pi)'

NESTING BIOLOGY. SOCIAL PATTERNS AND DISPLAYS OF THE MANDARIN DUCK, A_IX GALERICULATA

Richard L. Bruggers

A Dissertation

Submitted to the Graduate School of Bowling Green State University in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

December 1974

__ ' U J 591913

W A'W

.'X55’ ABSTRACT

A study of pinioned, free-ranging Mandarin ducks (Aix galericulata) was conducted from 1971-1974 at a 25-acre estate. The purposes 'were to 1) document breeding biology and behaviors, nesting phenology, and time budgets; 2) describe displays associated with copulatory behavior, pair-formation and maintenance, and social encounters; and 3) determine the female's role in male social display and pair formation.

The intensive observations (in excess of 400 h) included several full-day and all-night periods. Display patterns were recorded (partially with movies) arid analyzed. The female's role in social display was examined through a series of male and female introductions into yearling and adult male "display parties."

Mandarins formed strong seasonal pair bonds, which re-formed in successive years if both individuals lived. Clutches averaged 9.5 eggs and were begun by yearling females earlier and with less fertility (78%) than adult females (90%). Incubation averaged 28-30 days.

Duckling development was rapid and sexual dimorphism evident. 9 Adults and yearlings of both sexes could be separated on the basis of primary feather length; females, on secondary feather pigmentation.

Mandarin daily activity patterns consisted of repetitious feeding, preening, and loafing, but the duration and patterns of each activity varied with the social periods. Diurnal activity rhythms were obvious only during the summer and fall.

Females incubated 80% of the daytime and did not leave the nest at night. Males threatened conspecifics and waited for their females from loafing spots throughout incubation and attended their mates during their two daily inattentive periods. Mandarins (unless incubating) generally spent the night on the water, swim-feeding until 2200, then floating concealed near shore but alert. Some fed until 0200.

Copulations were initiated by the female. Adult sequences were more stereotyped than yearling sequences; the latter's were longer in duration (as was mounting) and contained more displays.

Social display occurred only in the female's presence, much of the display being directed to her. Females, through their movements, postures, and vocalizations, both initiated and intensified male social display. Females indicated their preference for a particular male by Inciting him against other birds. The season was important in both the method and length of time necessary for pairs to form. Fall or winter pairing took several weeks and involved social display, directed courtship, and changing partners. Late spring pairing(in the absence of social display) resulted from direct attempts by either sex, and bonds formed in a week. m

ACKNOWLEDGEMENTS

I am indebted to Mr. and Mrs. J.J. Schedel for allowing me access to their estate and waterfowl collection and to the Schedel Foundation for funding many of the project activities. The opportunity to observe and participate in the propagation of so many exotic waterfowl species in such a beautiful setting was unique.

Appreciation is likewise extended to Slim Bouldin, the estate caretaker, and to those individuals indicated in Appendix I. In particular, I am grateful to Dr. Charles Dane for his hospitality during my visit to Northern Prairie Wildlife Research Center, and his suggestions regarding ageing techniques; Mr. Ray Greenwood for providing nasal saddles; Mr. Charles Webster for providing some ducks as well as allowing me to visit his estate on several occasions and observe his waterfowl collection; and Dr. Janet Kear, for her helpful comments during the study. I also am grateful to Dr. Frank

McKinney for letting me view his waterfowl behavior films and for discussions which helped in the formulation of study procedures and interpretations.

Thanks are extended to the members of my committee, Drs. T.R.

Fisher, Stephen Vessey, William Baxter, Reginald Noble, and John

Paul Scott for their evaluation of the manuscript, and my adviser,

Dr. William B. Jackson, for his guidance and encouragement during the study and throughout my graduate career.

Finally, I must thank you, Jackie, for your help in collecting some of the data, and for preparing and illustrating parts of the manuscript. You were most patient and understanding (usually)I a iv

TABLE OF CONTENTS Page INTRODUCTION AND PURPOSE ...... 1

METHODS...... 15

RESULTS AND DISCUSSION...... 29

Reproductive and developmental biology...... 29

Copulation frequency and success...... 29

Nesting and incubation...... 35

Egg morphology...... 54

Duckling development...... 56

Ageing yearlings and adults...... 66

Time budgets...... 73

Spacing behavior...... 73

Population time budgets...... 77

Individual time budgets...... 83

Displays...... 92

Descriptions...... 92

Social displays...... 112

Copulation sequences ...... 124

Display motivations...... 145

SUMMARY...... 148

LITERATURE CITED ...... 166

APPENDIX ...... 172 V

LIST OF TABLES

Page Table 1 Number of individuals (pinioned) of exotic waterfowl species present on Schedel Estate, Elmore, Ohio during March, 1970-1974 ...... 17

Table II Predators trapped at Schedel Estate 1971- 1973 and predation between 20 April-30 May, 1972...... 18

Table III Chronology of male and female introductions into adult (experienced) and yearling (inexperienced) 3-male courting parties, winter, 1973...... 27

Table IV Proportion of Mandarin copulations during the pre-layino, laying, and renesting periods in 1971 and 1973 ...... 31

Table V Outcome of 47 copulatory sequences of yearling and adult Mandarins recorded during 1971 and 1973 ...... 34

Table VI Frequency of nest box inspection by 8 Mandarin pairs during pre-laying and laying periods in 1971 and 1973...... 36

Table VII Nest boxes used by Mandarins 1971-1974...... 38

Table VIII Comparison of the number of eggs and fertility in first and second clutches of Mandarin females (allowed to complete the clutch uninterrupted), whose initial clutch was destroyed by a predator or removed for incubation...... 43

Table IX Nesting biology of seven Mandarin females observed during 1973 and 1974 ...... 46

Table X Temperatures in middle of nests during incubation and at intervals during rest periods...... 49

Table XI Mandarin duck egg morphology under various conditions...... 55 vi Page

Table XII Age at which comfort movements first appeared in five Mandarin ducklings, hatched 17 June 1973 and kept in captivity...... 64

Table XIII Average number of comfort movements/ duckling/ten minutes at four, seven, eight, and nine days of age - June 1973 ...... 65

Table XIV Primary feather lengths of male and female Mandarin ducks...... 68

Table XV No. and % yearling and adult Mandarin females with "complete" or "incomplete" white markings on the distal ends of the secondary feathers...... 71

Table XVI Mandarin feeding patterns, based on the proportion of birds engaged in each type of feeding at 1/2 h intervals during 1973...... 84

Table XVII Percent time spent in various activities during two morning and three evening inattentive periods by a Mandarin pair during the incubation period, spring 1973 ...... 89

Table XVIII Description and use of female vocalizations...... 10 3

Table XIX Descriptions and use of male vocalizations . 104

Table XX Frequency of displays and agonistic encounters (pecks, chases, fights, threats), given by groups of 3 yearlings and 3 adult males alone and with the 30- min introduction of another bird...... 113

Table XXI Activities of a "known" female, and 3 adult and 3 yearling males in each of 6 x 30 minute introductions of this female during February and March 1973 ...... 117

Table XXII Frequency of responses by R M3 (

Table XXIV Occurrence of movements in the pre- and post-copulation behavioral sequences of yearling and adult Mandarins ...... , . . 127

Table XXV Average duration (and range) of each phase in complete reproductive sequences of adult and yearling Mandarins ...... 132

Table XXVI Percent occurrence of each pre­ copul atory display (movement) of three yearling and three adult males in successive sequences, spring 1973...... 138

Table XXVII Frequency of display of penned and free-ranging paired-males in social encounters...... 146

» vili

LIST OF FIGURES Page

Figure 1 Distribution of Mandarin duck in Asia (Savage, 1952) ...... 5

Figure 2 Distribution of Mandarin in England (Savage, 1952)...... 7

Figure 3 Map of study area showing 1.5 and 2.5 acre ponds, locations of out-door breeding pens, nest boxes, and over-wintering building...... 16

Figure 4 Frequency of copulatory behavior (n=26 sequences) relative to time of day between 3 April - 31 May 1973, the date of the last observed sequence ..... 30

Figure 5 Egg-laying patterns of Mandarins each spring 1971-1974 ...... 40

Figure 6 Time of day at which females were observed laying eggs, 1972-1974 ...... 42

Figure 7 Mandarin female inattentive periods, recorded during 1972 and 1973 . , . , , , , 47

Figure 8 Growth of Mandarin ducklings observed at one week intervals from hatching for ten weeks and again on 28 October 1973 ...... 58

Figure 9 Feather development and maturations! changes in 5 male and 8 female Mandarin ducklings observed at one week, intervals from time of hatching, 1972 61

Figure 10 Frequency distribution of the 7th and 9th. primary feather lengths of yearling and adult female Mandarin ducks, measured during fall 1972 and 1973...... 69

Figure 11 Comparison of distal markings of the 5th-l1th secondaries of 6 Mandarin duck females, as both, yearlings and adults, while in alternate plumage ... 72

Figure 12 Locations of free-ranging Mandarins during the first two weeks following release to the lakes on 2 April 1973, during laying period, and during fall (pre-alternate molt and alternate plumage periods) ...... 75 IX

Page Figure 13 Proportion of free-ranging flock engaged in various activities at 1/2 h intervals during full-day observations 1973 ...... 78

Figure 14 Proportion of birds engaged in each activity for the entire day in each of 7 periods, spring - fall 1973 ...... 80

Figure 15 Nighttime activity patterns of 5 - 8 Mandarins observed during the post­ incubation (4 June) and pre-basic molt (1 July) periods...... 82

Figure 16 Chronologic daily activity pattern of a free-ranging Mandarin pair during full- day observations in 1973...... 86

Figure 17 Activity patterns of Mandarin pairs during spring 1973......

Figure 18 Display sequences ...... , , , ,105

Figure 19 Display sequences ...... 106

Figure 20 Display sequences...... 107

Figure 21 Display sequences...... 108

Figure 22 Display sequences ...... 109

Figure 23 Vocalizations of adult female Mandarin ducks. . , . .110

Figure 24 Vocalizations associated with the displays of adult male Mandarin ducks...... Ill

Figure 25 Displays performed by yearling and adults male Mandarins during 6 introductions of a "known" female to each 3-male group...... 121

Figure 26 Order of movements in 13 Mandarin pre­ copulation sequences...... 128

Figure 27 Relative frequency of movements in adults and yearling Mandarin male pre-copulation sequences . . .131

Figure 28 Distribution of pre-copulatory movements from 13 sequences...... 134

Figure 29 Order of movements in 14 Mandarin post­ copulation sequences...... 141 X

LIST OF PLATES

Page Plate 1 Out-door breeding pens indicating types of nest boxes...... 153

Plate 2 Marking methods...... 153

Plate 3 Preferred nesting area...... 153

Plate 4 Adult female incubating in ground nest box...... 154

Plate 5 Male in Full-sail posture while female incubates in box 15...... 154

Plate 6 Mandarin clutch in nest box ...... 154

Plate 7 Feather development of Mandarin ducklings between 2-6 weeks of age...... 155

Plate 8 Secondary feather markings of yearling and adult Mandarin females...... 156

Plate 9 Male Burping during social encounter following loafing...... 157

Plate 10 Female leading male in positions typci.al of most spring movements...... 157

Plate 11 3-male social display party; female in foreground...... 158

Plate 12 Male Burping; female in "Ack" vocalization posture...... 158

Plate 13 Male Body-shake (note neck and head not outstretched as in Fig. 18); female in water...... 158

Plate 14 Male pecking female; female later paired to male at left...... 159

Plate 15 Female threatening B FI6; male in Full-sail posture . 159

Plate 16 Male display Preening, probably to female at left . . 160

Plate 17 Male display Preening; note clearity of nasal saddles ...... 160 XI Page

Plate 18 Three males in Alert posture, vocalizing (Pfrruib); females vocalizing (Ack)...... 161

Plate 19 Two (A-C and D-F) Drink-mock-preen sequences directed towards a female...... 162

Plate 20 Display situations ...... 163

Plate 21 Display situations...... 164

Plate 22 Display sequences...... 165 1

INTRODUCTION AND PURPOSE

Although information is available concerning the breeding biology and reproductive behaviors and displays of many waterfowl species,

"there are many gaps in the knowledge of this topic, which require prolonged observations on marked individuals of known age and breeding status" (McKinney, 1967). This lack of specific information is particularly acute for the Mandarin duck (Aix galericulata), despite its widespread aesthetic appeal and occurrence in most private breeding collections and zoos. Aside from general behavior studies and discussions of displays, little information has been published on the species.

The breeding biology and behaviors of a waterfowl species, its reproductive success or failure, and variations in daily activity patterns are better understood when the visual and auditory signals associated with pair formation arid maintenance are understood. There­ fore, the purposes of this study of the Mandarin were 1) to determine its breeding biology and behaviors, nesting phenology, and time budgets; 2) to describe in detail the displays and vocalizations associated with copulation patterns, social encounters and pair formation and pair maintenance; and 3) to determine the female's role in stimulating male social display and subsequent pair formation.

Although the study of waterfowl under wholly natural conditions is desirable, this method has limitations. Quantification of behavior is difficult to obtain without extended observations of marked birds of known age and sex. The waterfowl collection at the J,J. Schedel 2

estate, Elmore, Ohio readily permits a study of free-ranging, known-

age, pinioned birds in a semi-natural environment.

Mandarin duck

Taxonomy and description

The Mandarin and Wood duck (Aix sponsa) are the only two members of the genus Aix in the Cairinini tribe, subfamily Anatinae of the

Anatidae. Through the years there has been much discussion as to the relationship of these species and the validity of their placement in

Cairinini (Delacour, 1964). They are similar in: 1) markings of the downy young; 2) physical characteristics of adults; 3) ecology and tree hole habit; and 4) certain behavior patterns (Delacour and Mayer,

1945; Lorenz, 1941; Johnsgard, 1965).

The tribe, having attained world wide distribution, includes the

Spur-winged goose (Plectropterus sp.), Muscovy duck (Cairina muschata) native to Central and South America, Comb duck (Sarkldiornis sp.) of

South America, Africa, Madagascar, and India, Hartlaubs duck

(Pterorietta sp.), Pygmy goose (Nettapus sp.) of Africa, South American

Ringed teal (Callometta leucophrys), Brazilian teal (Amazonetta sp.), and the Australian Wood duck (Chenoretta .jubata) (Johnsgard, 1968).

Mandarin ducks in their native Asia are not sympatric with any close tribe members (Johnsgard, 1965), but artificial juxtapositions can occur at breeding estates. Heinroth (1911), although not able to produce a Mandarin-Wood duck hybrid under sympatric conditions, noted that an unpaired Mandarin male may associate with a paired Wood duck female. Although scattered reports of Mandarin hybrids have occurred, 3

Prestwich (1960) considered their evidence inconclusive, claiming chromosomal differences prevent hybridization. Karyotypes of one such recent "hybrid" indicated the bird was probably an abnormal

Mandarin rather than a hybrid (Shoffner, pers. comm.).

Mandarins throughout their range generally begin displaying in autumn, often in courting parties. Pair formation occurs during the winter with breeding in April and May. The average clutch size is

8-12 eggs, with incubation usually taking 28-30 days. Males remain in the area of the incubating female and feed with her during incubation breaks. Just prior to hatching of the clutch, usually in mid-June, the males congregate and molt into basic (eclipse) plumage. The pre-alternate molt into alternate (nuptial) plumage occurs in August._

The male Mandarin is perhaps the most beautiful of all the waterfowl. Heinroth (1911) acclaimed it as having the highest plumage development in all the Anatidae. Savage (1952) has provided descriptions of the plumages of male and female Mandarin and Wood ducks. The most striking features of the male Mandarin's breeding plumage are its highly modified, orange, secondary "sail" feathers and its facial features: white crest above eye and rust colored mane, all of which are erected during most displays. Mandarin and Wood duck males are most similar when in basic (eclipse) plumage; then only the

Wood duck's longer orange bill, red eye, and white chin distinguish it from the Mandarin. The female Mandarin, except for her smaller size and shorter forehead, is almost indistinguishable from the female Wood duck throughout the year and from the male Mandarin during his basic 4

plumage.

History and Asian distribution

Mandarins, native to China and Japan, have been symbols of love

and fidelity for many centuries. The Buddhists in 200 B.C. established

the Mandarin as a model of mutual affection, fidelity, kindness and

consideration; and these qualities have been depicted in odes and songs,

paintings, porcelain, embroidery, and literature.

While the Mandarin was facing extinction in China as a result of

breeding habitat destruction and use for food, plumage, and captivity,

its current status is uncertain. The species in Japan is thought,

however, to be increasing from post-war lows (Savage, 1952).

Although their main breeding grounds are in Manchuria (40-45 N

latitude), nesting Mandarins have been reported as far north as 55 N

latitude along river valleys. Manchurian winter temperatures approximate -36 C (-32 F) minimum, necessitating a southern migration; most birds go into Kiangsi Province (30-35 N latitude) near Shanghai.

In Japan, Hokkaido Island is the Mandarin's main breeding area, while it winters throughout the main island (Fig 1).

Introduction into Europe and propagation status

The Mandarin was introduced into Europe near the end of the 18th century and first added to the London Zoo in 1830, where it bred in

1834 (Savage, 1952). Initial introduction into the United States apparently occurred early in the 20th century (Allen, pers. comm.).

Although there are exceptions, the Mandarin nearly always has been difficult to breed in captivity (Savage, 1952). Propagation in English 5

Fig. 1: Distribution of Mandarin duck in Asia (Savage, 1952)

DUCK IN ASIA THIS MAP SHOWS THE PRESENT DISTRIBUTION OF MANDARIN DUCK (a/X CALEfUCULATA ui>™. . JeJ" e}1 ) IN THE FAR-EAST SOFAR AS CAN BE DETERMINED

Hi" IA° 6

estates, such as Walcott and Leckford (Stevens, Jones, pers. comm.),

and parks like Windsor Great Park (Savill, pers. comm.) has varied, with some breeders consistently and others only sporadically success­ ful. Perhaps the greatest breeding success has occurred at Sevrin

Wildfowl Trust, Slimbridge, England, where according to the Annual

Reports approximately 350 birds were raised between 1963-1970.

Consequently, a sizable feral population has been established in the vicinity. In other areas of England Mandarins escaped from private collections (Olney, pers. comm.), and as a result, the Mandarin has been included on Britain's Official Bird List (Kear, pers. comm.).

It is most common in Surrey and Berkshire (Fig. 2). The general ecology and behavior of British Mandarins approximate what is known from Asia, except that the milder temperatures of Britain do not necessitate migrations (Savage, 1952).

While many of the birds at these large English estates are free- flying, most of those kept by U.S. breeders are pinioned, i.e., have the terminal digits of one wing removed. Most breeders have enjoyed some success with either full-winged or pinioned birds.

Previous research

Considering the Mandarin's popularity and beauty, surprisingly little specific information is available, aside from discussions of the more ritualized and conspicuous displays and general behavior.

Most studies concerned correct taxonomic placement of the species, an important pioneering work being published by Heinroth (1911). With the exception of a general, naturalistic study of an unmarked, feral 7

Fig. 2: Distribution of Mandarin in England (Savage, 1952)

GENERAL DISTRIBUTION} S AND MANDARIN DUCK ^(SPORADIC RECORDS IN GREAT BRITAIN VKPEVY

bfleeoiHG colonies . . . FftlQVZHT WINTER VI5FTOR ISOVATCD RECORDS . . WATERFOWL COLLECTIONS WHICH HACK BEEN, OR ARE, SCURCES of free• FLYING MANDARIN DUCK. -.- THE DUKE OF BEDFORD . I«oo CrHA’AHE»?,...). SIR RICHARD GRAHAM, PRE - I?l4...... X U* GREY OF FALLODON , ...... 3 ALFRED CZRA Etc,.; ¡928 ONWARC3...... A NOEL STP.YENS £ai, I9J5 ONWARDS...... 5 LT &•£■ MEADE-WALDO, c 19«? -Fiji...... 6 ' •*» V 8 population in England by Savage(1952), no other studies of the daily activity patterns of birds of known age, as related to breeding biology and reproductive behaviors, have been published.

Although the behavioral patterns and displays of the Mandarin have been described in general by several workers, most notably

Heinroth (1911), Lorenz (1941), Johnsgard (1965), and Delacour (1964), there has been no quantification of displays and vocalizations. Dilger and Johnsgard (1959) considered the relationship of plumage patterns, mate selection, and species recognition in both Mandarins and Wood ducks. A few notes from estates, zoos, and gardens are available.

Sporadic breeding success has occurred at the Schedel Estate.

Between 1962-1966, the first years the species was bred on the estate, the annual Mandarin egg production varied between 13 and 61, with approximately 50% fertility and between 31-62% hatching success.

However, from 1967-1970, total egg production was 15, and no ducklings survived (Appendix II).

Development of hypotheses

A detailed study focused on the reproductive biology and social pattern of the Mandarin duck was initiated during 1971 with an intent of establishing a successfully breeding population at the estate.

Components of the program included:

a. A complete study of the breeding biology of known age, marked

individuals, including daily and seasonal time budget

determination and ageing methods.

b. Determination of the specific association between displays 9

and daily activities as related to pair bond formation,

maintenance, and duration, and copulation patterns in adults

and juveniles.

c. Delineation of the relationship of the female to male social

display and the role of social display to pair formation.

Nesting biology and time budgets

The chronology of waterfowl activity during the breeding season, of behavior variations related to time of day, or of interactions between breeding pairs is not known precisely (McKinney, 1967).

Usually species have their own optimum time budget adapted to local conditions (Verner, 1965).

Mandarins, as well as most waterfowl (Delacour, 1964), generally are most active during the early morning or evening (Savage, 1952).

However, recent research on feeding behavior in dabbling ducks

(Swanson and Sargent, 1972), including Wood ducks (Ball, pers. comm.), has shown intense nighttime feeding, particularly during the breeding season. Similar night activity may be characteristic of Mandarins, perhaps related to insect (chironomid and ephemeropteran) hatches or concentrations of certain plankters (cladocerans) at the surface during the night.

Although flight-associated activities or displays obviously are hampered by pinioning, most of a bird's reproductive behaviors are unaffected; and insight into the behaviors of wild birds can be gained by using pinioned birds (McKinney, 1967). The use of pinioned, free-ranging birds in the present study population was felt justified 10

because of the apparently low occurrence of aerial displays in the

Mandarin repertoire and the desire for successful reproduction with

the estate birds. McKinney, using captive, but full-winged shovelers

in flight pens (four pair per 607-763 m? enclosure), observed that daily' and seasonal activity patterns of the shovelers paralleled those of wild birds, but they had reduced nesting success. It is proposed that the daily activities and nesting biology of pinioned

Mandarins are expected to be similar to those reported for wild birds.

Reproductive displays

Displays are those peculiarly standardized and often exaggerated performances, including all vocalizations, movements, and postures, which have become specialized and modified as social signals or releasers (Moynihan, 1955). Displays usually are fixed action patterns whose function or signal value depends on the reaction of the display recipient, which may be behavioral, physiological, or both (Ficken and Ficken, 1966). While these displays may be innate, the responses may be learned (Marler, 1961). Many of the stereotyped movements serve as consumatory acts (Thorpe, 1956).

Displays are produced by dual or multiple motivational factors.

Agonistic displays are motivated by attack and escape tendencies

(Tinbergen, 1959), and courtship displays are motivated by sexual as well as attack and escape tendencies (Hinde and Tinbergen, 1965;

McKinney, 1961). Components of courtship displays are associated with either of these tendencies, and variations in their strengths and in the animal's responsiveness to the releasing stimili are thought to n

account for species differences in the displays (Hinde and Tinbergen,

1965; McKinney, 1961).

Such species variability has been demonstrated by Stoke's study

(1962) of agonistic behavior in the Blue tit in which given behavior

elements rarely led to a highly predictable subsequent behavior, since

the same element often represented tendencies to attack, stay, or

escape. Further, a seasonal relationship between elements of the

display and the subsequent behavior was observed. Variations in display components (particularly during the appetitive phase) among

birds may be influenced by environmental and seasonal variations as well as specific releasing situations.

Heinroth (1911) and Lorenz (1941) noted that certain displays

in ducks were more prevalent in certain situations, and other

investigators have since attempted to quantify these differences.

Weidmann (1956), using Mallards, demonstrated that some displays characterize high intensity rather than low intensity courtship, while Johnsgard (1960) has correlated the display type with the number of males in the courtship party. McKinney (1961) has shown that the displays of male Eiders, before and after copulation, during agonistic encounters, and in courting parties, were usually situation specific; but when the same display was characteristic of more than one of these situations its frequency of occurrence varied.

Although displays of Mandarins have been presented in general terms by a number of authors, only certain displays, usually the most ritualized (Drink-mock-preen, Grunt-whistle or Display-shake, Double­ grunt-whistle or Double-display-shake, and Preen-behind-the-wing) are 12

discussed in any detail. An analysis of Mandarin displays, including

detailed descriptions, situations used, and frequency of occurrence,

should indicate whether these displays are situation specific and

related to the experience (reproductive age) of the males involved.

Comparisons of displays will be made with Wood ducks and members of

the genus Anas.

The female's role in social display and pair formation

According to the classical view, social display (i.e., social courtship) promotes pair formation; males appear to "put themselves

in a favorable light" by displaying before females (Heinroth, 1910).

However, work by Lorenz (1941) with surface feeding ducks indicated much species variation in the interrelationship between social display and pair formation. Lorenz considered social display as highly developed in Mallards and particularly Mandarins, citing its occurrence in all-male groups of these species. He further regarded social display in Mandarins as an affair of males, who are in no way concerned with the females' presence; the role of mate selection is left exclusively to the females.

Social display consists of complex swimming maneuvers in which male members of the group maintain or adjust positions relative to each other and the female and employ characteristic displays and calls.

A social display bout, consisting of several bursts of synchronized displays by one or several males, may last from a few minutes to an hour (Weidmann and Darley, 19/la).

