THE WOOD DUCK, AIX SPONSA (LINNAEUS), AND ITS

MANAGEMENT

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Paul Alva Stewart, B.Sc., M.Sc.

The Ohio State University

1957

Approved by:

0 Adviser ACKNOWLEDGMENTS

Most of the field work in this study was done in Delaware,

Fairfield, Franklin, Licking, Perry, Pickaway and Ross Counties in central Ohio; several visits were made to Magee Marsh in Ottawa

County, and a visit was made to the Savannah National Wildlife Refuge,

Savannah, Georgia. The study was started in March 1955 and concluded in June 1957.

The study was conducted when the author was a research fellow of the Ohio Cooperative Wildlife Research Unit. Special thanks are due to Dr. Eugene H. Dustman, Unit Leader, for guidance and financial support throughout the course of the study.

I am indebted to Frederick C. Lincoln, U.S. Fish and Wildlife

Service, for a detailed history of hunting regulations applying to the wood duck. I am also indebted to Allen J. Duvall for making avail­ able to me all data of banded wood ducks in the files of the U.S. Fish and Wildlife Service, Conversation with Francis M, Uhler, U.S. Fish and Wildlife Service,was very helpful. Special thanks also go to E.O.

Mellinger and Horrie Womble for their co peration in my work at the

Savannah National Wildlife Refuge.

Delmar E. Handley and Karl Bednarik, Ohio Division of Wildlife, contributed much useful information and assistance. Personnel of game agencies in nearly all of the AS states replied to a questionnaire and provided helpful information. Wm. R. Miller, Vermont, and Dirck Benson,

New York, were particularly helpful. Milton B. Trautman, Ohio State Museum, aided in field work and contributed much useful information and encouragement. Robert W.

Sagar also contributed much-appreciated assistance and information.

George W. Blaydes aided in identification of seed specimens. Richard

D. Alexander aided in analysis of voice recordings. Donald J. Borror and Carl R. Reese aided by making tape recordings of wood duck calls.

Maurice L. Giltz and Ernest E. Good aided in the study of the wood duck's flight. Loren S. Putnam, Robert W. Winner, Robert W. Donohoe,

George J. Fhinney, Francis W. Kessler and Norman C. Negus aided in sundry ways.

By granting permission for me to erect nesting boxes on their property, many farmers contributed toward the success of this study.

The privilege of being permitted to work on the Wehrle Estate at

Buckeye Lake was particularly helpful.

Most of the statistical tests were performed in the Statistics

Laboratory, The Ohio State University.

In typing and retyping the manuscript and in proof-reading,

Esther G. Stewart made a major contribution toward completion of the report. TABLE OF CONTENTS

PART I. HISTORICAL AND BIOLOGICAL BACKGROUND

Chapter Page

I. INTRODUCTION AND HISTORICAL ACCOUNT 2

Historical account 3

Summary 7

II. TAXONOMY AND DISTRIBUTION 8

Taxonomy 8

Synonymy 8

Vernacular names 9

Distribution 9

III. BIOLOGICAL BACKGROUND 11

Plumages and molts 11

Natal down 11

Juvenal plumage 11

Postjuvenal molt 13

Plumage of adult male 13

Plumage of adult female 16

Sequence of molts and plumages 17

Field identification characteristic® 18

Voice 19

Growth 25

Weights 40

Body temperature 47

iv Chapter Page

Locomotion 51

Swimming speed 51

Diving 52

Skittering speed 53

Running speed 53

Flight 54

fi' Entering the nesting cavity 55

Malformations 56

Life expectancy 57

Summary 59

IV. ’RODUCTIVE BEHAVIOR 61

Courtship 61

-Selection of the nesting cavity 62

Nesting territory 63

Egg-laying 64

Incubation 69

Relative humidity in the nest 78

Turning the eggs 73

Injury-feigning 80

Hatching 80

Renesting 85

Hybridization 86

Summary 87

IE OF THE YOUNG 89

Movement of ducklings to water 89 vi

Chapter Pagg

Care of young 96

Summary 98

PART II. THE MOOD DUCK IN ITS ENVIRONMENT

VI. LOCAL MOVEMENTS 101

Brood Movements 101

Local movements of young after attaining ability of flight 111

Early fall dispersal of late summer congregations 116

Winter movements 118

Summary 122

VII. ANALYSIS OP BANDING DATA 125

Comparison of data from different states 125

Effect of hunting 135

Late summer radial dispersal 136

Southward migration 138

Northward migration 144

Return to natal area and earlier nesting place 146

Recovery of pen-reared birds 150

Summary 150

VIII. FOOD AND FEEDING 153

Food 153

Feeding 158

Summary 160 vii

Chapter ESfiS

IX. ROOSTING AND LOAFING 163

Roosting 163

Loafing 166

Summary 167

X. NESTING AND BROOD HABITATS 169

Summary 184

XI. DECIMATING AND LIMITING FACTORS 185

Insects 185

Fish 185

Bullfrog 186

Snapping turtle 187

Snakes 188

Hooded merganser as nest competitor 189

Starling as nest competitor 189

Miscellaneous avian nesting box competitors 190

Avian predators 192

Opossum 192

Mice 192

Raccoon 193

Squirrels 195

Mink 196

Hunting 197

Muskrat traps 198

Loss of nest-hunting females in chimneys 198

Frost-cracking of eggs 199 viii

Chapter Page

Nest desertion 201

Loss of ducklings in water-current 201

Filamentous algae 202

Parasites and disease 205

Weather 208

Summary 208

PART III. MANAGEMENT OF THE WOOD DUCK

XII. NESTING PLACES 212

Natural nesting places 212

Artificial nesting places 213

History of nesting box development 2 U

Further comments on raccoon-proofing nesting boxes 216

Loss of nests to experimental method 218

Test of different types of nesting boxes 219

Conventional or basic type box 219

Elliptical entrance 236

Tunnel entrance 238

Grease drum 238

Hole-in-lid ammunition case 24-2

Hole-in-side ammunition case 242

Galvanized pipe nest box 245

Nail keg box 246

Miscellaneous types of boxes 247

New versus old boxes 247 Chapter

Post versus tree mounting

Density of nesting boxes

Sawdust or other loose material in nesting box

Height and exposure of boxes

Vandalism

Checking nesting boxes

Summary

XIII. MARKING INDIVIDUAL WOOD DUCKS FOR SUBSEQUENT IDENTIFICATION

Bait trapping

Banding incubating female

Use of a dog in capturing flightless young

Imping feathers

Painting and dying

Color-banding

Toe-clipping and web-punching of ducklings

Wing-tagging of ducklings

Summary

XIV. CENSUS METHODS

River float counts

Brood counts

Bridge stops

Aerial Survey

Feeding flight counts

Roosting flight counts

Nesting box check X

Chapter Page

Summary 303

XV. LEGAL PROTECTION FOR THE WOOD DUCK 305

Benefit of a selective closed season 305

Protecting local northern populations from over shooting 307

The place of refuges in wood duck management 308

Legal shooting hours and baiting 309

Bag limits 310

Summary 311

XVI. ESTABLISHING A LOCAL WOOD DUCK POPULATION 312

Rearing and releasing 312

Extensive versus intensive management 3 H

Human interference in attracting wood ducks 315

Use of loafing logs 317

Summary 317

XVII. MISCELLANEOUS TECHNIQUES IN WOOD DUCK MANAGEMENT 319

Finding nests of wood ducks 319

Use of nest eggs for inducing nesting 320

Aging of eggs 320

Egg membrane counts 321

Obtaining data on sex ratios 326

Obtaining data on age ratios 328

Collecting crop contents from living ducks 329

Alimentary tracts from ducks in hunters' bags 330

Fishermen interviews 331

Summary 331 APPENDIX

List of scientific names

LITERATURE CITED LIST OF FIGURES

Figure Page

1. The immature female can be recognized by the white spot around the eye 12

2. The immature male can be recognized by the white cheek patch 14-

3. The white cheek patch of the immature male is some­ times relatively indistinct 15

4.. Audiospectographic analyses of a. call of caged male; b. peep of day-old duckling; c. peep of unknown number of ducklings in nest 21

5. Audiospectographic analyses of a. call of female when disturbed with young; b. call of female when dis­ turbed with young; c. call of caged female 23

6. Growth curve of captive-reared wood duck 27

7. Flight feathers of wild-reared woodduck at eight weeks of age 28

8. Flight feathers of captive-reared wood duck at six weeks of age 29

9. Flight feathers of captive-reared wood duck at seven weeks of age 30

10. Flight feathers of captive-reared wood duck at eight weeks of age 31

11. Captive-reared wood duck at one week of age 32

12. Captive-reared wood duck at two weeks of age 33

13. Captive-reared wood duck at three weeks of age 34.

14.. Captive-reared wood duck at four weeks of age 35

15. Captive-reared wood duck at five weeks of age 36

16. Captive-reared wood duck at seven weeks of age 37

17. Captive-reared wood duck at fifty-three days of age 38

xii X1XX

Figurw Page

18. Wild-reared wood duck at fifty-three days of age 39

19. That single clutches are sometimes laid by more than one female is indicated by differences in the size and shape of the eggs 65

20. Abnormally large eggs are sometimes laid 70

21. Automatic nesting activity recorders were operated on two nests 74

22. Record of attentive periods at a wood duck nest 76

23. The eggs are frequently turned and their relative positions in the nest are changed 79

24. In hatching, the duckling cuts a cap from the large end of the egg 82

25. The neck is curved backward on the duckling's belly with the egg tooth against the shell 83

26. "When leading the brood from the nest, the female paused at- the entrance before flying to the ground 91

27. Wood ducks left this pond when about two weeks of age 102

28. Broods gathered on this pond in both 1955 and 1956 105

29. This pond was occupied by brood congregations during 1955 and 1956 110

30. Wood ducks gathered at this pond soon after attaining the ability of flight in 1955 and 1956 115

31. Pools among the maiden-cane were used for loafing places but less frequently than pools among trees and bushes 120

32. Water pools among buttonbushes were much used for loafing places 121

33. Distribution of recoveries of wood ducks banded in Illinois 126

34. Distribution of recoveries of wood ducks banded in Maine 127 xiv

Figure Page

35. Distribution of recoveries of wood ducks banded in Massachusetts 128

36. Distribution of recoveries of wood ducks banded in Michigan 129

37. Distribution of recoveries of wood ducks banded in New York 130

38. Distribution of recoveries of wood ducks banded in Vermont 131

39. Corn and sand removed from digestive tract of trapped wood duck 161

40. Wood ducks roosted among spatter-dock near the cat-tails 165

41. Occasionally some of the birds swam slowly about for short distances 168

42. A woodland pond affords excellent wood duck habitat 170

43. Frost-cracked egg 200

44-* Ponds containing abundant surface vegetation are potential death traps because of filamentous algae beneath the surface 203

45. This raccoon readily climbed a steel pipe two inches in diameter 217

4 6 . Conventional or basic type box 234

47. Elliptical entrance nesting box 237

48. Tunnel entrance nesting box 239

49. Grease dum nesting box 240

50. Hole-in-lid ammunition case nesting box 243

51. Hole-in-side ammunition case nesting box 244

52. Wood duck nesting box mounted on steel fence post 249

53. The steel fence posts are sometimes bent by strong winds 250

54* Boxes were used as much as one-fourth mile from large trees 253 Figure

55. The nesting boxes were sometimes lost to vandals

5 6 . The six wood duck traps were easily moved from place to place

57. One of the traps in operation

58. Most incubating females could be readily removed from the nest for banding

59. Wing-tagged duckling

60. A flock of 20 wood ducks remained on this section of the Hocking River during the late summer and fall of 1956

61. Egg shell and membrane after hatching of duckling

62. Egg membranes after ducklings left the nest LIST OF TABLES

Table ?age

I. Weights of 64 adult wood ducks 41

II. Weight of 106 immature wood ducks 42

III. Food in digestive tracts of wood ducks 44

IV. Body temperatures of 143 wood ducks 50

V. Dynamic Life table for wood ducks 58

VI. Measurements of eggs in a wood duck nest 71

VII. Direct recoveries of banded wood ducks 132

VIII. Indirect recoveries of banded wood ducks 133

IX. Comparative utilization of nesting boxes in different habitats 171

X. Numbers of ducklings in broods at different stages of development in different habitats 173

XI. Brood counts for May and June in large river habitat with U-tests for differences 176

XII. Brood counts for small stream and large river habitats with U-tests for differences 177

XIII. Brood counts for farm pond and large lake habitats with U-tests for differences 178

XIV. Brood counts for large lake and small stream habitats with U-tests for differences 179

XV. Brood counts for farm ponds and small stream habitats with U-tests for differences 180

XVI. Brood counts for farm ponds and large river habitats with U-tests for differences 181

XVII. Brood counts for large lakes and large river habitats with U-tests for differences 182

XVIII. Summary of nesting box survey in 1956 220

xvi xvii

Table Page

XIX. Summary of nesting box survey in 1957 225

XX. Comparative use by wood ducks of different types nesting boxes 227

XXI. Comparative success in different types boxes 228

XXII. Relative use of different types of nesting boxes 229

XXIII. Comparative use by starlings of different types nesting boxes 231

XXIV. Relationship of nesting box usage by wood duck and starlings 232

XXV. Utilization of different types of nesting boxes at Magee Marsh, 1957 233

XXVI. Surveys of wood duck population on an 11-mile section of the Scioto River 281

XXVII. Variation in numbers of wood ducks observed on an 11-mile transect of the Scioto River in April 28£

XXVIII..Variation in numbers of wood ducks observed on an 11- mile transect of the Scioto River in May 285

XXIX. Variation in numbers of wood ducks observed on an 11- mile transect of the Scioto River in June 286 PART I

HISTORICAL AND BIOLOGICAL BACKGROUND

1 CHAPTER I

INTRODUCTION AND HISTORICAL ACCOUNT

The annals of American history are besmirched with shameful re­

minders of the wanton wastefulness of white savages. For unknown cen­

turies red savages lived and prospered along with the passenger pigeon

and the wood duck. With the white man*s unmanaged taming of the land,

these birds had to go, and the passenger pigeon left the scene never to

return. Because a few scattered individual wood ducks were able to sur­

vive through that initial period of unrestrained destruction, the wood

duck lives today. He lives as a reminder of the dawning of a new era —

the awakening of a new consciousness. He lives as a tribute to the

forward thinking of conservation pioneers, such as Thomas Jefferson,

Gifford Pinchot, and Theodore Roosevelt.

Conservation does not mean stock-piling; it means wise use. Wise use implies intelligent management. Intelligent management presupposes a thorough knowledge of the biology of the species to be managed, but much remains yet to be learned about the wood duck. I wish, therefore, to contribute my iota toward an increased understanding and possible improved management of the one duck which truly belongs to the United

States.

The esthetic appeal of the wood duck dates back to the time of the

American Indian. Wilson (1854*600) reported that the skin of the head and neck of the wood duck was frequently used to ornament the Indian*s calumet or peace pipe. The feathers have been used for making artifi­ cial trout flies (Bent, 1923:168), and the gorgeous coloration makes the males desirable as hunting trophies. Almost without exception,

2 authors writing about the wood duck acclaim the male of this species

the most beautiful of all American ducks; its relative rareness con­

tributes further to the charm of its beauty. Certainly, the esthetic

importance of this duck is a value not to be ignored.

Wood duck management, however, commands a great expenditure of

public funds because of the bird's importance as a game bird and not

because of its esthetic value. In Alabama, Arkansas, and New Hampshire,

this species is the second most prevalent duck in hunters* bags; in

Georgia and Indiana it occupies third place; in Kentucky, Maine, New

York, North Carolina, Pennsylvania, and Tennessee.it holds fourth place; and in Delaware and Louisiana it occupies fifth place. In Louisiana an estimated 15,000-30,000 are killed each year. In Indiana the 1955 kill was estimated to have been 8,000; in Tennessee the 1955 kill was estimated at 6,000; in Maine the 1953 kill was estimated at 5,700; in

Washington the 1955 kill was estimated at 500.

Historical Account

The primeval American forest undoubtedly afforded abundant nesting places for the wood duck. Although the same habitat must also have been favorable for predators, the wood duck was pre-eminently successful.

According to P.L.Hatch (quoted by Bent, 1923:159), writing of conditions in Minnesota in 1892, "they pour in until every woodland pool has been deluged with them."

The wood duck was used for food by the American Indian (Goslin,

194-5), but it was not until the advent of market hunting by the white man that such use became significant. Dawson (1903:600) quoted Howard Jones as having stated that wood ducks were killed by wagonloads every

spring in the vicinity of Circleville, Ohio, 30 years before Dawson*s

book was written. Small ducks of wood duck size were sold by market

hunters for 15 to 30 cents each (Mumford, 1954:21). The birds were in­

discriminately killed at all seasons. The females were shot as they flew

to their nests, and the young in the flapper stage were captured with

dogs (Wright and Harper, 1913:490). As recently as 1901 there was no

closed season on the wood duck in Alabama, Arkansas, Georgia, Kansas,

Mississippi, Oklahoma, South Carolina, and Texas (Fisher, 1901:457).

Through the destruction of both habitat and breeding stock, the wood duck was reduced to near extinction by the beginning of the twen­ tieth century. In 1905, reports from ten observers recorded by Forbush

(1905:27) indicated that the wood duck was near extinction. These ten observers spoke of the status of the wood duck as follows: extinct, two; nearly extinct, five; decreasing, two; decreasing until last two years, one. In 1907, Butcher (1907:189) sent inquiries to ornitholo­ gists in different parts of North America, and all replies indicated that the wood duck was alarmingly scarce.

Finally in 1918, the wood duck was given full legal protection through the passage of the Migratory Bird Treaty Act. Article IV of the treaty with Great Britain provided that special protection be given the wood duck by a closed season extending over a period of five years or by such other means as might be deemed advisable. Full protection continued until 1941. Between 1918 and 1941, wood ducks were reared and released in some states, notably Connecticut (Frank, 1948:128); the

United States Biological Survey conducted experiments in the use of artificial nesting places by the wood duck. The species made an enor­

mous response to efforts to save it from extinction.

