HARVARD UNIVERSITY LIBRARY OF THE Museum of Comparative Zoology TRANSACTIONS OF THE LINNAEAN SOCIETY OF NEW YORK Volume IV. Studies in the Life History of the Song Sparrow I By Margaret Morse Nice lit, . '7*'- A •\ A ■ ‘ f - N y » ’ '' Courtcsv Hmi-Lorc Maga::inc PLATE 1— SONG SPARROW {Melospiza melodia) Studies in the Life History of the Song Sparro>x VOLUME I A Population Study of ihe Song Sparrow By Margaret Morse Nice d.lH- To ?ny Friend Ernst Mayr TABLE OF CONTENTS FRONTISPIECE Plate I PAGES INTRODUCTION ------------------ i_2 CHAPTER I The Song Sparrow as a Subject for Study - -- -- -- -- 3.7 A. Suitability as an Object for Study. B. Resume of the Life History. C. Technique of the Investigation. D. Summary. CHAPTER II The Son(; Sparrow and its Environment - -- -- -- -- - g-iy A. The Climate. Chart I. B. The Habitat. Map i. Plate II. (F'acing page 16.) C. Relations to Other Animal Species. 1. Invertebrates. 2. Reptiles. 3. Other Birds. 4. Mammals. D. Summary. CHAPTER HI Weights and AIeasurements - -- -- -- -- -- -- - 18-28 A. Measurements. Chart II. B. Weights. Chart HI. 1. Weights Throughout the Day. Table I. 2. Weights Throughout the Year. Table II. a. Weights in Winter. b. Weights in Spring. 3. Weight and Age. C. Summary. CHAPTER IV Migratory Status of the Song Sparrow on Interpont ----- 29-42 A. The Transients. B. The Winter Residents. C. Residents and Summer Residents. 1. “Individual Migration.” 2. Stability of Migrating and Non-migrating Behavior 11 3- Inheritance of Migratory Behavior. Charts IV and V. 4. Numbers of Residents and Summer Residents. 5. Differences Between Residents and Summer Residents. 6. Comparison With Other Species. 7. Discussion of “Individual Migration” in the Song Sparrows. D. Summary. CHAPTER V Spring and Fall Migration - -- -- -- -- -- -- -- A. Spring Migration. 1. Migration in Relation to Temperature and Time of Year. Table III. Charts VI and VII. a. Discussion of Some Contrary Theories. 2. Migration of Individual Males. Table IV. 3. Migration of Individual Females. Table V. 4. Migration in Relation to Sex and -Age. B. Fall Migration. C. Summary. CHAPTER VI Territory Establishment - -- -- -- -- -- -- - A. The Establishment of Territory. B. Territory and the Development of Song. C. Summary CHAPTER VII Territory Throughout the Year - -- -- -- -- -- - A. Territory in the Fall. 1. Singing in the Fall. 2, Taking up of Territories. B. Behavior in Winter. C. Behavior in Spring. 1. Song and Temperature. Table VT. Charts VHI, IX, X. 2. Defense of the Territory. D. Summary. CHAPTER VHI The Territories From Year to Year - -- -- -- -- -- A. The Territories of the Adult Males. Maps 2, 3, 4, 5, 6, 7. B. The Returns of the Females. Map 8. C. Territories of the Males Banded in the Nest. Map 9. D. Territories of the Females Banded in the Nest. Map 10. E. Territories of Song Sparrows Banded in Our Garden. F. Some Family Histories. Maps ii, 12, 13, 14. G. Summary. PAGES 43-56 57-60 61-69 70-83 iii CHAPTER IX The Relations Between the Sexes - -- -- -- -- -- - A. The Situation During One Season. 1. Desertions. 2. Is There a Reserve Supply of Unmated Birds? B. The Situation from Year to Year. C. Bigamy. D. Sexual Selection. E. Summary CHAPTER X The Nests of the Song Sparrow - -- -- -- -- -- -- A. Position of the Nests. Plate III. (Facing page 32.) B. Security. C. Building Technique. D. Building of Old and Young Birds. E. Summary. CHAPTER XI The Start of Laying - -- -- -- -- -- -- -- -- A. What Factors Influence the Start of Laying? B. The Start of Laying in Relation to Temperature. Table VH. Charts XI, XII, XHI, XIV. C. The Start of Laying and Other Factors. D. The Start of Laying of Other Species. E. Dates of Laying of Individual Females. F. Summary. CHAPTER XII The Eggs of the Song Sparrow - -- -- -- -- -- -- A. The Number of Eggs in a Set. B. Time of Replacement of a Destroyed Set. C. The Color of the Eggs. D. The Size of the Eggs. Table VIII. 1. Average Size of Eggs in Relation to Various Factors. 2. Weights of Sets. E. Inheritance of Color, Size and Shape. F. Summary. CHAPTER XHI Incubation - -- -- -- -- -- -- -- -- -- - A. Role of the Female. 1. Length of Incubation. 2. The Rhythm of Incubation. Table IX. B. Role of the Male. Table X. C. Summary. pages 84-91 92-96 97-107 108-121 122- 129 IV CHAPTER XIV Care of the Young - -- -- -- -- -- A. Care of the Young in the Nest. Table XI. B. Intervals Between Broods. Table XII. C. Summary. CHAPTER XV Nesting Success and Failure - -- -- -- -- -- -- - A. The Number of Young Fledged Per Pair in One Season. B. Size of Sets and Size of Broods. Tables XHI, XIV. Chart XV. C. Completely and Partially Successful Broods. Table XV. D. Numbers of Young Raised in 21 1 Nests in Six Seasons. Table XVI. I. Comparison with Other Studies of Nest Success. Table XVII. E. Analysis of the Loss of Eggs and Young. Tables XVIII, XIX. Chart XVI. F. Summary. CHAPTER XVI The Cowbird in Relation to the Song Sparrow - -- -- -- - A. The Cowbird as Parasite. 1. Non-specialization of the Cowbird. a. The Incubation Period of the Cowbird. 2. Relations to its Own Species. 3. Relation to its Breeding Area. 4. Relation to its Hosts. a. The Eggs of the Cowbird. b. The Nestling Cowbird. c. Destruction of Eggs of the Host. B. The Song Sparrow as Host. C. The Effect of the Cowbird on the Song Sparrow. Tables XX, XXL I. The Incidence of Cowbird Parasitism on Interpont. D. The Success of the Cowbird on Interpont. E. Summary. CHAPTER XVII Survival of the Adults - -- -- -- -- -- -- -- - A. Survival of the Adult Males. Table XXH. Chart XVTI. B. Survival of the Adult Females. Table XXIH. C. Losses and Replacements in the Population. Table XXIV. D. Proportion of First-Year Birds in the Population. Table XXV. E. Summary. pages I 30- I 33 134-151 152-165 166-179 V CHAPTER XVIII Survival of thi: Young - -- -- -- - - -- -- -- - A. Return of the Fledged Young. 1. The N^umber of Nestlings that Returned. Table XXVI. a. The Sex Ratio of the Returned Nestlings. 2. The Distances from the Birth Place at which the Young- Birds Settled. 3. Do Young Return to their Birth Place? B. Survival of the Fledged Young. I. What Percentage of Fledged Young Should Survive till the Following Spring in order to Maintain the Population? Table XXVII. C. Summary. CHAPTER XIX Age Attained bv Song Sparrows - -- -- -- -- -- -- A. The Average .Age of Song Sparrows. B. Potential Age. C. Age Composition of a Song Sparrow Population. Tables XXVHI, XXIX. D. Mortality Factors. E. Summary. CHAPTER XX Some Population Problems - -- -- -- -- -- -- -- A. Factors Influencing the Size of a Song Sparrow Population. B. The Course of Events on Interpont. Table XXXI. C. Comparison with Other Studies. Chart XVHI. D. Some Theoretical Considerations on Population Problems. E. Summary. Summary - - - - - - - - - - - - - - - - - - - - - A. Response of the Song Sparrow to Climatic Influences. B. Relations to the Flora and Fauna of the Habitat. C. Territory. D. Relations between the Pair. E. Some Physiological Responses. PA Population Problems. PAGES 180-189 190-198 199-208 209-212 VI APPENDIX J PACES The Technique of the Investigation - -- -- -- -- -- 213-220 A. The Course of the Study. B. Trapping the Birds. C. Banding Methods. D. Nomenclature and Records. 1. The Banding Record. 2. Card Catalog. 3. Key Tables. 4. Daily Records. 5. Field Note Books. 6. Maps. 7. Nest Record. E. Plan of Work. APPENDIX II Some Statistics Concerning Song Sparrow Banding on Interpont - 220-221 A. Total Song Sparrow Population Given M and K Numbers. B. Returns of Birds Banded. 1928-1935. APPENDIX III Nesting Censuses on Upper Interpont ----------- 221-223 APPENDIX IV Further Data on Cowbirds : Returns and Weights ------ 223-224 A. Returns. B. Weights. 1. Weights of Young Raised by Song Sparrows. 2. Weights of Adults in Grams. APPENDIX V Meteorological Data for Columbus. Tables XXXII, XXXIII - - - 224-225 BIBLIOGRAPHY ------------------ 226-239 INDEX OF SUBJECTS ---------------- 240-243 INDEX OF SPECIES ---------------- 244-247 INTRODUCTION For the past eight years the writer has concentrated on the study of the life history of one species of bird. The technique was based on recognition of the individual in the field by means of colored bands, and on repeated censuses over Interpont to check the status of the community. The opportunity to examine at intervals the birds in the hand was an important feature of the work. The method has been almost entirely that of observation with a minimum of experimentation and no collecting, the hope being to find out what actually happens in a population of wild birds. The first year was devoted to intensive study of two pairs, an indis- pensable foundation for the later work, which at one time included observation of 75 banded males. Through trapping, banding, and continued search, individuals were traced throughout their lives, and family histories established, the place of residence of relatives determined, and the inheritance of migratory behavior, song characteristics, egg color, etc., investi- gated. The present volume is concerned with the population aspects of the study, leaving more detailed treatment for a second volume. Volume I deals with the Song Sparrow and its environment, its ecology, migration, territory, and reproduction, all from a somewhat statistical point of view, and finally with survival problems. Volume II will deal with the behavior of the Song Sparrow, including de- tailed observations on the technique of territory establishment, “court- ship,” song, and so on. Eight years' concentration on one species has brought results of undoubted value. Yet no one can be more aware of the deficiencies of the study than is the author — the meagerness of data on various points, the failure to find certain important nests, the uncertainty as to the exact course of events with particular pairs, and many other unfortunate gaps. The explanation lies in the difficulty and complex- ity of the problem and in the fact that it was undertaken by one person alone. Grateful acknowledgements are due to Mr. Edward S. Thomas for the loan of three photographs, and to Dr. Selig Hecht for plotting the curves of the temperature thresholds for the start of singing and laying. I am much indebted to Dr. Erwin Stresemann for asking me to write a report of my study in the winter of 1932-33 for the Journal fur Ornithologie; the preparation of this paper brought many problems to light and enabled me to work more intelligently during the last three years. Thanks are especially due to Dr. Lawrence E. Hicks and Dr. Paul L. Errington, and most of all, to Dr. Ernst Mayr and Mr. William Vogt, for their kindness in reading the present manu- script and for their helpful criticisms. The attempt has been made — by division into headings, by full summaries after each chapter, and by indices — to present the material in as clear and orderly a manner as possible. But a word of warn- ing to the reader may not be amiss. As the study proved to be a complicated one, so some of the tables may seem inconsistent, in- volving as they often do somewhat different sets of birds, for in- stance, the breeders on Central or Upper Interpont are not entirely the same as the total handed breeders on and near Interpont. I have tried to explain each table adequately, and in case of apparent con- tradictions, I hope the reader will study the captions and text care- fully before concluding that there are errors. In the summaries after each chapter, it must be remembered that their application is narrow, referring to the Song Sparrows that I studied and not to birds in general. Finally it must be kept in mind that “Upper Interpont” is merely a shorter way of saying “Central and North Interpont,” not a third area. It is my earnest hope that this work on the Song Sparrow will stimulate others to study intensively the biology of our common birds. 3 CHAPTER I The Song Sparrow as a Subject for Study Melospiza melodia “is a bird of very extensive geographic range, breeding throughout the temperate parts of the North American con- tinent including the plateau of Mexico. No other bird of the Nearctic Region has proven so sensitive to influences of physical environment, and as a result of this plasticity of organization it has become divided into a large number of geographic forms, some of extensive, others of very circumscribed range, the area of distribution in every case coinciding strictly with uniformity or continuity of physical conditions’’ (Ridgway, i6o). Some of the races are migratory, while others are resident. The Mississippi Song Sparrow {Melospiza melodia beata*) is the subspecies that nests throughout Ohio. A. Suitability as an Object for Study Although the Song Sparrow is an abundant, widely distributed, friendly and attractive bird, yet it has been almost wholly neglected as a subject for life history studies. Nevertheless, it has proved eminently suited for such investigations. Here on Interpont, there is a large population right at my door; no time is wasted in going to and from the field of study and I am able to keep track of my subjects all the time except when I am absent from Columbus during a portion of each summer. This pop- ulation shows much stability, little tendency to scatter, and a high proportion of returns of the young. The birds are much attached to their homes, so the pair can always be found on its territory during the breeding season, and resident birds remain in the near vicinity throughout the year. As a rule the birds can be trapped without too much trouble and are not disturbed by the experience. The bands can be seen fairly easily, except in the coldest weather when the feathers are so fluffed out that the legs are concealed. The parent birds will endure visits to their nests daily or even oftener without desertion. And finally, a point that makes the Song Sparrow of unique interest and value, the male shows great individuality in his songs. ♦Dr. A. Wetmore has renamed this subspecies euphonia (A New Race of the Song Sparrow from the Appalachian Region. 1936. Smithsonian Misc. Coll. 95 (17) 4 As to disadvantages, the first is the uncertainty of locating nests unless a great deal of time is spent with each pair. Persons studying hole-nesting birds, such as Tree Swallows, Starlings, and Titmice in Europe, have a great advantage here. The other difficulty is that both sexes and all ages after the post-juvenal molt are alike in plum- age. Male and female can usually be distinguished by measurement of the wing. In the breeding season, the sexes can be distinguished by their behavior, but this is not true in fall or winter. The juvenal males, when residents, can be known in January or early February by the character of their songs, but with juvenal summer residents this can be done but rarely. When I was in Berlin in July, 1932, Dr. Stresemann told me that juvenals of some species have pointed tail feathers and adult birds rounded tail feathers. After that I noted this character, finding in the fall that most of the males were easily classifiable into one category or the other, but in winter the difference is less pronounced, and in spring little reliance can be placed on it. All birds classified as juvenals by this method, in the fall, and later heard singing, corroborated my judgment by the warbling character of the song. As to the males considered adult, I made only one wrong diagnosis of those birds whose age could be checked by their manner of singing. Unfortunately the shape of the tail feathers is of no help in estimating the age of the females as birds of all ages have more or less pointed rectrices. During the nesting season the presence of the incubation patch in the female and its absence in the male is a sure criterion of sex with a captured bird. B. Resume of the Life History The Song Sparrow on Interpont is a strongly territorial bird from the time of taking up territory in the late winter or early spring to the end of nesting, but territory is not held during the molt, very little in fall, and not at all in winter. About half the nesting males are permanent residents ; the rest migrate south in October and return from late February through iMarch, but only about one-fifth of the females are residents. The resident males start to sing in late January or in February according to the weather; they sing almost constantly until joined 5 by a mate, when singing abruptly drops to almost zero. Territory is defended and acquired by a special ceremony that includes song, posture and fighting. The male “courts” or, better, dominates his mate by “pouncing” on her — i.e., suddenly flying down, hitting her, and flying away with a loud song. Although the male carries nesting material during preliminary nest-hunting stages, all the work on the real nest is done by the female. The male at this time starts to sing again, singing to quite an extent while his mate incubates. The female incubates for approximately 20-30 minutes at a time, stay- ing away from the nest for about 8 minutes. The eggs hatch in 12 to 13 days as a rule, and the young usually stay in the nest 10 days. The role of the male is that of guardian of his territory, mate, eggs and young; he feeds the last from the time they hatch, and takes the major responsibility for them soon after they have left the nest, when his mate is normally busy with a new nest. The young become independent at the age of 28 to 31 days. Song Sparrows in this region regularly make three to four attempts at nesting, some of them raising three broods. The food of the Song Sparrow consists largely of weed seeds and insects, 66 per cent of the contents of 401 stomachs examined by Judd, Qi, being vegetable matter, 34 per cent animal matter. Berries are eaten to some extent, as are spiders, snails and millipeds. From May to August insects “compose more than half the food.” C. The: Technique of the Investigation An account of the course of the study, of methods of trapping and banding the birds, an explanation of the nomenclature used, and a description of the system of record keeping, are given in Ap- pendix I. The most essential points for the understanding of the text will be given here. The spring and summer of 1929 were spent on intensive ob- servation of two pairs of birds — iM and K2, 4M and Ky. The next year the study extended over most of Central Interpont, while 6 beginning in 1931 North Interpont was also included. During the breeding season of 1932^ all the males on Upper Interpont — 69 in number — and most of the females were banded; while during the next two years all but two of the males were banded. During 1935 this was true of only 14 of the 25 males present, and in 1936 of 9 out of 18. Most of the birds had to be trapped on their territories. All the adults were given colored celluloid bands, the aluminum band always being placed on the left leg, while with nestlings it was put on the right leg. Individual “field numbers,” having no relation to the band numbers, were given to all the nesting adults, iM, 2M, etc., for the males, Ki, K2, etc., for the females. The plan of the work consisted in repeated censuses over Inter- pont (and on occasion over surrounding territory), in order to keep a careful check on the personnel of the population. D. Summary 1. The Song Sparrow ofifered a favorable object of study be- cause of its availability and abundance, and the individuality in song of the males; the chief disadvantage proved to be the difficulty of locating nests. 2. Sex can usually be distinguished by wing measurement, and in the breeding season by the presence or absence of the incubation patch. 3. Whether males are adult or young can be told in the fall from the shape of the tail feathers, and young resident males by the char- acter of their singing. 4. Male Song Sparrows are strongly territorial during the breed- ing season. 5. Part of the breeding population is permanently resident, the rest migratory. 6. Although both sexes are alike in plumage, the male takes no part in building the nest, incubating the eggs, or brooding the young. 7 7- The male defends his territory, mate, nest and young, and often does the major part in feeding the last. 8. The food of the Song Sparrow consists of two-thirds vege- table matter and one-third animal matter. 