HARVARD UNIVERSITY

LIBRARY

OF THE

Museum of Comparative Zoology

Tfie WIson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY WEST VIRGINIA U. • MORGANTOWN, W. VA.

VOL. 84, NO. 1 MARCH 1972 PAGES 1-116

The Wilson Ornithological Society Founded December 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President — Pershing B, Hofslund, Dept, of Biology, University of Minnesota Duluth, Duluth, Minnesota 55812.

First Vice-President — Kenneth C. Parkes, Carnegie Museum, Pittsburgh, Pennsylvania 15213.

Second Vice-President — Andrew J. Berger, Dept, of Zoology, University of Hawaii, Honolulu, Hawaii 96822.

Secretary — James Tate, Jr,, Laboratory of Ornithology, Cornell University, Ithaca, New York 14850.

Treasurer — William A. Klamm, 2140 Lewis Drive, Lakewood, Ohio 44107.

Elected Council Members — Ernest P, Edwards (term expires 1972) ; Elden W. Martin (term expires 1973); Robert D. Burns (term expires 1974).

Membership dues per calendar year are: Active, $8.00; Sustaining, $15,00;

Life memberships, $200 (payable in four installments).

The Wilson Bulletin is sent to all members not in arrears for dues.

The Josselyn Van Tyne Memorial Library The Josselyn Van Tyne Memorial Library of the Wilson Ornithological Society, housed in the University of Michigan Museum of Zoology, was established in concurrence with the University of Michigan in 1930, Until 1947 the Library was maintained entirely by gifts and bequests of books, reprints, and ornithological magazines from members and friends of the Society. Now two members have generously established a fund for the purchase of new books; members and friends are invited to maintain the fund by regular contribution, thus making available to all Society members the more important new books on ornithology and related subjects. The fund will be administered by the Library Committee, which will be happy to receive suggestions on the choice of new books to be added to the Library. William A. Lunk, University Museums, University of Michi- gan, is Chairman of the Committee. The Library currently receives 104 periodicals as gifts and in exchange for The Wilson Bulletin. With the usual exception of rare books, any item in the Library may be borrowed by members of the Society and will be sent prepaid (by the University of Michigan) to any address in the United States, its possessions, or Canada. Return postage is paid by the borrower. Inquiries and requests by borrowers, as well as gifts of books, pamphlets, reprints, and magazines, should be addressed to “The Josselyn Va i Tyne Memorial Library, University of Michigan Museum of Zoology, Ann Arbor, Michigan.” Contributions to the New Book Fund should be sent to the Treasurer (small sums in stamps are acceptable). A complete index of the Library’s holdings was printed in the September 1952 issue of The Wilson Bulletin and newly acquired books are listed periodically.

The Wilson Bulletin

The official organ of the Wilson Ornithological Society, published quarterly, in March, June, September, and December, at Morgantown, West Virginia. The subscription price, both in the United States and elsewhere, is $10.00 per year. Single copies, $2.50. Subscriptions, changes of address and claims for undelivered copies should be sent to the Treasurer. Most back issues of the Bulletin are available (at $2.50 each) and may be ordered from the Treasurer. Special prices will be quoted for quantity orders.

All articles and communications for publications, books and publications for reviews should be addressed to the Editor. Exchanges should be addressed to The Josselyn Van Tyne Memorial Library, Museum of Zoology, Ann Arbor, Michigan.

Second class postage at Lawrence, Kansas, U.S.A. 66044

Allen Press, Inc., Lawrence, Kansas 66044

THE

WILSON BULLETIN

A Quarterly Magazine of

Ornithology

George A. Hall

Editor

Editorial Advisory Board

Ornithological Literature Editor Peter Stetteniieim

William C. Dilger Douglas A. James William A. Lunk

Helmut C. Mueller Robert W. Nero Kenneth C. Parkes Glen E. Woolfenden

Andrew J. Meyerriecks

Volume 84

1972

Published

by

THE WILSON ORNITHOLOGICAL SOCIETY

THE WILSON BULLETIN

A QUARTERLY MAGAZINE OE ORNITHOLOGY

Published by The Wilson Ornithological Society

VoL. 84, No. 1 March 1972 Pages 1-116

CONTENTS

Laysan Albatross, Erontispiece photo by Harvey I Fisher 6

The Oceanic Distribution of the Laysan Albatross, Diomedea

IMMUTABILIS Harvey I. Fisher and James R. Fisher 7

Habits of the Crimson-Crested Woodpecker in Panama

Lawrence Kilham 28

Territorial Behavior in Savannah Sparrows in Southeastern

Michigan Peter E. Potter 48

Flocking Associates of the Pinon Jay

Russell P. Baida, Gary C. Bateman, and Gene F. Foster 60

On the Evolution of Sociality, with Particular Reference to

Tiaris OLIVACEA Ronald Pulliam, Barrie Gilbert, Peter Klopfer,

Dennis McDonald, Linda McDonald, and George Millikan 77

General Notes

FURTHER NOTES ON THE PINNATED BITTERN IN MEXICO AND CENTRAL AMERICA

Robert W. Dickerman 90

CHRONOLOGY OF HATCHING BY LAYING SEQUENCE IN CANADA GEESE

James A. Cooper and Jon R. Hickin 90

SPRING MIGRATION OF SWAINSON’s HAWK AND TURKEY VULTURE THROUGH VERA- CRUZ, MEXICO James R. Purdue, Charles C. Carpenter,

Dale L. MarcelUni, and Robert F. Clarke 92

AN UNUSUAL NEST OF THE SANDHILL CRANE Carroll D. Littlefield 93

variability of TAIL MOLT IN THE BURROWING OWL William D. Courser 93

ANOTHER RECORD OF A SHORT INCUBATION PERIOD FOR THE ROBIN

Henri C. Seibert 95

DISCOVERY OF THE NEST OF THE KAUAI AKEPA C. Robert Eddinger 95

MOBBING OF A FISH CROW BY PASSERINES Walter Kingsley Taylor 98

VESPER SPARROW NESTS ABANDONED AFTER SNOW . Max //. Schroeder 98

RECORDS OF THE SCARLET IBIS AND RED-BREASTED BLACKBIRD IN ECUADOR

Henry M. Stevenson 99

Ornithological News

100

Conservation Section, Bird Damage to Corn in the United States

IN 1970 Charles P. Stone, Donald F. Mott,

Jerome F. Besser, and John W. DeGrazio 101

Ornithological Literature 106

John A. Wiens, An Approach to the Study of Ecological Relationships among Grassland Birds, reviewed by D. Jean Tate; Phillip S. Humphrey, David Bridge, Percival W. Reynolds, and Roger Tory Peterson, Birds of Isla Grande {Tierra del Fuego), reviewed by Claes C. Olrog; Theodore C. Fitzgerald, The Coturnix Quail; Anatomy and Histology, reviewed by Robert D. Klemm; Peter Slater and others, A Field Guide to Australian Birds. Non-passerines, reviewed by Roy P. Cooper; Robert J. Raikow, Evolution of Diving Adaptations in the Stiff tail Ducks, reviewed by Lowell Spring; John S. Dunning, Portraits of Tropical Birds, reviewed by Stephen M. Russell; Jack McCormick, The Pine Barrens. A Preliminary Ecological Inventory, reviewed by Ernest A. Choate.

Publication Notes and Notices 99, 105, 116

Laysan Albatross in Flight. Photo by Harvey I. Fisher

THE OCEANIC DISTRIBUTION OF THE LAYSAN ALBATROSS, DIOMEDEA IMMUTABILIS

Harvey L Fisher and James R. Fisher

The purpose of this paper is to portray the oceanic distribution of the Lay- san Albatross (Diomedea immutabilis) as indicated by records in the literature and by recoveries of birds banded by us. An attempt is also made to understand the reasons for the general distribution, as well as for changes associated with season and age.

The distribution of breeding colonies has been reviewed by Rice and Kenyon (1962), but no one has yet attempted an analysis of the pelagic range of the species. Present knowledge of the range is based upon incidental sightings at sea and recoveries of a few banded birds. Several publications list Laysan Albatrosses observed during transects of the North Pacific Ocean (for example, Clark, 1946; Hamilton, 1958; and Cogswell, 1946), and there are regional surveys as by Sanger (1965) off the coasts of Oregon and Wash- ington, by Kuroda (1955) in the northwest Pacific Ocean, and by King (1970) near the eastern end of the Hawaiian Islands.

METHODS

Data used in the analysis consisted of 109 recoveries of birds we banded, 53 published records of birds banded by others, and 113 sight records. Of the 109 recoveries, 64 birds were less than 3 years of age, 23 were 3 to 7 years old, and 22 were adults, including 19 known breeders. No significance can be attached to the relative numbers of the different age classes; we banded several times as many young as juveniles or adults. The sightings date from 1897 (Kaeding, 1905), but most are since 1945. Sight records prior to 1897 were not included because of possible confusion between records of the Laysan and the Short- tailed Albatross (D. albatrus) prior to that date. Attempts have been made to verify all records and to eliminate questionable sightings, hut data collected over such a long period and by so many different persons are subject to some error. Gathering of data over three-fourths of a century does have one advantage; it tends to smooth out annual vagaries such as Ingham (1959) and Tickell and Scotland (1961) noted in the annual patterns of dispersal of Giant Petrels iMacronectes f^iganteus) .

The paucity of verified records (276) spread over the millions of square miles of the North Pacific Ocean is troublesome and in several instanc(‘s makes impossible more than tentative statements. The problem is ameliorated, however, by multiple records in certain regions. Another (juestion is whether our data reflect the distribution of albatrosses or of persons recapturing albatrosses. A minimum of 69 p(*r cent and a tnaximum of 89 p(‘r cent of the recaptures were made by .Japanese tuna fishermen; 9 per cent were taken as scientific specimens. Tin; unc('rtainty in actual figures aris(‘s because tin* codes us«‘d by the IJ.S. Fish and Wildlife S(‘rvice to indicate the method of n'coveiw are not mutu- ally exclusive.

This possible confusion as to the distribution being indicated is pcuhaps immat(*nal, for we can assume that most tuna fish(*rmen are where tuna ai(* or wlnue tuna can be

7

8

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

expected. The evidence also indicates a probable similarity in the distribution of tuna and albatrosses in the North Pacific because both derive a large proportion of their food from squid. Our studies on Midway indicate that at least 90 per cent of the Laysan’s diet consists of squid. Nakamura (1965) reported that the main molluscan food item of skipjack tuna {Katsuwonas pelarnis) in 1957-59 was squid. Waldron and King (1963) found that in Hawaiian waters squid constituted 35 to 83 per cent of the food items of: skipjack tuna; yellow fin tuna {N eothunnus macropterus) ; and bigeye tuna i Parathunnus sibi) .

In analyzing the variation in oceanic distribution with age, three categories were estab- lished: young birds (to 3 years of age) ; juveniles (3-7 years) ; and adults (7 plus years) . Separation into these classes is based upon differences in behavior. Until they are three or more years old, the young Laysans are at sea and seldom return to the breeding colony (Fisher and Fisher, 1969). Between three and seven years the juveniles establish patterns of return, territories, and pairs. They visit the breeding colony at intervals between January and June. After the age of seven, the birds can be expected to be breeders, or within a year of breeding. They tend to return to the colony initially between November and February. Such differences in the relationship between albatrosses of different ages and the breeding grounds may affect oceanic distribution despite the remarkable flight powers of the albatross.

All oceanic records in the immediate vicinity of the Hawaiian Island breeding colonies were omitted. Records associated with these colonies add nothing to our knowledge of oceanic distribution, and their inclusion in analyses of latitudinal and longitudinal move- ments or even of distribution introduces a bias. Breeding albatrosses are of necessity restricted in their oceanic travels, although perhaps less than many other species.

Sea-surface temperatures are 20-year means ( 1947-66) furnished by R. A. Schwartlose of Scripps Institution of Oceanography.

RESULTS

All 276 records reported here lie within the limits of 8 to 59° N lat. and 132° E to 116° W long. Published reports of occurrences within these limits in- clude: 1) 25 sightings off Japan made by Kuroda (1955). Macdonald and Lawford (1954) and Wilhoft (1961) reported incidental sightings in the western and central Pacific area, as did Clark (1946), Dixon and Starrett (1952) and Hamilton (1958) ; 2) 11 sightings around the Aleutian Islands made by Kenyon (1961), Kuroda (1955), Macdonald and Lawford (1954), and Murie (1959) ; 3) Sight records off the west coast of North America by Sanger (1965), Love (1958), Willet (1913), Stager (1958), Thompson (1951), McHugh (1950), Kenyon (1950), Fredrich (1961), Holmes (1964), Kaeding (1905), and Yocum (1947) ; and 4) Occurrences around the Hawai- ian Islands and other eastern North Pacific islands were recorded by Fisher (1948), Munro (1945, 1946), Hanson (1959), Jensen (1949), Cogswell (1946), Eastman and Eastman (1958), and Thompson (1951).

Few Laysan Albatrosses have been found south of approximately 28° N, except around the breeding colonies which are essentially between 28 and 22° N. According to Amerson (1969), Laysans are “accidental on islands in the

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

9

Fig. 1. Records of Laysan Albatrosses in the North Pacific Ocean: sight records and handed birds more than 3 years of age.

northern Marshalls [approx. 13° N] probably at-sea-visitor.” However, he reported a Laysan Albatross at Mejit Island in the Marshalls, 10° 17' N and 172° 52' E. And there is the lone record at 8°. Dixon and Starrett (1952) stated that Laysans are “Noted south of 30th parallel only to eastward of Wake Island.” Baker (1951) in his review of Micronesian ornithology re- ported no records of Laysan Albatrosses in the Micronesian Islands.

The plot of all the sightings of birds of unknown age and of recaptures of our banded birds more than three years of age (Fig. 1) indicates that the primary oceanic range of the Laysan Albatross lies between 28 and 52° N and

Fig. 2. Records of Laysan Alliatrosses in the North Pacific Ocean: handed l)irds 3 or fewer years of age.

10

THE WILSON BULLETIN

March 1972 Vol. 81, No. 1

Number of records

11 7 7 10 22 14 21 19 21 12 30 28

-I ^ ^ 1 1 \ 1 1

— h- 70	-J- 51	— t- 54	-H \ 1 h- 61 66 68 59	-H h- 60 60	41	— h- 45	H 52
			Water Temperatures (F)
Aug.-	Sept.-	I 4-' O 0	Nov.- Dec.- Jan.- Feb.-	Mar.- April-	May -	June-	Ju ly-

Fig. 3. Distribution of Laysan Albatrosses in the North Pacific Ocean by latitude, month and surface water temperature: birds of all ages.

between 140° E and 120° W. This area includes 83 per cent of all records. Within this general range are four areas of concentration: 1) east of Japan and the Kurile Islands; 2) south of the western Aleutians; 3) off the west coasts of British Columbia and the United States; and 4) at sea around the eastern end of the Hawaiian Islands.

Certain regions contiguous to the general range have few or no instances of sightings or recaptures of Laysans: 1) the Sea of Okhotsk and the Sea of Japan; 2) the Bering Sea; 3) west of lower California; and 4) a vast circle of ocean between the eastern Aleutians and the Hawaiian Islands, centering at 40° N and 170° W. The only evidence of Laysan Albatrosses in the seas west of Japan consists of the recovery of a banded bird off the southwest coast

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

11

Fig. 4. Distribution of Laysan Albatrosses in the North Pacific Ocean by latitude, month and age: banded birds less than 3 years of age.

of Japan and another off the city of Okhotsk. Dement’ev et al. (1951) re- ported that the Laysan is a casual straggler “in Russia” but listed as evidence only one Laysan obtained in Kamchatkan waters. Kenyon ( 1950 ) reported no certain records in the Bering Sea, and Arnold (1948) and Kuroda (1955) saw no Laysans north of the Aleutians.

Records of birds three or fewer years of age are concentrated ( 87 per cent ) in an area east of Japan and roughly bounded by 30 to 45° N and 140 to 160° E ( Fig. 2 ) . With only two exceptions, all birds recaptured at a year or less of age have been between 35 and 45° N and 140 and 175° E.

Seventy-two per cent of the 3- to 7-year-old birds recaptured (23) were in this area, and 17 per cent were nearby. One bird in the Aleutians and one in Hawaiian waters represented the records most distant from the concentration.

Although the 22 banded adults were recajitured in widely sejia rated places, two-thirds were in this same area east of Jajian.

The mean latitude of all recaptures or sightings is 38° N. The monthly mean latitude of these records and the 20-year means of sea-surface temperatures at these mean latitudes are shown in Figure 3. From May through November the albatrosses are most freijuently north of 10° N and in temjieratures of 11 to

12

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Degrees of Longitude East-I-West

OOOO OOOOOOOO CO^lDCD h-OOr^CDlO'^COCN

Fig. 5. Distribution of Laysan Albatrosses in the North Pacific Ocean by longitude and month: birds of all ages.

61° L (except in August). From December into April the majority of the albatrosses are south of 35° N and in M ater temperatures of 59 to 68° F.

Albatrosses less than three years of age exhibit essentially the same seasonal shift in latitude <Fig. 4i. HoMCver. in their first 12 months (excluding, of course, approximately 5 months in the natal colony) the young birds are found mostly north of 38° N. Although they are south of this during the M'inter months at the beginning of their second year, none have been retaken beloM' 30° N. They shift north a full month ahead of the older birds ( NIarch versus April, Figs. 3 and 4, respectively) .

When the recaptures and sightings of all albatrosses are plotted by month and mean longitude (Fig. 5i, the average longitude of occurrence appears to be 176° E. Seasonal shifts are apparent. From May through August and from

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

13

November through January the majority of albatrosses are found between 150 and 160° E; from February to April and from September-October most are found between 155 and 175° W.

DISCUSSION

General Considerations. — The Laysan Albatross ranges in significant num- bers over most of the North Pacific Ocean north of 28° N and exclusive of the contiguous seas to the west and north. Continental land to the northwest, east, and northeast is an obvious barrier to this pelagic species. Islands to the north and west may function similarly, as is discussed later. But no land masses, even intermittent ones, delimit the southern extent of the range.

It is suggested that food is the most important single factor in determining the southern limits of the range and the relative abundance of Laysan Alba- trosses within the range. Such a positive correlation between the occurrence of oceanic birds and their food supply is not new, of course. Kurochkin (1963), for example, regarded food as a primary determinant of distribution for many species including several procellariiform species. Voous (1965) stated that the distribution of many antarctic birds corresponded with the dis- tribution of surface plankton. The papers of Jameson (1961) and of Gibson and Sefton (1959) on the Wandering Albatross iD. exulans) ^ of Thompson (1951) on the Black-footed Albatross [D. nigripes) and of J. Fisher (1952), Salomonsen (1965), and Brown (1970) on the Fulmar {Fulmarus glacialis) also emphasized the importance of plankton.

With these views in mind, and recognizing that Midway Laysans obtain 90 per cent of their food from plankton-feeding squid whose distribution is less known than that of plankton, it is logical to relate the occurrences of plankton and these albatrosses.

Four factors directly and indirectly affect the volume of plankton in an area — nutrients, water movements, water temperature, and water salinity. Water movements, as in currents, convergences of currents, and upwellings, affect available nutrients, temperature, and salinity. Any type of turbulence that mixes deep and surface layers of the sea increases the availability of nutrients in the surface layers and lowers temperatures, and both actions are basically favorable to the growth of plankton. It is also established ( Marr. 1956, for example) that the larger euphausicls upon which both s(juid and albatrosses feed occur primarily in the near-surface, eutrophic waters and are virtually limited to cold currents. Thus, Laysans, plankton, s(juid. certain temperatures, and turbulence should coincide in their distribution. J he avail- able data support this view.

Laysan Albatrosses occur most frecjuently and in larger numbers where water temperatures range from 10 to 65° F (King, 1970, said helow 72° F).

14

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Fig. 6. Major water masses and currents in the North Pacific Ocean.

although the temperature variance over their general range is 36 to 84° F. And this temperature zone of preference coincides with the zone of highest plankton productivity — between 28 and 35° N (King and Iversen, 1962). In this zone, the south edge of the North Pacific Current, they obtained ap- proximately 14,500 organisms per hour of trawling. North of 35° and in the Aleutian area the catch was 9,500 per hour. Between 28 and 5° N (Hawaiian and North Equatorial waters) King and Iversen reported less than 500 orga- nisms per hour.

Therefore, the southern limit of the range of the Laysan Albatross appears to be formed by a major drop in the abundance of food organisms. Tempera- ture may be the primary factor, but salinity may also be significant as Sanger (1970) suggested for the offshore waters west of North America. King (1970:96) stated that albatrosses “. . . tended to be most numerous over high- salinity water. . . .” but later “. . . it appears unlikely that surface salinity is a significant limiting factor in the distribution of sea birds in the study area.” However, Seckel and Yong (1970:191) noted that “Hawaii is located in the vicinity of a relatively high salinity gradient that delineates the boundary of the North Pacific Central Water.” The southern limit of the range of the Laysan thus appears to coincide with major, though gradual, changes in tem- perature and salinity. The limit can also be identified as the northern edges of the westward-trending Equatorial Current west of the Hawaiian Islands and of the North Pacific Equatorial Water between Hawaii and Central America (Fig. 6 ) .

The correlation of turbulence and records of Laysan Albatrosses, mentioned

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

15

earlier, is discussed in connection with the four major concentrations of birds

(p. 15-18).

Extralimital areas. — The few records west of Japan and the Kuriles, north of the Aleutians, and west of Lower California, despite the presence of numer- ous fishing boats (Ommaney, 1963) which have been the main source of records elsewhere, support the hypothesis that these areas are indeed outside the regular range of the species.

The Japanese, Kurile, and Aleutian islands may form partial barriers to the seas behind them, since these albatrosses do not normally approach land other than that of the breeding grounds. Even more significant is the fact that to reach these outer, fringe areas an albatross would have to pass through or over rich seas which are presumably attractive feeding grounds for the species. It is probable that the Tsushima Current just west of the Japanese islands, with its warmth and its low productivity (Sorokin and Koblentz-Mishke, 1958), is a major deterrent to the Laysan Albatross. The warm North Pacific Equa- torial Water west of Lower California may be a similar factor.

The Central North Pacific area, in the south edge of and just south of the Pacific Subarctic Water (Fig. 6), also lacks significant numbers of records. Less than 10 per cent are from this several-million-square-mile region which is devoid of strong currents, turbulences or upwellings and which is lower in plankton productivity than the regions to the north or south ( King and Iversen, 1962). The occasional records within this vast expanse of sea are in either the fringes of the Aleutian Current or the eastward extension of the North Pacific Current, and the birds may be assumed to be vagrants from the richer areas near the source of these currents. However, the scarcity of records may reflect the lesser human use of this region, despite our earlier discounting of this possibility for other parts of the North Pacific.

