goldeneye. Notes made on Barrow's Goldeneye in this study also suggest that few significant differences exist between the brood behaviour i n these two species. Survival of Young on the Lakes. The ideal procedure for determining survival of young ducks under natural conditions i s to follow individual broods from hatching through to f l i g h t . This i s possible only on small, isolated bodies of water, but i t was attempted wherever feasible in this study. Where numerous broods occur, eolour-Hmarking has been used by some workers to trace the permanence of individual broods. This introduces some unknown factors; mortality induced by the dying procedure per se has been found to be slight (Evans. 1951; and others) but i t i s not known whether the "maternal drive" of the parent i s affected by this alteration i n appearance of the young. Evidence that i t may be so affected, for young dyed after leaving the nest, has been given by the k i l l i n g of their own dyed young by two Barrow*s Goldeneyes (Jackson, verbal), and dyeing of young was not carried out i n this study. - 70 -The adherence of broods to "home ranges" makes possible inference of the identity of broods seen i n the same area repeatedly. The most generally used method for following survival of young waterfowl i s by averaging numerous counts of carefully aged broods. Although somewhat greater precision i s attainable with experience, i n general, -three classes of flightless young are distinguished: I - downy young; i n Bufflehead these are up to about two and one—half weeks of age, and appear not more than one-quarter the size of the adult; I I — young bearing both feathers and down; these are from 2\ to 5\ weeks of age, and range from one-third to three-quarters the size of the adult; I I I - fully-feathered young; these are from 5-jp to 7^ weeks old, and are only distinguished with d i f f i c u l t y at a distance from groups of moulting adults and yearlings. The age at which young w i l l take wing is not known with accuracy for many species. Hochbaum (1944) and Mendall (1958) provide most of the data available for waterfowl, seven to eight weeks for the Ring-necked and Ruddy Bucks being about the shortest periods given for diving ducks. The Bufflehead broods on Pete Kitchen Lake were followed through to f l i g h t i n 1959. Toung l e f t two of the nests on June 9, and the survivors took wing for distances of 40 to 70 metres, when alarmed, on August 1, a period of 53 days, i f i t is assumed that these were the f i r s t f l i g h t s made by any of these young. Variations i n ease of take-off and length of f l i g h t suggested that some of these young had probably made short flights previously. Several other broods elsewhere had reached a comparable stage of development in the period July 30-August 7 in 1959, while few broods can have appeared prior to June 9, so a period of 50 to 55 days seems the best available for the flightless stages of Bufflehead. The age at which young w i l l take wing of course varies with the circumstances, some of the earliest flights being made when the birds were approached by a - 71 -boat during banding operations, while others took off into a strong wind. The overall picture of survival of Bufflehead young prior to fl i g h t i n B r i t i s h Columbia i s comparable to that obtained for other species studied i n detail elsewhere. The data on Bufflehead summarized below were obtained i n part from this study, and in part from broods reported i n the B.C. Nest Records Scheme during the years 1953 to 1957, mostly i n the same general areas. Data for the Common Goldeneye and the Ring-necked Duck are quoted from Carter (1958) and Mendall (1958), respectively. Table 25. Progressive Decline i n Average Size of Waterfowl Broods (No. of Records i n Parentheses). Stage Number of Young/Brood i n Various Species Bufflehead Common Goldeneye Ring-necked Duck Eggs/clutch Hatched Class I Class II Class I I I 8.6(105) 8.0(76) 6.3(172) 5.3(139) 4.8(8) 9(15) 8.7(not given) 6.2(127) 4.7(58) 4.8(49) 9.0(423) 8.4(357) 7.0(488) 6.0(361) 5.2(141) Although the very small sample of Class I I I Bufflehead broods casts some doubt on the trend i n that species, the parallel decline i n the three species i s striking (see Figure 12). Mortality i s clearly heaviest upon the small young of Class I, as might perhaps be expected, while Mendall (1958) attributed the continued decline i n Class II broods to a gradual increase i n the wandering of the young at this stage. Mortality i n small Class I young may in part occur before broods reach the water. In the example given of young leaving the nest (pp. 50-5l) one young was lost on a t r i p of about 40 metres to the water, while two later nestings in the same tree also apparently lost one and two young before reaching the water. Although most broods were not followed sufficiently closely to distinguish between mortality during the t r i p to the water and that occurring after reaching the water, at least ten broods were considered to have reached water without mortality, a l l but one of these being from nests within ten metres of the water. It may also be pertinent to mention that Furniss (1935, 1938, and i n l i t t . , 1959) noted that he never saw a Bufflehead brood, near Prince Albert, Sask., i n the '30*3, numbering more than six . He attributed this to smaller clutch-size than i n most ducks, the only clutch he counted numbering five , but i t seems possible that mortality during the t r i p to the water might have been greater there. Figure 12. Progressive Decline in Average Size of Waterfowl Broods. - 73 -In view of the smaller size of Bufflehead young i t i s perhaps surprising that their survival i s no worse than that of the larger species. This might i n part be due to the fact that i n British Columbia the Bufflehead i s i n one of i t s most favoured breeding areas, whereas the other species are probably nearer the limits of their distribution i n the areas where the studies quoted were made. Also, the data for Bufflehead are i n a considerable part taken from 1958, when brood survival of a l l species was quite strikingly enhanced over the usual levels, due perhaps to continuing fine weather. Few brood counts are available for Bufflehead from areas outside of Br i t i s h Columbia, scattered single observations being of l i t t l e help. The averages for 34 Alaska broods i n 1956 and 1958 (Hansen, i n l i t t . . 1959). and 70 California broods, mostly i n 1954 (Naylor, i n l i t t . , 1959), weres Alaska - Class I 7.0 (28 counts), Class I I 5.7 (6 counts); California - Class I 7.0 (26 counts), Class II 6.5 (42 counts), Class I I I 4.5 (2 counts). These values are somewhat higher than those obtained i n this study, i n spite of being from areas nearer to the limits of the range of the species. This may be a consequence of the smaller samples, which thus represent differences between years as much as or more than differences between areas. In this study data were secured of brood survival at Watson Lake i n both 1958 and 1959, for young aged up to about five weeks, while data from P h i l i l l o o Lake i n the same years permit comparison between areas as well as between years. The figures for Watson Lake are tabulated below as numbers of young per brood, while they are converted to percentages of the number of eggs incubated i n Figure 13(a), for comparison with the data from P h i l i l l o o Lake. to follow p.73, Figure 13. Survival of Young Bufflehead at Watson and Phililloo Lakes, 1958 and 1959. ' (a) Percentage of number of eggs incubated surviving as young to various ages. »00 "to o >* t o •xs - P 3 O W si a •H .rl = > • H j> O U - P SO a <D O U © 60 40 0 5 Hatching Date Watson. 195« Phililtoo. 195» Watson, 1959 \ Phililloo, 1359 10 15 20 25 30 Age (days) (b) Mean brood sizes on specific dates. 10 o 9 o u • u 7 PH hO 6 O >> 5 of 4 . o 3 525 2 195% Wotson Phililloo 19 59 _j» Wation "I - x P h i l i l l a * ' 1 mm 15 25 5 15 2 5 1 5 2 5 June July June 1 5 July 25 - 74 -fable 26. Survival of Young Bufflehead at Watson Lake. Stage Mean No./Clutch or Brood i n 1958 1959 Eggss during incubation Young: leaving nest 5 days old 10 " " 15 " » 20 * " 25 n n 30 " " 9.3 8.6 7.2 7.0 7.0 7.3 7.0 7.2 (9) (9) (10) 7.9 7.6 5.4 5.4 4.6 3.7 3.5 4.5 (15) (16) (17) (17) (17) (16) (10) (10) (8) (7) (5) (15) (12) (Number of clutches or broods given i n parentheses; for broods this number was obtained by adding broods for which no nest was located to the total number of nests known to have succeeded; this was necessary since the "pirating" of broods reduced the actual number of brood units present). Both Table 26 and Figure 13(a) show that the mean values for survival at 30 days are out of line with the remainder of the figures. This is most probably due to the fact that broods containing large numbers of young are less easily mistaken for groups of moulting birds, while smaller broods may have been omitted for this reason at this stage. It is possible that larger broods may have enhanced survival, since some "combined" broods had very small losses i n 1958, but data to adequately prove this are lacking. The data for P h i l i l l o o Lake do not permit a detailed analysis, since few v i s i t s were made there during brood season. Mean sizes of broods observed there i n 1958 and 1959 are plotted i n Figure 13(b), with the figures for Watson Lake on the same dates shown for comparison. Approximate percentages of survival (plotted i n Figure 13(a)), were calculated from the highest count on a single day i n each year, both of these counts f a l l i n g about 20 days after the mean hatching date. For 1958 i t was necessary to make an assumption on the number of eggs incubated. Although only six nests were actually located at _ 75 -P h i l i l l o o Lake i n 1958, eleven broods were seen there on one day that year; ten nests were known to have been successful in 1959, so i t i s unlikely that the number of eggs incubated was less in 1958 than i n 1959, and the numbers were assumed to have been the same i n each year. Both graphs i n Figure 13 show clearly that mortalities of young Bufflehead at P h i l i l l o o Lake paralleled those at Watson Lake, but were rather more severe i n both years. One can only speculate about possible causes for the higher mortalities at P h i l i l l o o Lake. A much higher proportion of the shores there are wooded, so mortality due to young straying from the brood on the way to the water may be more important than at Watson Lake. Aquatic vegetation, both of submerged and emergent forms, is more abundant at P h i l i l l o o Lake as well, so deaths due to drowning might also be more frequent there. At Pete Kitchen Lake, broods were followed through to flying in 1959. The observations there are summarized i n Table 27. It was f e l t that the three nests located were a l l that were occupied at this isolated lake, so this sample, although small, may also serve for comparison with those from Watson and P h i l i l l o o Lakes i n 1959. Table 27. Survival of Young Bufflehead, Pete Kitchen Lake, 1959. Nest Clutch Hatched No. Young Surviving on Date Size Date (No.) June July August 10 20 30 10 20 30 10 # 59 9 June 8 (9) H — ? 