9. The main features of the technique of the study consisted in trapping the subjects on their territories, in banding them with colored as well as aluminum bands, and in keeping track of them by repeated censuses. The first season was spent in intensive study of two pairs, the six subsequent years in extensive work. 8 CHAPTER II The Song Sparrow and Its Environment Certain fundamental environmental factors that alYect the Song Sparro’sv in this region will be discussed in the present chapter : temper- ature, precipitation, and sunshine, and also the flora and fauna of the habitat. A. The Climate Columbus is situated near the center of Ohio, latitude 40 degrees north, and longitude 83 degrees west. The elevation of Interpont is about 220 m. 9 The average monthly temperature and precipitation are shown in the climograph in Chart I. According to records of the United States Weather Bureau at Columbus, the annual mean temperature during the last 55 years has ranged from 9.4° C. (49° F.) in 1917 to 13° C. (55.4° F.) in 1931, averaging 11.2° C. (52.3°F.) The average mean temperature by months follows: January — 1.6° C. (29° F.) ; February — 0.8° C. (30.6° F.) ; March 4.4° C. (40° F. ) ; April 10.6° C. (51.1° F.) ; May 16.6° C. (62° F.) ; June 21.5° C. (70.6° F.) ; July 23.9° C. (75° F.) ; August 22.6° C. (72.6° F.) ; September 19.3° C. (66.8°F.) ; October 12.7° C. (54.9° F.) ; November 5.6° C. (42.1° F.) ; December 0.2° C. (32.4° F.). The absolute highest temperature reached has been 41° C. (106° F.) in 1934 and the absolute lowest — 28.9° C. ( — 20° F.) in 1884 and 1889. During the period covered by my study, every year but 1929 and 1935 was warmer than the average, 1931 having the highest tempera- ture of any year of record, largely because of unusually mild weather from September through December. The absolute highest tempera- tures ranged from 33.3° C. (92° F.) in July 1929 to 41° C. (106° F.) in July 1934. The absolute lowest ranged from — 9.4° C. (13° F.) in January 1931 to — 26.7° C. ( — 16° F.) in January 1936. Annual precipitation has ranged from 548 mm. (21.6 inches) in 1930 to 1,301 mm. (51.3 inches) in 1882, the average being 919 mm. (36.19 inches). During the period of study only one year — 1929 — surpassed the average, with a total of 1,074 mm. (42.27 inches). Precipitation in 1931 and 1935 almost reached the average, in 1933 it was 10 cm. short, in 1932 15 cm. short, while in 1934 there were only 560 mm., and 1930 was the driest year of record. The average of the seven years was 794 mm. (31.26 inches). The average amount of snowfall per year is 609 mm. (24 inches). The number of hours of daylight ranges from 9 hours 19 minutes in December to 15 hours i minute in June. During January the days lengthen about minutes per day, but for the next three months about 2I/2 minutes per day, slowing down in May to less than a minute a day. The amount of sunshine averages 54 per cent of the possible total. It ranges from a minimum of 33 per cent in December, in- 10 creasing each month and reaching a maximum of 70 per cent in July- Further details as to the temperature, precipitation, and amount of sunshine in Columbus will be found in Appendix V. B. The: Habitat The eastern flood plain of the Olentangy River between Dod- ridge Street and Lane Avenue Bridges I have named Interpont. Central Interpont, about 30 acres (12 ha.) in extent, has been the main fleld of my studies, but North Interpont, about 10 acres (4 ha.) has also been of much importance. These two tracts taken together I call Upper Interpont. South Interpont, being mostly occupied by a city playground, has less to offer the Song Sparrows; only about 10 acres of land are suit- able for their purposes and I have done little work there. Across the Olentangy above and below Interpont the river was bordered by country suitable for Song Sparrows; in some places there was only a narrow strip of such land, but in others it amounted to a width of some 300 m. Upper Interpont for the most part was waste land flooded periodi- cally and little used for purposes of cultivation. But in the late sum- mer of 1932 a squad of unemployed laborers “cleaned-up” all the cover but the trees along the river bank, while in the following spring the bulk of the land was converted into gardens, the destruction extending even to most of the dikes in the following year. In 1935 the dikes and river banks were left undisturbed. When I started the study, Interpont was largely covered with a rank growth of bottom land weed association, blue grass {Poa pratensis) being present in some places. There were large patches of elderberry (Sambuciis canadensis) and, near the river and along the bluff to the east, a number of trees. The most important weeds were: wild rye (Elymus canadensis), sweet clover {MeliloHis alba), smartweed (Persicaria) , tick-tre-foil (Meibomia) , Indian-cup (Silphium per- foliatum), cow-parsnip (Heraclinni lanatum), giant ragweed (Ambrosia trifida), burdock (Arctium minus), beggar-ticks (Bidens), teasel (Dipsacus), dandelion (Leontodon taraxacum), thistle (Cirsium), sunflowers (Helianthus) , golden rod (Solidago) and asters (Aster). The most abundant trees are the cottonwood (Populus deltoides), American elm (Ulmus americanus) , buckeye (Aesculus glabra), hackberry (Celtis occi- II Map I. Interpont in the Spring of 1932 12 dentalis), silver maple (Acer saccharinum) , red maple (Acer rubriim), box elder (Aeer negundo), sycamore (Platamis occidentalis) , and willows (Salix). Poison ivy (Toxicodendron radieans) and Virginia creeper (Parthenocissus qninquefolia) are common. The trees of¥er singing posts and look-outs for the male Song Sparrows, while the same is true of shrubs and large weeds. The shrubs, grasses and weeds provide the birds with food, protection and nesting sites. Interpont is not typical Song Sparrow country in that there is no permanent water supply except the river, the majority of the terri- tories offering no water whatsoever. The birds must leave their terri- tories several times a day to procure water for drinking and bathing purposes. C. Relations to Other Animal Species From 1930 to 1932 the Song Sparrow was the most abundant avian species on Interpont, but now first place goes to the Robin {Tiirdiis in. migratorius) . Yet among the birds nesting in the weeds and shrubbery Mclospiza still holds first place. Interpont used to sup- port a rich and varied bird life. In 1931 30 species and about 220 pairs, an average of 5.5 pairs ])er acre (14 per ha.), nested on the 40 acres, but in 1935 there were only 25 species and about 150 pairs, or 3.75 pairs per acre (9 per ha.). Nesting censuses for Interpont for four years are given in Appendix III. According to the censuses taken for the U. S. Biological Survey the number of nesting birds on farm land in northeastern United States averages i.i pairs per acre (2.8 per hectare), 41, 44. The relation of the Song Sparrow with the other animals on Inter- pont will be briefly touched upon. I. Invertebrates The role of the Song Sparrow in relation to insects and various other invertebrates is largely that of predator (see Judd, go, gi). As for invertebrates that prey upon the Song Sparrow, this species in this region appears to be comparatively free from parasites, the only one that I was able to procure being a Hippoboscid fly — Ornithomyia anchinenria. Peters, 146, reports the following external parasites as taken from Song Sparrows: 6 species of lice (Mallophaga) — Degeeriella vulgata, MachcerilcEmns mcBstiim, Menacanthus chrysophcuiim, Myrsidea incerta, Philopterus subflavescens, Ricinus melospizcB \ 2 bird-flies (Hippohoscidcz) — Ornithoica conjhienta, Ornith- otnyia anchineuria; 4 mites {Analgesidcs) — Analgopsis sp., Liponyssus sylviarum,. Trombicula bisignata, Trombicula cavicola; and 3 ticks (Ixodoidea) — Hczma- pliysalis leporis-palustris, Ixodes briinneus, Ixodes sp. During the spring of 1935 I collected all the Song Sparrow nests after the young had left and gave them to Mr. E. S. Thomas of the Ohio State Museum for examination for Protocalliphora. Negative results were obtained from all but one nest (Thomas, 188). This nest had had five young that I had weighed daily till the age of 7 days, four leaving the nest when ten days old ; during the first four days their weights were less than average, but after that they compared well with other nestlings of like age. In 1936, out of seven nests that raised young, two were found to be infested. Each of these nests had about the same number of pupae — 9 to 10 — yet all the birds left the nest at the average age. Manwell and Herman, 119, 119a, found by blood smears and inoculations that 22 out of 62 Song Sparrows at Syracuse, New York, were affected by one or more species of malaria parasite. 2. Reptiles There are many garter snakes (Thamnophis s. sirtalis) on Inter- pont and they may take toll of the eggs and young of the Song Sparrows. Dr. H. K. Gloyd has called my attention to the fact that this species is supposed to feed entirely on cold-blooded prey. Yet one in- stance {22a) has been reported from Iowa where a nestling Yellow Warbler (Dendroica a. cestiva) was taken by a 12 inch garter snake identified by A. G. Ruthven as T. parietalis. Gabrielson, 62, reports this species as swallowing the eggs of a Bobolink {Dolichonyx oryzivorus), and Ruthven, i6ya, states that it has been observed eating fledgling birds. 3. Other Birds Other birds, as they affect an individual Song Sparrow on Inter- pont, might be divided into four categories ; its own species ; more or less neutral species ; predatory species ; and the brood parasite. The Song Sparrow, by intra-specific hostility in spring and sum- mer, ensures a spacing of pairs, thus eliminating competition for food. 14 and interference in family affairs. In the fall and winter hostility is much diminished, but there never is pronounced gregariousness with my birds. As to the largest category of birds, all those species not predatory nor parasitic on the Song Sparrow, I believe there is little competition between them and Melospiza either for food or nesting sites, because there always appears to be an abundance of both, and also because the habitat niches of the different species are somewhat different. It is true the Song Sparrows during the nesting season are somewhat hostile to most other species not too large or indifferent (see Chapter VII under Defense of Territory), but I believe that this comes from an hypertrophy of the territorial instinct. It does not prevent the other species from nesting side by side with the Song Sparrows. I do not believe the abundance of Melospiza on Interpont has been prejudicial to the abundance of any other species. ]\Iany of these birds, on the other hand, may be an aid to the Song Sparrow by attracting some of the attentions of the Cowbird to them- selves. This seems to be especially true of the Northern Yellow-throat {Geothlypis trichas brachidactyla) and probably also of the Indigo Bunting {Passcrina cyaiiea). As to predators, one pair of Sparrow Hawks (Falco s. sparverius) nested until 1933 on North Interpont and another just across the river from Central Interpont, but they appear not to hunt in this region. In fall and winter Sparrow Hawks sometimes try to catch Song Sparrows, but I have never seen one succeed. The Song Sparrows have almost no fear of these Falcons. The Sharp-shinned and Cooper’s Hawks (Accipiter v. velox and A. coo peri), on the contrary, are greatly dreaded by all the small birds ; they undoubtedly get some of the Song Sparrows on their occasional visits in fall and winter. The Screech Owl {Otns asio naevius) is a resident in the region and may well be responsible for the disappearance of some of the nesting adults, espe- cially recently when cover was inadequate for protection. Blue Jays (Cyanocitta c. cristata) and Bronzed Crackles (Quis- calus qiiiscula aenens) may take some of the eggs and young. I once surprised a Ring-necked Pheasant (Phasianus colchicus torquatus) just after she had emptied a nest of two-day old nestlings, and I sus- 15 pected it was she that threw three four-day-old nestlings out of an- other nest. 4. Mammals The relations of the Song Sparrow to some of its mammalian neighbors are neutral — notably the cottontail rabbit (Sylvilagus flori- danus mearnsi), but it is a far different matter with others. We will consider the native predators, the introduced predators, and man. Unfortunately I have seldom been able to get good evidence of the destruction of nests by any particular predator. The list of native mammals that might prey on the Song Sparrows is not long, and there are few representatives of each ; the opossum (Didelphis v. virginiana), weasel {Mustela n. noveboracensis) , skunk {Mephitis nigra), red squirrel (Sciurus hudsonicus loqnax), and chipmunk (Tamias striatus fisheri). The three introduced predators are far more abundant and all, I believe, are much more inimical to the Song Sparrows than the native mammals; these are self-hunting dogs, the Norway rats that frequent the dumps at each end of Interpont, and most destructive of all, cats. The influence of man has many ramifications : the clearing of the land, at first beneficial to the species, later disastrous; the ploughing of occupied territories ; the introduction of new enemies — cat, rat, dog and Pheasant ; the activities of boys ; and finally, for this study, myself. Interpont was undoubtedly a much better habitat for Melospiza melodia in 1932 than it had been when covered with forest. But the cultivation of the land since then and the entirely unnecessary de- struction of cover on the dikes, in the ditches and along the river bank have wrought havoc with the area as a home for ground- nesting birds. Many of the Song Sparrows that come into Interpont during the nesting season — both males taking up territories and females joining males that have lost mates — have probably lost their homes in neighboring regions as a result of the activities of man. Melospiza melodia is an adaptable bird and will utilize, as its home, sites on the bluff in South Interpont and further down the river that are nothing but masses of tin cans and weeds. i6 The introduction of the House Sparrow (Passer domesticus) into this country has been a calamity to our native birds in many ways, not the least of which has been the prejudice it has cast on our native sparrows, in consequence of which boys think it a meritorious deed to shoot any small brownish bird. Since Interpont is within the city limits, it is illegal to use any rifle or shot gun here ; yet, in spite of all my efforts to educate the boys, telling them of the laws, and trying to interest them in the birds, they continued shooting the Song Sparrows. Finally in desperation, I procured a commission as Special Game Protector of the State of Ohio and then I found that my words of warning, backed up by the shining badge, were listened to with re- spect, and the shooting largely stopped, at least while I was in Columbus. As for myself, I have tried not to interfere with the course of events, not removing Cowbird eggs (except in 1934), nor killing natural enemies — i.e., native animals. I fear that dogs followed my tracks to at least two nests and destroyed them, but on the other hand, I saved various nests from destruction and moved threatened young to other nests when Interpont was plowed in June 1933. The nests found by me suffer fewer disasters on the average than those nests I do not find, as judged by the re-nesting of the birds. On the whole I feel that my activities are somewhat beneficial to the Song Sparrows. D. Summary 1. The annual mean temperature at Columbus, Ohio, averages 11.2° C. ; the annual precipitation averages 919 mm. (Chart I). 2. The period covered by the study showed a small excess of temperature and marked deficit in precipitation. 3. From 1928 till the spring of 1933 Upper Interpont was largely waste land supporting a rank growth of weeds, many shrubs and some trees. 4. Fifteen species of external parasites of the Song Sparrow have been reported. 5. The Song Sparrow’s relation to most of the birds about its size and smaller is one of hostility during the nesting season, although Plate II. PHOTO BV E. S. THOMAS Song Sparrow near Columbus, Typical Habitat on Interpont . ».} • •V f. . w * . .‘*^ ■ MM''*'- ,; tm'" -■. w V-' ' ■ '/' r V .V' ■ ’ ■• y. y r ' V' vi^' -V' 7.V| " ^V'«- L^- r.'- ✓ , -r' ' ‘r.S‘>‘ >V. >* > ,;^;^ .1^? -i* I? these birds probably do not compete seriously with it for food, and some of them relieve it of part of the burden of Cowbird parasitism. 6. Seven avian predators are mentioned, besides five native mam- malian predators, and three that have been introduced. The cat and Norway rat are considered the worst enemies of the Song Sparrows. y. The influence of man on the Song Sparrow is a complicated matter, some of it being beneficial, but much deleterious. NUMBER OF BIRDS i8 CHAPTER III Weights and Measurements The measurement of birds has been one of the foundation stones of systematic ornithology, but the matter of weights of birds has been much neglected. Yet the biological significance of the latter subject far surpasses that of the former. The taking of measurements is particularly suited to bird skins, but for most of the problems con- nected with weight it is essential to deal with live birds. A. AIeasurkments Ridgway’s method, i6o, was followed for the wing measurement with one point of the divider “resting against the anterior side of the bend, the other touching the extremity of the longest primary.” Tail 58 59 60 61 62 63 64 65 66 67 68 69 LENGTH OF WING IN MILLIMETERS Chart II. Wing Measurements of Breeding Song Sparrows. 137 Males; 123 Females 19 measurements proved so variable that little attention was paid to them, but the wing measurement became an important index of sex. The wing measurements of 90 resident and 47 summer resident males ranged between 62 and 69 mm., averaging 66.3, the residents averaging 66.3, the others 66.2. Tail measurements ranged between 62 and 72 mm., the median being 68. The wing measurements of 32 resident females and 91 summer residents ranged between 58 and 66 mm., the average of both classes being 62.1. The tails ranged be- tween 56 and 65 mm. with a median of 62. It will be seen in Chart II that the majority of the males had larger wings than the females, but that there was a certain overlap. Any bird with a measurement of 63 mm. or under is almost certainly a female ; any bird with a wing measurement of 65 mm. or over is almost certainly a male, but the birds with wings of 64 mm. were fairly evenly divided as to sex. There were 8 males among the breeding birds with wings of 64 mm., 3 of 63 and one of 62. Among the breeding females, there were 17 with wings measuring 64 mm., 3 measuring 65 and one 66 — the last four being quite exceptional, just as were the four smallest males. 131M — the male with a wing of 62 mm. in 1932, but 63 mm. in 1935 — I con- sidered the female of the pair until I observed him singing after being banded. K155 — the giant with the wing of 66 mm. — I could not believe was a female until I had watched her in the field. But as a rule there was no difficulty in assigning the bird to the proper sex. The measurements of the transients did not appear to be different from those of the breeding birds ; at least none was larger or smaller. I called all birds with wings 63 mm. or under females ; all those with wings 65 mm. or over males^ while of the 10 birds with wings of 64 mm., the three with tails 66 and 67 mm. were called males, and those with shorter tails females. ' I never attempted to take the total length, but Wetherbee, 197, reports an average of 150.7 mm. for adult summer resident males and 144.8 mm. for females of Melospiza in. mclodia (the Eastern Song Sparrow) in New England; the wing and tail measurements of her birds agree well with those of mine except that some of her breed- 20 ing males reached a wing length of 70.75 mm. and some transients 72 mm. B. Weights Beginning with September 1931 each bird captured was brought into the house, and after being banded and measured was placed in a small cloth bag and weighed on scales sensitive to one-tenth of a gram. The weight of the bag fluctuated with atmospheric conditions and had to be determined each day. Seven hundred and forty-six weights of some 455 individuals were taken. Almost no difference was found in the weights of residents, summer residents, transients, and winter residents taken at the same time of year. The noon weights of 126 males in March and April gave the following averages: 52 residents 22.8 g., 52 summer resi- dents 22.6 g., and 22 transients 22.4 g. The weights of Wetherbee’s, 197, 39 spring transients of the Eastern Song Sparrow were higher than those of any other category — 24 g. — ^but these were all weighed in March. The 34 breeding males taken from April 15 to August 30 averaged 21.8 g., the 21 breeding females 19.8 g., 121 immature birds from June 24 to September 5 averaged 19.7 g. and 81 fall transients 21.9 g. The weights of 18 male Song Sparrows collected by Dr. L. E. Hicks in various local- ities in Ohio from March 19 to July 8, 1935, ranged from 19.7 g. to 24.3 g., aver- aging 22.1 g., which compares closely with what I consider standard weight for my males — 22.4 g. TABLE I Song Sparrow Weights According to Time of Day 6-8 A. M. 9 A.M.-2 P.M. 3-6 P. M. Av. Sex Month Total No. No. Birds Av. Wt. No. Av. Birds Wt. No. Av. Birds Wt. Per cent of all Increase Wghts, Male February 82 18 23.5 49 24.2 15 24.6 47 24.1 Male March 148 23 22.6 90 22.9 35 237 4.9 23.0 Female April 78 19 21.0 48 21.4 II 22.0 4-8 21.4 Male II Mos. 463 105 22.3 263 23.1 95 23-4 4-9 23.0 Female 10 Mos. 267 75 20.8 140 21.5 52 21.7 4-3 21.3 I. Weights Throughout the Day An increase in weight takes place towards noon and especially late afternoon. This is shown in Table I where samples are given 21 of three months where sufficient numbers of birds were weighed at the three different periods to give trustworthy results, and also the total weights of all the males and all the females (except for 1 6 that were laying eggs). The percentage of daily increase is somewhat less than 5 per cent. Partin, 145, with some 740 weighings of adult House Finches (Carpodacus mexicanus frontalis) reported an average daily fluc- tuation of 3.5 per cent. Linsdale and Sumner, 108, weighed four captive Zonotrichia coronata three times a day for 35 days and found a daily fluctuation of “less than 4 per cent.’’ 2. Weights Throughout the Year The 746 weights of the Song Sparrows are shown distributed by sex and months in Table II and also in Chart III. Immatures are included in September and October, but juvenals less than a month old are omitted. TABLE II Weights in Grams of Adult Song Sparrows Throughout the Year Including Breeding Birds, Winter Residents and Transients — Males — — Females — Number Number Total Month of Birds Range Average of Birds Range Average Number September - i6 19.2-25.0 21. 1 17 17.0-23.0 20.1 33 October - - 45 18.8-24.4 22.0 25 18.9-24.3 20.8 70 November - 27 20.0-24.3 21.8 14 I9-4-23-2 20.8 41 December - 24 20.7-26.8 23.8 18 18.9-24.4 22.2 42 January - - 30 21.2-30.0 24.9 6 21.7-25.8 22.9 36 February - 82 20.1-29.0 24.1 22 20.0-24.4 22.7 104 March - 148 19.2-27.9 23.0 64 18.1-24.2 21.2 212 April - - - 63 19.6-25.8 22.4 87 17.1-26.0 21.7 150 (78)^ (17.1-25.7) (21.4) May - - - 20 19.6-23.4 21.2 23 18.2-25.1 21.8 43 (17)’ (18.2-22.7) (20.7) June - - - 6 19.9-23.6 21.7 7 17.6-23.4 20.5 13 ( 6)^ (17.6-21.7) (19.9) August 2 19.7-21.4 20.6 2 Total - 463 18.8-30.0 23.0 283 17.0-25.8 21.4 746 (267) (21.3) ‘Figures in parentheses represent values without the 16 laying females. 