Weidmann (1956) has suggested that the epigamic displays of 13

Mallards are of three types: social displays; pair-formation and maintenance displays, called "directed courtship" by von de Wall

(1965; cited in Weidmann and Darley, 1971a); and copulation-accompanying displays. Social display and directed courtship can be interrelated with their arousing the female's interest, as indicated by an increase in female Inciting and Following. As pair bonds form, the male will give preening and drinking displays, i.e., directed courtship.

The female's role in social display of ducks is still uncertain.

Von de Wall (loc. cit.) suggested her presence in social display.is of minor importance and that minimal interaction occurs between sexes.

He considered any female-directed displays to be elicited by other males. Likewise, Lebret (1961) doubted the significance of social display in pair formation, suggesting that it occurred in lieu of overt hostility, thereby maintaining flock unity. However, film analysis studies by Weidmann and Darley (1971a) and McKinney (1956a) on Mallards and Green-wing Teal, respectively, have shown that introduction of a female to penned males leads to social display and pair formation. Johnsgard (1960) also had previously suggested that the female played a major role in male display synchronization.

However, Weidmann and Darley (1971a) observed social display in the absence of females and noted that synchronization of these male- influenced displays was as good or better than in the presence of a female. Ultimately, the role of the female in social display and the role of social display in pair formation may have to be assessed solely on a species basis. 14

Lorenz (1941) has stated that Mandarin males in social display pay little attention to the female; however, her role in these parties has not been evaluated. If she were not essential to social display, similar results would be expected with the introduction of either a male or a female to a male courting party. It is postulated that although Mandarin social display may occur in the absence of a female, her presence should greatly enhance it.

Differences in intensity of courting party activity (i.e., the frequency and use of certain displays) also should occur relative to the age of courting party members (yearling or adult) and the activeness or passiveness of the female.

In summary, it is suggested that the reproduction biology and daily activity pattern should be similar to and supplement those reported for feral Mandarins. In addition, it is hypothesized that

1) displays associated with reproduction and social encounters are situation specific, and 2) the female stimulates or enhances male social display and resultant pair formation. 15

METHODS

Description of Schedel Estate

The 25-acre Schedel estate, along a wooded portion of the

Portage river (Ottawa County, Ohio), is unique in its blending of

more than 350 varieties of native and exotic trees and shrubs in flood

plain and upland habitats. Two interconnected lakes (1.5 and 2.5

acres) have controlled water levels and are fed by wells (Fig. 3).

These lakes presently are the major activity centers for approximately

15 species of exotic waterfowl (Table I), although in the past 13 years nearly 80 species have been raised or maintained at the estate.

Pinioned Ross geese, wild Mallards, and wild Wood ducks also nest on

the grounds during the spring, while other native waterfowl including

Canada geese frequent the ponds during migrations.

The lakes are highly eutrophic and contain bass (Micropterus spp.), sunfish (Eupomotis sp.), and muskel1unge (Esox masquinongy).

Plume grass (Miscanthus sinensis), cattails (Typha sp.), pines (Pinus spp.), bamboo (Bambusa sp.) and willows (Salix sp.) provide cover along the edges. Both ground and elevated lakeside nest boxes are available to the ducks. To minimize disturbance of nesting birds, only necessary maintenance work is done during the breeding season.

Study flock and propagation procedures

Prior to spring 1970 Mandarins were free-ranging and used lakeside breeding boxes. A high incidence of nest disturbance and adult mortality from predators (Table II) and constant agonistic interference 16 Fig. 3. Map of study area showing 1.5 and 2.5 acre ponds, locations of out-door-breeding pens, nest boxes, and over-wintering building.

50m 4

OVER-WINTERING BI BUILDING BUILDING O ELEVATED POLE BOX CZ3 ELONGATE GROUND BOX nn elongate elevated box O BURIED BOX ------FENCE

CONIFEROUS LOWLAND

DECIDUOUS HILL DUMP OUSE 17

Table 1. Number of individuals (pinioned) of exotic waterfowl species present on Schedel Estate, Elmore, Ohio during March , 1970-1974.

1970 1971 1972 1973 1974 Coscoroba Swans 7 8 12 8 4 (Coscoroba coscoroba) European Shelducks (Tadorna tadorna) 11 4 6 6 8 Australian Shelducks (Tadorna tadornoides) 63 14 21 9 8 Chiloe Wigeons (Anas sibilatrix) 8 5 4 1 1 Radjah Shelducks (Tadorna radjah) 2 2 3 1 2 Comb Ducks (Sarkidiornis melanotos) 0 2' Û 2 2 Ross Geese (Anser rossi) 1 4 3 3 3 Red-breasted Geese (Branta ruficollis) 2 2 4 0 0 Maned Geese (Chenonetta jubata) 2 0 1 2 2 Asheyhead Geese (Chloephaga poiiocephala) 2 1 2 1 1 Red-crested Pochards (Netta rufina) 43 0 0 0 0 Tufted Ducks (Aythya fuligula) 1 0 0 0 0 Wood Ducks (Aix sponsa) 8 0 0 0 0 Mandarins (Aix gal eri culata) 5 11 17 23 17

TOTAL 155 53 76 56 48

Other species usually present: 1-2 pair wild Wood Ducks 4-5 pair wild Mallards crested seriemas, black-necked screamer, silver and golden pheasants, peacocks (all terrestrial species) ' 18

Table II. Predators trapped at Schedel Estate 1971-1973 and predation between 20 April-30 May, 1972.

Great-horned Owls Raccoons Opossum** Cats Red Fox 71 72 73 72 73 72 73 72 73 72

January 1

February 1 2

March 3 1

Apri 1 2 2

May 3 2 4 1 3

June 1 2

July 1

August 2

September

October 1 3

November 1 1

December 1 1

Totals 14 10 1 9 1 15 2 1 3

20 April- 30 May* 4 4 3 1

*no. prey individuals killed (20 April-30 May) 9 Pheasants 1 Peacock 2 Mallards 2 Mandarins

records not kept on a monthly basis 19

from other waterfowl (Australian Shelducks in particular) necessitated

a shift to pen breeding. A small pen was constructed during February

1970 under the direction of Mr. Mel Block, Curator of Birds, Toledo

Zoo. Following correspondence with waterfowl breeders, primarily

in England, some changes from previous Mandarin propagation procedures were made for the 1971-1974 breeding seasons. (See Appendices III

-V for additional details.)

During January 1971, two adjacent pens, each with approximately

70 m^ of land and 55 m^ of water were constructed, replacing the smaller pens used in spring 1970. The number of Mandarin pairs per enclosure was reduced from four in 1970 to two thereafter. Several other pairs were maintained free on the lake (Appendices VI-IX).

The number of boxes per enclosure was increased from one per potential breeding pair to at least two more than the number of pairs of ducks. All boxes except one in each pen were the elongate variety used in Wood duck breeding programs (30 x 15 x 46 cm with a 10 cm di a. hole) and elevated (usually on poles) between .3 and 1.3 m with ascent ramps

(Plate 1). Only elongate boxes previously had been provided for penned ducks. Nesting material, previously sawdust, was changed to a combi­ nation of sawdust, peat, and timothy grass. Most eggs, (except for penned and most 1973 free-ranging females which incubated their own eggs) were either collected as laid or removed during early incubation and incubated under bantams.

All ducklings were raised in the indoor pens for six weeks without their parents, and allowed to swim for short periods until their feathers attained water repellency, before being put in an out-door pen or released to the lake for the summer. These procedures are 20 similar to those used by other waterfowl breeders (Webster, Macy,

Guest, Bell, pers. comm.), although the dimensions of the pens vary.

For the first 8-10 weeks ducklings were fed Purina Chick

Startina mixed with small amounts of cracked corn, after which their diet was the same as adult birds: supplemental scratch grain during summer, fall and spring, a scratch grain-corn conditioning diet during November and December while in the over-wintering pens, and scratch grain egg-laying mash from January through March.

Observational Procedures

Observation of both free-ranging and penned, color-marked (Plate

II) birds were made each spring 1971-1973, at two or three-day inter­ vals with 7 x 35 binoculars, either from a hill overlooking both lakes or from partially concealed positions around the lake perimeter

(Appendix X). Distances from the ducks usually did not exceed

150 meters.

During 1971 the general locornoter, social, and comfort activities of most penned pairs were recorded continuously during one to two hour periods; positions and status of all birds in the pens were determined at 15-minute intervals. From 16 June 1971 (the release date of the birds from the breeding pen to the lakes) to 21 July, observations were made less frequently; only group activities were recorded, since wing markers had fallen off, and individual identities were lost with summer molts.

Additional birds and the need for observations of social behavior necessitated having increased number of free-ranging but pinioned birds on the lake in 1972 and 1973. Observations on individuals or pairs were made at 30-second intervals for 10-minute periods, with an 21

activity recorded if it occurred during that 30 seconds (Appendix XI).

(Repetitive behaviors in the same 30 seconds also were sometimes

recorded). All birds, both penned and free-ranging, usually were

observed each session; the activities of those not under direct

quantitative observation also were noted.

Observations of pair formation displays were made of both adults

and yearlings in the outdoor pens during September 1972 following

completion of pre-alternate molt into the breeding plumage. However,

most behavioral observations were made in 1973 but were supplemented

by earlier data (i.e., feeding patterns and some reproductive biology)

that would not have been unduly influenced by the experimental

manipulations.

Time budget determination

In 1973, time budgets were determined during six, 15-hour dawn-

to-dusk observations made during the pre-laying, laying, incubation,

post-incubation, and the summer and fall molt periods. The occurrence

of behaviors and displays associated with social encounters and

reproduction, and time spent in these and other activities (preening,

feeding, loafing) of one free-ranging pair were continuously recorded.

The activities and locations of most other free-ranging birds also were noted at 30 and 60 minute intervals, respectively (Appendices XII-

XIV). Human disturbance usually resulted in movement by the birds to

a new location and resumption of previous activities.

Two all-night observations (using a Starlite Night Vision Scope) were made during the 1974 post-incubation and summer molt periods. 22

Activities and locations of all observable birds were recorded during

half-hour observations from 2100-0530.

Reproduction and developmental biology

Laying patterns.- To determine laying dates and intervals, nest

boxes were checked at least every other day during 1973, either at

the end of observations or while the birds were in another area of

the estate. Most eggs of both first and second clutches were numbered

with a rapidograph, measured with calipers, and weighed (nearest 0.1 g)

on a torsion balance at day 1 (fresh weight), day 15, and at pipping.

All eggs during the study were candled at mid-incubation to

determine fertility. Those not hatching were opened and observed for

embryo death. Following hatching, the shell membrane was removed, and

shell thickness measured with an Ames Thickness Measure modified for

curved surfaces. An average shell thickness was calculated from five

readings taken around the center of the egg (methods of NPWRC,

Jamestown, North Dakota).

Incubation.- The onset and duration of incubation and the time and

length of daily incubation periods and incubation breaks by each female were observed each year. Nighttime incubation patterns were determined

in 1973 using a Rustrak automatic recorder and kymograph paper. The

recorder, attached to a nest box ramp, and started after the female had

returned to the nest for the night, was triggered by a switch activated

by the duck's weight as she entered or exited from a box. Time of day was calibrated to chart speed. Kymograph paper, smoked black, was

attached to two ramps such that any movement across the paper would 23

leave tracks. This technique to record direction of activity has

been used in rodent movement studies (Justice, 1961).

Nest temperature.- Temperatures in the middle of the nest, taken

when females were on and off the nest, were recorded on four (1972)

and five (1973) females using a portable YSI manual temperature

recorder. A 30 m thermister lead was taped to an egg and the temperature

read from a concealed location. Nest temperatures also were read

during incubation inattentive periods at 30 min intervals until

the female returned. The presence or absence of down over the eggs

during the female's departure was noted. Ambient and box temperatures

of five bantam hens and a wild female Mallard were recorded for comparison.

Duckling development.- Morphological development (standard criteria: body weight, tarsus length, culmen length and width, and middle toe length, and feather developmental patterns of 13 ducklings

(5 8V and 8??) were monitored at hatching and at one week intervals from that time throughout the 1973 summer. The ducklings, having hatched within two weeks of each other, were banded and numbered with expandable plastic wire snaps. They were put in the outdoor pens at seven weeks. Measurements (to the nearest 0.1 mm) were made on the live birds with vernier calipers.

Five of the ducklings were held in a .6 x 1.7 m pen and allowed to swim each day and observed daily for 10 days. The occurrence of comfort movements was quantified (average no./bird/lO min.), and the maturation and development of motor patterns also was noted during ten minute periods at days 4, 7, 8 and 9. 24

Morphologic measurements were made of all adult birds on the

following occasions: prior to placement in the over-wintering pens

(October 1972); at the time of nasal marking (January 1973); prior

to their release to the lakes (29 March 1973). Summer molt patterns

in adult males also were observed.

Ageing adults and yearlings

During late fall 1972 and 1973, after completion of the pre­

alternate molt while the birds were in alternate (or breeding) plumage, eye color, primary feather length, and secondary feather markings were evaluated as indicators of age. Using methods of Dane (1968), yentral, flattened surfaces of the 7th, 9th, and 10th primaries of yearling and adult males and females were measured from the sharply defined edge of digit II and the partially fused digit III to the feather tip. Outer primaries were measured since they have been found to show greater variation in length than do proximal ones (Engles 1938; Baumel 1953; and Dane 1968).

White coloration on the margins of the 5th-9th secondary feathers also was examined. Mandarins have 16 secondaries. The first nine are grey with white outer margins. Numbers 10 and 11 are steel blue but usually lack the white tip. Secondaries 12 (a highly modified sail feather) through 16 sometimes are called tertials (Miller, 1925).

Feather length and coloration data were collected from individuals belonging to Mr. Schedel, Mr. Carl Strutz (North Dakota), and

Mr. Charles Webster (New York); six to eight females were examined as yearlings and again as adults. 25

Female's role in social display and pair formation

During winter 1972-1973 the role of the female in the social display and pair formation of experienced (adult) and inexperienced

(yearling) males was studied by introducing "strange" and "known" males and females to three-male courting parties and observing ensuing displays and behaviors. The technique of description and analysis from movies, initiated by Dane, Walcott, and Drury (1959) and further developed by McKinney (1951) and Weidmann and Darley (1971 a, b), was used in analyzing the positions of the birds and components of their displays.

All first year, inexperienced (six months in December 1972) and experienced birds were held in two of the large indoor, over-wintering pens from 30 September 1972 to 3 January 1973. Two of the small pens were prepared for courting party analysis by visually separating them with burlap. They were marked with paint at .8 m distances around the perimeter of the water to facilitate use in spatial analysis. Three yearling and three experienced males (one of each group either having paired or begun to pair as of 1 January) were put into each of the two smaller pens on 3 January 1973. Although the number of courting party males varies in the wild (Savage, 1952; Kear, pers. comm.), three-male parties were used in this study to allow comparison with data obtained from studies of other species.

Following completion of baseline observations of the 3-male groups, introductions between 25 January and 8 March 1973 of "strange" and previously "known" birds were made to determine the role of the female 26

in stimulating male activity and the effect of prior experience on

pair bond formation:

1. a "strange" female obtained from Mr. Charles Webster in

December;

2. a "strange" male also obtained at the same time from

Mr. Webster;

3. a "known" female, one to which a male in each party had

previously been paired; and

4. a stuffed female, floating in a semi-crouch posture.

To offset the effect of increasing photoperiod and temperatures with spring, most introductions were replicated in late February or early March (Table III). The same birds, with the exception of the previously paired females, were introduced into each courting party.

During the 30-rnin observations, begun at the time of introduction, descriptions of all activities, displays, behaviors, and relative positions were recorded on tape and later transcribed. The introduced birds were placed back in the larger pens following each observation period. Observations and movies (using a 16 mm Bolex) were made through a small hole in the pen doors, a maximum distance from the birds of 6.1-7.7 m. The sound of a water pump, used to keep open water in the adjacent lake and in the pens, was loud enough to mask the camera noise. Vocalizations were recorded at a speed of 19 cm/sec using a

Sony Tape recorder TC800B, and later analyzed on a Sona-graph, Power

Unit Model R (662B), Kay Electric Co., Pine Brook, N.J.

Following cessation of the experiment on 8 March, all males were returned to the large pens, from which all paired birds except one pair 27

Table III. Chronology of male and female introductions into adult (experienced) and yearling (inexperienced) 3-male courting parties, winter, 1973.

date adults yearlings 25 Jan. alone* alone 26 Jan. alone alone 1 Feb. alone alone 1 Feb. strange male strange female 3 Feb. strange female strange male 9 Feb. strange female stuffed female 22 Feb. known female known female 22 Feb. stuffed female' no Introduction 23 Feb. known female known female 23 Feb. no introduction strange female 27 Feb. known female strange male 1 Mar. known female known female 2 Mar. known female known female 2 Mar. alone alone 3 Mar. strange male known female 5 Mar. stuffed female stuffed female 8 Mar. strange female no introduction 9 Mar. known female known female

SUMMARY freq. .freq. alone 4 4 strange female 3 2 strange male 2 2 stuffed female 2 2 known female 6 6

*alone = baseline observation of males only 28 had been removed. Pair formation and pair maintenance were observed in the remaining birds, which in effect, represented male courting parties in the presence of two to four females. Differences in frequency of displays and preening between paired and unpaired birds were determined, as were the roles of paired birds in social display.

On 29 March 1973 all birds were collected, weighed, measured, and either put into the outdoor breeding pens or turned free to the lakes.

Statistical considerations

The following statistical analyses were performed according to

Sokal and Rohlf (1969), and are treated individually in appropriate sections.

1) Single classification analysis of variance (unequal sample

size) to determine differences a) in egg and egg shell

morphology with respect to the age (yearling vs. adult) of the

female; b) between clutch size of yearling vs. adult females;

and c) between primary feathers of yearlings vs. adults.

2) T-test of paired-comparisons to determine differences between

the primary feather lengths of females measured as yearlings and

again as adults.

3) Student's t-test for differences in mounting duration of

yearling and adult copulation sequences.

4) Mann-Whitney U and Wilcoxon two-sample tests for differences in

a) the pre- and post-copulation duration of yearling and adult

copulatory sequences; b) between the number of displays performed

by each age male in these sequences; and c) between the number of

courting party displays performed by yearling and adult males. 29

RESULTS AND DISCUSSION

Reproductive and developmental biology

A knowledge of the reproductive biology is important in under­

standing a species behavioral patterns. Changes in breeding status

(i.e., pre-laying, laying, incubation periods etc.) affect both daily

time budgets and social encounter responses. The following data are

based on intensive partial and whole-day (dawn-dusk) observations of

pinioned free-ranging birds, supplemented by data from penned birds.

Copulation frequency and success

Copulations were not observed in the over-wintering pens but did occur 17 and 3 days after release to the lakes in 1971 and 1973, re­ spectively. (Intensive post-release observations were not made in 1972 and 1974.) Copulations occurred in quiet, undisturbed areas of the lakes, most frequently during the morning (Fig. 4). Morning copulation was like that observed for semi-wild Canada geese by Col lias and Jahn

(1959), but unlike the random pattern of Shovelers observed by McKinney

(1967). Copulations were more frequent during the pre-laying than lay­ ing periods in both years: 65% in 1971 and 52% in 1973 (Table IV).

McKinney (1967) noted a similar pattern for Shovelers, relating it to the increased time spent by females at the nest during laying and hence their unavailability. This is an even more probable explanation for Mandarins, since laying also was most prevalent during the morning.

The greater percentage of copulations observed during the pre­ laying period in 1971 probably was a function of that year's extremely 30

Fig. 4. Frequency of copulatory behavior (n=26 sequences) relative to time of day between 3 April - 31 May 1973, the date of the last observed sequence. 2 0 0 0 j I

Table IV. Proportion of Mandarin copulations during the pre-laying, laying, and renesting periods in 1971 and 1973.

pre-laying laying renesti ng* year total no. % obs. h no. % obs. h no. % obs. h

1971 20 13 65 37** 7 35 15

1973 27 14 52 32 11 41 45 2 7 20

No renesting occurred in 1971

In 1971 a total of 53 h of observation were made during the pre-laying period, but only 37 h encompassed the period of copulatory activity (17 Apr-19 May); in 1973 copulations were observed within 3 days of release to the lakes. 32

late laying period. All the birds kept from previous years also had a history of poor and late breeding. Additionally, the six yearling birds purchased from propagators in North Carolina and Nebraska were X not received until just prior to their release and presumably did not acclimatize rapidly. This failure to nest following relocation is apparently quite common and perhaps related to the age of the birds

(Anon., 1965).

The pattern of copulations during 1973, when 41% of the observed copulations occurred during the laying period, was more representative of feral Mandarins in a natural nesting situation (Savage, 1952).

Multiple copulations were observed for seven females, with 48% and 33% of the copulations of these females occurring during their pre-laying and laying periods, respectively. Four females (19% of the copulations) copulated following removal or destruction of their first clutches.

Since two of these females later renested, their copulations were included in the calculations for both the pre- and post laying periods.

Frequent copulations by a pair during the pre-laying and laying periods probably are not necessary to insure fertilization, since

Elder and Weller (1954) reported sperm viability in Mallards of 12-15 days. Frequent copulations are perhaps more important in synchronizing reproductive physiology and strengthening pair bonds. There was no apparent relationship between frequency of copulation and age of males. Copulations decreased with the end of spring; only three sequences were observed after 21 May in 1973, and one stopped from male disinterest. 33

Sixty-two percent of the observed copulation sequences led to

apparent successful copulation (Table V). Better success (75-77%) was

achieved when both sexes were the same rather than different ages

(55%). This lack of success in mixed pairs probably resulted both

from the time of year in which these unsuccessful copulations were

observed (late May) and the age of the male involved. Although he

successfully copulated twice, all nine unsuccessful copulations

during 1971 involved the same adult male with either a yearling or

adult female. He was at least six years old and showed little

interest in pre-copulatory behavior, usually swimming away after the

female went Prone. He died during the summer 1973.

Five copulation sequences were interrupted by other birds (four occurring among penned birds) and nine stopped because of low

intensity male behavior. Low intensity female behavior was observed only on 24 October 1973. This female had successfully copulated at

0830, but after assuming the Copulation position at 1000, she slowly swam away.

As with Shovelers (McKinney, 1967), promiscuousness was rare and rape infrequently observed, characteristics undoubtedly indicative of strong pair bonds. This situation is unlike that known for waterfowl like Pintails (McKinney, 1973), Muscovys and Comb ducks (Weller,

1964), and others which have few displays or vocalizations and weak or no pair bonds (Johnsgard, 1965).

Copulation by one pair sometimes stimulated other birds in the area, particularly under penned conditions. In April, observed intervals between copulations of eight pairs of penned birds were 34

Table V. Outcome of 47 copulatory sequences of yearling and adult Mandarins recorded during 1971 and 1973.

yearling mi xed* adult totals birds bi rds bi rds

copulation occurred 12 12 7 31

copulation did not occur 4 10 2 16 reasons for failure

a) pair interrupted or disturbed 2 2 1 5 by other birds

b) low intensity female precopulatory 1 1 behavior /“ c) low intensity male precopulatory 2 0 1 9 behavior

d) mount without successful'--* 1 1 copulation

totals 16 22 9 rape*** 2 ]

*mixed - in all cases, an adult male and yearling female

Successful copulation - a sequence consisting of pre-copulation movements, mounting and apparent cloacal contact.

unsuccessful copulation - a sequence consisting of some precopulatory movements which may or may not lead to mounting. If mounting occurred there was apparent lack of cloacal contact.

***rape - see page 142. 35

15-45 minutes, less than the one hour interval often observed for

these birds. Successive copulatory behavior by different pairs of

free-ranging birds was observed only on 31 May 1973, when the pairs

had begun to re-associate, facilitating contagious behavior. This

infrequency was a result of the larger area occupied by free-ranging

birds and agonistic encounters tending to space them and reduce

visual contact.

Nesting and incubation

Nest site selection.- Nest box inspection occurred primarily

during the morning and late afternoon (Table VI). The male in Full-

sail posture accompanied the female. (See pages 92-102 for posture

description.) The choice of boxes appeared to be by the female, as

she entered the boxes 72% of the time. Although the male occasionally

approached a box and stood on the ramp prior to the female's entering,

he usually (74%) swam Alertly at the base of the ramp. Jones (1972)

also reported seeing a male enter a box during inspection, but this

does not appear to be typical behavior. Inspection time (n=ll) inside

the box by the female averaged two minutes (range 1-7 min.; 10 <3 min.,

5

During the pre-laying periods females inspected different nest

boxes (average 3.3, range 1-5 different boxes). They frequently re­

inspected previously visited boxes. During 1971 a female was observed

inspecting three different boxes, two of those five times each, and one three times, while another female in 1973 was observed inspecting five different boxes, three one day and four the next. One female was 36

Table VI. Frequency of nest box inspection by 8 Mandarin pairs during pre-laying and laying periods in 1971 & 1973. Dawn obser­ vations were begun as soon as light permitted recognition of birds and sometimes continued until dark, when birds no longer could be seen.

Time Behavior i' i1 0 • interval* obser. h enter box look in wait at base dawn-1100 40 female 15 1 1 male 3 0 8

1500-dark 70 fema1e 21 10 2 male 4 3 21

Totals 146 female 36(72%) 11(22%) 3(6%) male 7(18%) 3(8%) 29(74%)

bno nest box inspection was observed between 1100-1500 (36 observation h) 37

observed inspecting the same box (in which she later laid) on three

of five occasions; another was observed inspecting two boxes, each

twice. Five of seven females laid in the box they were last observed

inspecting. Savage (1952) states that Mandarins visit many potential

nest sites, choosing cavities in old oak trees as deep as eight feet and between 10-20 feet above ground. As with copulations, frequent nest box inspection by a pair, and its associated displays and behaviors, probably are important in pair bond maintenance and synchronization of physiological condition.