5 5 — 5 - . ^1 2 v 10v 10x # 5 8 10 " 8 (8) H — 8 8 ^ / z 3 2 2 8 ) Z 7 K # 80 7 ca.Jn.26 (4) H—" 3 3 (H = hatching date; ) = broods combined). 53 percent of the number of eggs in nests # 58 and # 59 produced young which attain-ed f l i g h t on August 1. If one assumes that no further losses took place in the - 76 -later brood the percentage for a l l three nests i s 46 (12 young from 26 eggs). These figures compare well with that of approximately 40 percent for Watson Lake i n 1959 (see Figure 13(a)), but are clearly higher than those for P h i l i l l o o Lake i n that year. I t was very noticeable that poor survival of young was correlated with weather conditions that might be expected to be adverse for small young. This should not be taken as proving causation, but the facts are striking (see Figure 11), In 1958 hardly any adverse weather was experienced during the entire period of small young, and survival even at P h i l i l l o o Lake (see Figure 13(a)) was higher than those for any of the three lakes studied i n 1959. In 1959 the peak hatch came just before the spell of cold, wet weather from June 24 to July 6, and many broods were never found after they l e f t the nest. At Pete Kitchen Lake the f i r s t two nests (# 58 and # 59) hatched very early, and their young were two weeks old by the start of the really adverse weather. The six young lost to broods there at ages of less than three weeks were a l l missed during weeks with adverse weather, whereas "three older birds disappeared during fine weather, presumably from some other causes. Brood Schedules. The breeding schedules of Bufflehead become progressively later from the south to more northern parts of the breeding range, this probably being related to 1he dates of spring breakup. Altitudes of breeding areas f a l l from south to north, but climatic amelioration due to this effect i s not sufficient to counter-act the trend toward later dates i n more northern areas. Except for one record, quoted without details by Phillips (1925), of young i n British Columbia May 16, the earliest date for a brood of Bufflehead is of one i n California May 27, 1921 (Say, 1921). At the other extreme, the f i r s t of five broods seen at Fort Yukon, Alaska, in 1956, was not noted u n t i l July 10 (Hansen, i n l i t t . . 1959). Details for these and other areas are given in Table 28. Table 28, Bufflehead Brood Bates i n Various Regions. Area Lat. Elev. (m.) F i r s t Broods Bates for Small Class I's, Range Peak Sources Calif . 41 1600 Oregon 44 1550 Mont. 46 950 B.C. 52 900 54 750 56 670 Alta, 52 850 56 670 59 250 Yukon 62 600 Alaska 63 500 66 150 N.f.T. 68 60 May 27, 1921 ca.June 1, 1925 ca.June 1, 1954 June 11, 1940 June 8, 1958 June 23, 1945 June 17, 1946 June 19, 1938 ca.June 15, 1896 June 18, 1914 ca.June 20, 1918 ca.June 12, 1956 June 27, 1944 June 25, 1933 June 21, 1944 July 2, 1957 June 28, 1958 July 10, 1956 June 1-July 25 ca.July 20, 1946* ca.June 10, 1917* June 8-July 31 June 12-July 16 ca.June 20 ca.June 15-30 ca.June 25 ca.June 15-25 ca.June 27 ca.July 5, 1904 Bay, 1921 Dixon, 1926 Naylor, i n l i t t . . 1959 Evenden, 1947 Saunders, 1921 Munro, 1942 see Table 29 Munro, 1949 Cowan, 1939 Macoun, 1900 Horsbrugh,.1915 Taverner, 1919a Smith, in l i t t . , 1959 Soper, 1949 Soper, 1933 Rand, 1946 Hansen, i n l i t t . . 1959 Preble, 1908 Single dates i n the "range" column represent observations of single broods, not mean dates. - 78 -The variation i n peak hatching dates i s much less than that for " f i r s t broods", most broods appearing between June 15 and July 1, regardless of l o c a l i t y . New broods continue to appear during a period of nearly two months i n California, and for four to six weeks i n B r i t i s h Columbia and Alberta. The ranges i n date of new appearances further north are poorly documented, but several workers (e.g. Munro, 1949; Band, 1946; Soper, 1949) have remarked on the rapid buildup i n numbers of broods i n days following their f i r s t dates. The f i r s t nearly full-grown young i n California i n 1954 were noted on July 22, while a fledging period of 50-55 days (see p.70) suggests that most young would have attained f l i g h t there early i n August, the peak hatch having fallen around June 20. Other records of well-grown or flying young include August 6, 1925 i n central Alberta (Munro, 1929b), and August 2, 1945, August 7, 1944, August 8, 1936, and August 18, 1940, i n various parts of British Columbia (Munro, 1937, 1947, 1949} Johnstone, 1949). Young s t i l l unable to f l y were noted i n the Yukon August 24, 1948 (Drury, 1953). For comparison between years, data for 1951 to 1959 have been assembled for the Cariboo-Chileotin d i s t r i c t of British Columbia, One nest i s known to have hatched before June 18 i n 1952, an early year, while i n 1954, a late year, the only known hatching date i s June 28, Data for broods aged during banding in 1951, 1954, 1955, a l l late years, suggest that most hatchings i n those years occurred between June 20 and July 15. The figures for 1956 to 1959 are given in Table 29. - 79 -Table 29. Brood Dates for South-Central British Columbia. Dates i n Various Tears 1956 1957 1958 1959 Source Myres Myres This Study This Study Stage 1st brood on water June 27 June 12 * June 8 June 9 Peak of 1st appearances July 2 June 16 June 18 June 18 Last small Class I*s July 15 Not given July 31 July 18 1st flying young Aug. 17« Aug. 2» Aug. 17* July 30 Peak of 1st fl i g h t s Aug. 22« Aug. 6* Aug. 1 1 Aug. 7 * The f i r s t brood i n 1958 was seen June 8, but another was calculated to have l e f t the nest by June 2; that brood could have flown as early as July 22, and i t i s certain that many others did so before August 17. 1 These dates are calculated on a basis of 50 days to attain f l i g h t after the f i r s t and peak appearances of broods on the water. It has already been shown (Table 9) that both 1958 and 1959 had unusually early springs, and the brood seasons i n these years were clearly longer than that of 1956. The hatching span i n the study area i s thus roughly five weeks, on the average. In late years this span i s somewhat shorter, and in early years broods appear over longer periods. Near the northern limits of the breeding range, broods batched after the end of July would not have time to attain f l i g h t before the lakes freeze in late September (cf. Kendrew and Currie, 1955, pp. 76-77). However, i t i s possible that the longer days available for feeding in high latitudes might somewhat reduce the fledging period there. Data on growth of young Bufflehead and on freeze-up dates for specific breeding areas are inadequate to either prove or disprove this possibility. - 80 -Recapitulation. The foods of both downy young and of adult Bufflehead consist mainly of insects, with Crustacea of local importance. An unsuccessful attempt was made to correlate abundance of invertebrates i n three lakes with their Bufflehead populations. In the study area i t seems rather unlikely that Bufflehead abundance and breeding success are limited by lack of adequate food, other factors probably having more direct effects upon their numbers. Brood territories are occupied by Bufflehead females and their young, for periods of four weeks or more. Such territories do not appear to be defended against other Bufflehead, but conflicts with female Barrow's Golden-eye, which may sometimes k i l l young Bufflehead, do occur. Brood transfer i s not uncommon in Bufflehead, but may be favourable to survival of young i f the latter are concentrated around those females most attentive to them. Predation, parasites, and drowning may a l l lead to mortality i n young Bufflehead, but l i t t l e indication has been obtained that any of these factors are of great significance. Correlation of juvenile mortality with adverse weather i s very good on a local scale. Under combinations of low temperature and continuous rain, which may rest r i c t feeding time, mortality of small young may exceed 80 percent in a given area. Averages of progressive mortality i n young Bufflehead in British Columbia appear quite comparable with those obtained for other species by other workers, but large variations occur from year to year. Brood schedules also vary from year to year within a region, phenology mainly varying with weather, and near the northern limits of the distribution such variations might lead to a total failure of breeding, i f young had not time to attain f l i g h t before freeze-up. - 81 -DISCUSSION OF THE BREEDING DISTRIBUTION, MAINLY ON THE BASIS OF FACTORS ACTING DURING- THE BREEDING SEASON The known breeding range of Bufflehead should be outlined before one proceeds to a discussion of factors limiting distribution. The distribution is noticeably discontinuous, but Bent's description (1925) is reasonably accurate i f supplemented with more recent records. The distribution maps of Phi l l i p s (1925) and Kortright (1942) neither indicate the discontinuities nor emphasize the casual nature of some peripheral records. Actual records of breeding, compiled from the sources li s t e d below, are plotted in Figure 14, Table 30, Sources of Distributional Data, Alaska; Baird et a l (1884), Dice (1920), Friedmann (1935), Dixon (1938), Gabrielson and Lincoln (1959) ; Hansen (in l i t t , , 1959). Yukon; Baird et a l (1884), Cantwell (1898), Bishop (1900), Rand (1946), Godfrey (1951), Drury (1953), Soper (1954). N.ff.T.; Baird et a l (1884), Preble (1908). A l l records for Wood Buffalo Park are l i s t e d under Alberta. British Columbia; Brooks (1903, 1917), Munro (1918, 1927, 1930a,b. 1937, 1942, 1947, 1948, 1949), Taverner (1919b), Kelso (1926), Williams (1933), Cowan (1939, 1948), Stanwell-Fletcher (1943), Johnstone (1949); Edwards (in l i t t . . 1959), Ritcey (in l i t t . . 1959), Helsitt (verbal); B.C. Nest Records Scheme; this study. Alberta; Macoun (1900), Preble (1908), Stansell (1909b), Brooks and Cobb (1911), Horsbrugh (1915, 1918), Taverner (1919a, 1928), Farley (1922), Barnes (1923), Henderson (1924, 1927, 1936), Raine (1927), Munro (l929a,b), Randall (1933), Soper (1942, 1949, 1951a,b), Eaton (1948), Godfrey (1952); Smith (in l i t t . . 1959), Burns (in l i t t . . 