22 Chart III. Weights of Song Sparrows Throughout the year Males: 463 Weights Females: 283 Weights ----- Females zvith 16 Laying Birds Omitted The weights are somewhat low in fall, reach their maximum in late December, January and early February, gradually decrease to what we might call a “standard weight” in April and from then on (except for laying females) decrease to a lower point than in fall. The January figure is ii per cent higher than the April one for the males, and 7.5 per cent higher for the females. It is unfortunate that July and August are practically unrepresented in my data. My birds normally molt in August and September and their weights are un- doubtedly lowest at this period. The fall weights are slightly lower than the late spring weights. If we compare the two main migration months — October and ]\Iarch — we find the average of the spring birds some 3 per cent heavier than the fall birds in both sexes. The males are heavier than the females in every month but May — when 6 out of a total of 43 birds of both sexes happened to be laying females. The average of the 463 males is 23.0 g., that of the 283 females is 21.4 g., or 93.0 per cent the weight of the male. Linsdale, 106, found female Fox Sparrows (Posscrella iliaca) 98 per cent as heavy as the male, and Partin the female House Finch 99 per cent as heavy as the male. If we omit the 16 laying females, as shown in the figures in par- entheses in Table II, we find that the males average consistently heavier 23 than the females, the differences ranging from 0.5 to 2 g. per month, and averaging 1.6 g. a. Weights in Winter The height of the curve for both sexes from December through February is striking, as are also many of the weights of individual birds, some gaining as much as 25 to 44 per cent of their lowest weights. The weight increase was not a matter of the birds growing fat on food provided by me, for the highest weights came from birds that seldom benefited from my bounty. The supply of weed seeds on Interpont is apparently ample for the needs of all the Song Sparrows. A few individual weights will be given; the bird’s year of birth being given in parantheses : 50M (1930) — 1932; Jan. 7, 27 g. ; Mar. 7, 24.2 g. ; May 20, 23.5 g. ; Dec. 16, 25-8 g. ; 1933: Apr. 21, 23.1 g. 4M (1926 P-1927?) — 1931: Sept. 7, 21 g. ; 1932: Jan. ii, 30 g. ; Mar. ii, 27.9 g. ; Mar. 22, 25.1 g. ; April 25, 24.2 g. ; May 3, 23.1 g. ; May 18, 22.3 g. ; Dec. 14, 26.6 g.; 1933: Apr. 29, 23.7 g. ; 1934: Mar. 31, 24.1 g. ; Apr. 23, 24.2 g. ; 1935: Mar. 29, 23.6 g. ; Apr. 2, 23.6 g. ; May 30, 23.4 g. 187M (1933)— 1933: Dec. 3, 24.5 g. ; Dec. 26, 26.3 g. ; 1934: Mar. 16, 23.6 g. 83M (1931)— 1931 : Sept. 7, 23 g. ; 1932: Feb. 5, 26.4 g.; Mar. 8, 23.3 g. ; Apr. 12, 23.9 g. ; June 6, 21 g. 86M (1931) — 1931: Nov. 6, 20.8 g.; 1932: Jan. 6, 25.3 g. ; Mar. 7, 20.8 g. 221M (1934)— 1934: Nov. 24, 22.3 g. ; 1935; Jan. 25, 25.5 g. ; Mar. 29, 21.9 g. ; Apr. 23, 21 g. ; May 17, 20.6 g. The curve for Partin’s House Finches in California is fairly similar to that of my Song Sparrows, especially in the marked in- crease from December to February. Other birds for which increased weight in winter has been recorded are Golden-crowned Sparrows (Zonotrichia coronata) ; Fox Sparrows {Passerella iliaca) , lop; P)ramblings (Fringilla montifringilla) ; Yellow Hammers (Bmberiza citrinella) ; and Fieldfares (Tnrdus pilaris) (Zedlitz, 216) ; House Sparrows (Passer domesticiis) , gg; Starlings (Sturmis vulgaris), j8 ] Slate-colored Juncos (Jiinco li. hyemalis) about 6% on Interpont, and Cardinals (Richmondena c. cardinalis) about 3% on Interpont. Kendeigh speaks of having data that indicate “that this same weight relation between summer and winter may be generally true for passer- ine species,” gg: p. 333. Zedlitz says, however, that Magpies (Pica pica) 24 lose in winter, and that he found no gain in the Northern Willow Titmouse {Pams atricapillus borealis) and the Marsh Titmouse {Parus p. pahistris) ; I found practically none in Tufted Titmice {Baeolophus bicolor) and Carolina Chickadees {PentJiestes c. carolinensis) . On the basis of 287 specimens of the Chinese Tree Sparrow {Passer montaniis satiiratus) taken throughout the year Shaw reports that the “seasonal variation is very slight,’’ iy6. Dr. Ernst Mayr sug- gests that the lack of gain may be correlated with the habit of sleeping in holes. Linsdale’s and Sumner’s captive Zonotrichia coronata “tended to gain weight during cool weather,” 108, but there seems to have been no question here of really cold weather. Hicks weighed nearly 3,000 Starlings between December 6 and March 28, finding the following average in grams: males — December 81.46, January 84.65, February 87.42, March 85.15; females — December 77.15, January 80.73, Feb- ruary 82.24, March 79.46. “Starlings commonly gain weight in cold weather if the ground is bare or the snowfall light. . . . Starlings at Columbus (winter of 1933-1934) did not lose weight during the near-zero weather of December 26th-29th and actually gained weight during the near-zero weather of Japuary 28th-3ist. However, the sub- zero weather of February 8th-ioth, followed by the long near-zero cold wave and snows of February 20th-28th, resulted in an average loss of 9.1 grams weight per bird. Many are known to have perished from starvation,” y8. I do not have much data on the relationship between weight and changes in winter temperature, but there is evidence of an initial loss of weight in some cases during a cold spell. In January 1935 four Song Sparrows caught from the 19th to 22nd, during a week in which the mean temperatures averaged 4.4° C. (8° F.) above normal, averaged 24.4 g. in weight, but the same birds on recapture from the 25th to 28th, during a week with temperatures averaging 6° C. (11° F.) belozv normal, averaged only 22.9 g. However, in the very severe weather of the following winter, when from January 19 to 31 the mean temperature averaged 11° C. (20° F.) below normal and from February i to 21 6° C. (11° F.) below normal, 8 weights of male Song Sparrows from January 19 to 25 averaged 24.6 g., and 12 weights of males from February i to 18 25.5 g. And these were not all birds that fed at my feeding shelf ; four males 25 that were captured February lo by the river, where they had been baited to only a small extent, averaged 25.4 g. So even though the ground was covered with snow and the nights were long (about 14 hours), these birds adjusted to the abnormally low temperatures by increased weight. The decreased temperature in winter with consequent increased appetite in the birds (see Kleiber and Dougherty, 96), the absence of territorial and reproduc- tive activities, and the sedentary habits of the Song Sparrow all favor an increase in weight. “Another factor that may be involved in giving birds a resistance to extremely low temperatures in the winter,” writes Kendeigh, 94, p. 336, “is a poten- tially more active endocrine system and the ability, when necessary, to increase greatly the rate of metabolism or heat production in the body. The regulation of heat production in the body is an involuntary function, controlled in large part through endocrine action (Baldwin and Kendeigh, 1932). The necessity for a rapid metabolism and consequently more rapid utilization of reserve food supplies is generally avoided at medium winter air temperatures by the substitution of a better insulating coat of feathers and fat. A greatly increased rate of metabolism in order to maintain the body temperature would be advantageous and necessary only during periods of unusually low temperature.” Kendeigh gives a table (p. 335) showing that plumages are appreciably heavier in fall and winter than in spring and summer, that of Passer domesticiis averaging 1.7 g. in fall, 1.5 g. in winter and 1.2 g. in summer (without rectrices and remiges). He found that “Heavier birds live longer than do lighter birds at low air temperatures” 94, p. 341. Riddle, 755, 136, and Haecker, 69, reported that the thyroids in birds are heavier in fall and winter than in spring and summer, while Kiichler, 98a, found that winter was a period of storing up of material in these glands. Wetmore, 198a, established by actual count the greater number of feathers on birds in winter than in summer. Six Song Sparrows taken in March had from 2,093 to 2,335 feathers, while a bird on July 2 had only 1,304. b. Weights in Spring It will be noted in Chart III that the males, after reaching their peak in January decline almost consistently till May, showing a slight upward peak in June which is of no significance because of the small number of birds involved. The figures by half months are : Feb- ruary 24.4 g., 23.2 g. ; IMarch 23.4 g., 22.7 g. ; April 22.6 g., 22.2 g.,— a steady loss of the excess weight of winter. This loss undoubtedly results in part from activities connected with territory matters, involving singing for perhaps half the day- light hours, but that is not the whole story. For W6 — the male that 26 wintered on Interpont for four seasons — also showed a spring loss, weighing 27.3 g. on February 5, 1932, 27.6 g. on January 29, 1933, and only 24.6 g. on March 27, 1933. In April the male has reached his “standard weight” — 22.4 g. A few males in 1933 reached it as early as February 9; perhaps this is not surprising when we realize that they had been proclaiming territory for nearly three weeks by this date. Linsdale and Sumner, log, found from 1,422 weights of the Golden-crowned Sparrow a peak in weight in January and a much higher one in May, while 71 1 weights on the Fox Sparrow showed similar peaks in December and May. “Sup- plementary records . . . indicate that high weight is maintained until arrival on breeding grounds.” Heydweiller, 77a, reports that the Tree Sparrow {Spizella a. arborea) attains its maximum weight “just preceding the spring departure during the first two weeks in March.” Kendeigh, 94, reports higher weights in spring than fall with White-crowned {Zonotrichia leucophrys) and White-throated Sparrows {Zonotrichia albicollis). My own figures on the latter on Interpont show an average fall weight of 25 g. (124 birds) and an average spring weight of 29.9 g. (35 birds), a 20 per cent increase. The Slate-colored Junco {Junco h. hyemalis) averaged somewhat heavier in winter and spring than in fall ; 19.5 g. in fall (75 birds), 21.2 g. in winter (81 birds), 20.6 g. in spring (12 birds). The weight curves of Golden-crowned and Fox Sparrows are very different in spring from that of my Song Sparrows that do not gain then, but lose steadily. The average weight in ]\Iay was low, because over half the males were feeding young at the time of capture. These ii birds averaged only 20.3 g., while 9 more care-free males averaged 22.3 g. In June the four fathers averaged 20.9 g., the other two 23.2 g. It is interest- ing how these hard working individuals lost some 9 per cent of their normal weight, an average loss of 2 g. The females show a similar decline from January to April, but after that the course of their weight depends on the stage of the nesting cycle. The 38 birds weighed in the second half of March averaged 21 g., while the 75 females (not within 4 days of laying) weighed in April averaged 21.3 g. The female at this stage spends a great deal of her time eating. Five birds, three to four days before starting to lay, averaged 22.8 g. ; 16 birds, one to two days before lay- ing and during laying, averaged 24.1 g. (ranging from 23 to 26 g.) ; 9 birds, during incubation, averaged 21.5 g., while ii birds feeding young averaged 19.3 g. 27 In the first two categories we see reflected the growth of the eggs, profoundly affecting the female’s weight when she is in the midst of laying a set. (See Chapter XII under The Size of the Eggs.} As to the matter of incubation, a slight gain is evident here. In three birds near the end of incubation there was an average increase of 4.5 per cent in comparison to their pre-nesting weight. Riddle and Braucher, 158, found that Doves gained some 8 per cent during the 15 or 18 days spent in incubation and attributed this to the relative inactivity of the birds. Perhaps incubation is a time of rest and re- cuperation for the Song Sparrow ; certainly it is not the “arduous duty” that some would have us think. Feeding the young, on the other hand, is a strenuous period as is shown by the drop in weight. Just as in the 15 males, the ii females showed a 9 per cent loss from the 21.3 g. average weight with which they started the nesting season. Heydweiller, 77a, found an even greater loss with the Tree Sparrow {Spizella a. arborea) in Manitoba while feeding young — “almost 20 per cent for the males and 10 per cent for the females.” 3. Weight and Age Does weight increase with age? WTittle, 2oy, and Wetherbee, igy, found the weights of immature Song Sparrows in summer slightly less than those of adults, the average of 23 weighed by the former coming to 21.37 121 weighed by the latter 19.68 g. I could not find any consistent difference in the fall between immature and adult birds. The heaviest weight in the three fall months of any bird was that of a young male, (99M), nearly through the post- ju venal molt on Sep- tember 13, at which time he weighed 25 g. As to older birds, 4M was heavy, but other long-lived males have not weighed more than average. Two birds at least four years old weighed as follows: 23M, 22.7 g. on April 5, and 131M, 19 g. on May 17 while feeding young. When nearly five years old, loiM weighed 22 g. on April 30; 57M, when nearly six years old, weighed 24 g. on February 10, 1936, but only 21.7 g. on February 18. Two of my old females were heavy birds with large wing meas- urements: Kii, when at least three years old, showed a wing measure- ment of 65 mm. and a weight of 23.3 g. on March 28 ; K24 at the same age, had a wing measurement of 64 mm. and weight of 24.2 g. 28 on April ii. The other females I was not able to capture late in their careers. Young birds in fall and winter have been reported as weighing less than adult birds by Hicks for Starlings, and Crows (Corvus brachyrhynchos, 8i), Zedlitz for Hooded Crows {Corvus cornix, 216), Haigh for Pink-footed Geese (Anser brachyrhynchos, 70), and Sumner for California Quail (Lophortyx calif ornica, 184a). In conclusion, let me emphasize the fact that my data on weights are almost entirely a by-product of capturing Song Sparrows for band- ing, not a deliberate attempt to investigate problems connected with weights. Nevertheless^ much valuable information has accumulated and light has been thrown on various aspects of the biology of this bird. The possibilities of investigations along this line are great, especially with species that enter traps more readily than does Melo- spiza with me. C. Summary 1. Measurement of the wing has been used as an aid in deter- mining sex, the great majority of males having wings 65 to 69 mm. in length, the majority of females 59 to 63 mm. (Chart II.) 2. Seven hundred and forty-six weights of adult Song Sparrows were recorded. WYight increases as a rule during the day, the increase reaching 4.3 to 4.9 per cent as shown in Table I. 3. WYight is at a minimum in late summer and fall starting to increase in December, reaching a maximum in January and decreasing again to a “standard” weight in April. (Table II, Chart III.) 4. Some of the Song Sparrows increased as much as 25 to 44 per cent in winter. The average January weight was ii per cent above standard for the males, 7.5 per cent for the females. 5. Increased weight in winter has been noted in a number of other birds, but in a few species there appears to be little or none. 6. IMales while feeding young averaged 9 per cent less than standard weight. 7. The weight of females increases markedly just before and during the deposition of eggs ; it appears to increase slightly during incubation, but decreases again (as much as 9 per cent), while young are being fed. 8. Some old birds were heavy, others not. 29 CHAPTER IV Migratory Status of the Song Sparrows on Interpont The Song Sparrows on Interpont have proved particularly inter- esting in the matter of migration, because part of the breeding popu- lation is resident and part migratory. By means of banding it has been possible to find out something about the behavior of brothers and sisters, parents and children, grandparents and grandchildren, in re- spect to migration and permanent residency. It has also been possible to observe the stability of the migrating character in the individual. The Song Sparrow population on Interpont is made up of four categories : residents, summer residents, winter residents and tran- sients. Of the 533 adults banded, 306 (158 males,. 148 females) be- longed to the first two categories ; 80 were assigned to the third and 147 to the fourth, or approximately 57, 15 and 28 per cent respectively. There is some uncertainty in regard to these figures for a potential resident, captured in the winter, that died before spring, would be counted a winter resident; a potential summer resident captured in March and never seen again would be called a transient. Of 353 nestlings banded from May 1929 to June 1935, 26 males and 14 females have joined the ranks of nesting birds, thus bringing the totals of banded breeding birds to 184 males and 162 females, or 346 in all, (See Appendix II for data on the birds banded through 1935.) Of the 886 birds banded not a single one has been reported away from Columbus. Hence, we do not know where the last two groups breed, nor do we know where our summer residents or transients spend the winter. (A Song Sparrow banded in May at Gates Mills, Ohio, was taken in December in Georgia; see Appendix II.) It is not possible to distinguish any of these classes by weights or measure- ments. But in the spring there is a difference in appearance, as the birds that arrive from the south are much cleaner and lighter in color than those that have been subjected to the soot of the city of Columbus. As to the sex ratio of the trapped adults, if we count each in- dividual only once each month (some birds were captured several times), males made up the following percentage: September 52; Octo- ber and November 70; December through February 74; March 67 and 30 April, May and June 45. This last figure does not represent the popu- lation present, which actually shows a small preponderance of males ; it merely means that most of my capturing of nesting females was in April and May while a large proportion of the males were caught earlier. The figures show that the wintering population consists largely of males. The preponderance of males in the migration months may be due to the fact that in the fall I get the late transients (having trapped in September only in 1931), and in spring the early transients, for the spring transients have seldom entered the traps except in the cold and snowy weather of March 18 to 20, 1934, when I caught 26 unbanded Song Sparrows, 15 of which I judged to be males and ii females. A. The Transients The transients arrive in March, the bulk of them from the middle to about the 25th of the month, a few still being recorded during the first week in April. On their return journey the first birds arrive the last of September, but October is the chief month of both arrival and departure. Transients do not differ in appearance from summer residents, except that a few appear especially light in coloring; their behavior is different, for they are quiet and inconspicuous, the adult males being silent, although the young birds warble. No transient has ever been taken in a later season. B. The Winter Residents These birds come in October and possibly early November; banded individuals have been recorded until February 18, 1930; February 12, 17, 22, March 27, 1931 ; ]\Iarch 3, 5, 18, 1932; March 27, 1933; March 7, 1935; March II, 1936. Although I have banded over 70 winter residents, only two birds have returned to Interpont — W 1 and W6. The former, wearing an aluminum and faded celluloid band, was seen in the fall and early winter of 1931 and again in 1932; I was not able to capture it, but it must have been a bird banded in 1930. (I had considerable trouble with the fading of the first bands, which I made out of celluloid toys.) W6 is a most exceptional character, since he settled four winters in the same spot on North Interpont. I first captured him Feb. 26, 1931, and heard him warbling from Feb. 28 to Mar. 27. In the fall I found him (November 2) in the 31 same region, noting him at intervals each month until his capture, Feb. 5, 1932, but never hearing him sing, except a little on Feb. 8. The next winter I did not locate him until Jan. 4, trapping him the 29th, and giving him the fine new colored bands of the Biological Survey. Interestingly enough, this year he sang quite a little from Feb. 18 to Mar. i, at first rather softly and from only half way up a weed, but later loudly. Immediately after this all the cover was destroyed on W6’s wintering home and I saw nothing more of him until I happened to trap him on Mar. 27-, 135 meters directly to the east along the bluff. In the fall I was delighted to see him on his old stamping ground on Oct. 17, 1933, and hear him sing his queer, dis- tinctive song. He sang again the next day, but that was the last I ever saw of him. The cover on Interpont had been so destroyed that his former winter home was no longer suitable ; either he was captured by some enemy, or he moved elsewhere. It is curious how two wintering birds should be so faithful while not one of the 70 others was recorded in a season after it was banded. Wharton, 200, records the return of two Song Sparrows to their winter quarters at Summer- ville, South Carolina. Winter residents are sometimes more numerous than the residents and sometimes present in about equal numbers. (Hicks and Chapman, 82, found the Song Sparrow ranking as the fifth most abundant winter bird in Ohio on the 392 Christmas censuses taken in the state for Bird-Lore from 1900 to 1932.) The proportion of females appears to be very small indeed, 14 out of 36 different birds of this sex captured December to February having been assumed to have been winter resi- dents, but a number of these were probably residents. Adult males seldom sing, but the Juvenals warble a considerable amount. In Oklahoma Song Sparrows were present only as transients and winter residents ; I recorded adult songs occasionally, but curiously enough, I never heard warbling. Ridgway, 759, says that in southern Illinois this species “is a winter sojourner, abundant, but very retiring, inhabiting almost solely the bushy swamps in the bottom-lands, and unknown as a song bird ” Howell, 87, in writing of this bird in Flor- ida, says it “is practically silent during its stay in the south.” C. Residents and Summer Residents When I started this study, I shared the common opinion that the breeding birds were summer residents and that all the birds present in winter had nested to the north. Therefore it was a great surprise to me in the winter of 1929-1930 to find 4M continuously present. 