Areas of eoneentration. — Four major concentrations of albatrosses are evi- dent on Figure 1: 1) east of Japan; 2) south of the western Aleutians: 3) west of North America; and 4) around the larger, eastern islands of Hawaii.

There are reasons for the Laysan Albatrosses to be numerous in each of these areas. In each instance the conditions within the region of the concen- tration are generally constant from year to year within a certain range of coordinates. Bourne (1967:141) has noted that seabirds . . are normalK restricted to very limited sea-areas by strict preferences for certain ty})es and temperatures of surface water. . . .” Bailey (1968) noted this same phenom- enon among seabirds in the western Indian Ocean. The Lavsans apparently respond to permanently profitable foraging areas and do not utilize inter- mittent, locally enriched seas. The first reason may he the time and sj)ace lag between surface enrichment and the resulting production of suitable food. Secondly, the exigencies of amount of food, of lime, and of distanct‘ probably

16

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

make it impossible for the Laysans to rely on spotty food resources during such critical times of the year as the immediate pre-egg stage and the chick- feeding period ( Eisher, 1967 ). And last, species in which each member is so closely restricted to its own island and natal colony for breeding (Eisher, 19716) might he expected to be similarly related to its feeding areas.

The major concentration of records is east of Japan where the cold waters of the Oyashio Current collide with the warmer waters of the Kuroshio Cur- rent ( Eig. 6). The resulting turbulence and many eddies occur between 35 to 40° N and 140 to 160° E ( Seckel, 1970 ) and subside into the North Pacific Current still farther east. This region has been identified by Koblentz-Mishke (1965, Eig. 2) as having the greatest primary production in the North Pacific Ocean. More than a third (36 per cent) of the records are within the longi- tudinal limits of the turbulence and 30 to 45° N. Nearly half ( 45 per cent ) are here, if sightings and recaptures on the immediate fringes of the region are included. Kuroda (1955) said of the Laysan in the northwest Pacific “This species was most plentiful 180-200 miles eastsoutheast of Shinshiru Island, where we saw 14 birds in one day.” This is within the area of turbu- lence. Because of the turbulence and the consequent abundance of nutrients, plankton and squid occur year-round, with only minor seasonal changes, and the albatrosses find a plentiful food supply which they exploit constantly.

Although the second area of concentration, the western Aleutian region, includes only 7 per cent of the records, we believe the Laysan Albatross uses this area more extensively than the data may indicate. One reason is that Laysans tend to move into cooler waters during the summer months where plankton and presumably squid, are seasonally more abundant ( King and Iversen, 1962 ). Another is that the Aleutian Current courses northeastward through the islands while the Oyashio comes southward in the western part of this region. Such flows may produce major eddies and turbulence and rich waters around islands (Wyrtki, 1967), as has been demonstrated behind the islands of Johnston and Hawaii ( Planar, 1969 ). Larrance (1971) has noted the higher primary productivity in Aleutian coastal waters, as compared to areas to the south. Bourne (1963:836) also noted the higher productivity of seas around islands. The Aleutian region thus has all the features of a major seasonal feeding ground.

Along the west coast of North America (120-140° W and 30-50° N) is the third concentration (25 per cent of the records). Only two of our banded Midway birds have been found here. Eew of these North American records are inshore, and winter reports are negligible. It appears that the Davidson Current, which in winter flows north along the coast to about 48° N, rather than the continental shoreline, forms the eastern boundary of the Laysan range in these latitudes. The Davidson and the various northward extensions of

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

17

the North Pacific Equatorial Water are probably responsible for the absence of Laysan Albatrosses in the area south of 30° N and east of 145° W. The warmth of this water and its low oxygen and salinity are not typical of waters frequented by albatrosses.

Considering the amount of ship traffic, the number of scientific voyages, and the number of persons involved here in the past 75 years, it is apparent that this is a lesser concentration of Laysan Albatrosses than occurs east of Japan. Although these offshore waters are reportedly rich and upwellings are prevalent from March through July (Sverdrup et al., 1942) , there are a number of reasons why albatrosses may occur here less frequently. 1 ) The richness of these waters, as in the Aleutians, may be only seasonal and also considerably less because of the admixture of warmer, southern waters. 2) It has been suggested that Laysans avoid waters of low salinity, as found at least off Oregon (for example, Sanger, 1970). 3) Prevailing winds and water currents are not as advantageous for movement to and from this area as to the waters off Japan, and the distances are greater. 4) Another possibility about which we known little is that Laysan Albatrosses breeding on different islands may go to different parts of the ocean. Tickell found some indication of this segre- gation in young D. melanophris (1967) and in D. exulans and D. epomophora (1968). Our data indicate that Midway birds are for the most part (90 per cent) recaptured west of 180°. However, nearly half of all known oceanic records of Laysans (117 of 276 ) are east of the date line. This could be an indication that Laysan Albatrosses breeding in the western end of the Ha- waiian Chain move northwest to sea and the Japanese or Aleutian concen- trations and that albatrosses breeding farther to the east move into either the Hawaiian or North American concentrations referred to earlier. However, the picture is clouded by the fact that virtually all Laysans now breed east of 180°. 5 ) There is still another explanation for the lesser number of re- captures and sightings off the North American coast. At least 61, and perhaps 85 per cent since reporting techniques vary, of the recaptures were made by fishermen. The Japanese tuna fishermen of the western and central Pacific are predominantly surface fishermen of the open sea, and they recapture albatrosses on their long-line sets of tuna hooks or in their surround nets. Fishing in the far northeastern North Pacific, aside from a minor component of surface trolling for sport fishes, is closer inshore and directed more toward deeper dwelling fishes. Hence we should not expect as many albatross recap- tures in these waters as in the Japanese area, even though the number of birds and fishermen were the same as in the western Pacific.

Nevertheless, a significant number of Laysan Albatrosses have been re- corded in the cool offshore waters of North America, waters formed by the California Current or the lower limits of the Aleutian Current.

18

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Ten per cent of our Laysan records are from the fourth area, the vicinity of the five large, easternmost islands of the Hawaiian Chain, all of which are east and south of the primary sea range indicated earlier. Several factors, some factors unique, some perhaps common to other concentrations, may in- fluence the number of birds observed here. This concentration is no more than 800 miles from the breeding grounds, compared to 2,000 miles for the Japa- nese concentration, 1,200 for the Aleutian area, and the nearly 3,000 miles between the American concentration and the breeding grounds. There are major water eddies, turbulence and subsequent rich waters in the lee of these large Hawaiian islands (Manar, 1969), which may provide adequate feeding grounds in the otherwise generally unproductive Hawaiian waters ( King and Iversen, 1962). A large fleet of sportfishing boats and an active Audubon Society probably increased greatly the number of sight records. But the concentration of Laysan Albatrosses appears to he factual, and the primary reason may be that the North Pacific Current, which turns southward at about this longitude (Fig. 8), brings cooler, more productive water and Laysan Albatrosses with it. And the prevailing northeast trade winds, moving essen- tially parallel to this current, may further influence albatrosses to move south- ward into this region.

Because of this Hawaiian concentration it is necessary to extend the primary oceanic range of the Laysan south to 20° in the region of 155° W long.

Distribution by age. — Eighty-seven per cent of the 64 young birds were recovered in and around the Oyashio-Kuroshio turbulence east of Japan. The J3 per cent retaken outside this area are considered to be exceptions. Other than the two occurrences west of Japan, the sites of recapture are ones to which inexperienced birds may well have been transported by ocean currents (Figs. 2, 6).

The offshore waters of Japan constitute a nursery area, at least for Midway birds, in which the young remain until they begin their annual visits to the breeding colonies (Fisher and Fisher, 1969). Tickell (1967) reported similar concentrations of D. melanophris in their first 3 years of age. While adults are commonly recaptured on both sides of the Atlantic, Falkland Islands young go to the western Atlantic and South Georgia young move into the eastern Atlantic. Within a few months of fledging. Fulmars from West Green- land and St. Kilda go to the Newfoundland Banks (J. Fisher, 1952:325; J. Fisher and Lockley, 1954:138) . And Robertson has suggested recently ( 1969) that young Atlantic Sooty Terns {Sterna fuscata) may congregate in the Gulf of Guinea in Africa while those of the Pacific assemble in the central Philip- pines. It seems probable that as we learn more of the oceanic distribution of seabirds of different ages we shall discover that many species exhibit at least a partial segregation by age which reduces intraspecific competition.

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

19

Factors other than the richness of these Japanese waters are important in attracting young Midway Laysans. The area is closer to Midway than is any other known and abundant food source, and it is more easily reached by inexperienced birds with relatively poor powers of flight. As the young bird leaves Midway it is subjected to westward trending water currents (Fig. 6) and generally consistent winds from the east-northeast sector (U.S. Naval Oceanog. Office, 1966). As it rests or feeds on the water it is carried west by south; its weak flight is greatly affected by the wind, and the bird in the air also drifts southwest. Should it drift too far to the south, another factor increases the rate of its westward trend — the westerly flow of the Equatorial Current. One might compare this latter current to a drift fence along which the young birds move until they reach the Kuroshio Current which then moves them north to the nursery area. Conditions in these critical months of July- September may be particularly advantageous to these birds. King and Hida (1957) obtained their largest plankton catches then and reported that surface catches at night were 1.5 times as large as in daytime. These data may mean that not only are the squid attracted to the abundant food, they are attracted presumably in greater numbers to the surface and at night. This, of course, means greater accessibility for the initial foraging efforts of the young birds. The abundant supply and greater availability of suitable food organisms to inexperienced birds could be expected to hold the birds there until initial sexual development stimulates them into migratory patterns.

The 23 juveniles were recaptured at sites more widespread than those of young birds, but 72 per cent were taken in the rectangle described as the nursery area. An additional 17 per cent were retaken nearby. Thus even juveniles, most of which have already made one or more trips to the Midway breeding colonies, return to the offshore waters of Japan.

The 22 adults were widely scattered at the time of recapture, but 68 per cent were secured off Japan. One was recovered off California, one in the Hawaiian area, and four in the Aleutians. Since 19 of these adults were known to have bred on Midway, it is plausible to suggest that a majority of Midway’s breeders return at times to feed in this area. It is unfortunate that we did not know the current breeding status of each of these adults, for then we might know whether they move between the breeding colony and this area during the nesting season. Other studies in progress show that Midway adults feeding chicks may fly east and north at least 1500 miles.

It is evident that Midway birds of all ages feed in the turbulent convergence of currents east of Japan, that a high percentage remain there for the first three years of life, that juveniles return there as they initiate j)eriodic visits to the breeding colony between the ages of 3 and 7 years, and that many of Midway’s breeding Laysan Albatrosses feed there at least from time to time.

20

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Fig. 7. Distribution of Laysan Albatrosses in the North Pacific Ocean by latitude, longitude and month: mean locations of birds of all ages.

Seasonal distribution. — Figure 5 illustrates the occurrence of four periods of longitudinal shift: eastward (into west longitude) in January and August and westward in April and November. In Figure 3 three latitudinal shifts are shown: northward in April-May and August-September and southward in October— December. Figure 7 is an attempt to portray graphically the com- bined results of the longitudinal and latitudinal shifts for birds of all ages. We believe the directions of movement are accurately shown, but the extent of the shifts is probably unduly influenced by the use of averages including extreme records.

Starting in August, the first month of the year in which no Laysans are involved with reproduction, there is a shift in concentration toward the north- east along the North Pacific Current and perhaps into Pacific Subarctic Water. This continues until October. The birds move from water near 70° F into 50-degree and presumably more productive water. In November the birds are back in the central reaches of the North Pacific Current where the sea-surface temperatures are now in the low 60s. In December and January the albatrosses are near 30-35° N and 170° E where temperatures gradually climb into the upper 60s. The birds then move eastward, with no change in latitude, to 165° W where water temperatures are 60° or below. The birds are probably in the same water as in December-January, water that has cooled as it moved eastward in the southern fringe of the North Pacific Current. In April the move is west and north to the 40-50° F waters east of Japan, waters brought to this temperature by the cold waters of the Oyashio Current. The albatrosses remain until late July. In all the above statements we are referring to “aver- age moves,” not necessarily to movements of the entire populations.

It might be suggested that Figure 7 reflects at least in part the movements to and from the breeding grounds. And one might justifiably speculate, on the

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

21

Fig. 8. Distribution of Laysan Albatrosses in the North Pacific Ocean: mean location of birds 4 or fewer years of age.

I basis of this study and published information on the biology of the Laysan Albatross, that: 1) virtually all of the birds less than 3 years of age are in the May-July plotting; 2) the August to November roundtrip is made mostly by breeders, along with some older juveniles; and 3) the December to May I plotting consists primarily of breeders, with 3- to 7-year-olds contributing to the March to May portion (Fisher and Fisher, 1969). It is perhaps equally logical, with regard to the December-May period, to suggest that incubating 1 birds (November-January) with their longer periods of relief from nest duties ( 8-20 days, Fisher, 1971a ) can subsist in less productive areas. But when the I food requirements of the chick are added to those of the foraging parents and I when the total time for travel and foraging between chick feedings averages j 2 days (February— April) the adults shift eastward to presumably better food sources. Also during this time, nestling mortality has released additional parents and the chicks are fed smaller amounts and less frequently, as evi- denced by the fact that their weight declines after May (Fisher, 1967 ) . These factors may permit failed or even current breeders to move northwest in April and May.

[ However, we believe that the movements of breeding birds are not signifi- cantly involved in the conclusions to be reached from study of Figure 7. First, the Laysans four or fewer years of age show the same directional shifting (Fig. 8), and most of these are not yet making their periodic visits to the natal colonies. Second, only 37 per cent of the handed portion of the sample is of an age to visit the colony or to breed. And third, Laysan Albatrosses in short periods of time are capable of traversing distances greater than those involved in the latitudinal shifts. Kenyon and Rice (1958), for example.

I showed that Laysans removed from eggs or young chicks home at the rate of I nearly 200 miles ]>er day.

22

THE WILSON BULLETIN

March 1972 Vol. 81, No. 1

Furthermore, evidence from the literature, though scanty, tends to support the concept of these movements and concentrations. King (1970:9) in record- ing Laysan Albatrosses at sea east of the Hawaiian Islands and between 10 and 25° N and 148 and 150° W, gave their status as “Uncommon visitor February-April, rare or absent May-January.” He also noted that the decline in numbers in April was accompanied by . a contraction of range to the northern end of the study area.”

Amerson (1969:293) wrote of the Laysan “Found at sea normally south to 15° N in the Central Pacific during the breeding seasons. . . [italics mine], thus indicating his belief in a seasonal shift. However, his reasons were not stated, and we now know that Laysans seldom penetrate this far south.

The northward movement of the albatrosses in summer may well be related to the seasonal northward advection of southern water into the latitudes of the Hawaiian Islands (Seckel and Yong, 1970). This advection brings warm water of lower salinity into the southern part of the albatross range.

The observations of McHugh (1950), Holmes (1964 ), and Sanger (1970) that Laysan Albatrosses were more abundant off the North American coast “in winter” lends some strength to our view of eastward shifts in February and March.

Hamilton’s observations (1958) during a June transit west to east and between 35 and 48° N tend to substantiate the midsummer concentration around the 40th parallel.

Austin and Kuroda (1954) believed that the Laysan was a regular visitor off the Pacific coasts of Honshu and Hokkaido from early spring to late autumn, and Kuroda (1957) said that it arrives off Japan in March. We presume he meant greater numbers were present there at that time. It is probable that this influx is of young birds which we think move north in March (Fig. 4). He further wrote (p. 448) of a “post-breeding movement northward in spring.” This may correspond to the April shift or perhaps the greater occurrence off Japan in July. He indicated this was a movement with the northward trend of the warm Bonin Island Air Mass. We regard it as a seasonal movement away from the increasingly warm waters of the Kuroshio Current. The average water temperatures in July drop from 81° F at 30° N to 53° F at 45° N in these longitudes, and Kuroda had earlier ( 1955) stated that the Laysan “. . . seems to avoid water above 13° C.” In 1960 he indicated that the Laysans congregated off the Kuriles in June and July at sea temperatures slightly above 40° F, which is in basic agreement with the data in Figure 3. He did not find many Laysans in either the colder or warmer waters of this region. Szijj (1967) noted that albatrosses in southern seas were most numerous at water temperatures between 6 and 13° C.

The implication is that Laysans seek out these temperatures, for one reason

Fisher and Fislier

DISTRIBUTION OF LAYSAN ALBATROSS

23

or another, probably food. It is probable that the Laysan adult, like the Fulmar (J. Fisher, 1952:325; Brown, 1970; and Salomonsen, 1965 ), regularly moves to a food source that is adequate, accessible and predictable on a time and place basis. This seasonal phenomenon is also reported for the Wandering Albatross (Tickell, 1968; Gibson and Sefton, 1959, 1960; and Jameson, 1961). Dixon Q933) and Tickell and Gibson (1968) believed that Wan- derers, especially those of pre-breeding age, had a regular migratory path between South Georgia and the sea off New South Wales. And Gibson ( 1963: 216) has said of the Wanderer: . . when free from breeding commitments

at their home islands, these birds returned regularly to an assured natural food supply, contrary to the generally held conception of a free-ranging ocean wanderer unbound by conventional migrations.” The Royal Albatross regu- larly moves between Campbell Island and South America (Dixon, 1933; Tickell, 1968). Falla (1963) noted that “several albatrosses” breeding in the Subantarctic moved into colder waters in late summer, a shift perhaps com- parable to the September-October move of Laysans.

Our data (Fig. 7) do not support Bourne’s view (1967) that seabirds tend to move clockwise around anticyclonic stationaries in the middle latitudes of . the Northern Hemisphere.

SUMMARY

All 276 oceanic records of Laysan Albatrosses are within the limits of 8 to 59° N and 132° E to 116° W in the North Pacific Ocean. The primary range, however, is between ^ 28 and 52° N and 140° E-120° W.

The northern boundary of their distribution is the Aleutian Islands and the relatively non-productive waters of the Bering Sea. The Kurile and Japanese islands, along with the warm Tsushima Current, constitute a western harrier. The North American continent with its warm inshore Davidson Current forms the eastern limit. The southern border is marked by warm equatorial waters of low salinity and low productivity.

Within these limits Laysan Albatrosses tend to congregate in four regions: 11 east of Japan (35-40° N and 140-160° E) ; 2) south of the western Aleutians (50° N and 165° E-175° W); 3) off the west coast of North America (30-50° N and 120-135° W); and 4) near the large, eastern islands of Hawaii (20° N and 150 — 160° W).

The Japanese region serves as a nursery foraging area for birds fewer than 4 years of age; seldom are they recaptured elsewhere. However, older juveniles and adults from Midway also return there to feed.

j There is evidence of seasonal shifts in concentrations; the birds move east in JanuaiT and August, west in April and November; they move north in April-May and in August- : .September, south in October-December, In general these movements are associated with j changes in surface water temperatures.

I Laysan Albatrosses tend to be associated with turbulent seas, eddies and currents;

^ the birds most frecpiently are in water tem|)eratures of 40 to 65° F. Such waters are I generally most productive, and it is sugg(‘sted that food is the ])rimary deteiininanl of the ! Laysan’s distribution.

24

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

ACKNOWLEDGMENTS

It is impossible to thank individually the many persons who have so obviously con- tributed to this study, but our appreciation is no less sincere. We do want to note partic- ularly the contributions of the unknown men of the Japanese tuna fleet, for without their cooperation in recapturing significant numbers of albatrosses at sea this study would have been impossible.

Original financial support for the banding of nestling birds in 1961-63 came from the Office of Naval Research (ONR 3479 (00)). Continuing support is being furnished by the Office of Graduate Studies and Research, Southern Illinois University, Carbondale.

LITERATURE CITED

Amerson, a. B., Jr. 1969. Ornithology of the Marshall and Gilbert Islands. Atoll Res. Bull. No. 127.

Arnold, L. W. 1948. Observations on populations of North Pacific pelagic birds. Auk, 65:553-558.

Austin, 0. L., Jr., and N. Kuroda. 1954. The birds of Japan, their status and distri- bution. Bull. Mus. Comp. Zook, 109:277-637.

Bailey, R. S. 1968. The pelagic distribution of sea-birds in the western Indian Ocean. Ibis, 110:493-519.

Baker, R. H. 1951. The avifauna of Micronesia, its origin, evolution, and distribution. Univ. Kansas Publ. Mus. Nat. Hist., 3:1-359.

Brown, R. G. B. 1970. Fulmar distribution: a Canadian perspective. Ibis, 112:44-51. Bourne, W. R. P. 1963. A review of oceanic studies of the biology of seabirds. Proc.

13th Internatl. Ornithol. Congr. Ithaca 1962, 2:831-854.

Bourne, W. R. P. 1967. Long-distance vagrancy in the petrels. Ibis, 109:141-167. Clark, T. 0. 1946. Bird observations to San Francisco aboard a transport. Elepaio,

7:1-3.

Cogswell, H. L. 1946. Bird study aboard a transport to the Western Pacific. Elepaio, 6:46-48, 53-54, 62-63.

Dement’ev, G. P., R. N. Meklenburtsev, and A. M. Sudilovskaya. 1951. Birds of the Soviet Union. Vol. 2.

Dixon, C. C. 1933. Some observations on the albatrosses and other birds of the southern ocean. Trans. Roy. Canadian Inst., 19:117-139.

Dixon, K. L., and W. C. Starrett. 1952. Offshore observations of tropical seabirds in the Western Pacific. Auk, 69:266-272.

Eastman, W., and K. Eastman. 1958. Bird study in the Hawaiian Islands is a stimu- lating experience. Elepaio, 19:1-5.

Falla, R. A. 1963. Distribution patterns of birds in the Antarctic and high-latitude Subantarctic. Biol. Antarctique, Prem. Symp., Sept. 2-8, Paris, 1962, pp. 367-376. Fisher, H. I. 1948. Laysan Albatrosses nesting on Moku Manu Islet off Oahu, T. H. Pacific Sci., 2:66.

Fisher, H. I. 1967. Body weights in Laysan Albatrosses, Diomedea immutabilis. Ibis, 109:373-382.

Fisher, H. I. 197U/. Incubation hatching, and associated behavior in the Laysan Alba- tross, Diomedea immutabilis. The Living Bird, 10:19-78.