Tree cavity nesting explains the preference of captive pinioned

Mandarins for elevated boxes (66%) in general, and for pole boxes in particular (60%) (Table VII). Greater use of unelevated boxes might have been expected because of their easier accessability to the pinioned birds, but this was not the case. (Females sometimes had difficulty ascending nest box ramps, with one falling off three times prior to entering.) There was no indication that the pole boxes were chosen by those females laying first and buried or ground boxes used by females laying later. Nest box choice appeared more related to environmental location, as 60% of the boxes used by free-ranging females (boxes no. 8-17, n=7) were along the water on the wooded, hilly, and most undisturbed east side of the large lake (Fig. 3;

Plate 3). During 1973 and 1974, the years with the greatest nesting pressure, 75% of the nests started by free-ranging females in these seven boxes; with five of the 12 (42%) begun in pole box 17. This box had less nest material and a deeper cavity than the adjacent and similar box 15. 38

Table VII. Nest boxes used by Mandarins 1971-1974. The number of boxes of each type was constant each spring.

No. clutches started each type type no. boxes ij sed year us ed destroyed of box* avail. numbei'’S 71 72 73 74 no. % no. %

buried box 11 9 2 6 17 5 83 16 1 10 1 4 1 14 1

tree or pole box 21 58 1 5

elevated 10 17 2 3 44 p** 1 1 13 P 1 1 43 P 1 1 34 P 1 1 1 15 1 1 46 P 3 1 20 1 elongate box 9 25 2 22

ground 7 39 P 1 49 1 40 1 1 37 P 2

elevated 3 31 1 1 0 0 29 1

% total - all ele'vated 24 66 1 4 - all groijnd or buried 12 34 7 58

■¡fe Elongate, ground or elevated box: 76 x 30 x 38cm; 15cm porch; 10 cm di a opening; Buried box: 53 x 53 x 38cm; 38cm covered entrance (tile); 18cm dia. opening Tree or pole box : 30 x 15 x 46cm; 10cm dia. opening; .31-1.3m elevation ** P indi cates location in out-door breeding pens 39

The preference for nesting in this area and the subsequent

greater concentration of nests, a phenomenon noted by Weller (1964)

for waterfowl generally, resulted in increased use of buried boxes

and occasional failure to complete clutches. However, buried boxes,

used by Australian Shelducks during the winter, were unsuitable as

Mandarin nest boxes, for 83% of the nests in them were destroyed by

predators or spring drainage.

Laying patterns.- Only five of 32 females (some present more than

one year) failed to lay eggs. Of these five, one just failed to lay,

two died early in the breeding season, and two were full-winged and

escaped. Even until their escape, the full-winged birds were

extremely active and nervous and showed no signs of nesting. This

was surprising since Funk (pers. comm.) had reported good nesting

success among penned, full-winged birds. Their nesting failure

may have been related to their recent relocation from Nebraska and

failure to acclimate readily.

The average time from release to the first egg each year was

15 days (Fig. 5). Forty percent of the 18 females laid their first

egg between 17-21 days following release. Extensive cloud cover and

consequent low temperatures probably were responsible for the 10-day

delay in the onset of nesting between 1973 and 1974. Temperatures in

spring 1974 were more variable, with the mean daily minimum temperatures during April 1974, less (0 C) than in 1973 (4 C). Additionally, only

one four-day period occurred in 1973 with temperatures less than 0 C,

(10-13 April), while three such periods (5-6, 8-10, 19-20 April) occurred during 1974. Maximum daily temperatures gradually increased Fig. 5. Egg-laying patterns of Mandarins each spring 1971-1974. Release date to outdoor areas indicated by arrows. Number of females layinq eqqs: 1971 - 3; 1972 - 5; 1973 - 10; 1974 - 9. S G G E

F O

R E B M U N

-pa o

MARCH MAY JUNE 41

during April 1973, but abruptly rose following a cold spell in 1974.

In both years birds first laid on days with temperatures between

21-25 C.

Although the time of laying of individual females varied daily,

as with most waterfowl (Weller, 1964), it usually occurred during the morning (Fig. 6) with the male in attendance outside the nest. Most

(68%) of the 26 eggs actually observed being laid were laid prior to

1100. Twelve additional eggs were found when the boxes were inspected before noon. These data seem indicative of the true situation, since nest boxes usually were checked upon my arrival and departure, and the number of morning and afternoon observations during this period were similar. The actual duration of laying, based on the time spent in the box by three females averaged.23 minutes, (range 15-31 minutes).

This time is considerably less than the two-six hours spent on the nests daily during laying period by Shovelers (McKinney, 1967) and

Gadwalls (Oring, 1969).

During the four years, first clutches were begun as early as

3 April in 1972 and as late as 30 May in 1971, with 75% of the eggs laid between 15 April and 15 May. The initial laying dates were like those of birds in North Dakota, New York, and England (Strutz, Webster,

Olney, pers. comm.), but slightly later than those known for Mandarins in Japan and Russia (Savage, 1952).

Seventeen of 21 Mandarin females laid in their first year (when

10-11 months of age), but with reduced fertility (78%) compared to adults (90%) (Table VIII). The average clutch size for 16 nests in which the eggs were not removed as laid, was 9.50 eggs, with yearling 42

Fig. 6. Time of day at which females were observed laying eggs, 1972-1974.

1972 - 5 eggs (H); 1973 - 20 eggs (•); 1974 - 4 eggs (0). y a d f o e m i t 43

Table VIII. Comparison of the number of eggs and fertility in first and second clutches of Mandarin females (allowed to complete the clutch uninterrupted), whose initial clutch was destroyed by a predator or removed for incubation.

lst clutch 2nd clutch lst clutch yearl date eggs date eggs adul t date eggs year ?? start no(fert) start no(fert) ?? start no(fert)

1972 F 0A23+ 12 Ap 10(10)

1973 F 0A27 16 Ap 8(5) 4 Ap 5(5) F 0A4-0 13 Ap 11(9) F 0A44 10 Ap 6(2) 13 Ma 9(9) F 0A12+ 25 Ap 8(7) F 0A31 21 Ap 12(9) 28 Ma 7(5) F 0A19+ 19 Ap 9(9) F 0A35 23 Ap 9(5) F 0A36* 30 Ap 10(?) F 0A32+ 27 Ap 8(6) F 0A30+ 8 Ma 9(8)

1974 F 0A46* 22 Ap 11(0) 14 Ma 7(7) F 0A34+ 26 Ma 7(7) F 0A41* 22 Ap 9(0) 11 Ma 5(5) F 0A49+ 23 Ap 12(11) F 0A48+ 30 Ap 13(12)

Avg. clutch 9.75 6.60 8.75

% fertility 78 94 91

+no attempt to induce renesting by these females by removing their clutches

*F 0A46 and F 0A41 laid prior to pair bond formation and fertilization so eggs are included in average clutch size but not % fertility. Eggs of F 0A36 were destroyed by a predator prior to candling for fertility. 44

clutch size (9.75 eggs) not significantly different from adult clutch

size (8.75 eggs). While the mean clutch size was similar in both

age groups, the variation was greater in yearlings.

Yearling females laid earlier in the spring than adults. The first

clutches of 10/11 yearlings during 1973 and 1974 were started prior to

30 April while only two of the six adult females' clutches during

these years were started this early. There is little specific

information on laying periods relative to age in any waterfowl, and

none for wild Mandarins. However, yearlings laying before adults is

contrary to that known for some of the Anatini, such as Pintails

(Sowls, 1955) which breed in their first year. In this species (and

others) adult females return and presumably breed before yearlings.

Anatini yearlings, in which plumage development and pairing occurs

later than adults, probably compete less successfully in pairing

situations and subsequently nest later (McKinney, 1965 b). Yearling

Mandarins do not face these disadvantages, as the pre-alternate molt

is complete in all birds during the fall and pairing usually stabilized

by March.

Most Anatids, including Wood ducks (Breckenridge, 1956) usually

lay one egg per day, with skips being relatively rare. This was the

case for only five of 14 clutches (average clutch size was 9.27 eggs).

The average completion time for these 14 clutches was 12.14 days, with an average interval of 2.15 days between the first two eggs.

Failure to lay an egg a day may have resulted from human disturbance,

the concentration of birds, or, as Erskine (1972) suggested for

Buffi eheads, the failure of the females to accumulate necessary food 45

reserves during early incubation to permit laying an egg every 24

hours.

Incubation patterns.- Mandarin females began lining the nest with

down during laying of the last couple of eggs and usually completed

it within the first two days of incubation (Table IX). They never

brought additional materials to the nest (Plate 4). Incubation

appeared to begin during the afternoon of the final day of laying.

The incubation period was between 28-30 days. Most females took

three inattentive periods during the initial days of incubation

(Fig. 7). For example, F 0A32 (31 May) began inattentive periods

ranging between 75-85 minutes, at 0615, 1210 and 1945. The time on

the nest during the first day for her and F 0A35 was shorter (4 h)

in the morning than afternoon (6-8 h).

Later, inattentive periods occurred only prior to dawn and at

dusk. Evening inattentive periods (avg. 97 min, range 55-180 min;

n=9) were longer than pre-dawn periods (avg. 80 min, range 65-120 min;

n=4). The smaller sample size during the morning occurred because

females usually had left the nest prior to sunrise or upon my arrival

about 0530, and complete periods were not seen. However, based on the

63 min average duration observed for the 11 such inattentive periods,

it may be surmised that the lengths of both periods are similar. The

daily incubation period for five females on six occasions averaged

10.25 h (range 9.5-12 h). Females usually were returning to or on the nest by sunset and based on event recorder and kymograph paper

information, did not leave during the night.

The onset of morning and evening inattentive periods, seemed 46

Table IX. Nesting biology of seven Mandarin females observed during 1973 and 1974'.

elapsed eggs at which pipping time to clutch down inc. dur. of start hatchi ng* year ?? no. size pulled start inc.(days) (hours) (hours)

1973 F 0A27 8 8 8 28 0700 24-30

F 0A32 8 6 6 29-30 1900 22-24

F 0A36 10 10 10

F 0A31 10 9 10 29 0700 24-30

F 0A19 9 9 9 ■ 33**

1974 F 0A48 13 11 12 30-31 0900 30

F 0A49 12 11 11-12 29-30 1300 35-40

* hatching time - duration between pipping (when first egg cracked) and when duckling is out of shell

female killed in first week of incubation - eggs incubated under bantam Fig. 7. Mandarin female inattentive periods recorded during 1972 and 1973. For mornings when onset of inattentative period was not observed, solid line indicates actual periods of observations.

date female 18 Jun F 0A31 16 Jun F 0A31 15 Jun F 0A44 15 Jun F 0A31 13 Jun F 0A44 11 Jun F 0A44 11 Jun F 0A31 4 Jun F 0A31 3 Jun F 0A31 1 Jun.F 0A44 31 May F 0A31 31 May F 0A32 31 May F 0A44 24 May F 0A19 24. May F 0A23 21 May F 0A30 21 May F 0A44 21 May F 0A32 15 May F 0A23 13 May F 0A36 11 May F 0A19 11 May F 0A35 10 May F 0A23 9 May F 0A23 4 May F 0A31 3 May F 0A31 2 May F 0A23 1 May F CA23 24 Apr F 0A18 2 2 0 0 48

closely tied to sunrise and sunset. Evening breaks began later as

spring progressed. The five inattentive periods (3, 4, 15, 16,

18 June) of F 0A35 were very similar in duration (60-75 min) and

generally began later each evening. This relationship could not be

as rigidly defined for mornings, since the actual beginning of the

inattentive periods were seldom observed. The birds were both off and back on the nest earlier in late May. Additionally, it was my

impression that females incubating in west-facing boxes began inattentive periods later in the evening than females in east-facing boxes. This was demonstrated on 11 and 15 June 1973 during which F 0A44, (box 34) left earlier than F 0A31 (box 15) (see Fig. 3 for box placements).

Incubation temperatures in nests of six females varied between

34.4 and 36.6 C (Table X). Comparatively, the incubation temperature of the female Mallard was 36.1 C. During a long inattentive period, one of 2.5 h (extended because of my presence in the area), the temperature of a down-covered nest dropped only 2.3 C degrees. The insulating quality of the female's down further was demonstrated by temperature drops of < .6 C in thred uncovered nests arid 2.2 C in one uncovered nest after 30 minutes. While female waterfowl usually cover the nest during these inattentive periods (Weller, 1964),

Mandarin females generally left the nest uncovered on breaks during the first few days and thereafter showed no consistent pattern. Three females left their eggs uncovered on one April morning. Another female covered her nest during a morning in June and on only one of two successive evening breaks on nights with the same ambient temperature.

When eggs of two clutches were pipping simultaneously, only one female Table X. Temperatures in middle of nests during incubation and at intervals during rest periods. Sensing accomplished with YSI portable thermister unit.

Temoerature (C) when nest at rest intervals (min.) date time female amb. box incub. 30 60 90 150 180 comments

1972 16 May 1500 0A18 19.0 28.9 36.1 35.7 nest covered 17 May 1715 0A18 26.6 36.7 17 May 1715 0A19 26.6 28.3 34.4 33.9 nest covered 17 May 1715 0A23 26.6 35.0

1973 11 Jun 1850 0A32 31.1 35.6 33.3 32.8 31.9* 35.5 nest uncovered 11 Jun 2025 0A31 31.1 35.8 35.3* 36.1 15 Jun 1853 0A34 31.1 32.2 35.8 35.6 nest uncovered (eggs pip) 15 Jun 1845 0A32 31.1 32.2 35.6 35.6 35.0 33.9 33.3* 35.6 nest covered (eggs pip) 16 Jun 0645 0A32 22.8 35.6 35.1* 38.1

*indicates resumption of incubation 50 left her nest covered.

On one occasion a male was observed incubating. After approaching the box with the female, the male quickly entered. The female then moved away and stood concealed under a pine tree. Both were feeding upon my arrival at 0600 the following morning. Although Rollin (1957) observed a male Wood duck in basic (eclipse) plumage incubating a clutch in July, male incubation is considered aberrant behavior in waterfowl.

Male behavior.- Savage (1952) indicated that males sometimes attended females into the brooding period. This was not determined in the present study, since clutches were removed at pipping. Pair bonds, subject to individual male variation, remained strong through most of the first incubation period. However, males left with the onset of any second clutches, beginning in late May. If by this time incubation was complete or the clutch destroyed, the pair remained together and began re-associating with other pairs by mid-June.

During early incubation the male waited for the female at a loafing or waiting spot close to and in view of the box (Plate 5).

From here he threatened or chased conspecifics of either sex entering the area. He attended the female during her rest periods. He did not feed but remained Alert while she fed, bathed, and preened. Such male waiting behavior had several advantages. First, fertilization for a second clutch was by the original male; and second, the male discouraged harassment by other males of the female during feeding or vocally warned her of approaching danger while she was incubating.

Males began to associate with each other as incubation progressed 51

or occasionally formed new, but weak pair bonds with another female.

But they usually continued to rejoin their original mates during inattentive periods. During late incubation the female sometimes vocalized (Kett) upon leaving the box. This attracted the male, who also vocalized (Pfrruib) and swam in Full-sail posture to meet her.

Most males had congregated by early June and were in the pre- basic molt. Its onset seemed to be partially related to the paired condition of individual males. An unpaired male kept in the open-air, over-wintering pens during spring 1973 began to molt 1-1.5 weeks before the paired free-ranging birds. Similarly, in 1974, two homosexual males and a third male separated from his penned female began molting by 8 June, nearly two weeks before the late June onset of molt in two other males attending incubating females until 12 June.

Renesting patterns.- Females sometimes renested following completion of the first clutch and its destruction by predators, or removal for incubation under bantams. In this situation five of six yearling females laid second clutches, all significantly smaller

(P<.001) (5.60 average) in size than their first clutches but higher in fertility (94%) than either their first clutch or the clutches of adult females. Adult females did not renest, presumably because of their initial later laying date. No females renested that had incubated initial clutches until pipping. The longest incubation period followed by renesting was 13 days.

Sowls (1955) stated that renesting intervals were directly proportional to the time spent incubating the first nest, averaging 52 three days and .62 days for each day of incubation. This renesting interval was close to that actually observed for four Mandarin females. In marked contrast was a female that laid six eggs but did not incubate and renested 22 days later. While only one of the five females renested in the same box, four nested in the same type of box.

Two females allowed to incubate their second clutches until pipping had incubation periods of 30 and 35 days. In the longer case, four of six ducklings had hatched and left the box (presumably being eaten by fish, since they were not with their mother) before my arrival at 0600. The other two ducklings were hatched by bantams.

The eggshells of the four escaped ducklings had been removed from the box, parts being found in the water and vegetation below the box.

During the four years three cracked eggs, also exhibiting bill punctures (presumably by the female) were found in water or vegetation outside of nest boxes. One apparently was cracked during laying and a replacement laid, while the other two were cracked during incubation.

Removal of damaged eggs or eggshells is reported for many waterfowl species (reviewed by Johnsgard and Kear, 1968), including Wood ducks, but has not previously been observed in Mandarins.

Summary of reproductive biology

Mandarin copulation behavior occurred primarily in the morning throughout the spring, except during the incubation period. Most incomplete sequences resulted from low intensity male behavior. Rape and promiscuousness were infrequent. 53

Nest box inspection averaged two minutes, occurred primarily

during the morning, and nearly always was by the female with the male

in attendance. Several potential nesting sites were inspected, with

females usually laying in the last one observed inspecting. Elevated,

elongate, pole boxes along the undisturbed wooded shoreline were

preferred.

Mandarin females laid when yearlings (10-11 mo of age) but with

less fertility (78%) than adults (90%). Initiation of first clutches

was between 3 April-30 May, with most begun between late April-

middle May. Yearlings began laying earlier than adults each year,

and were prone to initiate second clutches if the first were removed

or destroyed. The average clutch size was 9.5 eggs, the size not

being different between yearlings and adults.

Females began incubation (lasting 28-30 days) usually in the

afternoon the day the last egg was laid. They generally took three

inattentive periods during the first few days of incubation but only

two (pre-dawn and dusk) thereafter. They did not leave the nest boxes

at night. The nest temperature during incubation was 34.4-36.6 C

and dropped only 1-2 C during the female’s inattentive periods when

covered with down,

Mandarins form strong seasonal pair bonds which re-form successive years if both individuals live. The strength of the bond may be a result

of frequent copulation. Males attend and defend (usually) the female

throughout the spring, often waiting for her in view of the nest site

during incubation. 54

Egg morphology

Mandarin eggs are glossy and buff-white in coloration (Plate 6), typical of hole-nesting species. White presumably is easier for females to see in the dim light; and since the eggs are relatively inaccessable to predators, there is little need for cryptic coloration

(Lack, 1968). The average length and maximum diameter (or width) of 106 eggs from 11 different clutches was 52.86 x 38.64mm (Table XI).

This is smaller than the 56 x 39mm reported by Nakata (1965) for

Japan-reared Mandarins but slightly larger than average size (48.8 x 36.3mm) reported by Savage (1952) for 18 Mandarins reared in

England. Width was less variable than length, a condition also noted by Preston (1958) for the eggs of many birds species.

Fresh egg weights averaged 43.72g, not different from the 43g average obtained from the weights of 68 eggs provided by Kear in

England (pers. comm.). Weights and lengths of eggs laid by adult birds (46.26g and 53.74mm) were significantly greater (P c.05) than those of eggs from yearlings (42.68g and 52.52mm). Increased egg size with age also was found by Romanoff and Romanoff (1949) for chickens and reviewed for other species by Preston (1958).

The average fresh egg weights of some clutches were significantly different (P < .001) from others (e.g., F 0A44, 38.53±0.71g and

F 0A36, 44.45 t0.69). Batt (1970) found that large Mallard eggs were more tolerant to cold exposure than small eggs. Like other waterfowl,

Mandarins do not cover the eggs during the period of laying. Therefore, larger eggs might be advantageous to birds nesting in cold, northern latitudes. 55

Table XI. Mandarin duck egg morphology under various conditions. Means (±SD) of lengths, breadths, and weights are given. Statistic­ al analysis by Model I Single Classification ANOVA (Sokal and Rohlf, 1969).

age of female clutch egg state q S t 2nd yearling adult total fresh days pip n=67 n=39 n-106 n=14 n=14 n=87 n=17 n=57 length (mm) ±2’^ * 53.74 52.86 51.34 50.74 ±1.93 ±2.20 ±2.32 NS ±1.83

breadth(mm)38.27 NS 39.18 38.64 38.01 NS 37.11 ±1.23 ±0.98 ±1.18 ±1.11 ±1.00 weight(g) 43.72 38.96 39.14 ±4.05 ±10.29 ±4.17

length(mm) n=ll 52.75 first egg ±1.99 NS 52.06 last egg d2.02

breadth(mm) 38,33 ±1.14 first egg NS 38.21 last egg *1.41

fresh egg n=59 n=28 46.26 weighty * ±3.41

shell(mm) n=60 n=8 n=7 .2592 .2609 .2451 thickness NS ±.027 ±.020 ±.001 n = number of eggs * = differences significant (P-^.OS) NS = differences not statistically significant (P>.05)

I 56

Likewise, the ratio of day-old ducklings (26.9g) to the female's weight was a relatively high 5.3%. Delacour (1959) has stated that

this ratio is correlated to latitudinal distribution and therefore probably the rigorousness of the climate (Kear, 1970). Since feral,

Asian Mandarins breed at high latitudes (Lack, 1968), large ducklings might be expected and certainly would be advantageous for survival.

Egg weight decreased 10.5% (43.75 to 39.14 g) during development.

Weight loss primarily is from water evaporation and inversely proportional to the relative humidity (Romanoff, 1967). The large variation in weights at day 14 probably resulted from the small sample size and the variety of incubation methods used which resulted in non-uniform rates of embryo development.

The shell thickness of 60 eggs from 11 clutches averaged 0.259 mm

(range .191-.329 mm). The average thickness of shells from some clutches were significantly different from others (P < .05). The variation probably is a result of pooling the shells of both infertile and fertile eggs. Ducklings absorb some calcium from the shell during development (Rothestein, 1972) and subsequently the shell thickness of these eggs would be less than that of sterile eggs.

Duckling development

Body weight and linear growth.- Mandarin ducklings weighed 20-30 g between 12-36 h after hatching (also observed by Smart, 1965; and Kear, pers. comm.), with males being (29 g) slightly heavier than females

(25 g). The weight difference probably resulted from the small sample size. Both sexes doubled their hatching weight by one week 57 and were 13 times heavier after five weeks. Males were 40-100 g heavier than females through six weeks, a characteristic that appeared a useful sexing criteria during this time. Similar sexual dimorphism was evident in Mallard ducklings at six weeks (Prince, et. al.,

1970). All ducklings had nearly attained adult size at seven weeks

(Fig. 8; Appendix XV) except that adult weight was not attained before the end of the 1st pre-alternate molt in October. A similar development pattern has been observed for Wood ducks (Lee, pers. comm.).

The growth rate was more erratic during the physiologically stressful fledging period, body weight increasing only 8% for males and 11% for females during that time. Weller (1957) and Kear (1970) observed actual decreases in the growth curves of Redheaded and

Tufted ducks, cited similar findings in a number of other species of captive waterfowl, and suggested its occurrence in wild waterfowl under sub-optimum conditions.

Rapid tarsus development, characteristic of diving species

(Aythya) and presumably an adaptation for aquatic birds (Kear, 1970), was complete by six weeks. This rapid growth in Mandarins (and probably Wood ducks) also may be an adaptation to terrestrial locomotion. Although duckling time budgets were not determined because of the artificial brooding procedures, the birds seemed to spend considerably more time walking and picking at insects than swimming. Savage (1952) also noted this for wild Mandarin broods, citing the occurrence of mile-long overland expeditions the day of hatching. Stewart (1958) similarly found movements of wild Wood 58

Fig. 8. Growth of Mandarin ducklings observed at one week intervals from hatching for ten weeks and again on 28 October 1973. Mean ( + SO for body weight) indicated. Data from Appendix XV.

- 500

■ 400 ) g t

T • 300 H G I E W

, 200

» 100

- 0

AGE IN DAYS 59

ducks broods up to 3.5 miles from the nest.

Feather patterns.- The family Anatidae is among the more

difficult groups of birds regarding designation of various molts and

the replacement of certain feathers. Little is known about the

plumage sequence in the perching ducks, tribe Cairinini (Delacour,

1964).

Mandarin ducklings are brown dorsally and buff-yellow ventrally,

with two light dorsal spots. Pengelly and Kear (1970) noted that

these spots were darker in week old Blue duck (Mymenolaimus malacorphynchus) males than females, but this was not looked at in

Mandarins. The only evidence of sexual dimorphism, aside from the weight differential, was the reddening of male bills at 28 days

(usually not a reliable indicator for another one or two weeks) and change in the female's voice from the duckling contact call to the plaintive, more nasal 'Ack' at six weeks. The male's voice did not change to the whistle-like "Pfrruib" until the seventh and eighth week.

The bills of both sexes have a conspicuous orange tip, an episematic pattern, by which Nakata (1965) states parents recognize their ducklings. This may be supported by Kear's (1968) findings of little individual variation among the vocalizations of young Anatidae.

Facial markings consist of a dark post-ocular stripe and a second, less distinct parallel line just below the eye. Savage (1952) suggested the eyes are disproportionately large in the ducklings and develop faster than the rest of the body.

The post-natal molt into the juvenal plumage, similar to the basic 60

(eclipse) plumage of adults, was noticed first at two weeks of age.

Feathers usually appeared simultaneously for all individuals of

each sex, but some individual variation existed, probably resulting

from growth variation during the first few days (Fig. 9; Appendix

XVI). Description of the molting pattern was somewhat complicated

by inspecting only at weekly intervals, which precluded observance of

actual dates of first appearance of feathers. However, scapular,

tail and stomach feathers were the first to appear, with coverts and

secondaries breaking the skin a week later (Plate 7). Natal down

in the lumbar region and under the wings was lost between eight and

eleven weeks. Mandarin duck!ings apparently do not have a basic I

plumage, molting from the juvenal directly to the alternate plumage

as black chest stripes and green and rust feathers of the alternate

(nuptial) plumage began to appear on the head and mane of males at ten

to twelve weeks. The rectices but not remiges were replaced in the

pre-alternate .1 molt, which always was completed in all birds by the

end of October.