In 19-41, the season was opened on wood ducks in Alabama, Arkansas,

Delaware, Georgia, Kentucky, Lousiana, Maryland, Mississippi, Missouri,

North Carolina, Pennsylvania, South Carolina, Tennessee, Texas, and

Virginia, the season in these states being the same as for other water­

fowl, with a bag and possession limit of one bird. In 1942, the wood

duck was not mentioned in Regulation IV on open seasons, which means

that the season was nationwide; the bag and possession limit was con­

tinued at one bird. In 194-3, regulations were the same as in 194-2. In

1944, hunting was permitted in all states except Massachusetts and North

Dakota. Bag and possession limits in other states were continued at one bird. The 194-5 regulations were the same as those for 1944. In 1946,

Arizona, Colorado, Kansas, Nebraska, Nevada, Utah, and Wyoming were added to the states where these birds could not be legally taken. Bag and possession limits remained at one bird in other states. In 1947,

New Jersey and West Virginia were added to the states not permitting

shooting of the wood duck, with no change in the bag and possession limit in other states. In 1943, South Dakota was added to the states not permitting shooting, with the bag and possession limit continued at one bird in states where they could be legally taken. In 1949, New

Jersey was removed from the list of states prohibiting shooting; bag and possession limits continued at one bird. In 1950, regulations were similar to those in 1949* In 1951, Massachusetts and Nebraska were re­ moved from the list of states prohibiting shooting, with no other change in bag and possession limits. In 1952 and 1953, regulations pertaining to the wood duck remained the same as in 1951 • In 1954-, in the Atlantic

Flyway, the wood duck could be legally taken in all states, with one in

the daily bag and two in possession, except in New Hampshire where the

possession limit was one. There was no open season in any of the

states of the Mississippi Flyway. In the Central Flyway, with a bag and

possession limit of one bird, the season was open in all states except

Colorado, Kansas, North Dakota, South Dakota, and 'Wyoming. In the

Pacific Flyway, with a bag and possession limit of one, the season was

open in all states except Arizona, Nevada, and Utah, In 1955, the sea­

son was open in all states of the Atlantic Flyway, with a bag limit of

one and a possession limit of two birds. In the Mississippi, Central,

and Pacific Flyways, the season was open in all states, with a bag and

possession limit of one. In 1956, the season was again open in all

states of the Atlantic Flyway, with a bag limit of one and a possession

limit of two. It was closed in all states of the Mississippi Flyway.

In the Central and Pacific Flyways, the season was open in all states,

with a bag and possession limit of one bird.

Since 194-1, thousands of dollars have been spent for the erection

of nesting boxes and for research concerning the wood duck. Some nest­

ing boxes have been erected in all of the states within the wood duck's regular nesting range. Placing of nesting boxes has been particularly

important in Illinois, Massachusetts, Pennsylvania, Ohio and New York.

Forty-five hundred nesting boxes were erected in Massachusetts and 3,000 in Pennsylvania. Notable research on scientific management of the species has been done in Illinois, Iowa, Massachusetts, Maine, New York, and Vermont. It now appears not only that the wood duck can be retained as a member of our avifauna, but that it can also serve as an important

game species.

Summary

Destruction of habitat and market hunting had so reduced the wood

duck population that extinction appeared imminent in the first and second

decades of the twentieth century. In 1918, legal protection was enact­

ed, and notable increase of the population followed in the succeeding

two decades. In 194-1, the wood duck became legal game in some states, as it has been in all subsequent years. The species now makes an im­

portant contribution to the hunting bag in many states of the eastern

United States. A chronology of the hunting regulations is given. CHAPTER II

TAXONOMY AND DISTRIBUTION

Taxonomy

The family Anatidae contains the swans, geese, and ducksj the subfamily Anatinae contains the surface-feeding ducks which are also known as puddle ducks, dabbling ducks, and river and pond ducks. These ducks ordinarily feed by tipping rather than by diving. The hind toe of the Anatinae is slender and without a flap. Most of the Anatinae have a metallic speculum, a rectangular patch on the middle of the hind edge of the wing. When flushing, they spring directly into the air, rather than pattering over the surface of the water.

The generic name, Aix. is a greek word for a kind of waterfowl and was used by Aristotle. Aix differs from all other ducks in the silvery white edging of the primaries, the black and white terminal banding of the vermiculated flank feathers, and the broad, truncate tertials, the second of which is smaller than the third (Miller, 1925:

49). Sponsa is a Latin word meaning bride or betrothed. This refers to the brilliant plumage of the male, and actually it is the bridegroom who wears the bridal attire. The wood duck is the only North American representative of the genus Aix.

Synonymy. - The synonymy of the wood duck is as follows (chiefly from Phillips, 1925:^6):

,Anas sponsa Linnaeus, Systema Naturae, Ed. 10, vol. 1, p. 128, 1758.

Aix sponsa (Linnaeus), Boie, 0ken*s Isis, p. 329. 1828. 9

DencLronessa sponsa (Linnaeus), Swainson and Richardson, Fauna Boreali- Americana, vol. 1, p. 446. 1831, Lampronessa soonsa (Linnaeus), Wagler, Oken*s Isis, p. 282. 1832. Aia sponsa (Linnaeus), Eyton, Monograph of Anatidae, p. 120. 1838. Cosmonessa sponsa (Linnaeus), Reichenbach, Avium Syst., Natotores,

p. x. 1852. Aix promissa Brehm, Naumannis, (in synonymy of Aix sponsa) p. 298.

1855. Dendrocvgnus snonsa (Linnaeus), Brewer, Proc. Boston Soc. Nat. Hist., vol. 7, p. 308. 1860. Aex soonsa (Linnaeus), British Ornith. Union List British Birds, p. 123. 1883. Aix soonsa (Linnaeus), Allen, J.M. Bull. Amer. Mus. Nat. Hist., XXIII, p. 305. 1910. Vernacular Names. - In addition to the common name of wood duck, sometimes shortened to woody, Aix sponsa is also generally known as summer duck. Some names in local use are acorn duck (New Jersey,

Maryland, South Carolina), tree duck (Maryland, Michigan, Indiana), squealer (Arkansas, Louisiana), widgeon and wood widgeon (Connecticut), swamp duck (Louisiana), branchier (Louisiana), and black (Illinois).

Some book names are bridal duck, bride, bride duck, Carolina duck,

Carolina teal, crested wood duck, gray wood duck, king duck, rainbow duck, regal duck, whistler, and whistling duck (McAtee, 1923). In addition to names given by McAtee, Kortright (1942:221) reported the use of plumer and grey duck.

Distribution. - According to the American Ornithologists* Union

Check-list of North American Birds (1957:30), the wood duck breeds locally in almost every one of the United States and southern Canadian 10

provinces, from southern British Columbia, northwestern Montana, southern

Manitoba, and southern Ontario southward; it rarely breeds farther north.

It winters from southern British Columbia, central Missouri, southern

Illinois, and southern Virginia to Jamaica and central Mexico, rarely

north to Michigan and Massachusetts. It is casual in Bermuda. Wood

ducks have been found north as far as latitude 60° N. (Fisher, 1901:

455).

Up to 1939, the wood duck had been recorded in the Bird Lore

Christmas bird counts in 25 of the 49 states (Wing, 1947:39); Massachu­

setts, New York, Indiana, Illinois, Wisconsin, Minnesota, and Oregon

were the more northern states represented. Of the wood ducks recorded

on Christmas bird counts, by far the largest numbers were found on the

Savannah National Wildlife Refuge, Port Wentworth, Georgia. Of 5,249

wood ducks reported on all bird counts in 1941> 5,000 were observed on

the Savannah Refuge.

The wood duck is a fairly common breeding bird in all of the states

of the eastern United States west to the eastern edges of North Dakota,

South Dakota, Nebraska, Kansas, and Oklahoma. It occurs rather infre­

quently in the mountainous portions of the eastern United States. In

the western half of the United States, with the exception of Texas,

California, Washington, Oregon, and Idaho, the wood duck is rare as a

breeding bird. In Idaho the species is fairly common in the northern

part of the state; in Texas it is fairly common in the eastern part of the state. Distribution is further discussed in Chapter VII. CHAPTER III

BIOLOGICAL BACKGROUND

Plumages and Molts

Natal Down. - The upperparts of the downy duckling are dark brown, sometimes with a yellowish wash; the underparts are pale yellow; there

is a dark line back of the eye and a white line over the eye; the bill below and the tip above is yellowish.

By the time the ducklings are three weeks of age the postnatal molt has started, and feathers of the juvenal plumage commence to appear on the scapulars and flanks. Somewhat later, feathers appear on the tail, breast, and belly, later on the back and head, and finally on the wings.

The last of the down can be seen on the hind-neck and rump when the bird is almost fully grown.

Juvenal Plumage. - In the juvenal plumage the sexes are but slight­ ly different and somewhat resemble the adult female. In the juvenal female, the crown is dark brown; sides of head and forehead, dull grey, with white around the eye (Figure 1); stripe over eye and area between bill and eye, dull white; throat, white; upper surface of body, brown, with metallic lustre of purple, green and bronze; breast and belly whitish, mottled with dull brown; chest and sides, whitish, tinged with bright brown or buff.

The juvenal male differs from the juvenal female in having a greenish lustre to his brown crown, in having a less conspicuous white eye ring, in having the white of the throat extending in two bars up

11 Figure 1

The Immature Female Can Be Recognized

by the White Spot Around the Eye 13

into the cheeks and the sides of the neck (Figures 2 and 3), and in

having light brown rather than dark brown eyes.

Post.juvenal Molt. - Soon after the juvenal plumage has been attain­

ed, the postjuvenal molt commences to displace the juvenal plumage with

the first winter plumage. Following this molt, the mottled breast of

the young becomes white, the upper breast becomes rich chestnut, and the

sides of the body become vermiculatedj the adult head pattern is attain­

ed and much of the iridescence of the adult's back and scapulars is

acquired. In the first winter plumage the young males are essentially

similar to the adults, except that they are generally somewhat duller;

the lesser and middle coverts have less blue, and the tertials are

shorter* narrower, and dull brown rather than black glossed with metallic

blue.

Plumage of Adult Male. - In the adult plumage the male's head is

metallic green; the crest is iridescent green and bronze in front, dark­

ening to iridescent bronze, blue, and purplish black at the rear and

bordered below by a narrow white line; the sides of head have areas of

iridescent purple, violet, green and bronze; the nape is velvety black;

chin and throat, white; foreneck, white with two broad white bars ex­ tending upward, one ending below and behind the eye and the other ter­ minating in back of nape, a narrow white streak extending from base of

bill over eyes, into crest at nape; upper mandible, pinkish white in middle, bright red at base, with a narrow yellow border, and with ridge, tip, nail and lower mandible black; bill much shorter than head, narrow, height at base greater than width; nail, broad and sharply hooked; eye, bright orange red to vermilion; eyelid, orange yellow; back and rump, Figure 2

The Immature Male Can Be Recognized by the

"White Cheek Patch Figure 3

The White Cheek Patch of the Immature Male Is Sometimes

Relatively Indistinct dark bronzy green; scapulars, black, glossed with purple-green; upper

breast, glossy purplish chestnut, paler and spotted with white where it

meets the white breast; a vertical bar of white and one of black on

sides in front of folded wing; sides, pale buff, minutely vermiculated

with black; the longer side feathers broadly black and white; a large

patch of purplish chestnut on each flank near base of tail, the long

upper flank feathers blackish with a central orange-brown streak;

belly, white; feet, dull yellow to orange-yellow, with blackish webs;

claws, narrow, sharp and curved; upper surface of tail, black, glossed

with purplish green, bronze and blue; under surface of tail glossy

brown; upper tail coverts, black, glossed with purple and green; under

tail coverts, glossy dark brown; lesser coverts of wings, greyish

brown, with olive reflections; middle and greater coverts, glossed with

metallic green and purple and tipped with black; primaries, silvery grey

on outer edges, dusky on inner webs, with metallic-blue tips; secondaries,

with speculum of iridescent bluish green, bordered behind with narrow

black and white bars, in front with black bars, and proximally by rich

dark purple; inner tertials, black, with white tips, the outer ones,

black, glossed with metallic blue; wing lining and axillars, white, marked with dusky brown.

Plumage of Adult Female. - The adult plumage is acquired by the female following the postjuvenal molt. The adult female has a brownish- ash head; crown and crest, brown glossed with green; chin and throat, white; neck, brownish ashy; a white ring around eye tapering to a

streak behind eye; a narrow white ring around face near base of bill, sometimes faint or lacking; bill, bluish gray, ridge and nail, blackish 17 with white spot below nostril, edges pinkish; shape of bill same as that of male; eyes, blackish brown; eyelids, bright yellow; back, scapulars and rump, brown, glossed with bronze and greenish purple; upper breast and sides, olive brown, mottled with buff; breast and belly, white or creamy, the latter mottled with brown; feet, dull yellow with dusky webs; tail, olive brown, glossed with green; upper tail coverts, brown, sometimes faintly glossed with bronzy green; under tail coverts, white, speckled with greyish brown. The wings are similar to those of the male, except the colors are duller and less iridescent; the female has a small patch of reddish purple on inner greater coverts; the secondaries are more broadly tipped with white, mostly on the outer web; the tertials are brown with a bronze gloss.

Sequence of Molts and Plumages. - The winter plumage also con­ stitutes the breeding plumage and after the nesting season is lost with the postnuptial molt. The postnuptial molt starts in the male in June or July and continues into August. In this molt the flight feathers are all lost simultaneously, and the bird, now in the eclipse plumage, is unable to fly for some two weeks. The eclipse plumage re­ sembles the plumage of the juvenal male, except that the breat of the adult remains white and is not mottled as in the case of the juvenal; the back retains almost the full metallic lustre of the breeding plumage.

The molt out of the eclipse plumage into the winter plumage commences in August or September. In the early stages of the molt the green crown and crest colors develop; dark feathers begin to appear on 18 the scapulars, purple feathers on the upper chest, and the vermiculated feathers on the sides.

In August and September the females undergo the postnuptial molt.

A s with the male, the female loses all flight feathers simultaneously.

The female, however, has no eclipse plumage or double summer molt as has the male.

Field Identification Characteristics

With some practice, wood ducks can be readily identified in the field. When in flight, the relatively long square tail is perhaps the best field character for the species. The under surface of th-e wings is dark colored, and the belly is light colored. Because of its crest, the bird has a big-headed appearance which sometimes facilitates field identification. The white along the hind edge of the secondaries can usually be seen and serves to identify the species. The wood duck flies with its bill pointed downward, and it frequently turns its head from side to side.

When seen at close range, the adult male in winter, or breeding, plumage can be readily, distinguished from the adult female. When fly­ ing away from the observer, the back of the male appears somewhat darker than that of the female. When in flight and viewed from the side, the white vertical bars on the sides of the neck and the head of the male are the most easily seen distinguishing character. These white bars on the sides of the head and neck can also be used to distinguish the sexes of birds in the juvenal plumage when the birds are seen under favorable conditions. The white spot around the eye of the female can be seen 19 when the birds are viewed from the side.

Voice

Various moods of the wood duck are expressed by a variety of different calls, and to analyze the various calls and to correlate them with behavior is a major goal in itself. Since the study of the bird* s calls has been incidental to other objectives in this study, a comprehensive treatment is not possible at the present time, but it seems desirable to record the few observations which have been made.

Audubon (1843:274) stated that the male gives a guttural call when courting. Leopold (1951:212) reported that the male gives a low goldfinch-like call, sv-ee-ee-t. when he waits near the nest as the female is egg-laying. Peterson (1947:43) stated that the call of the male is a finch-like ieeee with a rising inflection. These authors may or may not have referred to the same call. The only call that I have heard which was positively known to have been given by a male was given by a caged bird. An audiospectographic analysis is shown in

Figure 4-A. This was a high pitched call audible only at close range.

It had a buzzy quality and lasted nearly three-tenths of a second. The frequency ranged from 3,500 to 5,000 cycles per second.

The call of the day-old ducklings sounds much like the call of other ducklings, such as the mallard. This call, commonly called a

"peep", is given chiefly when the young are separated from the brood or female, and it indicates some degree of distress. It can sometimes by induced by holding the ducklings by the feet, head downward. An audio­ spectographic analysis of the peep is shown in Figure 4-B» Each peep 20

was about 0.05 seconds in length. The peep started at 5,250 cycles per

second but quickly dropped to 4-,000 cycles. Through 0.025 o f a second

there was a buzzy quality in the peep which was between 4-, 100 and 5,100

cycles per second; then there was an upward slur reaching to 5,250

cycles per second. At the end of each peep there was a sharp downward

slur. Some of the peeps lacked the slight upward slur, but all were

concluded with a sharp downward slur. Six peeps were given per second

and 0.10 of a second elapsed between peeps.

Figure 4.-C is an audiospectographic analysis of a recording of a

peeping group of ducklings in a nest. A somewhat different pattern is

shown in these peeps than in those described above, but the analysis is

confused by the fact that the peeps of the individual birds cannot be

segregated with certainty. The single inverted V probably represents

the peep of one duckling. These peeps started at 2,000 cycles per

second, and in 0.025 of a second they increased 2,000 cycles or more

per second, falling again to 2,000 in 0.025 of a second.

"When calling the ducklings from the nest or calling the brood

together, the female gives a highpitched pe-e-e. This call was not re­

corded. When the female with young is disturbed, she gives a weak kuck-

kuck-kuck-kuck call, sometimes preceded by one or more hoo-eek calls.

The first syllable of this call was given at 1,600-2,100 cycles per

second and the second was slurred downward from 3,000 cycles per second

to 1,500 (Figure 5-A). This call lasted about 0.10 of a second. The kucks were between 1,500 and 2,000 cycles per second and lasted 0.025 of a second. A more protracted introductory call which lasted 0.20 of a second was sometimes given (Figure 5-B). This note did not close with Figure A

Audiospectographic Analyses of "Wood Duck Calls.

A, call of caged male; B, peep of day-old duckling;

G, peep of unknown number of ducklings in nest re EO CD 00 a ro ro rr CD— ro ro CD oo - e r FREQUENCY IN KILOCYCLES PER SECOND CD 00

o i\d o o ho 0.4 0.6 0. oo TIME IN SEC O N DS 23

Figure 5

Audiospectographic Analysis of Calls of Female:

A, when disturbed with youngj B, when disturbed with young;

C, caged female i ! { \ t t j I i j ■| ■| Hjj . / ..... - > - i L_ m X m \ M M . — [Y> OJ i. - — - i. __ D c 1 oj i i k k • 1 ■=. ■=. -JF, JP A — — no J : \ —J ______r ,

______o ^ u _ . . ,i M FREQUENCY IN KILOCYCLES PER SECOND - 1 - ' ~

■ % - - 3 E m -w 5 ■ ■ ro ro O 00 4^ o O TIME IN SECONDS 25

the same sharp downward slur as the above described note. The kuck

calls were more sharply slurred downward. Figure 5-C shows an audio-

spectograph of the call of a caged female. This call extended through

0.10 of a second and was of a frequency from 1,000 to 2,000 cycles per

second.

When disturbed on the nest, females gave a hissing call. Accord­

ing to Sibley (1955), hissing calls of nesting birds tend to protect

the nests through the resemblance of the birds* hiss to that of a

snake.