32 (Later I came upon four published instances where 8 banded Eastern Song Sparrows (Melospiza m. melodia) had been found to be resi- dent, two in Pennsylvania (Gillespie, dj, and Middleton, 124), one in New York State (Baasch, 8), and one at ]\Iartha’s Vineyard (Eustis, 5(5) ). The next season I discovered that half of my banded nesting males spent the entire year on Interpont, while the other half went south for the winter. At first I believed that all the females migrated,, but in the spring of 1931 I began to suspect that the few dark colored females that joined their mates from the second to the fourth week in February might be resident and the next winter I found this to be true. I. “Individual Migration” The situation with these Song Sparrows is that called ‘individual migration’’ by Thomson, igo, igi, where “individual birds belonging to the same species and native to the same area may behave differently as regards migration.” He cites Lapwings {V anelliis vanellns) nesting in Aberdeenshire and asks : “If the racial custom is similarly inherited by all the birds, what is it that stimulates it to greater activity, or to different activity, as between one individual and another? . . . Are there various gcntes not morphologically distinguishable but differing in constitution and temperament in ways not at present definable, as,, for instance, a resident gens, an Ireland-seeking gens, and a Portugal- seeking gens?” He also remarks that “evidence is lacking as to whether, in cases like this, any given individual behaves in the same way in successive years,” igo, p. 301. With my Song Sparrows I have been able to trace the “inherit- ance” of migratory behavior in many families, and I also have evi- dence on the status of a considerable number of individuals in suc- cessive years. Is the difference in migratory behavior a matter of the young wandering and the old remaining on their territories? Is it a matter of different gentes or strains? Or is it a matter of the migra- tory impulse being present in all the birds, and stimulated or inhibited by weather conditions and individual temperament ? The logical way to treat the subject would be first to examine the numbers of residents and summer residents present each year; and next the inheritance of migratory behavior, but the fact that this be- Plate III. Typical Ground Xest of Song Sparrow not far from Columbus; One Cowbird Egg and Five Eggs of the Owner. Typical Xest Off the Ground. Ki32’s Second Set of Unmarked Eggs; Xorth Interpont. 33 havior has not been entirely stable in all individuals necessitates a different approach, and proves that we cannot consider migrating or non-migrating as a definitely fixed character in any one bird. 2. Stability of Migrating and Non-Migrating Behavior The majority of my birds have been definitely migratory or definitely sedentary, but a few have changed their status. Twenty-four males have remained consistently resident — 18 during 2 winters, 4 during 3 winters, one for 6 winters and one for at least eight. Thirty- one males have been consistently migratory — 15 for 2 years, ii for 3 years, 4 for 4 years, and one for 5 years. But 6 other males and one female changed their status. A total of 55 males have thus retained their status, in contrast to the six that changed it. As for the females 5 remained consistently resident — 2 for 2 years, 2 for 3 years, and one for 4 years, while 37 Were consistently migratory — 28 for 2 years, 7 for 3 years, and 2 for 4 years. One changed status in contrast to 42 that retained it. I have only one certain record of a summer resident turning resident — 19M — although I thought this was the case also with 12 iM, that was noted as a light colored bird when first seen February 26, 1932, but remained the following winter. 19M migrated two winters, but remained the third — 1931-1932. The fact of a summer resident spending the winter is not so strange as that of residents migrating. I can find no instance of this in the literature. Nevertheless it has happened with 5 of my males and one female. 9M was a juvenal resident, banded Jan. 26, 1930. He settled near us and remained through the next winter, starting to sing Jan. 24. But in the fall of 1931 he disappeared, so I concluded he had been killed. On Mar. 2, I was greatly astonished to have him reappear on his territory in light, clean plumage. 54M, banded in the nest May ii, 1930, son of a summer resident — 12M — and brother of a resident — 52M — remained on his territory continuously until late Oct., 1932; he returned, bright and shining, Feb. 26, 1933. 96M was an adult resident when captured Feb. 8, 1932, but migrated in the fall, returning on the same day as 54M. lOoM and 107M were juvenals banded in the fall of 1931 ; they were recorded throughout the winter, with occasional captures, and nested in 1932, but both migrated the second winter, 107M returning Mar. 19, looM not until April. 34 In previous publications I recorded 131AI as supposedly a resident when first recorded on Mar. 17, 1932, because of his sooty appearance; he has migrated and returned three times since then, so it may well be he was migratory from the first. I thought he had moved onto Interpont from across the river ; perhaps he had arrived early and acquired a coat of soot. The one female that changed status was a resident with only one foot — K75 ; I captured her Feb. 15, 1932, and followed her history until June 14; I did not see her again until Mar. 18, 1933, the day she arrived from the south with no soot on her plumage. Unfortunately, every one of these birds that surely changed status, and 12 iM also, disappeared the summer following the change, so that I was not able to follow their histories further. 3. Inheritance of Migratory Behavior When I first found the difference in the migratory status of my Song Sparrows, I believed there must be two strains, with sons doing as their fathers had done. Therefore it was an exciting time in Feb- ruary and March 1931, capturing my 7 resident males that had been banded the previous summer, and reading their bands. And my fine theory was soon exploded ! Chart IV. Genealogies Shozving Inheritance of Migratory Behavior in the Song Sparrows. Circles indicate residents; rectangles summer residents. Where birds have changed status, the first status is indicated inside second. Numbers of birds not enclosed are of unknozon status. 35 The accompanying charts show most of the data obtained on the migratory status of relatives among my Song Sparrows. Chart IV gives a number of examples of resident sons of resident fathers, and of migratory sons of migratory fathers, but it also shows many exceptions. 87M, banded in the nest May 15, 1931, and cap- tured the following December, but never seen again, was a resident son of two migratory parents, 23M having migrated four winters. The two sons of the summer resident 12M were both resident for two years, but 54M as already mentioned migrated his third winter. 70M and all his known relatives — two mates and a son and daughter — were all migratory. The case of K46 is contradictory: with a migra- tory husband that later turned resident she had a resident son, but the next year with a resident husband, she had a migratory son. The children of 121M, nest mates, behaved in opposite ways — the brother remaining stationary, like his father, the sister migrating, like her mother. But a similar parental situation with 114M and K14 gave two resident children — brother and sister. The two resident females both had resident fathers but migratory mothers, while three other females on this chart and the next, with the same parental situation, were migratory. Three other genealogies, not shown on the charts, were as fol- lows: a resident father (14M) and mother of unknown status (K34) had a resident son (56M) ; a migratory pair (119M and K116) had a migratory son (169M) ; while a resident father (225M) and migratory mother (K211) had a migratory son (265M). The genealogies shown on Chart V are of especial interest. There is one straight summer resident line for three generations — 22M and his descendants — all five birds involved being known to have been migratory. The mates of K2 were both summer residents, as she was herself, yet she had one resident son (55M) and two resident grandsons. The history of the descendants of 25M and K28 is noteworthy be- cause of the remarkable fact of three young from one brood surviving. Unfortunately, the status of 145M is not known: he was caught some 50 meters to the south of our grounds, October 4, 1932, but never seen 36 Chart V. Additional Genealogies of Song Sparrows again. Possibly he had nested in town and had left his territory be- cause it no longer offered food and shelter. The two sisters nested here in 1932 (one 135 meters from her birthplace, the other 450) and the grandmother also; on Alay 25 I banded K28’s children in the nest, and on May 28 and 31 two broods of her great grand-children. The family histories of 24M and K51 are also of great interest. In 1930, 24M, a summer resident, and a mate of unknown status, had a resident son 57M, who lived to be almost 6 years old — my second oldest Song Sparrow so far as known. In 1932, 24M and K51, a resi- dent female, had a resident son, 155M. But the year before, K51 and a resident male, 48M, had had a son and daughter from the same nest that wintered along the fourth dike and mated in the spring. So far as I can find out, there is only one other instance of known brother and sister, in the wild, mating — Downy Woodpeckers {Dry abates pubescens medianus) in New Hampshire — as reported by Shelley, lyS. All three eggs of the first nesting of 88M and K80 hatched. 88M disappeared that summer, but K80 lived in the same locality until the spring of 1934, reaching an age of nearly three years. To sum up, we find that seven resident fathers had seven resident sons, and that four migratory fathers had four niigratory sons. But two 37 resident fathers had two migratory sons and five migratory fathers had seven resident sons. As to the daughters, four migratory pairs had migratory daughters ; two resident males and migratory females had three migra- tory daughters, while two other pairs with the same combination had two resident daughters ; a resident pair had a resident daughter, while 96M that changed status had a migratory daughter. In five cases nest mates were known to have survived : two migra- tory sisters and a brother of unknown status, two resident brothers, two instances of resident brother and sister, and one of a resident brother and migratory sister. It has been suggested that perhaps the migratory impulse was a recessive character in these birds, but theorizing as to the inheritance or non-inheritance of this character seems to me futile, when we have seen that the very same bird may migrate one year and remain the next, or more commonly, remain one winter (and in two cases two winters) and migrate the next. The instinct would appear to be present in all the birds, but for some reason is inactive in many of the birds most of the time. 4. Numbers of Residents and Summer Residents In 1930 the banded males on Central Interpont were equally divided as to migratory status. The proportion of residents among the breeding males on Upper Interpont rose from 50 per cent in 1931 to 59 in 1932, dropped to 40 in 1933, and to 35 in 1934,' reached 40 in 1935 and increased to 61 in 1936. (The numbers were as follows, the residents being given first: 1931, 24 and 24; 1932, 41 and 28; 1933, 18 and 26; 1934, 10 and 19; 1935, 10 and 15; 1936, ii and 7.) The increase of residents in 1932 was due to the large number of young birds that remained through the winter in i93i-’32. Among the summer resident males nesting on Interpont approximately a third were first-year birds, but this was true of more than half of the banded resident males — 23 out of 43. The decrease of residents the next year is partly accounted for by the change from resident to summer resident of five individuals, while only one bird made the contrary change. But by 1933 it became evident that the survival rate of the resident males was becoming much worse than that of the summer residents. 38 This ivas not due to the severity of the zvinters at Columbus ; in ^ the years from 1927 to 1936 I have not known of a single Song Spar- row coming to its end through cold and starvation here. I believe the explanation lies in increased ease of predation on Interpont during recent years, making this area a more dangerous place to winter on, than the wintering quarters in the South. But during the past winter — 1935-36 — the summer residents must have suffered heavy losses from the severe weather, for their percentage of the breeding population is the lowest in seven years. As to the females, the proportion of residents has fluctuated be- tween II and 33 per cent, averaging 19 per cent. In 1931, 5 of 46 breeding birds were residents (10.8%); in 1932, 14 of 65 (21.5%); in 1933, II of 41 (26.8%) ; in 1934, 3 of 25 (12%) ; in 1935, 4 of 25 (16%); and in 1936, 4 of 12 (33.3%). 5. Differenees Betzveen the Residents and Summer Residents It might be expected that the migratory Song Sparrow would have longer wings than his sedentary neighbor, and that the resident would be heavier than the summer resident. However, I have been able to And no significant difference between these two sets of birds in length of wing or tail, and only a very slight one in weights taken at the same time of year. As to coloring, there is no difference in the fall, but an artificial one in spring, the result of Columbus soot. In the matter of zeal in singing, there is considerable variation between males in this respect, some of the most enthusiastic being residents, and also some of the least so. Among the females, the two really energetic singers have been residents. (Females sometimes give harsh, unmusical songs early in the season before nesting begins, i^j, 198). Resident females occa- sionally start to nest earlier than some of the late-arriving migratory females, but on the whole there is little difference. Resident females do not differ from the others in the number of eggs laid. 6. Comparison With Other Speeies “Individual migration” has been found to occur in a number of species: the Cormorant {Phalacrocorax carbo sinensis), /j, the Lap- wing (Vanellus vanellus), 190, the Woodcock (Seolopax rusticola), 39 igiy Buzzard (Buteo buteo), 22, 186, Hooded Crow {Corvus cornix), 186, Greenfinch {Chloris chloris), 24, Song Thrush (Turdus phil- omelus), 21 1, European Blackbird (T. merula), 48, Starling (Sturnus vulgaris), p8, i8y, and California Shrike {Lanius ludovicianus gam- beli), 12 g. Details as to the correlation of sex and age with migrating and non-migrating are largely lacking, except for the general rule that females often winter further south than males, and that the young, in some species, wander while the adults are sedentary. With Robin Redbreasts (Brithacus riibecula), 2g, Cabanis’s Wood- peckers (Dryobates villosus hyloscopiis) , iii, Prairie Chickens {Tym- panuchus cupido americaniis) , 42, Chaffinches (Fringilla coelebs), 4P, European Blackbirds {Turdus merula), 4^, and Eastern Mocking- birds {Mimus p. polyglottos) , 100, looa, the males are sometimes permanent residents and the females migratory. According to Thienemann, 186, and von Lucanus, iig, the young of Titmice and Woodpeckers are much more migratory than are the old, and Wachs, 195, reports the same of the Blackbird (T. merula), the Buzzard {Buteo buteo) and the Sparrow Hawk {Accipiter nisus). Eaton, gi, found that first-year Herring Gulls {Lams argentatus) on the Atlantic Coast make very long migrations, second-year birds shorter ones, while older birds do not appear to migrate at all. Thomas’ data from 7,000 banded Starlings show that the birds of the year are responsible for the spread to new territory, 182. With the Buzzard and the Blackbird, young from the same nest were found to behave in opposite ways in the matter of migration, but the sex of these birds is unknown. The only case I can find of banded birds migrating one year and failing to do so the next, besides my Song Sparrows, is that of the Starlings banded by Thomas, i8y, in winter in Columbus, a number of which were taken in subsequent winters considerable distances to the northeast. Of 21 December and January returns, 14 or exactly two-thirds, were recovered to the northeast of Columbus, while 7 were taken to the southwest. “It would thus seem unquestionable that a large number of Starlings, after once having migrated at least as far south as Columbus, fail to do so in some subsequent year, remaining 40 as permanent residents in the north.” Dr. Hicks informs me that he has similar records with the Starlings banded by him. The Blackbird (Turd us merula) is reported as becoming resident in Holland during the last 6o years (Wolda, 212), in Hungary during the last 40 years (Csdrgey, 45), and recently in Sweden according to Dr. E. Lonnberg in a paper at the Eighth International Ornitholo- gical Congress at Oxford in 1934. 7. Discussion of “Individual Migration’' in the Song Sparrows It is clear that the difference between migrating and non-migrat- ing, with my birds, has nothing to do with age and also is not a matter of inheritance. The fact that seven birds changed their status shows that the character is not a hard and fast one. If we examine the years in which the birds changed status we find in the exceptionally mild fall of 1931, 19^! remained, as well as 54M, 96M, lOoM, 107M and K75, but that in the bleak fall of 1932 the last five birds migrated. 9^^! is an exception, for he migrated for the first time in 1931 — although probably not until quite late. But the behavior of the other five lends support to the theory that the weather had a good deal to do with the migrating or non-migrating of these birds. Rowan, zdj, in telling of the Mallards (Anas p. platyrhyncJios) some of which fail to migrate each fall from Alberta, says, “In year's in which the fall is late and open, a far larger number stay behind.” Mute Swans (Cygnus olor) sometimes fail to migrate in mild falls (v. Sanden, i6y), sometimes starving to death later in the season in such cases (Heinroth, y6, II, p. 147). In those Song Sparrows that change status the migratory urge cannot be very strong; perhaps the warm weather of October, 1931, nullified its promptings, while the bleak temperature of the following year gave sufficient stimulus to start the birds south. It has been suggested that perhaps the Song Sparrows that leave Interpont in October and return from February to April are not true migrants but “wander about somewhere in the vicinity.” It is entirely contrary to Song Sparrow char- acter as I know it to “wander” ; this bird settles down wherever it is, the winter- ing individuals both on Interpont and in Oklahoma remaining within an area of an acre or two for months. Aly birds may not go for more than a few hundred miles, but I believe that those that leave Interpont are as true migrants as any. The data on the spring migration given in Chapter V points to a true migration and not to an absence somewhere near Columbus. 41 Perhaps the migratory impulse is latent in all my Song Sparrows ; it functions normally in the majority of the individuals, but for some reason lies dormant in others most of the time. Miller, in discussing “individual migration” in California Shrikes, says, “For some reason, certain individuals, adult and first-year birds alike, fail to respond to the changing seasons. It is possible that psychic differences of the individual overcome what must be in Shrikes a relatively weak physi- ological migration drive, and thus permit certain birds to remain on their breeding territories,” 12^. Relationship between weather and fall migration is reported in the case of the Golden-crested Kinglet in Finland by P. Palmgren, Ueber den Massenwechsel bei Reguliis r. regiiliis (L.). 1936. Ornis Fennica, 13: 159-164. “The Golden-crested \\>en is a typical repre- sentative of those birds in which the migratory instinct functions in only a part of the population. ... At the period when migration may take place, low temperature stimulates the migratory impulse and sets the greater part of the population in motion. If cold weather comes after the waning of the migratory impulse, migration is no longer possible. Those birds that have remained have little prospect of sur- viving the winter.” D. Summary 1. There are four categories of Song Sparrows on Interpont; spring and fall transients, winter residents, summer residents and permanent residents. 2. The sex ratio of banded Song Sparrows shows a decided pre- ponderance of males from October, through March, being highest dur- ing the winter. 