1959). Saskatchewan; Raine (1892), Arnold (1895), Macoun (1900), Coubeaux (1900), Furniss (1935, 1938), Mowat (1947), Soper (1952), Hooper and Hooper (1954). Manitoba; Thompson (1890), Raine (1892), Macoun (1900), Godfrey (1953). Ontario: Forster (1772), Morden and Saunders (1882), Alberger (1890), Macoun and Macoun (1909), B a i l l i e and Harrington (1936); B a i l l i e (in l i t t . , 1959). Washington: Dawson (1897), Dawson and Bowles (1909). Oregon: Evenden (1947), Griffee (1958) (in l i t t . , 1959). to follow p.81. Figure 14. Breeding Distribution of Bufflehead, based largely upon the literature. Only records specifically mentioning nests, broods, or occurrence of breeding, at specific localities, are mapped. - 82 -California: Ray (1921), Dixon (1921, 1926), Labarthe (1922), Grinnell et a l (1930), Davis (1941); Naylor (in l i t t . . 1959). Idaho; Salter (1953) (in l i t t . . 1959), Steel et a l (1956), / s* Montana: Saunders (1921), Wyoming: Gary (1917), Rosche (1954). North Dakota: Barnes (1909), Wood (1923)? South Dakota: Stoudt (1949) (in l i t t . . 1959). Wisconsin: Cory (1909)* Iowa: Bent (1925)* Maine; Knight (1897)*, Brewster (1924), Palmer (1949)* (Listed i n chronological order within each geographic unit, except that a l l publications by one author dealing with that unit follow his f i r s t paper). Records marked * are compilations citing earlier records, which were not available i n the original publications. No striking changes i n range, comparable with the recent range extensions of the Gadwall (Hochbaum, 1955) and Ring-necked Duck (Mendall, 1958), are known for the Bufflehead i n historic times, with the possible exception of the disappearance of the alleged breeding population i n Maine, Certainly the nesting habits of Bufflehead are not such as to encourage pioneering i n unknown areas. Possible relationships between the breeding distribution and habitat, nesting biology, and survival of young, w i l l now be discussed in turn. Relationships of Breeding Distribution to Habitat Used. One of the basic problems here i s to show which phyto—ge©graphic regions provide acceptable nesting habitat, and this can best be attacked on the basis of plant ecology. Generally speaking, neither geological substrata nor soils , of themselves, show any striking correlations with Bufflehead distribution^ except on one point. The Bufflehead i s lacking from almost a l l regions of the Canadian Shield - 83 -(cf. Figures 14 and 15). In this recently glaciated area the scanty topsoil and immature drainage systems characteristic of such regions are accompanied by stunted vegetation usually unsuitable for nest-trees, and the Arctic and Alpine Tundra regions which show somewhat similar characteristics also lack Bufflehead, Climate acts upon the substratum and s o i l to produce conditions favourable for various plants, the species characteristic of any area depending both upon edaphic conditions and upon the particular combinations of temperature and precipitation found there. Climatic data have been brought together for some 45 stations, some within the breeding range of Bufflehead and others beyond i t s lim i t s , and these data are tabulated i n the Appendix. That material i s summarized in succeeding paragraphs, and data for selected stations are presented i n Table 31, The stations l i s t e d represent the major breeding areas i n B r i t i s h Columbia and Alberta, and stations just outside the breeding range to the south and north, respectively. Table 31. Climatic Data for Selected Weather Stations. Station Mean Daily Temperatures ( C) Mean No. Days Temp.)6°C Pptn. Annual (cm.) Summer 1« of Pptn. in Summer Dog Creek, B.C. 21 8 163 39.5 16.2 41 Beaverlodge, Alta. 21 7 ca. 145 42.5 15.4 36 Denver, Colo. 28 14 ca. 220 36.2 11.9 33 Norman Wells, N.W.T. 19 8 118 28.9 15*2 53 (Temperatures l i s t e d are mean daily maxima and minima averaged for June, July, and August). From these examples i t i s clear that beyond the southern limits of the range summer temperatures are much higher, and beyond the northern limits the growing season is to follow p.S3 Figure 15. Areas covered by the Canadian Shield. - 84 -much shorter, than i s the case i n the main breeding areas. Proportionately less precipitation f a l l s i n summer at Denver and more at Norman Wells than at the other stations, although the differences are not striking. It was shown earlier (pp.17, 19) that the southern limits of the distribution corresponded i n the west to the upper part of the Ponderosa Pine Zone, and i n the east to the southern edge of the Aspen groveland, South of, or altitudinally lower than, these zones, precipitations are similar to and temperatures higher than those of the major breeding areas, and those regions are characteristically grassland. To the north, the short growing season combined with the severity of the winters leads to a thinning-out of forests not far beyond the limits of the Bufflehead's breeding range, although the "forest-tundra ecotone" extends considerably further. On the west-facing slopes of the Rocky and Coast Mountains, higher precipitation coupled with moderate temperatures permits the establishment of forests, and the same is true for areas east of the range of Bufflehead. It i s thus not a lack of trees, such as prevails i n the Prairies and in the Tundra, that prevents the Bufflehead breeding in those areas, so other factors must be operative. Census data cited on pp. 20-22 suggested that no scarcity of Flicker holes existed either within the breeding range of Bufflehead or i n the eastern forests. Figure 16 presents the overall range of various Flickers, as given by Bent (1939). Comparison with Figure 14 shows that the only region where the limits of Flicker distribution do not l i e beyond those of Bufflehead is along the edge of the Canadian Shield northwest of Reindeer Lake, Saskatchewan, where both species extend almost to the li m i t of trees. However, as shown earlier, the abundance of Flicker holes may be significant locally i n limiting the number of pairs breeding at a given lake. Even where nest-holes are abundant there seems to be a lower li m i t to to follow p.84 Figure 16. Breeding Distribution of Flickers (Colaptes spp.). after Bent (1939), "Life Histories of Worth American Woodpeckers (Order Piciformes)." - 85 -the "areal" tolerance of the species (pp. 30-31), although i t appears to be the female Bufflehead that i s defended rather than the "territory". The element of food supply for adults as a regulating factor for density can hardly be dismissed completely, but i t may only act indirectly through the spacing-out of breeding pairs. Either abundance of nest-sites or spatial tolerance (or both) may set the upper lim i t to the breeding population of a given area, as far as can now be ascertained. Relationships of Breeding Distribution to Nesting Biology. Present knowledge is of l i t t l e help i n this case. As stated earlier, i t seems l i k e l y that the vast majority of Bufflehead pairs which secure a suitable nest-site w i l l succeed i n laying eggs, incubating, and hatching young. I t is possible that high summer temperatures i n regions south of the normal range may reduce hatehability of eggs, as has been suggested for some other species, but this remains to be proved. Other factors acting here seem unlikely to limit distribution. Relationships of Breeding Distribution to Factors affecting Survival of Flightless Young. Food supply may perhaps be limiting on the survival of flightless young i n the forested regions of the Canadian Shield. Breeding waterfowl of any species are exceedingly scarce there (Hanson et a l , 1949), and the only records for Bufflehead breeding are those from Favourable Lake and Albany River in western Ontario ( B a i l l i e , i n l i t t , 1959). It is possible that with poor soils and inefficient drainage l i t t l e leaching of nutrient minerals takes place in this region, and in consequence both aquatic vegetation and aquatic invertebrates might be expected to be scarce i n the lakes. Here too, the "lake-type" normally chosen may be rare or lacking, many lakes on the "Shield" being oligotrophic or dystrophic. Climate i s another factor which appears to be correlated with survival of young Bufflehead, and i t can also be shown that a good correlation exists with - 86 -distribution. I t seems unlikely that any but the most extreme weather conditions (e.g. hail) would have any effect upon the survival of adult Bufflehead, but low temperatures i n combination with precipitation might well lead to c h i l l i n g i n young, particularly when the latter are very small. This would be of major importance for newly-hatched young travelling through rain-soaked vegetation to the water, while young separated from the parent for any reason would probably not survive long under conditions of cold and wet. Although this has yet to be proven, i t seems probable that c h i l l i n g i s an important cause of death i n young here, as i n the Velvet Scoter studied by Koskimies (1955). Furthermore, prolonged wet spells even without low temperatures might prevent feeding activities i n small young, and this too could easily result i n weakness or death. 0 Mean daily minima i n a l l regular breeding areas are above 6 C i n a l l three summer months, but few have minima exceeding 10°C. Maxima likewise f a l l into a narrow range, from 18 to 24°C, The 50 cm, isohyet appears to correspond well with the limits of distribution i n the western mountain regions as well as in the east, while the 25 cm. isohyet roughly parallels the tree-line west of Great Slave Lake (see Figure 17). Another manner i n which climate might act indirectly to limit survival of young is through the degree of development of the shrub layer (see p.13 and compare Figures 18 and 4) . The precipitation i n central Alberta i s somewhat higher than in the main British Columbia breeding areas, perhaps as a result of thunder-storms, but this may perhaps explain the differences in shrub layer. In regions with s t i l l higher precipitation and temperature (e.g. the "Coast Forest" of B.C.) the understory assumes a tangled, "jungle-like" aspect, and Bufflehead breed casually or not at a l l in such regions. Loss of young on the way from nest to water might well be the crucial factor i n lack of nesting success in such areas, i f the data given earlier (p. 71 ) are typical. to follow p.86. Figure 1 7 . Isopleths of Precipitation and Length of Growing Season, compared to Breeding Distribution (shaded) of Bufflehead. to follow p.86 Figure 18. Bufflehead Nests in Populus woods near Leduc, Alberta (note dense shrub l a y e r around both nests). - 87 -Climate may also operate to l i m i t the breeding range of Bufflehead through the length of the ice-free period on the breeding lakes. If one allows one week from arrival on "fee