When coming into a loafing, feeding, or roosting area, females

usually uttered a querulous what. When at the loafing place, a slight­

ly higher pitched quov-e-e-e was frequently uttered. This call had the

quality of the killdeer*s call, but it was slightly lower in pitch.

Nest-hunting females give a te-te-te-te call. The most commonly heard

call of wood ducks is the hoo-eek call of the female given when the bird

flushes.

Growth

The ducklings at hatching, weigh slightly less than one ounce

(28.35 grams). Five ducklings which were weighed when dry soon after hatching showed a mean weight of 24.7 grams and ranged from 23.9 to

25.7 grams. These weights were taken at about 10:00 P.M., and when the same ducklings were weighed a short time before they left the nest the following morning at 8:00 A.M., they showed a mean weight of 23.5 grams, or an average of 0.8 of a gram per bird less than ten hours earlier. 26

A duckling which hatched from an egg in another nest weighed 26.5

grams soon after hatching and declined to 25.6 grams through the follow­

ing ten hours. This duck was reared in captivity, and its growth rate

is shown in Figure 6. An effort was made to capture marked wild duck­

lings to make comparisons of the growth rate at different stages of development of wild- and captive-reared birds, but none of the marked ducks was captured before the flying age had been attained. One marked bird was captured 53 days after hatching. The wing feathers of this bird were sufficiently developed (Figure 7) that flight was possible.

The development of the flight feathers of the captive-reared bird is

shown in Figures 8, 9, and 10.

When the wild-reared bird was first weighed at three days less than eight weeks of age, it was well filled with corn and weighed

487.1 grams. Twelve hours later it weighed 442,5 grams, though the crop still contained some corn. The captive-reared bird at a comparable age, weighed 434.7 gramsj there was little or no food in the crop when the captive-reared bird was weighed. These two birds showed closely similar weights at comparable ages, and if the amounts of food in the alimentary tracts of each had been identical, their weights presumably would have been still more nearly similar. While it cannot be known whether the growth of these two birds followed the same curve, they attained approximately the same weight by the time the ability to fly was attained at about eight weeks of age. Photographs of the captive- reared duck are shown in Figures 11-17. Figure 18 shows the wild-reared bird at fifty-three days of age. The backgrounds are graduated in inches. Weight (grams) 200 300 400 100 Growth Curveof Captive-reared Wood Wood Duck 3 Figure6 g (weeks)Age (

5

6 74 27 8 Figure 7

Flight Feathers of Wild-reared Wood Duck at Eight Weeks of Age m Figure 8

Flight Feathers of Captive-reared Wood Duck at Six Weeks of Figure 9

Flight Feathers of Captive-reared Wood Duck at Seven Weeks of Figure 10

Flight Feathers of Captive-reared Wood Puck at Eight Weeks of Age Figure 11

Captive-reared Wood Duck at One Week of Age Figure 12

Captive-reared Wood Duck at Two Weeks of Age Figure 13

Captive-reared Wood Duck at Three Weeks of Age Figure 14*

Captive-reared Wood Duck at Four Weeks of Age Figure 15

Captive-reared Wood Duck at Five Weeks of Age Figure 16

Captive-reared Wood Duck at Seven Weeks of Age Figure 17

Captive-reared Wood Duck at Fifty-three Days of Age Figure 18

Wild-reared Wood Duck at Fifty—three Days of Age Weights

Thirty-one adult male wood ducks weighed at different times of

day during the spring and summer showed a weight range from 531.1 to

836.3 grams; 33 adult females showed a range from 494.1 to 737.1 grams

(Table I). Weights of 248 males recorded by Nelson and Martin (1953:

39) varied from 680.4 to 907.2 grams; 163 adult females varied from

635.0 to 907.2 grams. The data used by Nelson and Martin were presum­

ably weights of wood ducks in hunters* bags and were thus from fall and

winter birds. This being the case, wood ducks are apparently heavier

during the late fall and winter months than during the summer and

early fall.

The body weights of 106 immature wood ducks showed considerable

variation (Table II). Many of these weights were taken during Sep­

tember when it might be expected that not all of the immature birds

would have attained fully adult weights, and comparison of 95 September

immatures with 37 September adults gave a t value of 6.501 when 1.979

is significant at the 5 per cent level. There is a significant differ­

ence in September between weights of birds of the year and older birds.

Not all of the September immatures, however, weighed less than some

fully adult birds, and the mean weight of 95 immature birds was only 56

grams below the mean weight of 37 adults. The immatures consisted of

birds of the year which were already able to fly.

In ruffed grouse, Bump efc al. (1947:95) reported that two-* to three-

year old birds weigh more than first-year birds. The age in years was not determined of any weighed adult wood duck, and it cannot be stated whether a similar condition exists in this species,

40 Table I

Weights (in Grams) of 64 Adult Wood Ducks

Mean with Month No. Individual standard Standard Coefficient of Weighed Sex weighed weights Range error deviation variation Male 5 581.1-737.1 626.2^28.18 63.0 10.24 Anril Female 5 /.QA.1-7PP.9 61 5. . 1 5 70.7 11.49 .

May Female 12 502.0-637.8 575.4-12.5 43.3 7.49 Male 1 83b. 3 July Female 1 737.1 Male 1 722.9 August Female 1 654.9 Male 23 595. >816.0 692.5-11.01 52.8 7.73 September Female 13 609.5-680.4 649.7- 6.96 25.1 3.86 Male 1 779. b October Table II

Weights (in Grams) of 106 Immature Wood Ducks

Mean with Month No. standard Standard Coefficient of weighed I Sex weighed Range error deviation variation Male 5 524.4-708.7 601.2-31.08 68.4 11.37 August Female 6 473.6-660.5 555.2- 2.79 67.0 12.12 Male 41 524.5-751.2 629.1- 6.38 41.5 6.39 September Female 54 289.0-669.0 607.2- 9.29 67.6 11.13 43

There is some evidence (Table III) that immature birds may have

a greater food capacity than adults. Three immature birds contained

food amounting to 17.0-21.8 per cent of their total weights; one seem­

ingly well-filled adult female contained food amounting to only 5.4- per

cent of her body weight. Stoner (1936:158) found that crops of young

bank swallows have a greater food capacity than crops of adult birds.

An important part of the observed variation undoubtedly depended

upon the extent to which the alimentary tract contained food, sand, or

other materials. One very small immature female contained 62.9 grams

(Table III) of material in her alimentary tract or 21,8 per cent of her

total weight. Another had in her tract material weighing 98.9 grams or

20.4- per cent of her weight. If the birds were trapped in the morning

before they had opportunity to feed, a slightly lower weight was found

than when they had ample opportunity to feed. Fourteen birds kept in

captivity with access to water but without access to food through 12

hours lost an average of 4-8.7 grams or 7.7 per cent of their maximum weight. An incubating female known to have remained on the nest through

24- hours lost 26.9 grams of weight, IJhen the birds were taken in traps

containing a plentiful supply of food, however, the amount of ingested food varied but slightly because the birds normally gorged with food when in the traps only a short time.

Most of the wood ducks weighed were taken in banding traps set where food had been available through three or more weeks before the birds were captured and weighed. Time and food were ample for the birds to have accumulated some body fat in connection with the abnormal food supply available in the traps. If there was a relative shortage of food Table III

Food in Digestive Tracts of Wood Ducks

3ody weight Per cent of before corn weight of From From Total c o m jas removed evening From crop .proventricuius gizzard per bird (grams) trapped birds No. No. (Whole Whole composed of Grams Grains Grams Grains Iranis food Immature Female September 4-1.5 86 10.0 16 11.4 12 62.9 114 289.0 21.8 13, 1956 Immature Female September 62.5 131 26.2 49 10.2 10 98.9 190 484.3 20.4 13, 1956 Immature Female September 51.2 104 23.0 44 3.0 1 77.2 149 453.6 17.0 13, 1956 Adult Female September 23.8 44 6.2 11 7.0 4 37.0 59 669.0 5.4 13, 1956

£ 45

before the birds found the baiting station, and if various birds first

started visiting the baiting station at markedly different times, the

amounts of fat deposition would vary with durations of the periods the

different birds were exposed to an abnormally large food supply.

Variations in fat deposits could be associated with seasonal

availability of food or other factors of a seasonal or periodic nature.

Great variation in body weight, however, was found within single sea­

sons. If availability of food and consequent variation of fat deposits

is an important factor explaining the striking weight variation, it

affects individuals within the seasons and does not appear as a seasonal

factor.

Incubation by females was suspected of being associated with a weight decline, and a significantly smaller weight was shown by females

weighed in May when weights of May females were compared with 14 Sep­

tember females. A t value of 2.147 resulted, 2.064 being significant

at the 5 per cent level. Miller (1950:5) observed that incubating fe­

males weighed approximately one-fourth of a pound less than fall-trapped females.

An extremely wide range of weights is shown among five April-trap­

ped females. The range, 496.1 - 722.9 grams, may have been partly the result of weighing both incubating and non-incubating females. Obvious­ ly these two classifications are not clear-cut: an incubating bird taken from the nest may recently have returned from feeding, or a bird taken in a baited trap may have just left the nest and not ingested food before being weighed. The light weight of incubating females may be associated with a general relative scarcity of food rather than with the fact that 4.6

the incubating bird does not feed regularly.

Comparison of 23 September males with 5 April males gave a t value

of 3.056 when 2.056 is significant at the 5 per cent level. Comparison

of weights of 23 males and 14. females trapped in September gave a t

value of .723 with 2.030 being significant at the 5 per cent level.

Males are significantly lighter in weight during the spring months than

during the late summer; there is no significant difference between the

weights of the sexes in September.

Only one molting bird was weighed, and the relationship of molt­

ing with the body weight cannot be determined. This molting male was

the heaviest of 170 wood ducks and weighed 836.3 grams.

As with adults, immature birds of the two sexes are not significant­

ly different in weight. A t value of 1.270 was obtained in comparison

of 4-1 immature September males and 54- immature September females. For

the appropriate degrees of freedom, a t value of 1.987 is significant at

the 5 per cent level.

The heaviest birds of both sexes were weighed in July. As far as the seasonal weight cycle can be followed, it appears that the minimum weight is obtained in the spring months; there is a gradual increase in weight until a peak is reached during the early winter months.

The ultimate practical purpose of a knowledge of the average body weight of a game bird is the possible application of weighing for an in­ dex to the general health of the species. The lightest wood duck which had attained the power of flight weighed only 289.0 grams. This bird died in a burlap bag where it was kept overnight. The contents of its alimentary tract weighed 62.9 grams, or 21,8 per cent of the bird's 47

total weight. The body of the bird devoid of food weighed only 216.1

grams, which is about the weight of a large, well-filled bob-white.

This small wood duck was posted for parasites along with five other

birds taken at the same time. The small bird contained a heavier in­

festation of parasites than the larger birds: tapeworm, 1; slender

duck louse, 3; common chicken louse, 2; louse, Menacanthus sp., 1.

Only one or two parasites or none were found on the five other ducks.

The part parasites played in the underweight condition and death of the

bird is not known, but it can be expected that a relationship existed.

Unfortunately, available data on weights of wood ducks are too

meager to permit the construction of curves of diurnal rhythm. It can

be inferred that the diurnal curve would show morning and evening peaks

as was described by Stewart (1937:328) and Baldwin and Kendeigh (1938:

426) in other species, except that the ascending curve should be

steeper if ample food is available. After reaching the flying age,

wood ducks feed chiefly or entirely during the morning and evening, and

this should result in daily morning and evening peaks with a gradual

decline through the periods between feeding.

Body Temperature

Through the use of a Taylor mercury glass thermometer, body temper­

atures were taken of 143 adult and immature wood ducks. The thermometer used was calibrated in degrees centigrade, and fractions of a degree were estimated. This thermometer was of a type in which the mercury went up or down with the temperature, and the maximum reading was taken as the thermometer protruded from the bird. The thermometer was intro­ 43

duced into the cloaca of the bird, and it was left inserted at least

one minute after the maximum reading was shown. A Leeds and Northrup

galvanometer was also used for the first few temperature readings, but

because of hazards associated with transportation of this equipment

into the field, its use was abandoned in favor of the glass thermometer

after closely similar readings were observed.

A range of 4.0.0°C. to 44*9°C. was found in the temperature deter­

minations of 143 wood ducks. In studies of the bank swallow, Stoner

(1936:174) noted individual variation amounting to 11.1°C. He observed

a rise of 5.1°C. in ten minutes in one adult individual under approx­ imately constant external conditions; a drop of 4«3°C. occurred in an­

other bird in eight minutes. These variations presumably stemmed from

individual differences in body metabolism before and during the time

the temperatures were taken. In the present observations of wood duck temperatures, maximum readings were taken, but declines were sometimes noted after the maximum readings had been made. The maximum drop noted was 1.5°C., and it is believed that an individual of the much larger wood duck does not ordinarily show such a wide temperature range as quickly as the bank swallow. This variation imposed by the individual differences in current metabolic activities, however, is present with the wood duck data presented here and may be the chief factor in causing the noted variation.

When the temperature of the individual fluctuates so much in such a short time, it is difficult to be certain that the maximum point was observed. Nor can it be known that the bird reached its maximum point when its temperature was being taken. On the other hand, it may be that the maximum temperature reached by a bird in the hand is an abnormal

temperature because the bird may be disturbed rather than behaving

naturally. The use of standard temperature (Baldwin and Kendeigh 1932:

22-23), or the temperature of basal metabolism, would alleviate con­

siderable variation, but the time involved in its determination and the

fact that the ducks nearly always contained well-filled digestive

tracts, made its use infeasible in this study. Usually when the tem­

perature of a bird in the hand was being taken, the mercury immediately

rose to a level where it remained constant, at least through several

minutes. This is the source of the temperature readings used in the

accompanying table (Table IV).

Unfortunately, only a small sample is available for comparison,

but available data suggest that incubating females maintain a slightly

lower temperature than active, non-incubating bird3. Five females

which were taken from nests showed a mean temperature of L, 1.6°C. while

three taken from banding traps the same season averaged 43.0°C, One

reading in each of the incubating and non-incubating groups was of the

same bird. This decline of temperature may be associated with de­

creased muscular activity when the bird is incubating. Stoner (1936:

163) demonstrated that muscular activity causes an increase in the body

temperatures of adult bank swallows.

Baldwin and Kendeigh (1932:106) observed that temperature control

of newly hatched chicks was less perfect than in chicks three days of

age. Similarly, temperature control of newly hatched ducklings is in­

completely developed. Eight ducklings removed from their nest and their temperatures taken immediately had an average temperature of 36.9°C. in Table IV

Body Temperatures (in °C.) of 143 Wood Ducks

Mean with standard Standard Coefficient of Month Aee Sex Number Ranee error deviation variation Male 2 42.2-42.8 42.5-.429 .600 1.41

Adult Female 3 1 2 , 8 - 1 3 . 3 A3.0i.2A5 .600 1.14 April Immature Adult Female .. 5 . A1 .1-42.1 A1.6i.195 .390 .93 May Immature Adult Male 2 4.1.1 July Immature Adult Male 3 42.5-43.5 43.li.249 .424 .98 Aueust Immature Male 22 41.0-44.9 42.7i.065 .308 1.5 Adult Female 20 A1 . 2 - 1 3 . 2 A2.8i.073 .329 .88 Male 40 41.2-44.0 42.6-.949 .600 1.47 September Immature Female A2 40.0-44.0 A2.A-.00A .240 .56 Adult October Immature Female ... ,4 ... A2.0-A3.1 A2.6i.265 ...... 530 ..... 1.24 51

an air temperature of 18.0°C.j after 30 minutes their average tempera­

ture was 32.2°C.j after 45 minutes their average temperature was 31.8°C.

The birds were then moved into an environment with an air temperature of

7.2°C. and held there through 15 minutes. At the end of 15 minutes,

the following readings were taken: 31.7°C., 31.9°C., 31.1°C., 30°C.,

31.1°C., 31.7°C., 31,9°C., 29.5°C. At about 24 hours of age, these

ducklings exhibited a control which tended to stabilize their tempera­

ture between 29.0°G. and 32,0°C.

Temperature control at two weeks of age was much better developed

than in newly hatched young but not sufficiently effective to maintain

the body temperature unchanged in a suddenly colder environment. A duck two weeks old with a temperature of 41»1°C. was taken from an air

environment of 24.5°C. and placed in the freezing compartment of a

refrigerator at —2°C. When removed at the end of five minutes, the

duck*s temperature was 40.8°C. It was returned to the refrigerator, and at the end of another five minutes its temperature was 40.0°C.

Locomotion

Swimming speed. - It is difficult to obtain data on maximum

swimming speeds of wood ducks because when pressed the flightless young either dive beneath or skitter over the surface of the water, and the fully feathered ducks take flight. As would be expected from its slowness, swimming alone is used for only leisurely escapes.

Three measurements of swimming speeds were obtained for day-old ducklings. For these measurements a lane was constructed of two pieces of chicken wire placed parallel in a vertical position 6 inches apart and extending from the shore toward the center of a pond. This lane

was 20 feet long and wire extended 4 inches above the surface of the

water. The ducklings were released near the shore into the lane, and

the time they spent in swimming at an approximately uniform speed was measured with a stop watch. When the birds ceased swimming and dived or skittered over the water, the watch was stopped and the swimming distance was measured. As the measurements were taken immediately be­

fore the ducklings resorted to a more 3peedy method of escape, it is believed that approximately maximum swimming speeds were recorded. The

speeds of eight ducklings ranged from 0.8 to 1.2 feet per second with an average of 0.9 foot per second or 0.6 of a mile per hour. Hochbaum

(1944*120) reported that flightless adult canvasbacks swim at a rate of

2-3 miles per hour.

Diving. - When pressed immediately on leaving the nest, the duck­ lings dived and swam under water for short distances. The length of time four ducklings just out of the nest remained under water was measured with a stop watch; the periods of submergence varied from 7 to 13 seconds.

A duckling one day old swam under water 16 feet in 7.5 seconds. This bird swam an average of more than two feet per second when under water.

By the time the ducklings were three to four weeks of age, they swam submerged through much greater distances. Unsuccessful attempts were made to determine the time spent and the distance moved under water t>y the ducklings at more advanced ages. Ducklings more than two to three weeks of age seldom reappeared on the surface of the water before reach­ ing protective plant cover. Spots 75 feet from the nearest shore were chosen for the release of several ducklings three to four weeks of age 53

in an effort to measure their time and distance under water, but in

each case the ducklings were not seen again. "When ducklings were re­

leased near the shore, they commonly dived, swam a short distance,

came out on the shore, and ran into vegetative cover. Presumably the

birds which were not seen after being released 75 feet from shore swam

all of the way to plant cover before they emerged from the water.