3. The transients pass through in March and October. There has been no return from approximately 150 transients banded. 4. The winter residents arrive in October and stay until March. There have been only two returns out of some 70 banded birds ; one of these being present three winters and the other four. 5. About half the nesting male Song Sparrows are permanent residents on Interpont, the others migrating south in October and returning in February or March. 42 6. The proportion of residents among the females has ranged from II to 33 per cent. 7. The majority of the Song Sparrows have been consistently migratory or sedentary, but six males and one female changed status. 8. One male migrated for two winters and remained the third. 9. Three males and one female remained one winter and mi- grated the next, returning the following spring; two males remained two winters, migrated the third and returned in the spring. 10. Charts are given showing information obtained on the migra- tory status of parents and children, and in three cases for three generations. 11. A brother and sister wintered in the same territory and mated in the spring. 12. The supposition of two strains of Song Sparrows on Inter- pont, one migratory and the other not, is not substantiated. 13. The survival of residents and summer residents was equally good during the first three years of the study, but after that the mor- tality of the residents increased, due, presumably, to lack of sufficient cover. 14. The percentage of summer residents dropped in 1936 to the lowest point in the seven years, apparently on account of the severity of the winter. 15. No significant differences could be detected between resi- dents and summer residents in length of wing, weight or coloring, except that the plumage of the former becomes darkened through soot. 16. “Individual migration” has been recorded in 10 other species; in at least 6 species besides the Song Sparrow males are sometimes per- manent residents and females migratory; while in still other species the young are more migratory than the old. 17. The migratory impulse is believed to be latent in all the Song Sparrows, functioning normally in the majority of the birds, lying dormant in most of the others, but perhaps capable of stimulation or inhibition in a few by weather conditions in October.. 43 CHAPTER V Spring and Fall Migration In the following discussion of the migration of the summer resi- dents, it must be remembered that the foundation of the study was recognition of the individual in the field, involving almost daily cen- suses of Interpont. In this way I was able to know the date on which each male and each female arrived, rather than depending on trapping records which in most cases would fall later than the real arrival. Fall data, naturally, are much less definite. But the arrival of a male on his nesting grounds, unless the weather has happened to turn bleak, is a conspicuous thing, for he himself sings his loudest and his neigh- bors, in turn, have to settle affairs with him with repeated territory establishment activities. The arrival of a female is the converse of the picture ; indeed, I often say to myself on nearing a territory where silence reigns over night, “Such and such a male must be either dead or married,” and upon careful search I find either two birds or none. A. Spring Migration The subject of the spring migration is a complicated one, due to the various categories of Song Sparrows that I distinguished, and to the vagaries of the weather, but it has proved a fascinating problem. We will first consider the migration in relation to time of year and temperature and later the data on individual birds. I. Migration in Relation to Temperature and Time of Year The arrival of the Song Sparrows on Interpont is shown in Chart VI where the mean temperatures at Columbus are given from Feb. 17 to Apr. 5 during five years; male and female summer residents and transients are differentiated, while the arrival of the breeding Cowbirds is also indicated. The records of 1930 and 1936 are not given because of their incompleteness; during the early spring of 1930 I was watching only a few pairs, while in 1936 I was absent from Interpont the first and last weeks of March. In 1930 there was an early migration of summer resident males in late February in connection with a marked warm wave with mean temperatures of 9°-i5.6° C. (48°-6o° F.), the first male being recorded Feb. 23 and the first female Mar. i. The main migration occurred between the 15th and 22nd of March. In 1936 the first transients were seen Feb. 28 after a warm spell from the 23rd to 25th with mean TEMPERATURE. 44 FEBRUARY MARCH APRIL 17 21 25 29 4 8 12 16 20 24 28 1 5 BREEDING MALE SONG SPARROW. i = MANY TRANSIENT SONG SPARROWS. ,t=a=l BREEDING FEMALE SONG SPARROW. cj^C= FIRST ARRIVAL OF MALE COWBIRDS. x=A FEW TRANSIENT SONG SPARROWS. ?C= FIRST ARRIVAL OF FEMALE COWBIRDS. Chart VI. Migration and Daily Mean Temperatures at Columbus TEMPERATURE, °C 45 temperatures from 9°-ii° C. (48°-52° F.). The first week of March with meant temperatures of 5.6°-7.2° C. (42°-45° F.) brought a few breeding males and one female, while four days from the 8th to nth with temperatures of 44°-i2° C. (40°-54° F.) brought more birds. It will be seen that there are usually two well defined migrations — an early migration and the main migration. Summer resident males^ have always arrived first with the exception of 1936. (Perhaps their absence this year from the earliest migration was due to the great lack of summer resident Song Sparrows at present' — only seven of these males having settled on Upper Interpont by April 6. It may have been that the birds that would have migrated earliest wintered farther north than the others and were killed by the severity of the past winter). Transients are the next to arrive, and the breeding females last. The main migration brings most of the males (except in 1932 when the majority came early), as well as the bulk of the transients and females, some of the latter not appearing until April. Song Sparrow migration is closely correlated with temperature. The early migration is absolutely dependent on a warm wave the last of February or the first of March, but the main migration is only rela- tively dependent on a rise in temperature. Severe cold waves stop migration short. It will be noted on the chart that early migration never took place except in connection with a decided rise in temperature of 9°-i6° C. (i7°-28° F.) above normal the last of February, or 7°~9° C. (i2°-iy° F.) in early March. The main migration started a few days before the middle of March with mean temperatures 4°-g° C. (8°-i6° F.) above normal, but in the absence of warm waves at this period (1931 and 1932) occurred at normal temperatures the 19th and 20th of the month. It is not only the migration of the Song Sparrows that depends on warm waves; the other February and March migrants on Inter- pont behave in a similar manner. Eight species (besides the Song Sparrow) appear here practically without fail either in February or March; they are; Eastern Mourning Dove {Zenaidura macroura Caro- line nsis) \ Killdeer {Oxyechiis v. vociferus); Bronzed Crackle (Quis- calus quiscula aeneus) ; Eastern Meadowlark (Sturnella m. magna) ; Red-eyed Towhee (Pipilo e. erythrophthahnus) ; Eastern Fox Spar- 46 row (Passerella L iliaca) ; Northern Flicker (Colaptes auratiis luteus) ; and Eastern Cowbird (Molothrus a. ater). Let us take the median date of arrival of these nine birds from 1928 or 1929 to 1935 and see how their migration agreed with one another each year. In 1930 every one was early; in 1931 8 were late and i on time; in 1932 8 were early and i late; while in 1934 all were late. In 1933 5 were early and 4 late ; possibly the wholesale destruction of cover that was started on Interpont the first of March drove away some birds. In 1935 5 were late, 3 early and i on time — a record that corresponds well with fhe course of the temperature that year. And it is not only the very early arrivals that are influenced by weather. The Brown Thrasher (Toxostoma rufum) arrived on Interpont between April ii and 13 each spring from 1930 through 1934, but after the extraordinarily warm weather in late March and early April, 1929, it came on April 3, and during the cold April of 1935 it did not appear till the 21st, while during the present cold April (1936) the first bird was heard on the i8th. The House Wren {Troglodytes a. aedon) came very early in 1929 — April 5 — and very late in 1935 — April 24 — its other arrivals falling between the 12th and 2ist. (For temperatures in April see Charts XI, XII and XIII.) In order to study the relationship of migration with temperature and time of year, we will examine the data concerned with the migra- tion of the breeding males, since their arrival is more conspicuous than that of the females or the transients, and the results are more con- sistent. TABLE III First Arrival of the Breeding Male Song Sparrows on Interpont Average Temperature Early Males Later Males in February Departure Departure Year Month C.° F.° Last 10 Days C.® F.° Date Mean Temp. C.®F.° from Normal C.° F.° Date Mean Temp. C.° F.o from Normal C.° F.® 1930 4.7 40.4 11.1 52.0 2:23 13.3 56 13.3 24 3:15 5.6 42 1.7 3 1931 1.6 36.8 3.7 38.5 2:28 8.3 47 7.2 13 3 :17 0.6 33 —3.4 —6 1932 4.2 39.6 5.1 41.2 2:26 13.3 56 12.8 23 3:18 2.2 36 —2.2 —4 1933 0.2 32.4 5.6 42.0 2:24 10.6 51 10.1 19 3 :13 12.2 54 9.0 16 1934 —5.6 22.0 —8.4 16.9 3:5 7.2 45 5.6 10 3:13 8.9 48 5.6 10 1935 0.2 32.4 —1.1 30.1 3 :2 10.6 51 9.4 17 3:11 6.6 44 3.9 7 *Avg. 1.0 33.9 2.7 36.8 2:27 10.6 51 10 18 3:15 6.1 43 2.3 4 tNor’l —0.6 30.8 0 32.0 2:27 0.6 33 3:15 3.8 39 •Average temperature for the particular dates during the six years. tAverage temperature for these dates during 57 years. 47 It is at once apparent in Table III that the early males migrate at markedly higher temperatures than the later ones — at an average of 10.6° C. (51 °F.) in contrast to 6° C. (43° F.). The difference is even more marked when we note the amount above normal — 10° C. (18° F.) for the early individuals and only 2.3° C. (4° F.) for the main migration. It is evident that the later birds are not waiting for higher temperatures. This same relationship is also shown within the two groups. The three earliest dates — Feb. 23 to 26 — occurred at mean temperatures of io°-i3° C. (i9°-24° F.) above normal, while the three from Feb. 28 to Mar. 5 took place at 5.6°-9.4° C. (io°-i7° F.) above normal. Turning now to the later males, we find the four arrivals from Mar. II to 15 coinciding with temperatures i.7°~9° C. (3°-i6° F.) above normal, while the two on the 17th and i8th took place despite tem- peratures 2.2°~3.4° C. (4°-6° F.) below normal. Correspondence between temperature and early migration is also shown in Table III where the average mean temperatures of the last ten days of February are given. These show consistently that the warmer the weather, the earlier the migration, and the colder the weather, the later the migration. This is graphically shown in Chart VII. Chart VII. Average Mean Temperature of Last Ten Days of February and First Date of Arrival of Male Song Sparrows Further evidence of the decreasing temperature threshold is given by the average mean temperatures for the ten days up to and include ing the day of arrival. They were as follows: Feb. 23, 6.4° C. (43.4° F.) ; Feb. 24, 4.8° C. (40.7° F.) ; Feb. 26, 3.4° C. (38.2° F.) ; Feb. 28, 3.6° C. (38.5° F.) ; Mar. 2, —1.3° C. (29.7° F.) ; Mar. 5, —2.8® C. (27° F.). 48 It is evident from the charts and Table III that an average tem- perature during the last ten days of February 5°-ii° C. (g°-20° F.) above normal will bring some of the breeding males and transients, but that no migration will take place when these temperatures run below normal. The high temperatures of late February in 1930 and 1932 brought the first of the breeding females on Mar. i, but in other years the temperatures were not sufficiently high to bring a female before the 6th of March (1935). The decreasing temperature threshold for the migration of the males is graphically shown in Chart VIII in Chapter VII, where the mean temperatures of the days of arrival of the males as given in Table III are indicated. Disregarding the two values for Mar. 13 which were evidently much higher than necessary, a tentative line has been drawn to indicate the threshold. The formula for such a curve would be: Tm.=53° F. — 0.7 d. (or 11.6° C. — o.39d.). Tm.rrrthe average temperature at which migration took place, d.= day, and o.y^the constant indicating the slope of the curve. Or in other words, the threshold of migration is set at 53° F. (11.6° C.) on Feb. 23, and decreases about % of a degree Fahrenheit (about 2/5 of a degree Centigrade) each day for a month. If we return to Chart VI we notice a very large migration in February in 1932 and a medium one in 1933, while only one bird came in February, 1931. These diflferences correspond with the height and duration of the warm waves in late February, but a further possible explanation for the contrast between 1932 and 1933 lies in the tem- peratures for the whole month of February. In 1933 this was only slightly above normal, but in 1932 this month was “persistently warm, with two exceptions, the warmest February of record,” 2a. It may be that the Song Sparrows had migrated part way north during the warm weather of 1932. As to 1931, the temperatures in late February and throughout March were characterized by an unusual lack of fluctua- tions; and although February averaged 3.3° C. (6° F.) above the nor- mal and March only 1.4° C. (2.5° F.) below normal, yet the early migration was almost nil, and the main migration was the latest of any year from 1929 through 1936. 49 It may be objected (see Cooke, 4^), that the temperature at the place of arrival is not what should be studied, but that at the place of departure. But we are in the unfortunate situation of not knowing where these Song Sparrows winter. Since they sometimes arrive on the second of two warm days (they are night migrants) it seems probable that the last stage of their journey was made from a region approximately 160 km. (100 miles) south of here. The mean tempera- tures at Portsmouth, Ohio, 160 km. south of Columbus, of the days before the arrival of the early males averaged 9.4° C. (49° F.), and before the arrival of the first males of the main migration 7.8° C. (46° F.). (The average mean temperature at Portsmouth for Feb- ruary is 3.4° C. (6.1° F.) above that at Columbus, and the average for March 3.3° C. (5.9° F.) above.) Warm waves are not local affairs here, but sweep up from the Gulf of Mexico. I believe that the study of the temperature at the place of arrival on the day of arrival has much to teach us in the case of early migrants. The Song Sparrow is sometimes said to follow the isotherm of spring or 35° F. (1.7° C.) in its spring migration, 161; in reality, a study of the weather maps shows that in central Ohio its migration typically follows an isotherm about midway between 40° F. and 50° F. (4.4° and 10° C.). This results in getting the first birds here at a time when the “normal” temperature is not far from 35° F. The Song Sparrows can stand much lower temperatures than this in March, yet a “normaV'' temperature of F. does not stimulate them to mi- grate. In late March they need no stimulus from temperatures higher than average, but a decided drop in temperature inhibits migration. That time of year is one of the fundamental factors in Song Sparrow migration is shown by the fact that high temperatures in December, January, and early February, although lasting one to three weeks with means of 5.6°-io° C. (42°-5o° F.) never bring a migration. Migration is conditioned by both lengthening days and tempera- ture. None of the birds is ready to come to its breeding grounds until late February no matter how high the temperature is. A few of the 50 males will respond to marked warm waves in late February and early March, but most stay away until the middle or last third of March, when they will migrate on a moderate rise in temperature. a. Discussion of Some Contrary Theories Although many observers have mentioned the importance of warm waves in connection with migration, others have failed to realize this relationship, perhaps because they are searching for a rule applic- able to all migratory birds. Cooke, 4^, stated that '‘Birds prefer to migrate in spring during a rising temperature,” and cited ‘the case of the Robin {Turdus migratorins) that arrived at Lanesboro, Minnesota, from 1885-1890 at an average temperature of 41° F. (5° C.) on the average date of Mar. 16, the normal temperature of which is 31° F. ( — 0.6° C.). But because he also worked with dates of the Baltimore Oriole {Icterus galbula) and other late migrants, he could find no consistent relationship between weather and migration as a whole. Rowan, 162, i6g, largely rules out temperature as a factor in inducing migration, and Lincoln, 104a, goes so far as to say, “The state of the weather at any point has little if anything to do with the time of arrival of migratory birds.” These authors fail to distinguish between Wettervdgel and In- stinktvdgel as Weigold calls them, iq6, — the early migrants that are strongly influenced by the weather and the later migrants that are far less affected by temperature. They try to average the whole spring with the whole migration, and the result, naturally, is a failure. Table III shows that we cannot take even the average for one month and have it correlate with the arrival date of one species ; compare 1933 and 1935. If we add the average temperatures of February and March in 1931 and again in 1932 we find their totals are the same, but in the former year all but one of the February and March migrants on Interpont were late, while in the latter all but one were early. Yet if the weather records are studied in detail, excellent correla- tion is found between temperature and migration. As Kendeigh, 94, has pointed out, it is impossible to single out one factor that is responsible for inducing migration. My results with the Song Sparrows support his conclusion that “The regulation of migra- tion as to time is controlled in the spring by rising daily maximum 51 and night temperatures and changing relative proportions daily of light and darkness.” 2. Migration of Individual Males The dates of spring arrival of 22 banded males, for which there are records for two to five years, are given in Table IV. These figures (with a few exceptions enclosed in parentheses) are believed to give the actual dates on which the birds arrived, as determined by daily censuses over Interpont. TABLE IV Dates of Spring Arrival of 22 Summer Resident Males Birds 1930 1931 1932 1933 1934 1935 2M - - Feb. 26 Mar. 20 Mar. I Mar. 8 loM - - Mar. 15* Apr. 3-5 Mar. 26 Mar, 18 Mar. 16 23 M - - Apr. 3-5 Mar. 21 Mar. 18 24M - - Mar. 23 Feb. 26 47M - - Mar. 19 Feb. 26 62M - - Mar. 23 Mar. 19 64M - - Mar. 30* Feb. 26 68M - - Mar. 26 Feb. 27 Mar. I 70M - - Feb. 28 Feb. 26 Mar. 2 iiiM - - Feb. 27* Mar. 15 Mar. 28 Mar. 17 112M - - Feb. 27* Mar. 13 115M - - Feb. 26 Feb. 24 Alar. 17 119M - - Feb. 26 Alar. 2 120M - - Feb. 26 Feb. 24 123M - - Mar. 5 Mar. 16 Mar. 18 134M - - Apr. I* Mar. 13 131M - - (Mar. 16) Mar. 16 Mar. 31 Alar. II 141M - - (Apr. 19) Mar. 14 Mar. 16 176M - - (Apr. 4) Mar. 28 Mar. 17 183M - - (Alar. 27) Mar. 30 Mar. 15 185M - - Mar. 16* Mar. 2 204M - - Mar. 27 Mar. 6 ( )=first record, but perhaps not the first arrival. =present as breeding bird. * = a first-year bird. A study of the table will show us that the so-called punctuality to the calendar does not exist with my Song Sparrows, which are strongly influenced by the weather prevailing each spring. The dates 52 of arrival of one bird may range over more than a month (64M and 68M), although usually there is considerably less difference. How consistently do these birds appear early, late or with the bulk of the migrants ? On the whole they fall rather definitely into early and late groups, exceptional behavior in a certain season being usually explainable by variations in the weather. For instance, because of the lack of marked warm waves in 1931, many of the birds that in other years came early did not appear until the height of the migration, and in 1934, because of the exceptionally cold February and meager warm wave in early IMarch there was really no early migration at all. On the other hand, in 1932 the exceptionally warm weather of the month of February brought more than half of the males to Interpont in late February. Thus some males that in other years came early, in 1931 delayed till the 19th or 20th of March, while others that came in February in 1932 waited the following year till the middle of March. Birds belonging in the early group are 70M, 115M, 119M, and 120M. 2M was moderately early except in 1931. Birds consistently late were loM (except in 1934 when he came with the majority), 23M, 123M and 131M. 176M and 183M were late arrivals except in 1936 when the former was found on ^lar. 8, the latter Mar. 10 ; apparently the main migration was very early this year. 62M arrived both years at the height of the migration. Other birds that typically belonged here were iiiM and 112M, that were brought early their first year by the exceptional weather of 1932. We cannot be sure where 24M and 47M really belonged, whether they were early migrants delayed by the lack of warm waves in 1931 and migrating at their proper time in 1932, or whether they be- longed with the bulk of the migrants and were tempted north unduly early in 1932, as iiiM and 112M appear to have been. 64M, 134M and 185M came late their first year and much earlier the next. The chief difficulty in deciding upon the status of many of the males comes from the small number of dates available. It is an interesting problem as to why one Song Sparrow should regularly migrate early and another regularly migrate late. It may be a matter of physiological constitution, and again it may be concerned with the location of the winter home, or perhaps there are other factors. 