nesting lakes u n t i l the start of laying, an average of 13 days for laying, 30 days for incubation, and 55 days for fledging, a minimum of 105 days must elapse between "breakup1* and "freeze-up" in order that any young may survive to flying age. A further two weeks would probably be necessary to ensure that young be produced i n sufficient numbers and be sufficiently strong awing that some might survive to perpetuate that breeding population. Near the northern limits of the range " f i r s t broods" appeararound July 1, while "freeze-up" occurs i n late September on the average. Any young that survive to f l i g h t must have hatched by about July 20, within two or three weeks of the f i r s t brood. Probably about 130 days represents a minimum average length of ice-free period, since i n late years breeding may be delayed up to ten days after the average dates, and the 130-day isopleth is plotted i n Figure 17, The limi t of trees closely parallels this line, but l i e s somewhat further north, probably nearer the 120-day isopleth, A Comparison of the Breeding Distributions of Bufflehead and of Related Species, Like many other North American waterfowl, the Bufflehead has an essen-t i a l l y western breeding distribution, but no other species precisely parallels i t s range. The Common and Barrow's Goldeneyes together occupy i n the west much the same area as the Bufflehead, but the Common Goldeneye also breeds east to the Atlantic coast. The Barrow's Goldeneye has also an apparently separate population i n Iceland and perhaps i n Labrador, a discontinuous distribution somewhat similar to that of the Harlequin Duck, It i s possible that reports of Bufflehead breeding in Maine i n the 19th Century were the last records of an eastern population i n this species also. If this were the case any theory of precipitation limiting distribution either directly or indirectly would need to - 88 -be modified. The breeding range of the White-winged Scoter in the west is somewhat similar to that of the Bufflehead, but the latter i s far more abundant through most of the area. Before considering in more detail the evolution of the present breeding distribution of the Bufflehead i t may be well to look at the evolution of this species, as far as this may be deduced from present evidence. Systematic Position of the Bufflehead. Delacour and Mayr (1945) placed the Bufflehead i n the tribe Mergini, closest to the Goldeneyes and somewhat more distant from the Mergansers. In adult plumage, the drake Bufflehead resembles the Goldeneyes and the Hooded Merganser most, while the head pattern of the female is similar to those of the White-winged and Surf Scoters and the Harlequin Duck. The pattern of the downy young is also most like that of the Goldeneyes, but young of a l l species i n the Mergini are somewhat similar. Displays vary widely i n the group, but Myres (1959b) has indicated that those associated with copulation i n Bufflehead resemble i n some elements those of the White-winged and Surf Scoters, while other important elements are also found, some only i n vestigial form, i n the Goldeneyes. In nest-site used the Bufflehead i s similar to but perhaps more rig i d l y selective than are the other species which nest i n tree-holes, and i t s nesting schedule differs very l i t t l e from other species i n the Mergini. Parental care of the young is similar to that of the Goldeneyes, differing clearly from that shown by Mergansers and at least i n detail from that of the Scoters. The fledging period is probably shorter than that of most other species in the Mergini. The more detailed information now available confirms the close similarity (and presumably the relationship) between the Bufflehead and the Goldeneyes, but suggests that at least as many elements of taxonomic importance are shared with the Scoters as with the Mergansers. The most striking difference between the Bufflehead and a l l - 89 -other ducks in the Mergini i s i t s size (see p. 5), and this i s associated with many of the ways i n which i t deviates from i t s relatives. I t seems most probable that the deviation in size i s primarily an adaptation to a type of nest-site, since a l l other species are larger and have more generalized requirements i n this respect. One may speculate that the "ancestral Merginid" was of a size comparable to most of the present members of the tribe, and nested i n cavities both on the ground and i n trees. Through reduction i n size, the Bufflehead became able to u t i l i z e nest-holes of the "ancestral Flicker", which offered enhanced protection for the nest as well as being more numerous than were the larger cavities used by other forms. However, smaller size cannot be considered an unmixed blessing. The young of Bufflehead are small compared to those of other species, so they expose proportionately greater areas for heat loss relative to their bulk. This may be expected to make them more vulnerable to exposure, while loss of young on the way from nest to water may also be greater for the smaller species. Another behavioural t r a i t characteristic of the Bufflehead, as of most of i t s relatives, may also place this species at a disadvantage. The Bufflehead i s the smallest species of duck which leads i t s young to open water to feed; other species which do so are larger, as are the Goldeneyes or Scoters, or exercise much greater parental care, as do the Grebes. Other water birds of comparable size (Teal, Ruddy Ducks, Coots) frequent more marshy areas where food may be secured with less exposure. Evolution of the Present Breeding Distribution. The f o s s i l record of birds i s of l i t t l e help for plotting past distributions, since migrant species may be preserved far from their breeding ranges. Moreau (1951) considers that migration i s of very ancient origin, and that the Pleistocene glaciations did no more than set the present patterns of f l i g h t routes. It may be worth - 90 -looking at present migration patterns for indications, however slight, of former ranges, since the f o s s i l record of Bufflehead extends at least to the Pliocene (Lane, 1946), Banding returns alone can only indicate two points on a migration route and, even when numerous returns between two areas are secured, may not indicate more than the areas wherein a species i s easy to trap and easy to shoot or recapture. However, a picture of migration routes based upon banding may be accepted without much question when i t i s supported by "visible migration". The concept of "standard direction" in migration has been discussed by many workers (e.g. Thompson, 1953} Hochbaum, 1955), and i t i s generally accepted that "standard directions" may be modified by topography. In the present discussion i t w i l l be postulated that "standard directions" may represent the directions i n which birds flew from their breeding grounds to the wintering grounds during the last glaciation. Banding returns, illustrated in Figure 19, suggest that "standard directions" i n Bufflehead trend to the southwest and southeast from each breeding area. At present nearly the entire breeding range is closer to the Pacific coast than to the Atlantic, and one might expect that the standard direction leading to the Pacific would be favoured, at least where mountains do not present too much of a barrier. Most birds banded i n the interior of British Columbia presumably follow the major valley systems, which follow a generally southwesterly direction to the Pacific coast. Birds banded i n the Peace River basin also go predominantly to the west coast, crossing the Rocky Mountains at their lowest point just southwest of the Peace River parklands. Some birds banded i n central Alberta also appear on the Pacific coast, and one return, on a line from Ministik Lake to the Peace River country, suggests that these birds follow the same route through the Rockies as do the Peace River birds. to follow p.90# Figure 19. Major Migration Directions of Bufflehead, as indicated by Banding Returns. Banding Localities, and Number of Returns from each. A. Minto Lakes, Alaska ( l ) ; B. Tetlin, Alaska (5); C. Cariboo-Kamloops d i s t r i c t , B.C. (106, not including 38 direct returns from int e r i o r B.C.); D. La Glace, Alta. (10); E. Athabaska delta, Alta. (4); F. Ministik Lake, TofieId, A l t a . (15); G. Yorkton, Sask, (4); H. Dafoe, Sask. (1); J. Brickson, Man. (4); K. Baldur, Man. (1); L. Branchport, N.Y. (1). Ma.jor Wintering Grounds^ and Number of Returns to each. I. Georgia Strait, Puget Sound, coastal Washington (74)j II. coastal Oregon and California (28); I I I . coastal Hew England and middle Atlantic States (2). - 91 -Nearly a l l birds banded elsewhere in the Prairie Provinces have moved i n generally southeasterly directions from the points of banding, although a considerable scatter occurs. A few birds apparently follow the eastern foothills of the Rockies south, even as far as Mexico; since most of these birds were banded as adults i t seems unlikely that they were individuals which had failed to find their way through the Rockies to the west coast. When a l l returns are considered i t i s found that about five percent of birds banded i n B r i t i s h Columbia or i n the Peace River area have moved to the southeast, as have about half of the birds banded at Ministik Lake, and nearly a l l of those from more eastern points of banding. In Alaska also most returns from Tetlin had moved to the southeast, although one out of five direct returns was to the southwest. Sight records supporting these standard directions include Bufflehead moving southeast on the Prairies i n f a l l migration (Barnes, 1912; Swallow, 1941), while heavy movements of this species were noted i n early November at Alta Lake, Br i t i s h Columbia, i n one of the main valleys running southwest from the interior to the Pacific coast (Racey, 1948). Post-breeding dispersal to the north, such as occurs i n Barrow's Goldeneye banded in Br i t i s h Columbia, appears not to be characteristic of Bufflehead, the only return north of the point of banding being that of the bird banded at Ministik Lake and recovered towards the Peace River. Movements to the southeast from the British Columbia breeding areas must cross the Rockies, as must birds moving to the Pacific from Alberta, even i f the latter do cross the mountains at their lowest point. It is d i f f i c u l t to imagine anything, other than an inherited directional tendency, leading birds to cross a major mountain range when easier alternative routes exist. For the Bufflehead to have evolved two such standard directions in the Pleistocene, one must postulate that the wintering range at that time lay partly to the - 92 -southwest and partly to the southeast of the breeding