Skittering Speed. - Skittering is accomplished by the duck elevat­

ing the anterior part of its body higher than in swimming and by running,

with much splashing, over the surface of the water. One measurement of

the skittering speed of a day-old duckling was made in a method similar

to that of determining swimming speeds. This duckling skittered at the

rate of 8.5 feet per second or 5.8 miles per hour for a distance of 17

feet. The skittering speed of an adult male in the flightless stage

of the postnuptial molt was 75 feet in 5.4- seconds, a rate of 14.1 feet

per second or 9.6 miles per hour. This is faster than the maximum

terrestrial speed of 13.3 feet per second or 9.1 miles per hour for

the red squirrel reported by Layne and Benton (1954.: 103). Hochbaum

(1944-:120) reported a skittering speed for flightless male canvas-

backs of 8-10 miles per hour.

Running Speed. - Tests of running speed were made on an almost

level plot of closely grazed pasture land. A liberal application of

paint was smeared on the bottoms of the ducklings* feet before release.

Measurements of strides included the distance from the anterior end of one track to the anterior end of the next track. The time travelled

in an approximately straight line was measured with a stop watch. In spite of the fact that ducks in general do not'appear well adapted for 5 4

running, young wood ducks can run rather rapidly, and one bird three

to four weeks of age ran 10.4- feet per second for 24 feet; during the

first several yards, which are not included in this calculation, its

strides were only about 5 inches long, but after running several yards

its strides lengthened to 8 inches. Three other ducklings three to

four weeks of age each ran 8 to 9 feet per second. The strides of two

of these ducklings were 7 to 8 inches long, but one duckling made

strides 12.5 inches long. This duckling did not move correspondingly

faster and covered only 8 feet per second. These ducklings made 7.7

to 15.4- strides per second, or each stride required one-eighth to one-

fifteenth second of time. They ran at the rate of 5.5 to 7.1 miles

per hour. The maximum observed terrestrial speed of a wood duck,

a bird six to seven weeks of age, was 12 feet per second or 8.2 miles

per hour.

The maximum running speed of 8.2 miles per hour is not markedly

lower than running speeds recorded for some terrestrial species of

birds. Cottam, Williams, and Sooter (1942:131) recorded maximum running speeds of 10 miles per hour for three ring-necked pheasants, and 12 and 15 miles per hour for two road-runners. Even among mammals, maximum running speeds of only 10.6 miles per hour for the chipmunk,

9.1 for the red squirrel, and 17.0 miles per hour for the gray squirrel have been reported by Layne and Benton (1954:103).

Flight. - The flight speeds of eight wood ducks, two immature males and two immature females, over a distance of 204 feet was measured with a stop watch. The watch was started when the birds had flown approxi­ mately five feet after being released. There was little or no wind when 55 the measurements were made, and the birds flew almost horizontally.

The flight speed of these eight birds averaged 4-5.8 feet per second or 31.0 miles per hour with a range of 27.8 to 34-•8 miles per hour.

Because of variations under which the birds were flying as well as variations in making measurements, it is difficult to find published data on flight speeds for comparison. Speeds from 26 to 72 miles per hour have been recorded (Cooke, 1937) for various species of ducks.

McLean (1930:1-2) noted that the speed of a Cinnamon Teal increased from 32 to 59 miles per hour when the bird was chased. A higher speed could probably be attained after flight had been in progress for a longer period of time than was the case in this experiment with wood ducks.

Motion pictures were taken of four flying wood ducks in order to determine the number of wing beats per unit of time. These birds were followed with the camera several hundred feet after they were re­ leased. The pictures were projected at a slower speed and the wing beats of the birds on the screen were counted. These four birds made

7 to 7.5 wing beats per second. Aymars (1935:144-) reported that the golden-eye makes nine wing beats per second. At 31.2 miles per hour, the wood duck moves approximately 6.5 feet at each wing beat.

Entering the Nesting Cavity. - I talked with many farmers and hunters in connection with my studies of the wood duck in central Ohio,

Immediately after I had mentioned the wood duck, a usual comment by these persons had to do with the manner of entrance of the female into the nesting cavity. I was repeatedly told that the wood duck flies directly into its nest at a high speed and that it stops by crashing against the inside of the nesting cavity opposite from the entrance.

This widespread misconception had apparently been the result of a tele­ vision program from Columbus, Ohio. Forbush (1912:110) also reported that when the entrance hole is sufficiently large to accomodate her easily, the female appears to fly directly through the hole, "striking the plumage of her breast against the lower edge of the entrance to break the force and speed of her descent."

Bellrose (1955b:30) stated that the female flies directly into a house with a large opening, rather than cling.ing momentarily to the opening. Gigstead (1932:605) observed that the wings closed just as the duck arrived at the entrance to the cavity. Leopold (1951:214) observed that the female brakes so close to the nest that it is sur­ prising she can stop. In more than 20 observations, Dreis (1951:25) noted that the female perched at the bottom of the entrance using her tail as a prop before entering the cavity. The female enters the nest so quickly that details of her movements are not readily observed, but seemingly she enters with little loss of the horizontal position of the body in flight. When the bird enters a small cavity, or when she has been disturbed, she normally pauses at the entrance of the box, perching woodpecker-fashion at the opening, before she enters. Except that she behaves differently when disturbed, the problem might be further eluci­ dated through the use of high-speed photography.

Malformations

Abnormal feet were often noted on trapped wood duck. The abnor­ mality consisted chiefly of missing parts, parts of toes, holes through the webs, and in one case a female had lobed rather than webbed feet.

Some of these abnormalities undoubtedly resulted from mechanical

injuries. Disease seems a more probable cause of loss of the webs from

the feet, but no actively diseased feet were found.

Life Expectancy

Using the method demonstrated by Hickey (1949), data from 2,458 wood ducks were used in the construction of a dynamic life table

(Table V). Because of my inability to get permission the recoveries from Illinois were not used. Of young birds banded during the summer months, only those which survived until September 1 were used.

The life expectancy indicated by this table is such as would be duplicated in another period under similar conditions. Because of changes from year to year in the hunting regulations, it could not be anticipated that the life expectancy values in another period of equal length would show a similar distribution. These data extended only to January, 1957; since most of the birds were banded after 1950, the number of birds in the upper age groups is disproportionately small.

Of the 2,458 wood ducks in the sample, 69.3 per cent were shot within the first year after banding. After the first year, the mean annual mortality rate was 35.7 per cent. The mean annual mortality rate for all years was 38.3 per cent. One bird lived to age 13-14 years. 58

Table V

Dynamic Life Table for Wood Ducks

Mean Reported Alive at Mortality mortality Age shot start of year rate per year rate

0-1 1714 2458 69.3 69.3 1-2 418 744 60.6 2-3 179 326 54.9 3-4 66 147 44.9

A-5 45 81 55.5 5-6 19 36 50.0 6-7 5 17 29.4 35.7 7-8 6 12 50.0 /

8-9 4 6 33.6 9-10 0 2 10-11 1 2 50.0 11-12 0 1

12-13 0 1 13-14 1 1 100.0

Total 2458 3834 *38.3%

*mean mortality rate per year Stunmarv

The molts and plumages of the wood duck are described, and field

identification characteristics are given.

Some of the calls of the wood duck are described, and audiospec-

tographic analyses of several calls are presented. Calls of the male and young have frequencies up to about 5,000 cycles per second; calls of the female have frequencies up to 3,000 cycles per second.

At hatching, ducklings weigh slightly less than one ounce. The ability of flight is attained at about eight weeks of age when the body weight is 4-50-500 grams. Adults are lightest in weight during the spring and heaviest during the winter, but a wide range in weight can be found in a single season. As much as 21.8 per cent of the total body weight consisted of food in the alimentary tract. The crops of young birds may have a greyer food capacity than those of adults.

There appears to be no sexual difference in weight.

Body temperatures from 40.0°C to 44-.9°C. were found among 143 wood ducks. Variation in the readings probably resulted from differ­ ences in metabolic activity. Incubating females showed a slightly lower body temperature than trapped birds. Temperature control in newly hatched ducklings is imperfectly developed, but their temperatures tend to stabilize between 29.0°C. and 32.0°C.

Day-old ducklings swam at a rate of 0.6 of a mile per hour. Day- old ducklings dived and remained submerged as much as 13 seconds. One day-old bird swam at a rate of about two feet per second when under water. A day-old duckling skittered over the water at a speed of 5.8 miles per hour; a flightless adult male skittered at a speed of 9.6 miles

59 60 per hour. A duck 6-7 weeks of age ran 8,2 miles per hour. Strides were as long as 12.5 inches. The maximum measured flight speed was

34.8 miles per hour. Flying birds made a maximum of 7.5 wing beats per second, moving approximately 6.5 feet per wing beat.

The female sometimes alights, woodpecker-fashion, on the entrance of the nesting cavity, but she usually enters in the horizontal posi­ tion of flight.

Ducks with missing toes and torn webs were frequently found.

Of 2,4-58 banded wood ducks, 69.3 per cent were shot within the first year after banding. After the first year the mean annual mor­ tality rate was 35.7 per cent. The mean annual mortality rate for all years was 38.3 per cent. One bird lived to age 13-14 years. CHAPTER IV

REPRODUCTIVE BEHAVIOR

Courtship

According to Forbush (1912:108), wood ducks pair during the months

of December, January, and February in the South. Audubon (1843:272) re­

ported that pairing takes place about March 1, or sometimes a fort­

night earlier in Louisiana. Bennett (1938:40) reported the blue-winged

teal to be courting by the end of December in Mexico. Lincoln (1935:

27) reported that mallards and black ducks occur in pairs as early as

January. Cooke (1906:12) reported that mating begins in North Carolina

by February 15, among mallards, black ducks, gadwalls, baldpates, green-

winged teals, blue-winged teals, shovellers, and pintails.

I spent the period of January 2-9, 1957, observing wood ducks on

the Savannah National Wildlife Refuge, Port Wentworth, Georgia. During

this time the ducks moved about with seeming randomness in regard to sex

companions. No evidence of heterosexual pairing was noted.

The courtship performance has been described by Audubon (1843:

273) and by Townsend (1916:15-16). It is described as taking place on

the water, and it is said to start by the male swimming close to the fe­ male. As they swim along side by side, occasionally the male touches

the head of the female with his bill. The two then stop, the male jerks

his head up and down several times, and at the same time he utters a low

whistle. The crest of the male is elevated from time to time. 6 2

Schreiner and Hendrickson (1951:6) observed wood ducks courting on

April 22, 1950, in Iowa. Five females were swimming in a group which was surrounded by five drakes. The courtship display of the drake con­ sisted of raising and lowering the crest while arching the neck upward and then lowering the head quickly. The drakes frequently made menac­ ing dashes at each other, but these dashes ended short of actual contact.

I observed what was presumably an abbreviated courtship performance at Buckeye Lake, Ohio, on April 20, 1956. The male and female stood side by side on a stump protruding from the water, and each took turns in series of head bobbings. The head bobbings consisted of forward swings of the head so that the head and neck were brought into a plane nearly horizontal with the body. The nesting status of these birds was not known.

Phillips (1925:70) reported that the male wood duck is more or less polygamous when in captivityj the females are probably monogamous.

The nature of the mating habits of wild-breeding birds has not been fully ascertained. The fact that birds pair in the South and the fe­ males return to former hatching and nesting places suggests that poly­ gamy is improbable.

Selection of the Nesting Cavity

In March, soon after wood ducks arrive from the South at their northern breeding grounds, they can be seen in the trees exploring potential nesting cavities. In Iowa, Leopold (1951:211) noted that appearance at the nesting site normally was delayed several days to a week after the birds arrived in the vicinity. 63

In the breeding season, the female usually takes the lead as the pair moves about, and this is particularly true during cavity explora­ tions. Sometimes two or more pairs go together on the nest-hunting foray, and the female of each pair is followed by her mate. In Cali­ fornia, Dixon (1924.! 46) reported that the male takes the lead in ex­ ploring for cavities and that he enters the cavities first, though the female always makes the final selection. In all cases observed of wood ducks nest-hunting in Ohio, the female took the lead and entered the cavities, and Bellrose (1955b:6) reported a similar condition in

Illinois. After alighting on a horizontal limb of a tree, the female frequently stretches out her neck in various directions in a plane horizontal with the body as she appears to be searching for a cavity.

Having found a cavity, she moves to it and explores it from the out­ side, and later she may enter it. Sometimes the pair flies away with­ out the female having entered the cavity, and the later failure of occupancy of such cavities suggests that they had been judged unsuit­ able. Most of the nest hunting is done in the morning before 9:00 A.M., and the pair sometimes continues to return to the tree daily in the morning, sitting in the tree and entering the cavity. The pair spends much time on limbs outside the cavity.

Nesting Territory

Wood ducks are essentially non-territorial, and the male does not defend an area around the nest against the approach of other ducks of the same species. When the female is incubating, the drakes gather into small groups and spend the days together some distance from their nests. 64

The female is sometimes accompanied by several males when she returns to the nest. The males may fly away after the female has entered the nesting cavity or they may remain together on water near the nest.

Females frequently enter and lay eggs in wood duck nests other than their own, apparently without being disturbed by the original occupant of the nest, Bellrose (1943) photographed two female wood ducks incubating side by side in the same cavity. This, of course, does not in itself indicate that wood ducks are non-territorial. Hoyt (1948:

195) observed two mallard ducks with nests side by side, but both ducks were attended by the same drake. Hochbaum (1944:93) reported two shoveller nests only 9 inches apart.

Egg-laving

Soon after mid-March in central Ohio, some of the female ducks start laying eggs, even though they may not yet have found a nesting cavity to their liking. The eggs that are laid early are sometimes dropped on the ground, but they are usually laid in a cavity of a tree or nesting box. As many as several dozen eggs may be laid in one cavity.

Such nests are known as Mu m p ” nests, and they can be recognized by the fact that the eggs are left without being covered, and down is not intro­ duced. Usually the eggs in dump nests are never incubated, but some­ times a female will take over the nesting cavity, add her eggs, and in­ cubate the dumped eggs along with her own eggs. Such nests can be recog­ nized by differences in the size and shape of the eggs (Figure 19).

Bellrose (1944) reported a case of a sparrow hawk incubating wood duck eggs, and this was presumably such a dump nest. ’Wood ducks also Figure 19

Differences in the Size add Shape of the Eggs Indicate that Single Clutches Are Sometimes Laid by tore than One Female 6 6

sometimes lay their eggs in other nests of the same species after incu­

bation has started, and introduced eggs are readily accepted by the

incubating female. I once added nine eggs to a clutch already con­

taining nine eggs, and incubation continued without interruption. Eggs

in various stages of incubation can frequently be found in individual

duck nests. Barnes (1948:449) reported such a case where three eggs re­

moved from a nest of hatching young required two additional weeks of in­

cubation before ducklings hatched from them. In such cases, the intro­

duced eggs are left partially incubated when the original clutch hatches

and the ducklings leave the nest.

When the laying place is not a dump nest, the female covers the

egg or eggs when she goes off the nest from the time the first egg is

deposited. At first the loose material in the cavity is used to cover

the eggs, but after part of the clutch has been laid, the female starts

to pluck nesting down which is used to cover the eggs. According to

Dreis (1951$20) the first down is introduced after the second or third

egg has been laid. Leopold (1951$212) stated that introduction of down

started with the laying of the fourth to the tenth egg; in 70 per cent of the cases, it started with the laying of the sixth to eighth egg.

Gigstead (1938:607) reported that some nests were observed where down was not placed in the nest until after two weeks of incubation; one clutch hatched without any down having been introduced into the nest.

According to Hochbaum (1944$4-8) the canvasback introduces the first down with the third or fourth egg.

The first down in the nest may consist of only one or two wisps neatly centered in the cavity. The amount of down added increases daily 67

until, in the first nestings, about 200 cubic inches of loose down is

present. This makes a layer of down about two inches thick over the

top of the eggs. When the first few eggs are being laid, the female

spends a shorter time on the nest than when there are more eggs in the

nest. The additional time is presumably spent in plucking more down

and in covering more eggs. In nests which represent second nesting

attempts, the amount of down becomes considerably reduced, and it may not be in excess of 25 cubic inches. An increased portion of the nest­ ing material is breast feathers in nests representing second attempts.

The first evidence of utilization of a nesting box is a shallow depression in the loose material which is in the box. After egg-laying has commenced, the eggs are covered and the depression is no longer pre­ sent. Normally one egg is laid each day before 9:00 A.M. Apparently a day is sometimes skipped; Leopold (1951:212) reported 13 days were skipped in the laying of 297 eggs.

When egg-laying was disturbed by an observer flushing the bird or by a starling building a nest over earlier laid eggs, the bird did not lay an egg in the nest on that day. Egg deposition in the nest was re­ sumed the next day, or in one case when the starling*s nest was removed, the second day after the last egg had been laid. Sowls (1949:264.) and

Andrews (1952:49) also observed that ducks of several species did not lay an egg in their nests on days they were flushed from the nests before egg-laying. With ducks other than wood ducks, Sowls (1955:132) found that eggs are dropped elsewhere than in the nest on such days.

In central Ohio in 1956, 34 nests contained 449 eggs or 13.2 eggs per nest. C.E. Knoder advised me (personal conversation) that he had 63

been able to induce captive females to lay 30 eggs in a season. An

interval occurred between each series of ten eggs. Audubon (1343:

273) stated that there are 6-15 eggs in a clutch, according to the

age of the bird. Richdale (1949:95) has shown that four year old

yelloweyed penguins lay slightly larger clutches than two year old

birds.

McLaughlin and Grice (1952:253) found that eggs will remain

viable only about 16 days before incubation starts. If this is the

case, a female should not lay more than 15 eggs if all are to hatch.

Miller (1950:6), however, noted that one egg hatched though it had re­

mained in a nest more than a month before incubation started. In this

Ohio study, the smallest successful clutch contained seven eggs of

which only one hatched. In New Hampshire, Lee (1956:8) found a female

which incubated a clutch of only one egg.

In the northern part of the birds* range only one brood is reared

each year. This may also be true for the southern part of the range,

though Baynard (1913:241) stated, "I thought they sometimes rear two

broods in the same year." Howell (1932:143) reported that William

Brewster noted small ducklings on the Wekiva River in Florida on March

19 and 20, 1377. Allowing eight weeks for these ducklings to attain

the age of flight, the female would be free to start another nest in

late May.

The first nesting occurs progressively later northward. Audubon

(1843:273) stated that nesting starts in the "middle states" about the beginning of April, in Massachusetts it starts early in May, and in

Nova Scotia it starts early in June. Nichols (1937:533) reported a nesting in New Jersey in which the first eggs must have been laid in

about mid-March. Beyer, § t al. (1907:320) reported that nesting begins

early in April in Lousiaha,

In the last week of March, wood ducks in central Ohio start egg-

laying in nests where the eggs are later to be incubated. In northern

Ohio, this start is made about 10 days later. In Indiana, Mumford

(1952:2A) likewise found that egg-laying in northern parts of the state

starts about 13 days later than in the southern part of the state.