Fisher, H. I. 19716. Experiments on homing in Laysan Albatrosses, Diomedea im- mutabilis. Condor, 73:389-400.

Fisher and Fisher

DISTRIBUTION OF LAYSAN ALBATROSS

25

Fisher, H. L, and M. L. Fisher. 1969. The visits of Laysan Albatrosses to the breeding colony. Micronesica, 5:173-221.

Fisher, J. 1952. The Fulmar. Collins, London.

Fisher, J., and R. M. Lockley. 1954. Sea-birds. Houghton Mifflin Co., Boston. Fredrich, L. a. 1961. An occurrence of the Laysan Albatross on the northwestern coast of Oregon. Condor, 63:506.

Gibson, J. D. 1963. Third report of the New South Wales albatross study group (1962) summarizing activities to date. Emu, 63:215-223.

Gibson, J. D., and A. R. Sefton. 1959. First report of the New South Wales albatross study group. Emu, 59:73-82.

Gibson, J. D., and A. R. Sefton. 1960. Notes on some albatrosses of coastal New South Wales. Emu, 60:128-130.

Hamilton, W. J. HI. 1958. Pelagic birds observed on a North Pacific crossing. Condor, 60:159-164.

Hanson, C. 1959. [Field notes]. Elepaio, 19:67.

Holmes, R. T. 1964. Notes on the occurrence of the Laysan Albatross near the Cali- fornia Coast. Condor, 66:302-303.

Ingham, S. B. 1959. Banding of Giant Petrels by the Australian National Antarctic Research Expeditions, 1955-58. Emu, 59:189-200.

Jameson, W. 1961. The Wandering Albatross. Rev. ed.. Doubleday and Co., New York. Jensen, S. 1949. Field notes: from Johnston Island. Elepaio, 9:66.

Kaeding, H. B. 1905. Birds from the west coast of Lower California and adjacent islands. Condor, 7:105-111.

Kenyon, K. W. 1950. Distribution of albatrosses in the North Pacific and adjacent waters. Condor, 52:97-103.

Kenyon, K. W. 1961. Birds of Amchitka Island, Alaska. Auk, 78:305-326.

Kenyon, K. W., and D. W. Rice. 1958. Homing of Laysan Albatrosses. Condor, 60:3-6. King, J. E., and T. S. Hida. 1957. Zooplankton abundance in the Central Pacific, Part II. Fish. Bull., 57:365-395.

King, J. E., and R. T. B. Iversen. 1962. Midwater trawling for forage organisms in the Central Pacific 1951-1956. Fish Bull., 210:271-321.

King, W. B. 1970. The trade wind zone oceanography pilot study. Part VII: Obser- vations of sea birds March 1964 to June 1965. U.S. Fish and Wildl. Serv. Spec. Sci. Rept. Fish. No. 586.

Koblentz-Mishke, 0. J. 1965. [The magnitude of primary production of the Pacific Ocean]. Oceanography (Russian), 5:325-337.

Kurochkin, E. N. 1963. [Distribution of some seabirds in the North Pacific.! Zool. Zhur., 42:1223-1231.

I Kuroda, N. 1955. Observations on pelagic birds of the northwest Pacific. Condor, 57: 290-300.

' Kuroda, N. 1957. A brief note on the pelagic migration of the Tuhinares. YamashinaY

I Inst. Ornithol. and Zool. Misc. Rept. No. 11:436-439.

I Kuroda, N. 1960. Analysis of sea bird distribution in the Northwest Pacific Ocean.

; Pacific Sci., 14:55-67.

j Larrance, j. D. 1971. Primary production in the mid-subarctic Pacific region, l%6-68.

I Fishery Bull., 69:595-613.

^ Love, C. M. 1958. Preliminary report: Brown Bear Cruise 188 off tlie coasts of \\ asli- j ington and Vancouver Island, 19-28 February 1958. IJniv. Washington Dept. Oceanog.

I Ref. 58-16. (ditto).

26

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Macdonald, J. D., and P. A. Lawford. 1954. Sight records of birds in the Pacific: compiled from the bird log kept during the recent cruise of H.M.S. Challenger. Emu, 54:7-28.

Manar. T. a. 1969. Progress in 1967-68 at the Bureau of Commercial Fisheries Biologi-

cal Laboratory, Honolulu. Bur. Comm. Fish. Circ. 321.

McHugh, J. L. 1950. Increasing abundance of albatrosses off the coast of California. Condor, 52:153-156.

Marr, j. W. S. 1956. Euphausia superba and the Antarctic surface currents. Norsk. Hvalfangst-Tid., 45:127-134.

IMunro. G. C. 1945. The Laysan Albatross on Kauai. Elepaio, 5:70.

Munro, G. C. 1946. [Field notes.] Elepaio, 6:79.

Murie. 0. J. 1959. Fauna of the Aleutian Islands and Alaska Peninsula. N. Amer.

Fauna No. 61.

Nakamura. E. L. 1965. Food and feeding habits of skipjack tuna, Katsuwonus pelamis, from the Marquesas and Tuamotu Islands. Trans. Amer. Fish. Soc., 94:236-242.

Ommanney, F. D. 1963. The Fishes. Time Inc., New York.

Rice. D. W.. and K. W. Kenyon. 1962. Breeding distribution, history and populations of North Pacific albatrosses. Auk, 79:365-386.

Robertson. W. B., Jr. 1969. Transatlantic migration of juvenile Sooty Terns. Nature, 223:632-634.

Salomonsen. F. 1965. The geographical variations of the Fulmar (Fulmarus glacialis) and the zones of marine environment in the North Atlantic. Auk, 82:327-355.

Sanger, G. A. 1965. Observations of wildlife off the coast of Washington and Oregon in 1963, with notes on the Laysan Albatross {Diomedea immutabilis) in this area. Murrelet, 46:1-6.

Sanger, G. A. 1970. The seasonal distribution of some seabirds off Washington and Oregon, with notes on their ecology and behavior. Condor, 72:339-357.

Seckel, G. R. 1970. The trade wind zone oceanography pilot study. Part VHI: Sea- level meteorological properties and heat exchange processes July 1963 to June 1965. U.S. Fish and Wildl. Serv. Spec. Sci. Rept. Fish. No. 612.

Seckel, G. R., and M. Y. Y. Yong. 1970. Harmonic functions for sea-surface tempera- tures and salinities, Koko Head, Oahu, 1956-69, and sea-surface temperatures, Christmas Island, 195T-69. Fish. Bull., 69:181-214.

Sorokin, U. L, and 0. J. Koblentz-Misciike. 1958. The primary production of the Sea of Japan and the part of the Pacific adjoining Japan, during spring 1957. Papers Acad. Sci. U.S.S.R., Moscow, 122:1018-1020.

Sverdrup, H. U., M. W. Johnson, and R. H. Fleming. 1942. The Oceans. Prentice- Hall, New York.

Stager. K. E. 1958. A record of the Laysan Albatross from Southern California. Condor, 60:404^405.

Szijj, L. 1967. Notes on the winter distribution of birds in the western antarctic and adjacent Pacific waters. Auk, 84:366-378.

Thompson. D. Q. 1951. Notes on distribution of North Pacific albatrosses. Auk, 68: 227-235.

Tickell. W. L. N. 1%7. Movements of Black-browed and Grey-headed Albatrosses in the South Atlantic. Emu, 66:357-367.

Tickell. W. L. N. 1968. The biology of the great albatrosses, Diomedea exu/ans and Diomedea epomophora. Antarctic Res. Ser., 12:1-55.

Fisher and Fislier

DISTRIBUTION OF LAYSAN ALBATROSS

27

Tickell, W. L. N., and J. D. Gibson. 1968. Movements of Wandering Albatrosses. Emu, 68:7-20.

Tickell, W. L. N., and C. D. Scotland. 1961. Recoveries of ringed Giant Petrels, Macronectes giganteus. Ibis, 103:260-266.

United States Navy. 1966. Pilot chart of the North Pacific Ocean, No. 1401, July 1966. U.S. Naval Oceanog. Office, Wash., D.C.

Waldron, K. D., and J. E. King. 1963. Food of skipjack in the Central Pacific. Proc.

World Sci. Meet. Biol. Tunas and Related Species, FAO Fish. Rept. No. 6:1431-1457. WiLiiOFT, D. C. 1961. Birds observed during two crossings of the Pacific Ocean. Con- dor, 63:257-262.

WiLLET, G. 1913. Pelagic wanderers. Condor, 15:158.

Wyrtki, K. 1967. The spectrum of ocean turbulence over distances between 40 and 1000 kilometers. Deut. Hydrog. Z., 20:176-186.

Yocom, C. F. 1947. Observations on bird life in the Pacific Ocean off the North American shores. Condor, 49:204-208.

Voous, K. H. 1965. Antarctic birds. Monogr. Biol., The Hague, 15:649-689.

Unfortunately, we did not know of the significant report by V. P. Shuntov (Zook Zhurnal, 47:1054-1064, 1968) until after our paper was in press. His study of the distri- bution of the Laysan Albatross, based on approximately 800 records obtained at sea over a 10-year period, is in nearly complete agreement with ours in relation to basic distri- bution and its correlation with land masses, oceanic currents and primary productivity, to major areas of concentration, to temperature preferences, and to seasonal movements. The major difference is that Shuntov found a summer and fall penetration of the western and southern parts of the Bering Sea by significant numbers of Laysans and lesser num- bers throughout the Sea of Okhotsk in these seasons. We thank Dr. Isaac Shechmeister of Southern Illinois University for translating this article for us.

DEPARTMENT OF ZOOLOGY AND SCHOOL OF MEDICINE, SOUTHERN ILLINOIS UNI- VERSITY, CARBONDALE, ILLINOIS 62901, 9 JUNE 1971.

HABITS OF THE CRIMSON-CRESTED WOODPECKER IN PANAMA

Lawrence Kilham

I studied the Crimson-crested Woodpeckers {Campephilus [Phloeoceastes] melanoleucos) in the Panama Canal Zone in February 1965 and from November 1970 to February 1971, a period which included the end of the rainy season when nesting began and the onset of the dry season when young were fledged. The behavior of this species resembles that described by Tanner (1942) for the Ivory-billed Woodpecker {Campephilus principalis) and has not hitherto been the subject of any detailed reports, with exception of notes by Short (19706), as far as I am aware. In Short’s opinion (1970a), Phloeoceastes should be merged in Campephilus and I have adopted this terminology.

While the aim of present studies was to learn as much as possible about the total behavior of C. melanoleucos, the problems raised by its similarity in size and coloration to the sympatric Lineated Woodpecker (Dryocopus lineatus) were kept in mind, thanks to the ideas of Cody (1969 ) on why this parallelism exists. Actual field observations, however, failed to support his interesting theories, which are dealt with in greater length in a final discussion.

STUDY AREAS

I studied Crimson-crested Woodpeckers in five localities of which three, Madden Forest Reserve, Limbo Hunt Club, and Barro Colorado Island (BCI), were, for the most part, mature forests. Of the other two areas, one was of second growth forest 10 to 20 m in height at Cardenas Village where I lived and the other at Frijoles, an area under partial cultivation opposite Barro Colorado Island. The Crimson-crested and Lineated Wood- peckers were sympatric in all five of these localities, as indeed they are in much of South America.

METHODS OF COMMUNICATION Instrumental Expressions

Drumming. — Drumming is typically a strong blow followed by short, weak, vibratory roll, “DA-drrr.” Such bursts usually come at a rate of one to two per minute, three per minute being a fast rate. This drumming serves a num- ber of functions. Single “DA-drrs,” given occasionally throughout the day, enable members of a pair to keep in touch as they travel through woods together; duets of them continuing for periods of up to 20 minutes may occur at the height of courtship and just prior to copulation; while louder drumming, delivered against a resonating stub, is usually related to territorial disputes and assertions of dominance. This abbreviated drumming of C. melanoleucos.

28

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

29

Fig, 1. Female Crimson-crested Woodpecker drum-taps at rim of recently completed j nest hole as mate approaches.

which at times can be no more than a single “DA,” appears to be the same as that described by Tanner (1942) for the Ivory-billed and by Short (1970a and b) for the Magellanic (C. magellanicus) and other Campephilus woodpeckers ! in South America. Although both sexes of C. melanoleucos drum, males drum i far more than females during the nesting season.

Pileated Woodpeckers (D. pileatus) (Kilham, 1959) strike a sharp rap with I their bills against any surface they happen to be on when nervous or excited. According to Bock (1963), the genus Campephilus is an offshoot, phylogeneti- cally, of Dryocopus and one might wonder whether the single drumming of Campephilus is not derived from the rapping of the latter genus. An observa- tion of Tanner (1942) on the Ivory-billed Woodpecker is of interest here, I for he noted that “The adults always were disturbed and excited whenever I I first found a nest.” In addition to giving calls they “often double-rapped or pounded on stubs or limbs of the nest tree and nearby trees.” Thus, the drumming was done in the same context as the rapping would l>e done for j D. pileatus.

' Drum-tapping. — As discussed in a preceding report (Kilham, 1959), most

j woodpeckers tap at a regular and countable rate at the time of excavation of I a nest hole. Pileated Woodpeckers, on the other hand, have a more rapid

30

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Fig. 2. Female Crimson-crested Woodpecker backs down from nest hole to touch hills with her mate who reaches up toward her.

roll taking the place of tapping, which I have called “drum-tapping.” Crimson- crested Woodpeckers drum-tap in the same manner as Lineated and Pileated Woodpeckers, both at the rim of the nest hole at time of excavation ( Fig. 1), but also down inside the nest at time of relief at the nest, a habit also described by Sielmann (1958) for the European Black Woodpecker {D. martins).

Wing noises. — Crimson-crested Woodpeckers can fly silently. They often, however, make a heavy sound, even in flying short distances, that doubtless keeps each member of a pair informed when the other moves and in what direction. Heavy wing noises are a feature of conflicts.

Displays

Bill-touching. — At times of most active courtship, the woodpeckers of a pair may come close to one another, crests raised and even curled forward, then fence gently, making contact for roughly half the length of their bills. This interest in bills at time of courtship may be related to the way a male pecks down at the bill of the female while copulating. Ivory-billed Wood- peckers touch bills in courtship according to Tanner (1942) and Allen and Kellogg (1937) wrote that as a female climbed up a pine toward her mate “he bent his head downward and clasped bills with her.” Although I noted a

Lawrence

Killiam

CRIMSON-CRESTED WOODPECKER

31

NEST NOV. 1970 NEST JAN. 1971

® KNOLL (Courtship)

® CONFLICTS WITH RIVAL MALE JAN. 1971

NEST EXCAVATION D.LINEATUS

Fig. 3. Territory of Pair A of Crimson-crested Woodpeckers at Cardenas Village.

I similar bending down on several occasions, as illustrated in Figure 2, I never I: observed bill clasping with C. melanoleucos.

Vocalizations

Alarm notes. — Notes of moderate disturbance made, for example, when one f comes too close to a nest excavation are ca and ca-wa-rr-r often repeated. A sharp, high-pitched ca given alone was the only vocalization heard in several conflicts between males. Shrill, piping put put puttas given by both males and females are expressive of high excitement. These may be kept up for minutes on end. On the whole, however. Crimson-crested Woodpeckers are relatively silent birds, giving way to alarm notes with far less frequency than the related Lineated Woodpeckers.

Intimate notes. — These low notes are expressive of closeness of pair bonds, , being given just prior to coition and at times when one partner relieves the other at excavating. Variations include ivuk ivuk, ivrr wrr. wun ivun. and uh I uh among others.

I Main breeding:^ call. — A tree-frog-like kwirr kwirr-ah or squeer squeer-ah- ' hah.

'' Comparisons to other species. — Short (19706) records a three-noted call j wink-at-chew for C. melanoleucos in Argentina. Vocalizations of C. magellani- I CHS (Short, 19706) are given in more detail and here the double-noted calls, j ivieeer and kee-argh (harsher, more drawn out) appear somewhat similar to I the kwirr-ah and ca-wa-rr-r notes described above for C. melanoleucos.

32

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

COURTSHIP AND COPULATION

The following activities of Crimson-crested Woodpeckers, as well as the excavation of nest holes, with exception of Pair E, took place in the rainy season.

Pair A. — The woodpeckers of Pair A frequently came to a knoll IFig. 3) at the edge of woods by Cardenas Village for early morning courtship and preening, the male (MA) always to a special place on his tree and the female (FA) to hers. These trees were about 14 m tall and 10 m apart. Here the two began a duet of drumming at 06:10 on 22 November 1970. After 15 minutes of low bursts, one or two per minute from each, MA flew to FA’s tree and I heard low notes then as well as five minutes later when FA moved out onto a horizontal limb. Here she crouched low in a crosswise position as MA approached. He mounted in full coition, pecking gently down at her bill four or five times as he gradually fell to the left in establishing cloacal contact.

This copulation suggested that the pair must have a nest nearing completion and on 26 November I was led to it at 16:00 by the sound of FA excavating. She took alarm and flew out, then drummed on nearby trees as if disturbed. 1 returned again in the late afternoon. The woodpeckers were feeding in trees close-by when, at 17:20, FA flew to the hole and clung to its lower rim. When MA alighted a meter below, she drum-tapped on the rim of the hole (Fig. 1) and as he hitched upward, she moved down to meet him, bending over to fence bills (Fig. 2) as he stretched upward. All now looked well for actual nesting. The way a pair of Collared Aragaris ( Pteroglossus torquatus) dispossessed the woodpeckers of their completed hole 20 minutes later is described beyond.

Five days later, on 2 December, the two Woodpeckers, now without a nest, had returned to the knoll (Fig. 3). MA drummed at an uneven rate of 11 times in 10 minutes at 06:23, but FA, on her tree, did not respond. When he flew to her, however, the two fenced lightly with their bills. FA moved on a less horizontal part of the trunk and crouched low, but MA, taking no apparent interest in this invitation to coition, flew away. On the fol- lowing morning MA drummed again, with only one burst in reply from FA. The two preened in leisurely fashion for 35 minutes, then left. It thus appeared on succeeding days that, with loss of their nest, the woodpeckers gradually lost interest in courtship.

I now felt there would be little to observe further with this pair when on 15 December I heard kwirr-kwirra notes by the knoll when MA alighted on the trunk of a slender tree, to be joined almost immediately by FA, both being at the same level as they bent heads together to touch bills. The crests of the two were raised to the full and curled forward. They returned to their original positions, only to bend heads together on the other side of the trunk to fence again. Both now flew to the knoll and drummed a brief duet before a longer period of leisurely preening. Had the woodpeckers found a new nesting site as the renewal of courtship activities suggested? 1 had no further indication of this until 1 January when at the comparatively late hour of 18:45 1 saw the pair on a bare tree near the knoll in full copulation. From here, after feeding for five minutes, they flew east. With this direction as a clue I was able to find their nest, in which they were to hatch young, a few days later.

Pair B. — At 06:10 on 26 November drumming led me to find a male Crimson-crested Woodpecker on the limb of a dead tree above Madden Forest. A female Lineated Wood- pecker alighted briefly on the same limb but after she had left, the female Crimson-crested, whom I could not see, drummed five or six times, then flew to her mate. I heard low notes, then witnessed full coition lasting possibly 10 seconds. Afterwards the two birds

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

33

preened in leisurely fashion for 10 minutes, then flew to feed in the forest. At 07:00 the female gave a single kwirr-a. After a pause, the male flew to her, there were more low notes, and a second copulation, not well seen, followed. This was 35 minutes after the first one.

EXCAVATION OF NEST HOLES

Trial nest stub. — Not all pairs of Crimson-crested Woodpeckers were able to find suitable nest stubs. Pair C, for example, had already tried to excavate one stub when, at 09:25 on 24 December, I found the male carving an entrance in a second one. His mate remained nearby making querulous wer wer and wiik wuk notes until she took over the excavating at 09:40. Her interest, however, waned after five minutes and she circled up the stub, pecking here and there as if to test the nature of the underlying wood, which was probably too hard to excavate, before flying away. She thus appeared more critical of the stub as a nesting site than her mate. Little further progress was ever made with entranceway and by early Eebruary Pair C had still failed to find a place to nest. From such observations I came to believe that suitable stubs in this and other parts of Barro Colorado Island, as well as other localities, were generally in short supply.

Successful nest stubs. — Both males and females excavated and their greet- ings at times of changing over were expressive of attachment to the nest hole as well as to each other. At 10:45 on 26 December, for example, the male at Nest D drum-tapped when inside the hole with head still visible, made low notes, then drum-tapped again when his mate flew over to take his place. She tossed some sawdust from the entrance but spent most of her time looking out. When Male D returned in 15 minutes, she immediately disappeared to drum- tap at the bottom of the cavity. He peered in at her several times before sbe squeezed out by him to fly away. MD then tossed out sawdust. As with ED, however, he was soon looking out idly and I believed from this and subse- quent behavior that the nest was ready for egg laying.

The woodpeckers of Pair E were late in excavating in comparison with Pair D, for they did not begin until late in January. Their hesitancy to use the stub finally chosen was probably due to the mass of epiphytes at the top, together with the lianas that might have encouraged arboreal mammals or other un- wanted neighbors. Tbe pair had, however, carved an entrance by 22 January. The female did most of the excavating at this nest and change-overs, when they did occur, were much the same as for Pair I) with one exception. Jliis was on 31 January. Male E had been excavating when his mate flew to the hole making low notes. Instead of dropping out of sight to drum-ta|). ME remained by the hole to meet his mate directly and the two touched bills a number of times before be flew away.

34

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

NESTING

Greeting ceremonies in the first few days of incubation were much the same as in the previous period in regard to drum-tapping, but changes soon became evident when the birds became silent, increasingly undemonstrative, and no longer looked out from entrances. When MD came to his nest hole at 16:00 on 8 January, his mate swung out of the hole in silence, the two wood- peckers resting side-by-side below the hole momentarily without other cere- monies before she flew away.

Times between change overs are long for C. melanoleucos. In waits of one and a half hours or more I never saw a woodpecker return after leaving. Skutch (1969), however, in waiting for extended periods at a nest of the closely related Pale-billed Woodpecker [ C. guatemalensis ) , noted the female as spending four and a half hours on the eggs and her mate as remaining on them from 12:15 until dark without being relieved.

Nestling period. — The behavior of the woodpeckers at Nest D changed with the hatching of the eggs, their greater restlessness being exemplified by the following observations: On 15 January MD, after looking out from the hole for five minutes, flew out at 13:45 to preen for a few minutes on an adjacent tree, then re-entered to brood the nestlings. The longest time he spent away from the nest in the course of two and a half hours was 10 minutes.