range. Of the various refugia advanced for the Pleistocene survival of North American boreal species (see e.g. F l i n t , 1947), those i n either the Eocky Mountain region or i n the Appalachians w i l l f i t "this hypothesis. That in Alaska requires migration across the ice-sheets to the southeast, and that i n the Pacific coast mountains migration across the Rockies, neither of which seems l i k e l y as a general practice. F l i n t (1947) has demonstrated that the major climatic changes of the Pleistocene can be explained on the basis of mountain-building during that epoch affecting the former paths of weather systems, and much evidence has been presented (e.g. Chaney, 1940) that north-south climatic zonation has existed since at least Eocene times. Numerous controversies have arisen concerning the extent to which vegetational zones were deflected to the south or reduced to narrower bands by the advance of the ice-sheets (e.g. Braun, 1951; Potzger, 1951), but i t seems well established that the same genera and probably the same speeies of trees existed i n various longitudes i n the Pleistocene as they do today. Forest associations of that epoch probably bore considerable ecological resemblance to those of the present, and with that i n mind one may question whether the Bufflehead would have had i t s major refugium i n the Appalachians, since i t is now found i n neither the deciduous nor the coniferous forests of that region. At present there are i n that region few areas with lakes suitable for much breeding of waterfowl, and this may also be taken as evidence against the Appalachian refugium. Since the last (Wisconsin) retreat of the ice-sheets, usually dated as having begun about 25,000 years ago, climates are considered to have become warmer and drier u n t i l the climatic "optimum" about 6,000 years ago, and since then a slight cooling has occurred. Vast lakes of "melt-water" persisted i n Canada after the retreat of the ice, and un t i l these dried up at - 93 -the time of the climatic "optimum" much of Alberta, northern Saskatchewan, and the Mackenzie Basin i s believed to have been unforested (Raup, 1947). Hansen (1952, 1953) favours the persistence of forests on the east slopes of the Rockies i n northeast Br i t i s h Columbia, and also in the unglaciated portions of Alaska, throughout the glacial period, but other workers disagree. In any case, i t seems unlikely that forest cover i n this area can have been sufficiently extensive to have provided a major refugium for tree—nesting species of birds. One line of speculation has i t that the breeding range of the Barrow's Goldeneye was s p l i t i n two by the Pleistocene glaciation, with the Common Goldeneye subsequently colonizing North America from Eurasia while the Barrow's Goldeneye was s t i l l blockaded i n the Rocky Mountain refugium. Competition between the two closely related species prevented the subsequent range expansion of the Barrow's Goldeneye, while the lack of a related competitive species might have permitted the Bufflehead to spread east of the Rockies, i f i t too had a formerly trans-continental range divided by the ice-sheets. However, no competitor exists to prevent the spread of the Bufflehead to the Atlantic, while an ecological barrier might serve this function. The question i s far from closed, but as presented here the evidence seems to favour a western distribution of the species since before the Pleistocene, breeding in the east having been sporadic or casual i n nature. SUMMARY Conclusions to be drawn from this study may be summarized under the four topics of i t s major sections, namely: breeding habitat, nesting biology, survival of flightless young, and factors limiting breeding distribution. For breeding the Bufflehead selects water bodies i n forested or partly forested areas. Eutrophic lakes of moderate depth are probably favoured, and dense growths of surface or emergent vegetation are avoided. Regions with a "parkland" type of forest cover provide the major breeding areas, and poor development of the "shrub layer" i s probably desirable. Nest-sites are usually selected close to the water, and differ l i t t l e in form from typical Flicker holes. No other hole-nesting duck in North America can enter most of the holes used for nests by Bufflehead. The breeding biology of Bufflehead in the study area is quite similar to that of the Goldeneyes. Laying commences soon after the ice leaves the breeding lakes, and eggs are deposited at intervals of usually greater than one day. Clutch-size averages about 8,6 eggs for original nestings, and perhaps two eggs less for re-nests. Incubation usually occupies 29 to 31 days, and the young depart from the nest 24 to 36 hours after hatching. Departure from the nest may be greatly prolonged in comparison to those described for related species. Losses during the "nest stages" are relatively l i g h t , about 80 percent of nests succeeding and over 90 percent of eggs in successful nests giving rise to young. Survival of flightless young Bufflehead appears to be well correlated with favourable weather. Ihen low temperatures and continuous rain coincided with the period in which broods f i r s t appeared in 1959, up to 85 percent of young disappeared without evidence being secured to indicate the operation of - 95 -other potential mortality factors,, Under favourable weather conditions i n 1958 about 70 percent of young survived to at least one month of age. The importance of other mortality factors remains to be proved, but i t was f e l t that losses of young between the nest and the water might assume serious proportions under some circumstances. The smaller size of Bufflehead young possibly makes them more vulnerable to exposure as well as to loss while travelling through vegetation to the water. However, the Bufflehead seems to have nicely balanced the disadvantages of having smaller young against the benefits of more abundant nest-sites, since i t has survived as a species since at least the Pliocene and seems i n l i t t l e danger of extinction at the present time. The present distribution of the Bufflehead is in the western part of North America. To the north the li m i t of trees would set a limit to the breeding range, but i t seems probable that the length of the ice—free period on the lakes limits distribution somewhere short of the tree-line. To the northeast trees are sparse upon the Canadian Shield, and suitable lakes and food supply may be scarce or lacking i n that region. Lack of nest-sites probably limits the range to the south, where "parkland" grades into the Prairies and the Great Basin, while a lack of suitable lakes makes breeding habitats in the western mountains scarce south of the Canadian border. The breeding distribution to the east and i n the mountains of British Columbia seems to end at the 50 cm. isohyetj i t seems possible that this may operate through increase i n the luxuriance of shrub layers under higher precipitations, thus reducing the proportion of young reaching the water. The evidence presented seems to favour a Rocky Mountain "refugium" for the species during the Pleistocene, i t s distribution having probably been western in nature since before that epoch. - 96 -Further studies on the distribution of this species might most profitably concentrate upon factors limiting spread to the south and east, since these seem less well explained than are the limits in other directions. Further studies on the Bufflehead might do well to utilize its habit, peculiar among ducks, of nesting in Flicker holes. Such sites are readily located, and are fixed in location, thus obviating a fresh search for nests each year, while around favoured lakes quite high densities may occur. The females are readily trapped upon the nest, and survive for long enough periods that they may be retrapped in several consecutive years. For these reasons the Bufflehead is particularly suited to studies of territory (home range), nest—site tenacity, and re-nesting, and is to be preferred to some waterfowl used for such studies in the past. - 97 -REFERENCES CITED References marked with an asterisk have distributional significance only, in this thesis. Nearly a l l major English-language journals dealing chiefly or solely with natural history were consulted, and coverage for breeding distribution and biology was as complete as possible. References dealing with the Bufflehead on migration and on i t s wintering grounds were also surveyed, although somewhat less completely, but have only been included here when pertinent to the discussion. The form of citation follows that of "Auk", except that names of individual authors are written out in f u l l once only. References cited only i n the Appendices are not included. * Alberger, A.H. (1890), Ornith, and Ool,, 15: 87-88 (reference not seen) * American Ornithologists' Union (1957), "Check-list of North American Birds, F i f t h Ed,". Arnold, E. (1895), "An outing i n Assiniboia, 1895", Oologist, 12: 168-170. Audubon Society, National (1936-1958), F i r s t to Twenty-First Breeding Bird Censuses, published annually in "Bird-Lore", "Audubon Magazine", and "Audubon Field Notes". * B a i l l i e , J.L., and P. Harrington (1936), "The distribution of breeding birds in Ontario. Part I", Trans, Roy. Canad. Inst., 21: 1-50, Baird, S,F,, T.M, Brewer, and R. Ridgway (1884), "The water birds of North America", vol. 2; L i t t l e , Brown, and Co,, Boston. * Barnes, R.M. (1909), "The Bufflehead duck", Oologist, 26: 73-74. — (1912), "A western Minnesota early blizzard", Oologist, 29: 381-382. * (1923), Preface, Oologist, 40: 141, Bent, A.C. (1925), "Life histories of North American wild fowl (Order Anseres)", U.S. Nat. Mus., B u l l , 130: 1-316. - - - (1939), "Life histories of North American woodpeckers (Order Piciformes)", U.S. Nat. Mus., B u l l . 