Freak eggs are sometimes laid, and a large double-yolked egg was

found in one nest (Figure 20). This egg did not hatch as is usual with

double-yolked eggs.

"Wood duck eggs vary considerable in size and shape (Table VI) among

different individual wood ducks. The shape varies from short subellip­

tical to subelliptical. The color of the eggs is buffy white or ivory.

When not nest-stained, the shell is somewhat glossy. It is relatively

hard and does not break readily.

Dixon (1924:4-6) reported a case of an egg being cracked while

still in the bird's oviduct presumably as a result of a bird squeezing

through a small opening to the nest cavity. The cracks were sealed over

by addition of shell before the egg was laid.

Incubation

On the day the last egg in the clutch is laid, the female normally commences incubation. According to Leopold (1951:213) the female some­ times spends several nights on the nest before the clutch is completed.

Dixon (1924:56) reported that one female started incubating her clutch Figure 20

Abnormally Large Eggs Are Sometimes Laid 71

Table VI

Measurements in Millimeters of 17 Eggs in a Wood Duck Nest

58 x 40 . 55 x 39

58 x AO 53.5 x 38.5

57 x 4-1 52.7 x 40

57 x 39 51.7 x 40

57 x 38 *49 x 39

56 x 39 *48.5 x 38.5

55 x 38 *48 x 40

55 x 39 *47 x 38.5

55 x 39

^infertile 72 when nine eggs were laid although she laid a total of twelve eggs.

Breckenridge (1956:17) reported that a female remained on the nest all night three nights before incubation started, Hochbaum (194/-*50) ob­ served that canvasbacks sometimes commenced incubation a day or two before the last egg was laid.

W h e n a bird starts an attentive period the down covering the eggs is moved to the sides of the nest, and the bird's breast is placed directly against the eggs. Because the down used to cover the eggs was removed from the bird's breast, the feather-covering of the breast is thinner than if the down had not been removed, and body heat is readily radiated from the breast. Thus surrounded with down and covered with the warm breast of the female duck, the incubating temper­ ature closely approaches the body temperature of the female. Tempera­ tures around the eggs of from 38.0°C. to 4.0°C. were recorded during incubation. The temperature readings beneath the down of the birds' breasts were usually only .5 to 1 degree centigrade above the tempera­ tures among the eggs, and the cloacal readings were only .5 to 1 degree above the readings made beneath the breast feathers.

At air temperatures of from 10°G. to 13.3°C., the temperature in the box above the down was only one or two degrees above the outside temperature. At a temperature of 12.2°C. in the nesting box over the nest, the temperature beneath the down among the eggs dropped 5.5°C. the first ten minutes after the incubating bird had left the nest; it dropped

3.9°C. in the next one and one-half hours. Two hours after the bird had left the nest, the temperature among the eggs was still 13.3°C. higher than the temperature above the down. Breckenridge (1956:20) noted that 73

the nest temperature dropped 7.5°C. in one hour and twenty-two minutes

during an inattentive period with the air temperature outside the box

between 33.9°C. and 35.0°C.

At one nest where I placed an automatic recording device (Figure 21)

after incubation had been in progress 11 days, the female did not return

to the eggs until 49 hours after the recorder had been placed at the

nest. All of the eggs in the nest later hatched. Miller (1950:7) re­

ported that eggs hatched after the incubating female had been absent

from the nest as long as a day. Breckenridge (1956:18) observed that

females sometimes remained off the nest all night at both early (3

days) and relatively late (15 days) stages of incubation. Hamilton

(1943) reported a case where a bluebird remained off its nest for 49

hours at 9 days incubation, and four of the 3ix eggs later hatched.

Shreiner and Hendrickson (1951:7) reported that females do not

incubate their eggs during the hot part of the day in June and July.

Breckenridge (1956:18) reported a tendency to incubate shorter periods

on hotter days. He considered this an adjustment of the female to meet the needs of the eggs. It seems to me more probable that this be­ havior is related to some physiological variable in the female, such as

greater water requirements during hotter days.

Incubation periods of the wood duck vary between 25 and 33 days.

Kortright (1942:226) reported incubation periods of 28 to 31 days, and

Leopold (1951:216) reported incubation periods of 27 to 33 days for 35

clutches. Through the aid of an automatic activity recorder, Brecken­ ridge (1956:20) established an incubation period of 25 days, and he

suggested that the incubation period should be considered 25 or more days. Figure 21

Automatic Nesting Activity Recorders

Were Operated on Several Nests 75

He assumed that the striking disparity among authors has resulted from variations in attentiveness of the incubating birds and differences in the amount of down surrounding the eggs. Girard (194-1) also reported variation in the incubation period of the mallard amounting to two to three days either above or below an average of 28 days.

In the present study, automatic nesting activity recorders were operated on two nests throughout the incubation periods. Because of disturbances by fishermen, one of the birds remained off the nest all of one night. The other nest was in a more secluded location and the female had an undisturbed incubation period as indicated by the record from the automatic recorder. This bird was removed from the nest for banding, but when restored she remained on the nest 20 hours after banding. The incubation period of this clutch was 31 days.

There was no regularity in the time of the attentive and inatten­ tive periods of this female (Figure 22). After incubation commenced, the female left the nest eighteen times afte'r 3:00 P.M., seven times before 9:00 A.M., and eight times between 9:00 A.M. and 3:00 P.M. This shows a tendency for more inattentive periods to occur in the late after­ noon rather than at other times of the day. Breckenridge (1956:17) found irregularity in the attentive periods of incubating females.

Girard (194-1:238) noted that mallards typically left the nest twice daily. Hochbaum (1944-: 50) found a lack of regularity in attentiveness among canvasbacks. Low (194-5:43) found irregularity in attentiveness among female redheads. Rest periods occurred at irregular hours of the day and night. Sowls (1955:103) found that various species of ground- nesting ducks left their nests once or twice daily, and the number of 0

1 11111 j H ' N * H i n n | i i i i i | i i i i i H H i i | h h H i h i i 1 i i i i i |"i i i i i | i i i i i | i i i i i h h i i | m i i i [ 22 23 24 25 26 27 28 29 May

| I I n I | H I I I | I I 1 I I 1 I I I I I I +1 I | I I I I I | I I I | I 1 II I I i 1 I I I II | I I I i I | i II t I | It i I I | I I I I I | 29 30 31 I 2 3 4 5 May June

| I H M j I I I I I I I I I I I 1 I M M 14 I f I 1 I I II I I I I M 1 I I I I I H I I II I I I I II M I 1 H I'M '-H H I | I 1 I I I | 5 6 7 8 9 10 II 12 June — Attentive Period

H H'H i 11 f-H ii 1111 m 111 \ i't fH+fi4 11111 h 111 m M -| M i 111111 1111111 f 11 Inattentive Period 12 13 14 15 16 17 June (hatched)

Figure 22 - Recording of Attentive Periods at a Wood Duck Nest O 77

times varied with individuals.

Attentive periods varied in length from 2 hours and 10 minutes to

60 hours. There were three periods about the middle of the incubation

period, in which the female remained on the nest more than 55 hours.

No comparatively long periods of attentiveness were found by Brecken­

ridge (1956). Richdale (1942) reported that a royal albatross remained

continuously on its eggs for 14 days. Hochbaum (1944s50) found that

with canvasbacks departures from the nest became less frequent and of

shorter duration with advancing incubation.

During these long periods of attentiveness, the incubating fe­ males never voided excrement in the nest unless they were disturbed.

’When they were disturbed, they always voided excrement on the eggs.

Considerable excrement accumulates in the cloaca during protracted

periods of incubation. That which could be gathered after being voided into the nest by one bird weighed 7 grams. A portion nearest the anus was entirely desiccated.

The duration of inattentive periods in this Ohio study varied from 15 minutes to 5 hours and 15 minutes. Breckenridge (loc. cit.) reported a variation in periods of inattentiveness from 13 minutes to

4 hours and 40 minutes.

Incubation is normally performed entirely by the female. Rollin

(1957) reported a case of a captive drake in eclipse plumage incubating eggs in a nest deserted by the female of the pair. The drake was first seen to enter the nesting cavity during the second week of July, about two weeks after desertion of the nest by the female. Since this drake was in eclipse plumage, it was flightless and gained access to the box 78

by means of a log. It was reported that the eggs were actually warmed

by the drake. He covered the eggs with nesting material when going off

the nest. This drake continued setting on the eggs until July 26.

Attentiveness covered periods of both night and day. It is very dubious whether the behavior of this male should be considered incubation. The

breeding period had passed, and this drake was in the process of molting.

Little is known about the behavior of drake wood ducks during the post­

nuptial molt. Perhaps the males sometimes seek retreat in cavities

during their flightless stage.

Relative Humidity in the Nest. - Mayhew (1955:4-6) has suggested

that moisture is so important to successful hatching of mallards that

during dry years hatching success is progressively poorer as nests are

located farther from waterj presumably with nests nearer to water there is less time for the plumage of the female to become dry as she goes from water to the nest. Romanoff (1943) stated that the best hatches of domestic runner ducks occurred when the immediate environ­ ment of the eggs was maintained at 80 per cent relative humidity.

Little success was had in determining the relative humidity in wood duck nests. At the single nest where a result was obtained, the relative humidity was 3 per cent higher in the box over the nest than outside the box. The relative humidity, however, was rather high (60 per cent) outside the box. The relative humidity in the nest among the eggs was not determined.

Turning the Eggs. - The incubating female does not permit her eggs to remain long in one place nor in one position. The eggs are fre­ quently moved from place to place in the nest (Figure 23)j all eggs May 15, 4:30 P. M. All T's Up. May 16, 2:00 PM.

^YlOG iT 1

wt

May 18, ICOO A.M. May 19, 9:00 A.M,

Figure 23

The Eggs Are Frequently Turned and Their Relative

Positions in the Nest Are Changed

(T=top, B=bottom, R=right, L=left, G=great and, S=small end) 80

are therefore more or less uniformly heated. Frequent moving of the

eggs by the female and the insulation of down in the nest may be

responsible for the ability of the wood duck to hatch large clutches

of as many as 28 eggs. Variation in the size of clutches may be

partly responsible for the wide range of incubation periods (25 to 33

days) reported in the literature.

There is no available information on the method used by the wood

duck in turning the eggs, but Hochbaum (1944s50) observed that this was

done with the bottom of the bill by captive canvasbacks.

Injury-feigning. - The incubating female usually does not flush when disturbed, but when she does flush she normally gives the broken wing ruse. The ruse is given the first 100-150 feet as the female leaves her nest.

Hatching

At least a day before the first pipping of the eggs was notice­ able, movement of the ducklings could be heard within the shell, and as much as 48 hours before hatching the first sign of pipping could some­ times be seen. At a point slightly nearer the larger end of the egg than the largest cross-sectional diameter, a bulge appeared on the out­ side of the egg. Depending on how the shell had cracked, this bulge was three-sixteenths of an inch or more in diameter. The duckling continued to press outwardly against the inside of the shell until its egg tooth could be seen at the center of the bulge through the broken egg shell and membrane. The time spent thus far since the first evidence of pipping was 24 to 48 hours. Then, after a scrambling within, a bulge appeared about one-eighth of an inch to the right (with the pointed end

of the egg toward the observer) of the original bulge, and outward push­

ing continued until the shell and egg membrane were broken at the new

point. More struggling occurred and the bird turned inside the egg so

its bill and egg tooth had moved about another eighth of an inch toward

the right. This alternate turning within the egg and breaking small

sections of egg shell and membrane continued until the egg was almost

encircled and a cap was cut from the larger end of the egg (Figure 24-)•

The egg was not entirely encircled, but the remaining shell cracked and

the short section of membrane acted as a hinge. Pressure of the neck

against the cap caused the latter to separate from the cup end of the

egg, and the duckling soon straightened its neck (Figure 25) and crawled

out of the cup. Hatching of a black duck was described by Allen (1892) and the process was similar to that of these wood ducks.

After pipping of the eggs started, increased movement of the in­

cubating female was apparent. This was also observed by Breckenridge

(1956:19) for the wood duck and by Girard (194-1) for the mallard.

Usually all of the ducklings hatched from a clutch of eggs within the same hour. Sometimes, however, some of the ducklings hatched enough later that they were somewhat weaker than those hatching earlier and were left behind in the nest or were later lost because of their inability to follow the brood. Webster (1955) reported some ducklings hatching after the brood had left the nest. The female will continue to incubate beyond the normal incubation period if no ducklings hatch, and

Leopold (1951s215) reported a case of a female incubating a clutch of infertile eggs 62 days before the rotten eggs broke. If only one duckling Figure 2 U

In Hatching, the Duckling Cuts a Cap from the large End of the Egg oa Figure 25

The Neck Is Curved Around on the Duckling*s Belly with

the Egg-tooth Against the Shell 2 hatches, the female will take it and abandon the nest.

Thirty-four nests examined in central Ohio in 1956 contained 449

eggs or an average of 13.2 eggs each, and an average of 11 ducklings

left each nest. The hatching success was 100 per cent in only 12 of

the 27 nests. Seventeen nests contained 4.1 (9.1 per cent of total)

infertile eggs; 13 nests contained 26 eggs (5.8 per cent of total) re­

quiring from one day to four weeks more incubation. One nest contain­

ing all infertile eggs was found at Magee Marsh in 1957. The hatching

success of clutches varied from 0 to 100 per cent. In 12 nests in New

York, Klein (1955:76) noted variation in hatching success from 25 to

100 per cent. Checking 640 eggs in New Hampshire, Lee (1956:4-) found

2.5 per cent were infertile and 3.6 per cent failed to hatch because

of incomplete incubation. In Vermont, Miller (1951:1) found a hatch­

ing success of 63.4- per cent for 868 eggs. He found that 6.9 per cent

of 693 eggs were infertile. In Iowa, Leopold (1951:214) observed an

80 per cent hatch of 868 eggs; 4- per cent were infertile. McLaughlin and Grice (1952:253) found 8 per cent of 12,180 eggs infertile and 7 per cent fertile but not producing ducklings to leave the nest. Hatch­ ing success may be somewhat lower among wood ducks than among some other species of birds; Kiel (1955:197) reported that 99 per cent hatch­ ed of 1,394- eggs of the coot.

In central Ohio, duckling broods were first observed on May 12,

1955, May 10, 1956, and May 9, 1957. In New York, Klein (1955:76) re­ ported the first hatching on May 17. In southern Indiana, Mumford (1952

26) found the first brood on April 14 and in northern Indiana on May 9.

In central Ohio there is a distinct issue of young from the nests slightly before mid-May. In a sample of 21 broods on the Scioto River in 1956, this early issue represented about one-third of the broods for the season. There was a second distinct issue in the period June 20-

25, both in a Scioto River transect in Ross and Pickaway Counties,

Ohio, and at the Olentangy Wildlife Research Station, Delaware County,

Ohio, consisting of about 50 per cent of the broods. Two females of early broods at the Olentangy station had been banded the preceding year as incubating adults; one female of the late broods had been banded as an incubating adult; five females of the late broods had been banded as locals or immatures. The limited data suggest that older females nest earlier than the year-old females. The late-nesting adult female may have been renesting.

Renesting

In the present study in Ohio in 1955-1957, known renesting by marked females occurred in only two instances. One bird at Buckeye Lake had 12 eggs in a first clutch, and these were deserted as a result of my disturbance on May 28, after 24 days of incubation. Sixteen days later this female started laying eggs in a nearby box. The second clutch totaled ten eggs, and it too was deserted on June 25, after being incu­ bated six days. This desertion did not result from my disturbance> al­ though I had removed the bird from the nest to determine its band number.

The second nest was about 200 feet from the first. The male and female remained together near the nest during laying of the second clutch as during earlier egg-laying. This bird presumably did not rear a brood in 86

In the second case of renesting, at the Olentangy Station, the

first clutch contained eleven eggs, and the second contained only seven

eggs. Six of the seven eggs were infertile, and one duckling left the

nest on July 19, 1955. Adult males were last seen in the area May 30.

A1 owing 30 days for incubation and one day for the laying of each egg,

the males left the area at about the time this bird started egg-

laying.

Barnes (194-8:449) reported a case of a wood duck renesting where

the first nest contained twelve eggs, and the second nest contained

eight eggs. Only three young hatched from the eight eggs of the second

nesting. Sowls (1955:130) observed that second clutches of several

species of ducks in Manitoba were significantly smaller than first

clutches, but large second clutches were larger than small first

clutches. Seubert (1952:327) observed that second clutches of ring­ necked pheasants were smaller than first clutches.

Hybridization

Bagg (1948:451) observed a male wood duck following a female mallard while the wood duck repeatedly whistled and squeaked, and Bagg inter­ preted this behavior as courtship. He concluded that, "Apparently the wood duck may be included among those...species with which the mallard crosses.” Copulation was not observed, and no hybrid ducklings were ob­ tained. It seems doubtful, therefore, if this observation should be accepted as evidence of hybridization of the mallard and wood duck.

Coues (1903:911) stated that when in confinement with other ducks, the wood duck hybridizes freely with other species. Phillips (1925:71-72) reported that this duck has been crossed with the mallard, redheads,

rosy-billed duck, Brazilian teal, Chilean.pintail, white-eyed duck, and

yellow-billed duck.

Summary

Wood ducks pair when on the wintering ground in the South. The

courtship performance is described; this can sometimes be observed in

the North after pairing has taken place. Before pairing, the birds move about with seeming indifference to sex companions; after pairing,

the male follows the female in all movements.

Wood ducks are non-territorial, and females frequently lay eggs

in other wood duck nests. Egg-laying starts in late March in central

Ohio and about ten days later in northern Ohio. After several eggs have been laid the female plucks down from her breast and uses it to

cover the eggs, as much as 200 cubic inches accumulating in early nests.

Thirty-four nests contained an average of 13.2 eggs per nest. Only one brood is normally reared in a year.

Incubation is performed entirely by the female and commences when the last egg is laid or 2-3 days earlier. The female covers the eggs with down when leaving the nest. This down covering prevents the loss of heat, the temperature in one nest being 13.3°C. above the outside air temperature two hours after the incubating female had left the nest.

Incubating temperatures from 38.0°C. to 4-0.0°C. were recorded. After being disturbed one female remained off her nest 49 hours, and all of her eggs later hatched; when not disturbed, inattentive periods varied from 15 minutes to 5 hours and 15 minutes. Attentive periods varied from SB

2 hours and 10 minutes to 60 hours. There was no regularity in the time of day of the inattentive periods. Incubation periods from 25-33 days have been recorded. Artificially incubated duck eggs hatch best when incubated in an environment with a relative humidity of about 80 per cent. The eggs are frequently turned by the incubating female and their positions in the nest changed.