Primaries were first observed breaking the skin at three and four weeks in females and males, respectively, with all ten complete in

both sexes by ten weeks. Flight occurred in poorly pinioned birds

of both sexes, at seven weeks, with the first-seventh primaries cleared at this date. This fledging date was one week later than reported by

Lack (1968), but one and three weeks earlier than those given for

Mandarin and Wood ducks, respectively, by Savage (1952).

This fledging period is similar to that of the Gadwall, (Anas

streperà) (Oring, 1968) and other dabbling ducks (Anas spp. ) (Weller, Sli Fig. 9; Feather development and rcaturational changes in 5 male arod: 8 female Mandarin ducklings observed at one week intervals from time of hatching, 1973. Beginning and end of lines; indicate first appearance and full emergence of particular feathers in each bird.

BILL

VOICE

PRIMARIES

SECONDARIES

SCAPULARS

BREAST AND STOMACH

HEAD

I vi»; I —.—.n| i ..... I in u ■ T n— ; ii 111

7 14 21 28 35 42 43 56 S3 70 AGE IN DAYS ‘female's bill color does not change 62

1957). Streseman (1931) and Heinroth (1928) [both cited in Kear

(1970)], have suggested that the shortened fledging period of the less aquatic dabblers (relative to the divers) is related to their vulnerabi1ity to predators and pressure to shorten the period of relative helplessness.

Behavioral patterns.- The ability of day-old perching ducks to climb from a tree cavity nest as high as 50 feet and jump to the ground upon the call of the female is unique and well-known (Gottlieb,

1968; Johnsgard and Kear, 1968). Mandarins sometimes climb as much as six feet inside a tree cavity from the nest to the entrance

(Savage, 1952). Day-old Mandarins were able to escape from a bushel basket in two jumps of six to eight inches each by digging their long

(4mm average), sharp, middle claw into the wood, hanging a few seconds, then springing, and grabbing the top, again with the claws. Meade-

Waldo (1912) similarly noted two-day old Mandarin ducklings leaping

18" from the water to a branch. The anatomical features, primarily thigh and leg muscle size related to this ability in perching ducks have been discussed by Rylander and Bolen (1970). In comparison,

Pengelly and Kear (1970) noted that Blue ducks, non-perching ducks which inhabit fast-moving mountain streams, climb using their chins, and could jump six inches only after six days.

Mandarin ducklings at hatching were extremely inquisitive and active, pecking at most moving objects. By putting meal worms on the duckling starter food, feeding activity could be initiated in solitary or small groups of newly hatched ducklings during the first three days in the brooders. This was a most critical time period, since failure 63 to eat or feed consistently, resulted in death. It was not as great a problem in larger groups of ducklings, since feeding behavior, once initiated, spread more rapidly in the form of contagious behavior.

In the wild, feeding probably is learned from the mother.

Despite a two week difference in hatching dates, ducklings were a close group, They frequently nibbled at each other's neck and head, and at least until release to the out-door pens at six weeks, slept touching each other. Little agonistic behavior was exhibited until attainment of the alternate plumage.

The rate and sequence of appearance of comfort movements

(shaking, stretching, preening, bathing, and other such movements) was like that observed for Mallards, Anas piatyrhynchos (McKinney, 1965c), and Blue ducks (Pengelly and Kear, 1970). Most were expressed in the first two days (Table XII).

Newly hatched ducklings performed the same preening movements as adults. Nibbling and preening sessions, consisting primarily of movement to the belly, neck, and wings, were frequent but brief during the first three days. Perhaps this is related to rest pauses and their unbalance. As with Mallards (McKinney, 1965c), movements were made in the direction of the preen gland during the first two days, but fully developed oiling sessions, with associated head-rolls, were not observed until the fourth day (Table XIII). Dashing-and-diving, a contagious and rather complex motor activity, occurred in one Mandarin duckling at six days and the others at eight days. This appearance was considerably earlier than the 13-14 days reported for Mallards and may indicate the advanced nature of Mandarin development relative to 64

Table XII. Age at which comfort movements first appeared in five Man­ darin ducklings, hatched 17 June 1973 and kept in captivity. Unless otherwise stated, the day indicates the time movements were seen for all ducklings.

age (days) comfort movements

1 nibbling movements-belly, neck, vent, wings and occasionally in vicinity of oil gland

head-shake, body-shake, wing-flap, tail-wag, jaw-stretch, head scratch, leg-arid-foot-peck

2 swim-shake

3 both-wing-stretch and wing-shuffle

4 oil preen with head-roll on back and side (one duckling)

6 dashing-and-diving (two ducklings)

7 all nibble at oil gland while swimming (first time observed)

8 dashing-and-diving (all ducklings), high intensity bathing (pecking at sides, head, neck and oil gland while head- dipping) Table XIII. Average number of comfort movernents/duckling/ten minutes at four, seven, eight, and nine days of age - June 1973. Observations on five ducklings began ten minutes after swimming sessions.

days 4 7 8 9

Preen or nibble vent 15 3 1 1 wing and side 4 5 7 4 back 1 0 2 1 belly 5 2 1 4 neck and chest 20 30 22 30 feet and legs 1 2 2 1 oil gland 4 7 5 9 dorsal head roll 1 4 4 3 side head roll 0 3 1 2

Both-wing-stretch 0 0 1 0

Wing-flap 1 2 1 0

Body-shake 4 3 5 6

Head-shake 3 4 6 2

Scratch 10 13 10 15 66

dabbling waterfowl.

Summary of developmental biology

The buff-white Mandarin eggs averaged 52.86 x 38.64 mm with the width less variable than length. Eggs from adult females were larger than those from yearling females. Egg weight decreased

10.5% (43.75 g - 39.14 g) during development. The average shell thickness was 0.259 mm, but variable both within the same egg and among eggs in a clutch.

Ducklings weighed 20-30 g at 12-36 h post-hatching, with subsequent development rapid. Sexual dimorphism was evident between three and six weeks, as males were 40-100 g heavier than females and had red bills. All ducklings were nearly adult size .by seven weeks, except for body weight which was attained after the pre- alternate molt in October. Voice changed in females at six weeks and in males at seven-eight weeks. The ducklings were very active and inquisitive, spending more time walking and picking at insects than swimming. They performed most adult comfort movements within the first two days. Mandarins apparently do not have a basic I plumage, but molt directly from the juvenal to alternate plumage.

Fledging occurred at seven weeks.

Ageing year!ings and adults

Separation of yearlings (birds having completed only one pre­ alternate molt) from adults (birds having completed at least one post­ alternate molt) is necessary for determining the influence of age on reproductive capability. Such data are important and useful both in 67

management of wild waterfowl populations and in propagation of exotic waterfowl, such as the Mandarin duck. The'monetary importance becomes more evident in light of the following data:

1) the present market value for the Mandarin is nearly $50.00/pair.

2) females of both ages usually laid between 8-12 eggs/clutch

and often renested if the eggs of the initial clutch were

artificially incubated.

3) the yearling female Mandarin fertility rate (78%), was less

than adult female fertility rate (90%).

Many morphological characteristics have been evaluated in ageing waterfowl: cloacal characters, tertial and covert markings, bill spots, primary feather length (summarized by Dane, 1968), and eye color (Trauger, 1974). When some of these methods were applied to known-age Mandarin ducks, best results were obtained using primary feather length and secondary feather markings. Eye (iris) color remained brown from hatching through adult stage in both males and females and could not be used as an age indicator.

Primary feather lengths.- Yearlings and adults could be separated on the basis of primary feather length. All primaries of adult males and females were significantly longer than those of yearling males and females, respectively, despite individual overlap (Table XIV).

However, use of the 7th and 9th primaries resulted in greatest ageing accuracy, since the difference in mean length of these feathers was significant (P <• .05, t - test for paired comparison) when feathers of the same female were measured when she was a yearling and again as an adult (Fig. 10). ’ Table XIV. Primary feather lengths of male and female Mandarin ducks. Measurements (in mm) made during fall 1972 and 1973 while birds were in alternate plumage. Statistical analysis by Model I Single Classification ANOVA (Sokal and Rohlf, 1969)

males females mean + S.D. range N P mean ± S.D. range N P

7th Primary yearlings 154.75 ± 5.23 146-161 16 149.05 + 5.67 140-158 22 .001 .01 adults 162.89 ± 4.94 157-169 9 156.60 + 5.06 149-164 10

9th Primary yearlings 167.75 + 3.37 160-173 20 162.88 ± 4.13 155-169 26 .001 .001 adults 174.40 + 3.03 170-179 10 170.58 ± 4.70 165-178 12

10th Primary yearlings 168.39 ± 4.58 162-177 18 165.62 + 2.80 160-170 21 .001 .00001 adults 175.50 + 3.92 169-180 10 169.54 ± 2.38 166-173 11 69

Fig.10: Frequency distribution of the 7th and 9th primary feather lengths of yearling and adult female Mandarin ducks, measured during fall 1972 and 1973. yearlings-individuals not having completed a post-alternate molt, adults-individuals having completed at least one post-alternate molt. Paired-comparisons indicates the same individuals measured as yearlings and adults.

7th primary length 9th primary length

180

178 - 1

176 1 1 2 174 -

172 2 2

170 1 1 168 1 1 1 3 3 166 1 1 4 1 2 1 m 1 1 n 164 1 1 1 1 4 n i 2 2 1 162 h 1 t 1 1 g 1

n 160 e l

’ 158 1 1 1 1 156 1 1 1 2 1 1 154 1 1 1 152 - 2

150 3 1 1 1 148 1 1

146 1 144 4 3 142 •

140 » nm— 2 1 year. adult year, adult year, adult year, adult total paired- total paired- sample comparison sample comparison 70

The mean primary lengths of all three primaries of yearling and

adult males were longer than those of the same age females. Such

sex differences also have been found for other avian species by

Stewart (1963), Dane (1968), and Johnson (1974).

Secondary feather markings

The tips of the secondaries of yearling females generally were

more speckled and less complete (i.e., white not on both sides of the

rachis) than those of adult females (Table XV; Plate 8). Seventy-

eight percent of the yearling females showed speckled or incomplete

white markings, while markings on 75% of adult females were complete.

However, observations again on a paired-comparison basis were less

convincing. Of six females, three whose secondaries were speckled

or incomplete as yearlings were complete as adults; but the other

three when adults, retained the same markings as when yearlings

(Fig. 11). These markings were not at all effective in separating yearling and adult males, as the white edges were complete in both

age groups. Nutrition probably was not a factor in the feather length

or marking differences, since most birds were from the Schedel

population and the supplemental diet provided all birds from the three

distinct populations was similar.

i 71

Table XV. No. and % yearling and adult Mandarin females with "complete" or "incomplete" white markings on the distal ends of the secondary feathers. No. of individuals tested from each of the three populations is indicated.

secondary feather markings no.* complete incomplete % complete yearlings 27 6 21 22

adults 12 9 3 75

★ yearlings n=27 (21 Schedel, 3 Strutz, 3 Webster) adults n=12 (7 Schedel, 2 Strutz, 3 Webster) 72

Fig. II: Comparison of distal markings of the 5th-llth secondaries of 6 Mandarin duck females as both yearlings and adults, while in alternate plumage. Observations made during fall 1972 and 1973. 73

Time budgets

Time budgets of the individual pair and the entire free-ranging population are based on six continuous (dawn-dusk) observations during different periods in 1973, supplemented by shorter observations of both penned and free-ranging birds during the same year and in 1972.

Two all-night observations were made during summer 1974. The behaviors of the pair are believed to be indicative of the entire population. Likewise, the activity patterns of both this pair and the entire flock, within the limitations imposed by the habitat and pinioning, probably are representative of the activities of feral birds.

Spacing behavior

Savage (1952) found nests of feral Mandarins widely dispersed and observed pairs feeding in fields as far as one mile from their nests.

His brief description suggest the operation of spacing mechanisms and territoriality in Mandarins. Both concepts are reviewed for waterfowl in general by McKinney (1965b), who categorized the responses of ducks during encounters as a) visible and/or audible displays; b) attack; c) escape and avoidance; d) sexual pursuit and/or rape; and e) sociability. I observed all of these responses during Mandarin social encounters, with certain types predominating during nesting.

Sociability and territoriality.- Mandarins were very social during the first days after release, a time period presumably comparable to arrival on breeding grounds by feral birds, initially

75% of the free-ranging birds loafed next to penned birds, but this 74

sociability decreased with the onset of pre-laying and laying activity

(Fig. 12). With the break-up of the flocks, males began to defend

only the area occupied at that particular time by the female. This

area was flexible and moving, with encounters between birds often

at close range, e.g., at a nest box during inspection or laying.

The clumped nesting pattern indicated that pre-laying encounters

were ineffective in spacing pinioned birds on the lakes. The

agonistic behaviors characteristic of these encounters were primarily

threats, influenced by the close proximity of birds and presence of

the female. Males primarily performed Bridles and Bill-jerks, Head-

shakes, Open-bill threats and Burps; females Incited and Head-

shook. Head-shake and Bi 11-jerk threats were performed by both

sexes of all pairs inspecting boxes (and usually the intruders as

well) on all six occasions that I observed interruption of nest

inspection. Likewise when pairs met while leaving loafing spots to

enter the water, females Incited and males Burped and Bridled

(Plate 9).

Most agonistic encounters and all observed fights occurred among

penned birds. One pair dominated each pen, and these males

continually Bridled or Jabbed at each other through the fence while their females Incited. On one occasion the dominant male in

a pen killed another male. Both birds behaved in a manner similar to

that described for mice by Scott (1972; p 74). The subordinate male returned the threats for nearly two weeks until his female, who also had been continually chased, stopped Inciting. He then began Fig. 12. Locations of free-ranging Mandarins during the first two weeks following release to the lakes on 2 April 1973, during laying period, and during fall (pre-alternate molt and alternate plumage periods). Points are based on birds' locations between 1200-1300. Out-door breeding pens are indicated by box and contain 8 birds.

PRE-LAYING PRE-LAYING LAYING PRE-ALTERNATE 5 April • 9 April • 3 May • |O-Sept • 6 April ■ 13 April POST-ALTER NATE

<_n 16

to retreat but was continually chased. Since he was unable to escape, he passively laid in the water and submitted to the attacks.

Although he was removed from the pen, he did not eat, became blind, and died.

Incubation.- The onset of incubation and presence of males in individual waiting areas reduced the general activity and hence the number of social and agonistic encounters.' Except while accompanying females on their rest breaks, the kinds of encounters changed to long range threats and slow pursuit. From their waiting areas, males usually threatened intruders by assuming Full-sail posture and performed either Head-shakes or Bill-pointing. One male routinely patrolled the entire end of the lake in which the female nested.

On these patrols he often lowered his head near the water and slowly pursued the intruders. These pursuits were extended and often were followed by displays directed toward his female. Aerial chases, common spacing behaviors among most waterfowl (McKinney, 1965b), did not occur because of pinioning, this may have been an important factor in the changed nesting pattern.

Waiting males throughout the spring were more tolerant of unpaired males than they were of either solitary females, paired males, or other pairs. This also was observed for Gadwalls by Dwyer (1974), and

Erskine (1972) reported conflicts between male Buffleheads only in the presence of a female. As incubation progressed, tolerance to paired males increased, except in the presence of the female. 77

Population time budgets

General daily patterns.- A diurnal activity rhythm was not as evident during the spring as in summer and fall (Fig. 13). During pre-laying and laying periods intense feeding began prior to dawn, continued until 0730, followed by nest box inspection and then intermittent feeding during the remainder of the day. Extended periods of inactivity, in which more than 50% of the birds either loafed or preened, occurred only between 0830-1030 and 1430-1600 in the pre­ laying period and during the afternoon of the laying period.

Individual pairs followed feeding with short loafing and preening sessions.

The activity of the birds (excluding two pair still incubating) during the post-incubation period (i.e., prior to the onset of the pre-basic molt) was similar to that occurring with the assumption of the basic (eclipse) plumage. Feeding during these two periods still was evident usually until 0930, but began later in the afternoon than during the laying period and extended to twilight (1730-2100).

Evening feeding during these two periods primarily was grazing, usually with both birds participating. Between 1615-1700, most birds swim-fed, while from 1730-1830 grazing was most prevalent with the others eating corn or continuing to swim-feed. Nearly exclusive loafing occurred during mid-day.

In marked contrast to the post-incubation and basic molt periods were the activities during the pre-basic and to a lesser extent, the pre-alternate molt periods. Nearly all birds, both penned and free- ranging, preened between 0600-1030, slept until 1330, and resumed 78

SWIM FEED PREEN SWIM

□ LOAF GRAZE EAT CORN

PRE-LAYING

LAYING

POST INCUB.

PRE-BASIC MOLT

BASIC PLUM.

PRE-ALT ER. MOLT !

IOO

ALTERNATE PLUM, so

TIME

Fig. 13. Proportion of free-ranging flock engaged in various activities at 1/2 h intervals during full-day observations 1973. (n=5-ll) eating and preening. They presumably fed prior to the initiation of observations at 0600. The intensive preening in July and

September probably was related to feather replacement irritation; in September it may also have been related to resumption of the breeding plumage and pairing.

Display preening, unless highly ritualized (such as Preen-behind- the-wing), was difficult to distinguish from normal comfort preening.

Some preening during September and much of it following the pre­ alternate molt (October) seemed different from that of the post­ incubation period. The birds did not pick or pull feathers (few feathers being found in the loafing spots) but instead touched and rearranged them. The belly, chest and wings were preened more frequently than others. The preening movements were sometimes accompanied by Head-flick displays. Social facilitation of this activity was evident as birds in the group preened simultaneously

(see 24 October; 0930, 1400, 1430, 1530 and 1700), males more vigorously than females. Most adults had paired by this date as indicated by mutual Caressing movements and mutual preening. The pair-drake led during many fall movements, opposite of the spring.

The change in total flock activity patterns with each period is shown by Fig. 14. Loafing, lowest during the pre-laying and laying periods, and feeding generally were inversely related. General swimming, associated with nest box inspection and pursuits, decreased with the end of the laying season. Increased swimming time during the pre-basic molt and basic plumage periods occurred as a result of the change in social patterns from spring pairs to the flock Fig. 14. Proportion of birds engaged in each, activity for the entire day in each of 7 periods, spring - fall 1973. Data based on recordings at h h intervals throughout the whole day watch.

iOO«|

o00 81 condition, as the birds entered the water for sleeping earlier.

Nighttime patterns.- During the post-incubation and pre-basic molt periods, Mandarins entered the lakes at dusk and fed either actively or intermittently until 2200 (Fig. 15). They swam slowly or quietly along the shoreline, picking among the algae mats and occasionally snapping at emerging insects. One pair of Mandarins continued to feed with Chileo Widgeons and Australian and European

Shelducks as late as 0200 h on 4 June, but all feeding had stopped by 2400 h on 1 July. Based on these observations, their similarity with dusk surface-feeding patterns of 1973, and nighttime feeding observations on other waterfowl by Swanson and Sargeant (1972),

Mandarins probably actively fed until early morning during the spring.

After feeding, the birds floated, heads drawn in and bill foward, concealed under the over-hanging shoreline vegetation. This was a different position from the head-tucked posture characteristic of most sleeping waterfowl (Johnsgard, 1958) and from Mandarins sleeping on the shore and may have been an alert resting position. Most predation on Mandarins occurred during the night, especially in the early spring and fall (Table II).

Two females in their third week of incubation during the 4 June observation did not leave their nests at night, further supporting the 1973 event recorder and kymograph paper data (see Reproductive and developmental biology, p45). Both pair-drakes stood or swam near their female's nest throughout the night.

Feeding patterns.- Mandarins utilize a variety of feeding 82

7

Fig. 15. Nighttime activity patterns of 5 - 8 Mandarins observed during the post-incubation (4 June) and pre-basic molt (1 July) periods 1974. Based on proportion of birds engaged in each activity at 1/2 h intervals.

« w -I S H LOAF *4 W PREEN SWIM FEED A » ■ « (SLEEP) GRAZE □ SWIM (activlly) % 4 JUNE

I JULY 0 5 3 0

TIME 83 techniques but are predominantly surface feeders and grazers (Table

XVI). The specific nature of their food is difficult to determine without stomach analysis, but some conclusions seem possible. The feeding activity of individuals of a pair during the spring and the flock during summer are closely related. Feeding patterns were influenced by changes in social patterns, particularly the onset of nesting. Feeding, predominately surfacing feeding, occupied a far greater percentage of the day during spring. Mandarins generally fed during the early morning and late afternoon, times when invertebrates are most available (Verbeek, 1972). They swam along the shore and among algae mats, presumably picking at both vegetation and invertebrates and snapping at those insects that flew up. During terrestrial grazing they also seemed to search and pick, again probably taking both vegetation and invertebrates. While moving under vegetation they picked insects, probably emerging odonata, from the leaves.

I have also seen all members of the flock during June rapidly zig-zagging across the surface, apparently catching emerging insects.

I have never seen them dive for food although Savage (1952) indicated that it infrequently occurred. Corn was provided for ducks on the estate but was eaten for only a few minutes each evening and was a relatively minor part of their diet. In England they prefer Beech mast» acorns and chestnuts (Savage, 1952).

Individual time budgets

Pre-laying and laying.- Birds must spend a certain amount of time

BOWLING GREEN STATE UNIVERSITY LIBRARY 84

Table XVI.Mandarin feeding patterns, based on the proportion of birds engaged in each type of feeding at 1/2 h intervals during 1973.

feeding type no. h h surface tip- in­ 0* obser. graze feed fish corn sects* alert Pre-laying (4-20 Apr.) 56 15 37 0 5 5 0 38 males females 16 35 0 7 6 0 36 average 16 36 0 6 6 0 37

Laying (21 Apr-10 May) 100 males 26 30 0 2 7 5 32 females 25 43 2 4 8 6 12 average 25 37 1 3 8 6 21

Post-incubation (11 May-10 Sept) 115 males 31 37 0 2 10 2 18 females 31 42 2 4 11 2 8 average 31 40 1 3 11 2 13

♦insects - from leaves or air 85 in self-maintenance with migration, reproduction, and molt requiring extra energy (Pitelka, 1958). The amount of time devoted to different activities influenced survival (Orians, 1961). The daily behavior of waterfowl in general, and Mandarins in particular, consisted primarily of repetitious feeding, preening, and loafing, with variations in these patterns related to the different social periods of the Mandarin.

Observations of a pair throughout the various periods supported the findings of the flock (Fig. 16). The pair, usually led by the female (Plate 10), was active most of the day during pre-laying and laying periods, but the extended mid-day inattentive period, characteristic of summer and fall, was eliminated. The female spent nearly two-thirds of the day feeding. Extensive feeding during the nesting period also was observed for Gadwalls by Dwyer

(1974) and related to obtaining the food reserves necessary for laying.

Males fed 38% and 14% less than their female during the pre­ laying and laying periods (Fig. 17). Bengston (1972) and Dwyer (1974) studying Harlequin ducks (Histrionicus histrionicus) and Gadwalls, respectively, found that the male fed almost half as much as the female during the time the pair was together. Dwyer (1974) concluded that the drake provided the seclusion necessary for the increased feeding at these times by chasing other ducks. Perhaps just as important for an isolated, non-social bird like the Mandarin, he also was alert to warn the female of approaching danger or to chase intruders, allowing her to intensely feed. Fig. 16. Chronologic daily activity pattern of a free-ranging Mandarin pair during full' day observations in 1973. Agonistic encounters indicated by dots. SWIMFEED RM2 PREEN PRE LAYING SWIM LOAF WF22 GRAZE INCUBATE R M2 EAT CORN LAYING COPULATE NEST BOX INSP. W F22 WALKING (ALERT) LAYING

co CD TIME Fig. 17, Activity patterns of Mandarin pairs during spring 1973. Proportions derived from the number of min spent in each activity by a pair during full-day (dawn-dusk) observations.

swim feed preen 100 walk ' 7 6 ’ graze swim 80 ooo incubate loaf eat corn o• • • • « » • • nest box insp. 60 t n e c r e p 40

20

R

Incubation.- The percent time spent in the various activities by males was similar throughout the incubation period (Fig. 17).

Except for periods at noon and late afternoon, they remained constantly alert while loafing, and even patrolled the nesting area and engaged in several agonistic encounters. They fed (usually grazing) while the female incubated. Savage (1952) has reported similar male activity rhythms for wild Mandarins. Although the male remained attentive to the female during the fourth week of incubation, his restlessness at this time was evident by his many brief activity periods and his partial association with other birds. His preening frequency increased 10%, probably associated with the pre-basic molt.

The females incubated and fed about 80% and 10%, respectively, of daylight hours each day throughout incubation. They were very active during their incubation inattentive periods, loafing or sleeping on only one of six recorded occasions. Incubation breaks consisted of feeding, bathing (usually), and preening; the prevalence of feeding decreasing during evening inattentive periods from 80% during the first day of incubation to 40% in the last two weeks

(Table XVII). This initial extensive feeding pattern may be a carry over from the great amount of time spent feeding during pre-laying and laying periods. Males attended their females during these inattentive periods but participated in feeding only when their mates grazed

(usually during the evening inattentive period). The rest of the time

(57-90%) males remained alert.

Feeding patterns.- Morning and afternoon feeding patterns 89

Table XVII.Percent time spent in various activities during two morning and three evening inattentive periods by a Mandarin pair during the incubation period, spring 1973. Data are based on the number of minutes :spent by each bird of the pai r i n each activity.

Percent time morni ng evening

2nd week 4th week 1st day 2nd week 4th week 0 0 0 + o’ + cf ? cf + cf ?