174: 1-334. Bird, R.D. (1930), "Biotic communities of the Aspen parkland of central Canada", Ecology, 11, 356-442. * Bishop, L.B. (1900), "Birds of the Yukon region, with notes on other species", N. Amer, Fauna, 19: 47-96. Braun, E.L, (1951), "Plant distribution in relation to the glacial boundary", Ohio Jour. Sci., 51: 139-146. - 98 -* Brewster, ¥, (1924), "The birds of the Lake Umbagog region of Maine. Part I", B u l l . Mus. Comp, Zool. Harvard, 66: 1-210. Brit i s h Columbia, Dept. of Agric. (1951-59), "Climate of British Columbia, Tables of temperature, precipitation, and sunshine". Brooks, A. (1903), "Notes on the birds of the Cariboo d i s t r i c t , British Columbia", Auk, 20: 277-284, * (1917), "Birds of the Chilliwack d i s t r i c t , B.C.", Auk, 34: 28-50. * Brooks, W.S., and S. Cobb (l91l), "Notes from eastern Alberta", Auk, 28: 465-469. Burns, F.L. (1900), "A monograph of the f l i c k e r (Colaptes auratus)", Wilson Bul l . , 12; 1-82. * Cantwell, G.G, (1898), "Birds of the Yukon t r a i l " , Osprey, 3: 25. Carter, B.C. (1958), "The American Goldeneye i n central New Brunswick", Canad, Wildl. Service, Wildl. Man. Bu l l . , Ser. 2; 9: 1-47. Gary, M. (1917), "Life zone investigations in Wyoming", N. Amer. Fauna, 42: 1-95. Chaney, R.W. (1940), "Tertiary forests and continental history", Bull . Geol, Soc. Amer., 51: 469-488, Cook, G.L. (1930), "Runt eggs", Oologist, 47: 80. * Cory, C.B. (1909), "The birds of I l l i n o i s and Wisconsin", Field Mus. Nat. Hist,, Zool, Ser., 9: 1-767. Cottam, C, (1939), "Food habits of North American diving ducks", U.S. Dept. Agric., Tech. B u l l . , 643: 1-139. * Coubeaux, E. (1900), "Contributions to the natural history of the northwest territories, I. The birds of southern Saskatchewan", Ottawa Nat., 14: 24-31. Cowan, I. McT. (1939), "The vertebrate fauna of the Peace River d i s t r i c t of British Columbia", Occas. Papers, B.C. Prov, Mus., 1: 1-102, — — — (1948), "Waterfowl nesting ground studies in British Columbia -1948", U.S. Fish and Wildl. Service, and Canada, Dom, Wildl. Service, Spec. Sci. Rpt., 60: 26-39, - - - and J. Hatter (1952), "A trap and technique for the capture of diving waterfowl", Jour, Wildl, Mgt., 16: 438-441, Dane, B., C. Walcott, and ff.H, Drury (in preparation), cited i n Dane, B., C. Walcott, and W.H. Drury (1959), "The form and duration of the display actions of the Goldeneye (Bucephala clangula)", Behaviour, 14: 265-281, - 99 -Davis, D.E0 (1955), "Breeding biology of birds", in "Recent studies in avian biology", Amer.Ornith.Union; Univ.111.Press, Urbana, pp.264-308. Davis, H.P. (1941), "Nesting of Bufflehead duck at Lake Almanor, California" Condor, 43: 294. * Dawson, W.L. (1897), "A preliminary l i s t of the birds of Okanogan County, Washington", Auk, 14: 168-182. * and J.H.Bowles (1909), "The birds of Washington", vol.2, The Occidental Publ.Co., Seattle, Delacour, J., and E. Mayr (1945), "The family Anatidae", Wilson Bull,, 57: 3-55. - - - and - - - (1946), "Supplementary notes on the family Anatidae", Wilson B u l l . , 58: 104-110. Dice, L.R. (1920), "Notes on some birds of interior Alaska", Condor, 22: 176-185. * Dixon, J.S. (1921), "The Bufflehead breeding in California", Condor, 23: 165. (1926), "The Bufflehead breeds i n California", Condor, 28: 47-48. * (1938), "Birds and mammals of Mount McKinley Park", U.S. Nat.Parks, Fauna Ser., 3: 1-236. Drury, W.H. (1953), "Birds of the Saint Elias quadrangle in the southwestern Yukon Territory", Can.Pield-Nat., 67: 103-128. * Eaton, S.W. (1948), "Bird distribution along the Peace, Slave, and L i t t l e Buffalo Rivers of Canada", Auk, 65: 345-352. Erskine, A.J. (1959a), "A joint clutch of Barrow's Goldeneye and Bufflehead eggs", Can.Field-Nat,, 73: 131. - - - (1959b), "A method for opening nesting holes", Bird-Banding, 30: 181. - - - (i960), "Further notes on inter-specific competition among hole-nesting ducks", Can.Field-Nat., 74: in press, Evans, CD. (1951), "A method for color marking young waterfowl", Jour. Wildl.Man., 15: 101-103, * Evenden, F.G. (1947), "The Bufflehead nesting in Oregon", Condor, 49: 169. - 100 -* Farley, F.M. (1922), "Summer birds of the Lac l a Biche and Fort McMurray region", Can. Field-Nat., 36: 72-75. Fisher, J. (l95l), "Bird recognition. 2. Birds of prey and waterfowl", Penguin Books, Harmondsworth, Sussex. F l i n t , R.F. (1947), "Glacial geology and the Pleistocene epoch", John Wiley and Sons, New York. * Forster, J.R. (1772), "An account of the birds sent from Hudson's Bay; with observations relative to their natural history; and Latin descriptions of some of the most uncommon", P h i l . Trans., 62: 382-433 (reprinted by the Willughby Society, 1882). * Friedmann, H. (1935), "The birds of ivodiak Island, Alaska", B u l l . Chicago Acad. Sci., 5: 13-54. Furniss, O.C. (1935), "The duck situation in the Prince Albert d i s t r i c t , central Saskatchewan", Wilson Bull., 47: 111-119. (1938), "The 1937 waterfowl season in the Prince Albert d i s t r i c t , central Saskatchewan", Wilson Bu l l . , 50: 17-27. * Gabrielson, I.N., and F.C. Lincoln (1959), "The birds of Alaska", The Stackpole Co., Harrisburg, Pa., and Wildl. Mgt. Inst., Wash., B.C. Godfrey, W.E. (1951), "Notes on the birds of southern Yukon Territory", Nat. Mus. Canada, B u l l . , 123: 88-115. * - - — (1952), "Birds of the Lesser Slave Lake - Peace River areas, Alberta", Nat. Mus. Canada, Bull., 126: 142-175» * - (1953), "Notes on birds of the area of intergradation between eastern prairie and forest in Canada", Nat. Mus. Canada, Bull., 128: 189—240. * Griffee, W.E. (1958), "Notes on Oregon nesting of American Merganser and Barrow's Goldeneye", Murrelet, 39: 26. Grinnell, J., J. Dixon, and J.M. Linsdale (1930), "Vertebrate natural history of a section of northern California through the Lassen Peak region", Univ. Cal. Publ. Zool., 35: 1-594. Halliday, W.E.D. (1937), "A forest classification for Canada", Canada, Dept. Mines and Resources, For. Serv. Bull., 89: 1-50. Hansen, H.P. (1952), "Postglacial forests in the Grande Prairie - Lesser Slave Lake region of Alberta, Canada", Ecology, 33: 31-40. (1953), "Postglacial forests in the Yukon Territory and Alaska", Amer. Jour. Sci., 251: 505-542, - 101 -Hanson, H.C. (1949), "Waterfowl of the forested portions of the Canadian Pre-Cambrian Shield and the Palaeozoic Basin", Can.Field-Nat., 63: 183-204. * Henderson, A.D. (1924), "With canoe and camera on some Alberta lakes", Oologist, 41; 14-18. (1927), "New tenants in old nests", Oologist, 44: 166-168. (1936), "Field notes of 1925", Oologist, 53s 129-131. Hochbaum, H.A. (1944), "The Canvasback on a prairie marsh", amer.Wildl. Inst,, Washington, D.C. - - - (1955), "Travels and traditions of waterfowl", Univ.Minn, Press, Minneapolis. * Hooper, R., and D. Hooper (1954), "A preliminary l i s t of the birds of the Somrae district, Saskatchewan", Yorkton Wat.Hist.Soc., Contr., 3: 1-13. Horsbrugh, C.B. (1915), "Ornithological notes from the Alix and Buffalo Lake districts, Province of Alberta, Canada, 1914", Ibis, ser. 10; 3: 670-689. * - - - (1918), "Further notes on birds observed at Alix, Buffalo Lake, and Red Deer in the Province of Alberta, Canada, in 1915 and 1916", Ibis, ser. 10; 6: 477-496, Jackson, M.F. (in preparation), Ph.D. thesis, Univ.British Columbia. Jennings, A.R., and E.J.L. Soulsby (1958), "Disease in a colony of Blackheaded Gulls Larus ridibundus", Ibis, 100: 305-312. * Johnstone, W.B. (1949), "An annotated l i s t of the birds of the East Kootenay, British Columbia", Occas.Papers, B.C. Prov.Mus., 7: 1-87. * Kelso, J.E.H. (1926), "Birds of the Arrow Lakes, West Kootenay district, British Columbia", Ibis, ser. 12; 2: 689-723. Kendeigh, S.C. (1934), "The role of environment in the l i f e of birds", Ecol. Monogr., 4s 301-417. Kendrew, W.G., and B.W. Currie (1955), "The climate of central Canada", Queen's Printer, Ottawa. * Knight, O.W. (1897), "A l i s t of the birds of Maine", Univ.Maine, Dept. Nat.Hist., Bull., 3: 1-184, Kortright, F.H. (1942), "The ducks, geese, and swans of North America", The Stackpole Co., H&rrisburg, Pa., and Wildl.Mgt.Inst., Wash., D.C. - 102 -Koskimies, J. (1955), "Juvenile mortality and population balance in the Velvet Scoter (Melanitta fusea) in maritime conditions", Acta XI Congr. Int.Ornith., (1954): 476-479. - - - (1957a), "Polymorphic v a r i a b i l i t y in clutch size and laying date of the Velvet Scoter, Melanitta fusca (L.)", Orn.Fennica, 34: 118-128. - (1957b), "Variations in size and shape of eggs of the Velvet Scoter, Melanitta fusca (L.)", Arch.Soc."Vanamo", 12: 58-69. Krajina, V.J. (1959), "Bioclimatic zones in British Columbia", Univ.Brit. Col., Bot.Ser., 1: 1-47. * Labarthe, J. (1922), "Further remarks on the occurrence of the Bufflehead at Eagle Lake", Condor, 24: 68-69. Lack, D. (1954), "The natural regulation of animal numbers", The Clarendon Press, Oxford. Lane, H.H. (1946), "A survey of the f o s s i l vertebrates of Kansas. Part IV, Birds", Acad.Sci., 49: 390-400. Lynch, Brother D. (1955), "Ecology of the Aspen groveland in Glacier County, Montana", Ecol.Monogr., 25: 321-344, McLaren, W.D. (in preparation), M.Sc. Thesis, Univ.Brit.Col. McLaughlin, C.L., and D, Grice (1952), "The effectiveness of large-scale erection of food Duck boxes as a management procedure", Trans.N.Amer. Wildl.Conf., 17: 242-259, * Macoun, J . (1900), "Catalogue of Canadian birds. Part I. Water birds, Gallinaceous birds, and pigeons", S.E. Dawson, Printer to the Queen's Most Excellent Majesty, Ottawa, * and J.M. Macoun (1909), "Catalogue of Canadian birds", Govt, Printing Bureau, Ottawa. Mendall, H.L. (1958), "The Ring-necked Duck in the Northeast", Univ.Maine Bul l . , 60: 1-317. Mill e r , A.H. (1951), "An analysis of the distribution of the birds of California", Univ.Cal.Publ.Zool., 50: 531-643. * Morden, J.A., and W.E. Saunders (1882), Can.Sportsman and Naturalist, 2. (reference not seen). Moreau, R.E, (1951), "The migration system in perspective", Proc. X Int, Ornith.Congr., (1950): 245-248. - 103 -Moss, E.H. (1932), "The vegetation of Alberta. IT. The poplar association and related vegetation of central Alberta", Jour.Ecol., 20; 380-415. * Mowat, P.M. (1947), "Notes on the birds of Emma Lake, Saskatchewan", Can.Field-Nat., 61: 105-115. Munro, J.A. (1918), "The Buffle Head in the Okanagan Valley, British Columbia", Oologist, 35: 12-13. * (1927), "Concerning August birdlife in the Okanagan Valley, British Columbia", Can.Field-Nat., 41: 108-110. * - - - (l929a,b; 1930a,b), "Glimpses of little-known western lakes and their bird l i f e " , Can.Field-Nat., 43: 181-187, 200-205; 44: 12-17, 40-45. - - - (1937), "Studies of waterfowl in the Cariboo region, British Columbia", Condor, 39: 163-173. - (1942), "Studies of waterfowl in British Columbia. Bufflehead", Can.Jour,Research, 20D: 133-160. — — — (1947), "Observations of birds and mammals in central British Columbia", Occas.Papers, B.C. Prov.Mus., 6: 1-165. * - - — (1948), "Waterfowl breeding conditions in British Columbia, 1948", U.S. Fish and Wildl.Serv., and Canada, Dom.Wildl.Serv., Spec.Sci.Rpt., 60: 13-25. (1949), "The birds and mammals of the Vanderhoof region, British Columbia", Amer.Mid.Nat., 41: 1-138. Myres, M.T. (1957), "An introduction to the behaviour of the Goldeneyes: Bucephala islandica and B.clangula (Class Ave3, Family Anatidae)", M.A. Thesis, Univ.Brit.Col. - (1959a), "The behaviour of the sea ducks", Ph.D. Thesis, Univ. Brit.Col. (1959b), "Display behaviour of Bufflehead, Scoters, and Goldeneyes at copulation", Wilson Bull., 71: 159-168. Nice, M.M. (l94l), "The role of territory in bird l i f e " , Amer.Mid.Nat., 26: 441-482. (1954), "Problems of incubation periods in North American birds", Condor, 56: 