On 22 January ED had been brooding for a half hour when she flew out, leaving the young unattended for 45 minutes before MD arrived and entered. ED returned almost immediately, replacing him within a few minutes. Her attention to the nest was closer than that of her mate on this and succeeding days, as she would generally stay near the nest when not in it, whereas MD might, at times, be away for more than three hours.

Other events of special interest at Nest D during the nestling period can be summarized as follows:

1) Neither parent ever brought visible prey to feed their young. It is con- ceivable, however, that they might have done so had the young survived longer, for Tanner (1942) describes Ivory-billed Woodpeckers as carrying large grubs to well-grown nestlings.

2) ED and MD were both together in the nest on two occasions. Thus, on 30 January MD entered while his mate was inside, only to leave a few minutes later and on the following day, under similar circumstances, he remained inside with her for five minutes.

3) MD became increasingly apprehensive as the nesting period progressed, delaying each entry by much looking about and bowing into the hole, only to withdraw. Whether the predator that finally destroyed the nest (if predation was the cause) was in the vicinity I did not know, but on nearly every visit

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

35

to Nest D I did see a pair of Spectacled Owls [Pulsatrix perspicillata) almost within sight of it.

I found Nest D devoid of activity when I came to it on 7 February. The entrance was undamaged and I could find no clue as to why the nest had failed. In looking about in woods nearby I was able to locate the parents and to follow them for one and a half hours. Their behavior was now much as in the pre-nesting season, with no sign that they any longer had young to feed.

Tanner (1942) speaks of three nests of Ivory-billed Woodpeckers from which the young disappeared mysteriously and the nest of the Pale-billed Woodpecker observed by Skutch (1969) also failed. He noted a large black snake in the vicinity. It would seem that snakes may be likely predators of such woodpeckers when entrances are undamaged.

YOUNG AFTER LEAVING NEST

I was watching a male Crimson-crested Woodpecker digging out grubs from I a well-rotted stub on 22 January on Barro Colorado Island when a second wood- pecker in adult female plumage alighted 25 cm away. She made no effort to feed herself but preened lightly, making k-da k-da begging notes much of the time. The male paid no seeming attention until, on encountering a huge larva , (4 cm long and 1^4 cm in breadth), he leaned over and fed it to her. Mean-

while, a second female, I believed the mate of the male, joined the other two. This second female, unlike the first one, dug out her own food. The two ; females got along peacefully although later on I had evidence of a brief con- flict between them.

The male did not feed the begging female again in the course of the hour that I followed them. She foraged for herself occasionally but much of the time she followed him so closely that she was almost at the tip of his tail, whether he was feeding along the under or top side of a limb. It seemed possi- ble that she was a young of the year before and while this may seem a long time for a young one to stay with parents, it does fit a situation described by Short (19706) for C. melanoleucos in Argentina as well as by Tanner (1942) who wrote of C. principalis as follows:

“The young birds usually leave or are driven away by the following nesting season, but the single male that was raised by the John’s Bayou birds in 1938 stayed in that I territory through the following spring. The female of the pair frequently tried to drive him away, but he would only dodge, sulk, and return. The old male paid little or no : attention to his yearling son.”

: The first juvenile I encountered on Barro Colorado was on 51 January

when continued k-arr k-arr disturbed notes attracted me to one at the edge of a gap in the forest. Its mother in the same clearing gave her k-wirr k-wirr-a I notes, then flew off. The young one followed and later I found it close beside

36

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

her as she dug for grubs in a dead stub. The only other encounter with juveniles was on 19 Eebruary 1965 when, as described below, one adult female, attended by two juveniles, had a confict with a second female.

TERRITORIAL AND AGONISTIC BEHAVIOR

Female vs female. — Two females alighted low on a series of trees on the morning of 19 February 1965, shifting around trunks as one tried to strike the other, or made a display of doing so for over 15 minutes. The presence of several juveniles indicated that the conflict, possibly a territorial one, had come at the end of the nesting season.

Male vs male. — Sounds of much drumming had come from Territory E on the morning of 27 December. When I followed these into the woods at 13:00, I found two males, one pursuing the other in short, heavy-sounding flights from one to another of four trees centering on a tall stub, to which they often returned. The stub, although unsuitable in a number of aspects, was large enough for nesting. When the woodpeckers came to rest, I noted two types of more direct conflicts: In one that lasted five minutes, one male clung almost upside down below a large limb, while the other, perched on top, half-opened his wings each time the first one tried to come around from below. When the two flew, it was to continue with an even milder type of encounter on a tree trunk nearby. Here one backed down as the other retreated backward. Finally both flew in opposite directions with the territory owner going to a large dead limb where he drummed in slow but resounding fashion for six minutes. He then attacked the intruder again. All of the fighting was silent except for two sharp ca notes. The males were still engaged when I left 50 minutes later.

A somewhat different and even milder conflict between males took place on 12 January at Cardenas Village, where Female A was probably incubating eggs. MA was preening and occasionally drumming at “the knoll” ( Fig. 3 ) on what was usually FA’s drumming tree when, at 07 :0o, a second male ar- rived on MA’s usual drum tree 10 m away. MA did not appear disturbed. He continued to preen and drum as before, giving about five bursts to every single one given by the intruder. The latter clung almost immobile the whole time. Possibly, being well within the territory of MA, he was intimidated. This was suggested when he suddenly flew toward MA, then changed his mind in mid-air, and returned to his original position. Five minutes later he again flew, but this time in an opposite direction.

The intruder again returned some minutes later to a tree close by MA. The conflict ended at the knoll when MA left soon afterward, followed in a minute or two by the second male. This was possibly the first of a series of encounters taking place on subsequent mornings between the two males.

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

37

The above encounters were all mild in nature. There were no accompanying vocalizations, no bill-wavings, or for the most part any threat displays, and no direct clashes such as one can observe among temperate zone species such as the Pileated Woodpecker (Kilham, 1959). It may be that tropical species, being under more pressure from predators, cannot afford to attract attention to themselves when engaged in conflicts. Short ( 1970b ) , however, gives a description of a more conspicuous conflict between two males of C. robustus.

INTERACTIONS WITH OTHER SPECIES

Collared Aragaris. — The mildness and seeming lack of aggressiveness of Crimson-crested Woodpeckers was exemplified, in a different context, when several Collared Aragaris took over the just completed nest hole of Pair A on 26 November. After drum-tapping and bill-touching by the hole, these wood- peckers had appeared comfortably on the way to nesting when MA entered the hole to roost for the night at 17:33. He was soon looking out, however, as if nervous. Seven minutes later he slipped out and moved around to the rear of the stub, being joined by his mate as a toucan flew to the hole and put its bill in several times. The woodpeckers made a few low krr notes but gave no sign of resistance. They simply flew off and as far as I know they never returned. A feature of this performance was that the toucans did not appear too confident. They did not roost in the hole on the 26th and when I returned to the nest stub at 17:25 four evenings later, I found them still chary about entering, for they rested nearby for 20 minutes as if looking the situation over before doing so. A few nights later, on the contrary, they arrived at dusk and entered directly. They had thus won the hole without any show of aggressiveness.

Reaction to a marmoset. — On 24 February 1965 I watched a male Crimson- crested Woodpecker feeding in a mass of vines at the top of a tall stub in company with two marmosets (Oedepomidas ^eojfreyi) . A marmoset came down a vertical liana on which a woodpecker was working. Neither species ])aid any attention to the other, even though they passed within 5 to 7 cm of each other on either side of the vine. Crimson-crested Woodpeckers did, however, become much excited by monkeys on one occasion. This was when the members of Pair E were excavating a nest not far from a cage of Cebus monkeys on Barro Colorado on 26 January. Loud screaming from the cage upset both birds to the extent that they made almost continuous put-put-piitta notes for 10 minutes.

On the whole, however, I found Crimson-crested Woodpeckers relatively unexcitable as compared with Lineated Woodpeckers. I heir tameness in fact was of great aid in observing them. It would seem that J'anrier ( 19 12 I had a similar exj)erience in noting that Ivory-billed Woodpeckers became used

38

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

to people so rapidly that “in a day or so (they) would pay little or no attention to one a moderate distance away.”

FEEDING BEHAVIOR

Methods of Foraging. — The feeding behavior of Crimson-crested Wood- peckers was separable into the following categories:

Pecking. — The uncovering of prey with relatively few blows against bark of super- ficial layers of wood.

Percussion. — While a woodpecker may deliver many blows per minute in pecking, not all of these are to uncover prey. Some appear to be exploratory, given here and there without digging into the wood, either to cause a wood-boring larva to move within its tunnel and thus reveal its location or to sound out difference in resonance between a hollow tunnel and solid wood.

Scaling. — When working on limbs that have been dead for some time. Crimson-crested Woodpeckers may combine pecking with sidewise, glancing blows that dislodge sizeable pieces of loose bark and other debris that may shower to the ground as the woodpecker moves along. On the other hand, almost nothing may fall when a woodpecker is working on the closely adherent bark of a dying limb; the powerful, rapid, occasionally prying blows involved in its straightforward pecking being sufficient to uncover prey.

Probing. — Putting the bill into natural cavities or clumps of epiphytes, etc., presum- ably to explore their interstices with their tongues, although these are seldom visible.

Digging. — When working on well rotted stubs for deeper lying prey. Crimson-crested Woodpeckers may dig cavities 10 cm or more deep, seizing and tossing larger slivers of rotten wood to the ground as they do so. The sizes of such cavities are usually no larger than those made by Hairy Woodpeckers (Dendrocopos viUosus) and never as large as the deep troughs dug by Pileated Woodpeckers in North America. This doubt- less reflects the fact that conditions of decay and location of insects are different in tropical climates.

The listing of these categories of feeding and foraging does not provide a full picture. As pointed out lucidly by Bock and Miller ( 1959 I the Campephi- lus group of woodpeckers have remarkable adaptations not only in the for- ward direction of all their toes, but particularly in having legs directed away from the center of the body in such fashion that the full tarsus can be pressed against trunks and branches. The result is that such a species as C. melano- leucos, in whatever position it is working, whether on the underside of a limb, on the smooth bole of a large tree, or out on smaller branches, appears to be solidly stabilized for delivering powerful blows.

Feeding of non-breeding pairs in dry season. — Observations on a pair with- out attendant young, followed for 140 minutes on 24 February 1965 on Barro Colorado Island, brought out some aspects of feeding when the woods were relatively free of leaves. The two birds were usually within 15 m and often much less of each other as they moved through the mature forest feeding at heights varying from 6 to 25 m, the latter height bringing them close up under

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

39

the canopy of larger trees. Most of their feeding, however, was at intermediate levels. An almost constant feature with this pair was the greater activity of the female, for she was not only the first to fly on the six occasions when the woodpeckers flew from one part of the woods to another, but she also moved along a greater extent of limbs and tree trunks in feeding than the male. At one time, for example, both woodpeckers flew to a dead limb 4 m long. During the next 10 minutes she progressed nearly the whole length of the limb in knocking off bark and debris while he moved only a fifth as far as he probed thoroughly in a limited area, which he continued to do after she had left. His inclination to work one place thoroughly was again exemplified later in the morning. This time he was on a tree with a relatively smooth bole where he found large numbers of grubs under a strip of discolored bark and fed on them for 15 minutes. When any fell, he would press his belly against the bark to recover them.

Foraging in the rainy season. — The dry season arrived late in Panama in 1971 so that essentially all observations made from November into the latter part of January were made in the rainy season. They were divisible into two categories of which the first was in the second-growth woods at Cardenas Village. Here at 16:40 on 4 January, when Female A was presumably incu- bating, I found MA working alone on a small semi-dead tree, 3 to 4 m above the ground and at the level of my eye as I stood on a slope above. At times he moved out onto branches of 2.5 cm in diameter. Clinging securely by grasping small branchlets, two of his forward-directed toes on one side and two on the other, he pecked steadily on the still adherent bark, as if finding

I considerable amounts of prey. At one time, for example, I saw him extract a larva grub about 3 cm long. At another time he clung to the underside of a slightly larger branch, his forward directed toes serving well for hanging in this position. It is likely that insect larva are particularly abundant on the ! underside of limbs and branches where moisture collects and persists longer than on the uppersides. The male also worked on a limb of 10 cm in diameter. Here I could see that he delivered three or four powerful pecks in one place, . then moved along to another, pecking rapidly and nowhere penetrating deeply into the wood. With a background of watching woodpeckers in the temperate I zone, I would have thought the branches more suitable for a Hairy or even a Downy {D. pubescens) Woodpecker, than for a large species such as C. melanoleucos. Short (1970a) noted C. magellanicns feeding on small branches I in a similar manner.

Crimson-crested Woodj)eckers are versatile feeders whether in second- I growth woods, such as those in Cardenas Village, or on Barro Colorado where I the mature forest contained many large stubs and branches. Methods of I feeding in these habitats are illustrated in the following examples. 1 ) Feeding

40

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

directly under bark. MA delivered hard blows on the firmly adherent bark of a dead stub near Cardenas Village and as bits of bark came loose. I could see tunnels of wood-boring larvae directly below. The woodpecker’s blows were straight on, followed by a few at slight angles, together with prying motions. 2) Excavations into wood. A pair of Crimson-crested Woodpeckers on BCI dug holes 3 to 5 cm deep in a large stub finding not only small grubs, but also several large ones measuring approximately V2 by 4 cm. 3 ) Tree with smooth bark. A female fed under strips of loose bark on the unusually smooth bole of a large tree by splaying her legs well out to the side. 4 ) Possible feeding on termites. A lone female fed for 40 minutes on a dead stub arising from a small, understory tree. She dug so industriously into its basal portion that the upper part broke off and fell to the ground. Later examination of this portion revealed that it contained many termites along with a few tunnels, all old, of large larvae.

The foraging habits of Crimson-crested Woodpeckers were easy to observe for several reasons. First, the woodpeckers would often move from one tree to the next, finding plenty to look for without taking long flights from one good tree to another as is often the case with other woodpeckers, such as the Pileated or Hairy in northern woods; and second, when feeding high up on dead limbs, they would often move along the underside where one still had a good view of their activities.

COMPARATIVE FEEDING BEHAVIOR AND INTERACTIONS WITH D. LINEATUS

Crimson-crested and Lineated Woodpeckers fed in the same locations and occasionally on the same trees on Barro Colorado Island without signs of hostility or indeed special reactions of any kind.

I heard, for example, vocalizations of both species, then found the four woodpeckers of two pairs intermixed as they fed among a small group of trees on 24 December. When the Crimson-cresteds left, the Lineated Woodpeckers moved into the tree where they had been. Here the male probed into holes and crevices of a dead limb, then moved out onto a dead branch 2.4 cm in thickness that one would have thought suitable only for a smaller species. I had noticed a female Lineated Woodpecker doing much probing a short while before and an impression that this method of foraging was a characteristic habit of D. lineatiis, more so than of C. melanoleucos, was re-enforced by further observations on 5 February. Thus, at 09:00 I heard both the kwirr-as of Crimson-crested and the wer-wer- wer notes of Lineated W'oodpeckers coming from close by a trail. Sounds of digging then led me to a male Lineated. He pecked only briefly, then began probing a spot on the upper side of a large limb, 15 cm below a decaying branch stub. This was probably a ramifying area of decay, for the male turned and twisted his head for the next five minutes, as though reaching into deep tunnels or interstices wdth his tongue, the whole performance being identical with what I have witnessed with Pileated Woodpeckers on many occasions. It seemed probable, therefore, that the Lineated Woodpecker was for- aging on ants and their larvae.

At 09:10 a male Crimson-crested suddenly alighted only a meter below the male

Lawrence

Killiam

CRIMSON-CRESTED WOODPECKER

41

Lineated Woodpecker, As the Crimson-crested Woodpecker hitched upward, neither he nor the Lineated Woodpecker raised their crests in even mild excitement. When the two were within 30 cm, the Lineated flew to a limh a meter away, remaining there quietly while the slightly larger Crimson-crested Woodpecker took over his feeding place. The latter gave only a few pecks, as though finding nothing of interest, then moved on to drum once on the broken branch stub, preen briefly, and leave. The Lineated now returned to continue at his feeding spot for another 10 minutes.

j A number of aspects of this episode were noteworthy. First, the Crimson- crested Woodpecker had not replaced the Lineated in a supplanting attack, for there was no sign of hostility, the situation appearing to be one of simple dominance at a food situation. The Crimson-crested was the larger wood- pecker and this, plus having a longer, heavier bill, may have explained his dominance.

A second feature of the episode was that whereas the Lineated Woodpecker had started making put-air notes when I had arrived, he stopped making these I notes when the larger woodpecker replaced him, appearing thus, if anything, to have become calmer, rather than more excited. What was the most signifi- cant feature of the encounter, however, was the light it threw on the feeding habits of the two species. The Lineated obviously found much to feed upon in the one spot, for he was able to feed there actively for a total of 15 minutes,

I quite possibly on ants and their larvae. On the other hand, the tree itself provided feeding places of a different kind, such as decaying dead limbs, attracting C. melanoleucos, for I had watched the male and female feeding here a few weeks before. These observations suggested the two species of i woodpecker, instead of having the similar “ecologies” needed to support , Cody’s ( 1969 ) theory, can forage on the same trees for quite different sorts of prey. While they do undoubtedly overlap in some of their feeding habits,

' as indeed Tanner (1942) showed for Pileateds and Ivory-bills, this is not of sufficient degree to interfere with their being sympatric.

That the Lineated Woodpecker is specialized is seen most clearly, as is well discussed by Skutch ( 1969 ) , in its attacks on Cecropias and the colonies of Azteca ants harbored in their hollow trunks and branches. These trees grow in abundance nearly everywhere and their prevalence at edges of woods may explain why Lineated Woodpeckers come to these situations. On 9 January, for example, I found a male digging into a Cecropia at the edge of the lahora- I tory clearing at Barro Colorado Island. He worked first on the trunk where I it was 7 cm in diameter, then on a limb of half that thickness. Although I had j many more observations on C. melanoleiicos than on I). lineatus. I never saw I it even alight on one of these fast growing trees which, in general, hear little dead wood.

) I found it more difficult to observe the feeding habits of D. lineatus than ! those of C. melanoleucos for three reasons, namely that I). lineatus was more

42

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

easily disturbed, often starting put-air notes on seeing me; that it was more thinly distributed, being only one-third as numerous as C. melanoleucos on Barro Colorado Island; and finally that, whereas C. melanoleucos usually foraged from one tree to another close by, D. lineatus, after spending some time on one tree, might take a long flight to another and was hence easily lost to further observation. The last two situations suggested that D. lineatus requires larger feeding territories than C. melanoleucos. Whatever the dif- ferences between the two species, it was striking that they thrive together whether in old and mature woods such as at Madden Forest, Limbo, and Barro Colorado, or the second growth ones at Cardenas Village and Frijoles. It is of interest here, finally, that Slud (1964) found D. lineatus less common than C. guatemalensis in Costa Rica.

COMPARATIVE BREEDING BEHAVIOR OF LINEATED WOODPECKERS AND INTERACTIONS WITH C. MELANOLEUCOS

Skutch (1969) has provided a general account of the Lineated Woodpecker. Additional aspects based on recent observation are given below to bring out mainly how it is that D. lineatus and C. melanoleucos can live in sympatry without undue competition or overlap in any aspects of their lives. Reproduc- tive isolation is, of course, complete. Not only are patterns of plumage colors about tbe head different, but also, and this may be of special importance, C. melanoleucos bas a bright yellow iris while the iris of D. lineatus is strikingly white. This situation is depicted in color for D. lineatus and C. guatemalensis by Sutton (1951) . The latter woodpecker forms a superspecies with C. melano- leucos and is also similar in plumage to D. lineatus with which it is sympatric. Short ( 1970b ) noted that the eyes of an immature female of C. melanoleucos were white.

The drummings and vocalizations of C. melanoleucos and D. lineatus are also different. Thus, in C. melanoleucos the main call is a kwirr-a while in D. lineatus it is, according to Skutch (1969), a flicker-like wic ivic wic. I have found, however, that this latter is actually part of a spectrum, becoming at high intensity a wuk wuk wuk of about 17 notes, falling off at the end, that one recognizes at once as being similar to the high call of the Pileated Wood- pecker (Kilham, 1959), while at low intensities the notes become a wer iver wer that one might never consider as coming from a woodpecker. The drum- mings differ to an equal degree. Thus, much of the communication between members of a pair as well as between rivals in C. melanoleucos is by their peculiar drumming, vocalizations being infrequent. Comparable communica- tions of D. lineatus.) on the other hand, are more by vocalizations, while the long rolling drum, again like that of D. pileatus, is used less frequently.

Nest excavation. — Crimson-crested and Lineated Woodpeckers are further

I

Lawrence

Killiam

CRIMSON-CRESTED WOODPECKER

43

Table 1 Simplified Outline of Differences in Breeding and Other Habits of Crimson-crested AND Lineated Woodpeckers that Permit Sympatry Without Undue Competition*
Differences	C. melanoleucos	D. lineatus
Main breeding season	Last of rainy season and first of dry season (Nov.-Jan.)	Last of dry season (March-April)
Relative size of territories	Small	Large
Type of stub used	Large (45-50 cm diam.) ;	Stubs or tops of stubs
for nesting (optimal)	substantial	small in diam. (18-23 cm) ; more risk
Food *	*Larvae of wood-boring beetles primarily and possibly termites	Azteca and other ant larvae, overlaps with C. melanoleucos otherwise in feeding on beetle larvae
Temperament	Relatively tame	Easily alarmed

* It should be emphasized that this outline is based on observations of relatively few individuals.

** Special adaptations of feet and legs (Bock and Miller, 1959) make C. melanoleucos especially efficient at extracting this type of prey. ( See text. )

isolated reproductively by the timing of their nestings, that of C. melanoleucos I coming at the end of the rainy season and that of D. lineatus toward the end

j of the dry season (Table 1). W. John Smith (pers. comm.), for example,

found a pair of P. melanoleucos nesting at Frijoles on 27 January 1967 not I far from where a pair of D. lineatus had nested in May 1966, and Chapman ( 1929 ) mentions the young of a pair of C. melanoleucos as leaving their nest on Barro Colorado in February. Skutch (1969) stresses that the closely re- lated C. guatemalensis, which replaces C. melanoleucos northward of Panama and is also sympatric with D. lineatus^ is an unusually early nester. Although Lineated Woodpeckers nest later than Crimson-crested, they may, in some cases, start trial nest excavations early in January, as indicated by the fol- lowing observations: On 2 January I found a pair of Lineated Woodpeckers excavating a hole in the dead top (Fig. 1) of a living tree, one of the Bom- I bacaceae. The cavity was already deep but the two birds continued to toss out sawdust from the entrance until 4 January, when the excavation afipeared I to have been completed. Yet with exception of a brief view on 5 January 1 } never saw the pair by the hole again. Strong winds came with the beginning

I of the dry season later in the month and on 1 February I found that the top of

j the tree had broken off where the cavity of the woodjieckers had weakened it ' (Fig. 4).