3. Migration of Individual Females The dates of the arrival of 19 banded females for which there are records for two and three years are given in Table V. 53 TABLE V Dates of Spring Arrival of 19 Summer Resident Females Birds 1929 1930 1931 1932 1933 1934 1935 K2 -Mar. 15* Mar. 15 Kii - Apr. 3-5 Mar. 25 K14 Apr. 8* Apr. 3-5 Mar 29 K24 - - Mar 26-28 Mar. 28 K41 - Mar. 24 Mar. 20 K46 - Apr. 3-5 Mar. 25 K52 - Apr. I Mar. I K58 - Mar. 24 Mar. 3 Mar. 14 K60 - Apr. 3-5 Mar. 28 K89 - - Mar. 26-28 Mar. 15 K90 - Mar. 30 Mar. 15 Mar. 18 Kioi - - Mar. 19 Mar. 18 K102 - - Mar. 20 Mar. 16 Kiio - - Mar. 23 Mar. 13 K117 - - Mar. 28 Mar. 16 K125 - - Apr. 1-3 Mar. 19 K131 - - Apr. 18** Mar. 23 Apr. 3 K165 - - Apr. 22 Mar. 17 Mar. 19-20 K181 - - Apr. 2** Apr. 2-3 ^Believed to be Young from Egg Quota. **Known to be Young from Bands. Present as a Breeding Bird. Age appears to make more difiference in arrival with females than with males, first year birds sometimes arriving very late the first year and much earlier the next. It is not easy to classify most of these birds into early and late groups. K58 was a consistently early bird, while several (Kii, K14, K24, K46, K131, K165) usually came late. The females, as well as the males, show the lateness of the migration in 1931 in comparison with 1932 and 1933. 4. Migration in Relation to Sex and Age It is evident that most of the males come before most of the females, but it is also true that quite a number of females come before the last of the males. As to age, the individual often comes later the first year than in after years, although here the weather may reverse matters. In the population on Interpont many first-year individuals arrived before some of the adults. 54 There appears to be no special advantage in early or late arrival. The early Song Sparrows in 1932 experienced the coldest weather during the entire winter, but did not appear to be any the worse for it. The fact that the nights were comparatively short at this time — about 12 hours instead of almost 15 in December — made the severe temperature less of an ordeal than it would have been at mid-winter. The late bird is usually able on the day of his arrival to wrest his territory from any first-year resident that has settled on it. The late arriving female usually finds her old place pre-empted, but there are almost always other males that still lack mates. B. FalIv Migration The transients pass through on their return journey from late September to late October. The breeding females can still be found the last week in Septem- ber and early in October, always in the vicinity of their nesting terri- tories. The summer resident males usually leave about the middle of October. The exceptionally early molt in 1930, over two weeks earlier than usual (see Chapter XV), did not affect the date of migration. There is some evidence that mild weather during the first half of October tempts the birds to stay longer, while bleak weather hastens their departure. During October, 1930, the weather was unusually mild till the 17th when a sudden drop in the temperature took place ; three males were last recorded on the 15th, one on the i6th and one on the 17th. The next year October was mild throughout the month, the average temperature of the fourth week being 4® C. (7° F.) above normal; one male was last seen on the 15th, five on the i6th (in- cluding one juvenal), and two others exceptionally late — October 28 and 31. In 1932, on the contrary, the first half of October was cold and bleak and the birds apparently left early ; after the 12th I could locate only one male, and he stayed till the i6th. The first 12 days in October 1935 were cold, but after that there was a marked warm spell for 10 days ; I recorded one summer resident until the 20th. As to October dates for the nesting females : in 1930 four birds were seen from Oct. 2 to 8; in 1931, two birds, Oct. i and 2; in 1932, two, Oct. 7 and 9; in 1933. one on Oct. 13, and 1934 three individuals, Oct. 2, 10 and 13. The latest date I have for a migratory female happens to be for a young bird, banded Sept. II, 1931, as she was finishing her molt and re-trapped Oct. 16; she returned the next spring. Mar. 27 or 28. 55 2M in 1930 migrated shortly after losing his tail, returning the following spring. I also saw a transient that fall, in the same condition, that disappeared within a day or two. C. Summary 1. The spring migration normally shows two main flights: an early migration of breeding males in late February or early March, and the main flight of breeding males and females, and also transients the middle of March (Chart VI). 2. The early migration is absolutely dependent on a warm wave the last of February or the first of March, but the main migration is only relatively dependent on a rise in temperature. Severe cold waves stop migration short. 3. The early males migrated at markedly higher temperatures — • an average of 10.6° C. — than the later males — average of 6° C. (Table III). 4. The migration of 8 other February and March migrants cor- responded well with that of the Song Sparrows. 5. The decreasing temperature threshold is shown by the average mean temperature of the last 10 days in February (Table III and Chart VII) and also by the average mean temperature of the 10 days up to and including the day of arrival. 6. A formula for the temperature threshold for the migration of the males is suggested, viz.: Tm.=53° F. — 0.7 d, as shown in Chart VIII; i.e., migration may occur on Feb. 23 at an average mean tem- perature of 53° F., its threshold decreasing about % of a degree Fahrenheit each day for a month. 7. The migration of the early males appears to follow an isotherm of some 45° (7° C.). 8. High temperatures in December, January, and early February have never brought a flight. 9. Migration is dependent on both increasing day-length and rising temperature. 10. Some males will migrate in late February if strongly stimu- lated by a decided rise in temperature, but most of the birds fail to migrate till mid-March, when they will migrate on only a slight rise in temperature. S6 11. Some authors do not distinguish between Wettervogel and Instinktvdgel, and fail to recognize the important role played by tem- perature in the case of the early migrants. 12. Fifty-seven migration dates of 22 banded males are given. Five birds came consistently early and six consistently late. Others showed considerable difiference in different years depending on the weather and also on age (Table IV). 13. Forty-three dates of arrival of 19 banded females are given (Table V). First-year females sometimes arrive very late. 14. The fall migration of the transients takes place in late Sep- tember and throughout October. 15. Summer resident females have been recorded as late as Oct. 16, and males as late as Oct. 31, but normally the former are not seen after the 13th and the latter after the i6th. 16. Bleak weather tends to hasten the fall migration, mild weather to delay it. 57 CHAPTER VI Territory Establishment Melospiza melodia, in my experience, is a typically territorial bird, behaving very much as does Howard’s classic example — the Reed Bunting (Bmberiza schoeniclus) . Territory is of fundamental im- portance to the Song Sparrow on Interpont — the basis of its individual and social life for more than half of the year. Special ceremonies are ■concerned in the establishment of territory; the matter of song is closely bound up with territory, while males show a strong and lasting attachment to their individual territories. A. The Establishment of Territory That territorial behavior is deeply ingrained in my birds is evi- denced from two things : the elaborate ceremonies that are involved in its maintenance, and the part it plays in the change from ju venal to adult singing. When a new male Song Sparrow arrives in spring, the neighbor- ing males at once try to drive him ofif. If he is a transient, he flies, but if a candidate for a territory, he stands his ground — and then the ^‘territory establishment” begins. The complete procedure consists of five parts ; assuming the role ; staking out the claim ; the chase ; the fight ; and finally the proclamation of ownership of each bird on his own bit of land. In the first part the two birds show diametrically opposed be- havior. The invader — puffed out into the shape of a ball, and often holding one wing straight up in the air and fluttering it — sings con- stantly but rather softly, the songs being given in rapid succession and often being incomplete. The defender, silent and with shoulders hunched in menacing attitude, closely follows every move of the other bird. The newcomer continues to sing flying in this peculiar pufifed out shape from bush to bush that he wants to claim. Soon the owner begins to chase the intruder, but the latter, if determined, always re- turns to the spot he wants to claim. The chasing continues and at last finishes with a fight on the ground. After this the new bird is either 58 routed or both males retire to their respective territories, and sing loud and long^ answering each other. In less serious encounters the chasing and fight are omitted, the first and last parts only being indulged in. When, however, affairs are in deadly earnest, as in the spring when a summer resident returns and finds a resident has adopted his old territory, there is little wing fluttering and pufflng, merely the singing, chasing and fight. With a thickly-settled Song Sparrow population, territory estab- lishment ceremonies of all degrees of seriousness may be seen through- out the year except during the molt; in fall and winter they are not common and occur only on mild days. At these seasons a bird will go some distance to start a “territory establishment” with another male with whom there is no question of real conflict over boundaries. In such cases the roles of despot and underling are freely interchanged. Excluding the very mildest territory establishment manifestations that are indulged in only by a young bird in the fall on some occasions when another Song Sparrow alights on a branch above him, the less serious the encounter, the more prominent is the posturing, bluff taking the place of action. As Howard, 86, p. 37, says “violent wing- action and violent contortions of the body are associated with post- poned reaction.” When a summer resident returns to And his old territory pre- empted by another bird, at first the new arrival takes the role of the invader and is pursued by the bird in possession, but it does not take long for an old bird to reverse matters ; after a fight or two he be- comes the defender and drives his rival. Burkitt, 2g, tells of an old Robin Redbreast (Erithacus rubecida) being driven from his territory by a young bird ; but this has not happened to my knowledge with the Song Sparrows ; with them the old bird usually drives off the inter- loper, although sometimes he will take a neighboring territory. But as this sometimes happens under no pressure from other males, we cannot be sure that the old male was really intimidated by the young one. Territory establishment ceremonies have not been worked out in such detail with any other species so far as I know. Howard writes of “butterfly-like” and “moth-like” flights, and of rapidly vibrated 59 wings, and Pickwell, 147, describes the boundary quarrels of the Prairie Horned Lark (Otocoris alpestris praticola) which show much resemblance to those of my Song Sparrows, except that the fight takes place in the air. But neither of these authors clearly dififerentiates between the behavior of the two participants, perhaps because they worked with unbanded birds. The Micheners describe what they think may be “a ceremony marking territorial lines” with Mockingbirds {Mimus polyglottos leucopterus) , where one of the owners of the territory “came to the fence and approached the unbanded bird facing it and bowing and bobbing. One would step forward and the other back and then they would reverse,” p. 126. Closely similar behavior is reported by Laskey, looa, with Mimus p. polyglottos. It is reasonable to expect strongly territorial species to have special instinctive reactions by which territory questions can be settled. In order to observe and understand these, however, one must have in- dividuals plainly differentiated; one must study the birds from the first taking up of territory; one must study two or three pairs inten- sively at first and finally there must be a sizeable population, so that territory establishment behavior can be shown. In 1935, for instance, when there were very few Song Sparrows, I saw almost no activity of this nature, although I was especially on the look out for it. B. Territory and the Development of Song Volumes could be written on the matter of song and territory, but I will confine myself to a brief treatment of two features. With Melospiza melodia song is the chief means of proclaiming territory ; the taking up of territory in late winter and the beginning of zealous singing coincide ; while the main season of Song Sparrow song on Interpont is in March before the arrival of the females. Territory has a powerful influence on the development of the Song Sparrow’s juvenal warble into the short separate songs of ma- turity. A young bird may be warbling along peacefully by himself, but the moment a territory rival appears, the singing becomes almost typically adult. In late February a young bird may warble in low situations on his territory, but when he sits high in a tree proclaiming ownership^ his songs are adult in form. The young transient males that pass through in March warble freely, but I have never heard a 6o young summer resident male warble in the spring on Interpont; upon the arrival at the nesting grounds the bird reacts as an adult. With the young residents the warble is given up in late February and never reappears, all of the late summer and fall warbling coming from young birds. C. Summary 1. The Song Sparrow has a special ceremony consisting of pos- ture, song and fighting for the procuring and defending of territory. 2. The new bird takes a humble, subservient role, the owner a dominating, threatening attitude. 3. The complete ceremony consists of five parts : assuming the role ; staking out the claim ; the chase ; the fight ; the subsequent procla- mation of ownership. 4. Song is Melospiza melodia's chief means of proclaiming ter- ritory. 5. The young male has a continuous song of warbling character; but in territorial situations this is changed to the adult form of song. 6i CHAPTER VII Territory Throughout the Year The actual breeding season of the Song Sparrow lasts from 3I/2 to 5 months, but the territory is inhabited by the summer resident male from 6% to 8 months and by the resident throughout the year. It is not, however, defended during the molt, nor the cold of winter, and only to a limited extent in fall. A. Territory in the Fall The Song Sparrows normally molt in August and September, an occasional bird not finishing till October. Because of my absences from Columbus at this season I do not have much data on the molt of the adults. Wharton, /pp, in Groton, Mass., says the molt of his local Song Sparrows begins during the second 10 days in August and lasts from 40 to 45 days, but from my scattered observations I should ex- pect it to last longer. Magee, ii8a, states that the wing molt of Purple Finches {Carpodacus p. purpurens) takes 10 weeks on the average. In 1930, perhaps in some way due to the unprecedented drought, the birds started to molt the middle of July and were through molting more than two weeks before their usual time. I. Singing in the Fall With the adult males there is a recrudescence in fall, in a lessened degree, of spring behavior so far as territory is concerned. Young males that have settled unmolested during the molt of the owner, are now driven ofif with appropriate territory establishment procedure, although other Song Sparrows are tolerated. Singing is heard again from some of the adult residents, while others are practically silent. During normal years the singing from summer residents is of irregular occurrence, but in 1930 there was a wonderful amount from both classes of males. With many of the birds entirely through the molt the loth of September instead of the last of the month as usual, with fine weather in September and an extraordinarily mild early October, and with the migration not taking place until its usual time in mid-October, we enjoyed a most unusual treat of Song Sparrow music. The summer resident iM in 1929 sang Sept. 28, 29 and Oct. 4, but in 1930 from Sept. 17 to Oct. ii. Song was recorded from another summer resident — loM —Sept. 10 to Oct. II in 1930; Sept. 28 and Oct. 4, 1931; Oct. 9, 1932, and Sept 28, 1933. 62 4M’s early morning singing has started on the following dates: Sept. 29, 1929; Sept. 10, 1930; Sept. 28, 1932 (we returned to Colum« bus the day before); Sept. 28, 1933; Sept. 30, 1934; and Sept. 29, 1935. Considerable warbling is heard from juvenals in the fall — from residents, summer residents, transients and winter residents. 2. Taking Up of Territories Many young residents take up their territories in their first fall and keep them for the rest of their lives ; others try to do the same but are driven out by the owner when he completes his molt; still others do not settle down until February. I do not know whether this difference depends on age or other factors. Some young summer residents also choose their territories in their first fall and return to them the following spring. 185M was caught in our garden Aug. 3, 1933, in juvenal plumage and was noted warbling 50 meters to the south from Oct. 4 to 6; on Mar. 16 he returned to the very same spot. In 1931 a right-banded bird warbled constantly west of our garden on Sept. 28 and Oct. 15, but I was not able to trap him; on Feb. 27 a right-banded bird returned and took up his territory in this same spot (112M). On Oct. I I banded 134M and found him Oct. 6 warbling south of the third dike; on Apr. I he returned and took up his territory about 100 meters to the south of this place, which at this time was entirely filled by other males. Burkitt’s, 28, young Redbreasts (Brithacus riiheciila) took up territories in July and August; Aliller, I2g, found that with California Shrikes (Lanius liidovicianns gambeli) fall is the main time for taking up of territories; the Micheners report that young Mockingbirds (Mimus polyglottos leucopterns) do so in August and September, 12^, while British Stonechats {Saxicola torquata hibernans) settle in pairs on their territories in October, loia. But all these species defend their territories throughout the year. It is interesting to find the Song Sparrow, which defends his territory only during the breeding season, settling on it so early in life. B. Behavior in Winter It may be largely a matter of habit that keeps the adult residents of both sexes in the vicinity of their territories throughout the winter, if sufficient food and cover are present. Similar behavior is shown by the winter residents, in a few cases for a number of years, as with W6, as told in Chapter IV. 63 At this season the male resident may range over an area approxi- mately 150 by 225 meters, a district six to ten times as large as the breeding territory. In cold spells birds may come unusual distances for brief visits to my feeding station, several from 270 meters, while two traveled more than 500 meters (57M and 58M, see Maps 9 and 13). In cold, snowy weather Song Sparrows are apt to form into small flocks, the organization of which is very loose. On Jan. 16, 1931, I watched 50M leave his regular flock in our graden and join another below the first dike, the birds here paying no special attention to him. After staying with them for five days, he returned to his former com- panions. These flocks on Interpont are not made up of ‘‘family parties” nor of “neighborhood groups,” since they are composed of both resi- dents and winter residents, and family ties are broken with the young when the latter are a month old ; while mates, even if both are resident and winter near together, apparently pay no more attention to each other in fall and winter than they do to strangers. C. Behavior in Spring In late January or early or mid-February, depending on the weather, the resident Song Sparrows begin to take up their territories — isolating themselves through hostility to other members of their species and making themselves conspicuous by song. I. Song and Temperature Song gradually comes to an end in November, and no matter what warm and pleasant weather may occur in December, only occasional snatches of song are heard. (There have been three warm spells in December of three days duration and one of six days during the period of this study; mean temperatures ranged from 7.2°-i4.4° C. (45°~58° F.), or 7.2°-i5° C. (i3°-27° F.) above normal, the median temperature being 10° C. (50° F.).) But in January song usually be- gins again, there having been from 4 to 16 days per month on which a fair amount of song was recorded from 1930 through 1935. Table VI shows the mean temperatures at which the Song Sparrows started singing. 