173-197. Northcote, T.G., and P.A. Larkin (1956), "Indices of productivity in British Columbia lakes", Jour.Fish.Res.Bd.Canada, 13: 515-540. * Palmer, R.S. (1949), "Maine birds", Bull.Mus.Comp.Zool.Harvard, 102: 1-656. - 104 -Palragren, P, (1936), "Uber die vogelfauna der birmengewasser Alands", Acta Zool.Fennica, 17: 1-59. Peterson, R.T. (1947), "A field guide to the birds", Houghton Mifflin Co., Boston. Phillips, J.C. (1925), "A natural history of the ducks", vol,3; The Riverside Press, Cambridge, Potzger, J.E. (1951), "The fossil record near the glacial border", Ohio Jour. Sci., 51: 126-133. Pough, R.H. (1951), "Audubon water bird guide", Nat.Audubon Soc; Doubleday and Co., Inc., Garden City, N.I. * Preble, E.A. (1908), "A biological investigation of the Athabaska -Mackenzie region", N.Amer.Fauna, 27: 1-574. Racey, K. (1948), "Birds of the Alta Lake region, British Columbia", Auk, 65: 383-401. Raine, W. (1892), "Bird-nesting in northwest Canada", Hunter, Rose; and Co., Toronto. * (1927), "Nest of Solitary Sandpiper, Yellowlegs, and Bufflehead duck", Oologist, 44: 4—5. Rand, A.L. (1946), "List of Yukon birds and those of the Canol Road", Nat. Mus,Canada, Bull., 105: 1-76. * Randall, T.E. (1933), "A l i s t of the breeding birds of the Athabaska district, Alberta", Can.Field-Nat., 47: 1-6, Raup, H.M. (1947), "Some natural floristic areas in boreal North America", Ecol.Monogr., 17: 221-234. * Ray, M.S. (1921), "On the occurrence of the Bufflehead at Eagle Lake", Condor, 23: 192-193. Richdale, L.E. (1949), "The effect of age on laying dates, size of eggs, and size of clutch in the Yellow-eyed Penguin", Wilson Bull., 61: 91-98. * Rosche, R.C. (1954), "Notes on some birds of Yellowstone National Park", Wilson Bull., 66: 60. * Salter, R.L. (1953), "Waterfowl breeding ground survey in Idaho", U.S. Fish and Wildl.Serv., and Canad.Wildl.Serv., Spec.Sci.Rpt., Wildl. No. 25: 152-154. * Saunders, A.A. (1921), "A distributional l i s t of the birds of Montana", Pacif,Coast Avif., 14: 1-194. 105 -* Soper, J.D. (1942), "The birds of Wood Buffalo Park and vicinity, northern Alberta and District of Mackenzie, N.W.T., Canada", Trans.Roy.Can.Inst., 24: 19-97. — - - (1949), "Birds observed in the Grande Prairie - Peace River region of northwestern Alberta", Auk, 66: 233-257. * - - - (1951a), "Waterfowl and related investigations in the Peace -Athabaska delta region of Alberta, 1949", Can.Wildl.Serv., Wildl.Mgt. Bull., ser.2; 2: 1-63. * (1951b), "The birds of Elk Island National Park, Alberta, Canada Can.Wildl.Serv., Wildl.Mgt.Bull., ser.2; 3: 1-60. * (1952), "The birds of Prince Albert National Park, Saskatchewan" Can.Wildl.Serv,, Wildl.Mgt.Bull., ser.2; 4; 1-83. * - - - (1954), "Waterfowl and other ornithological investigations in Yukon Territory, Canada, 1950", Can.Wildl.Serv., Wildl .Mgt.Bui1., ser.2; 7: 1-55. Sowls, L.K. (1955), "Prairie ducks", The Stackpole Co., Harrisburg, Pa., and Wildl.Mgt.Inst., Wash., B.C. Stansell, S.S.S. (1909a), "Birds of central Alberta", Auk, 26: 390-400. * (1909b), "Birds of Stony Plain, Alberta", Ottawa Nat., 23: 125-127. * Stanwell-Pletcher, J.P., and T.C. Stanwe11-F1etcher (1943), "Some accounts of the flora and fauna of the Driftwood Valley region of north central British Columbia", Occas.Papers, B.C. Prov.Mus., 4: 1-97. * Steel, P.E., P.D. Dalke, and E.G. Bizeau (1956), "Annotated l i s t of the avifauna in and around Gray's Lake, Idaho", Murrelet, 37: 4-10. Stewart, R.E. (1955), "Virgin bottomland White Spruce forest", Aud.Field Notes, Nineteenth Breeding Bird Census - census 9; 9: 415-416. * Stoudt, J.H. (1949), "Waterfowl breeding ground survey in the Dakotas - 1949 U.S. Fish and Wildl.Serv., and Canada, Dom.Wildl.Serv., Spec.Sci.Rpt., Wildl. No,2: 24-29. Swallow, H.S. (1941), "Rain of ducks at Foam Lake, Saskatchewan", Can.Field-Nat., 55: 130. * Taverner, P.A. (1919a), "The birds of the Red Deer River, Alberta", Auk, 36: 1-21. * (1919b), "The summer birds of Hazelton, British Columbia", Condor, 21: 80-86. - 106 -Taverner, P.A. (1928), "Ornithological investigations near Belvedere, Alberta, 1926", Nat.Mus.Canada, Bull., 50: 84-104. Tinbergen, N, (1953), "The Herring Gull's world", Collins, London, Tisdale, E.W., and A. McLean (1957), "The Douglas Fir zone of southern interior British Columbia", Ecol.Monogr., 27: 247-266. Thompson, Sir A.L. (1953), "The study of the visible migration of birds: an introductory review", Ibis, 95: 165-180. * Thompson, E.E. (1890), "The birds of Manitoba", Proc. U.S. Nat,Mus,, 13: 457-643. Williams, M.I. (1933), "Biological notes, covering parts of the Peace, Liard, Mackenzie, and Great Bear River basins", Can.Field-Nat., 47: 23-31. * Wood, N.A. (1923), "A preliminary survey of the bird l i f e of Worth Dakota", Univ.Mich., Misc.Publ,, Mus.Zoo1., 10: 1-94. Name of Lake Lat.-Long. Elev. Size Coordinates (ft.) (mi. or ac,) Tree Cover Nests and Remarks "Bailey 1s Pond" 521-1220 2750 ca.4 ac. Bridge Lake Campbell Lake Colpitt Lake Gummings Lake Disdero Lake Doctor's Lake E l l i o t Lake Pear Lake "Plat Lake" "Goose Lake" (nr. Knutsford) fd-Sp/LP one side Highway other 513-1204 3650 4 2 mi. Wd-As/LP 503-1201 3400 2 \ mi. Grv.-As 520-1220 2950 \ mi. by | mi. 520-1220 2350 _ mi. by i mi. Wd-As/LP one side DP other, open at ends Wd-DF one side, range other 504-1200 3050 1 mi. by { mi. Wd-DF one side, 520-1221 2950 ca.2 ac.(1954) ca.l ac.(1959) 514-1212 2900 ca.20 ac. 520-1221 3050 ca.2 ac. 503- 1202 3450 \ mi. by i mi. 504- 1202 3150 ca.15 ac. range other Wd-LP one side, range othe r Wd-As/LP one side range other Wd-DF one side, range other Grv.-As/PP Grv.-As/DF/PP road one side No nests, few ducks No nests, moult-ing lake # 15, 16, 17, 100, 101 # 12, 105, 106 # 77 # 19. No ducks 1959 # 35, 45 # 18 # 75, 87 # 76 Goose Lake 515-1213 2800 1 mi. by % mi, (nr. Lac la Hache) Lac la Hache 515-1213 2650 9 mi. by 1 mi. Lilypad Lake Loch Lomond Maze Lake and nearby potholes Moss Meadow Lake "Murphy Lake" (142 Mile) 101 Mile Lake 103 Mile Lake 105 Mile Lake 149 Mile Lake Pete Kitchen Lake 514-1213 2900 512-1212 3700 520- 1214 3100 521- 1220 3050 520- 1215 2300 514-1212 3200 514-1212 2900 514- 1212 2900 521- 1220 2400 515- 1212 2950 I2 mi, by % mi. 1 mi. by mi. 1 mi, by j mi, plus 10 & 5 ac, 2 mi. by i mi, i mi, by i mi, ca,10 ac, \ mi, by % mi, 1 mi. by j mi, i mi, by i mi. i mi, by 4 mi. Phililloo Lake and 515-1214 3100 2 mi. by mi. nearby ponds Wd-Df # 72, 73 Grvi-As/DP one side Wd-DF other Wd-As/DF Wd-LP, bare flats in parts Wd-DF Wd-As/LP, open at ends Grv,-As/LP one end, range rest Grv,-As/LP Wd-Sp/DF one side Grv,-As other Grv.-As GrT.-As Grv.-As Wd-DF No nests, few ducks No nests, few ducks No nests, moulting lake # 61, 84, 85, 86, 102 # 60 No nests, migration lake No nests, one brood seen # 1, 24, 25, 63 # 98 No nests, migration lake # 58, 59, 80 # 37, 38, 39, 40, 41, 42, 43, 44, 62, 70, 81, 82, 83, 88, 89, 90, 103, 104 Riske Creek potholes 520-1222 3300-(north range) to 520-1223 3450 Riske Creek potholes (south range) Rocky Lake (nr. Big Creek) Soda Lake and ponds and slough nearby "Turnoff Slough" (nr. Colpitt Lake) Watson Lake Westwick Lake Williams Lake 515-1223 513- 1231 515-1212 520-1221 514- 1212 520- 1221 521- 1220 3000 4700 3000 3000 2900 3000 1950 30+ ponds, i ac. to \ mi. by \ mi. 10+ ponds, \ ac. to \ mi, by \ mi. ca.20 ac. 1 mi. by i mi. plus 10, 3 + 3 ac. ca.2 ac. (1954) ca.i ac. (1959) 2 \ mi. 1 mi. by % mi. 5 mi. by 1 mi. Grv.-As and range Grv.-As/DF and range Shrubby f l a t s , Wd-LP beyond Wd-Sp/DF one side range other Wd-As/LP Grv.-As/LP Grv.-As Wd-DF/As one side Grv.-As/DF other # 51, 52, 53, 54, 55, 56, 57, 64, 65, 66, 67, 68, 69, 91, 92, 93, 94, 95 # 50 # 107 # 46, 47, 48, 49 # 13. No ducks 1959 # 3, 4, 5, 6, 7, 8, 9, 10, 11, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34, 36, 71, 74, 78, 79, 97 # 21, 22 No nests, migration lake Abbreviations used here: mi. — miles, ac, - acres; Grv, - Groveland, Wd - wooded (see e.g. Figures 4 and 5); As, LP, PP, DF, Sp, for tree species as in text (Table 6), Nests referred to by numbers in nest catalogue. to follow p.109. Map of British Columbia, to show Locations of Study Areas listed in Appendix I. (letters on map refer to the localities listed for each). A. Riske Creek potholes; B. Colpitt, Doctor's, Fear, and Westwiok Lakes, and "Turnoff Slough"} C. Cummings, Moss Meadow, Murphy, and 149 Mile Lakes; D. Goose, Maze, and Phililloo Lakes, and Lac la Hache; E. Bridge, Ell i o t , Lilypad, 101 Mile, 103 Mile, 105 Mile, Pete Kitchen, Soda, and Watson Lakes; F. "Rocky Lake"; G. "Flat" and "Goose" Lakes (nr.Knutsford); H. Campbell and Disdero Lakes; J. Loch Lomond. "Bailey's Pond" and Williams Lake are near the village of Williams Lake. - 110 -Appendix II. Procedures and Results of Invertebrate Sampling. (a) Description of lakes. All three lakes sampled (Watson, Phililloo, Lilypad) were about 1.5 miles long and 0.4 miles wide, with depths of 4-5 metres at the deepest, averaging 3 metres, and the long axes of a l l three ran in a northwest-southeast direction. Beds of emergent Scirpus (probably S. acutus) were present in a l l three, while beds of Typha were present in one bay at Lilypad Lake, Nymphaea covered considerable portions of the surface at Lilypad Lake, Submerged aquatic plants were scarce in Watson Lake, except at one end, and also in the "main lake" at Phililloo Lake, pH measurements with indicator papers gave values of 8.5, 7.0, and 7.2 for Watson, Phililloo, and Lilypad Lakes, respectively, (b) Plankton work. Horizontal plankton hauls were made for two minutes each, at depths of from six inches to three feet, with a six-inch # 2 mesh net. The sample was drained for one minute, and the volume was determined by differential readings. Large Amphipods (Gammarus fasciatus) were counted and their volume measured, and the relative abundance by volume of other organisms was estimated by eye. The results are tabulated below, the Groups being numbered from the end away from the outlet. Lilypad, May 21, 1958 Groups of 6 hauls Aver. Vol. per Haul (ml.) L.Am. Average Volume SoAm, Clad. (ml,) per Cop 9 Haul of Hydr. Chao. Depth (m.) I 7.4 5.7 0.94 0.25 0.22 0.30 2.0 II 1.8 0.95 0.18 0.22 0 4 3X 0.13 3.7 III 2.1 0.1 0.95 0.55 0.44 3.9 IV 1.4 0.1 0,03 0.66 0.32 0,29 3.0 also snails (0,52) in Group I, and Ostracods (0,07, 0.02, 0.06, 0.01, respectively) in a l l Groups. Phililloo, May 27, 1958 I 5.0 0.1 trace 2.80 1.25 0.25 0.55 2.5 II 4.9 0.1 2.50 1,27 0.49 0.44 2.6 III 3.7 - - 1,92 0.92 0.48 0.37 3.6 IV 3.2 0,1 trace 1.79 0.80 0.48 0.35 4.1 also Dytiscids (0.10 and 0.05, respectively) in Groups III and IV, and Ostracods ( trace) in Group IV. Watson, June 6, 1958 I 13.7 0,3 0.14 12.7 0,14 0.01 0,42 2,4 II 10.2 0.1 0.20 9.3 0.10 0.03 0.42 3.5 III 12.6 0.1 0.12 11.9 0.12 0,01 0.38 3.2 IV 8.7 0.1 0.26 7.7 0,17 0.02 0,31 2.6 also Corixids (traces only) in Groups I-III, and Culicid pupae (0.01 and trace, respectively) in Groups I and II, - I l l -Abbreviations used in Plankton Table: L.Am, - Large Amphipods, S.Am. - Small Amphipods (Hyalella?), Clad. -Clodocera, Cop. - Copepoda, Hydr, - Hydracarina, Chao. -Chaoborus (Culicidae) larvae, (c) Bottom Drags. This procedure was only qualitative„ A weighted net was dragged along the bottom, or as near to the bottom as the often dense vegetation permitted. Six Groups of four drags each were distributed in approximately the same pattern in each lake, Large Amphipods and other forms large enough to be conveniently handled with forceps were counted, and smaller forms were estimated by eye. The results are tabulated below, the figures for each lake representing the number of Groups in which each animal was found. Lilypad Phililloo Watson May 22, 1958 May 28, 1958 June 7, 1958 Hirudinea 2 2 1 Gastropoda Lymnaea 6 2 1 Physa 6 - -Planorbidae 5 3 1 Pelecypoda 4 3 _ Crustacea Gammarus 6 6 5 Hyalella (?) 