44

THE WILSON BULLETIN

March l‘J72 Vol. 8;, No. 1

c 21cm D. lineatus

'

"c 45cm

C. melanoleucos

Fig, 4. Contrasting sites of nest excavations of Lineated and Crimson-crested Wood- peckers. (The entrance hole of the Lineated’s nest having been under the curving arch of the limb above, is shown as seen from below, looking up.)

On 8 January on BCI a pair of Lineated Woodpeckers were excavating a hole they had pirated from a pair of the smaller Black-cheeked Woodpeckers ( Melanerpes pucherani) . The hole was in an arching limb 18 cm in diameter at the top of a tall dead stub. Both members of the pair of larger woodpeckers could enter their excavation completely by 9 January. The male was still excavating a week later but on 17 January the entire stub crashed to the ground.

In summary of these and other observations it would seem that D. lineatus differs from C. melanoleucos in the locations as well as in the timings of its nest excavations. Thus, while C. melanoleucos is particular about finding a large stub ( Eig. 1 ) that will be a secure place to nest and appears wary about even attempting to nest otherwise, D. lineatus is attracted to inherently more risky situations, whether in stubs or in dead tops of trees so narrow that the nest cavity is barely accommodated. Its walls, therefore, are necessarily thin, offering too little support in case of wind or storm. Advantages of using such situations, however, must outweigh disadvantages. They may include such things as freedom from competition with the sympatric C. melanoleucos for

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

45

nest stubs, locations so high above ground as to be less within reach of usual climbing predators, and in being less attractive in their fragile nature to nest hole competitors of various species such as Collared Aragari.

Finally it should be noted that this habit of making nest excavation in places that would seem too narrow and too risky is not confined to D. lineatus. As previously described (Kilham, 1959) the same situation holds for Pileated Woodpeckers in central Florida where, in absence of any large trees, they may nest in narrow pole-like dead pines where a full nest cavity may be sup- ported by little more than outer bark, Truslow (1967), who happened to be present at the dramatic moment, has recently photographed the breaking up of one such nest under only a light wind.

DISCUSSION

The Crimson-crested and Lineated Woodpeckers, whose breeding and feed- ing habits have now been compared, are a remarkable pair of species in being alike in size and general coloration, yet sympatric within the same monsoon- rain forest habitat. Thanks to Cody’s article ( 1969 ) , I became interested in studying these species concomitantly. If it were true, as Cody claimed, that these two are so alike in habits that they can coexist sympatrically only by means of an interspecific territoriality promoted by convergence in size and plumage patterns, then here was a remarkable biologic phenomenon. Unfortu- nately, I could find no evidence supporting Cody’s ideas, for I was struck, as also was Karr (1971), that the two species are mutually tolerant. Every time I encountered Lineated Woodpeckers on Barro Colorado Island, for example, they were within the territories of one or another of pairs of Crimson-crested Woodpeckers under study. At no time did I observe conflicts such as might arise from mutually exclusive territoriality. The general peacefulness between the two species was notable not only when pairs happened to be feeding on adjacent trees, but also on one occasion when a male Crimson-crested, coming close to a male Lineated Woodpecker, temporarily displaced it from a feeding spot without show of hostility on the part of either the dominant or of the submissive species.

Having concluded early that interspecific territoriality did not exist, 1 wondered whether Cody’s theory might not be modified to apply to spacing out in relation to nest sites. This hypothesis, however, likewise became unten- able in the light of experience. The two species are divergent in such impor- tant aspects of their lives as the time of their breeding, the nature of nesting sites they look for, as well as in their feeding habits, as summarized in Table 1. Such a situation is, of course, the usual outcome of natural selection. Wbat is unusual, if not very rare, would seem to be interspecific territoriality based on any long term evolutionary process.

46

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

An interesting example of limited interspecific territoriality among very closely related woodpeckers is given by Selander and Giller (1959). They found that, seemingly due to man’s interference with natural ecologic barriers, morphologically similar members of the same superspecies, Centurus carolinus and C. aurijrons, met in Austin, Texas, and, in a limited area of sympatry, held mutually exclusive territories. This situation would appear different from what must be the historically long sympatry that has existed between C. melanoleucos and D. lineatus.

From one point of view an instructive example of a species pair comparable in some ways to the Crimson-crested and Lineated Woodpeckers, and even more alike in plumage although dissimilar in size, are the Hairy and Downy Woodpeckers. I have found (Kilham, MS ) that in spite of wide differences in prey and feeding habits, in type of nesting sites, as well as in the time of onsets of breeding behavior, these species must still be acted upon by many selection pressures in common, such as predation, survival over winter months when trees are bare of leaves, and many others, in relation to which their plumages represent one of many optimal compromises for survival. While the selection pressure may differ from tropical rain forest to north temperate woodlands, the principles of why certain birds are similar in plumage would seem to be the same.

SUMMARY AND CONCLUSIONS

Reproductive and feeding habits of Crimson-crested Woodpeckers were followed in mature as well as in second-growth woods of the Panama Canal Zone.

The double drum DA-drrr, characteristic of Campephilus woodpeckers, was a main method of communication, whether used to express mild alarm, territorial dominance, or in duets between members of a pair at time of courtship.

Copulations and excavations were seen in November but most pairs had difficulty finding suitable nest stubs and either began nesting in December or January or, in some cases, failed to nest.

Territorial conflicts between rival males were marked in January, the intrusions being largely by males of pairs that were failing to establish nest holes.

Both sexes excavate and the bird excavating drum-taps on the inside or outside of the cavity on the arrival of its mate. This drum-tapping ceremony is identical in Campephilus and Dryocopus.

Bill-touching or fencing between members of a pair takes place at the nest excavation or elsewhere at the height of courtship.

Crimson-crested Woodpeckers become silent and difficult to observe in the incubation period, sitting on their eggs for prolonged periods without looking out from nest holes.

After hatching, either sex may look out and in the first few days when brooding young, drum-tap on the arrival of a mate. Prey was never visible in the bills of parents coming to feed young in the first three weeks.

A bird in adult female plumage, seeming by her begging behavior to be a young one of the year before, was seen accompanying a pair of Crimson-crested Woodpeckers in January. The male fed her a large grub on one occasion. Juveniles of recent nestings were first seen late in January and in February.

Lawrence

Kilham

CRIMSON-CRESTED WOODPECKER

47

Crimson-crested Woodpeckers have remarkable adaptations of legs and toes which I enable them to cling securely when feeding in such difficult situations as the undersides

I of limbs, small branches, or on boles of large trees. Larvae of wood-boring insects appear

to be their chief prey.

Crimson-crested Woodpeckers live in the same woods and even feed in the same trees I with Lineated Woodpeckers, which appear remarkably like them in size and general I coloration. The two species differ in feeding habits, in time of onset of nesting, and in types of nest sites chosen. No signs of interspecific hostility or territoriality were observed.

I ACKNOWLEDGMENTS

I am much obliged to Lester L. Short for going over my preliminary manuscript and I also to my wife, Jane Kilham, who greatly aided these studies in finding nests of Crimson- erested Woodpeckers and helping to watch them, as well as in re-drawing the field sketches shown in Figures 1 and 2.

I LITERATURE CITED

I Allen, A. A., and P. P. Kellogg. 1937. Recent observations on the Ivory-hilled Wood- I pecker. Auk, 54:164-184.

' Bock, W. J. 1963. Evolution and phylogeny in morphologically uniform groups. Amer.

I Naturalist., 97:265-285.

I Bock, W. J., and W. D. Miller. 1959. The scansorial foot of woodpeckers with com- j ments on the evolution of perching and climbing feet in birds. Amer. Mus. Novitates,

no. 1931:1-95.

Chapman, F. M. 1929. My tropical air castle. Appleton, New York.

Cody, M. L. 1969. Convergent characteristies in sympatrie species: A possible relation to interspecific competition and aggression. Condor, 71 :222-239.

Karr, J. R. 1971. Ecological, behavioral, and distributional notes on some Central Panama birds. Condor, 73:107-111.

’ Kilham, L. 1959. Behavior and methods of communication of Pileated Woodpeckers. Condor, 61:377-387.

Selander, R. K., and D. R. Giller. 1959. Interspecific relations of woodpeckers in Texas. Wilson Bull., 71:107-124.

SiELMANN, H. 1958. Das jahr mit den spechten. Verlag Ullstein, Berlin.

Short, L. L. 1970a. The habits and relationships of the Magellanic Woodpecker. Wilson Bull., 82:113-240.

Short, L. L. 19706. Notes on the habits of some Argentine and Peruvian woodpeckers lAves, Picidae). Amer. Mus. Novitates, no. 2413:1-37.

Skutch, a. F. 1969. Life histories of Central American liirds III. Pacific Coast Avifauna, no. 35:1-580.

Slud, P. 1964. The birds of Costa Rica. Bull. Amer. Mus. Nat. Hist., 128:189. j Sutton, G. M. 1951. Mexican birds. Univ. Oklahoma Press, Norman.

Tanner, J. T. 1942. The Ivory-hilled Woodpecker. Res. Rept. No. 1, Natl. Audubon I Soc., New York.

j Truslow, F. K. 1967. Egg-carrying l>y the Pileated Woodpecker. Living Bird. 6:227- I 236.

I DEPARTMENT OF MICROBIOLOGY, DARTMOUTH MEDICAL SCHOOL, HANOVER. NEW J HAMPSHIRE. 17 MAY 1971.

TERRITORIAL BEHAVIOR IN SAVANNAH SPARROWS IN SOUTHEASTERN MICHIGAN

Peter E. Potter

The Savannah Sparrow {Passerculus sandivichensis) is a bird of open grasslands, bogs, coastal marshes, and tundra. In southeastern Michigan its thin insect-like song is heard wherever farming has produced pastures and fallow fields. It migrates south in late summer and fall and returns in April and early May. For three successive breeding seasons (1965—67) I observed the territorial behavior in Savannah Sparrows in a field five miles west of Ann Arbor, Washtenaw County, Michigan. The population ranged from about 18 pairs in 1965 to 12 pairs in 1967.

METHODS

The study area was measured off in a grid, with tape markers placed along border fences and metal ground markers at the grid intersections in the field. Song perches were marked with colored pipe cleaners, some with colored foam plastic balls attached. Adult birds were netted and marked with aluminum and color-coded plastic bands. Sex was determined by behavior since there is no discernible difference in appearance. Nestlings were marked only with aluminum bands. ( Only one bird banded as a nestling later returned to the study field to breed.) Fifty- two adults were banded in 1965, 12 in 1966, and 6 in 1967, a total of 70. Seventy-five young were banded in 1965, 29 in 1966, and 26 in 1967, a total of 130. ( Banding in 1966 and 1967 was more selective, aimed at birds evidently linked to a territory. In several instances, females on their nests were flushed into nets posted near them. Only one non-resident Savannah Sparrow was caught in each of those years, contrasted to 19 in 1965. )

I observed the birds mostly on Fridays and Sundays from 06:00 to 12:00. Occasionally, I made evening visits. In all, I spent 490 hours in observation.

Because Savannah Sparrows spend so much of their time on the ground, it was im- possible to determine their territorial boundaries where vegetation was dense. “Walking” the birds around their territories was not feasible since they would leave their territories when pressed. Neither did many territories touch others, where the males might have clashed and revealed the borders. It was necessary, therefore, to fall back on the device of marking the males’ singing perches to provide an approximation of the territorial areas. When singing was done on the ground, usually during pauses in foraging among the hummocks of grass, adjacent grass clumps or weed stalks were marked.

Gradually the accumulation of markers described areas the edges of which appeared to be defended consistently. Furthermore, the birds did not appear to go much beyond these markers to defend their territories. Thus, the variation between the edges of those areas described by markers and the actual territorial boundaries seemed slight enough to make the location of the territories clear and the measurement valid.

STUDY AREA

The study field contained 4.74 hectares (11.72 acres) and was essentially level and poorly drained. It was bounded on the south by a gravel road and a brushy field, to the north by cropland, and on either side by wet pastures.

48

Peter E. Potter

SAVANNAH SPARROW TERRITORIES

49

I

Most of the study field was covered by bluegrass (Poa pratensis) fallen over or blown down in successive layers to form hummocks 30 to 50 centimeters in diameter and up to I 30 centimeters high. The bluegrass and interspersed timothy (Phleum pratensis) grew up to 45 centimeters tall by mid-June. In widely separated locations were slowly-spreading I circles of sedge (Carex stipata) ; chickweed iStellaria grarninea) was also prevalent.

I The northern half of the field was free of woody plants except for a small copse of willows iSalix sp.) up to 4.5 meters tall at one place along the northern fence. The southern half contained scattered clumps of willow iSalix petiolaris) from one-half to I two meters tall. The field had occasionally been used as pasture for cattle in previous j years, including the year immediately preceding the study period, but no cattle were ! there during the study period itself. In those three years there was an increase in the I amount of thistle (Cirsium sp.) , goldenrod (Soli dago sp.), spirea (Spirea sp.) and asters (Aster sp.) .

ARRIVAL DATES

The earliest recorded dates of the birds’ spring arrival at the field during the study period were 9 April in 1965 and 1967, and 15 April in 1966. Males were singing on those dates.

Twenty-two males color-banded in 1965 were first observed in 1966 from 15 April to 13 May, and 17 color-banded males in 1967 from 9 April to 7 May. A color-banded male first seen as late as 21 May 1967 was not seen again.

In both years most of the returning males ( 20 out of 22 in 1966 and 16 out of 18 in 1967) arrived within a ten-day period in April (9-18 April 1966 and 15-24 April 1967. )

In 1966 and 1967, the first color-banded females were seen on 1 May and 30 April respectively. The earliest estimated start for nesting in any year was 30 April 1967. Returning color-banded females were first seen in 1966 as late as 14 June and in 1967 up to 27 May. Usually inconspicuous unless alarmed by the observer’s proximity to a nest or fledgling, some females could have been in the field several weeks before being seen for the first time.

TERRITORIAL DEFENSE

Singling. — Males began singing on arrival in their territories or shortly thereafter. In all three years of the study some singing, however limited, had begun by 15 April. In two of those years the field was full of song on that date. In the third year ( 1965 ) full song came on 23 April.

Singing did not appear to be done by other than territorial males. I never heard a female sing or make any other sound other than a chip of alarm and a buzz when rejecting the advances of a male.

Songs differed from one bird to another and in one bird’s repertoire, but I have no detailed notes on this. I did time one singing individual and recorded 25 songs in four minutes — an average of one song every 9.6 seconds.

I

50

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Borror (1961) found that individual Savannah Sparrow songs last two to three seconds.

Several birds were usually singing by 06:00 in April. They ceased as late as 19:40 in late April, and as late as 20:20 by the end of June. Singing tapered off after 09:00 and the birds were usually still after 12:00. Singing in the evening was less than in the morning but occurred regularly. It was also less or absent in strong wind or rain.

Song was sometimes distorted by wind, making the birds difficult to hear or locate, especially when they sang from behind hummocks on the ground during pauses in foraging.

Singing decreased by mid- June, since the male stopped singing during the incubation period and did not resume until the fledglings were on their own. (He also used the perches less frequently and was less frequently seen. ) When a nest was lost through predation, the male soon resumed singing.

Singing occurred mostly from perches in thistle, goldenrod and willow, and on the barbed wire fence around the field. Certain perches were used more than others.

Fighting. — The ultimate defense of Savannah Sparrow territory is a fight between males, but fights were infrequent. (No female was seen in a fight or any other defense of a territory. ) Typically, the two birds rose straight up about a meter above ground and went back down, breast to breast and clawing all the way. The fights were of short duration — I never saw a rise repeated — and the birds quickly went their separate ways. I heard no sound during the fights.

Chases. — Chases by territorial males were more common than fights, espe- cially early in the season when the territories were first established. They ceased with molt.

In all chases in which I was able to identify the pursuer, the chase was made by the territorial defender and ended at the border or shortly past it. The pursuer usually made a buzzing noise during the chase. In one instance the defender rose almost straight up about 6 meters to intercept and chase a Savannah Sparrow flying over its territory.

The pursuer often ended the chase by flying to a perch in his territory and making a chipping noise or singing. One pursuer, apparently agitated by the chase, flew from a grass clump out in his territory to a fence at the border, then back and forth two more times, singing constantly.

If the chased bird flew through more than one territory, the chase some- times became a relay event, the first defender stopping at his border and the neighboring defender taking up the pursuit.

On three occasions a week apart in April, 1967, I saw gang chases involving as many as five or six male Savannah Sparrows. The first incident began with

Peter E. Potter

SAVANNAH SPARROW TERRITORIES

51

a two-bird chase, the rest converging and all going down into the grass. The birds started scattering before I arrived, but I was able to identify four from their color bands. The second chase involved four birds, only one identifiable. The third incident involved five or six birds, one or two flying in from as far away as 15 meters. It broke up quickly but not before a fight occurred.

The location in all three incidents was the same “no-man’s-land” between several territories. The birds identified were all territorial residents in that area. I was unable to determine if they were ganging up on a bird from outside the area — a transient, perhaps, or a new arrival — or whether a single chase between two area residents excited others into general aggression.

Border-crossings did not always end in chases, perhaps because even Savan- nah Sparrows have difficulty finding each other in tall grass.

On 15 May 1966, for example, a territorial male flew onto a grass tuft and, his crown feathers raised, looked around quickly in many directions but started no chase. Another Savannah Sparrow soon flushed from the base of of a nearby fence post and flew off, whereupon the first bird, his crest now down, perched quietly on the fence and no longer looked around so rapidly.

Generally, however. Savannah Sparrows stayed within their territories throughout the breeding season except when the momentum of chasing an intruder carried a male into an adjoining territory or when a parent accom- panied a wandering fledgling across boundaries.

Other defenses, — Most adjustment of borders between the few territories that touched occurred without either fights or chases. Instead, the opposing males sang on either side of the line, about a meter apart, silently crowded each other back and forth across the line, or walked along the line side by side, a few centimeters apart. There were also combinations of these.

Examples:

1) M-44 was challenged at his border hy another male, M-39. The birds ran side by side, occasionally buzzing and fighting. At times they were only 30 cm apart and both singing.

2) I flushed M-64, and he flew to a grassy area at his boundary. He was instantly met there by M-38 of the adjoining territor>’. Both then walked side hy side, sometimes only centimeters apart, along their border. At one point M-64 stopped and M-38 walked on, whereupon M-64 crossed the “line.” M-38 immediately rushed hack at M-64 and buzzed; M-64 returned to his side and the side-hy-side walking resumed. M-64 occasionally

I sang as he walked. After a few minutes I moved away and M-38 flew to a perch in the center of his territory and sang, ending the confrontation.

3) M-53 resisted intrusions hy M-40, who had part of M-.53’s territory' as his own the 1 previous year. On one occasion M-40 sang from the ground in M-53*s territory hut was ! escorted hack across the border. That is, M-.53 flew to the ground about 30 cm , from M-40 and followed M-40 as the latter walked hack into his own territory. There was

no audible sound.

' .Among encounters on fences bordering adjacent territories, one observed 12 May 1%7

52

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

was typical. M-29 and M-33 approached each other, facing first one way, then another as they perched crosswise on the barbed wire. They fluttered their wings slightly, fanned their tails, raised their body feathers as if swelling, teetered forward with their heads lower than their tails, and opened their bills. At times they were only 30 cm apart.

One would hack up after depressing his body feathers, while the other advanced. Then the action would be reversed. The birds see-sawed a distance of not greater than 1.5 m, more often within a one-half-to-one-meter span. All was done silently except for a few very soft buzzes.

The confrontation ended when M-33 hopped up onto a fence post a little farther away and sang. M-39 hopped down into the grass a short distance in the opposite direction and began foraging.

Other encounters on fences lacked the buzzing, wing movements and feather-raising, but the see-sawing and teetering were the same. None of the encounters resulted in fights.

Immunity from defense. — Parent birds apparently could follow their fledg- lings anywhere without being attacked by territorial defenders. The parents were very excitable at this stage, both birds (but particularly the male) perching closer to the observer than usual and chipping rapidly and loudly.

In June, 1966, female F-69 from an adjacent territory, possibly foraging for her nestlings, perched and chipped in M-64’s territory without being chased out. But when her mate, M-18, also intruded, M-64 approached him and buzzed and M-18 retreated to his own territory.

Six days later, however, the situation changed. The nestlings had left the nest and were being tended by M-18 and E-69. The parent birds again moved into M-64’s territory. Although I was unable to see whether they were fol- lowing their fledglings, this time neither bird was bothered by M-64. On the contrary, M-18 approached M-64 and buzzed.

Interspecific aggression. — Aggression toward birds of other species was ob- served in only a few instances.

A territorial male was seen chasing a Field Sparrow (Spizella pusilla) which shifted only a meter or two at each rush but eventually left the territory.

A Savannah Sparrow landed beside a Song Sparrow ( Melospiza melodia ) and buzzed until the latter flew away, but in another case a Savannah Sparrow flew when approached by a Song Sparrow. In all other encounters, these two species appeared to ignore each other.

Goldfinches iSpinus tristis) and Bobolinks (Dolichonyx oryzivorus) nested in the field without being approached. On the contrary, I once saw a Bobolink chasing a Savannah Sparrow.

I saw no cases of Savannah Sparrows being aggressive toward other animals except in pursuit of insects for food.

Cessation of defense. — Nesting activity tapered off in late July, accompanied by lessening and cessation of territorial defense. The females left the study area, none being seen despite repeated inspection walks throughout all terri-

Peter E. Potter

SAVANNAH SPARROW TERRITORIES

53

tories. The males went into molt, stopped singing and skulked through the brush. When flushed, they flew only a short distance and disappeared into the brush again. Any chipping was low in volume and not persistent. Eventually the males also left the field.

The earliest date on which molt was noticed during the study period — that is, when the males first looked ragged — was 17 July. For some males it was noticed 30 July. In all cases molt was accompanied by a cessation of territorial activity. In no case was molt seen as long as the male was still tending fledglings.

I was never able to observe molt in a female. Quite often a female would appear to be in sleek plumage while her mate looked ragged. Generally the females left earlier and may have molted during this dispersal.