Pipping of the egg occurs as much as 48 hours before hatching.

The process of hatching is described. Because of infertile eggs and late-deposited eggs, hatching success is normally low or less than 80 per cent.

Broods in central Ohio were first observed on May 9, and about one- third of the broods in a study area emerged before mid-May. There was a second distinct issue of broods in mid-June. It appeared that the early issue was from older birds, and the second issue was from year- old birds. Sixteen days after deserting her nest, a female started laying eggs in another nest.

The wood duck in captivity is known to hybridize with several other species of ducks. CHAPTER V

CARE OF THE YOUNG

Movement of Ducklings to Water

Forbush (1925:229) states that the young of wood ducks have ex­

ceedingly sharp, hooked claws and a hooked nail at the end of the bill.

A careful comparison of the claws of day-old wood ducks with the claws

of shoveller ducklings did not confirm the belief in such a difference.

Though I have watched numerous ducklings climb up the sides of boxes, I

have never observed one make use of the bill in climbing.

The female first appears at the entrance to the nesting cavity,

and after a short pause there, if disturbances are absent, she flies to

the ground. On the ground she utters a weak pe-e-e-e call. This acts

as a "follow me" call to the ducklings. They start jumping above the

nest, and with their sharp claws they cling to the sides of the nesting

cavity. The first jump puts them about eight inches above the nest.

If they are successful engaging the other foot with the side of the

cavity, a second hitch moves them about four inches higher and like­

wise with successive hitches until the opening is reached. The jumps

are frequently ineffective and the ducklings fall beck into the nest

only to immediately repeat the trial. The ducklings leave the nest in

rapid succession, and all usually are out of the box in less than two minutes. Depending on the deepness of the cavity and the ease with which the young climb out, it may sometimes take as long as ten minutes.

Gigstead (1938:606) reported that a duckling which had been left behind

89 90

in the nest left the nest alone in the morning of the following day after

the remainder of the brood had gone.

The ducklings respond to the "follow me" call of the female by

peeping and by following her. Unless the auditory acuity of young wood

ducks is greater than that of other ducks (Trainer, 194-6), this call of

the female would be inaudible at distances which sometimes separate the

ducklings from the.female as they leave the nest and move to water.

With the young in a tree cavity 20-4-0 feet above the ground and with a

strong wind blowing, it hardly seems probable that the ducklings in the

nest could hear the call of the female on the ground. It seems likely

that the first "follow me" calls are given by the female before she

leaves the entrance of the nesting cavity.

On the somewhat windy day of May 10, 1956, a brood of ducklings

was observed leaving their nest. After sitting with her head out of

the nesting box entrance several minutes (Figure 26), the female flew

down and alighted on the water of a nearby pond. She first sat on the

water about 10 feet from a point directly beneath the nesting box, but

after five ducklings had promptly joined her she moved to a distance of

25 feet away. At this new position the female was separated from the

newly emerged ducklings by a steep bank about A feet high, and the wind

was blowing from the ducklings toward the female. The second five duck­

lings which then emerged from the box failed to move toward the female

for some time. On the contrary, they huddled together in a group near where they landed when they fell from the box. The female with part of

her brood continued to wait, while the last-emerged ducklings simply re­ mained in their huddle. After 10 minutes, fearing for the welfare of the Figure 26

When Leading the Brood from the Nest, the Female Paused at the

Nesting Box Entrance before Flying to the Ground 92 ducklings, I approached them and drove them toward the female, where­ upon they promptly joined her, and the complete brood swam away. The fact that the female continued to wait suggested that she heard the calls of the lost ducklingsj whereas, the failure of the ducklings to follow the female suggested that the ducklings did not hear the calls of the female. Presumably the direction of the wind imposed sufficient dis­ advantage on the ducklings that they could not hear, although they were heard.

The instinctive behavior of very young ducklings which causes the brood to maintain a compact group is a useful characteristic in pre­ venting loss of ducklings. Where one duckling can go, the remainder of the brood usually can go, and this is about the manner in which recently emerged ducklings behave when not disturbed. Because of differences of hatching time, however, some of the young are sometimes weaker than others and become lost on the overland trek to water. Also, when attacked, the ducklings scatter; then the faint "follow me" call of the female or the peep of the ducklings may not always be loud enough to be heard at the distance separating the female and the most distant duck­ lings, and some may be lost. Parts of broods have been observed separat­ ed by distances of 500-600 feet.

Audubon stated (1843:273) that the young leave the nest the moment they are hatched, but this is hardly true. The ducklings normally leave the nest sometime in the forenoon of the day following hatching, Gigstead

(1938:606), however, reported a case where the undisturbed brood left the nest shortly after 3:30 P.M. of the day after hatching. 93

Leopold (1951:217) observed that the female leads the brood from

the nest earlier on warm, sunny days than on cool, rainy days. The time

of leaving the nest may not be determined alone by the female, as duck­

lings hatched from eggs I had placed under a domestic fowl left the

hen and wandered about in a nearby garden in mid-forenoon of the day

following hatching.

Eye witness accounts of several methods of movement of young wood

ducks from the nest to water have appeared in the literature. The pre­

ciseness of the descriptions leave no alternative but acceptance of the

postulate that deviations sometimes occur from that which many wood duck observers have repeatedly witnessed. It can be stated only that wood ducks usually do thus and so and then hope that a freak duck will not behave in such a manner as to disprove earlier conclusions.

Catesby (1731:97) stated that the young are carried on the back of the female from the nest to water, and that they are regularly removed from danger by being carried to safety after attaching themselves to the back of the female. Kingsford (1917) reported having observed a female duck make 11 or 12 trips from her nest 50-60 feet up in a tree to a water area 200 feet from the base of the tree, and each time she carried a duckling by the neck in her beak. Audubon (18^3:273) and

Bent (1923:162) quoting Langille also suggested that the female carries her young to water when the nest is located "some distance" and "a little distance" from water. Neither of these authors gives corrobor­ ating details of these observations. Wilson (1854-:599) reported obser­ vations of a female carrying thirteen young, one by one, from a nest to the foot of the tree below the nest. After the ducklings were all 94

safely transported to the ground, they were led in a group to water.

Mousley (1924:581) reported observations of wood ducks carrying their

young in their bills from the nest to the ground below or to water.

These observations were not made by the authors themselves, but were

reported to them by observers thought to be reliable.

Bent (1923:163) reported an observation by E.F. Pope where the

ducklings were transported on the back of the female to the water 28

feet below the nest. The observation was described thus: "She emerged

from the cavity in the stub with a young duck on her back and simply

dropped straight down into the water, using her wings to check the

speed of her descent. When she arrived within a foot or two of the

surface she suddenly assumed a vertical position which caused the duck­

ling to slide from her back into the water. She rose quickly, circled

a time or two, reentered the stub, and at once repeated the performance

until the whole brood of 10 were on the water." W.S. Cochrane described

a similar performance (Bent 1923:164): "She visited the nest several

times and after circling around the woods returned and rested on the edge

of the nest which was in a hollow stub of the oak. After resting there

about 10 minutes she flew down toward the water with her wings slightly

elevated, and when about 10 feet from the water she began flying in an

upward position, allowing one of the young which she was carrying on her

back to slide off over her tail into the water. She went through this

performance 14 times."

Although Audubon (loc. cit.) and Bent (loc. cit.) suggested that the ducklings were sometimes carried to the water by the female, these authors reported that the ducklings also get to water by walking after 95

first jumping to the ground from their elevated nest. Recently,

Bellrose (1955b:67) published photographs of newly hatched wood ducks

jumping from the opening of their nesting cavity. Bellrose does not report having observed the young leaving the nest in any other way than by jumping. Dreis (1951*36) and Shreiner and Hendrickson (1951:7) observed ducklings leaving their nests only by Jumping to the ground.

All of the numerous ducklings which I have watched leaving the nest jumped from the box entrance to the ground.

The ducklings are very light and strike the ground with little force. Dixon (1924.:62) computed that the ducklings weigh 3.2 grams per square inch of the surface exposed to resistance against the air in falling. Even though they may sometimes land on solid ground or stones, dropping to the ground does not appear to harm the ducklings. I dropped five ducklings, when a day old, from a height of 20 feet onto a hard limestone drive, and after a moment of delay they all ran from the spot.

One duckling started running with a wobbly gait suggesting that it had been slightly dazed by the fall, but it soon joined the others and showed no further signs of injury. These ducklings all voided excre­ ment soon after striking the ground, as several ducklings were observed to do when jumping into water. The ducklings always landed on their bellies, and they bounced scarcely any when they struck the hard drive.

Ducklings observed jumping from their nest onto bare, loose ground bounced little or not at all, M.B. Trautman (personal conversation) advised me that he had observed ducklings jumping from their nest onto hard bare ground and that these ducklings bounced several feet. Mussel- man (194-Bs 199) reported observing ducklings which fell 15 feet onto grass 96

and bounced an inch or two. Leopold (1951 s218) observed that the young

sometimes bounce several inches on striking the ground.

In discussing integration of the brood by the female, Bent (1923:

164.) quoted notes of Manly Hardy in which it was stated, "It was plain

that she cuuld count enough to know if one was missing," and it was im­

plied that the female did not leave her scattered brood until all had

been gathered to her. This viewpoint is somewhat too anthropomorphic

for ready acceptance, and many ducklings doubtless become temporarily

or permanently separated from their original broods or female ducks.

Such loss is particularly probable when the broods are making their

initial trip to water from a nesting place some distance from water.

At Burlington, Iowa, Leopold (1951:220) observed a loss of 61 ducklings

from 189 which moved down a steep bluff and about one city block toward

water. At the Olentangy Station in 1955, three unusually small broods were known to have not come from boxes, but the distances they moved

from their natural cavities was unknown.

After the ducklings leave the nest, the female soon leads them to

concealing plant cover near water. In following the female, the ducklings

string out closely behind her or follow in a compact group. The young maintain a compact group when slightly disturbed, but when not disturbed

and feeding the young spread out somewhat more.

Care of Young

The ducklings seem to start feeding instinctively and pick insects from the vegetation without coaching from the female. Females do not pick up a morsel o'f food and seemingly call their young in the manner 97 followed by domestic chickens. A recently hatched brood was observed to continue feeding soon after being separated from the female.

Audubon (18,43:274) stated the young are taught to hunt food, but I can find no evidence to support this belief.

The female will simply lead her brood to concealing plant cover, if approached, but not closely pressed when on open water. When the duck­ lings are hidden, the female quietly flies away as the disturbance approaches more closely. If the unconcealed family is closely pressed when on the water, the ducklings will skitter over the water surface to plant cover, and the female will give the cripple ruse, usually uttering several "distress" calls when she first starts to move from the brood. If she is followed by the intruder, the ruse will continue for as much as one-quarter to one-half mile. If the ducklings are very closely pressed they will dive and move under water to plant cover.

The female will frequently remain hidden with her young among vegeta­ tion instead of flushing. On one occasion two female wood ducks re­ mained hidden among tall grass until I was within a few feet of them where I flushed a quacking mallard which flew from among her brood.

'When the mallard noisily flushed, the two wood ducks also flushed but without uttering an alarm note.

Observations of captive ducklings suggested that the young are brooded by the female much of the time during the early period out of the nest. It appears, however, that the ducklings can survive con­ siderable chilling. A duckling was known to have been lost and unpro­ tected on a pond through four hours on a cool rainy day. This duckling was captured and successfully reared. 98

The young are usually brooded among protective cover. A brooding

female with her family of 15 recently hatched ducklings was found,

however, on the Scioto River in the early morning on May 16, 1956,

This brood was on top of a railroad tie floating among other debris at

the edge of the river. The brooding female was exposed to view from all

directions. Droppings of ducklings indicated that the brood had spent

the night on the railroad tie.

As indicated by observations of captive birds, the ducklings

frequently come from beneath the breast of the brooding bird and get

drinks of nearby water. It was not determined how long ducklings can

survive without water, but accessibility of water may be a critical

factor in duckling survival if nests are located a long distance

from water.

The young are able to fly at about eight weeks of age. Young-of-

the-year were first seen flying on July 11 in central Ohio. Mumford

(1952:12) reported young flying as early as June 30 in Indiana.

Summary

In response to a weak pe-e-e-e call of the female, the ducklings

usually leave the nest by climbing up the sides of the nesting cavity

and jumping to the ground or water below. The brood is then led to water

or to protective cover near water. Published accounts indicate that the young are sometimes carried to water in the bill or on the back of the female.

The female when disturbed flies quietly away if the young are con­

cealed; if the young are not concealed, she gives the cripple ruse. The 99 female sometimes remains hidden with the brood among vegetation.

The young are brooded by the female among vegetation during the early period out of the nest. The young during brooding frequently come from beneath the breast of the brooding bird and get drinks of water.

The young are able to fly at about eight weeks of age. PART II

THE WOOD DUCK IN ITS ENVIRONMENT

100 CHAPTER VI

LOCAL MOVEMENTS

Brood Movements

The female leads the brood to water soon after it is out of the

nest. In two cases observed of nests somewhat distant from water,

approximately the most direct course to water was chosen. This appears

to be the usual procedure, and it results in wood duck broods being

observed in seemingly strange situations such as on city streets, as

has been noted by many bird watchers. According to the distance of

the nest from water, this trip may mean for the day-old ducklings an

over-land trek of a mile or more. Apparently this feat is accomplished

on the first day out of the nest.

The ducklings are secretive, moving about relatively little dur­

ing the first few days out of the nest; consequently they can be present on a pond without being observed. Before two weeks of age, the habitat requirements of the brood are simpler than after this age, and usually the ducklings spend this period at any available water area near their hatching place (Figure 27).

Wood duck broods generally spend their first two weeks somewhat isolated from other broods of the same species. If more than one brood hatches near a given habitat within the two weeks limit, some congre­ gating of broods may already be apparent. At about two weeks of age, a trend toward increased congregating of broods occurs. Maximum congre­ gations frequently consist of five and six broods.

101 Figure 27

Mood Ducks Left This Pond Mhen About Two Meeks of Age 103

In congregation movements, the broods moved both cross-country

and along water courses, and the movements did not always consist of a

simple process of drifting up or down a water course to a favorable

habitat. Two color-marked broods at the Olentangy Station appeared

on a pond about one-half mile from their natal ponds. The three ponds

were separated by woods and fields but were connected by a road, and

tracks of small ducks observed in the mud on this road suggested that

the road was used by the ducklings, Hochbaum (1944:104) observed that

various species of ducks follow roads in overland movement to new

water areas.

Movements of broods on streams usually followed the stream courses.

Two color-marked broods were found on creeks at the mouths of tributary

streams leading to the general localities of the ducks' natal boxes,

and both these broods had moved approximately a mile down the tributary

streams. The movement downstream was necessary to take the ducks to a

larger water course, and the birds followed the stream rather than taking

the shortest route to the ultimate water habitat.

Usually all of the broods within a certain radius moved into these

congregation areas, but the length of this radius was variable and unpre­

dictable. The extent of the area from which the ducks moved to congre­

gation points was somewhat dependent on the lack of other equally satis­ factory habitat. In one case, on the Scioto River in Ross County, Ohio, two congregations formed within a half mile of each other. Trapping a brood with a banded female showed that it moved 3.5 miles up the river to the congregation area. The air-line distance was only slightly more than half the river distance, but it is assumed that the ducklings reached 104.

their ultimate brood habitat by following the course of the river.

Those birds which arrived at concentration points as day-old ducklings

sometimes did not move to new areas at two weeks of age. The distances

moved by broods to congregation points varied from zero to 3.5 miles.

The total daily range of ducklings within these congregation

areas was normally small. When the congregation area was a farm pond

surrounded with herbaceous plants, the range of movements was limited

to the pond and an area 50-100 feet around all or part of its margin.

Where the pond contained a relatively large area of emergent vegetation,

the width of the marginal strip used for escape cover was somewhat

narrower and less continuous, or the movement of the ducks was confined

to the water area.

The daily range of river broods was normally restricted to a short

section of river, the limit of which was defined by the extent of the

escape cover. Before moving into these congregation areas, individual

broods sometimes swam along the shore ahead of a slowly moving boat as much as one-fourth mile before coming to cover overhanging the water

or running over the shore to protective cover.

Although as many as five and six broods sometimes occupied small

ponds less than half an acre in area (Figure 28) or a short section of river, not all of the broods of a given congregation combined into a

single group. The individual broods continued to maintain a moderate degree of brood purity, and mixing of broods occurred to only a limited degree. "When ducklings became lost from the original broods, they joined other broods when these were available. Some broods contained one or two ducklings of an age several weeks younger or older than the Figure 28 105 Broods Gathered on This Pond in Both 1955 and 1956 106 rest. Frequently two broods merged and moved about with one or both females, but never were more than 18 ducklings observed following one or two females in a single group.

When two broods merged, they were usually led by one female, but sometimes by both and sometimes by none. The fact that females seldom entered the traps with their young broods, together with the fact that females known to have broods were trapped at corn-baited sites some dis­ tance from their brood areas, indicated that the females fed elsewhere than with their young, and they still maintained their morning and evening feeding schedule when leading broods. The young were often left alone to fend for themselves or to join other broods.

Very young ducklings which became separated from the female con­ tinued feeding and moving about in a compact group. One brood found on the Scioto River swam out into the current of the river and was soon carried some 3,000 feet down the river after the female flew up the river. I approached this brood to within 15 feet and followed it with a boat for more than an hour. At another time a solitary duckling, apparently lost on the Scioto River, approached and followed my boat, showing almost a complete absence of fear. Fishermen on the Scioto

River also reported that solitary lost ducklings several times came to them at their boat docks and remained for hours. Collias and Collias

(1956:399), working chiefly with captive ducklings, showed that on the first day out of the nest ducklings have a strong tendency to follow any moving object. In wild ducklings under natural conditions another object apparently substitutes for the female when the ducklings become lost from the female. This trait has survival value only if the chances 107

of the lost duckling following a benefactor are greater than the chances

of its following a malefactor. When once in the water in the midst of

a fairly dense wood duck population, the chances are favorable of a

lost duckling joining another brood.

The females started spending more time away from their broods

about two weeks before the young were able to fly, about six weeks of

age. On June 30, 1956, post-brood flocking of females was first noted

when two groups of three adult females were seen on the Scioto River

in midafternoon. Because adult wood ducks normally feed in the morning

and evening and because of the absence of feeding behavior, it was

believed that these females had not left their broods for feeding pur­

poses. The females gave their broods less attention after about six

weeks of age, but they did not ordinarily completely abandom them at

this age.