5 6 6 Cladocera Not counted 3 6 Copepoda 2 Not counted Hydracarina - 5 3 Inseeta Ephemeroptera larvae _ 1 1 Odonata Dragonfly larvae 2 2 1 Damselfly larvae 2 2 4 Hemiptera Corixidae 1 - 2 Trichoptera (cases) 2 2 4 Diptera Chironomidae larvae 3 6 6 Culicidae Chaoborus larvae - 1 2 Mosquito pupae - - 2 Unidentified, larvae 1 on - 112 -Appendix III. Climatological Data for Selected Stations. The Sources from which these data were collected include the following? British Columbia, Dept. of Agriculture, "Climate of British Columbia, Tables of temperature, precipitation, and sunshine", publ, annually. United States Dept. of Commerce, Weather Bureau, "Monthly climatic data for the world", publ. monthly since 1948. "Climatological data. National summary", publ. monthly since 1950. "Climatological data. Annual summary", publ. annually. Canada Dept. of Transport, Meteorological Branch, "Monthly record, meteorological observations in Canada", publ. monthly. W.G. Kendrew and D. Kerr (1955), "The climate of British Columbia and the Yukon Territory", Queen's Printer, Ottawa. W.G. Kendrew and B.W. Currie (1955), "The climate of central Canada", Queen's Printer, Ottawa. Por each of the 46 stations listed, the following data are given: latitude-longitude "coordinates" (to nearest 10 ), elevation (feet); fir s t and last months each year in which the daily mean temperature for the month exceeds 32°F, daily mean maximum and minimum temperatures for June, July, and August, mean number of days annually with daily mean temperature greater than 42°P (calculated from monthly means except in 8 cases); mean precipitation of a l l forms (inches) for the year, for the months of June, July, and August, and for these three summer months combined, and the summer precipitation as a percentage of the annual total; the climax tree species characteristic of the region (although doubtless many stations are situated in edaphic climaxes or serai stages with different trees); the status of the region as a breeding area for Bufflehead (major, minor, acc idental, absent). It is hoped that these materials may be useful for other studies, in the future. Locality Lat—Long Coordinates Elev. 1st Month Daily Mean over 32° Daily Mean Max. Temperature June July & Min. s Aug, Last Month Daily Mean over 32° No, Days Annually Daily Mean over 42° 1. Madeline, Calif. 410-1203 5270 Mar.(35) 65-45 72-52 73-53 Nov.(37) (ca. 195) 2. Denver, Colo. 395-1045 5332 Mar,(39) 80-53 85-59 85-58 Nov.(41) (ca. 220) 3, Great Falls, Mont. 473-1112 3657 Mar,(34) 74-51 83-55 81-54 Nov.(36) (ca. 195) 4, Carmi, B.C. 493-1191 4084 Apr.(40) 64-42 73-47 72-46 Oct.(41) 175 5. Chute Lake, B.C. 494-1193 3916 Apr.(36) 65-39 73-41 70-43 Oct.(39) (ca. 155) 6. Cranbrook, B.C. 493-1155 3013 Mar,(33) 71-44 82-47 80-45 Oct.(43) (ca. 180) 7. Edgewood, B.C. 495-1181 1404 Mar.(33) 71-46 79-48 76-49 Nov,(34) (ca. 180) 80 Hemp Creek, B.C. 520-1201 2500 Apr,(41) 69-43 72-43 67-41 Oct,(38) (ca, 170) 9. Mamette Lake, B.C. 502-1205 3300 Apr.(38) 64-40 73-45 69-46 Oct.(39) (ca. 155) 10. Dog Creek, B.C. 514-1222 3370 Apr.(39) 66-44 72-47 71-46 Oct.(41) 163 11, Big Creek, B.C. 514-1230 3100 Apr,(38) 66-40 67-39 61-41 Oct.(39) (ca. 155) 12. Tatlayoko Lake, B.C. 514-1244 2700 Apr.(40) 65-41 71-44 69-45 Oct,(44) (ca. 175) 13. Kleena Kleene, B.C. 520-1250 2950 Apr,(37) 67-40 72-41 71-41 Oct.(39) (ca. 155) 14. New Hazelton, B.C. 552-1273 1150 Mar.(32) 68-42 70-48 68-49 Oct.(41) (ca. 175) 15, Vanderhoof, B.C. 540-1240 2093 Apr.(37) 66-40 69-45 69-44 Oct.(40) (ca. 160) 16. Dawson-Creek, B.C. 554-1201 2186 Apr.(33) 67-44 70-46 67-47 Oct.(38) (ca. 150) 17. Fort Nelson, B.C. 585-1224 1230 Apr,(36) 69-46 74-50 71-46 Oct.(36) (ca. 130) 18, Watson Lake, Yukon 601-1285 2248 May (46) 68-44 70-47 66-43 Oct.(34) 129 19, Whitehorse, Yukon Dawson City, Yukon 604-1351 2289 Apr.(32) 66-43 67-45 64-43 Oct.(34) 130) 20. 640-1393 1062 May (46) 70-43 73-47 68-42 Sep.(43) 133 21. Fairbanks, Alaska 645-1475 454 May (47) 67-41 72-49 68-43 Oct.(38) (ca. 140) 22 o Aklavik, N.W.T. 681-1345 25 June(49) 58-39 66-47 59-42 Sep,(38) 95 23. Norman Wells, N.W.T. 652-1265 209 May (42) 66-46 70-50 64-46 Sep.(42) 118 24, Fort Simpson, N.W.T. 615-1212 415 May (44) 68-45 74-50 69-46 Sep,(46) 133 25. Fort Smith, N.W.T. 600-1115 680 May (45) 69-42 74-48 69-44 Oct.(32) (ca. 130) 26. Lac la Biche, Alta. 545-1120 1835 Apr.(34) 66-48 72-54 66-52 Oct,(40) (ca. 160) 27. Beaverlodge, Alta. 551-1192 2500 Apr,(38) 67-44 72-47 70-45 Oct.(39) (ca. 145) 28. Edmonton, Alta. 533-1133 2158 Apr,(40) 69-45 74-50 72-47 Oct.(41) (ca. 170) 29. Vermillion, Alta, 532-1105 2037 Apr.(36) 69-42 74-49 67-50 Oct.(39) (ca. 160) 30. Calgary, Alta. 510-1141 3675 Apr.(40) 69-43 76-47 74-45 Oct.(42) (ca. 175) 31. Prince Albert, Sask, 531-1055 1432 Apr.(37) 71-46 76-51 73-48 Oct.(39) (ca. 170) 32. Regina, Sask. 503-1043 1884 Apr,(38) 73-47 79-51 77-48 Oct,(39) (ca. 175) 33. Swift Current, Sask. 502-1074 2440 Apr.-(41) 73-48 81-52 78-50 Oct.(42) (ca. 180) 34. Norway House, Man, 540-975 720 May (45) 67-45 74-53 70-50 Oct.(36) 149 35. Rivers, Man, 500-1002 1553 Apr.(38) 70-48 79-51 77-52 Oct.(43) (ca. 175) 36. Bismarck, N.Dak. 465-1004 1660 Apr.(43) 76-52 84-56 80-57 Oct,(45) (ca. 185) 37. Sault Ste. Marie, Mich.463-842 724 Apr,(37) 66-50 76-53 73-52 Nov.(33) (ca. 175) 38. Longlac, Ont. 495-863 1035 May (41) 72-48 75-50 70-49 Oct.(39) (ca. 145) 39. Kapuskasing, Ont. 492-823 752 May (46) 68-46 75-51 71-49 Oct.(40) (ca. 155) 40. North Bay, Ont. 462-792 1210 Apr.(38) 72-53 78-57 72-55 Oct.(47) (ca. 175) 41. Bagotville, P.Q. 482-710 536 Apr.(35) 69-50 72-56 72-53 Oct.(43) (ca, 160) 42. Goose Bay, Labrador 532-602 144 May (40) 62-40 69-53 66-50 Oct.(38) (ca. 130) 43. Chatham, N.B. 470-653 112 Apr.(37) 70-49 77-58 75-56 Nov.(34) (ca. 175) 44. Dartmouth, N.S. 444-633 136 Apr.(40) 67-49 73-57 73-56 Nov.(39) (ca. 195) 45. Columbus, Ohio 400-825 833 Mar.(41)* 82-59 88-63 86-60 Nov.(42)* (ca. 240) 46. Nashville, Tenn. 361-864 605 Mar,(50)* 86-67 90-69 90-67 Nov.(49)* over 270 * Months marked thus are not necessarily the f i r s t and. last with monthly means of daily mean temperatures greater than 42°P, but are listed here for comparison since no other stations given have such mean temperatures earlier than March or later than November. Locality Mean Precipitation Year June July (inches) Aug. Summer Percent in Summer Climax Vegetation Bufflehead Breeding Status 1 , Madeline, Calif. 1 7 , 8 1 . 3 1 . 3 0 . 5 3 . 1 (17) Ponderosa Pine Minor 2 , Denver, Colo. 1 4 , 3 1 . 3 1.8 1.5 4 . 7 ( 3 3 ) n I I Absent 3. Great Palls, Mont, 1 6 , 0 3,4 1.6 1.2 6 ,2 ( 39 ) t i it " ( ? ) 4. Carmi, B.C. 21.1 2 . 4 1.6 1 . 3 5.3 ( 2 5 ) D.Pir-E,Spruce " (?) 5 . Chute Lake, B.C. 2 4 . 4 2,1 1.4 1 . 3 4.8 ( 2 0 ) Douglas Fir " (?) 6. Cranbrook, B.C. 1 6 , 0 2 . 4 0 . 9 1.1 4.4 ( 28 ) D.Fir-P.Pine Minor 7 . Edgewood, B.C. 23.9 3 , 1 1.5 1.8 6 ,4 ( 27 ) Hemlock-Cedar it 8, Hemp Creek, B.C. 21.8 3.2 2,0 2.3 7.5 ( 34 ) tt i t » 9. Mamette Lake, B.C. 12,2 1 . 6 0 . 9 1.2 3.7 ( 3 0 ) Douglas Fir n ( ? ) 1 0 . Dog Creek, B.C. 1 5 , 6 2.9 1.4 2,1 6.4 ( 4 1 ) t ! It Major 1 1 . Big Creek, B.C. 1 2 . 4 1,8 1.4 1.6 4 , 8 ( 3 9 ) H I I Minor 1 2 . Tatlayoko Lake, B.C. 1 6 . 9 1.4 1 . 1 1.3 3,8 ( 23 ) II It " (?) 1 3 . Kleena Kleene, B.C. 1 4 . 0 2.3 1.1 1.8 5 . 2 ( 37 ) It II t i 14. New Hazel ton, B.C. 1 8 , 3 1,8 2.1 - 1.5 5.4 ( 30 ) Hemlock—Spruce Accidental 1 5 . Vanderhoof, B.C. 14,2 1.7 1.2 1 . 0 3 , 9 ( 27 ) Douglas F i r (?) Minor 1 6 . Dawaon Creek, B.C. 19.3 2,2 3,6 1.7 7 . 5 ( 39 ) White Spruce Major 1 7 . Port Nelson, B.C. 16.7 2.6 2.4 1.5 6,5 ( 39 ) Absent (?) 1 8 . Watson Lake, Yukon 16.7 2,0 1.9 1.9 5.8 (35) i i t i Minor 1 9 . Wiitehorse, Yukon 1 0 . 6 1.0 1.6 1.5 4.1 (39) II n t i 2 0 . Dawson City, Yukon 11,8 1.2 1,6 1.7 4 .5 (38) n n Absent (?) 21. Fairbanks, Alaska 1 1 . 9 1.4 2.0 2 . 0 5,4 ( 45 ) it i t Minor 2 2 , Aklavik, N.W.T. 9.2 0.8 1.4 1.4 3.6 ( 39 ) Tundra shrubs Absent 2 3 , Norman Wells, N.W.T. 11.4 1 , 4 2 . 0 2.6 6 . 0 ( 53 ) White Spruce Accidental 2 4 . Port Simpson, N.W.T. 1 3 , 0 1.5 2 , 0 1.5 5 . 0 ( 38 ) i i i t Minor 2 5 . Fort Smith, N.W.T. 12.1 1,5 2,0 1.6 5.1 ( 4 2 ) i t II i t 2 6 , Lac la Biche, Alta, 1 6 . 9 2 , 5 2.9 2.4 7,8 (46 ) n i t n 2 7 , Beaverlodge, Alta, 1 6 . 8 2,1 2.2 1.8 6,1 ( 3 6 ) i t n Major 2 8 . Edmonton, Alta, 1 8 . 0 3.2 3 . 3 2 . 4 8.9 ( 49 ) II i t t i 29, Vermillion, Alta. 1 6 . 4 2.6 3 , 0 2 , 9 8 . 5 ( 52 ) Aspen Groveland n 3 0 , Calgary, Alta, 1 6 , 4 3.2 2.5 2.3 8 , 0 (49 ) Short-grass Prairie Absent 3 1 . Prince Albert, Sask, 1 6 . 1 2.8 2,1 1 . 9 6.8 (42 ) Aspen Groveland Minor 32. Regina, Sask. 14.7 3,2 2.3 33. Swift Current, Sask. 14.4 2.9 2 o 2 34. Norway House, Man, 15.7 1.9. 2.2 35. Rivers, Man. X 2 # 1 1.3 1.1 36, Bismarck, N.Oak. 16.0 3,3 2.3 37. Sault Ste. Marie, Mich, 28.1 2.6 2.8 38, Longlac, Ont. 26.4 3.3 2.8 39. Kapuskasing, Ont, 27.8 2.6 3.3 40, North Bay, Ont, 39.2 3.4 3.9 41, Bagotville, P.Q. 38.7 4,6 4.2 42, Goose Bay, Labrador 29.3 2.5 3,5 43. Chatham, N.B, 40,2 3.9 2.8 44, Dartmouth, N.S. 54.3 4.1 3.6 45. Columbus, Ohio 34.6 3.5 3,4 46. Nashville, Tenn. 45,3 4,0 3,4 1.7 7.2 (49) Long-grass Prairie Absent 1.8 6.9 (48) n n ti I I 2.3 6,4 (41) White Spruce » (?) 0.9 3,3 (27) Aspen Groveland Minor 1.8 7.4 (46) Long-grass Prairie Absent 2.6 8.0 (28) White Spruce " (?) 2.9 9,0 (34) I I it Minor (?) 3.1 9,0 (32) I I it Absent (?) 3.5 10,8 (28) Beech-Maple-Hemlock n 3.6 12,4 (32) White Spruce I I 2.8 8.8 (30) I I it I I 3.2 9.9 (25) White and Black Spruce I I 4,5 12 « 2 (22) Red " *' " I I 3,2 10,1 (29) Beech—Maple n 3.9 11 e 3 (25) Oak-Hickory I I Some of the assignments of "climax vegetation" and of "Bufflehead breeding status" may be changed, when more accurate information becomes available on these topics. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items