The cessation of territorial defense throughout the field seemed to occur within a week’s time except for a few birds still busy with nestlings or fledg- lings. In each of the three years there came a particular day when I noted that territorial behavior seemed to have ended. Twice it was on 25 July and once on 31 July.

DEPARTURE

The females usually left the study area within two weeks after the end of their last nest, whether the end was from predation or fledging and although both males and females tended fledglings. While they no longer defended their territories, the males stayed on as long as a month and a half, the average being about a month. By 31 July, most had gone, but a few stayed on until mid-August. One was seen as late as 10 September in 1965.

The last resident birds of 1966 were seen on 14 August. Observations in 1967 ended on 31 July, with four males and three females remaining, repre- senting only 22 per cent of the full adult population that season. The seven birds included three pairs with late broods.

In general, the females left gradually through June and July, while most of the males left the last two weeks in July.

NATURE OF TERRITORIES

Shape. — The territories varied considerably in shape from almost square to long and rectangular and roughly triangular, with no apparent correlation between territory shape and success in attracting a mate.

Although the fields adjacent east and west were breeding areas, the I Savannah Sparrows I observed generally adopted the barbed wire fences not only as much-used singing perches, but also as territorial boundaries. J he birds did not cross the fences except when ap})roached by me or for a short I distance in pursuit of an intruding Savannah Sparrow. 1 also recorded one

54

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

instance in which a female apparently followed her fledglings into the adjacent field.

One of the two exceptions to adoption of the fences as boundaries was a Savannah Sparrow which frequently sang from a small sapling about two m beyond the fence, although the bulk of his territory was in the study field. Another bird clearly had territories which straddled the fence line in 1965 and 1967.

The fences were observed as boundaries even when they merely separated open grassland rather than being paralleled on one side by something dif- ferent, such as a road, a ditch or a thicket.

Nest location. — Nests occurred anywhere in a territory, even at the border. In 1965 I discovered two nests only 2.2 m apart in adjacent territories. Both nests were successful.

Size. — Eifty-eight per cent of 62 territories marked during the three years ranged from 601 to 1200 m“ — about one-sixth to one-third of an acre. Fifteen per cent were smaller, 27 per cent larger.

The average for the 62 territories was 1,068 m“ (0.26 acre). For the 27 territories in which no nest was found, the average size was 845 m-; for the 35 in which nests were found it was 1,239 m“.

The literature on the size of sparrow territories is limited. What there is indicates the Savannah Sparrows I observed had territories considerably smaller than the other species noted. I found reports of territory sizes for ten species in addition to my own figures for the Savannah Sparrow.

A comparative list follows, all figures translated into square meters:

Savannah Sparrow (Passerculus sandwichensis) — From 120 to 2,920 m", averaging 1,068 nF (0.26 acre). Present study.

Grasshopper Sparrow ( Ammodramus savannarum) — 4,850 to 13,330 m", averaging 8,200 nr ( 2.03 acres) . Smith, 1963.

Baird’s Sparrow i Ammodramus bairdii) — 4,730 m“ (1.17 acre). Cartwright, et ah, 1937.

LeConte's Sparrow (Passerherbulus caudacutus) — 1,020 to 6,300 nr, averaging 3,320 m~ ( 0.82 acre) . Calculated from maps by Murray, 1967.

Henslow’s Sparrow (Passerherbulus henslowii) —Average of 3,238 nP (0.80 acre). Robins, 1971,

Sharp-tailed Sparrow (Ammospiza caudacuta)- — Female less than 4,047 m" (1 acre), males not territorial. Woolfenden, 1956.

Seaside Sparrow (Ammospiza maritima) — Nesting area, 5,830 m“; shoreline feeding area, 4,170 nr ; total, 10,000 m‘ (2.47 acres). Woolfenden, 1956.

Tree Sparrow (Spizella arborea) — 5,580 to 39,100 m^ (1.38 to 9.66 acres). Heydweiller, 1935.

Chipping Sparrow (Spizella passerina) — 4,047 to 6,070 m“ (1 to 1.5 acre). Walkinshaw, 1944.

Field Sparrow (Spizella pusilla) — Less than 3,640 to 8,094 m“ (“less than 0.9 acre” to 2 acres). Walkinshaw, 1945.

Song Sparrow iMelospiza melodia) — For mainland, 2,000 to 6,000 nr (0.5 to 1.5 acre),

Peter E. Potter

SAVANNAH SPARROW TERRITORIES

55

Distribution of Territories	Table 1 According to	Size and Presence of Nests
Size (m2)	Territories without nests	Territories with nests	Total no. of territories in size range	% of all territories	Nests found	% of territories in size range with nests
0-600	8	1	9	15	1	11
601-1200	16	20	36	58	27	56
1201-1800	1	8	9	15	12	89
1801-2400	1	4	5	8	5	80
2401-3000	1	2	3	5	2	67
Totals	27	35	62	101	47

ij Nice, 1937; for lakeshore, 1,250 to 2,750 (0.31 to 0.68 acre), Suthers, 1960; for island,

160 (0.04 acre). Beer, et al., 1956.

Nest occurrence. — Eifty-four Savannah Sparrow nests were found. Behavior by adult birds indicated the probable existence of 15 more nests, for a total of 69. Thus the nests found represented about 80 per cent of those believed to have been in the field.

A breakdown of territories by size and known presence of nests is presented in Table 1. Only 47 of the 54 nests found are included. The other seven were in five territories also not included because of inadequate marking or because the nests were discovered too late to map the territories. Figure 1 shows the territories for the three years of the study.

As might be expected, most of the nests were found in the size range which also included a majority of the territories — 601 to 1,200 m“. But a comparison of the percentages of nest occurrence in the several size ranges revealed a roughly similar distribution (56 to 89 per cent) except where territories were smaller than 601 m“. Only one of the nine territories in that range had a nest, a distribution of only 11 per cent.

I F emale occurrence. — The same pattern of distribution could he applied to

! the presence of female Savannah Sparrows in the territories. This was so ■ because in only nine out of 45 territories in which adult females were known I to be present were there no nests found, and even in eight of those nine hehav- i ior of the adult birds indicated the probable existence of nests.

It appeared, therefore, that the size of the territory had some influence on ; the attraction of a female, with territories larger than 600 m- being more I attractive.

Territorial compression. — Two males experienced severe territorial com- pression.

In 1965, M-21 attracted a mate, F-23, to a territory originally 890 m- in

56

THE WILSON BULLETIN

March 1972 Vol. 81, No. 1

Fig. 1. Savannah Sparrow territories in field near Ann Arbor, Michigan, 1965-67. Numbers = identified males. U’s i= unidentified males. Dots = nests. Broken circles surround nests found too late to map territories. Broken territory at bottom, 1967— Male 38, is estimated from partial sightings.

Peter E. Potter

SAVANNAH SPARROW TERRITORIES

57

I

I size. Much of this was later used by another male as part of his own territory, , and M-24’s area shrank to 200 m“. M-24 and F-23 apparently nested once but I abandoned the field after 26 June. The following year, M-24 returned to the same spot, established a territory only 360 m^ in size and did not mate.

The other male, M-40, first established (in 1965) a 970 m“ territory, and F-35 became his mate. Encroachments by other males establishing adjacent territories compressed M-40’s area first to 360 m“ and then to only 200. Never- theless, there was at least one nest and probably two. In 1966, however, M-40 returned to the same spot, established a territory only 120 m“ in size and did not mate. He was surrounded by five other territories, the males all aggressive.

I Neither M-24 nor M-40 returned to the field in 1967.

(The original sizes of their territories in 1965, before compression, are used in Table 1, since these sizes existed when the females were attracted to the areas and began nesting.)

! Territorial expansion. — With the exceptions just discussed, early-arriving [| Savannah Sparrows did not seem consistently to claim large areas that were ! later scaled down by population pressure, although there sometimes was con- I siderable border adjustment at the beginning. On the contrary, there seemed to be room between most territories for the small expansion the male frequently I indulged in at the onset of a second nest.

However, 1 was not able to determine whether part of the old territory was i abandoned so that the total area remained the same size. This was because ! once his territory was established, each male favored only certain perches.

I Late arrivals (there were attempts to establish new territories even in July) i would sometimes choose unclaimed areas between territories and attempt to : crowd their way in, expanding to either side and reducing the sizes of the [ adjacent territories. If the unclaimed spaces were small to begin with and the 1 attempts at expansion failed, the late arrivals were often gone the next day hut sometimes stayed as long as two weeks.

Abandonment of territories. — Abandonment by one mate or the other is impossible to prove except when a missing bird shows up elsewhere. Otherwise,

, predation is assumed to be the cause of disappearance. During this study no males were proven to have abandoned well-established territories, although three disappeared, all in 1967. One of them had a mate, which disappeared nine days before the male and long before the usual departure time.

After having successfully reared a brood in 1966, F-69 followed her fledg- lings into the adjacent territory of M-6 1 and remained there to mate with him for a second, successful nest. Deserted M-lo spent the rest of the season singing in his own territory hut did not accjuire another mate. A similar occurrence was noted among Field Sparrows by Walkinshaw ( 1915).

I Another female, F-20, disappeared after her first nest in both 1965 and 1966

58

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

with the same male, M-21, although the male remained each time. Oddly enough, after M-2Us first mate of 1967 (not E-20) disappeared after laying four eggs, F-20 reappeared to mate with him again for the second nest, which was successful.

SUMMARY

Territorial behavior of Savannah Sparrows in a field in southeastern Michigan was ob- served for three successive breeding seasons.

Most males arrived within a 10-day period in mid-April. Females arrived over a one- month period starting at the end of April. The males established territories immediately, often returning to the same area of the field claimed in previous years.

Males defended their territories by singing from border perches, chasing intruders, walking side-by-side along the boundaries with males of adjoining territories or by as- suming threatening postures face-to-face at the borders.

Birds seldom left their territories except under stress, but adults accompanying fledg- lings could cross boundaries with impunity.

Fifty-eight per cent of the territories ranged in size from 601 to 1,200 m^ Fifteen per cent were smaller, 27 per cent larger. There was some enlargement between nestings.

The Savannah Sparrow territories observed were considerably smaller than those of 10 other species of sparrows reported in the literature.

The success of attracting a mate was apparently linked to territorial size, with a better chance in territories larger than 600 nr.

Nesting activity tapered off in late July, territorial defense ended and the males molted. The females left the study area usually within two weeks after their last nests were emptied, through June and July. The males usually remained about a month after the last nests were emptied, most of them departing the last two weeks in July.

ACKNOWLEDGMENTS

My thanks go to Harrison B. Tordoff for his guidance and encouragement in this study. In addition, he read the manuscript, as did Harold Mayfield, and both have my gratitude. Robert S. Butsch made helpful suggestions about mapping the study area. Edwin G. Voss and Rogers McVaugh assisted with identification of vegetation. James Baird provided many references, including his own manuscript on the Savannah Sparrow, at the beginning of my study. Library assistance was provided by Norman Ford and Sheldon Miller of the Josselyn Van Tyne Library at the Lfniversity of Michigan Museum of Zoology. I am also indebted to Edwin Aprill, who permitted the use of his field for this study.

LITERATURE CITED

Beer, J. R., L. D. Frezel, and N. Hansen. 1956. Minimum space requirements of some nesting passerine birds. Wilson Bulk, 68:200-209.

Borror, D. J. 1961. Songs of finches (Fringillidae) of eastern North America. Ohio Journ. Sci., 61:172.

Cartwright, B. W., T. M. Shortt, and R. D. Harris. 1937. Baird’s Sparrow. Trans. Roy. Canadian Inst., 21: Part 2:163-197.

Heydweiller, a. M. 1935. A comparison of winter and summer territories and seasonal variations of the Tree Sparrow. Bird-Banding, 6:1-11.

i

I KiL'r'' SAVANNAH SPARROW TERRITORIES 59

1 Murray, B. G., Jr. 1967. A comparative study of the LeConte’s and Sharp-tailed Spar- rows with comments on the ecology of sympatric species. Unpubl. Ph.D. thesis, Univ. j Michigan, Ann Arhor.

' Nice, M. M. 1943. Studies in the life history of the Song Sparrow. II. Trans. Linnaean Soc. New York, 6:152.

Robins, J. D. 1971. A study of Henslow’s Sparrow in Michigan. Wilson Bull., 83:42-48. Smith, R. L. 1963. Some ecological notes on the Grasshopper Sparrow. Wilson Bull., 75:159-165.

! SuTiiERS, R. A. 1960. Measurement of some lake-shore territories of the Song Sparrow. Wilson Bull., 72:232-237.

Walkinshaw, L. W. 1944. The Eastern Chipping Sparrow in Michigan. Wilson Bull., 56:193-205.

Walkinshaw, L. W. 1945. Field Sparrow 39-54015. Bird-Banding, 16:1-12. WooLFENDEN, G. 1956. Comparative breeding behavior of Ammospiza caudacuta and i A. maritima. Univ. Kansas Puhl. Mus. Nat. Hist., 10:45-75.

2518 E. HAMPTON ST., TUCSON, ARIZONA 85716, 1 MARCH 1971 (ORIGINALLY RE- CEIVED 6 OCTOBER 1969) .

FLOCKING ASSOCIATES OF THE PINON JAY

Rl ssell P. Balda, Gary C. Bateman, and Gene F. Foster

h-|^HE Pinon Jay [ Gymnorhinus cyanocephalus ) is a noisy, restless bird that normally forms large flocks. Our investigations of the annual flocking cycle of this corvid (Baida and Bateman, 1971) showed that several other species regularly joined and foraged with flocks of Pinon Jays.

In most interspecific flocks of the North Temperate Region reported on to date (Odum, 1942; Wing, 1941; Austin and Smith, in press; and the exten- sive review hy Morse. 1970 ( the species involved are mainly insectivorous, flocks form primarily in the fall or winter, and the “flock leaders” or “nuclear species” are not present in overwhelming densities compared to the associate species. By contrast, this report deals with five associate species that join relatively large flocks of Pinon Jays: Hairy Woodpecker { Dendrocopos

villosus). Downy Woodpecker {Dendrocopos pubescens ) . Red-shafted Flicker iColaptes cafer) . Clark’s Nutcracker { Nucifraga columbiana ) , and Starling iSturnus vulgaris). The Pinon Jay flock is maintained in a number of forms throughout the year, thus permitting interspecific association the year around. The main foods of the Pinon Jay during the fall and winter months when attendant species are most numerous are seeds of ponderosa {Finns ponder- osa) and pinon ( P. edulis) pine, and occasional arthropods ( pers. observ. I .

The efficient procurement of food has often been used as at least a partial explanation for flocking (Miller, 1921; Rand. 1954; Short, 1961; Morse. 1970). Our observations were made on two Pinon Jay flocks, one on its un- disturbed home range and the other when it visited a local feeding station where food was diverse and super-abundant. Comparisons were made of the foraging and agonistic behavior of the jays and associates in both situations.

STUDY AREAS AND PROCEDURES

We studied intensively a flock of about 250 Pinon Jays on a home range of eight square miles located 10 miles NE of Flagstaff, Arizona for over 480 hours from February 1968 through January 1971. Movements, foraging sites (ground, trunk or branch, tip of foli- age), and intra- and inter-specific social interaction were recorded. At periodic intervals foraging sites were recorded hy counting all birds foraging at each site. Six hundred and forty-nine counts of the entire flock were made in this way. Aggressive encounters, either “supplantings” (overt chases) or “displacings" (retreats) were recorded (after Willis. 1966). We also noted reactions to potential predators and stuffed Great Horned Owls ( Bubo virginianus) .

The third author has a 0.25-acre feeding station which was visited almost daily by a flock of about 70 Pinon Jays for the past five years. On some fall and winter days the flock visited the station up to four times daily; during spring and summer groups of young and adults often spent hours at the station. A number of different foods including

60

Baida, Bateman and Foster

pinon jay flocking associates

61

Table 1 Foraging Locations of Members of the Large Pinon Jay Flock Year (in per cent)	Throughout the
Foraging Location	Jan., Feb., March	April, May, June	July, Aug., Sept.	Oct., Nov., Dec.
Ground	40.0	59.4	48.9	39.7
Foliage	31.8	23.3	42.1	35.6
Crevices	29.1	12.8	4.1	24.7
Hawking In Air	0.0	4.4	4.9	0.0
Total Number
of Counts Made	213	172	116	148

mealworms, sunflower seeds, Spanish peanuts, commercial pigeon grain, bacon grease- bread crumb-popcorn mix, white millet, pinon pine seeds, raisins, and suet were always available at the feeding station. Qualitative notes were kept concerning the behavior and aggressive interactions of the Pinon Jays and four of the associates. The Clark’s Nutcracker did not visit the feeding station.

FLOCKING CYCLE OF THE PINON JAY

Descriptions of interspecific flocks often include a designation of one or more species as nuclear species without adequately describing the movements and behavior patterns of these important species in mixed flocks. We have described the flocking cycle of the Pinon Jay elsewhere (Baida and Bateman. 1971). Here we will only summarize and enlarge on behavior patterns essen- tial to understanding the role of Pinon Jays as a nuclear species in mixed flocks.

Fall and early winter. — During this period blue adults and gray first-year birds formed a loosely organized flock which foraged primarily in ponderosa pine forest. During early morning feeding the flock moved at an average rate of about one mile per hour. Short flights below tree-top level advanced the birds in either leapfrog fashion or as a broad front with all members simul- taneously moving in one direction. Longer flights taken over large meadows often consisted of rolling and swirling movements and were accompanied by loud calling of the flock members. Flocks moved up to 13 miles })er day while foraging. In the forest some of the flock walked on the ground, either probing for insects and/or pine seeds or caching pine seeds, while other mem- bers of the flock foraged off the ground. Some of these gleaned in the foliage, hammered open pine cones to extract seeds or tore out the tender new growth at the tips of the branches. I he rest of the birds j)icked food items out of crevices on the trunks and branches, or hammered vijiorouslv to flake bark to

62

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Table 2 Number of Individuals and Frequency of Occurrence THE PiNON Jay Flock	of Attendant	Species with
	Jan., Feb.,	April, May, J	lily, Aug.,	Oct., Nov.,
	March	June	Sept.	Dec.
	n = 49*	n = 77	n = 29	n = 64
		Hairy Woodpecker
Average Number
(when present)	5(2-7)=*=*	1(1)	0	4(1-7)
Frequency of
Association (%)	100***	14	0	88
		Downy Woodpecker
Average Number	2(1-3)	1(1)	0	1(1-2)
Frequency of
Association	80	6	0	42
		Red-shafted	Flicker
Average Number	6(4-9)	5(3-7)	2(1-4)	5(3-9)
Frequency of
Association	100	71	62	81
		Clark’s Nutcracker
Average Number	1(1-2)	2(1-3)	9(6-15)	7(4-12)
Frequency of
Association	12	16	45	67
		Starling
Average Number	7(3-14)	9(5-16)	0	4(2-7)
Frequency of
Association	39	74	0	28

* Number of visits.

** Range.

*** Per cent of visits when associates were present.

extract food (Table 1) . All five associate species foraged with the large Pihon Jay flock at this time (Table 2) .

During this period, the feeding station was visited from one to four times daily by a flock of 70 Pihon Jays. While at the station the birds fed on pihon seeds, bacon grease-bread crumb-popcorn mix, peanuts, sunflower seeds, suet, and mixed small grains in that order of preference.

Winter and early spring. — During courtship which commenced in mid- December adult blue birds radiated out in pairs from the feeding flock leaving the gray first-year birds plus a few blue birds to forage as a unit. The foraging flock varied in size from 35 to 70 birds. The foraging birds spent about equal

Baida, Bateman and Foster

pinon jay flocking associates

63

time on the ground and in the foliage. The birds feeding at arboreal sites divided their activities almost equally between foliage gleaning and cone feeding, and crevice picking and bark flaking. The entire flock reassembled periodically and moved to a new feeding location at a loud rapidly repeated krawing signal given by most individuals in the feeding flock. The resultant din could be heard for over three-quarters of a mile. During the nest building period the flock fed as a unit in the morning and evening. However, small groups of 4^12 birds often formed autonomous feeding units. Throughout the incubation period the main flock was divided into incubating females, a flock of adult males seeking and bringing food to the females and a flock of gray yearling birds. This latter subunit foraged quietly and moved rather long distances per flight.

When nest building began the visits of the smaller flock to the feeding sta- tion diminished to one each morning and evening. Small groups of jays, however, visited the station throughout the day. Later, males visited the feeding station regularly.

Late spring and summer. — After the young fledged, family groups of adults and juveniles foraged together as a unit. Adults failing in their first nesting attempt formed smaller nesting colonies and subsequent family-group feeding flocks. By late July the single winter flock was divided into a flock of year- old birds that did not breed, and five or six independent feeding groups. In late July or early August these flocks moved into the pihon-juniper woodland, where the birds opened pinon pine cones, extracted seeds and carried them into the ponderosa pine forest where they were cached. From this time on the birds remained together as a large flock.

ANTI-PREDATOR BEHAVIOR

Protection from predators is often described as a benefit of inter- and intra- specific flocking (see Morse, 1970 for discussion). Pinon Jays have two be- havioral mechanisms which can be termed anti-predator behavior. 4 hese are in addition to the protection afforded the birds by their mere presence in a flock ( Allee, 1938; Tinbergen, 1953).

Sentinels. — Throughout the year each subflock (feeding grou}); gray year- ling flock) and the entire flock when assembled was commonly surrounded by sentries as reported by Cary (1901). The number of sentries was rather constant around feeding aggregates and the yearling flock (3-5 birds) hut varied greatly (3-12 birds) around the large feeding flock that existed during the non-reproductive period. Sentinels were positioned at high vantage j)oints. either exposed or concealed in foliage. At the a|)proach of an aerial or terres- trial intruder the sentinel(s) gave a loud rhythmic krawk-krau-krawk which was often repeated. On occasion, a ground-feeding bird also gave this warning

1

64

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

call. This call was often sufficient to cause an immediate cessation of feeding and flight of all individuals up into the trees. During these rapid ascents, birds flew in all directions and it would seemingly be difficult for a predator to focus on and capture an individual. If the warning call was not repeated the flock resumed feeding. Individuals foraging in the trees when a warning call was given simply stopped feeding and remained still. The associate species re- sponded to the warning calls in the same manner. Even though Steller’s Jays did not participate in the activities of the mixed flock they responded to the warning calls. Pihon Jays in turn responded to the shook call (Brown, 1964 I given by the Steller’s Jay at hawks or owls.