Bent (1923:162), and Kendeigh (1952:186-187) reported that the

males help take care of the young, but Mumford (1952:28) reported see­

ing a male with a female and her brood on only two occasions. Bellrose

(1955b:8) reported that drakes leave the females shortly before the eggs

hatch and join others of their sex on secluded woodland ponds or in

dense swamps where they molt. Observations in this study indicated that

males gathered into increasingly larger groups near the end of the in­

cubation period or otherwise withdrew from association with the females

at that time. One male came to a pond and remained several minutes on

the morning of the day young came from a nest box on the pond; otherwise no males were seen with or near broods. 108

When the ducklings hatched, the females assumed full responsibility

for their care. Meanwhile, local drakes gathered into larger concen­

trations, and on May 30, 1955, a post-nesting concentration of drakes numbering 28 birds was last seen at the Olentangy Station. Where the

group went for the postnuptial molt was a mystery to me, but Richard

Reece, of the Ohio Division of Wildlife, later advised me (personal con­ versation) that he had observed numerous feathers and a number of flightless wood ducks on an isolated woodland pond eight miles north of the pond where I had last observed the flock. Wishing to trap and band flightless males, I operated a trap at this pond throughout the summer of 1956, but none was captured, nor was evidence of their presence found there in 1956. In the Scioto River study area during 1956 a marked decline in the number of males likewise occurred on about May 30, and disappearance of all males soon followed.

The only molting male captured in the two years of this study was in 1956 at the chief congregation pond at the Olentangy Station. Bellrose

(I955b:8) stated that males join others of their sex on secluded woodland ponds or swamps where they undergo the postnuptial molt. It is reason­ ably certain that there were no other, or at least few, molting males on this pond during the molting period. Presumably some males undergo this molt isolated from others of the same status.

The extent to which the selection of the brood congregation location is influenced by habitat offerings and the prior presence of duckling broods is not known, but both factors probably exert some influence. The gregarious nature of the wood duck is such that solitary adults are seldom seen except when incubating, and females with broods probably experience 109

a positive reaction to the presence of other adult individuals of the

same species. Many of the broods emerged from the nests in mid-May,

and nearly all of the males disappeared from the nesting localities by

May 30. Brood congregations began to form at the same time the males

disappeared. The female can seek the company only of other females

among adults of her own species, and at the disappearance of the male

she joins other females. There is only a nominal association with other

females of the group.

Except for one pond utilized in 1955 that was not used in 1956, the same four ponds were used by brood congregations in both years at

the Olentangy Station (Figure 29), although the distribution of occupied nesting boxes was different each year. The presence of two congrega­ tions, one of five broods and one of three broods, less than a half mile apart on the Scioto River, suggests that available habitat may be an important factor influencing the locations of congregations. Traps were operated at both of these congregation areas, and no exchange of ducklings between stations was noted although 4-1 ducklings were banded and 12 repeats recorded through June and July. Creek surveys in Ohio indicated that these congregations may be associated with discontinuous habitat, and congregations of broods were less apparent in two 10-mile sections of 'Whetstone and Alum Creeks' in central Ohio than in a 25-mile section of the lower Scioto River in Ross and Pickaway counties. Willow- herb and other low-growing vegetation in these creeks form more nearly continuous cover than the willow groves which provide cover only at iso­ lated sand bars and islands along the lower Scioto River, A brood re­ mained alone on one isolated farm pond where there was no known additional Figure 29

This Pond was Occupied by Brood Congregations During 1955 and 1956 111

wood duck habitat or brood within two miles. Undoubtedly, the discon­

tinuity of preferred habitat plays an important part in causing these

brood congregations.

Local Movements of Young After Attaining Ability of Flight

When the young ducks had attained the ability of flight, at about

eight weeks of age, further movement and concentration occurred. This

movement was first noticed at the Olentangy Station in 1956 on July 11,

the same date that a flying bird-of-the-year was first seen. This

duck moved a distance of eight miles from its native pond to the con­

gregation pond. Another duck made the same flight and was trapped two

weeks later. There was also movement of ducks on the Scioto River at

or soon after the time that the ducks first attained the ability of

flight. On August A, 1956, a wood duck was trapped five miles down the

river from the place the bird had been banded thirty-five days earlier

at six to seven weeks of age. Four other birds taken eight miles north

of the trapping site on the Scioto River probably moved at the same

time, although they were not retrapped until mid-September.

Not all wood ducks moved at eight weeks of age, but this movement

seemed to involve a more general shuffling of the population than the

earlier movements. Several wood ducks were trapped in late September

only a short distance from where they had been earlier banded as flight­

less young, and two were shot on the Scioto River on October 15, 1956,

only a short distance from the spot where they had been banded in June as pre-flight young. 112

Movement occurred both up (north) and down (soutl) the river from given points. Four ducks moved up the river 8 miles from one trapping

station, and one duck moved 5 miles down the river from this station.

It was not definitely determined whether the movements followed only the course of the river, but late summer and fall surveys of the duck population on the lower Scioto River indicated that there had been no major change in the total number of ducks present. The pre-hunting season movement of ducks on the Scioto River was probably confined to the course of the river.

There appears to be some segregation in the fall of locally reared young from adults. In 1956 at Buckeye Lake, in Licking, Fairfield and

Perry Counties of Ohio, none of the seven females banded earlier as in­ cubating birds were among the 84 wood ducks trapped during August and

September. There were, however, approximately IS birds which were not trapped but were known to be present on the lake. Sutton (1928:71) stated, "In the fall the locally reared young congregate, seemingly by themselves, and the single males and females, or small groups of the adults, wander about aimlessly."

The size of the fall congregations was not controlled alone by the numbers of ducks within a standard radius of the congregation area.

Six immature wood ducks were known to have flown approximately 8 miles from their brood sites to their post-brood congregation areas. Two fall groups maintained their isolation on farm ponds throughout September, although they were separated by slightly less than 2 miles. One of these groups consisted of only nine ducks, presumably a family group. This seemingly unusual situation may have resulted from the fact that the 113

ducks were being artifically fed at both stations. At Buckeye Lake,

which is 7.5 miles in length, a small group of 12 wood ducks persisted

throughout September, 1955, although a group of 160 birds gathered 6

miles distant near the opposite end of the lake. This was the largest

congregation observed in central Ohio. Three congregations of 100-120

birds were observed on the Scioto River, but the largest group observed

at one time at the Olentangy Station was 80 ducks.

The extent of the diurnal movements of these late summer groups was

not determined in detail. At Buckeye Lake, the ducks of the larger group

roosted only about one-half mile from the feeding site. This feeding

site was artifically maintained in connection with a banding program, and before the banding program was initiated the ducks had been feeding

in a farmer's hog lot about one-half mile farther from the roosting

place.

In the morning the ducks flew directly from the roosting place to the feeding site, and in the evening they flew directly from the feeding site to the roosting place. The morning exodus from the roosting site was made before light intensity favored detailed observations, but the ducks generally left the roosting site in the morning in larger groups than those in which they returned in the evening. The evening roosting flight extended through about half an hour and consisted of small groups of from one to twenty birds. At Buckeye Lake, where the feeding site was located on a mud bar near the mouth of Honey Creek, small groups of ducks started to gather near the feeding site in mid-afternoon. Feeding, however, did not occur until the regular feeding time. 114

When the morning and evening feeding flights took the ducks away

from their water habitat to farmers* hog lots, as was the case with

two groups on the Scioto River, the feeding flights were made by all

the ducks in one flock. Flocks started to assemble for such flights

in mid-afternoon. The interval between morning feeding and assembly

for evening feeding was spent in small groups sitting on the water, on

logs, and in low trees over the water preening and loafing.

The relative importance of the various environmental factors for

inducing late summer congregations was not determined. The congrega­

tions at the Olentangy Station were at the same pond in 1955 and 1956

(Figure 30). In 1955, the emergent was chiefly cat-tail 3-4 feet tall,

but in 1956, a higher water level was maintained so that there was an

increased water area covered by vegetation to a height of about 18

inches. There was a fallen tree in the shallower portion of the pond

as well as several standing trees. The feeding place, after artificial

feeding was discontinued, was not located.

Duck congregations on the lower Scioto River occurred near their

feeding places, and there usually seemed to be a wide bare sand or

gravel bar associated with congregations in this area. The birds loafed

on these bars. JJhere the congregations occurred at Buckeye Lake, an

abundance of emergent vegetation provided some isolation from fisher­ men: an extensive cat-tail marsh cut off the land side, while spatter-

dock beds cut off the water side against the approach of fishermen. This

furnished a roosting and loafing retreat, and nearby farmers* hog lots furnished a food supply. Artificial feeding was thought to have been responsible for holding two small groups on private ponds through September. Figure 30

Wood Ducks Gathered at This Pond Soon After Attaining

the Ability of Flight in 1955 and 1956 116

Known post-brood congregations during 1955 and 1956 on a pond at the Olentangy Station, on a private farm pond, on Buckeye Lake, and on the Hocking River in Hocking County, Ohio, persisted until about the first of October. The exact dates of dispersal of late summer congrega­ tions was not ascertained in every case but was known to have been within a few days before or after October 1. On September 30, 1955# the major­ ity of the flock of some 160 ducks suddenly left Buckeye Lake; on October

3, 1956, the number of wood ducks in the roosting flight declined from an earlier 145 to 15. A small flock at a farm pond near New Albany,

Ohio, was last seen on October 1, 1955.

Early Fall Dispersal of Late Summer Congregations

I first thought that the late September or early October decline of wood ducks was caused by southward migration. In 1955, 230 wood ducks were banded in Ohio as adults, flightless young, and immatures, and the subsequent recovery of thirty of these birds indicated something of the pattern of wood duck dispersal after the breakup of late summer congregations. Twenty-two of the thirty birds were banded at Buckeye

Lake, fourteen being birds shot in Ohio before the end of October.

These fourteen birds were shot within 40 miles of the late summer con­ gregation site, thirteen were shot within 15 miles, and twelve of the fourteen were shot within 10 miles of the late summer congregation site.

Only two were reported shot on Buckeye Lake. Of the twelve birds from

Buckeye Lake shot in Ohio elsewhere than at Buckeye Lake, all were re­ ported shot on rivers and creeks. All fourteen were immature birds which were probably hatched at Buckeye Lake. 117

Only four of the sixty-six wood ducks banded at the Olentangy Station during 1955 were recovered in Ohio the following fall, and two of these were known to be locally reared birds. The two locally reared birds were shot on streams within ten miles of the late summer congre­ gation site.

From recoveries of birds banded at Buckeye Lake and at the Olentangy

Station, it is apparent that when late summer congregations of wood ducks disband and leave lakes and ponds, the birds scatter in all directions and enter stream habitats. These early October movements seemed to consist of general dispersal rather than congregation of numbers as was characteristic of earlier movements. Only five of these eighteen birds shot in Ohio had moved in a southerly direction, and flight occurred in all directions.

If river habitat is the goal of the early October movement of wood ducks, there should be no movements of duck populations which earlier occupied river habitats. In the creek and river habitats under ob­ servation, there appeared to be no exodus of ducks before October 15, which was the date in 1956 of opening of the duck hunting season. In spite of the fact that there was no open season on wood ducks in Ohio during 1956, there was considerable duck shooting on the Scioto River during the early part of the hunting season. A survey of the population status of wood ducks on the lower Scioto River on October 20, 1956, indicated that the congregations had left the river. There is little available evidence to indicate where these ducks went, but the recovery of a crippled banded bird at Deer Greek near Williamsport, Pickaway

County, Ohio, in mid-November indicated that there was at least some 118

movement from the Scioto River to nearby creeks. More recovery data

are needed from ducks banded in river habitats.

Paul Gibson, a Union County, Ohio, duck hunter, advised me that

he had noticed a difference, correlated with shooting, in the times of

the fall when wood ducks left Big Darby Creek, In 1956, when there was

no known shooting of wood ducks at two congregations, these birds re­

mained at least until October 27. During years of shooting they normal­

ly moved out a few days after shooting started or almost two weeks

earlier when shooting occurred than when it did not. Available evidence

suggests that these birds simply moved to other nearby creek habitats.

An examination of the banding data in the files of the U.S. Fish

and Wildlife Service indicates that some birds travel several hundred miles in this late September radial dispersal. Banding recoveries

indicate that the heaviest southward migration occurs in the last few

days of October and early November.

Winter Movements of Wood Ducks

Wood ducks sometimes form extremely large congregations. Gabriel-

son (1943:171) reported congregations of as many as 10,000 on the 12,685 acres of the Savannah National Wildlife Refuge, Port Wentworth, Georgia.

Gigstead (1938:608) reported a flock of A,000 in the vast lotus beds of

Duck Island Lake, in Illinois. Errington (1937:53A) estimated 2,04.0 wood ducks were scattered through a 39-acre marsh in Iowa.

I spent a week, January 2-9, 1957, observing the 2,000-2,500 wood ducks then on the Savannah Refuge. The ducks were feeding largely on acorns, and they were feeding principally in two oak groves. The evening 119 feeding flight started about one and one-half hours before sunset, and

scattered ducks continued coming to the feeding area until darkness had fallen. Most of the birds came as small groups of less than 20, but

there were several flocks of 200-250 birds.

Ducks started to leave the feeding sites and moved to the roosting

place at one-half hour before sunset. The roosting flight continued until darkness made observation from a distance impossible.

The ducks scattered over the refuge during the day and spent the time on small open pools among the maiden cane (Figure 31), beneath the buttonbushes (Figure 32), and on the water in woodland swamps.

Some five hundred ducks fed, loafed, and roosted in approximately a square mile. The birds were concentrated in one or two groups only at feeding and roosting times. During mid-day they remained in relatively small groups on the water or perched on logs and the lower branches of buttonbushes and preened or loafed.

The males showed a lesser tendency to follow the females during this pre-mated period than was observed in Ohio after mating had taken place. In general, roosting, feeding, and loafing behavior were simi­ lar to that observed in Ohio.

A peculiar behavior of wood ducks I did not notice elsewhere than at the Savannah Refuge was the practice of about one hundred birds spend­ ing entire days in the midst of a cornfield scattered among thousands of mallards. Hawkins et al. (194-1 s395) observed an estimated two thousand wood ducks with a flock of at least twenty-five thousand mallards in a cornfield in Illinois. Whether these Georgia wood ducks adhered to the usual pattern of morning and evening feeding or whether they fed at Figure 31

Pools Among the Maiden Cane Were Used for Loafing Places but Less 120 Frequently Than the Pools Among Trees and Bushes (by E.O. Mellinger) Figure 32

Water Pools Among Buttonbushes Were Much Used for

Loafing Places (by E.O. Mellinger) 122 irregular periods of the day was not determined. Wood ducks in Ohio have been observed through the middle of the day sitting on shelled corn through long periods without eating the corn when it had been placed on their regular loafing logs. Corn placed at a loafing station on the Savannah Refuge likewise was not used through seven days by the ducks which continued to use the loafing place.

Summary

Broods usually remained at water nearest their natal sites through­ out their first two weeks out of the nest but moved to new habitats at about two weeks of age. Such moves frequently resulted in the forma­ tion of congregations of five and six broods. The ducks in these con­ gregations occupied the same areas, but broods usually did not merge.

Broods on ponds moved cross-country from one pond to another, but broods on small streams usually moved downstream to larger water courses. In one case two congregations formed within one-half mile of each other.

The longest distance moved by a brood to a congregation area was 3.5 miles, assuming that the brood followed the water course rather than moving cross-country. The daily range of wood duck broods within con­ gregations was small, and it was largely restricted by favorable escape cover. When two broods merged they were led by one, two, or no females.

Very young ducklings lost from the female did not show their normal fear and followed various moving animals and objects. They readily joined wood duck broods other than their own at such times and some mixing of broods thus occurred, but a reasonable degree of brood purity was main­ tained . 123

Females leading broods made their usual morning and evening feed­

ing flights to areas other than those occupied by the broods. Females

started gathering into small groups when their ducklings were about six weeks of age. Males left the breeding areas on or about May 30, pre­

sumably going into isolation as they underwent their postnuptial molt.

Brood congregations are probably caused by a positive reaction of fe­ males to other adults of the species as well as, and more importantly,

b y the discontinuity of preferred habitat.

A shifting of population occurred when the young had attained the ability of flight, and this resulted in further congregating of wood ducks. Six birds moved at least 8 miles. Birds at the congregation center did not move but furnished the nucleus of the congregation. The largest congregation observed in central Ohio contained 160 wood ducks.

Fall flocks frequently fed in farmers' hog lots, and feeding flights to hog lots were made en masse; feeding flights to areas more remote from human disturbance extended over longer periods and were made by individual ducks or small groups.

Local shifting of populations during 1955 and 1956 occurred within five days of October 1 at all of the Ohio areas studied except the lower

Scioto River. The birds left ponds and lakes and moved to nearby creeks and rivers. Movement usually amounted to less than 15 miles, and ducks moved all directions from the late summer congregation sites. A complex of factors affected the selection of the location of late summer congre­ gation sites. These factors were not always known, but nearness of food was sometimes an important one. 124

Eighteen of thirty recoveries of wood ducks banded in Ohio during the summer of 1955 were of birds shot in Ohio during the hunting season following banding. Population shifts occurred as a result of local duck shooting, and river and creek ducks left areas after being shot at a few times. There is little evidence as to where these ducks went before the end of October, but available evidence suggests they moved to other nearby creeks. Southward migration during 1955 and 1956 took place during the last few days of October or the first few days of

November.

Congregations of 10,000 wood ducks sometimes occur in limited areas in the United States. The birds are actually together in one group only when feeding and roosting. CHAPTER VII

ANALYSIS OF BANDING DATA

The research on the wood duck has increased along with the ducks

population. Much of this research has involved banding, and many

recoveries and returns have been accumulated by the United States Fish

and Wildlife Service. The present analysis makes use of 3,343 recovery

and return records, which were obtained from the Fish and Wildlife

Service during January, 1957. In order to derive most benefit from

this analysis, the reader is advised to make frequent use of the dis­

tribution maps (Figures 33-38) and Tables 7-8.