Feeding (total) 9 37 35 69 41 81 6 40 15 41 aquatic 0 20 0 34 0 32 0 12 0 0 graze 9 9 35 31 36 38 0 22 15 38 corn 0 8 0 4 5 11 6 6 0 3

Loaf or alert 63 31 21 0 20 0 33 0 29 0

Preen 12 15 32 23 14 11 49 40 35 51

Swim or walk 15 17 13 8 25 8 12 20 21 8

Duration of inattenti ve period (min) 65 12¡0 56 85 14 3 90

differed. During the morning inattentive periods females surface-fed,

presumably on aquatic insects (Swanson and Sargeant, 1972), and

grazed. However, 67% of 140 evening feeding minutes for the three

observations were spent grazing; only 21%surface feeding. Corn,

provided for all waterfowl, decreased in use with incubation and

accounted for the remaining 12%.

The change in over-all feeding patterns from 50-90% surface

feeding during pre-laying and laying to 27% during incubation, supports

the observations of Holm and Scott (1954) and Swanson and Nelson

(1970) that breeding females require nigh protein diets, and consume

a much greater proportion of aquatic invertebrates before and during

egg-laying. A shift from morning surface feeding to entirely evening

grazing, also evident during early spring, presumably resulted in

greater inconspicuousness and hence reduced predator susceptibility

during a time of high owl activity.

Renesting,- Activity during the renesting (re-laying) period,

following removal or loss of first clutch, was strikingly different

from that during initial laying (Fig. 16). During reriesting, the

pair spent somewhat more time preening; the female, about half the

time feeding. Unless renesting females actively fed throughout the

night, a distinct possibility based on early summer all-night observations, this decreased daytime feeding is not easily understood.

Krapu (1974) found that renesting female Pintails (Anas acuta) weighed

25% less than their prenesting condition (presumably due to loss of

fat reserves during initial incubation). This should increase the dependency of the renesting female on available food and result in 91

increased not decreased feeding.

Summary of time budgets

The daily activity patterns of free-ranging, pinioned Mandarins generally were like those observed for feral British birds and consisted of repetitious feeding, preening, and loafing (variations

in the amount of each related to the different social periods).

Activity continued throughout the day during the pre-laying and laying periods. Females spent 8 h/day feeding (primarily surface feeding); their mates, considerably less. They instead remained alert, presumably reducing harassment to the female by other drakes and allowing her to feed extensively.

Diurnal activity rhythms became obvious during the summer and fall as the birds fed and preened during the morning and late afternoon and loafed the rest of the day. Renesting time budgets were different from those of the initial nesting period, as females preened more and fed less.

During incubation females were on the nest. 80% of the day and did not leave the nest at night. They generally had pre-dawn and dusk inattentive periods each day. Males threatened conspecifics and waited for their incubating females at loafing spots (often in view of the nests) throughout incubation and attended them during their

inattentive periods. Females fed 90% of the time, preened, and bathed during these periods.

Mandarins surface-fed, grazed, tipped-up, and caught fish and flying insects. The general feeding patterns changed during the day 92

(swim-feeding in the morning and grazing in the evening) and during

the season (from nearly exclusive to only one-third surface feeding

in pre-laying and incubating periods, respectively).

Preening sessions nearly always followed bathing and feeding.

Preening was greatest during the pre-basic and pre-alternate molts

and after assuming the alternate plumage. Intense preening during

the first two periods undoubtedly was related to irritation of

feather replacement; perhaps during the latter period, to pairing.

During the summer evenings Mandarins entered the lake and swim-

fed along the shoreline until 2200. After feeding they floated

alertly, concealed under over-hanging shoreline vegetation the

remainder of the night. One pair, however, continued to feed until

0200. Night feeding also may occur during the laying and renesting

periods.

Displays

Descriptions

Displays are defined by Moynihan (1955) as ritualized behavior

patterns that have signal function and have become specialized in adaptation to that function. He lists four sources of visual display: displacement activities, redirected and intention movements, and

ritualization. McKinney (1965c) has discussed in detail the

ritualization of displays from comfort movements (i.e., Tail-wags,

Preens, Shakes, Bathes, Wing-flaps, and Head-flicks) in Anatidae courting parties, during hostile encounters, and in pre-copulation sequences as having originated from movements reacting to wetting of 93

the feathers, alighting, or pre-flight signals. The terms "motivation"

and "tendency" are used in the same manner as McKinney (1961) as "mere

shorthand descriptions of the state of an animal as judged by the

movements it makes" (Tinbergen, 1959). Displays usually are named

for their most obvious component (Tinbergen, 1959). The designation

of certain Mandarin behaviors as displays has been based on the

relative positions of the birds during the performance of the movements, changes (i.e., exaggeration) in the nature of the movements

from normal comfort movements, and their use and frequency of occurrence

in various situations. These are at best indirect methods, but as

McKinney (1970) has stated, there is no experimental proof of signal

function.

Since I observed some displays not previously attributed to

Mandarins, I have used the names of analogous displays given by

Lorenz (1941), Johnsgard (1965) and McKinney (1961, 1965c, 1970) for other species and have followed the convention of these and other authors in capitalizing them. Where there is no risk of confusion and unless otherwise stated references are to these citations. In addition, references to those displays previously described for the

Mandarin follow the display name. The term pair-drake (Weller, 1967) is used to designate the male of a pair.

Behaviors.- Mandarins displayed in courting parties, during encounters between pairs, upon the approach of a single male, and before and after copulation. Most displays were used in several of these situations; others were situation specific. A list and description of these displays and their accompanying vocalizations 94

follows. (See Figures 18-22 and Plates 11-27 for illustrations and

photographs of many displays).

Introductory-shake (Lorenz)

Called Swim-shake when performed in the water, during which the male simultaneously back-peddles, shakes, and slightly rises. Plays a minor role in Mandarin display (Lorenz). Similar to comfort movement body shake when performed on land except more stiffly executed. Not observed in male social display but was one male's first response to his female's assuming the Prone position.

Display-shake or Grunt-whistle (Lorenz)

Performed during male social display, following agonistic encounters, and sometimes after copulation. Consists of the head (with hood erect) being lowered, rapidly thrust upward, then shaken together with the tail. Accompanied by a call, uttered with the upward movement, and designated by Lorenz as a whirring "F^W" (the "w" and "r" pronounced together). Similar to the corre­ sponding display in the Wood duck (Johnsgard).

Double-display-shake or Double-grunt-whistle (Lorenz)

Performed only in the water. Consists of two movements. In the first stage the neck is stretched forward, and the bill dipped, then rapidly thrust upward; the body rising.from the water. The upward head movement, accompanied by sneezing whistle described by Lorenz as "Gnk-zit," is followed by the head being pulled back to the resting position. In the second movement, the bill is again dipped into the water rapidly, closer to the chest. The neck then thrust straight upward, with an accompanying vocalization and tail-wagging. As in the Drink-mock-preen, the white margins of the secondaries are temporarily exposed.

Occurred in high intensity display situations, always in the presence of females, though not always directed to them. Apparently ritualized from the general body shake but somewhat intermediate between the simpler Display-shake and the more complex Double­ grunt-whistle of Anas (Lorenz; Johnsgard).

Mock-drink (Lorenz)

Performed by males while standing with or swimming broadside I 95

to the female. Bill dipped into the water, held momentarily, then rested on the chest slightly to the side of the female. Differs from normal drinking in the male's apparent deliberate positioning of himself slightly ahead and lateral to the female and in his erected crest and sail feathers. Isolated Mock- drinks occur or immediately precede the Drink-mock-preen display. Probably functions primarily in pair maintenance.

Drink-mock-preen (Lorenz)

Also Preen-behind-the-wing (Johnsgard). An absolutely linked behavior sequence in male Mandarins (Lorenz) in which an exaggerated Drinking movement (with water droplets sprayed during the upward head movement) is followed by touching the bill to the inside of the raised sail feather; the normally concealed white edges and blue specula of the secondaries exposed. Performed primarily by a pair-drake, often while standing with his female at the edge of water; the pair may simultaneously Mock-drink (Lorenz). Observed frequently during social display and following- agonistic encounters. Clearly directed to the female; performed from a position ahead of and lateral to her with the lifted sail feather and Mock- preen invariably to her side. One male performed six such displays within eight seconds, re-adjusting himself to the female's position between each display. Mock-preen by itself occasionally was performed by the female to the male.

Burp (Lorenz)

Performed both on land and in the water. A sudden, sometimes jerky, upward thrust of the head in which "it looks almost if the bird must put his neck out of joint" (Lorenz). The hood erected and the crest spread; the pupil minute and the white eye ring very conspicuous. A "nasal" 'Pfrrruuiehb' is uttered exactly at the moment of maximum extension of the neck. Various degrees of intensity; 'sails' not always completely erected nor neck jerk always exaggerated.

Occurs as an introduction to social courtship in Anatidae (von de Wall, 1963) and thought to combine the function of call- note and warning (Lorenz; McKinney, 1965c). Observed in the same situations in Mandarins, occurring primarily during female Inciting, following agonistic encounters, or in social display bursts.

Rebuff (Savage)

Called Bridling (McKinney, 1970). Nearly always performed 96

on land. Accompanied by Bill-flicks and continuous "Jibbering"; functioned as a threat when birds came into close contact; upon leaving loafing sites, during social display, or following agonistic encounters. Given in the last two situations by a pair-drake standing close to his Inciting female and probably directed at the intruding males. Head moved slightly backward, the bill pointing upward but resting on a protruding and expanded chest.

Bi 11-jerking

Frequently performed Mandarin display, which as Johnsgard suggested for other waterfowl, was used as a greeting or sexual display toward females and a threat display toward males. Accompanied Bridling. Began with the head tilted slightly downward; followed by a rapid vertical jerk of the bill with the hood expanded. Occurred primarily during social display. Given as greeting by pair-drake to his female upon her approach during an incubation break and by male of a pair inspecting nest boxes upon approach of intruders.

Aiming movements (Lorenz)

Also called Head-pumping (Johnsgard) and Neck-craning (McKinney, 1965c). Major component of pre-copulation display in which slowly repeated forward and upward, rather rotary head movements are performed. Given by drake while swimming closely around the Prone female and by female as the initial movement prior to assuming the Prone position or in low- intensity, pre-copulatory situations. In the latter case seemed to re-stimulate male Bill-dipping and Neck-craning.

Bill-dip (Johnsgard)

Two forms observed. In each, the bill briefly dipped, sometimes entering the water but sometimes stopping above it. Occurred most frequently in conjunction with Neck-craning movements, performed by the male over the outstretched neck of the female. Also a mutual display leading to assumption of the Copulatory position by the female.

Turn-back-of-tne-head (Johnsgard)

Male swam or moved in front of female, conspicuously holding his head so that his nape and white crest stripes were directed to her. Sometimes given by the pair-drake during social courtship but more often immediately followed copulation. Performed 97

frequently with Wing-display (Johnsgard).

Head-nodding

Head and neck slowly swung forward and backward. Performed by both sexes while slowly surface feeding or changing locations. Performed by the female in pre-copulatory situations immediately prior to and linked to her initial Neck-crane movement before assuming the Prone copulatory position. Like Nod-swimming described for green-winged Teal (McKinney, 1965a) and similar to that described by Johnsgard for some Anas spp. in which the female swims among drakes jerkily nodding her head, illicitin.g male displays. Was not however, observed during Mandarin social display.

Chin-tickling (Savage)

Chin-tickling, described for Mandarins only by Savage, probably functions in pair-maintenance. Observed on only two occasions: once as pair returning to the nest box after the inattentive period and again following a display burst. Each time the pair swam together closely in a zig-zag course, the female leading with the drake turning in the opposite direction, her tail just brushing under his chin.

Wing-and-tail-flash (Johnsgard)

Indicated in Mandarins as similar but not homologous to the Head-up-tail-up of Anas spp. males (Johnsgard). Consists of a sudden upward jerk of the closed wing and tail; in Anas performed by courting males. I did not observe or recognize it for Mandarins.

Dip-and-point (Savage)

Observed on only two occasions by both individuals of an established pair while slowly surface feeding. As Savage noted, the individuals swam closely together but made movements independently. The head dipped as if in Mock-drinking, but then stretched stiffly parallel to the water, and maintained for up to 15 sec. The duration of the Point distinguished it from the female's linked Neck-stretch and Prone pre-copulatory behavior.

Bill-up

Presumably analogous to Chin-lifting described for the Wood 98

duck by Johnsgard and Eiders by McKinney (1961) and similar to Neck-stretching observed in Canvasbacks by Johnsgard. Neck stretched vertically and the bill tilted upward for several seconds. Because of the observation distance, any accompanying vocalizations not heard. "Jibbering and clucking" noises probably accompany the gestures as McKinney (1965c) noted 'chittering' accompanied this threat in Arias spp. Was associated with overt hostility and given by both males and females during nest box inspection and occasionally during the pre-basic molt period with the approach of other birds.

Bathe

Consists of repeated Head-dipping movements followed by Wing-flapping and Tail-shake. Performed by both males and females nearly always following copulation and agonistic encounters. In the former situation, males usually performed a Turn-back-of-tiie-head display prior to Bathing; females immediately began to Bathe. Both birds then briefly Preened; the female occasionally Bathed a second time. The number of Head-dipping and Wing-flapping movements was variable and apparently not related to the intensity of the previous situation.

Prone (Johnsgard)

Female flattened (sometimes gradually) along the water with head and neck outstretched. Position held during the entire male pre-copul atory display or assumed several times, presumably' illiciting continued male interest.

Inciting

Performed by a female standing next to the favored male or pair-drake. Involved highly ritualized, sideways, threatening movements in which the bill was lowered close to and moved along the ground or water and quickly raised. Repeated several times; often accompanied by an "Ack" vocalization. As in other waterfowl (McKinney, 1965c) directed at rejected males or intruders and indicated pair-bond formation.

Pursuit flights (Savage)

Infrequently observed by Savage during social display. A female suddently takes flight and is immediately followed around the lake by a group of males. The pinioned condition of my study flock precluded observance of such behavior. Attempted flight 99

observed on two occasions by males of a displaying party upon the introduction of a female. Although the head and bill positions were not noted, the birds rose nearly vertically, flew only a few feet, and fell into the water. May be analogous to the short Jump-flights described by McKinney for Teal (1965a), Eiders (1965d), and other waterfowl.

Mock-feeding

Called Lateral dabbling (McKinney, 1970) and Gabbling (Lorenz) in other waterfowl. McKinney (1970) observed performance by male Teal lateral to females during social display. In Mandarins, was associated with female Head-nodding, as the pre­ liminary activity for 11/13 pre-copulation sequences. With the male swimming closely behind in Full-sail posture she dipped bill into the water and made dabbling movements, which seemed slower but typical of normal surface feeding.

Head-shake

The bill pointed upward and shook. Derived from a movement apparently used to remove materials from the bill, functioning in many species as pre-flight signals (McKinney, 1965c). Common in pre-copulation sequences and courting parties of many waterfowl species but occurred only on encounters between Mandarin pairs during nest box inspection; the display being given by the resident birds (both sexes) and sometimes the intruder. On those occasions was linked to Bill-jerk and occasionally performed while facing away, indicating threat function. This agreed with Lorenz1 interpretation as a displacement activity performed in conflict situations preceeding flight.

Caressing movements

Performed by paired males standing or sitting with their female; an indication of pair-bond formation. Were "soft" Bill- dips by male close to and along his chest on the side of the female. Caressing movements sometimes given prior to a female's entering a nest box for incubation or inspection and upon her emergence.

Dashing-and-diving

Described by McKinney (1965c) for other waterfowl. Also called Playing-chase (Heinroth, 1910). Contagious behavior on both personally observed occasions. Occurred at mid-day and TOO

following evening feeding. Female of pair skittered across the water and dove; followed by the male. When surfaced, he Preened her and displayed (3 Drink-mock-preens, 3 Double-display-shakes and 1 Display-shake during three repetitions). Sequence ended with female Bathe and male again displaying (4 Drink-mock-preens, 3 Double-display-shakes, 2 Display-shakes and 1 Preen-behind-the- wing).

Savage saw variations of this by free-ranging flocks and by a pair in the London Zoo. In one case the dives were led by the male, but Savage did not mention such intensive display in either case. Its performance by paired birds and the associated display suggests its role in pair bond strengthening or maintenance.

Preening

Both sexes performed Dorsal, Belly and Chest, and Wing (both behind and outside) preen displays. Preens occurred in social display, or as brief isolated nibbling movements in conjunction with intensive body maintenance preening sessions. The latter situations seemed distinguishable only by their apparent greater frequency. Wings, belly, and shoulder region of the back received more than 85% of preening movements performed by pair-drakes standing with their females (five preening sessions); 60% of these male wing-preens to the side of the female. In contrast during three oil-preening sessions in which 12 different movements made, these same three areas received only 55% of 165 movements; 13% to the wing on the female's side. Additionally mutual preening evident, particularly by the male to the female's head and neck, and characteristic of pair formation.

Alert

The Alert posture, used by both sexes, but more conspicuous in the male because of his plumage characteristics. Neck stretched, crest and head erect and sails often erect. Often accompanied by Head-flicks with each sex's basic vocalization: "Pfrruib" (as many as 38/30 sec) by male, and "Ack" by females. Performance by a pair-drake while swimming with or to meet his female further suggested ritualization. 101

Vocalizations.- Mandarins vocalized infrequently, except during social display and when disturbed. Aside from the obvious calls, some of which have been recorded (Fig. 23 and 24), both sexes used quiet vocalizations. For example, females "hissed" when disturbed during late incubation, and males made soft whispery "ppffhhtt" noises to their females prior to her entering a nest box to inspect, lay, or incubate. I have either not heard or recognized several vocalizations attributed to Mandarins by Savage: the female's Brood and Exodus calls and the male's Alarm and Flight calls. All descriptive sounds except male "Jibbering" and female "Cluca-cluca-cluca" are those of either Lorenz or Savage (Tables XVIII and XIX).

The female's "Kett"(coquette call) was sharper arid slightly higher in pitch than her frequently repeated "Ack" vocalization.

Delacour (1964) stated that this single-note coquette call also was louder than the corresponding call in female Wood ducks. Her

"Clucking" was more rapid (as many as six/.5 sec) and slightly lower in frequency than the male's "Jibbering".

All male vocalizations except "Jibbering" were variations of a sharp, rising whistle accompanied by a lower snorting, nasal component.

The calls generally ranged between 85-6000 cycles, although a series of overtones extending to nearly 12000 cycles accompanied the Burp call. All were rapid, usually completed within .2 sec, and were repeated as often as two/sec ("Pfrruib") or four/.5 sec ("Jibbering").

There seemed to be slight individual variation in most calls relative to the presence of the lower nasal component. This individual variation was demonstrated by the difference between the first four 102

and last three "Clucking" calls, given by two different males.

Similar variation was present in the "Pfruuib" call sonograms of

different males and may have been related to the intensities of

the calls.

The lower frequency sound between the initial rising whistle

and secondary grunt in the Display-shake would seem to give credence

to Lorenz1 "F^W" whirring whistle description of this call. This

call was, however, recorded only once, and the intermediate sound was not clearly distinguishable on the sonograms. 10 3

Table XVIII. Description and use of female vocalizations.

Kett or Ke The coquette call (Lorenz; Savage). A loud, sharp, single cough-like note, given with a quick dip and thrust of the head forward and downward. Seemed to stimulate onset or increase intensity of social display also given by solitary females (e.g., those whose drakes were killed by predators), and following incubation inattentive periods. In the last situation it attracted the male, who responded with continu­ ous, whistling "Pfrruibs".

Ack. Plaintive call given almost continually while in flight (Savage). Observed in conjunction with Inciting and when disturbed (as often as 53 times per minute when flushed from a nest box) or nervous (prior to and after nest box inspec­ tion). May be analogous to the decrescendo call of Anas spp.

Cluca-cluca-cluca A soft, scarcely audible, repeated clucking-like vocalization (perhaps homologous to the soft, rapid "tetetetet" go-away call of the Wood duck). Always was given upon my entrance into the pens, by the female during male Bridling. Described as "twitter" by Savage, who indicated its use in pre-flight situations.

R-r-r-ruck The female brood call, described by Savage as "deliberately pronounced, but deeper than the normal voice." Although not heard by me, his discussion suggested its performance from the Alert posture. • 104

Table XIX. Descriptions and use of male vocalization

Pffrrruuiehb A long drawn-out call accompanying Burping (Lorenz).

Pfrruib The basic male call note, given with the hood ruffed (L’orenz). Performed prior to and during social display (as often as 50 times per minute) and while swimming to meet a female following her incubation break. Often has associated Head-flicks.

Gnk-zit A call similar in sound to a half repressed sneeze, given during Double-display-shake, and accompanied by a short sneezy whistle at the finish of the same display (Lorenz).

Uib The alarm call, uttered with the onset of flight (Savage).

Wriick A sharp whistle given during flight (Savage).

Jibbering Soft, repeated, twitter-like utterances, given when dis­ turbed in situations similar to the female's "cluca-cluca- cluca..." vocalization. May be the male's "go-away" call. Was performed during Bridling, while the bill rested on the chest.

FytyPfW A whirring whistle accompanying Display-shake with the "w" and "r" pronounced simultaneously (Lorenz). 105

Fig.is: A-C. Sequence illustrating Display-shake, performed on land. D. Male 1n Alert posture, from which the basic "pfrruib" vocalization is uttered. 106

Fig.19: A-E. Sequence illustrating Double-display-shake. F. Double-display-shake performed to female. 107

Fig.20: A. Male performing Mock-drink display. A and B. Sequence illustrating Drink-mock-preen display, C. Male performing Preen-behind-the-wing of Drink-mock-preen display. D. Male Preening female's head. 108

Fig. 21: A. Male Bridling. B. Bill-flick display (associated with Bridling). C. Male Burp display accompanied by female Inciting. D. Lateral view of male Burp display. 109

Fi g.22: A. Female in Prone copulation posture. B. Tum-back-of-the-head display performed on water following copulation. C. Turn-back-of-the-head display performed on land. D. Male in Full-sail posture. no

Fig. 23. Vocalizations of adult female Mandarin ducks.

i2ooo-

ACK ) s e l 85- c 2.0 y c ( q e r f

time (sec) m

Fig. 24. Vocalizations associated with the displays of adult male Mandarin ducks. ) s e l c y c ( q e r f

time (sec) 112

Social display

The presence of a female should greatly enhance male social

display, and result in pair formation. Differences in display

intensity relative to the age of the courting party males should

occur. To evaluate the hypothesis, two, 3-male goups (courting

parties) one of yearlings and one of adults, were formed. A male

and female from another (strange) population, a stuffed female,

and a female to whom one of the members of each courting party had.

previously associated (known female) was introduced into these male

groups several times during February and March for 30 minutes

(Table III). Ensuing behaviors were recorded on tape and film.

Introductions

Males when by themselves spent 80% of the time loafing and were

not distracted by vocalizations of displaying Mandarins in other pens.

Introductions of other birds into these 3-male parties resulted in

increased activity and agonistic encounters, but the displays were

low intensity except when known females were used (Table XX).

Agonistic behavior, pecks and threats, directed at the strange male upon his introduction by resident males decreased during the observation; period, since he stood or sat apart from the others. No vocalizations were given by resident males.

Introduction of the strange female resulted in Body-shakes and few Display-shakes by the males. This behavior, similar to that observed during the early introductions of the known female into yearling males, may eventually have led to pair formation had not pair -113

Table XX. Frequency of displays and agonistic encounters (pecks, chases, fights, threats), given by groups of 3 yearlings and 3 adult males alone and with the 30-min introduction of another bird.

no. agonistic .displays**. encounters experimental no...... no. no/1 min condition intro Dr Br B-s D-s D-d-s D-m-p Burp Total intro interval

ALONE Yearlings 4 4 4 1 1 Adults 4 1 1 8 3

STRANGE MALE Yearlings 2 15 15 22 8 Adults 2 18 18 16 10

STRANGE FEMALE Yearlings 2 26 2 28 26 8 Adults 2 19 19 25 17

STUFFED FEMALE Yearlings 2 14 14 40 4 Adults 2 19 19 5 10

KNOWN FEMALE Yearlings 6 1 15 18 13 2 6 33 88 25 12 % 1 17 20 15 2 8 38 100 Adults 6 2 21 1 9 33 26 14 % 6 64 3 27 100

*average frequency = total displays by all males/number of introduction

Dr = drink; Br = bridle; B-s = body-shake; D-s = display-shake; D-d-s = double-display shake; D-m-p = drink-mock-preen 114

formation begun earlier with the known female.

The stuffed female, in a semi-crouch posture known to illicit

male display when performed by live females, generally seemed ignored.

Although males performed Body-shakes, they made no movements toward

her.

Introduction of a known female resulted in social display and pair formation in yearling males and pair re-formation in the adults.

The differences between the two groups of males in numbers and types of displays performed were striking (Table XX). Yearlings performed significantly (P<.05) more displays than adults. In the two groups, only the number of Body-shakes was similar, perhaps expected since they often followed chases, which occurred with equal frequency in both age groups.

Intense adult social display with the introduction of the known female did not occur, perhaps for three reasons: 1) The female re-paired with her drake of the previous summer, reactivating an established pair bond which presumably was much stronger than any previous association between the yearling birds, 2) Pair re-formation was with the dominant male, perhaps inhibiting display by other males, and 3) Paired males generally did not take an active part in winter social displays.

Social display burst characteristics

During social display males continually adjusted their positions to each other and the female and performed bursts of synchronized, characteristic displays and vocalizations. Since adult males formed 115 pair bonds earlier than yearlings, the discussion of the female's behavior and the role of social display in pair formation will be based on observations of yearlings.

Penned yearlings.- Overt hostility (chasing and jabbing) occurred frequently during the introductions but was separate from the short, intense display bursts. During six minutes of an introduction in which the G MIO - W F22 pair bond strengthened, G MIO performed

1 Body-shake, 4 Display-shakes, 5 Double-display shakes, 1 Bathe,

5 Drink-mock-preeris, and 1 Preen-female: the other two males performed only 2 Head-jerks and 4 Burps. The only 2 chases occurred suggests a shift in motivation during courtship from agonistic to intense social display.