Mobbing,. — After the rhythmic warning call was given a number of birds (3-15) including the sentinel! s) often approached the intruder, circling it if it was perched or on the ground. If flying or running the intruder was chased. During this performance the mobbing Pihon Jays called loudly, often at- tracting numerous other birds including Steller’s Jays, Red-shafted Flickers, Grace’s Warblers, Chipping Sparrows, Acorn Woodpeckers, and j uncos. Hawks and owls that flew off in response to this harassment were always chased by the Pihon Jays. The Sharp-shinned Hawk {Accipiter striatus) and Cooper’s Hawk {Accipiter cooperii) often evaded the jays by flying an erratic but rapid course then landing and sitting quietly in a camouflaged location. Red-tailed Hawks {Buteo jamaicensis) and Rough-legged Hawks (Buteo lagopus) usually left the area by gaining elevation rapidly and then moving off. Great Horned Owls, however, seldom flew long distances and could not evade the jays. Consequently, Pihon Jays often mobbed them for up to 45 minutes.

FLOCKING ASSOCIATES

The following accounts are only for the five attendant species ( Moynihan, 1962) which regularly occur with the Pihon Jay (passive nuclear species, cf. Moynihan, 1962 ) flock at least for a portion of the year but are not important for the maintenance of the flock.

Hairy Woodpecker. — One to seven individuals of this species were constant members of the jay flock from late October through early March (Table 2). An occasional individual accompanied the non-breeding flock during the spring and early summer. During this woodpecker’s nesting period it did not associate with the flock. Nesting alone, however, cannot explain its seasonal appearance, as it left the flock before it began courtship and did not enter the flock until well after all its nesting duties were completed. During the period of association, however, the foraging pattern of the Pihon Jay was similar to that of the Hairy Woodpecker.

During fall and winter the jays spent considerable time searching the deep

Baida, Bateman and Foster

pinon jay flocking associates

65

crevices of the ponderosa pine trunks, hammering and flaking bark to extract food from old dead branches and stumps, and extracting seeds from ponderosa pine cones. Whether foraging alone or with the jay flock. Hairy Woodpeckers used these same sites. Stallcup (1969 ) reported Hairy Woodpeckers spending 64.5 per cent of their time extracting seeds from ponderosa pine cones in Colorado. Stallcup’s figures indicate that Hairy Woodpeckers spent about 83 per cent of their time foraging in the three sites listed above for the Pinon Jay. He noted that feeding on cones occurred mainly from mid-December through February, the very time these woodpeckers associate with the Pinon Jay flock in central Arizona. He reported as did Morse (1970) that the Hairy Woodpecker foraged throughout the winter in mixed flocks. Short ( 1961 ) reported the Hairy Woodpecker as a member of mixed flocks in Oaxaca, Mexico.

Hairy Woodpeckers were seen with the jay flock from sunrise to sunset and remained within the flock as it moved about in the forest and woodland. Interaction at foraging sites was minimal except at pine cones, where the jays successfully drove off the woodpeckers. Of 29 aggressive interactions ob- served, single jays or groups of jays were able to supplant the Hairy Wood- peckers 20 times. Nineteen of these encounters occurred at pine cones. Hairy Woodpeckers supplanted jays on 9 occasions. At other feeding sites wood- peckers of this species were always displaced by groups of seven or more jays.

During the short movements of the jay flock, the Hairy Woodpeckers always followed the Pinon Jays. The woodpeckers did not follow the flock when it made longer flights across fields, but remained in the trees at the edge of the field, calling loudly as the flock departed. Once, after the flock crossed a field one-quarter mile in width, three Hairy Woodpeckers rapidly flew around the edges of this small field to rejoin the jays. We have followed individual woodpeckers that spent four consecutive hours and traveled at least five miles with the flock.

At the feeding station, resident Hairy Woodpeckers fed alongside the Pinon Jays until the jays became too numerous at one location. Then displacement occurred and the woodpeckers perched silently in the trees until the jays left the station. The jays clearly dominated the woodpeckers; on one occasion an adult Pinon Jay took eleven peanuts, consecutively, from the hill of a Hairy Woodpecker. The woodpeckers did not come to the feeding station with the jays nor did they leave with them.

Downy Woodpecker. — This species associated with the jay flock during roughly the same months as did the Hairy Wood|)ecker (Jahle 2). Its for- aging mode was somewhat different, however, as it spent most of its time on the trunks and branches of the ponderosa pines and on the dead trt‘es. where it gleaned and flaked hark in search of food. Often it picked through piru‘

66

THE WILSON BULLETIN

March 1972 Vol. 81, No. 1

cones on the ground, but it seldom worked on cones up in the foliage. Downy Woodpeckers occur in low densities in this area; consequently, more than three individuals were never seen with the jay flock at one time. Of 14 interactions between jays and this woodpecker, the latter was displaced 12 times and sup- planted twice. Seven of these interactions occurred while this woodpecker fed on some object either on the ground or a short distance from it. Most direct conflicts were avoided because the Downy Woodpecker managed to stay away from Pihon Jays when they fed close together in groups. In other respects this species acted similarly to the Hairy Woodpecker. The calls of both species evoked no noticeable reactions from the jays.

At the feeding station the Downy Woodpecker did not feed at its usual sites when jays were present. It always left the area when the jays entered the station and returned when the jays left.

Red-shafted Flicker. — Flickers were the most regular associates of the Pihon Jay flock. Even during their breeding season a few flickers were almost always with the non-breeding gray bird flock ( Table 2 ) . During fall, winter, and spring as many as nine individuals were in constant association with the jay flock. One individual that was specifically followed spent seven hours with the jay flock and moved about nine miles with it.

The Red-shafted Flickers spent most of their time foraging on the ground among the jays. Their soil-probing activities greatly resembled those of the Pihon Jay. During slow movements through the forest and woodland the flickers flew with the group and were never segregated at the periphery or rear of the flock. During the winter months. Red-shafted Flickers spent con- siderable time probing into decaying logs for immature insects. This activity strongly resembled that of the Pihon Jay when caching food in these sites. Aggressive encounters were observed when jays and flickers foraged on the ground; groups of jays were observed driving flickers from cache sites in decaying logs. The jays either pointed their bills at the flickers or flew up at them. When a single jay came in contact with a flicker (n = 48 ) Pihon Jays were displaced or supplanted 46 per cent of the time, while jays dominated flickers 54 per cent of the time. When the jay flock moved over large fields some flickers often accompanied them, but others stayed behind, calling loudly as the flock departed. When sentry jays along the edges of the feeding flock gave their rhythmic kratvk-kraw-krawk, signaling the approach of a potential predator, the flickers responded immediately by flying up into the trees in the same manner as the Pihon Jays. When the warning calls subsided, the Red-shafted Flickers returned to foraging on the ground with the jay flock. Thus, their movements between feeding sites, as well as their movements within the flock when it was stationary, were carried out in synchrony with the Pihon Jays and in a similar fashion.

Baida, Bateman and Foster

pinon jay flocking associates

67

During the non-breeding season the Red-shafted Flickers appeared to be paired, a male and a female often foraging near one another. On one occasion in May, a feeding group of jays moved through an area where a pair of flickers was excavating a hole. The birds stopped working, flew into the aggregate, and foraged with them for at least the next hour.

At the feeding station Red-shafted Flickers fed near the jays but did not enter or leave with them. The flicker used bill pointing and thrusting to sup- plant Pinon Jays when it was not badly outnumbered, however a flicker always retreated from groups of 11 or more jays.

Red-shafted Flickers are strongly attracted to Pinon Jay flocks (Table 2), and during the non-breeding season it was rare to find a solitary flicker or pair of flickers far from the jay flock. J. D. Ligon (in litt.) observed the same phenomenon in New Mexico. Short (1961) described the Red-shafted Flicker as an irregular attendant of mixed flocks in Oaxaca, Mexico. Its be- havior in the vicinity of Pinon Jays appears to be quite different.

Clark’’ s Nutcracker. — Nutcrackers descended the slopes of the San Francisco Peaks in late August to collect pinon seeds and carry them up the mountains to about 10,500 ft, where they were cached. During this period of seed col- lecting the nutcracker opened the green cones in such a manner that in poor light it was impossible for us to distinguish nutcrackers from Pinon Jays. The jays and nutcrackers worked on the pinon cones in close association, yet no aggressive interactions were noted. Johnson (1902) commented on such an association in central Utah. On one occasion a yearling Pinon Jay watched from a distance of about one meter as a Clark’s Nutcracker opened a cone. At intervals spanning seven minutes the young jay fluttered its wings and begged softly while facing the nutcracker. The latter did not react to this begging. As the jay flock moved between feeding sites up to 15 nutcrackers moved with the flock. They responded to the danger krawks of the Pinon Jay by dropping the cones they were extracting seeds from and flying up to the tops of trees. They returned to seed collecting when the jays did. Twice the jay flock left the woodland and flew more than two miles to a watering hole, with eight Clark’s Nutcrackers accompanying them. During these flights, the low throaty calls of the nutcrackers could be distinguished from the krawks given by the jays. The nutcrackers were always in the rear half of the flock during these flights.

During the spring and summer of 1969 from one to three nutcrackers were often with the yearling flock and also with feeding groups. I he usual raucous calls given by nutcrackers during foraging and flight were not heard from these individuals. While foraging on the ground, thev j)erformed |)robing. insect capturing, and seed opening much as did the Pinon javs.

Starling. — Beidleman and Enderson (1961) first described tlu‘ associalioti

68

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

of Starlings with a flock of Pihon Jays in Colorado. In central Arizona, from 3 to 16 Starlings were associated with the jay flock during March, April, and May, and again during October and early November. Most Starlings left the jay flock during the nesting period and again in early winter when they be- came rare in rural central Arizona and conversely very common in the cities. Throughout the spring and summer months Starlings nested within the home range of the Pihon Jay flock but did not associate with it.

In late winter of 1968 the Starlings were first observed with the Pihon Jay flock when the male jays were roosting as a group and the females were incu- bating. During this period the Starlings roosted in holes, and on three morn- ings they stayed in their holes until the male jays called loudly and moved out of the forest to feed for the first time. The Starlings’ initial response to these calls was to look out of the holes, squawk loudly, and fly directly to the flock of feeding jays.

Late in the winter of 1969 Starlings were first noted in the Pihon Jay flock at the time courtship activities had commenced. After feeding in a very deliberate fashion with the jays on the ground for an hour in the morning, the Starlings began courting. Pairs segregated from the jay flock and courted high in the foliage and examined old woodpecker holes. The Starlings’ initi- ation of courtship agreed closely with the beginning of the daily courtship of Pihon Jay pairs. Courting activities were noted for six to ten Starlings each morning, and indicate not only a strong attraction to the jay flock, but also a close synchronization of daily events. The synchrony may be coinci- dence but also suggests the Darling effect ( Darling, 1938 ) .

During foraging the Starlings walked slowly and probed for seeds and insects in the same manner as the jays. Not only was their gait similar to that of the Pihon Jay, but in short flights made between feeding sites the Starlings displayed a very similar pattern of flight. At take off, both species beat their wings rapidly, but during sustained flight strong wing beats alternate with gliding. Neither of these species undulates in flight as do most woodpeckers, as the wings are partly outstretched during the glide phase of the flight. Jen- sen (1926) and Wetmore (1920) have pointed out these behavioral similar- ities. Under cloudy conditions, or when the jays and Starlings moved through heavy foliage, it was difficult to tell them apart.

Aggressive encounters between Pihon Jays and Starlings were not common as a Starling was not easily displaced by the mere presence of a Pihon Jay. Of 51 aggressive encounters observed, the Pihon Jay supplanted or displaced the Starling 57 per cent of the time; at least five other encounters resulted in both individuals leaving the area.

At the feeding station Starlings associated with the jays from November through mid-April. During the early winter. Starlings commonly entered

Baida, Bateman and Foster

pinon jay flocking associates

69

Solitary Species Occurring	Table 3 IN THE Habitats used by Pinon Jays
Species	Mean Weights (g)
Selasphorus platycercus	3.4**
*Colaptes cafer	110.9
Sphryapicus varius	50.7
*Dendrocopos villosus	64.3
*Dendrocopos pubescens	27.8
Tyrannus vociferans	47.0
Contopus sordid ulus	13.7
Tachycineta thalassina	10.6
Cyanocitta stelleri	113.0
Purus gambeli	28.2
Sitta carolinensis	18.1
Certhia familiar is	7.8
Regulus calendula	6.5
Lanius ludovicianus	47.0
Dendroica auduboni	12.6
Dendroica graciae	7.8
Vireo solitarius	17.0
Piranga ludoviciana	29.7
Piranga jlava	37.6
Chondestes grammacus	26.1
Spizella passerina	13.2
Total Number of Species = 21
Number of Associates r= 3

* Indicates associates of Pinon Jay flocks.

** Sources for weights in this table are Baldwin and Kendeigh (1938), Hartman and Brownell (1961), Miller (1955), Poole (1938), Salt (1957), Hubbard and Ligon (in litt.). Whenever possible weights were obtained from specimens in the Museum of Northern Arizona and the Northern Arizona University Museum of Vertebrates.

and departed from the station with the jay flock. However, in late winter and early spring Starlings were much more prone to stay at the station. Early on winter mornings Starlings gathered just outside of the station but would not enter until the jay flock entered. If the jays did not appear by 09:30 the Starlings left without feeding at the station. When feeding at the station, Starlings mingled with even the largest groups of Pinon Jays and were not displaced.

On two occasions during the winter of 1969, groups of Pinon Jays were seen associating with an urban flock of Starlings. On both occasions, the flocks contained about 55 Starlings and eight to ten yearling Pinon jays. I he flocks moved silently through a forested area.

70

THE WILSON BULLETIN

March 1972 Vol. 84, i\o. 1

Table 4 Gregarious Species Occurring in the Habitats used by Pinon Jays
Species	Mean Weights ( g )
Zenaidura mucroura	122.8**
Melanerpes jormicivorus	66.0
Eremophila alpestris	43.0
Corvus corax	969.0
Conus brachyrhynchos	479.0
*Nucifraga Columbiana	142.2
Psaltriparus minimus	5.8
Sitta pygmea	9.9
T Urdus migratorius	80.7
Si alia mexicana	24.6
Sialia currucoides	34.7
Bombycilla cedrorum	32.9
*S turn us vulgaris	81.9
Sturnella magna	145.0
Molothrus ater	50.5
Euphagus cyanocephalus	64.8
Hesperiphona vesper tina	53.6
Carpodacus cassinii	27.5
Spinus pinus	12.2
Spinus psaltria	10.4
Junco hyemalis	21.0
Junco oreganus	17.4
Junco caniceps	19.7
Total Number of Species = 23
Number of Associates = 2

* Indicates associates of Pinon Jay flocks. ** Same as Table 3.

DISCUSSION

Of the five species that associated with the jay flock, three are usually solitary, whereas the other two are often found in intraspecific associations (pers. observ.; Tables 3 and 4). Some of the species listed as solitary in Table 3 form intraspecific flocks at times of the year when not in the vicinity of Pinon Jays. Moynihan (1960) suggests that “many but not all species” that tend to form intraspecific flocks may also form interspecific flocks. Our data show, however, that 14 per cent of the solitary species and 9 per cent of the gregarious species that come in contact with the jay flock do associate with it. Innate social attraction cannot be used to explain interspecific flock- ing with Pinon Jays.

Balila, Bateman and Foster

pinon jay flocking associates

71

The ability of associate species to mingle and remain with the Pinon Jay flock is probably enhanced by the lack of intraspecific aggression among Pinon Jays. At any one time, less than 5 per cent of the jay flock was involved in intraspecific hostile behavior. Pinon Jays displayed similar ag- gressive behavioral patterns both intra- and interspecifically. These entailed crouching slightly, pointing the bill at the agressee and lunging, or flying up at an approaching intruder with legs extended and calling loudly. A direct thrust with the bill is also used to supplant other birds. These patterns could be easily learned and adjustments readily made. The Red-shafted Flicker and Starling used these same general agonistic behavior patterns to displace Pinon Jays. If the aggressive behaviors are easily learned or already in the behavioral repertoire of the species, actual combat that can result in injury and/or exhaustion is reduced or avoided (Moynihan, 1962). Once an inter- specific association is established, the Pinon Jays tolerate the associate species and act with the same low level of aggressiveness towards them as to conspecifics. Therefore the associate species can efficiently reap what benefits are available without expending undue energy. In this regard, the Starling which arrived in northern Arizona in the early 1960’s (pers. observ., G. F. Foster) has had only 10 years to learn and adjust to the behavior patterns of the Pinon Jay. Yet in many respects the Starling has the highest degree of behavioral compatibility with the jay flock. This must be due to the behavioral plasticity or preadaptation of this species.

The numerically superior Pinon Jay is also the socially dominant species in mixed flocks, in part because with superior numbers it can displace those associates individual Pinon Jays could not dominate. The associate species rank in an interspecific hierarchy ( based partly on compatibility and tenacity when faced with large numbers of jays) as follows: Red-shafted Flicker, Starling, Clark’s Nutcracker, Hairy Woodpecker, and Downy Woodpecker. The more abundant associates tend to bave higher ranks.

Although the Pinon Jay is largely passive in its behavioral relations with the five associate species, it does possess many of the traits discussed by Moynihan (1960, 1962 ) which promote both intra- and interspecific gregari- ousness. The general noisiness and restlessness of the jay flock tend to focus attention on it. The neutral, rather drab blue coloration of the Pinon Jay may act as an attractant to species that are normally repulsed by a shar})ly contrasting plumage. The dorsal blue or blue-gray coloration is similar to that found in species that form mixed flocks in the Andes and Bolivia (Moynihan, 1968). The associate sj>ecies show some of these same traits and others, including striking flash patterns on either wings, rump, or tail.

In our opinion the most important characteristics promoting this association are similarities of foraging strategies and similarities in size. Lsing the

72

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Table 5

Indices of Specialization (J') of Foraging by Pinon Jays and Sum Frequency of Association

	Jan., Feb., March	April, May, June	July, Aug., Sept.	Oct., Nov., Dec.
J'	0.992	0.757	0.716	0.983
Sum* Frequency	3.31	1.81	1.07	3.06
* From Table 2;	Average number of	species to be seen	with the flock	during this period.

suggestions of Pielou (1966) we calculated the foraging diversities (H'j and indices of specialization (J') (after Willson, 1970) of the Pinon Jay flock for four different periods of the year (Table 5). The higher the Y the less specialized and consequently more diverse the foraging pattern. J' was then compared to the sum frequency (see Table 2; expected number of associate species to be found with the Pinon Jay flock) and a very high positive corre- lation results. That is, when Pinon Jays are most diverse in their foraging sites, the number of associates is highest.

Numerous workers have pointed out the similarities in body size and weight of members of interspecific flocks. Tables 3 and 4 list weights for the species that occur in the home range of the Pinon Jay at least a portion of the year. The average weights of the associates range from a low of 28 g for the Downy Woodpecker to a high of 142 g for the Clark’s Nutcracker. If we eliminate the Downy Woodpecker from this comparison because of its low numbers and obviously low social status, as indicated by the outcomes of interspecific hostile interactions, the weight range for the other four species is 64-142 g. The average weight of 27 adult Pinon Jays is 108 g, almost exactly intermediate to the weight of the associates. This range includes five potential associates. Mourning Dove, Acorn Woodpecker, Steller’s Jay, Robin, and Brewer’s Blackbird that do not associate. The Steller’s Jay is found in high numbers year round, but appears to maintain definite winter home ranges. The other four species are either present in very low numbers through- out the year or are present only during the nesting season when they show strong affinities for nests or territories. Rather than join the flock, these birds all show signs of alarm when the jay flock comes into proximity with them. The typical response was to scold loudly and leave the area. During the warm winter of 1970-71 flocks of Robins occasionally mingled with the jays at watering or feeding sites but did not follow them. Thus, size must be only a secondary factor in determining flocking associates.

Bahia, Bateman and Foster

pinon jay flocking associates

73

Flocking of the five species with Pihon Jays is probably due to their join- ing the jay flock when food is scattered widely throughout the habitat. The associates are then assured a share of the food. When food is super-abundant, as at the feeding station, attraction to the jay flock is not as evident. This assurance is best demonstrated in those cases of a species associating with the jays when they performed a specific type of foraging. The woodpeckers are most closely associated with the flock during the winter when many jays forage off the ground by flaking bark, probing crevices and opening ponderosa pine cones. The flicker associates most of the year, and there is always a portion of the jay flock feeding on the ground. The nutcracker shows a bond with the jay flock during the time both species are caching pihon pine seeds. Austin and Smith (in press) have shown that some flocking species increase their foraging diversity in winter. This is true in the Pihon Jay. Morse (1970) demonstrated that the associates modify their area of foraging in the presence of socially dominant species whereas Austin and Smith (in press) believe the numerically dominant species may alter their foraging pattern to accommodate the associates. We believe the Pihon Jay increases its foraging diversity during the more demanding winter months in order to obtain an ample supply of food. This, in turn, attracts the associ- ate species. The Pihon Jay is probably more diverse in its foraging patterns than the associate species. This relationship between nuclear and associate species was also shown by Morse (1970) and Austin and Smith (in press).

The tendency of the associates to form mixed flocks is probably a species- specific trait, or set of traits expressed when advantageous, but not necessary for survival except under special conditions imposed by the local ecological situation. Harvesting of vast quantities of food by Pihon Jays may make it advantageous for other species to join them. At the feeding station, only Starlings actively joined the flock. Here food was constantly renewed and the woodpeckers and flickers did not move with the flock when it left the station but stayed to harvest the replenished food items.

Comparing the behavior of the associates at the feeding station to that of the flock in a more natural habitat, suggests that participation in the flock by the associates is directly related to the density and obviousness of the food items. When food is abundant, obvious, and easily obtained the tendency to form mixed flocks decreases. This has also been suggested for insectivorous flocks by Gibbs (1960) and Hinde (1952) .

The advantages to be obtained from the association herein described are in all probability food and protection gained by mechanisms similar to those described and reviewed by Morse (1970), for insectivorous flocks. J he above author rarely, if ever, observed ra])tors near or attempting to enter mixed flocks. In contrast, we observed raptors being scolded or mobbed on 12 per

I

74

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

cent of the observation periods, and observed potential predators on 84 per cent of our visits to the home range of the Pihon Jay flock.

When sentries gave the rhythmic danger call, associates responded by flying up into concealing foliage and remaining still. This action was spontaneous and took less than five seconds to complete. The associates never lagged behind the jays in this movement and appeared to recognize the danger call as quickly as did the Pihon Jays. Although Pihon Jays were quick to mob potential predators, only the Red-shafted Flicker and Clark’s Nutcracker par- ticipated in this behavior. Their participation in scolding and mobbing poten- tial predators was meager as they joined the jays on less than 20 per cent of the scolding and mobbing performances. Thus, the associates gained appreci- able protection from the actions of the Pihon Jays.