64 TABLE VI Lowest Mean Temperatures That Started Singing Date of Start of Singing Mean Temperature of Day of Start and Two Previous Days Centigrade Fahrenheit Normal Temperature of Day of Start C. F. Jan. 7, 1930 - - - 3.8 9.4 12.2 39 49 54 —1.7 29 Jan. 8, 1935 - 00 10 12.2 47 50 54 —1.7 29 Jan. 13, 1930 - . - — 2.2 5.6 8.9 28 42 48 —1.7 29 Jan. 13, 1932 - - - 0 8.9 13.3 32 48 56 —1.7 29 Jan. 19, 1933 - - - 7.2 6.6 8.9 45 44 48 —2.2 28 Jan. 21, 1934 - - - 2.2 4.4 8.3 36 40 47 —2.2 28 Jan. 24, 1931 - - - — 2.2 1.1 6.1 28 34 43 —2.2 28 Feb. 2, 1930 - - - — 3.3 0.6 4.4 26 33 40 —1.7 29 Feb. 2, 1932 - - - — 6.6 —4.4 2.2 20 24 36 —1.7 29 Feb. 2, 1935 - - - — 2.8 1.1 0 27 34 32 —1.7 29 Feb. 7, 1934 - - - — 1.7 —3.3 -2.8 29 26 27 —1.7 29 Feb. 9, 1935 - - - — 3.3 3.8 2.8 26 39 37 —1.1 30 Feb. 11, 1934 - - - —16.8 —8.9 —2.2 2 16 28 —1.1 30 Feb. 14, 1936 - - - — 6.6 1.1 0 20 34 32 —1.1 30 Feb. 24, 1936 - - - — 5.5 2.2 8.9 22 36 48 0 32 There has been some singing on the 7th and 8th of January fol- lowing two warm days, and from the 13th to 21st following one warm day. From Jan. 21 singing has started in earnest when the previous day was only 3.3° C. (6° F.) above normal; by Feb. 2 singing has been heard on the first warm day, and by the 7th may reappear after an interval of bleak weather at a temperature slightly below normal. Singing appeared Jan. 7 and 8 at temperatures 14° C. (25° F.) above normal; from the 13th to 21st at io°-i5° C. (i9°-27° F.) above normal; on the 24th at 8° C. (15° F.) above; on Feb. 2 from 2°-y° C. (3°-i2° F.) above and Feb. 7 and ii at 1.2° C. (2° F.) belozv. In 1936 when there had been no previous singing, it started on Feb. 14 at 1.1° C. (2° F.) above normal, and restarted on the 24th at 9° C. (16° F.) above. That singing appears at progressively lower temperatures is clearly shown in Chart VIII, for which Prof. Selig Hecht of Columbia Uni- versity kindly drew the curve and gave me its formula. Ts.=54.2° F. — o.yd. (12.3° C. — o.39d.). Ts.=the temperature at which singing starts, d.=day, o.7=the constant indicating the slope of the curve. Or in other words the threshold of singing was 54.2° F. (12.3° C.) on Jan. 7 and decreased about % of a degree Fahrenheit (about 2/5 of a degree Centigrade) each day. 65 Chart VIII. Threshold of Singing of the Residents and Migration of the Males. Dates of Start of Singing, 1930 to 1936, as shown in Table VI. Dates of Migration of Breeding Males, 1930 to 1935, as shown in Table III. It is of great interest that the curves for the threshold for the start of singing and for migrating should start at approximately the same temperature and have a similar slope, but the dates are a month and a half apart. Singing and territory activity are well established the fourth week in January at a mean temperature of 6° C. (43° F.). This is also the average temperature at which the main migration of the males took place (Table III). It is of interest to note that 100 years ago De Candolle found that 6° C. or 43° F. was the threshold for growth with wheat and other plants. This “has formed the base used by Merriam (1894) in working out his life zones. This is also the base commonly used by meteorologists” (Shelford, 177). Temperatures at which the birds will start singing and those at which they will sing after once being well started are two very dif- ferent things. If the Song Sparrows are once well started, they will sing to some extent at surprisingly low temperatures for a day or two. But a sudden drop in temperature, especially if accompanied by a bleak wind may stop singing temporarily, even as late as Mar. 6 (1932). There is also a difference between restarting and making the first start, as was shown in 1936. The birds that have been well 66 Chart IX. Average Temperature, Percentage of Sunshine, and Number of Days on which Song Sparrows Sang in January from 1930 to 1936 Started, and then stopped by a bleak spell, begin more readily than did those in 1936 that got no chance to sing until Feb. 24, except for one day — Feb. 14. (During the last half of January, 1936, the highest mean temperature was 3.8° C. (39° F.) ; after that there was nothing but cold weather till Feb. 13 and 14, after which there was another cold spell lasting till the 23rd.) Singing in January is not an automatic response to a certain tem- perature ; it is influenced by the temperature of the previous days, and also by other weather conditions, being inhibited by strong wind, and Chart X. Average Temperature, Percentage of Sunshine, and Number of Days on which Song Sparrows Sang in February from 1930 to 1936 67 sometimes apparently by cloudiness. It also depends on the individ- uality of the bird — some males starting to sing much earlier than others — , and upon whether or not he has already been singing. The influence of light upon the breeding cycle has been much emphasized by Bissonnette, 22b, Cole, 40b, Rowan, /dj, and Witschi, 2iia. Let us see whether the percentage of sunshine appears to affect the singing of the Song Sparrows. In Chart IX the percentage of sunshine in Columbus in January from 1930 through 1936 is given, as well as the average temperature of these months and also the num- ber of days on which singing occurred, while corresponding data for February are given in Chart X. The amount of singing correlates very well with the average tem- perature of these two months throughout the seven years, but does not correlate with the percentage of sunshine. An interesting case that bears on this point of the effect of temperature versus lengthening days is given by Laskey, looa: a certain banded Mockingbird (Mimiis polyglottos) in Nashville, Tenn., began to sing on Feb. 26 in 1933 and on Mar. 4 in 1934, but on Jan. 10 in 1935. “The temperature during January was unusually high and the excess for the month up to the 12th was 99 and reached 162 by the 19th.” For 12 days he sang and courted his last year’s mate, and then the tem- perature “dropped in one day from 59 to 14, followed by snow. Each bird retired to its own territory and they took no further interest in one another until Mar. 3.” Excessive temperature had started courting activity at a time when the days had barely begun to lengthen. 2. Defense of the Territory Hostile behavior towards territorial rivals begins at the time that singing is well established. Other Song Sparrows, that is, juvenal residents that have not yet started to sing, and winter residents, are tolerated. Perhaps this is a matter of personal acquaintance. 4M showed no hostility to two different young winter residents that stayed on his territory through February and one until Mar. 7. Warbling, being as it is, an expression of youth and of entire lack of intention to establish a territory, does not antagonize an adult male. By March, however, all Song Sparrows are driven off, as are most other birds unless they are too large or too indifferent. House Sparrows {Passer domestiens) and Goldfinches {Spinus tristis) ignore the threats of the Song Sparrows that learn in turn to ignore these 68 species. Field Sparrows {Spizella pusilla) are driven off with special vigor; nevertheless, two pairs used regularly to nest on Interpont in the midst of the Song Sparrows. The Song Sparrow pair dominates most of the species that come into the territory; transients usually fly away, while the nesting birds merely avoid the threatened attack. The species driven off by both male and female Song Sparrows include: Juncos {Jiinco hyemalis), Tree Sparrows {Spizella arhorea), Field Sparrows {Spizella pusilla), White-throated Sparrows (Zonotrichia albicollis), White-crowned Sparrows {Zonotrichia leucophrys), Fox Sparrows {Passer ella i. iliaca), female Cardinal {Richmondena c. cardinalis) , Red-eyed Tow- hee {Pipilo e. erythrophthalmus) , Indigo Bunting {Passerina cyanea) . Grey-cheeked Thrush {Hylocichla minima aliciae), Olive-backed Thrush {Hylocichla nstulata swainsoni) , Hermit Thrush {Hylocichla guttata faxoni), Northern Yellow-throat {Geothlypis trichas brachP dactyla), House Wren {Troglodytes a. aedon), Alder Flycatcher {Bmpidonax t. trailli), and Ruby-crowned Kinglet {Corthylio c. calendula) . The approach of a Cowbird {Molothrus a. ater) is greeted with the anxiety note; if the enemy comes near the nest site it may be attacked by both of the Song Sparrows. Territorial zeal is stated by Meise, 121, to show a recrudescence at the beginning of each new nesting cycle, but this has not been my experience with the Song Sparrow. Territorial zeal typically dimin- ishes as the season advances, unless a new territorial situation arises, such as the arrival of a new male, or as in the case when K2 nested outside of her mate’s — iM — territory in the territory of her neighbor 4M. D. Summary 1. Some of the male Song Sparrows sing regularly in the fall. 2. In 1930 there was an exceptional amount of autumn singing, with all the birds through the molt two weeks or more early, and unusually warm weather in October. 3. 4M has shown a remarkable regularity in the beginning of his singing each fall from 1929 to 1935 with the exception of 1930. 4. Some young males, both residents and summer residents, take up their territories in their first fall. 69 5- Song Sparrows flock to a certain extent in cold, snowy weather. These flocks are not made up of family parties nor exclu- sively of neighborhood groups. 6. The resident males start their singing and take up their terri- tories during warm weather in late January or early February. 7. Mean temperatures of 10° C. (50° F.) on Jan. 7 and 8, and of 9° C. (48° F.) on Jan. 13 will bring some singing, while singing will be well established in late January at temperatures from 8°-6° C. (47°- 43° F.), and on Feb. 2 at 2° C. (36° F.), as shown in Table VI. 8. The threshold of singing was 54.2° F. on Jan. 7 and decreased 0.7 of a degree Fahrenheit each day, as will be seen in Chart VIII. This is similar to the threshold for migration, but occurs a month and a half earlier. 9. The number of days on which singing was recorded in January and February from 1930 through 1936 correlates well with the average temperature of these months, but not with the percentage of sunshine (Charts IX and X). 10. Song Sparrows try to drive from their territories most other species except those decidedly larger. 70 CHAPTER VIII The Territories from Year to Year The question of the return of birds to their homes is one of per- ennial interest. How faithfully do adult birds — males and females — return to their territories? How far from their birth place do young birds settle? Over how much ground does one family scatter? An- swers to these questions can be given in regard to the Song Sparrows on Interpont. A. Thk Territorie:s of the Aduet Males On Maps 2, 3, 4, 5, 6, and 7 we see the territories of the male Song Sparrows on Central Interpont during 6 seasons. The last five give the status at the beginning of the nesting season Apr. 6, but 1 ^ f © I ^ \ i '@1 C m ] ' '1 IT- E3 ^ # r© \ 0 /i] '1 33 k. \ ^ I 3‘t ! aIap 2. Territories on Central Interpont, June, 1930. 33 males A circle means a resident, a square a summer resident, a cross a first-year bird, A bird present the previous year is underlined, a line being added for each subsequent year. (Map 2-3 by courtesy of the Journ. f. Ornithologie.) 71 Map 3. Territories on Central Interpont, April 6, 1931. 31 males Map 4. Territories on Central Interpont, April 6, 1932. 44 males. 72 Map 5. Territories on Central Interpont, April 6, 1933. 29 males Map 6. Territories on Central Interpont, April 6, 1934. 19 males 73 Map 7. Territories on Central Interpont, April 6, 1935. 17 males Map 2 represents conditions in June of 1930. The reason for this is that the June map shows two males that were not present in April, while the four that had disappeared between April and June of course could not be present the following years. In the first map there are 33 males (35 in April), in the second 31, in the third 44, in the fourth 29, in the fifth 19, in the sixth 17. Unfortunately a number of the 1930 birds were never banded, hence I do not know how many of the 6 numbered from 31 to 37 are in- cluded among those numbered 61, 62, and 66-69. I believe that 42M was the same bird as 48M, but cannot be sure. The map of 1931, when there was a scarcity of Song Sparrows and yet no destruction of cover, is of interest in showing that the birds did not spread out any more than usual. I have often noticed that a new arrival in spring — a first year bird — will try and try to establish a territory among a group of Song Sparrows, at the same time ignoring equally favorable land at a little distance, that is entirely unclaimed. Do the males have exactly the same territories year after year? This has been true of many birds, notably 2M, 12M, 20M, 23M, 28M, 74 4oM, 41 M, 50M, 52M, 54M, 58M, 13 1 M, and others, for periods rang- ing from two to four years. But some change of territory has been a common occurrence ; often the new and old partly coincide, but at other, times they do not. In some cases this may perhaps be due to the exigencies of the situation a summer resident finds on his arrival ; occasionally an old bird apparently adopts a slightly different territory rather than driving out the birds already established; this was true of 24M and 47M, in 1931. But a resident or an early summer resi- dent may shift his territory with no pressure from other birds ; this was the case with 4AI, 18AI, 19AI, and iiiAI; while 185M moved from the first to the second dike, on his second return, although only one male was in residence along Dike I. 4M I believe to have nested in much the same territory shown in Alap 3 as early as 1928, although I did not band him until 1929. He has had a somewhat different territory each year, and in the winter of 1931-32 he moved 30 meters to the west, although there was no question of any Song Sparrow driving him. In his early years he was a pugnacious bird, the tyrant of the neighborhood, and kept iM continually stirred up defending his boundaries. In 1932, however, 4M spent much less energy in picking quarrels and allowed iioM — a summer resi- dent juvenal — to settle down in iM’s former land with hardly a protest. The next winter he moved even further west over into pM’s former territory, and there he nested for three years, but in 1935 he came back into our garden. As to changes of territory following destruction of cover, in March, 1933, four banded males were driven from Upper Interpont; two first-year birds left the region and were never seen again; but two adults made short moves : 96AI settled across the river and later in the season returned to his old territory, while 90M moved 180 meters to the south, settling just south of Dike 3. Territories may range in size from 2,000 square meters to nearly 6,000 (half an acre to one and a half acres), depending partly on the pugnacity of the owner and partly on the amount of space available. B. The Returns of the Females The female that has nested before tries to return to her former home, but this is often impossible because another bird may have pre- empted her place. In that case she often settles next door, but some- times joins a male at a distance from 200 to even 700 meters, even though there may be unmated males near her old territory. 75 In 54 instances involving 41 birds I know the territories of females two years in succession; 20 of these were the same, 16 were neighbor- ing territories, in 9 cases the birds moved about 100 meters, in 7 from 150 to 250 meters, in two 400 meters, and one 700. On Map 8 territories are shown of 14 females: two for four years in succession, three for three years, and nine for two years. Maps 11-14 show the locations of four females two years in succession, four for three years and one for four years. There are 20 other females whose residences two years in succession are known, but these five maps present all kinds of situations from those females Map 8. Territories of 14 Females, two, three and jour years in succession. A broken line indicates a change of residence during one season. K24 and Kis5 were present four years, K14, K58 and K165 three seasons, 9 others tzvo sea- sons. Kys changed status from resident to summer resident. ;6 Map 9. Territories of 13 Males in Relation to Birth Place. 9 residents, 4 summer residents. that stayed on or returned to the very same territories for two or in two cases (K135 and K165) three years, to the birds that moved the longest distances, with a fair sample of short distance moves besides. C- Territories of the Males Banded in the Nest The territories of 13 young males banded in the nest are shown in Map 9 with arrows connecting these territories with their birth places. Territories of eight other young males in relation to birth 77 place are shown on Maps 11-14. 106M was hatched just below Dike I in 1931 and found in 1933 nesting 1,400 meters south of his birth place. Three right-banded males that established territory (two across the river and one below Lane Avenue) were not captured despite re- peated efforts on my part ; although all three were present in April only one survived till June and he disappeared later that summer. The parentage of these birds is not known. In one case — 145M — the parentage is known, but not the bird’s territory (Map 14). The distances that the 22 young males settled from their birth places ranged from 100 to 1,400 meters, the median being 280 meters.. Map 10. Territories of 6 Females in Relation to Birth Place^ 2 residents, 4 summer residents 78 The 15 residents settled from 100 to 660 meters from their birth places, the median being 330 meters; the 7 summer residents settled from 155 to 1,400 meters from their birth places, the median being 270 meters. D. Territories of the Females Banded in the Nest Fourteen females banded as nestlings survived to start the fol- lowing breeding season, but two right-banded birds disappeared be- fore they could be captured. K66 banded in 1930 just south of Dike I was found in 1933 almost one mile south. Territories of five of these birds in relation to birth place are shown on Maps ii, 13 and 14. The territories of the other six are given on Map 10. The distances from the birthplace of the territories on which the 12 females settled ranged from 45 to 1,300 meters, the median being 270 meters. Of the 40 nestlings that survived to adulthood only five — four males and one female — were recaptured in our garden; all the others were located and trapped on their territories. E. Territories of Song Sparrows Banded in Our Garden The majority of Song Sparrows trapped in our garden, that were later found nesting, have not scattered widely. Of 20 males caught in the fall of 1931 and spring of 1932, 18 settled between 120 and 550 meters from our house, one 700 meters, and one 1,600 meters. Careful censuses over the intervening region failed to show any other banded birds. Six males trapped in the garden the following fall and winter took up territories at the following distances : one in the garden, and the others, 90, 225, 300, 450 and 900 meters, the median distance being 260 meters. Eight females captured during these same periods settled from 90 to 550 meters away. To sum up, of these 34 birds, 31, or 91 per cent, made their homes within 550 meters of our garden, while 26, or 77 per cent, did so within 360 meters. This illustrates how little these Song Sparrows wander, either in fall or spring, before settling down. The sedentary character of this species once it has taken up its territory is shown by the fact that only three of these 34 birds were later recaptured in our garden ; the others were located by repeated 79 searches and recognized by their colored bands. Probably most of them were young when banded. It is not possible to judge of the survival of a territorial bird like the Song Sparrow — either adult or young — by the birds retrapped at a central point. F. Some Family Histories Genealogical trees of several families have been given in Chart V in Chapter IV. Let us see where these different relatives settled on Interpont. Maps ii to 14 show the direct descendants in each of the lines, and the mates of these descend- ants if any offspring are known to have survived to breed, or if anything i* known of the previous or subsequent history of these mates, in which case the earlier or later territories are shown. The territories are given of eight young males and five young females, in relation to birth place ; and the territories of Map II. Ke and her Descendants. The date gives the year of nesting 8o four males two years in succession, of four for three years in succession, and one for four years. Map II shows the descendants of K2, a summer resident female that had two summer resident mates, a summer resident daughter, resident son, and two resident grandsons. The son (55M) died during his second summer. His son (95M) in his first winter sustained a broken leg that never healed properly; he was deserted by his mate before nesting began and did not survive his second winter. The daughter (K17) raised only the first of her three broods and did not return the following year. Her son (50M) returned almost to the territory of his grandfather and here he lived to be a little over three years old. Twice I banded great-grandchildren of iM and K2; soM’s five young in June, 1931, and three young May 17, 1933, but none of these survived, to my knowledge. On Map 12 the territories are given of my only straight summer resident line for three generations — 22M, his son and grandson, and other nesting places of the mates of each of the males. Both 64M and his son 112M nested two sea- 8i Map 13. Ksfs and 24M’s Descendants. See Map 14 for 24M’s and i26M’s mates in 1932. K117 rejoined 120M in 1933, but after his death joined 57M sons, but only one brood of young was raised by either bird in the three years before we left Columbus in June. On Map 13 there is no third generation, but a number of half brothers and one case of full brother and sister. A summer resident (24M) has had two resi- dent sons that settled in opposite directions from home. 57M is a particularly interesting bird, because he has always been retiring, almost never singing, yet he survived to be almost six years old, obtaining mates during each of the five seasons, and raising young at least once (1932) and probably several times. (In 1934 his mate laid five eggs; two were taken by Cowbirds and the other three were sterile.) His mate in 1933 (K117) had remated with her former mate (120M), but upon his death joined 57 M. K51 was the mother of 88M and K80 — • the brother and sister that mated ; none of the young from their first nest survived. The descendants and different residences of K28 are given on Map 14. For two years this bird lived in a pretty, tangled spot on Central Interpont, but the third year I found her 700 meters to the south nesting on a dump below Lane.