Short preening sessions, with certain movements ritualized as in other waterfowl (McKinney, 1965a, c, 1970), occurred during most introductions. Males Belly and Wing-preened 52% and 11% of the 54 preen movements, while females preened these areas 55% and 30% of

29 occasions. Preening occurred more frequently in later introductions but was not an integral part of intense social display. Its performance by the female somtimes led to directed courtship displaying by the male.

Free-ranging.- Display parties of feral Mandarins vary in size from small groups of three and four males (Kear, pers. comm.) to larger groups of 8-12 (Savage, 1952). I have observed free-ranging display parties of three birds (2 dcf, 1 ) to 16 birds (7 and 9 ^5).

All occurred in October, soon after assumption of the alternate plumage. In the largest party the males seemed to break into groups 116

of three or four and displayed around the females. The same displays

were performed in both situations and, except for the lack of

Bridling, were like the social display of the yearling males.

Similarly, McKinney (1965a) has observed Bridling in social display

of feral birds only when groups were close to shore and males could

move onto the shore. The low frequency of Body-shakes in these

free-ranging situations confirms its use primarily in low intensity

display situations. Unlike their non-involvement when in the over­

wintering pens, two paired males participated in this fall display.

Female's role

As found by Weidmanr. and Darley (1971a) for Mallards, most (58%)

adult male display occurred in the two introductions (No. 2 and 5)

in which the female Incited, chased, and vocalized (Table XXI).

Similarly W F22 was very active (Preening, Bathing, vocalizing, or

Inciting) during the previously mentioned 6.5 minutes of yearling male display. Yearling males were very aware of her presence. Some displays in these bursts obviously were directed to her and probably were stimulated by her activity, indicated by the males' deliberate positioning relative to her. (Such observations contrast to the contention of Lorenz that males display with little attention to the

female.) Others seemed spontaneous, occurring as a result of encounters between males during their continuous attempts to follow her. Social

display was never observed in the absence of a female, and role was

further accentuated by failure of the stuffed female to illicit

display. 117

Table XXI. Activities of a "known" female, and 3 adult and 3 yearling males in each of 6 x 30 minute introductions of this female during February and March 1973. introductions 1 2 3 4 5 6 total MALE ACTIVITY % time with + ** Adults M2* 8 50 35 22 44 52 35 Ml 14 5 59 8 10 1 15 M4 13 8 46 24 1 3 15 Yearlings MIO* 17 20 36 48 54 67 60 Ml 3 5 3 0 1 7 0 2 M14 10 5 19 0 10 0 7 no. of cf displays Adults M2 11 27 10 13 42 12 115 Ml 2 2 7 2 8 3 24 M4 8 12 9 3 23 3 58 total 21 41 26 18 73 18 197 Yearlings MIO 9 24 51 62 43 44 233 Ml 3 8 40 7 30 41 21 147 Ml 4 13 46 3 19 40 24 145 total 30 no 61 111 124 89 525 peck or preen + Adults M2 1 3 1 17 0 2 24 Ml 1 5 4 8 2 7 27 M4 6 4 17 14 0 3 44 Yearlings MIO 0 4 0 1 6 6 17 Ml 3 1 0 0 1 0 1 3 Ml 4 2 1 1 0 0 2 6 FEMALE ACTIVITY Adults Incite 0 7 0 0 20 8 35 Bathe 2 1 0 1 2 1 7 Chase 0 2 0 3 0 0 5 Vocalize 0 17 0 0 9 0 26 Preen males R M2 0 0 0 0 0 0 0 R Ml 0 0 0 0 1 0 1 R M4 0 1 0 0 1 0 2 total 0 1 0 0 2 0 3 Yearling Incite 0 6 0 7 16 8 47 Bathe 3 0 5 7 2 2 19 Chase 1 0 0 5 2 7 T5 Vocalize 0 30 17 16 25 12 100 Preen males G MIO 0 4 0 0 2 0 6 G M13 1 0 0 6 1 0 8 G M14 2 2 0 2 1 0 7 total 3 6 0 8 4 0 21 indicates 'pairing male ** average percent time for all introductions 118

Vocalizations.- Female vocalizations (80% Ketts, 20% Ack) were

conspicuous parts of many yearling display bursts. Yearling males

displayed in 85% of the 69 one-min intervals in which she vocalized

but in only two-thirds of the 111 intervals in which she was silent.

While vocalizing did not always initiate intensive display, it

(particularly Ack) usually resulted in the males assuming Full-sail posture and intensified display already in progress.

Head-nodding.- These movements, considered highly stimulating to male Mallards (Weidmann and Darley, 1971a) were performed on three occassions only by the known female when introduced into the adult males. The low frequency of these movements was confirmed by obser­ vations of social display of free-ranging birds on 24 October 1973.

While two males displayed around a female, she only vocalized

{Kett), Neck-craned, and assumed a semi-Prone position.

Inciting.- The relationship of female Inciting to social display seemed inconsistent and was difficult to analyze, because only parts of each introduction were filmed. Additionally, Inciting occurred in the same one-min intervals with female vocalizations, chases, and

Preens.

Inciting seemed to both illicit social display and result from it.

Yearling males displayed in 80% of the 37 intervals in which the female

Incited, but also in half of the 143 intervals in which she did not

Incite. It illicited male Full-sail posture and vocalizations

(Pfrruib), most frequently by the preferred male. Female Inciting was directly tied to pair formation, her choice of males evident from her

Inciting a preferred male against other males. Inciting seemed to 119 link directed social display and pair formation. For example,

Inciting resulted in female-directed displays by G MIO who performed

Preens, Drinks, and Shakes in 85% of the Inciting intervals and

Preen-female or Drink-mock-preen following 10 of 18 Inciting situations.

Likewise, pair maintenance activities, such as Mutual Preening and

Caressing movements, and simultaneously Bathing and Wing-flapping occurred during Inciting intervals.

Mandarin females through their movements, postures, and vocalizations both initiated and intensified male social display.

Inciting and vocalizing probably were the most influential behaviors.

Relationship to pair formation

Yearling pairing.- Social display resulted in pair formation.

In the yearling pairing process, the female indicated her preference, but the male had final choice. During the first introduction, she

(W F22) moved among the males, randomly threatening and pecking them, with intra-male threat displays and chasing resulting. The males did not chase her until a pair bond formed but did peck (sometimes very hard) and preen her at this time. Twice, she briefly Preened

G M14, the male with whom she had associated prior to the experiment.

She presumably was rejected, however, since he threatened, pecked, and spent little time with her.

During the second introduction she Preened G MIO four times and

Incited against the other males when he was chased. G M14 displayed frequently during this introduction, which seemed only to stimulate female Inciting against him. G MIO returned to her after being chased, 120

a position from which he Burped and Bridled. He Preened her neck,

spent more time with her, and performed Display-shakes, some directed

at her. In subsequent introductions she Incited, Vocalized, and

Bathed, resulting in social display, and directed courtship by

G MIO. The pair bond strengthened as they spent increasingly more

time together and performed Mutual Preening and Caressing movements

(Table XXI).

Throughout the yearling introductions, G M14 was dominant and

G MIO subordinate. As the G MIO - W F22 pair bond formed, the kinds of agonistic behaviors directed by G Ml4 at G MIO changed from threats and pecks to chases. Pairing modified subordination.

G MIO's reactions changed from retreating to displays typical of a paired male (Burping, Bridling, and Bill-jerking, usually next to and supported by his female's Inciting), and occasional chasing

(sometimes directed at G Ml4 but usually re-directed at male 13).

G MIO also performed most of the directed displays (Fig. 25).

Alternate pairing.- Pairing also occurred without social display from direct attempts by each sex. This was most prevalent with unpaired yearlings during late spring when most adults have already paired, and during late summer when pairs re-formed. In one case during early March, an unpaired female (W F21) chose a paired male

(R M3) and for one week constantly followed him, Incited at other males, and chased his female. The male initially responded by occasionally standing with her but later started to Burp and Bridle during her Incitings. No pair bond formed, however, as he spent more time with his own female and nearly always displayed when she chased 121

YEARLING MALES ADULT MALES CJgio E23gi3 CZ3 R2 EZ3 R I

Fig. 25. Displays performed by yearling and adult male Mandarins during 6 Introductions of a known female to each 3-male group. 122 or Incited at W F21 (Table XXII). After an unsuccessful week of attempting to pair with R M3, W F21 then paired with G M14 by 22 March.

This male had been returned from the 3-mals courting parties to the flock on 8 March, which was during the initial week in which W F21 followed R M3. During a 15 min observation in which he never left the female, he (G M14) Preened her twice and displays on all 11 occasions she Incited or vocalized..

In another example, during late March, a male continuously followed a female, who initially Incited at and chased him. This - activity resulted in agonistic encounters but little social display, presumably because three of the five males in the pens had paired. Display-shakes again were the first female-directed display.

The pair bond was obvious after about a week when the female Incited next to the male and he Preened her 15-20 sec on each of six occasions during a two hour observation.

These data suggest that the time of year plays an important role in both the method and length of time necessary for pair bonds to form. Pair bonds forming late in the spring, presumably prior to the break-up of the flock in wild birds, but in the presence of paired males, and hence less potential social display, resulted from directed courtship and formed in a week. Pairs formed in late fall or early winter involved social and directed courtship, some changing of partners, and took several weeks. Therefore, social display facilitated Mandarin pair formation and, if participated in by paired males, probably resulted in a stronger, more stable bond. 123

Table XXII. Frequency of responses by R M3 (cf) to Inciting by his pair- female (B F16) and to a female (W F21) attempting to pair with him. Time (min:sec) spent with each female is indica­ ted. Data are based on 3 x 15 min observations in March 1973.

Frequency Bird No. activity/display 15 March 16 March 20 March

W F21: Inci te 10 8 3

R M3: Full-sail 6 4 1 Burp 0 1 1 Preen $ 1 0 0 No display 3 . 4 2

B F16: Incite •5 7 5

R M3: Full-sail 5 7 5 Burp 5 3 1 Preen 2 0 1 1 No display 0 0 0

R M3: time w/B FI 6 10:00 9:45 8:00 W/W F21 3:15 3:00 ' 4:30 alone or between 2? 1:45 2:15 3:30 124

Summary of social display

Introduction of a "known" female to each of two (yearling and adult) 3-male "courting parties" resulted in intensive yearling social display and pair formation, and minimal adult social display and re-pairing. Social display was absent in each courting party during the introductions of a strange male and both a strange and stuffed female.

Social display consisted of short bursts of rapid, synchronized displays, many directed at the female. Social display was never observed in the absence of females, who through their movements, postures, and vocalization both initiated and intensified social display.

The season of year was important in both the method and length of time necessary for occurrence of pair formation. Fall or winter pairing took several weeks and involved social display, directed courtship, and changing of partners. Late spring pairing (in the absence of social display) resulted from direct attempts by either bird with bonds formed in a week. Females indicated their preference for a particular male by Inciting movements directed at other birds; an "accepting" male continually maintained his position next to the female and reacted by assuming the Full-sail posture, Bridling and

Burping. Mutual Preening, Drinking, and Caressing movements indicated pair bond completion.

Copulation sequences

The occurrence of displays associated with all observed copulation 125

sequences of both free-ranging and penned pairs were recorded.

Descriptions and analysis of these behaviors were based on the age

of the males involved, since yearling females were involved in six

of the seven quantified adult copulatory sequences, and in all six

of the yearling sequences. The relative frequency of each display

movement has been calculated by totaling the counts of each movement

from all sequences then comparing the frequency of each movement per

sequence (Table XXIII). The percentage of sequences in which a

particular movement was observed also has been calculated (Table

XXIV); and the order of movements has been diagrammed (Fig. 25).

Sequences have been designated as complete or incomplete on the

basis of mounting.

Initiation

Pre-copulatory behavior nearly always was initiated by the

female, occurring most often (11/13 times) while a pair was slowly

swimming, the female Head-nodding and apparently Mock-feeding, and

the male swimming closely behind in Full-sail posture. The female would stop (in one case directly under the male's chin) and perform a linked Neck-crane then Prone sequence.

One of the two pre-copulation sequences not started by the female began when the male swam in front of the female and began Neck- craning. This resulted in mutual Neck-craning and assumption of the

Prone position by the female. The other sequence began when the male, who was in front of the female who was preening on land, suddenly swam toward her. She almost immediately entered the water and went Prone. 126

Table XXIII. Display patterns in complete and incomplete pre-copulation sequences of yearling and adult Mandarins.

COMPLETE SEQ INCOMPLETE SEQ adult yearlings adults & yearlings (n=7, 4 pr.) (n-6, 3 pr.) (n=5) male female male female male female fr. % fr. 0io/ fr. % fr. % fr. otof fr. %

Bill-dip 28 39 0 - 21 23 9 19 20 41 3 12

Bi 11-dip over $ 9 11 0 - 6 7 0 - 0 - 0 -

Neck-crane 31 44 29 73 50 56 23 49 23 48 16 64

Preen other 3 4 0 - 9 10 0 - 3 6 0 -

Preen self 1 1 0 - 3 3 0 - 0 - 0 -

Head flick 0 - 0 - 0 - 0 - 3 6 0 -

Prone 0 - 11 27 0 - 15 32 0 - 6 24

Total 72 40 89 47 49 25

Avg displays/seq 10 6 15 8 10 5 fr. = total of the movements of each type in each sequence.

I = number of movements of each type/total number of movements in all sequences for each classification. n = number of sequences 127

Table XXIV.Occurrence of movements in the pre-and post-copulation behavioral sequences of yearling and adult Mandarins. Frequency is the percent of sequences in which a movement occurred and not. its frequency of occurrence in each sequence as presented in Table XXIII.

Adults Yearlings n=14 ri=10

Pre-copulation sequences

start, both in water 100 8Q female Bi 11-dips 0 20 male Bi 11-dips 64 90 female Neck-crane 64 70 male Neck-crane 71 60 male Bill-dip-over-female 36 60 male Preen female 29 50 male Drink 10 7 revolve 50 80

•Post-copulation sequences n=7 n=6

male rapid dismount 71 50 male Turn-back-of-the-head 100 83 male Drink 14 33 male Bill-dip 0 17 male Display-shake 14 0 male Bathe 57 83 female Bathe 70 100 female Preen-wing to male 0 33 both leave water and Preen 70 67 both continue swimming 43 33 128

Fig. 26. Order of movements in 13 Mandarin pre-copulation sequences. Combined, adult, and yearling sequences are illustrated. Figures should be entered from rectangular blocks and arrows followed until Mounting. Solid arrows indicate frequent exchange between display. N = number of sequences.

TOTAL (13) Preen

ADULTS (N = 7) YEARLINGS (N= 6) 129

Social display stimulated copulatory behavior only once; following assumption of the alternate plumage when the birds were still very social. During an intensive display burst involving

16 birds (7 and 9??) a pair from the previous spring slowly moved to the fringe of the displaying group and performed pre­ copulation displays. Although the male had participated in the display burst, no typical pre-copulatory displays were given by either individual prior to their movement to the periphery, the female's assumption of the Copulatory posture, and mounting.

The sight of one pair copulating at times stimulated the same behavior in other pairs. Consequently, on one occasion three penned pairs copulated within 45 minutes.

General copulation patterns

Copulatory behavior was most frequent during the pre-laying and laying periods of spring. Its occurrence along quiet shorelines and under overhanging vegetation and the unsociability of pairs during this time often precluded observing the initial movements. Complete sequences were seen and quantified on only 13 occasions during the study. Movements were recorded but not quantified for 24 additional copulations. Except for two rapes, copulation always occurred between established pairs.

Copulation always occurred in the water. The most obvious

(but not the first movement) was the female's adoption of the Prone position. The male then slowly swam next to or around her, making

Bill-dipping and or Neck-craning movements and occasionally Preening 130 her neck and head. He then would nudge her with his chest; both slowly revolved in a tight circle just prior to his mounting and holding onto her head. The female was almost entirely submerged during copulation. After dismounting males usually performed a

Turn-back-of-the-head display, then Bathed and Wing-flapped; the female usually immediately began Bathing. Both birds then either went to the shore and Preened or swam away.

Pre-copulation display analysis

Complete sequences.- It is evident from these data that the pre-copulation sequences of adults were more stereotyped (i.e., less variable) than those of yearlings (Table XXIII; Fig. 27). Although waterfowl displays are innate (Johnsgard, 1968), they seemed to be modifiable with experience. The same movements generally occurred in the sequences of males of each age, but movements were more frequent in yearling sequences. The duration of the pre-copulation sequence (Table XXV) and the average number of movements performed in each sequence by yearling males (14.5) and females (7.8) were greater than that performed by adult males (9.6) and females (5.7)

(Table XXIII). This is particularly evident for female Heck-crane and Preen. Yearling females also Bill-dipped and assumed the

Copulation posture more frequently than adult females. Although

Mandarin females generally were passive prior to male mounting, this was not always the situation.

Further evidence of the variation between the sequences of different aged males and/or the variability in yearling male copulation 131

Fig. 27. Relative frequency of movements in adult and yearling Mandarin male pre-copulation sequences. Based on data of Table XX. N = number of sequences.

B-d = Bill-dip; B-d-g = Bill-dip-over- ¡¡J-neck; N-c = Neck-crane; H-f = Head-flick;

% 132

Table XXV.Average duration (and range) of each phase in complete reproductive sequences of adult and yearling Mandarins. N indicates number of sequences observed.

Adults Yearlings stage n time(sec) p* n time(sec) pre-copulation 6 30 (15-55) NS 5 37 (15-60)** mount 7 10 ( 5-15) .01 7 20 ( 5-35) post-copulation 7 15 (10-20) .05 6 35 (15-90)

*duration of pre- and post-copulation sequences,analyzed by a Mann- Whitney-U test; mounting duration by Student's t-test.

one abnormally long sequence of 180 seconds not included 133

movements becomes evident from the distribution of movements in

sequences (Fig. 28). The data for this table were determined by calculating the relative frequency of each display in each one-

third of all sequences. While only the last one-third of the adult sequences contained a greater variety of movements than the yearling sequences, the last movement prior to mounting still was slightly more stereotyped in adults (Bill-dip 67%, Preen female 17%, Bi 11- dip-over-female 17%) than yearlings (Bill-dip 50%, Neck-crane 17%,

Preen female 17%, Bi 11-dip-over-female 17%)

McKinney (1961) found the last movements of 32 European Eider

(Somateria mollisima) male pre-copulatory sequences nearly equally divided among behaviors with tendencies to behave sexually toward, to attack, and to escape from the female. In 13 completed Mandarin copulation movements [except perhaps Neck-craning which McKinney

(1965c) suggests may be derived from forward swimming movements] were sexual in tendency. However, Neck-cranes were the only male movement in two pre-copulation sequences. The movement immediately preceding mount always as sexual (Bill-dip, seven times; Bill-dip over female, twice; Preen female, twice; Neck-craning, once). That

Bill-dip occurred immediately prior to mounting in 9/12 cases may indicate ritualization, although perhaps not as much as the "Mock preening Complex" in Goldeneyes (Dane, Walcott and Drury, 1959) or the Shake in Steller Eiders (McKinney, 1965c), both of which nearly always immediately precede mounting.

Only in situations in which pairs actually were interrupted during pre-copulation and only after pre-flight Chin-lifting and female 134

Fig. 28. Distribution of pre-copulatory movements from 13 sequences. Percents calculated by determining the relative frequency of each display in each one-third of all sequences.

PREEN SELF

PREEN

BILL-DIP- OVER- j

BILL-DIP

NECK-CRANE

ADULTS YEARLINGS (N= 7) (N= 7) 135

Inciting were Mandarin agonistic behaviors associated with repro­

duction. These sequences did not continue once interrupted. e

Such differences in pre-copulatory movements between European

Eiders and Mandarins may be due both to the female's role in pre­ copulation and the social situations under which most of the copulations are performed. While female Eiders are known to attack sub-adult males after unsuccessful copulations (McKinney, 1961),

this was not observed in Mandarins.

European Eider copulations occurred following disputes, frequently among birds in diving flicks. Their pre-copulatory displays averaged

27 movements/sequencs and took an average of two minutes (range 0.5-

5.5 min). The presence of the flock presumably reduced predation susceptibility allowing such long pre-copulatory sequences, this theory of flock protection is further demonstrated by the rapid, inconspicuous pre-copulatory behavior of Steller Eiders (Polysticta stelleri), performed several hundred yards from the flock. Steller

Eiders are highly susceptible to agle predation (McKinney, 1965c).

Mandarin pairs, except for those in pens, usually isolated themselves during copulations. This isolation was verified by

Savage (1952), who despite watching as many as 84 birds engaged in social display at one time, apparently observed copulation only once and that was for an isolated pair. The elaborate coloration, small size, and open (surface and grazing) feeding habits of Mandarins made them susceptible to predation by raptors; presumably incon­ spicuous and short pre-copulation sequences are advantageous.

Savage (1952) indicated heavy predation by Black-eared Kites on Asian 136

Mandarins, and many of my birds were either "missing" or caught by

raptors.

With less chance of conspecific interruption, minimal conflict

in tendencies in Mandarins might be expected. This might account

for their relatively short sequences, particularly those of only

10-15 sec, consisting only of a few displays. As in most waterfowl,

the Mandarin female presumably is physiologically ready with assumption

of the Prone position; and, as suggested by McKinney (1961) for

European Eiders, it is probably that the continuous sequence of male

behavior and not each particular display has meaning for the female.

Since assumption of the Prone posture is one of the first

movements of female Mandarins, the duration of the sequence may

be related mainly to "self-stimulation" efforts by the male necessary

to overcoming any inhibitions preventing his immediate mounting of

the female. In the situations in which females perform multiple

Prone or mutual Neck-cranes, the duration also may be necessary for synchronization of reproduction physiology.

Incomplete sequences.- The small number of incomplete sequences observed and recorded, necessitated pooling of the data. This has made analysis difficult because of variations in duration (10-130 sec),

involvement of different males and causes for terminating the sequence. The reduced occurrence of some movements and increase in others (e.g., Neck-crane) may therefore only have been related to the shorter time period. However, it is evident (Fig. 27) that fewer male Bill-dips and no Bill-dips-over-female occurred, perhaps suggesting their necessity (in some cases at least) in overcoming 137 mounting hesitancy in males. Head-flicks, characteristic of pre­ flight in many waterfowl and frequently observed in long pre-copulation sequences of Red-crested Pochards (Netta rufina) (McKinney, 1965c),, were present and perhaps indicated fleeing tendencies.

McKinney (1961) observed an increase of Preening in low intensity

Eider male pre-copulation behavior and suggested its occurrence under weak sex tendencies. This could not be substantiated in Mandarins, since Preen did not differ in frequency between complete and incomplete sequences.

The approach of other birds resulted in incomplete sequences and induced Shakes, Drink-mock-preen, and Burp displays by the resident male that seemed to indicate conflicting tendencies. These displays, not part of normal pre-copulation sequences, were performed only after the pre-copulation sequence had stopped. Some accompanied female Inciting and others were directed to the female as the resident pair swam off.

Changes with spring.- Copulations of three pairs of birds of both age groups were observed twice. The occurrence and frequency of movements and the duration of each sequence relative to the age of the males was analyzed (Table XXVI). Despite the small sample size, some trends seem evident. The total number of movements per sequence remained almost identical in adults (32 and 35) but nearly doubled

(32 and 59) in yearlings, yet the duration of each sequence in both age groups was nearly the same.

The only second copulation appreciably longer than an initial one occurred between yearlings on 31 May and was of low male intensity. Table XXVI.Percent occurrence of each pre-copulatory display (movement) of three yearling and three adult males in successive sequences, spring 1973.

Adult males Yearling males Avg. Avg. pair• 1 pair■ 2 pair■ 3 pai r 1 pai r • 2 pai r 3 total total 12 31 6 3 21 11 6 31 3 28 11 13 Ma Ma Ap Ma Ap Ma 1st 2nd Ap Ma Ap Ma Ap Ap 1st 2nd

Neck-crane 56 11 75 67 47 44 53 37 0 14 38 87 100 100 44 61

Bi 11-dips 11 67 25 33 32 39 25 46 45 41 44 3 34 17

Bi 11-dips over $ 33 0 11 17 16 11 11 9 3 8

Preen ? 0 22 5 0 3 6 33 27 19 10 19 11

Preen self 5 0 3 0 11 9 0 3

Total no. displays 9 9 4 3 19 23 32 35 9 22 16 33 8 4 32 59

Duration (min:sec) :15 :35 :15 :15 :55 :45 :40 3:00* :55 :50 :20 :15

*long duration resulted from low intensity male behavior

cCoO 139

In conjunction with female Neck-cranes and repeated assumption of

the Prone posture (probably an attempt to re-stimulate male interest),

this male made 14 more presumably self-stimulating movements than in

his first copulation, including Preen-female. The female's movements

resulted in the return of the male, who on several occassions had

begun to swim away. In conjunction with his disinterest, the

increased duration of this second copulation can be explained in

part by the increased number of displays, female in particular. This

sex distinction seems important, since the number of displays

performed by the male of pair 2 doubled, yet the time slightly

decreased.

A decrease in Bi 11-dips and increase in Neck-cranes of yearling

males (opposite that of adults) was evident, perhaps an indication

that the necessary stimulation (or maintenance of high exciting levels)

for mounting in late spring sequences of yearlings was more difficult

than in adults.

Mounting

Mounting time was significantly (P^.Ol) longer in yearlings than in adults, averaging 20 and 7 seconds, respectively (Table XXV).

Males of both ages dismounted immediately following copulation, sc that the extended yearling mounting time was prior to cloacal contact and perhaps related to inexperience. Two different yearling males made multiple attempts at copulation. One male remained mounted during three attempts at cloacal contact. Another did not hold onto the female's head and slipped off, but immediately remounted, grabbed 140

her head, and apparently successfully copulated.