Associate species and also species that did not associate with the Pihon Jay flock were often, if not always, stimulated by social induction or facilitation (Rand, 1954) to feed when the jays were present. On numerous occasions Steller’s Jays and j uncos were observed to feed intently with the jays as they passed but these non-associates did not follow the flock when it departed, or only followed a short distance. Westcott (1969) made similar observations on Steller’s Jays following a Pihon Jay flock in southern Arizona. Feeding activities of these non-associates ceased when the flock departed. This behavior suggests that all birds may derive some protection from the well organized sentinel system of the Pihon Jay flock. Not only are other species induced to feed in the presence of the jay flock, but they can do so intently because the predator warning system established by the jays allows these species to con- centrate solely on feeding. One would suppose that this concentration would increase feeding efficiency. Thus, it is difficult to separate the benefits of associating with the jay flock into protection and feeding efficiency, as both appear to be important but not clearly distinguishable from each other (see Lack, 1968).

SUMMARY

The Hairy and Downy Woodpeckers, Red-shafted Flicker, Clark’s Nutcracker, and Starling were observed to form interspecific flocks with the highly gregarious Pihon Jay. The general noisiness and restlessness of the jay flock, plus the drab coloration of its members probably acted to attract the associate species. The Pihon Jay flock was intact throughout the year, although in a number of different forms, thus offering attend- ant species an opportunity to participate in mixed flocking year round. The frequency of occurrence and numbers of associates varied with season and foraging site diversity of the Pihon Jay flock. A strong positive correlation exists between foraging site diversity of the jays and frequency of the associates.

The ability of the associates to remain in the Pihon Jay flock is enhanced by the lack of intraspecific aggression among the jays.

Baida, Bateman and Foster

pinon jay flocking associates

75

The important characteristics promoting this association are similarities of foraging strategies and secondarily similarities in size.

The benefits derived by associate species as a result of interspecific flocking are prob- ably more effective utilization of the total food resources, indirectly resulting from effi- cient protection from predators while feeding and directly as a result of the greater ability of numerous individuals to locate scattered, but locally abundant, sources of food.

ACKNOWLEDGMENTS

S. Vander Wall is thanked for his valuable field assistance and J. Hubbard for supplying some of the bird weights. Earlier drafts of this paper were read and criticized by F. A. Pitelka, J. L. Brown, J. D. Ligon, and T. A. Vaughan. We thank these reviewers for their valuable suggestions and comments.

LITERATURE CITED

Allee, W. C. 1938. The social life of animals. Norton & Co., New York.

Austin, G. T., and E. L. Smith. 1972. Winter foraging ecology of mixed insectivorous bird flocks in oak woodland in southern Arizona. Condor, 74:17-24.

Balda, R. P., and G. C. Bateman. 1971. Flocking and annual cycle of the Pinon Jay (Gymnorhinus cyanocephalus) . Condor, 73:287-302.

Baldwin, S. P., and S. C. Kendeigh. 1938. Variations in the weight of birds. Auk, 55:415-467.

Beidleman, R. B., and J. H. Enderson. 1964. Starling-Pinon Jay associations in southern Colorado. Condor, 66:437.

Brown, J. L. 1964. The integration of agonistic behavior in the Steller’s Jay Cyanocitta stelleri (Gmelin). Univ. California Publ. Zook, 60:223-328.

Cary, M. 1901. Birds of the Black Hills. Auk, 18:231-238.

Darling, F. F. 1938. Bird flocks and the breeding cycle. Cambridge Univ. Press.

Gibbs, J. A. 1960. Populations of tits and Goldcrests and their food supply in pine plantations. Ibis, 102:163-208.

Hartman, F. A., and K. A. Brownell. 1961. Adrenal and thyroid weights in birds. Auk, 78:397^22.

Hinde, R. a. 1952. The behavior of the Great Tit {Pams major) and some other re- lated species. Behaviour Suppl., 2:1-201.

Jensen, J. K. 1926. The Pinon Jay {Cyanocephalus cyanocephalus). The Oologists’ Record, 6:41-43.

Johnson, H. C. 1902. The Pinyon Jay. Condor, 4:14.

Lack, I). 1968. Ecological adaptations for breeding in birds. Methuen & (^o. Ltd.,

London.

Miller, A. H. 1955. The avifauna of the Sierra del Carmen of Coaluiila, Mexico. Condor, 57:154^178.

IVIiLLER, R. C. 1921. The flock behavior of the Coast Bush-tit. (Condor, 23:121-127. Morse, 1). H. 1970. Ecological aspects of some mixed-species foraging flocks of birds. Ecol. Monogr., 40:119-168.

Moyniiian, M. 1960. Some adaptations which help to i)romote gregariousness. Proc. XII Internatl. Ornithol. Congr. : 523-541.

Moyniiian, M. 1962. The organization and jirobable evolution of some mixed species flocks of neotropical birds. Smithsonian Mise. 0)11., 143:1-140.

76

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Moynihan, M. 1968. Social mimicry; character convergence versus character displace- ment. Evolution, 22:315-331.

Odum, E. P, 1942. Annual cycle of the Black-capped Chickadee. Auk, 59:499-531.

PiELOU, E. C. 1966. The measurement of diversity in different types of biological col- lections. J. Theoret. Biol. 13:131-144.

Poole, E. L. 1938. Weights and wing areas in North American birds. Auk, 55:511-517.

Rand, A. L. 1954. Social feeding behavior of birds. Fieldiana Zook, 36:1-71,

Salt, G. W. 1957. An analysis of avifaunas in the Teton Mountains and Jackson Hole, Wyoming. Condor, 59:373-393.

Short, L. L., Jr. 1961. Interspecific flocking of birds of montane forest in Oaxaca, Mexico. Wilson Bull., 73:341-347.

Stallcup, P. L. 1969. Hairy Woodpeckers feeding on pine seeds. Auk, 86:134-135.

Tinbergen, N. 1953. Social behavior in animals. Methuen & Co., London.

Westcott, P. W. 1969, Relationships among three species of jays wintering in south- eastern Arizona. Condor, 71:353-359,

Wetmore, a. 1920, Observations on the habits of birds at Lake Burford, New Mexico. Auk, 37:221-247; 393-412.

Willis, E. 0. 1966. The role of migrant birds at swarms of army ants. Living Bird,

5:187-231.

Willson, M. F. 1970. Foraging behavior of some winter birds of deciduous woods. Condor, 72:169-174.

Wing, L. 1941. Size of bird flocks in winter. Auk, 58:188-194.

DEPARTMENT OF BIOLOGICAL SCIENCES, NORTHERN ARIZONA UNIVERSITY, FLAG- STAFF, ARIZONA 86001. ADDRESS OF THIRD AUTHOR: 420 W. OAK AVE.,

FLAGSTAFF, ARIZONA 86001, 17 MAY 1971.

ON THE EVOLUTION OF SOCIALITY, WITH PARTICULAR REFERENCE TO TIARIS OLIVACEA

Ronald Pulliam, Barrie Gilbert, Peter Klopfer, Dennis McDonald, Linda McDonald, and George Millikan

The behavior of the Yellow-faced Grassquit [Tiaris olivacea) apparently ranges from social and nonaggressive on the Central American mainland to territorial and very aggressive on the island of Jamaica (Pulliam, 1970). Why these differences?

This paper reports observations on the population size, habitat distribution, and social behavior of the Yellow-faced Grassquit on the island of Cayman Brae, West Indies, and speculations on factors influencing social behavior. Cayman Brae is a very small island (20 square miles) and this population of grassquits is extremely isolated from other populations, the nearest being found on Grand Cayman (80 miles southwest) and on Jamaica (190 miles southeast) . The third of the Cayman Islands, Little Cayman Island, is about ten miles west of Cayman Brae, but grassquits are very rare or absent there perhaps because of a lack of suitable habitat.

The observations reported here are based primarily on a two-week field study beginning 27 November, 1969. Additional observations must be made at other times of year for confirmation of our findings. However, the social organization of the species has been noted by one author (Pulliam) to be stable throughout the year in Jamaica and Skutch (1954) indicates that Costa Rican grassquits can be found in flocks during all seasons of the year.

THE EVOLUTION OF SELFISH BEHAVIOR

Hamilton (1964) has demonstrated that kinship selection can limit the expression of behavior which decreases the fitness of a neighbor more than it increases the fitness of the actor (i.e., selfish behavior). Kinship selection encompasses the notion that an individual’s overall fitness includes not only the effects of his genotype on his own ability to leave descendants hut also the effects of his genotype on the fitness of relatives who carry some proportion of genes identical by descent to his own. Although Hamilton’s model is formally correct, it is applicable only if the selfish behavior of a population is determined by the gene frequencies at one locus. We contend that aggression or selfish behavior is not coded at a single chromosomal locus (see Klopfer. 1969) but that the degree of aggression in an individual must he thouglit of as resulting from the interaction of the animal’s environment with the epislatic effect of a large number of genes at very many loci. Thus, in an almost

77

78

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

homozygous population, selfish behavior might be selected against because it decreases one’s own inclusive fitness.

In this paper we argue that some forms of aggressive and territorial behav- ior constitute “selfish behavior” and, thus, their occurrence in natural popula- tions must be restricted to relatively heterozygous populations.

An aggressive territory holder can decrease the fitness of a nonaggressive bird by excluding it from optimal habitat. It is less obvious that the decrease in fitness of the nonaggressive bird is greater than the increase in fitness of the aggressor. However, the territorial bird does lose some of the advantages of social behavior (whatever they are) and must spend considerable time defending bis territory, time which might otherwise be applied towards main- tenance and reproduction. The amount of time which the average aggressive individual spends defending his territory must necessarily increase as the proportion of the bird population which is territorial increases. Hence, the question: why are some grassquits territorial?

Suppose territorial individuals do have a lower reproductive capacity than social individuals would have in the absence of the former. This would result in a territorial population maintaining lower numbers than a social population even though the territorial individuals were superior in competition with the social individuals! If, for a given bird species, the social populations were shown to maintain a significantly higher population density than the territorial populations, we would have evidence that territoriality is a selfish behavior for that species.

Pulliam (1970) censused, during the breeding season, 11 similar habitats that appeared suitable for Yellow-faced Grassquits in both Jamaica and Costa Rica. Each habitat was visited twice. In Costa Rica, on a total of 25.9 acres, an average of 20.5 grassquits were seen. In Jamaica, on a total of 18.0 acres, an average of only 6.9 grassquits were seen. In both Costa Rica and Jamaica there were grassquits in four of the eleven habitats visited. The number of grassquits per acre in those sites containing some grassquits was 2.9 in Costa Rica, as compared to 0.7 in Jamaica. The increase in the density of the Costa Rican grassquits is especially surprising since there were many more individ- uals and species sharing sites with grassquits in Costa Rica than there were in Jamaica. Thus, it appears that the social grassquits of Costa Rica are able to maintain a population density two to three times as great as that of the terri- torial Jamaican grassquits. This accords with our supposition.

Very little is known about the degree of heterozygosity in natural popula- tions of birds and we are not yet able to predict the degree of heterozygosity that might permit selfish traits to evolve. However, we do know that both isolation and population size exert considerable influence on the degree of genetic diversity of natural populations. In very small populations, random

I’ulliam et al.

SOCIALITY IN GRASSQUIT

79

drift can lead to fixation or loss of genetic variability. This decay of genetic variation is counter-balanced by the forces of mutation and immigration. Soule (1971) presents arguments and evidence that for lizards large popula- tion size and migration between adjacent populations is necessary for the maintenance of genetic diversity. Soule showed that lizards from small, iso- lated island populations showed less variation in electrophoretically detectable isozymes than lizards from large island populations. The decrease in enzyme variation was correlated with a deerease in morphological variance. This result indicates that isolation and small population size result in a decrease in genetic diversity and could, therefore, limit the expression of selfish be- havior traits.

Tiaris olivacea is an abundant inhabitant of the subtropical plateau region of Costa Rica (Slud, 1969). However, the grassquit is a bird of secondary growth habitats, never found in the dense forest, and is therefore restricted in distribution to areas near human habitation and agriculture. The human population of Costa Rica is largely limited to areas in close proximity to roads or rail lines. Thus, habitat suitable for grassquits is discontinuously distri- buted along the few roads and railroads in eastern Costa Rica. In May of 1969 Pulliam searched for grassquits along the road from San Jose to Tur- rialba and along the railroad between San Jose and La Lola Farms, which is about 30 miles west of Port Limon on the Gulf of Mexico. This journey made an east-west transect across almost the entire range of Tiaris in Costa Rica. Grassquits were first noted along the roadsides about 5 miles east of Cartego. From Cartego to Turrialba, grassquits were frequently recorded in suitable habitats but these habitats were distributed in patches. Along the railroad, grassquits were noted from Turrialba to La Lola Farms, where they were common. Suitable habitat along the railroad was distributed in discrete patches and often interrupted by many miles of forest habitat. In addition to the patchwork character of suitable habitat, the presence of a dozen or more sympatric seed-eating finches may further limit the distribution of grassquits. This combination of a patchwork habitat and many competitor species would tend to result in Tiaris being found in isolated groups of small size in Costa Rica. We expect their social behavior to be related to a high degree of genetic I homozygosity maintained because of the patchiness of their distribution.

Tiaris olivacea is found in all parts of Jamaica with the possible exception I of the very dry Southeast. Throughout the range of grassquits in Jamaica there are numerous roads and, therefore, much more habitat suitable for Tiaris than in Costa Rica. This suitable habitat is virtually continuous over the entire island except in the high mountains which are s})arsely settled by I humans. Also, in Jamaica there is only one other sj)ecies of finch which feeds exclusively on grass seeds. The two factors combine to })roduce a continuous

80

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

and therefore very large grassquit population. We expect such a population to be genetically more diverse than the discretely distributed Costa Rican population and, thus, to permit the occurrence of selfish behavior. In fact, the Jamaican birds, in contrast to those of Costa Rica, are territorial, as noted above.

These arguments are conjectural and were largely developed ex post facto, after our studies in Jamaica and Costa Rica. If, however, the argument is correct we would expect to find that any isolated, small populations of grassquits would exhibit social rather than selfish behavior, and be more similar in their social structure to the Costa Rican population than to the Jamaican population. With this idea in mind, we attempted to ascertain the population size and social structure of the isolated grassquit population on Cayman Brae Island.

ESTIMATE OF GRASSQUIT POPULATION SIZE ON CAYMAN BRAG

Data for population size estimates were collected by locating and then, only once, walking slowly through suitable habitats and recording all birds heard or seen. “Suitable habitat” was defined as those areas where trees and shrubs covered less than 80 per cent of the ground and where there was some grass growing. This definition of suitable habitat was consistent with our observa- tions in Costa Rica and Jamaica that grassquits were found only in grassland and old-field habitats and the observations of Skutch (1954) in Costa Rica and Wetmore (1927) in Puerto Rico that the diet of grassquits consisted almost entirely of grass seeds. However, on Cayman Brae we often found male grassquits singing from the upper branches of trees and shrubs near the edges of fields. Figure 1 illustrates that the grassquits in trees were always very close to a grassy field. The data for Figure 1 were collected by pacing along a path which ran all the way across the island from North to South. The loca- tion of the bird is plotted as the location at which the bird was estimated to be at right angles to the path. Thus, those birds which appear, in the figure, to be in the fields may actually have been singing from trees and shrubs on the east or west sides of the fields. At any rate, the data presented in Figure 1 are consistent with our belief that the grassquits are found only in or near field habitats. Since the maintenance of such habitats on Cayman Brae de- pends entirely on their being accessible to people (due to the rapidity of suc- cessional growth), we felt confident that most such habitats could be found by traversing all roads and paths on the island.

One of the assumptions of the model (presented in the Appendix) used to estimate population size is that the probability of a call in any interval of time is constant throughout the time of observation. It is well known, however, that many birds show a pronounced decrease in singing in the middle of the

Pulliam et al.

SOCIALITY IN GRASSQUIT

81

cCCG ^£!brCCp

^hcCbcO

ciArrO-

575

150

t ^ 1725

Crr^r^Q--Q~(}-i(\

CxX)cC>rCf^£lit^

^ 2300

G

2875

<iss^?cm) cctpteoQ ,, n

^ ^ T t 3450

G G G G

Oi

4025

^ 4200 G

Fig. 1. Observations on the location of birds along a transect across Cayman Brae Island. The symbol G indicates the position of grassquits along the transect and the numbers on tbe right indicate the distance from the start of the transect, (irassy fields are indicated by the clear areas and forests and garden are indicated diagrammaticalK .

(lay. Thus, the probability of recording a bird in the middle of the da\ might be lower than, say, in the early morning. Table 1 shows the number of songs per thirty-minute interval for seven individual grassquits sampled at difftMcnt times of the day. It appears from this sample of singing arti\it\ that th(*re

82

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

Table 1

The Number of Bird Songs in Thirty-minute Intervals for Seven Individual Grassquits Watched at Different Times of Day.

The times on the left indicate the beginning of each thirty minute interval.

1 2 3 4 5 6 7 Average

7:00	60		20		40
7:30	38		18		28
8:00	48		20		34
8:30	118	3	61		60
9:00	23	17	80		40
9:30	0		82		41
10:00			85	36	60
10:30				17	17
11:00				1	1
11:30				46	46
12:00				3	3
12:30				40	40
13:00				57	57

13:30 40 40

14:00 76 76

14:30	28		28
15:00	18		18
15:30	21		21
16:00	4	43	23
16:30	9	56	32
17:00		51	51

may be a slight decrease in singing rate in the middle of the day. Since the sample size is so small, particularly for the mid-day period, this is not certain. Even if there is a decrease in singing rate at mid-day we believe it does not seriously effect our results, since the decrease appears to be small and less than 10 per cent of our censuses were taken in the mid-day period ( between 10:00 and 14:00).

Lor three of the seven birds for which data are given in Table 1, we were able to record the occurrence of each song to the nearest second. Erom these data we could assess the reliability of our census technique (see the Appendix) . Figure 2 indicates that the probability of recording a bird does not differ significantly from one time of day to the next.

For the total census we recorded 190 male and 24 female grassquits. Of the 190 males, 161 were heard singing and 29 were only seen. If we assume the sex ratio to be equal and that there must have been some suitable habitat which we did not locate, then we must conclude that there were at least 400 grassquits on the island. However, this is undoubtedly an underestimate since many

Pulliam et al.

SOCIALITY IN GRASSQUIT

83

TIME (MINUTES)

Fig. 2. The probability of recording a bird as singing as a function of the length of time that an observer is within hearing range of the bird. See Appendix for estimation procedures.

birds must not have been recorded even though we located the fields in which they resided. From the estimates of the probability of recording a bird in Figure 2 we can get some idea as to how accurate our census was. A singing male grassquit can be heard from 75 to 100 feet away. If we assume that our walking speed through the fields was between one and two feet per second, it follows that an observer was within hearing range of each bird for from one to three minutes if the field where the bird resided was actually located.

; Taking a very liberal estimate of the population size we assume that each

I bird was in hearing range for only one minute and thus, from the lowest esti- mated probability of recording a bird when it is within hear-range for one minute (From Bird No. 3, Fig. 2), we estimate that only 55 per cent of the male birds were recorded by being heard. Thus, a liberal estimate of popula- tion size is about 300 male birds (or approximately 600 birds, total). Ibis estimate may still be too low since there may have been first-year male birds which were not singing. Assuming there may he as many as one non-singing male for each singing male we can boost the total estimated population size to about 1,200. Finally, there were the birds in the fields that we did not locate and assuming that we may have not found as much as 20-25 per cent of

84

THE WILSON BULLETIN

March 1972 Vol. 84, No. 1

the suitable habitat, we reach a figure of 1,500 birds. It should be realized that in arriving at this estimate of population size we took the extremes of all estimation parameters so as to give an absolute upper limit. At the other end of the scale we could assume that we observed all of the male grassquits on the island. Taking the two extremes we can state fairly confidently that there were between 400 and 1,500 Yellow-faced Grassquits on the island at the time of our census.

SOCIAL BEHAVIOR OF THE YELLOW-FACED GRASSQUIT

In Jamaica, the Yellow-faced Grassquit is strictly territorial. Nine terri- tories in optimal habitat measured in June-July, 1968, near Treasure Beach, Jamaica, averaged only 0.25 acres each and aggressive encounters between males on adjacent territories were frequent. Although Jamaican grassquits never occur in flocks, individuals of both sexes are known to aggregate occasionally at artificial feeding stations and when this happens males seem to spend more time fighting than feeding.

Skutch (1954) describes the Yellow-faced Grassquit in Costa Rica as lacking “that pugnacious jealousy so prominent and characteristic in many members of the finch family” and as “a most pacific bird. I have never noted any fighting or discord among them.” However, males do defend a small area in the immediate vicinity of the nest from which other males of the same species are expelled. Skutch describes this defense as follows: “all the terri- torial male does is fly mildly in the direction of the intruder who retreats without necessity of conflict.” Grassquits which are not nesting are normally found in large feeding flocks which often contain thirty to forty individuals, with both sexes represented. Pulliam (1970) noticed no signs of aggression within flocks but did note occasional conflicts between grassquits and other seed-eating finch species during a three-week field study during the breeding season in 1969 near Turrialba, Gosta Rica.

The contrast between the highly social behavior of Costa Rican grassquits and the strictly territorial behavior of the Jamaican grassquits is typical of the differences in social behavior of a number of passerine bird species from Costa Rica and Jamaica. Pulliam (1970) compared tbe social behavior of all resident bird species of the families Fringillidae, Thraupidae, and Icteridae for which data could be found for Jamaica and Costa Rica. He found that 18 of the 26 Costa Rican species showed some form of social tolerance (family groups or flocking) compared to only two of the 11 Jamaican species. [The definition of “no social tolerance” is that at all times of the year individuals are either alone or in the company of a single adult of the opposite sex and/or juvenile birds up until a short time after fledging.] This is consistent with the supposition that continuously distributed species are more likely to be

NUMBER OF BIRDS IN EACH CLASSIFICATION

Pulliam et al.

Fig. 3

SOCIALITY IN GRASSQUIT 85

GROUP SIZE

. The sizes of groups of grassquits observed for three different populations.

86

THE WILSON BULLETIN

Marcli 1972 Vol. 84, No. 1

genetically polymorphic, and thus aggressive, since Jamaican birds seem to he more continuous in their distributions