- 82 Avenue. And there were still several bachelors in the vicinity of her former home. The fourth year she returned to the dump, but disappeared soon after. Her daughter (K63) was the mother of a brood of five, three of which survived to adulthood, my only example of such a happening. 145M was caught 50 meters to the south of our grounds on Oct. 4, 1932, and never seen again; he certainly had not nested on Interpont, nor in the vicinity, unless to the east in town. The sisters K123 and K131 settled in opposite directions from their birth place ; the latter was present three years, having the same mate during the last Map 14. K28: her Descendants and Residences during jour Seasons. Nesting place of 145M unknozmi; trapped in place indicated, Oct. 4, 1932 two years. In May, 1932, I banded K28’s children and also two broods of her great-grandchildren, but none, unfortunately, were found in subsequent seasons. If all the young of all the birds shown on the maps could have been banded each year, the genealogical tables undoubtedly could have been continued for some of the lines. But that was not possible and the known history of all these families has come to an end. Of all the 83 birds shown on the four maps, not a single one is alive at the date of writing — April, 1936. G. Summary 1. Some male Song Sparrows keep the same territory year after year, while others make slight changes, as shown in Maps 2-7 where the territories on Central Interpont are shown from 1930 to 1935. 2. Females have returned to their former nesting territories in 20 of 54 cases, have settled next door almost as often and in the other instances, have settled at distances from 100 to 700 meters (^lap 8). 3. Twenty-two males banded in the nest have taken up territories from 100 to 1,400 meters from their birth places, the median distance being 280 meters (^lap 9). 4. Twelve females banded in the nest have settled from 45 to 1,300 meters from their birth places, the median distance being 270 meters (Map 10). 5. Of the 40 nestlings known to have survived to adulthood, only five were captured in our garden. 6. Thirty-four Song Sparrows trapped in our garden have settled from o to 1,600 meters away, 77 per cent of the birds within a distance of 360 meters. Only three of these individuals have been recaptured in the garden. 7. The territories of the four families whose genealogical trees are shown in Chart V in Chapter IV are shown in Maps ii to 14 with other territories of the mates of the birds involved. 84 CHAPTER IX The Relations Between the Sexes The mating-pattern of the Song Sparrow follows the second type as described by Lorenz, 112, p. 327, where the male dominates his partner and yet there is for the time being a strong bond between the pair. The role of the male is that of guardian of his territory, mate, nest and young. By his singing he evokes a negative reaction in other males and a positive one in the female. He dominates his mate by “pouncing”; that is, he suddenly darts down at her, collides with her, and flies away with a loud song. Pouncing is evidently analagous to “sexual flight” in the Reed Bunting and Yellow Bunting as described by Howard, Tj bo ws Young Pledgee W U .52 u to C/3^ u boE bo Young Pledgee in in O ” o in g V a. Baron 15 England 11 71 265 160 124 60.4 46.7 b. Nicholson 143 Scotland ? 156 687 420 300 61.1 43.7 c. Praeger 149 Scotland ? 240 99 41.2 d. Clabaugh 39 California 13 38 187 103 76 55.0 40.6 e. Clabaugh 40 California 17 39 168 104 68 62.0 40.5 f. Pickwell 147 Illinois 1 30 102 79 46 77.4 45.1 g-. Potter 148 Pennsylvania 18 113 60 53.1 h. Walkinshaw 195a Michigan 1 46 20 43.5 i. Nice 127a Oklahoma 34 268 118 44.0 i. Nice (1930-1932) Ohio 1 147 585 389 243 77 66.5 41.5 52.4 k. Nice (1933-1936) Ohio 1 76 321 147 80 30 45.8 24.9 39.5 Total of Eggs in 6 d, e, f, j) - - Total of Nests in 5 h, i, j) - - - Studies (a, b, Studies (c, g. 481 814 1,994 1,225 857 374 61.4 43.0 45.9 Britain and the rest in this country from the eastern to the western border. Since all of them except the last four years with the Song Sparrows are so consistent, it seems as if we can place considerable reliance upon this ratio of 40 to 46 per cent success for open nests of Passerines in the North Temperate Zone. No definite studies appear to have been made of this matter either in the Tropics or in the Arctic. The number of successful nests is often easier to keep track of than the numbers of successful eggs. If we consider the five studies in which data on the success of nests are given (omitting the atypical results during the last four years with the Song Sparrows), we find that in 374 of 814 nests young were raised — 45.9 per cent of success. In various hole-nesting species the percentage of fledging is con- sistently higher. The percentage of success of 268 nests of the Bluebird (Sialia s. sialis) during three years was 68.6, 127; of 133 nests of the House Wren {Troglodytes a. aedon) 68, 94; of 54 nests of the Tree Swallow {Iridoprocne bieolor) in 1933 and 1934 71.5, 38; of 175 nests of the same species 57 in two years, but only 38 the third year, 114. Statistics on 283 eggs of the House Finch (Carpodaens mexicanus frontalis) — a species nesting in enclosed places — give 59 per cent of success, jp. The publications of the Phytopathological Service at Wageningen, Holland, ofifer a wealth of data on hole-nesting birds. 144 chiefly Titmice, but also Redstarts {Phoenicurus phoenicurns) and Starlings, the percentage of success ranging from 55 to 76, but the great majority of cases falling near 65, 156*, 214. It is evident that birds nesting in the open suffer from many more nesting disasters than do those nesting in holes. How do the former keep up their numbers? In many cases they attempt more broods than the hole-nesters, although this is not universally true. The Corn-Buntings {Embcriza c. calandra) studied by the Ryves show a very high percentage of fledging for ground-nesting birds, some 60 per cent for over 400 eggs, 165, 166. These birds are typically single-brooded; their late nesting date, chiefly July, is evidently more favorable than one early in the season. E. Analysis of the Loss of Eggs and Young In Table XVIII and Chart XVI the loss of eggs and young is analyzed for each of the six seasons from 1930 through 1935, the total number of eggs laid each year being the basis for all the percentages. Under “Cowbird” come those cases where eggs were eaten, and young Song Sparrows crushed or crowded out, not cases of starvation TABLE XVIII Analysis of Loss in 211 Song Sparrow Nests Cow- Parents Parents Young Flood Predator bird S&A* Failed Man Killed Starved Total Loss bo bo bo bo bo bo bo C in c in c m C in c c in c V- bo bo 3 bo 3 bo bo 3 bo 3 3 bo 3 rt bo bo 0 bo 0 bo bo 0 bo 0 bo 0 0 bo 0 0 u; {xl W W W W W W C 1930 Nos. 0 48 49 4 3 5 1 5 12 0 5 0 2 75 59 134 % 0 20.3 20.7 1.7 1.2 2.2 0.4 2.2 5.0 0 2.2 0 0.9 31.8 25.0 56.8 V931 Nos. 0 23 29 2 0 9 3 3 0 0 3 5 1 40 38 78 % 0 16.1 20.2 2.9 0 6.3 2.1 2.1 0 0 2.1 3.5 0.7 28.0 26.5 54.5 1932 Nos. 0 36 26 8 1 18 2 4 10 0 7 0 18 81 49 130 % 0 17.5 12.5 3.9 0.5 8.9 0.9 1.9 4.8 0 3.4 0 8.9 39.4 23.8 63.2 1933 Nos. 21 23 43 5 0 5 1 0 6 0 9 0 2 70 45 115 % 14.8 16.2 30.2 3.5 0 3.5 0.7 0 4.2 0 6.4 0 1.4 49.3 31.7 81.0 1934 Nos. 0 8 3 15 0 7 1 0 2 0 1 0 1 34 4 38 % 0 15.4 5.8 28.9 0 13.5 1.9 0 3.8 0 1.9 0 1.9 65.4 7.7 73.1 1935 Nos. 4 18 6 12 0 5 0 1 1 0 4 0 2 44 9 53 % Totals iNOS. 5.4 24.0 8.0 16.0 0 6.6 0 1.3 1.3 0 5.4 0 2.7 58.7 12.0 70.7 25 156 156 46 4 49 8 13 31 0 29 5 26 344 204 548 % 2.9 18.2 18.2 5.5 0.5 5.7 0.9 1.5 3.7 0 3.4 0.6 3.1 40.3 23.9 64.2 ’Sterile and addled eggs. 145 when Cowbirds were present. “Parental Failures” include the occa- sional disappearance of single eggs (8 cases), young carried off while hatching (7 cases), young pulled out of the nest (3 cases), or the last young deserted in the nest (3 cases). Under “Man” come two cases of nest robbery, the other damage having been done by plowing. When parents were killed and the nest undisturbed, boys were probably responsible in some cases and other enemies in others. A heavy rain drowned two nestlings in 1929, but no further damage occurred through this factor until the great floods of May ii and 13, 1933 which must have destroyed the majority of the ground nests in the river valleys in central and southern Ohio. In 1935 a smaller flood affected only one nest that I had found. As to the comparative loss of eggs and young, 344 eggs were lost and 204 nestlings, or 40.3 per cent loss of the original 854 eggs as eggs and 23.9 per cent as young. The egg stays in the nest from 12 to 18 days (the average of which is 15), the nestling 10 days; the average daily loss during the first three years was 2.1 per cent for the eggs and 2.5 per cent for the young. But the figures for the six years give a daily loss of 2.7 per cent for the eggs and 2.39 for the young. PER CENT q ip 2,0 3,0 4,0 5,0 6,0 70 8,0 9,0 100 EGGS 1930 YOUNG RETURNS >< XIXfXTXIX ^ III EGGS \ XIXMXDOO(]XI^jXIXIXIXIK[xI)^^ hdih A |C|M 1931 YOUNG RETURNS III EGGS X XMX XlxIX X X^X XIXIX XDK XFX[xI>^^ 5 1 RETURNS V Sr rv s/T-TvAavTs/TStTv' vT^TvTs/Ts/TV^vTs/' vl A 1 1 c LOOo I933YOUNG aiXIXLa/LAX AI/.IXiXIaIAi/'JA A C 1 M I r RETURNS EGGS >M3>(XI><><]X]X^^ A 1 c 1 M 193 4 YOUNG ATxWjiin 1 RETURNS EGGS 1 1 1 1 1 I'l 1 1 1 1 1 1— — M A 1 C I M 1 r I935YOUNG m m ATxMxiATxIMBiH si RETURNS m \ 1 1 Chart XVI. Analysis of Loss of Eggs and Young in 211 Song Sparrow Nests in Percentages of Numbers of Eggs Laid. X=eggs hatched, young fledged, and young "returning'' the following spring. Solid black— loss from predators. A— addled and sterile eggs; C—loss from Cowbirds; F=loss from flood; M=miscellaneous factors — parental failures, man, and parents killed; .•{^starvation. See Table XVIII. 146 The factor that affects only the eggs is that of sterile and addled eggs ; the factor affecting only the young is that of starvation. The Cowbird also does considerably more damage to eggs than to young. The fact that floods and plowing and the killing of parents have destroyed many more eggs than young is largely a matter of chance. Predators have taken exactly as many young as eggs — an average of 1.2 per cent for the 15 days the eggs are in the nest, and an average of 1.8 per cent for the 10 days the young are in the nest. Some enemies (cats for instance) prefer young to eggs. Often some enemy carries off the Song Sparrow eggs, leaving the somewhat larger Cow- bird egg, although the Cowbird nestling is taken as readily as the Song Sparrow. Some predators may be attracted to the nest by the begging note of the young during the last few days of nest life. Yet at this same time they must be immune from the attacks of the smaller of their enemies, such as little snakes that would eat the eggs. The loss from predators was high the first year — 41 per cent, dropped the next two years to 36 and 30 per cent, reached its maximum in 1933 with 46 per cent, and fell again the last two years reaching only 21 and 32 per cent. In Table XIX we see the loss for eggs and young during the six years and also the losses for both during the first two years and last four years. TABLE XIX Summary of Loss to Eggs and Young Divided Into Three Periods 1930-1935 1930-1931 1932-1935 854 Eggs 379 Eggs 475 Eggs i T . I ianf T r\cc Eggs Young Total Total Total Flood ----- 2.9 0 2.9 0 5.3 Predator - - - - 18.2 18.2 36.4 39.3 344 Cowbird - - - - 5-5 0.5 6.0 24 8.6 Sterile and Addled Eggs - - - 57 0 57 37 74 Parental Failure - 0.9 1.5 2.4 3-2 1.9 Man ----- 37 0 37 3-2 4.0 Parents Killed - 34 0.6 4.0 3-3 4.2 Starvation - - - 0 3.1 31 0.8 4.9 Total - - - 40.3 23-9 64.2 55-9 70.7 147 The total loss of eggs and young during 1930 and 1931 reached 55.9 per cent; during 1932 through 1935 70.7 per cent. (It must be remembered that most of the data was on the early part of the season, and that some of the great losses — at any rate in 1933 — were made good later in the season.) What is the reason for the great increase in mortality during the last four years? In the first place two factors have played a smaller role in the later years than earlier — predators (except in 1933) and parental failures. Wild predators have become increasingly scarce on Inter- pont — opossums, weasels, skunks and snakes. But all other factors have shown a large increase, and an entirely new one has been added, that of flood which amounted to a 5.3 per cent loss during the four years. Loss due to the killing of j^arents increased one-fifth, damage by man (entirely plowing) increased one-fourth, the percentage of sterile and addled eggs doubled, damage by Cowbirds mort than tripled, while death due to starvation increased six fold. In normal seasons there is a very small loss of nestlings from starvation, and even in 1930 the drought did not injure the young in the nest. But in 1932 no less than 18 nestlings died, apparently from lack of food. This condition was correlated with a heavy Cowbird infesta- tion and with a disastrous drought in May, for this month was the driest May in Columbus in the 54 years of the Weather Bureau’s record, less than an inch of rain falling — 2 cm. Thirteen of the 18 Song Sparrows had Cowbird nest-mates. There was no loss from starvation in the nestlings fledged before May 20, but in the next ten days only two broods were entirely successful, while eight suffered loss. In June no parents raised all the young hatched. No real rain came until June 27. Under normal conditions Cowbirds are usually raised without loss to the Song Sparrow young that are present ; hence this early drought must have caused a lack of insect food that resulted in the death of 18 young Song Sparrows. To return to the subject of predators. It may well be that the diminishing amount of damage by them reflects a real decrease in these animals. The very high loss in 1933 followed directly upon extended disturbance of the nesting area by plowing June 6 to 9. If I had not rescued the young they would have been destroyed by the tractor plow, but my activities in removing the threatened nestlings 148 (8 in number) and adding them to families in safe situations did little good, as most of these nests were robbed and only two of the trans- ferred nestlings were fledged. Several nestlings were apparently taken by Bronzed Crackles (Quiscalns quisciila aeneus) that came in throngs to follow the plow. Here we have a similar situation to those reported by Errington, 55, and Bennett, ly, where disturbance due to man’s interference and to exposure of nests was followed by a marked in- crease in predation. Summing up the factors responsible for the excessive losses of the last four years, we And that man and the weather are chiefly to blame. Drastic reduction in cover brought about an unbalanced condition between Cowbirds and hosts, and exposed the adult Song Sparrows to increased dangers, while the late plowing brought an influx of new predators to prey upon the young. As mentioned in Chapter IX a higher proportion of well concealed nests have suc- ceeded than of poorly concealed — namely 55 and 36 per cent respec- tively (omitting nests destroyed by floods or plowing). Extremes of weather — too little rain or too much — accounted for perhaps 8 per cent loss during the last four years (assigning half the loss through starvation to drought and half to Cowbirds). The list of factors in Tables XVIII and XIX is a compromise, the best that can be done under the circumstances, but we know that in reality man is responsible for far more than 3.7 per cent loss. On account of his disturbing activities he should be charged with much of the predator loss (cats, rats, dogs, and in June 1933 Crackles), much of the Cowbird loss (see Chapter XVI), some of the killing of parents, and perhaps even the flood. Indeed, it is only drought, sterile and addled eggs, parental failures and part of the predator and Cowbird damage that cannot ultimately be laid at his door. F. Summary 1. In 1929 two pairs of Song Sparrows each made four at- tempts at nesting, raising 9 and 5 young respectively. 2. In 1930 fifteen pairs raised from o to 10 young each, totalling 64 and averaging 4.3 young. 149 3- The great drought in 1930 cut the breeding season of the Song Sparrows short and brought on the molt more than two weeks early. 4. The proportion of five-egg sets in the early broods has been fairly consistent during the six years, ranging from 28 to 39 per cent, averaging 30 per cent. 5. The average number of eggs per nest in the early broods has been practically the same during the six years, ranging from 4.0 to 4.3 and averaging 4.1. 6. The average size of the broods hedged has decreased very much ; in the first two years it was 3.6, in the next 2.7. 7. During 1930 and 1931 nearly two-thirds of the successful nests raised their full quota, but in 1932 only 13 per cent did so and in 1933 40 per cent. In 1934 and 1935 not a single nest was com- pletely successful. 8. The average number of eggs per nest was 4.2 with both com- pletely and partially successful nests. The former raised 4.2 young per nest, the latter 2.4. 9. The percentage of eggs hatched was 70 in the first two years, 52 in the next four years, and 60 during the six years. 10. The percentage of eggs raised and fledged was 44 during the first two years and 29 during the next four, 36 during the six years. 11. The average number of young raised per total nest was 1.7 during the first two years and 1.2 during the next four, averaging 1.4 for the six years. 12. The average number of young raised per successful nest was 3.6 during the first two years and 2.6 during the next four, aver- aging 3 for all six years. 13. Comparison is made with 9 other studies of Passerine birds building open nests; of 1,994 eggs 62 per cent were hatched and 43 per cent fledged. Of 814 nests 45.9 per cent raised young. A per- centage of success of some 41 to 46 per cent of eggs and nests ap- pears to be normal in temperate North America. 14. In hole-nesting species the percentage of success averages around 65. 15. The losses of eggs are analyzed for each of the six seasons under 8 headings. 16. The total loss of the original eggs amounted to 40 per cent as eggs and 24 per cent as young — 64 per cent in all. 17. The average daily loss of eggs was 2.7 per cent and of young 2.4 per cent. 18. The percentage of the 854 eggs lost by the different factors was : flood 2.9 ; predators 36.4 ; Cowbird 6 ; sterile and addled eggs 5.7; parental failures 2.4; man — nest robbing, but chiefly plowing 3.7; parents killed 4; starvation 3.1. 19. The total loss during the first two years was 55.9 per cent; during the next four 70.7 per cent. 20. The damage done by predators decreased from 1930 to 1932, but reached a maximum in 1933 when conditions were greatly disturbed by the plowing of the whole of Interpont in June. 21. Floods occurred during the nesting season only in 1933 and 1935. 22. The heavy losses from 1932 to 1935 were due to inter- ference by man, severe parasitism by the Cowbird, drought and flood. 23. The results in 1936 were not included in this chapter, as the changes made by the addition of 52 eggs are so slight as not to warrant the reworking of the tables. In brief, they are as follows : 12 nests, 5 with 5 eggs, 6 with 4 eggs, i with 3 eggs; 2 raised 4 young, I raised 3 young, 2 raised 2 young, and 2 one young. Fifty- two eggs; 26 hatched (50%); 17 young were fledged (32.7%). Average number eggs per nest 4,3 ; number young raised per total nest 1.4, per successful nest 2.4. Wholly successful nests only one (14.3%). Loss of eggs: 13 taken by predators, 6 by Cowbird, 2 addled, 5 destroyed by man. Loss of young: 6 taken by predators, 3 starved (with Cowbird nest-mates). 24. The total loss to eggs and young of the 906 eggs laid dur- ing the 7 years was: flood, 25 (2.8%); predators, 331 (36.7%); Cowbird, 56 (6.1%) ; sterile and addled eggs, 51 (5.6%) ; parental failure, 22 (2.4%); man, 36 (4%); parents killed, 34 (3.7%); young starved, 28 (3.1%) ; a total loss of 40.9% of the eggs as eggs, and 23.5% as young — 64.4% in all. 152 CHAPTER XVI The Cowbird in Relation to the Song Sparrow Next to the European Cuckoo the Cowbird is the most famous Ijrood-parasite ; but although an immense literature exists on the former bird, the latter has been strangely neglected as a subject for study. A. The; Cowbird as Parasite Ciiculus canorus is highly specialized as a parasite, but this is not true with Molothrus ater ater. I, N on-specialization of the Cowbird The European Cuckoo is a large bird — the female weighing TOO g. — yet it parasitizes small birds, and its egg is comparatively small, averaging about 3 g., 74, 1^4. The Cowbird female, on the •other hand, weighs about 39 g. (see Appendix IV), and her egg is just about the same size as that of the Cuckoo. The Cuckoo is specific in its parasitism, its egg is abnormally small and hard shelled, and the nestling evicts eggs and young from the nest. The ■Cowbird is not specific in its parasitism, its egg is of normal size and shell-texture, the nestling does not “intentionally” evict its mates and finally the incubation period is not shorter than that of some of its relatives. As to the relative size of Cowbird eggs and other Icteridae, 'Cowbird eggs on Interpont average 8.9 per cent of the weight of the bird that lays them. It is a difficult matter to check on the weights of American birds and their eggs due to the scarcity of data, but from two different articles by Bergtold, 20, 21, and from information given me by Dr. L. E. Hicks, I found that the weight