Males mounted the female on her left side on 16 of 19 recorded

observations and most often copulated by passing their tails under

hers, also from the left side. McKinney (1965c) noted left-sided

copulation in 39/40 instances, among 24 species of Anatidae, and

explained it by Hochbaum's (1942) observation of the penis being

on the left side of the cloaca in the Anatidae.

The effect of experience on successful copulation is evident from

the changes in mounting duration on successive attempts, averaging

8 and 7 sec, and 28 and 15 sec, respectively, for adults and yearlings. This 50% reduction in yearling mounting duration probably resulted from no males slipping; all held the female's head.

Post-copulation

Much age variation also existed in post-copulation movements, adult behavior being more stereotyped than that of yearlings (Fig. 29).

Adult males rapidly dismounted on five of seven occasions, always performed the Turn-back-of-the-head as the next movement then

Bathed-Wing-flapped (usually). However, Display-shakes, often associated with agonistic encounters and social display, were performed with Turn-back-of-the-head, and may, as McKinney (1961) suggested for

European Eiders have indicated strong tendencies to flee.

Yearlings dismounted slowly as often as they did rapidly and performed Turn-back-of-the-head as their first movement only half of the time. Despite the variation in the sequence to this point, only one pair did not leave the water to Preen. An attempt to explain 141

Fig. 29. Order of movements in 14 Mandarin post-copulation sequences. Combined, adult, and yearling sequences are illustrated. Figures should be entered from dismount and followed to out and Preen, or swim away. N = number of sequences.

TOTAL (14) Head-flick dismount

out and Preen

ADULTS (6) YEARLINGS (8) Keazd-flick

T-b-h (Turn-back-of-head), D-s (Display-shake) 142

the variation in dismounting speed as a function of the intensity

of copulation (based on the frequency of Bathing) was unsuccessful.

Bathing and the number of Head-bobs per Bathe, occurred as frequently with rapid as with slow dismount.

In 12 of 14 post-copulation sequences both adult and yearling

females Bathed txhen Wing-flapped, usually immediately after the male dismounted. Females occasionally Bathed a second time, often

simultaneously with the male's first Bathe. The entire post­ copulation sequence averaged 15 sec for adults and 45 sec for yearlings.

The significantly (P< .05) greater duration in yearling sequences probably has related to their'slower dismounting rather than their greater number of movements (average - yearling males 5 and females

3; adult males 4 and females 2). Multiple Bathings were frequent.

Either individual of yearling pairs Bathed more than once in half of the sequences with one yearling male Bathing three times in one sequence. An adult Bathed twice only once. It is therefore, apparent that post-copulation patterns, like those of pre-copulation, are more variable in yearlings than adults, but presumably become more stereotyped with age and experience.

Rape

Rape was observed on two occasions, (both on 13 April 1973), and involved the same female (F 0A35), who at the time was unpaired but associating with a pair. As Lebret (1961) found for Mallard trios, the pair bond between the male Mandarin and the second female of a trio was not as strong as the bond with the preferred female. The 143

drakes of both species always followed their original preferred female.

F 0A35 was raped by a paired drake and by a solitary drake while walking or swimming apart from her pair past these males.

On one occasion she was continuously vocalizing (Kett and Ack) and suddenly dashed ahead of her related pair, which was being slowly pursued by another pair. On the other occasion she was chased from the bank into the water and pursued. As with Dashing-and-Diving behavior, the pursuing male immediately dove and chased her.

After he surfaced and swam toward her, they both dove, and resurfaced mounted. There were no pre-copulatory displays by either sex in either situation. In one case the only post-copulation displays were female Bathe Wing-flap. While in the other rape, the male performed Bathe Wing-flaps, Double-display-shake, and left the water and Preened.

No records were kept of the number of times unpaired or isolated females swam or walked past males and were not raped. However, these were the only observed times a female Dashed-and-dove while being pursued and significantly, the only times rape resulted from an agonistic encounter. Male Dashing-and-diving in many waterfowl species normally occurs with agonistic encounters during social display or in some Aytha between individuals of a pair during strictly agonistic encounters with another pair (McKinney, 1961). McKinney further states that the presence of a female in agonistic situations may stimulate sexual tendencies, with copulations often immediately following. Therefore the association of Dashing-and-diving with 144

rape probably further demonstrated the close relationship between sexual and agonistic tendencies.

F 0A35 had not been observed copulating with the male of her trio, yet she began laying fertile eggs soon after being raped.

Infrequent rape seems advantageous in such cases.

Summary of copulatory behavior

Copulations were most frequent during the pre-laying and laying periods, in quiet areas along the shoreline. They always occurred in the water and appeared initiated by female Head-nodding, and

Mock-feeding. They also resulted from social display and were sometimes contagious.

Pre-copulation sequences generally were short (10-15 sec) and consisted of the female assuming a Prone position and male Neck- craning, Bill-dipping, and Preening the female. Short sequences may have been related to the Mandarin's predator susceptibility.

Adult copulation sequences were more stereotyped than yearling sequences. The duration of pre and post copulation sequences and mounting was longer and contained more displays in yearlings than adults. Likewise the efficiency of yearling sequences increased with experience; the mounting time of second copulations was one- half that of the first. Both data suggests that reproductive sequences are modifiable with experience.

Pre-copulation movements were nearly entirely sexual in tendency.

Bi 11-dipping may be ritualized as the final movement prior to mounting. Post-copulation consisted of movements with escape 145 tendencies.

The duration of incomplete sequences, terminated by the approach of other birds and male disinterest, ranged from 10-130 sec and consisted of movements with conflicting tendencies to leave and behave sexually. Rape was infrequent. Only a few post-copulation displays were performed after rape.

Display motivation

There are three major types of sexual displays in Mandarins:

1) pair-formation or courtship (social) displays; 2) pair-maintenance displays; and 3) copulatory displays. Pair formation displays were the most elaborate (i.e., Double-display-shake and Drink-mock-preen) and presumably (Johnsgard, 1968) the most species specific; being necessary to prevent mismating. Pair-bond maintenance and copulation displays were less elaborate and as in other waterfowl (Johnsgard,

1968) involved mutual responses such as Preening, Drinking, and

Caressing movements. Additionally, Mandarin pair-bonds appeared strengthened by continual close contact of the individuals throughout the year, and displays associated with frequent copulation and social encounters.

Analysis of the major Mandarin displays associated with sexual and social encounters provided clues to the underlying motivation and predominant tendencies in each situation (see Table XXVII ).

Agonistic tendencies (indicated by Bridle, Bill-jerk, and probably

Burp) dominated over escape and agonistic (pre-flight Head-rattle,

Bridle, and linked Bill-jerk) during pursuits and territorial Table Frequency of display of penned and free-ranging paired-males in social encounters. Social displays of XXVII. free-ranging males (pair status uncertain) also indicated, n = number of observations.

free-ranqing flock over-wintering, penned flock during territoriality -copulation social pair - 2 following other pursuits interrupted display I nei ting chases males *** pursuers* residents* intruders** residents** chaser chased (n=3) (n=6) (n-5) (n=3) (n=4) (n=44) (n=5) (n=3) f % f % f % f % f % f % f % f % f % Body-shake 1 3 1 1 1 2 Display-shake 3 14 3 8 5 14 3 10 11 9 1 2 1 8 1 14 4 17 Double-dis-shake! 4 18 8 23 8 22 4 13 22 19 2 29 1 4 Burp 5 23 13 37 2 5 3 10 24 21 28 52 2 17 3 13 Drink 1 3 1 3 2 2 1 8 1 14 Dr-mock-Preen 1 4 7 20 17 47 12 40 42 36 6 50 11 48 Preen self 1 3 11 3 2 7 Preen female 3 6 2 9 Bi 11-jerk 7 32 1 3 5 9 1 8 1 14 Bridle 1 4 1 1 11 20 Turn-bk-head 3 10 3 3 1 14 2 9 Wing-flap 5 4 1 8 1 14 Head-dip 2 5 Head-flick 1 4 1 3 1 3 Full-seil(only) 5 9 Bathe 1 3 6 5 Total 22 35 36 30 117 54 12 7 23 $ Incite X X X X X V Vocalize X X X Chases X X X X *displays performed during the encounter. x = occurrence of a par- **displays performed following the approach of the other pair, usually while swimming away. ticular. activity. ***male in another pen or not directly associated with the social group. 147

encounters. Conflicting tendencies during social display, resulting

from hostile encounters between males attempting to stay near the

female, seem evident. Shakes (Body-shake, Display-shake, Double-

display-shake), often directed to the female, being of relatively

equal (21-36%) frequency in all situations except the purely

agonistic (Chasing), suggested a slight predominance of agonistic

over escape and sexual tendencies in social display.

Drink, Preen, and Drink-mock-preen, occurred in social display

situations as directed courtship, and were typical of pairing

situations, probably indicating a predominance of sexual tendencies.

This relationship is further suggested from their use by males

following interruption of pre-copulation sequences or when not

directly involved in agonistic encounters. While I did not observe

copulation by a retreating pair after such encounters, as has

McKinney (1961), on a few occasions the female did assume a semi-

Prone posture.

The frequent linking of Drink-mock-preen and Double-grurit- whistle, particularly during Dash-and-diving (a display with apparent chasing and sexual motivation), provides additional evidence for

Moyhnihan's (1955) theory that the balance of tendencies is dynamic

and can frequently change during most situations.

Pre-copulation displays seemed to result primarily from purely

sexual motivation, while escape tendencies predominated during post­

copulation. There was an increase in the strength of sexual

tendencies with the pre-copulation sequence as indicated by the

shift from Neck-cranes to Bi 11-dips, probably evolved from Drinking 148

and characteristic of pre-copulation in many ducks.

That Mandarins engage in pre-copulation displays largely distinct

from those of social encounters, suggests strong sexual motivation.

This segregation of behavioral responses in turn result iri a

shorter copulation sequence. In contrast Common Eiders perform

the same displays (with varying frequencies) before and after

copulation, during agonistic encounters, and in social display

(McKinney, 1961).

SUMMARY

A study of pinioned, free-ranging Mandarins (Aix galericulata) was conducted from 1971-1974 at the 25-acre estate of Mr. J. J. Schedel

Elmore, Ohio. The purposes of the study were to 1) document the

Mandarin's breeding biology and behaviors, nesting phenology, and.

time budget; 2) describe in detail the displays and vocalizations associated with copulation patterns, pair formation and maintenance, and social encounters; and 3) determine the female's role in male social display and subsequent pair formation.

Mandarins were observed for periods of one to three hours each spring nearly every other day; less often during the other seasons.

Several full-day (dawn-dusk) and all-night observations were made, and nest boxes were checked periodically to determine laying sequences.

Eggs were weighed and measured, and shell thickness also was recorded. •

Female incubation temperatures and daily incubation rhythms were determined.

Morphological and behavioral development, and feathering patterns 149

of ducklings were recorded weekly for 10 weeks. Additionally, the

eye color, secondary feather pigmentation, and primary feather

lengths of Mandarins from several populations were evaluated as

ageing characters.

The female's role in social display was analyzed through a

series of introductions of males and females to yearling and adult

male "display or courting parties," with ensuing displays and

vocalizations recorded on film and tape. Similarly, the frequency

of displays associated with copulation behavior, social encounters,

and pair formation were noted and their use in these situations

analyzed.

Mandarins formed strong seasonal pair bonds, which re-formed

successive years if both individuals lived. Rape and promiscuousness

were rare. Copulation behavior and nest box inspection were frequent,

with both usually occurring in the morning. Females, attended by

their pair-drake, inspected several potential nesting sites prior

to laying, with elevated pole boxes in undisturbed areas preferred.

Most clutches (avg 9.5 eggs) were begun between late April - mid May

and were initiated by yearling females earlier and with less

fertility (78%) than adults (90%).

Females began incubating (lasting 28-30 days) usually the afternoon the last egg was laid and generally took two inattentive periods daily (pre-dawn and pre-dusk). Males attended and defended

(usually) the female throughout the spring.

Mandarin eggs averaged 52.85 x 38.64 mm, with adult eggs larger 150

than yearling eggs. Egg weight decreased 10.5% during development.

The average shell thickness was 0.259 mm, but variable both within

the same egg and among eggs in a clutch.

Duckling development was rapid, with sexual dimorphism evident,

and fledging and adult size nearly attained by seven weeks. Most

adult comfort movements were performed within the first two days

after hatching. Ducklings apparently molted directly from the juvenal

to alternate plumage. Adult and yearlings of both sexes could be

separated on the basis of primary feather length, females on the

basis of secondary feather pigmentation, but neither sex on eye

coloration.

Mandarin daily activity patterns consisted of repetitious

feeding (as much as 8 h/day), preening (greatest during the pre-

basic and pre-alternate molt periods), and loafing. Variations in

the amount of each activity were related to the different social

periods. Activity was continuous throughout the day during the pre­

laying and laying periods, but diurnal activity rhythms were obvious

during the summer and fall (feeding in morning and evening and

loafing during mid-day). Mandarins utilized a variety of feeding methods, with the general patterns changing during the day, and during

the season.

During incubation, females were on the nest 80% of the day and

did not leave the nest at night. Males threatened conspecifics and waited for their incubating females from loafing spots throughout

incubation (usually) and attended them during their rest periods.

Females fed, preened, and bathed during these inattentive periods. 151

During summer evenings, Mandarins swim-fed along the shoreline

until 2200, then floated alertly, concealed under over-hanging

shoreline vegetation the remainder of the night. Some birds

continued to feed until 0200, probably also the pattern during the

laying and re-nesting periods.

Copulations always occurred in the water and were most frequent

during the laying and pre-laying periods. They usually were initiated

by the female, sometimes were contagious and occasionally followed

social display. Pre-copulation sequences generally were short

(10-15 sec), perhaps related to the Mandarin's predator susceptibility, and consisted of the female assuming a Prone position and male

Neck-craning, Bi 11-dipping, and Preening the female. Pre-copulation movements were mostly sexual in tendency with Bill-dipping as the final movement prior to mounting, perhaps ritualized. Post-copulation consisted of movements with escape tendencies.

Adult copulation sequences were more stereotyped than yearling sequences with the latter's pre and post copulation sequences longer in duration (as was mounting) and containing more displays. The efficiency of yearling sequences increased with experience suggesting the modifiability of such sequences with experience indicated by the mounting duration of second copulations being one-half that of the first.

Social display occurred only in the presence of the female with much display obviously directed at her, and resulted in pair formation.

Females through their movements, postures, and vocalizations, both initiated and intensified social display. 152

The season of year was important in both the method and length of time necessary for occurrence of pair formation. Fall or winter pairing took several weeks and involved social display, directed courtship, and changing partners. Late spring pairing (in the absence of social display) resulted from direct attempts by either sex with bonds formed in a week. Females indicated their preference for a particular male by Inciting him against other birds. An

"accepting'' male continually maintained his position next to the female, and reacted to the threats of conspecifics by assuming the

Full-sail posture, Bridling and Burping. Mutual Preening, Drinking, and Caressing movements strenghtened the pair bond. 153

Plate 1: Out-door breeding pens indicating types of nest boxes

Plate 2: Marking methods

Plate 3: Preferred nesting area 154

Plate 4: Adult female incubating in ground nest box

Plate 5: Male in Full-sail posture while female incubates in box 15

Plate 6: Mandarin clutch in nest box Plate 7: Feather development of Mandarin ducklings between 2-6 weeks of age

2 weeks 3 weeks 4 weeks

5 weeks 6 weeks 156

Plate 8: Secondary feather markings of yearling and adult Mandarin females

Yearling female

Adult female 157

Plate 9: Male Burping during social encounter following loafing

Plate 10: Female leading male in positions typical of most spring movements 158 ~*f^

Plate 11: 3-male social display party; female in foreground

Plate 12: male Burping; female in "Ack" vocalization posture

I------Plate 13: Male Body-shake (note neck and head not outstretched as in Fiq. 18): fpm^io •i n i*ia in vi 159

Plate 14: Male pecking female; female later paired to male at left

Plate 15: Female threatening B F16; male in Full- sail posture 160

Plate 16: male display Preening, probably to female at left

Plate 17: male display Preening; note clearity of nasal saddles 161

Plate 18: three males in Alert posture, vocalizing (Pfrruib); females vocalizing (Ack). Male in foreground in Full-sail posture. 162

Plate 19: Two (A-C and D-F) Drink-mock-preen sequences directed towards a female.

A. n)cOe in dtsjleiy position;note D. Drink (male positioning bill on side to female similar for Drink, ^lone)

E. Mock-preen while female Incites (male may be vocalizing).

C. Mock-preen behind raised sail F. Mock-preen behind wing and towards female raised sail feather towards female. 163

Plate 20: Display situations

A. non-pairing males Burp; female’s posture for "Ack" vocalization.

B. pair-drake (left) Pull-sail; E. female Preening by pair-drake female Incites at non-pairing (left); Tum-back-of-head by non­ males (right). pairing male (far right).

C. pair-drake Wing-flaps (fore­ F. Bill-jerk threat by non-pairinj ground); Bridle and Bill-jerk male (right); pair-drake (left by other males. Bridles. 164

Plate 21: Display situàtions

A. pair-drake threatening males.

B. beginning of Display-shake E, Burp and Bridle by two males (female preening;. behind female;(males vocalizing (Jibber).

C. Alert posture (right) from F. female vocalizing (Kett) which male vocalizes (Pfrruib) between pair-drake (foreground) and Head-flicks. and non-pairing males. 165 Plate 22: Display sequences A and B: Display-shake C: Burp D-F: Double-display-shake

A. initial stretch!, forward (male D, First Bill-dip by male in water in foreground).

B. body jerked upward (male E. vocalization (Gnk-zit). land).

C. Burp (male on land); female F. second Bill-dip out of view to right. 166

REFERENCES CITED

Anon. 1965. Breeding stock. Modern Game Breeding 1(6): 38-42.

Bait, B.D.J. 1970. The effects of cold on Mallard embryos. M.A. Thesis, University of Florida. 69 p.

Baumel. J.J. 1953. Individual variation in the White-necked raven. Condor 55(1): 26-32.

Bengtson, S.A. 1972. Breeding ecology of the Harlequin duck, Histrionicus histrionicus (L.). in Iceland. Ornis Scand. 3: 1-19.

Breckenridge, W.J. 1956. Nesting study of Wood ducks. J. Wild!. Mgmt. 20(1): 16-21.

Collias, N.E. and L.R. Jahn. 1959. Social behavior and breeding success in Canada geese (Branta canadensis) confined under semi-natural conditions. Auk 76(4): 478-509.

Cotter, Jr., W.B. 1957. A serological analysis of some Anatid classifications. Wils. Bull. 69(4): 291-300.

Dane, C. 1968. Age determination of Blue-winged Teal. J. Wild!. Mgt. 32(2): 267-274.

Dane, B., C. Walcott, and W.H. Drury. 1959. The form and duration of the display actions of the Goldeneye (Bucephala clangula). Behaviour 14: 265-281.

Delacour, J. 1959. The Waterfowl of the World. Vol. 3, Country Life, London. 270 pp.

Delacour, J. 1964. The Waterfowl of the World. Vol..4, Country Life, London. 364 pp.

Delacour, J. and E. Mayr. 1945. The family Anatidae. Wils. Bull. 57: 3-55.

Dilger, W.C. and P.A. Johnsgard. 1959. Comments on "species recognition" with special reference to the Wood duck and the Mandarin duck. Wils. Bull. 71: 46-53.

Doty, H.A. 1972. Hatchability tests with eggs from captive Wood ducks. Poultry Sci. 51(3): 849-853.

Dwyer, T.J. 1974. Social behavior of breeding Gadwalls in North Dakota. Auk 91(2): 375-386. 167

Elder, W.H. and N.W. Weller. 1954. Duration of fertility in the domestic Mallard hen after isolation from the drake. J. Wildl. Mgmt. 18: 495-502.

Engles, W.L. 1938. Variation in bone length and limb proportions in the coot (Fulica americana). J. Morphol. 62(3): 599-607.

Erskine, A.J. 1972. Buffleheads. Canadian Wildl. Serv. Mono. Series No. 4. Information Canada. 240 pp.

Ficken, R.W. and M.S. Ficken. 1966. A review of some aspects of avian field ethology. Auk 83(4): 637-661.

Gottlieb, G. 1968. Species recognition in ground-nesting and hole­ nesting ducklings. Ecology 49: 87-95.

Heinroth, 0. 1910. Beobachtungen bei einem Einburgerungsversuch mit der Brautente (Lampronessa sponsa (L.)) J.F. Ornith. 58(1): 101-156.

Heinroth, 0. 1911. Beitrage zur Biologie, namentlich Ethologie und Psychologie der Anatiden. Verh. V. Ornith. Kongr. (Berlin). 589-702.

Hinde, R.A. and N. Tinbergen. 1965. The comparative study of species- specific behavior, pp. 58-70; Readings in Animal Behavior (T.E. McGill, ed.). Holt, Rinehart and Winston. New York. 592 pp.

Hochbaum, H.A. 1942. Sex and age determination of waterfowl by cloaca! examination. Trans. N. Am. Wildl. Corif. 7: 299-307.

Holm, E.R. and M.L. Scott. 1954. Studies on the nutrition of wild waterfowl. New York Fish and Game J. 1: 171-187.

Johnsgard, P.A. 1960. A quantitative study of sexual behavior of Mallard and Black ducks. Wils. Bull. 72: 133-155.

Johnsgard, P.A. 1965. Handbook of Waterfowl Behavior. Comstock Pubi., Cornell Univ. Press. Ithaca, New York. 378 pp. johnsgard, P.A. 1968. Waterfowl: Their Biology and Natural History. Univ. Nebr. Press. Lincoln. 138 pp.

Johnsgard, P.A. and J Kear. 1968. A review of parental carrying of young by waterfowl Living Bird 7: 89-102.

Johnson, N.K. 1974. Molt and age determination in Western and Yellowish Flycatchers, Auk 91(1): 111-131.

Jones, T. 1972. Bird farm - 1971 Breeding season at Lackford. Avic. Mag. 78(1): 22-25. 168

Justice, K.E. 1961. A new method for measuring home range of small mammals. J. Mamm. 42: 462-470.

Kear, J. 1968. The calls of very young Anatidae. Beihefte der Vogelwelt 1: 93-113.

Kear, J. 1970. Studies cn the development of young Tufted duck. Wildfowl 21: 123-132.

Krapu, GJ... 1974. Feeding ecology of Pintail hens during reproduction Auk 91(2): 278-290.

Lack, D. 1968. Ecological adaptations for'breeding in birds. Methuen and Co. Ltd., London. 409 pp.

Lebret, T. 1961. The pair formation in the annual cycle of the Mallard, Anas pi atyrhynchos (L.) Ardea 49: 97-1.58.

Lorenz, K. 1941. Vergleichende Bewegungsstudien an Anatinen. J.f. Ornith. 89: 194-294. reprinted 1951-1953; Comparative studies on the behaviour of Anatinae. Avic. Mag. 57: 157-182; 58: 86-94, 8-17, 61-72, 172-184; 59: 24-34, 80-91.

Marler, P. 1961. The evolution of visual communication. P. 96-121 in Vertebrate speciation (W.F. Blair, ed.). Austin, Univ. of Texas Press.

McKinney, F. 1961. An analysis of the displays of the European Eider Somateria mol 1isstma mol 1issima (L), and the Pacific Eider Somateria mollissima v. nigra Bonaparte. Behaviour Suppl. VII. 124 pp. ~

McKinney, F. 1965a. The displays of the American Green-winged Teal. Wils. Bull. 77(2): 112-121.

McKinney, F. 1965b. Spacing and chasing in breeding ducks. Wildfowl 16: 92-106.

McKinnev, D.F. 1965c. The comfort movements of Anatidae. Behaviour 25: 120-220.

McKinney, F. 1965d. The spring behavior of wild Steller Eiders. Condor 67(4): 273-290.

McKinney, F. 1967. Breeding behaviour of captive Shovelers. Wildfowl Trust 18th Ann. Rep. 108-121.

McKinney, F. 1970. Displays of four species of blue-winged ducks. The Living Bird 9: 29-64. 169

McKinney, F. 1973. Ecoethological aspects of reproduction. Pp. 6- 21 in Breeding Biology of Birds. (D.S. Farner, ed.). Washington, D.C., Nat. Acad. Sci.

Meade-Waldo, E.G.B. 1912. Mandarin ducks at liberty. Avic. Mag. 3(4): 122.

Miller, W. De W. 1925'. The secondary remiges and coverts in the Mandarin and Wood ducks. Auk 4: 41-50.

Moynihan, M. 1955. Remarks on the orioinal sources of displays. Auk 72(3): 240-246.

Nakata, Y. 1965. Mandarin ducks. Modern Game Breeding 1(4): 29-31.

Oring, L.W. 1968. Growth, molts and plumages of the Gadwall. Auk 85(3): 355-380.

Oring, L.W. 1969. Summer biology of the Gadwall at Delta, Manitoba. Wils. Bull. 81: 44-54.

Orians, G.H. 1961 The ecoloqy of Blackbird (Aoelaius) social systems. Ecol. Mono. 31: 285-312.

Pitelka, F.A. 1958. Timing of molt in Stellar jays of the Queen Charlotte Islands, British Columbia. Condor 60: 38-49.

Pengelly, W.J. and J. Kear. 1970. The hand rearing of young Blue duck. Wildfowl 21: 115-121.

Preston, F.W. 1958. Variation of egg size with age of parent. Auk 75(4): 476-477.

Prestwich, A.A. 1960. On Mandarin duck hybrids. Avic. Mag. 66: 5-8.

Prince, H.H., P.B. Siegel and G.W. Cornwell. 1970. Inheritance of egg Droduction and juvenile growth in Mallards. Auk 87(2): 342-352.

Rollin, N. 1957. Incubation by drake Wood duck in eclipse plumage. Condor 59: 263-265.

Romanoff, A.L. 1967. Biochemistry of the Avian embryo. John Wiley and Sons, Inc. New York. 398 pp.

Romanoff, A.L. and A.J. Romanoff. 1949. The Avian Egg. John Wiley and Sons, New York. 918 pp. Rothestein, S.I. 1972. Eggshell thickness and its variation in the Cedar waxwing. Wils. Bull. 84(4): 269-747. 170

Rylander, M.K. and E.G. Bolen. 1970. Ecological and anatomical adaptations of North American tree ducks. Auk 87(1): 72-90.

Savage, C. 1952. The Mandarin Duck. Adam and Charles Black. London. 78 pp.

Scott, J.P. 1972. Animal Behavior. !Jniv. Chicago Press. 349 pp.

Smart, G. 1965. Body weights of newly hatched Anatidae. Auk 82(4) 645-648.

Sokal, R.R., and F.J. Rohlf. 1969. Biometry. W.H. Freeman and Co. San Francisco. 776 pp.

Sowls, L.K. 1955. Prairie Ducks. Wildl. Mgmt. Inst... Washington, D.C. 193 pp.

Stewart, P.A. 1958. Local movements of Wood ducks (Aix sponsa). Auk 75(2): 157-168.

Stewart, I.F. 1963. Variation of wing lenath with age. Bird Study 10(1): 1-9.

Stokes, A.W. 1962. Agonistic behaviour among Blut Tits at a winter feeding station. Behaviour 19: 118-138.

Swanson, G.A. and H.K. Nelson. 1970. Potential influence of fish rearing programs on waterfowl breeding habitat. Pp. 65-71 in A symposium on the Management of Midwestern Winterkill Lakes. (E. Schneberger, ed.). Madison, Wisconsin, Special Publ., Amer. Fisheries Soc., North Central Div.

Swanson, G.A. and A.B. Sargeant, 1972. Observation of nighttime feeding behavior of ducks. J. Wildl. Mgt. 36(3): 959-961.

Thorpe, W. 1951. The definition of terms used in animal behaviour studies. Bull Anim. Behav. 9: 34-40.

Thorpe, W. 1955. Ethology as a new branch of biology, pp. 34-49. iri Readings in Animal Behavior. (T.E. McGill, ed.) Holt, Rinehart and Winston. New York

Tinbergen, N. 1959. Comparative studies on the behaviour of gulls (Laridae): a progress report. Behaviour 15: 1-70.

Trauger, D.L. 1974. Eye color of female Lesser Scaup in relation to age. Auk 91(2): 243-254.

Verbeek, N.A.M. 1972. Daily and annual time budget of the Yellow- billed Magpie. Auk 89(3): 567-582. 171

Verner, J. 1965. Time budget of the male Long-billed Marsh Wren during the breeding season. Condor 67: 125-139.

von de Wall, W. 1953. Bewegungsstudien an Anatinen. J.F. Ornith. 104: 1-15.

von de Wall, W. 1965. "Gesellschaftsspiel" und Balz der Anatini. J.F. Ornith. 106: 65-80.

Weeks, J.L. 1972. An improved patagial streamer for waterfowl. Bird-Banding 43(2) : 140-141,.

Weidmann, U. 1956. Verhaltensstudien an der Stockente (Anas piatyrhynchos L.) I. Das Aktionssystem. Z. Tierpsychol. 13: 208-271.

Weidmann, U. and J.A. Darley, 1971a. The role of the female in the social display of Mallards. Anim. Behav. 19: 287-298.

Weidmann, U. and J.A. Darley. 1971b. The synchronization of signals in the "social display" of Mallards. Rev. Comp. Anim. 5: 131-135.

Weller, M.W. 1957. Growth, weights and plumaqes of the Redhead, Aythya americana. Wils. Bull. 69(1): 5-38.

Weller, M.W. 1964. The reproductive cycle. Pp. 35-79 in the Waterfowl of the World, vol. 4(J. Delacour, ed.). London, Country Life Ltd.

Weller, M.W. 1967. Courtship of the Redhead (Aythya americana). Auk 84(4): 544-559.

Welty, J.C. 1962. The Life of Birds. W.B. Saunders Company. Philadelphia. 546 pp. 172

Appendix I: Curators of Birds and private waterfowl breeders with whom communication regarding propagation or habits of Mandarins was made, or from whom Mandarins were purchased during the study (1970-1974).

Allen, G. A., Jr. editor Gazette, and director, Game Bird Preservation Center, Salt Lake City, Utah Beard, R. J. Private waterfowl breeder, Charlotte, North Carolina Bel 1, J. Curator of Birds, Bronx Zoological Park, New York Zoological Society, New York Block, M. Curator of Birds, Toledo Zoological Gardens, Toledo, Ohio Bruning, D, Assistant Curator of Birds, Bronx Zoological Park, New York Zoological Society, New York Butler, P. R. Curator of Birds, Stagsden Bird Gardens, Stagsden, Redford, England Funk, D. Private waterfowl breeder and President, Cain Funk Game Farm, Waterloo, Nebraska Greenwood, R. Wildlife Biologist, waterfowl propagation, USDI, and Lee, F. Northern Prairie Wildlife Research Center, Jamestown, North Dakota Guest, W. Private waterfowl breeder, Puddle Duck Farm, Islip, New York Hamilton, D. Private waterfowl breeder, New Bern, North Carolina Johnston, S. T. Curator of Birds, The Wildfowl Trust, Slimbridge, England Jones, T. Private waterfowl breeder, Leckford Estate Limited, Stockbridge, Hants, England Kear, J. Waterfowl propagation and research, The Wildfowl Trust, Slimbridge, England Macy, W. K. Private waterfowl breeder, Islip, New York Olney, P. J. Curator of Birds, London Zoological Gardens, Regent's Park, London, England Savi 11, E. Director of Gardens, Windsor Great Park, Berks, England Stevens, N. Private waterfowl breeder, Wallcot Hall Estate, Barbers, Martley, Worcester, England Strutz, C. Private waterfowl breeder and Secretary, Northern Prairie Wildlife Research Center, USDI, Jamestown, North Dakota Webster, C. Private waterfowl breeder, Islip, New York 173

Appendix II. Summary of reproductive success of Mandarin ducks on Schedel Estate from 1962-1970. Data are interpretations from partially complete records kept by Mr. Mel Block, Bird Curator, Toledo Zoo. Full information was not avail­ able on all clutches, and not all columns are complete.

Number of adult ducks Number of eggs and young killed or killed missina in- re- or year V? total fertile hatch 1 ease missing

1962 6 9 0 0 16 9 5 5 1

1963 16 13 0 1 40 21 17 0 0

1964 12 8 3 2 13 4 9 7 3

1965 13 13 1 1 49 27 22 18 9

1966 19 12 3 3 61 31 38 31 17

1967 15 7 5 4 4 4 0 0 0

1968 13 7 2 0 4 2 2 0 2

1969 6 4 6 3 3 2 1 0 1

1970 4 4 2 0 4 4 0 0 0 174

Appendix III: Propagation proceedures - 1971

1. Two adjacent pens, each with 70 m2 land and 55 m2 water were constructed.

2. Water and soil samples were taken from the lakes and

submitted to Ohio State University for pesticide analysis.

3. The total number of waterfowl was reduced from 155 in spring 1970 to between 23-55 during each remaining year of the study.

4. All birds were marked with plastic wing strips, and the number of birds per enclosure was reduced from four pair (1970) to two pair with one pair maintained free on the lake.

5. To induce renesting, the eggs of one clutch were removed from the nest box on the second day of incubation and placed under a bantam hen for incubation. The eggs of another clutch, also incubated by a bantam, were removed as laid and replaced with plastic eggs to stimulate continued filling of the clutch. 175

Appendix IV. Propagation proceedures - 1972

The pens and breeding proceedures remained identical to those of 1971 with the following exceptions:

1. The top of one outdoor breeding pen was covered with

V chicken fencing to maintain two pairs of full-winged birds. Two pairs of pinioned birds were kept in the adjacent pen, and three pairs were free-ranging.

2. Patagial markers, made from Saf-Flag material and tested at Ottawa Federal Wildlife Refuge, Oak Harbor, Ohio (Weeks, 1972) were attached to the birds using dental floss immediately prior to their release on 15 March. This release date was advanced from 1971 since precopulatory behavior was observed.

3. Most females were allowed to incubate their own eggs despite the predation risk, yet two hatching methods were evaluated. After pipping and just prior to hatching (on two clutches), and after putting food and water in the nest boxes, the entrance to each box was covered with screen while the female was incubating. This hatching proceedure had been used success­ fully on the estate with Wood ducks and Mallards. The eggs of another clutch, when pipping, were placed under a bantam for hatching, a proceedure used by many waterfowl breeders.

4. Following hatching all ducklings were put in .6 m x .3m brooder boxes for two weeks. They were then transferred to larger

7.5 x 8.5 m indoor pens in the over-wintering building, which contained four pens of this size and eight pens 8' x 24'(2.4 m

x 7.5 m).; 176

Appendix V. Propagation proceedures - 1973 and 1974

Except for the following minor changes, breeding proceedures for

1973 and 1974 were similar to those of 1972:

1. Seven (1973) and six (1974) pinioned pairs of Mandarins,

marked with colored vinyl plastic nasal saddles (obtained from

Ray Greenwood, USOI, Northern Prairie Research Center) and numbered with

Fearing Tuff-Flex Marking Paint (Fearing Mfg. Co., St. Paul, Minn.), were released from the over-wintering pens to the estate grounds on

29 March and 5 April, respectively. Two additional pairs were put in each of the two outdoor pens each spring.

2. The dividing fence between the pens was covered with burlap and grass to visually separate the birds and reduce agonistic encounters.

3. Females were allowed to incubate until pipping (usually), but the presence of predatory fish in the lakes necessitated a switch to bantams or an incubator at hatching. Any eggs not incubated by females or remaining after nest disturbances, were treated in the same manner.

4. In 1973, to induce a second clutch and allow a study of re­ nesting behavior and reproductive success, the eggs from both an experienced and inexperienced female were removed five days after the onset of incubation and incubated under bantam hens. This period has been found by Doty (1972) to increase hatching success in Wood ducks. 177

Appendix VI. Summary of Mandarin pairs under observation in 1971. Birds were released from winter quarters on 29 March.

age* Pairs (years) Obtained Comments

Pen A M 0A8 1 Funk, Mar 71 removed after M 6A10 changed pens and paired with F 0A9 F 0A9 1 Funk, Mar 71 missing, fall 71

M 0A10 1 Hami1 ton, Feb 71 released from Pen A to lake, May 71 F 0A7 1 Hamilton, Mar 71 released from Pen A to lake, May 71

M 0A1 1 Hamilton, Mar 71 removed after F 0A6died; missing, fall 71 F 0A6 1 Hami1 ton, Feb 71 died "shock," 31 Mar 71

Pen B M 3A3 9 Schedel killed by owl; Sept 71 F 3A8 10 Schedel missing, Sept 71

M 6A10 5 Schedel jumped to pen A F 6A9 5 Schedel died 27 April**

UnDai red M 5A11 5 Schedel

*Ages for individuals in Pen A are exact, while those of Pen B may be ± 1 year

**Death diagnosed as "infected bowels" by Mel Block, Curator of Birds, Toledo Zoo ***

***A11 birds released from breeding pens to lake, 16 June 178

Appendix VII. Summary of Mandarin pairs under observation in 1972. Birds were released from winter quarters on 15 March.

age Pai rs Color code (years) Obtained Comments

Pen A, full-winqed M 0A14 royal blue 1 Funk, Jan 72 killed by owl; Aug. 7 F 0A13 royal blue 1 Funk, Jan 72 escaped to lake;4 May

M 0A16 pink 1 Funk, Jan 72 escaped & left; 4 Apr F 0A15 pink 1 Funk, Jan 72 missing; 30 September

Pen B, pinioned M 0A11 or-red ? auction, Nov 71 alive F 0A12 or-red ? auction, Nov 71 escaped and missing; 30 September

M 0A38 dk-blue 2 Funk, 71 alive F 0A18 dk-blue T1 Webster, Mar 72 died "shock"; 19 May

Free-ranging, pinioned M 0A20 orange 1 Webster, Mar 72 killed by owl; 31 May F 0A17 orange 1 Webster, Mar 72 killed by owl; 4 May

M 6A10 plain 6 Schedel missing, August F 0A19 green 1 Webster, Mar 72

M 0A21 white 2 Webster, Mar 72 killed by owl; August F 0A23 purple 1 Webster, Mar 72 missing, July

Unpaired males M 0A24 light blue 1 Webster, Mar 72 released to pen A on 3 May, and to lake on 5 May; alive M 0A22 purple 2 Webster, Mar 72 kept in indoor pen until July and then released to Pen A; alive M 5A11 plain 6 Schedel kept in indoor pen until July and then released to Pen A; alive 179

Appendix VIII. Summary of Mandarin pairs under observation in 1973. Birds were released from winter quarters on 29 March.

age Pairs Color code (years) Obtained Comments Pen A, covered M 0A33 gr 14 1 raised, spring 72 F 0A32 wh 21 1 raised, spring 72 M 0A22 red1 3 3 Webster,, Mar 72 blinded by M 0A33i on 19 Apr and removed from pen; died 22 Apr F 0A26 wh 24 1 raised, spring 72 released to lake 22 Apr Pen B, uncovered M 0A34 gr 11 1 raised, spring 72 F 0A44 wh 19 1 raised, spring 72

M 0A29 gr 12 1 raised, spring 72 F 0A27 wh 17 1 raised, spring 72

M 0A25 gr 9 1 raised, spring 72 crippled Jun 72;died Apr 73 F 0A35 wh 23 1 raised, spring 72 released from Pen B to lake, 22 Apr: missing, June 72 Free-ranging M 0A28 gr 10 1 raised, spring 72 missing, 17 Apr F 0A31 wh 22 1 raised, spring 72

M 0A24 red 5 2 Webster, Mar 73 F 0A19 bla 7 2 Webster, Mar 72 killed by cat; 2 May (after laying) M 0A39 red 6 2 Webster, Jan 73 killed by owl; 4 Apr F 0A30 wh 18 1 raised, spring 72 killed by cat; 29 May

M 0A38 red 2 3 Funk, 1971 F 0A36 wh 20* 1 raised, Spring 72

M 0A41 red 7 2 Webster, Jan 73 F 0A40 bla 16 2 Webster, Jan 73

M 5A11 red 4 7 Schedel 1965 paired during 1972 F 0A12 bla 15** ? auction, fall 72 missing, 2 May

M 0A38 gr 13 1 raised, spring 72 missing, 23 May; paired with wh 20 prior to 23 May M 0A11 red 1*** 2 *F 0A36 associated with both red 4 (while wh. 23 was laying) and red 2 (while wh 22 was laying) **F 0A12 associated with M 0A4i and F 0A40 until 2 May when she was missing. ***M 0A11 was not released from over-wintering pens, as he did not pair and was in poor physical condition. 180

Appendix IX. Summary of Mandarin pairs under observation in 1974. Birds ...... were released from winter quarters on 5 April.

Age Pairs (loi or code (years) Obtai ned Comments Pen A, covered M 0A43 blue E 1 Schedel, 73 F 0A49 red A 1 Schedel, 73

M 0A46 red 2 3 Funk, 70 F 0A48 red H 1 Schedel, 73

Pen B, uncovered M 0A34 2 gr 11 Schedel, 72 also paired in spring, 73 F 0A34 wh 19 2 Schedel, 72

M 0A29 gr 12 2 Schedel, 72 also paired in spring, 73 F 0A27 wh 17 2 Schedel, 72

Free-ranging M 0A4Ì red 7 3 Webster, 73 also paired in spring, 72 at F 0A40 blue 16 3 Webster, 73 Webster's and 73 at Schedel's

M 5A11 red 4 8 Schedel, 65 also paired in spring, 73 F 0A36 wh 20 4 Schedel, 70 missing, 5 July 74

Unpaired F 0A41 red C 1 Schedel, 73 first clutch (22-30 Apr) in­ fertile first paired (red 7) 28 Apr second clutch. (11-15 May) fer­ tile consort with M 5A11 (14 May); missing, 15 June F 0A46 red V 1 Schedel, 73 first clutch (22 Apr-7 May), infertile first paired (10 May), with M 0A41 second clutch (14-21 May), fertile F 0A32 wh 21 3 Schedel, 71 consort with M 5A11 and F 0A4I (14 May) M 0A47 blue A 1 Schedel, 73 missing, 5 July

M 0A42 blue C 1 Schedel, 73

*all penned birds released to lake, 3 June. 181

Appendix X. Time (h) spent observing Mandarins 1971-1973.

dawn- 1100- 1500- 1059 1459 dark totals

1971* Mar 0 0 2 2 Apr 3.5 10.25 22.75 36.5 May 3.5 0 26.’ 5 30 June 4.5 6 5.5 16 July 3 4 3 10 totals 14.5 20.25 59.75 94.5 1972** pre-laying 15 Mar-2 Apr .5 3.5 4 laying 3 Apr-20 Apr 5.5 2.5 2.0 10 post-laying 21 Apr-26 May 9.5 9.5 7.0 26 totals 15.5 15.5 9.0 40 1973*** pre-1aying 3-19 Apr 9(16) 14(23) 9(11) 32(50) laying 20 Apr-15 May 26(43) 14(21) 16(28) 56(92) post-laying 15 May-31 May 11(20) 10(18) 8(13) 29(51)

Total - Spring 46(79) 38(62) 33(52) 117(193)

Total - Apr-Oct 64 68 64 195

*1971 - primarily on penned birds. **1972 - since several penned birds could be observed simultaneously the actual observation time was approximately one-half that from which some data are derived. ***1973 - parentheses enclose number of 1/2 h intervals during each period from which some data are derived. 182

Appendix XI. Data sheet for activities of birds 1971-1973. Activities recorded each 30 seconds.

Date_____ Temp______Time______Birds______Behaviors 1 2 3 4 5 6 7 8 9 10 Vocalize V prone H-flick (circum) T-wag (circum) Body-shake Disp-shake Burp Wing-flap Dbl-disp-shake Drink Dr-mock-preen Caressing mv'ts Dbl-w-str/s-w-str pre-flight Bridle Agon is tics V Incite

Acti vi ty

loaf(si t/stand) preen wal k feed swim skitter (circum) bathe (no. H-bob) n.b.i./Iaying ) Copulation 183

Appendix XII. Data sheet for activities of birds at 15 or 30 min intervals - 1973 and 1974. Individuals were identified and appropriate code indicated.

time sleep preen agon, feeding females n.b.i other intervals loaf enct. inc. lay M F M F M F M F 184

Appendix XIII. Code for use with data gathering sheets.

BIRDS - 1973* VOCALIZE FEEDING BODY-SHAKE

Free-ranging 1 - Kett 1 - tipping 1 - water

G MIO W F22 2 - Ack 2 - fish 2 - land

R M5 BL F7 3 - Pfrruib 3 - surface 3 - side-to-9 str-to-V R M6 W FI 8 4 - others 4 - graze

R M2 W F20 5 - insects from air

R M7 B F16 6 - corn

R M4 B F15 PREEN AGONISTIC

Penned-covered 1 - oil 1 - threat G M14 W F21 2 - behind-wng-? 2 - jab R M3 W F24 3 - outside-wng-$ 3 - chase open 4 - chest or belly 4 - slow pursuit G Mil W F19 5 - neck G M12 W FI 7 6 - dorsal G M9 W F23 7 - others

*altered when individuals died and/or pairs changed 185

Appendix XV. Growth parameters of Mandarin ducklings observed at one week intervals from hatching for ten weeks, and again on 28 October 1973. Measurements of adults in October, Jan­ uary, and March also included. All weights in grams and lengths in mm ± SD.

males (n=5) females (n=8) age tarsus mid culmen culmen tarsus mid culmen culmen (days ) weight In toe In In wd weight In toe In In wd 29.64 24.64 24.92 11.26 7.42 25.29 24.34 24.00 10.78 7.40 1 ±2.88 ±1.52 ±2.25 ±0.52 ±0.52 ±2.69 ±1.10 ±1.21 ±0.54 ±0.34 53.60 29.48 29.60 14.34 9.14 56.86 28.02 29.17 14.38 9.04 7 ±12.53 ±2.24 ±1.55 ±1.09 ±0.72 ’ ±10.41 ±2.49 ±3.10 ±1.12 ±0.86 141.40 40.06 40.20 19.40 11.64 102.29 34.75 35.15 18.46 10.68 14 ±20.62 ±1.89 ±1.48 ±0.91 ±0.47 ±30.14 ±3.78 ±4.10 ±1.97 ±0.86 213.60 43.40 43.10 22.62 12.78 161.88 38.95 39.38 21.39 11.80 21 ±22.03 ±1.95 ±1.47 ±0.97 ±0.49 ±46.75 ±4.15 ±3.65 ±2.24 ±1.01 321.40 44.80 46.62 25.02 13.60 240.38 41.94 43.75 21.00 13.19 28 ±36.83 ±1.30 ±1.69 ±0.33 ±0.73 • ±65.25 ±3.84 ±4.04 ±6.55 ±0.99 or 393.25 45.17 47.25 26.13 14.70 322.63 43.19 45.05 25.27 13.73 35 ±35.66 ±0.85 ±0.50 ±0.50 ±0.60 ±68.31 ±3.42 ±2.55 ±2.34 ±0.81 430.75 48.05 49.00 27.63 14.72 334.63 46.80 47.66 26.60 14.22 42 ±42.70 ±1.28 ±3.06 ±1.09 ±0.62 ±45.48 ±8.09 ±1.81 ±1.64 ±0.65 436.50 47.63 48.63 28.30 15.00 397.80 45.11 47.41 27.42 14.61 49 ±44.46 ±0.48 ±3.09 ±0.50 ±0.58 ±44.91 ±1.93 ±1.89 ±1.92 ±0.48 460.75 48.70 48.05 28.90 15.15 432.00 45.64 47.81 28.54 15.00 56 ±44.42 ±2.58 ±2.19 ±0.82 ±0.93 ±42.26 ±1.41 ±2.39 ±1.92 ±0.74 495.00 47.75 49.20 29.22 15.65 438.13 45.76 47.59 28.17 15.29 oJ ±47.93 ±1.32 ±2.59 ±1.52 ±0.57 ±35.42 ±1.39 ±2.56 ±1.54 ±0.72 542.18 47.70 49.31 29.27 15.71 479.60 45.60 47.72 28.39 15.50 /u ±40.1J3 ±1.18 ±1.89 ±1.36 ±0.62 ±41.32 ±1.65 ±2.30 ±1.31 ±0.62

133 611.43 47.64 50.62 28.88 15.90 567.67 45.53 48.76 28.90 15.82 (Oct) ±30.20 ±1.55 ±0.95 ±1.14 ±0.65 ±47.29 ±1.49 ±1.66 ±0.83 ±0.46 Adults

28 Oct 577.33 47.26 51.89 29.65 16.51 in-P)530-29 47.07 50.86 29.71 15.79 (n=8) ±28.56 ±0.95 ±1.31 ±1.00 ±0.46 (n_8'±38.28 ±1.38 ±2.07 ±1.11 ±0.53

11 Jan 580.40 48.66 54.30 29.52 16.54 /n_?\546.00 48.90 51.75 30.00 16.35 (n=5 ±37.36 ±0.44 ±4.79 ±0.72 ±0.46 1 '±19.80 ±0.99 ±4.17 ±0.00 ±0.64

29 Mar 622.29 49.17 53.17 30.36 16.20 <„_q^501 .00 47.80 53.93 29.73 15.87 (n=7) ±89.89 ±0.88 ±1.24 ±0.65 ±0.43 1 J ±32.70 ±1.25 ±4.19 ±1.27 ±0.64 186

Appendix XVI. Feather development of Mandarin ducklings observed for ten weeks, at one week intervals from hatching - summer 1973. Date, unless otherwise indicated, is for the first appear­ ance through the skin of the particular feathers in any of the birds of each sex. Lengths are maximum and fractions indicate the variation.

age (days) Males (n=5) Females (n=8)

14 Scapulars, breast, stomach Scapulars, breast, stomach Tail - 23mm: 1/2 unsheathed Tail - 22mm: 1/2 unsheathed

21 Primaries, secondaries Primaries, secondaries Secondary coverts Secondary coverts Tail - 42mm: 2/3 unsheathed

28 Primaries - 32mm: 1/5 Primaries - 31mm: 1/7 unsheathed unsheathed Secondaries - feathers starting Secondaries - feathers starting to unsheath to unsheath Bill starting to turn pink Voice starts to change Head feathers Head feathers Stomach and breast complete Stomach and breast complete Tail - 47mm: 5/6 unsheathed Tail - 60mm: 5/6 unsheathed

35 Primaries-76mm: 1/3-1/2 Primaries-83mm: 1/3-1/2 un­ unsheathed sheathed, 9-10 start to unsheath Secondaries-56mm: 1/2 Secondaries-59mm: 1/3 un­ unsheathed sheathed, 6-10 start to unsheath Tail and scapulars complete Tail completely unsheathed:76mm Bill completely pink in 4 birds Voice changed Neck, scapulars, stomach com­ plete

42 Primaries 1-8: completely Primaries-112mm: 1/3-3/4 unsheathed, 9-10mm: 1/2- unsheathed 2/3 unsheathed Secondaries-75mm: 2/3 Secondaries-83mm: 1/3-2/3 unsheathed unsheathed . Head complete FI i ght

49 Flight Head complete Secondaries completely Secondaries completely unsheathed unsheathed Voice changed Primaries-8-10 still sheathed 187

Appendix XVI (continued).

age (days) Males. (n=5) Females (n=8)

56 One bird completely feathered One bird completely feathered (except for lumbar down) Voice changed in second bird

63 All primaries unsheathed in all Primaries completely unsheathed birds in two other birds

70 Natal down present under wing All primaries unsheathed in all and in lumbar region birds Natal down present under wing and in lumbar region Appendix XIV. Data sheet for location of birds during the study.

DATE-

TIME-