Comparison of Data from Different States

One should use caution in making comparisons of numbers of ducks

taken in the different states. The figures probably have some compara­

tive value, in spite of differences in hunting pressure and differences

in the rate of reporting bands in different states. Perhaps the most

important variable is the extent of the area of wood duck habitat in

the different states. North Carolina, for example, has a longer coastal

strip of habitat 100-125 miles wide than South Carolina, Consequently,

proportionately more birds should be taken in North Carolina if the wood duck population is the same and hunting pressure in both states is

comparable. More wood ducks banded in Vermont, for example, were taken

in South Carolina than in North Carolina. This may mean either that

there were more wood ducks from Vermont in South Carolina than in North

Carolina, or they were in South Carolina a longer period of time. This

125 126

\03 \

Figure 3 3

Distribution of Recoveries of Wood Ducks Banded in

Illinois (Vertical figures are used for direct

and slanted figures for indirect recoveries) 127

T“>

Figure 34-

Distribution of Recoveries of Wood Ducks Banded in

Maine (Vertical figures are used for direct and

slanted figures for indirect recoveries) 128

Figure 35

Distribution of Recoveries of Wood Ducks Banded in Massachusetts

(Vertical figures are used for direct and

slanted figures for indirect recoveries) 129

- " V "

.— K______^r

o

Figure 36

Distribution of Recoveries of Wood Ducks Banded in

Michigan (Vertical figures are used for direct

and slanted figures for indirect recoveries) 130

4 0 K \ o

w O'

Figure 37

Distribution of Recoveries of Wood Ducks Banded in

New York (Vertical figures are used for direct

and slanted figures for indirect recoveries) 131

rj r

*• “'Cl

— L _

T— '

Figure 38

Distribution of Recoveries of Wood Ducks Banded in

Vermont (Vertical figures are used for direct

and slanted figures for indirect recoveries) T a b l e I I I

Wood Ducks Recovered Before Second Migration Period After Banding 133

Table VIII

Vood Ducks Recovered After Second Migration Period Following Banding + [ I n d i r e c t '

JS—2

fan.

Klee,

N.I. Ohio 13 A

in turn may mean better habitat, perhaps more accessible food, in South

Carolina than in North Carolina.

An important variable which influences the size of the bag from

the banded sample in the banding states is the conditions around the

trapping station and the location of the station. No birds banded in

Kentucky were reported bagged in that state. The birds were banded on

a refuge at the southern boundary of the state, and as soon as they

left the refuge on their southward migration they were out of the state.

There are probably various degrees of variation in the extent to which banded samples are exposed to hunting pressure, dependent on the loca­

tion in the state of the trapping station. Sixteen birds banded in

West Virginia were banded near the Ohio boundary, and only 31 per cent of these were bagged in West Virginia during the first hunting season.

Three additional birds were killed near the banding site on the oppo­ site side of the Ohio River in Ohio. Fifty per cent were bagged the first hunting season in West Virginia and in Ohio near the banding station.

Another important variable is difference in the open season. There was no open season on wood ducks in the Mississippi Flyway in 1956, but there was an open season in the Atlantic Flyway. This increased the probability of birds being reported from the Atlantic Flyway states as compared with the Mississippi Flyway states. In the comparison of num­ bers bagged in Florida and Lousiana a disproportionately large contin­ gent from Ohio was present in Florida in the period covered by the data as compared with a period when the season was open every year in both states. Effect of Hunting

The proportion of the ducks bagged the first year in the state of

banding as compared with other states varied from zero in Kentucky to

67 per cent in Ohio. As has been pointed out, the Kentucky birds were

banded on a refuge at the southern boundary of the state, and they were not exposed to hunting in the state. Explanations were usually apparent when hunters in the state of banding bagged fewer than 50 per cent of the total birds bagged the first year. Most of the recoveries from Ohio were of birds banded at Buckeye Lake and the Olentangy Station, and these are areas of relatively heavy hunting pressure. However, nearly all of the birds left the banding areas before the opening of the hunt­ ing season and moved onto streams where the hunting pressure was prob­ ably near the average for the state.

"When the birds leave the state of banding, the proportion of the total bag from that sample subsequently available to hunters in the state of banding is reduced because many of the birds do not return to their natal area, earlier breeding plaee, or earlier migration route.

Of the banded birds bagged in years after the first year by hunters in the state of banding, Michigan hunters bagged 12 per cent of 82; New

York hunters bagged 16 per cent of 287; Massachusetts hunters bagged 25 per cent of 98; and Maine hunters bagged 30 per cent of 14.0.

If randomness of both the banding and bag samples is assumed, a total of 13 per cent of the Ohio wood duck population was shot in 1955, and 8 per cent of that population was shot in Ohio. If the actual kill was 2.5 times greater than the kill indicated by the reported bands as suggested by Bellrose (1955a:75) for other species of ducks, more than

135 136

32 per cent of the Ohio wood duck population was bagged in the first

hunting season after banding; and more than 19 per cent was bagged in

Ohio. If the cripple loss was 0.5 of a bird for every bird bagged as

reported by Sowls (1955:166) for other species of ducks, A# per cent

of the Ohio wood duck population was lost to hunters in 1955; 27 per

cent was lost to hunters in Ohio. About 70 per cent of the wood duck

recoveries available from the Fish and Wildlife Service were bagged the

first year after banding; 30 per cent were bagged in subsequent years.

Allowing for unreported bands and cripple loss, 21 per cent of the

population was lost to hunters in years subsequent to the first hunt­

ing season after banding; 12 per cent was lost to Ohio hunters. Hunters are thus accountable for killing 68 per cent of the Ohio wood duck pop­ ulation; Ohio hunters are responsible for 39 per cent.

The above estimates are admittedly somewhat theoretical, but all of the calculations are based upon the best available information.

Sixty-eight per cent lost to all hunters and 39 per cent lost to Ohio hunters may be a reasonably good guess of the effect of hunting on wood ducks under conditions existing in 1955. This applies to the entire population to the extent that the sample is representative of that population.

Late Summer Radial Dispersal

Certain summer concentrations of wood ducks in the northern part of the United States disband in late September and early October and disperse in all directions. Most ducks move only 10-25 miles at this time, but some move much farther. An unaged and unsexed bird moved 800 137

miles northwestward, from Barrington, Illinois, to Milestone, Saskatch­

ewan, between September 27 and October 9; an immature female moved 4.00

miles southwestward between September 11 and October 11; an immature

female flew 150 miles northeastward between September 16 and October

A; an adult female traveled 600 miles southward between August 19 and

September 23.

There is a similar radial dispersal of Lotosiana-banded ducks

which, because of the nearby Gulf of Mexico, is restricted to movement

northward, eastward and westward. The radial movement in Louisiana,

however, takes place in October and November, a month later than that

in the North. Louisiana ducks travel to Alabama, Arkansas, Mississippi,

and Texas. The list of illustrations supporting radial dispersal could

be greatly extended with data from various states. This flight seems

to be made rather hurriedly; one bird traveled 800 miles northwestward

in 12 days.

The radial dispersal appears to consist of birds of all ages and

of both sexes. Both ex-breeding adults and locally reared juvenals

participate. The number of females involved and the extent of flights

of females to the northward appears to be more restricted than that of males, though both adult and immature females were recovered at least

100 miles to the northward, eastward, and westward from the banding

place following this radial dispersal. Southward Migration

Southward migration starts in late October, and many of the wood

ducks are south of 40° N . by November 1. Many along the Atlantic

Coast remain in Virginia, North Carolina, and South Carolina through

November and December, while others continue leisurely southward. Two birds banded in South Carolina in October were taken in Georgia and

Florida, respectively, in December. The average dates on which New York wood ducks were bagged in the various states was progressively later southward. The average date for the bagging of 160 ducks in New York was October 22; in Pennsylvania the average date for bagging 15 ducks from New York was November 2; in Virginia the average for 13 was December

3; in South Carolina the average for 47 was December 8; in Georgia the average for 23 was December 14; in Florida the average for 18 was

December 19. The fact that the average date for bagging New York wood ducks in Virginia is nearly a month later than that for Pennsylvania does not mean that the ducks spent a month in traveling the distance between the two states. A much later average date is to be expected in

Virginia, the first state to the southward from New York where many ducks remain throughout the winter.

Not all of the wood ducks left the northern part of their range in the late October exodus; some remained through December. Banded wood ducks were shot in New Brunswick as late as December 29; Quebec, December

19; and Massachusetts, December 20. Some wood ducks apparently remain rather far north through the winter. Some sample winter records are as follows: Massachusetts in January (Morgan and Emery, 1956:235); 50 in

Ontario in mid-January (Baillie, 1951:201); and two at each of two

138 139

stations in Idaho in late January (Low, 1952:208). M.B. Trautman

(personal conversation) observed two males on the Hocking River in

Hocking County Ohio, in mid-January of 1957.

Some of the ducks locally reared in Massachusetts presumably left

the state before the opening of the hunting season. Thirty per cent

of 33 birds bagged in Massachusetts during the first hunting season

after banding were locally reared birds; of 83 bagged outside of the

state, 4.1 per cent were locally reared birds. The young birds are

relatively more vulnerable to shooting at the beginning of the season,

and a significant proportion probably left the state before shooting

started. Evidence that immature birds are more vulnerable to shooting

than adults early in the season is shown by Maine data where the in-state

ratio of adults to immatures was 1:11 as compared with a ratio of 1:4-

outside the state. If the immatures were more vulnerable to shooting

early in the season, the ratios in the natural population would, of

course, be different inside and outside of Maine. The ratio of 1:4

may approximate the true ratio of the wild population at the season

represented.

There was some indication, particularly among ducks from Vermont

and Michigan, that the migrating birds followed waterways. Relatively

large numbers of ducks from the Upper Peninsula of Michigan were taken

in northern Illinois, rather than in Ohio and Indiana more directly

southward from the banding place; relatively large numbers from north­ western Vermont were recovered southwestward along the Saint Lawrence

River, Lake Ontario, and Lake Erie. Most eastern-banded ducks which were recovered when migrating were along the coast within 125 miles of the Atlantic Ocean. Wood ducks migrating along the Atlantic coast from

Maryland southward seem to be largely restricted to the Princess Anne,

Pamlico, Talbot, Penholoway, Wicomico, and Sunderland Terraces. The

greatest swamps are on the Princess Anne and the Pamlico Terraces, but

the Talbot, Penholoway, Wicomico and Sunderland Terraces also have their

share of swamps (Fenneman, 1933:31-32). At about 32° N., the ducks

spread to the westward across Georgia into Alabama. This westward

swing corresponds with the Coharie Terrace which, although characterized by its gravel substratum, contains some extensive swamps (Fenneman,

1938:30-41). Movement along the Atlantic Coast is chiefly associated with favorable woodland swamps, rather than with the Atlantic Ocean per se. The same relationship probably holds for migration along the

Saint Lawrence River and Great Lakes, as well as movement cross-country independent of waterways.

When the coastwise nature of the migration is considered, it must be remembered that the wood duck does not occupy salt water. When wood ducks occur near the coast, they are actually far enough from the ocean to be in fresh water.

The majority of the wood ducks produced in southern Canada and the northern United States east of 80° W. winter along the Atlantic Coast from Maryland southward. As the ducks proceed southward through Virginia,

North Carolina, South Carolina, and Georgia, numbers continually drop from their ranks. The numbers dropping from the migrating group increase progressively southward through Virginia, North Carolina, and South

Carolina^ thus numbers remaining through December increase southward through these states. The largest numbers remain in South Carolina and U1

Georgia through December. Rather large numbers of ducks continue to

29° N., but very few proceed beyond this line. An occasional bird moves slightly farther southward into Florida, and one bird from Ver­ mont was taken in the Bahama Islands. This species is a common resi­ dent over most of Florida (Howell, 1932:143) except the extreme southern portion,

The limit of southward movement of large numbers of wood ducks from

New England into Florida coincides with the limits of the range of the water oak. Several other species of oaks reach the southern limits of their range in northern Florida: yellow oak, Spanish oak, swamp Spanish oak, black jack oak, willow oak, white oak, post oak, over-cup oak, and basket oak (Hough, 1907:157). There are, however, other species of oaks with ranges continuing into southern Florida: southern red oak, laurel oak, and live oak. The wood duck feeds on acorns other than from oaks with southward distribution limited to northern Florida, but the south­ ward movement of large numbers of ducks is probably correlated with available food in the form of acorns. This point cannot be incontro- vertibly established, however, with the available information.

Wood ducks on the Pacific coast do not migrate nearly as far south­ ward as those on the Atlantic coast. Birds from British Columbia,

Montana, Oregon, and Washington migrate southward only into northern

California. Their southward movement extends little beyond 38° N.

In the East this would be equivalent to all of the birds wintering in northern Virginia and Maryland or farther northward. The majority of the eastern population migrates southward beyond 33° N,; much of it migrates beyond 35° N. U 2

Wood duck migration on the Pacific Coast is markedly coastwise,

but the coastal lane is not invariably followed. A notable exception

is a male banded in California on March 11, 1939, and recovered on the

Platte River in Nebraska, about 1,100 miles eastward from the banding

place, on November 30, 194-0. A southward migration had intervened the

two dates of capture, and it is not known whether the bird had first

flown eastward in a postnuptial radial dispersal or whether it first

migrated to the southern part of the United States and later northeast­

ward to its ultimate destination. This is the only one of 104 recoveries

of wood ducks banded along the Pacific Coast which hints that the Pacific

Coast migration is not strictly coastwise.

As is indicated by recoveries in western Pennsylvania, Ohio, West

Virginia, and Tennessee, of birds banded in Maine, New York, and Ver­ mont, some wood ducks take the direct route to the south-central United

States rather than following the Atlantic Coast terraces. It cannot

readily be determined how far westward birds continue after moving south­

ward along the Atlantic Coast to southern Georgia, but apparently some ducks take approximately the shortest route from Maine and Vermont to

Louisiana and Mississippi. Likewise, some of the birds from Illinois apparently travel southeastward to Georgia and Florida, instead of down

the Mississippi Valley and eastward.

Wood ducks in the central part of the United States migrate south­ ward on a less well-defined lane than that used by Atlantic Coast ducks.

This lane for birds migrating from Illinois may be said to include ex­ treme western Kentucky, extreme western Tennessee, Missouri, eastern

Oklahoma, and Arkansas, taking most of the birds to Louisiana, Mississip­ pi, and Texas. The lateral extensions of the migration vanguard are much 143 broader in the central than in the Atlantic lane. Birds from Illinois fly to Colorado and northeastern North Carolina on their first south­ ward flights, thus forming a migration vanguard the lateral extension of which moves at an angle of 14-0°.

Available data indicate that wood ducks from Ohio migrate in about equal numbers to the south-southeastward, including chiefly Florida, and to the south-southwestward, including chiefly Louisiana and

Mississippi. Ohio-reared birds which leave the state may be about equally divided between the Atlantic and the Mississippi Flyways, and the Mississippi Flyway population is not entirely safe-guarded when protection does not extend into the Atlantic Flyway. There was a higher percentage of an Ohio-banded sample bagged in Florida, for example, in a season when hunting was not permitted in Ohio (1956) than was bagged in Florida in a season when hunting was permitted in

Ohio (1955). In 1956, 1,64 per cent of 366 Ohio-banded birds were shot in Florida. This represents a relatively small difference, but it must be remembered that this difference of nearly 1 per cent must be con­ sidered in relation to 13 per cent, or a figure representing the total bag if the bird had not been protected in some states.

Of 36 birds recovered following more than one migration period after banding from Michigan, 13 were taken outside of the Mississippi

Flyway: six from the Atlantic Flyway and seven from the Central Flyway.

Both males and females migrate in the lateral extensions of the migrating vanguard. It cannot be determined whether females once migrat­ ing from New York to Louisiana, for instance, follow the same route in successive years. Males, of course, do not follow the same routes in 1 u

successive migrations. There is some indication that immatures have a

stronger tendency to travel in the lateral extensions of the migration

lanes than adults, though adults as well as immatures travel laterally.

Indirect recoveries of birds banded in South Carolina during the

fall and winter months indicate that female wood ducks tend to return

to the same wintering area and to the same migration route in successive years. Results from South Carolina, however, may be somewhat meaning­ less, since nearly all of the birds from the northeastern United States migrate southward to or through the eastern side of South Carolina.

The majority of wood ducks travel less than 1.300 miles from their nesting to their wintering places. One bird, however, travelled at least 1,600 miles, from Penobscot County, Maine, to Livonia,

Louisiana.

Travel was sometimes relatively rapid. A bird recovered in

Illinois on October 18 had been banded on the Upper Peninsula of Michi­ gan on October 13, five days before. This bird had travelled about

325 miles south-southwestward and averaged 65 miles per day. It is not known how long after being banded this bird delayed the start of its flight from Michigan or how long it had been in Illinois before being recovered.

Northward Migration

Vood ducks probably migrate northward in pairs or small groups of pairs. Sowls (1955:22-23) observed that mallards and pintails migrate northward in groups of from 1-35 birds. The northward movement of mallards proceeds slowly, taking two months for some of the birds to travel from 145

Louisiana to Manitoba.

Available data on wood ducks do not indicate whether the early

arrivals failed to move as far south as the later arrivals, started

earlier from the wintering area, or proceeded more rapidly in their northward movement. Early arrivals probably consist of birds that winter far to the north of the area of concentrated abundance. North­ ward movement of wood ducks extends from late February to late April or even May in the northern part of the bird*s range.

With a few wood ducks wintering far to the north of the area of

concentrated abundance.it is difficult to determine when northward move­ ment actually starts. The first northbound migrants reach central Ohio by about February 20. The earliest arrivals in 1957 were seen on the

Scioto River and on Blacklick Creek, where swift currents prevented freezing of the water. Mumford (1952:19) reported that early arrivals reached Hovey Lake, Indiana, February 13-24.. Early arrival dates given by Bent (1923:170-171) are March 16 for central New York, March

10 for central Ohio, March 24 for Heron Lake, Minnesota, March 25 for

Ottawa, Ontario, and April 2 for southern Manitoba. Most of the arrival dates reported by Bent were probably from observations made when wood ducks were scarce, and this may account for the relative lateness of his dates. They also were of birds incidentally seen in connection with general bird observations, rather than in connection with special research on wood ducks. Bent*s data, however, indicate that the early arrival dates are progressively later northward, which is as might be expected.

Sowls (1955:15-17) observed that locally breeding pintails and mallards, rather than northbound transients, were the first to arrive 14-6

at Delta, Manitoba, in the spring. I obtained little direct evidence

on this point for the wood duck, but the fact that some of the earliest

arrivals were continuously observed at nesting sites later occupied

suggests a similar pattern for the wood duck.

Return to Natal Area and Earlier Nesting Site

Adult females have a strong tendency to return to the same nesting

site year after year. This is almost, if not entirely, invariable. I

took one female from nests on the same pond at the Olentangy Station in

1955, 1956, and 1957. Miller (1952:4) reported that six of seventeen

females nested in the same box through two seasons. Thirty-one females

banded in Vermont produced a total of seventy-two returns as follows:

twelve females returned one and two years after banding; four returned

one, two, and three years later; one returned one, two, three, and

four years later; four returned two and three years later; two returned

two, three, and four years later; one returned two and four years later;

two returned three and four years later; two returned three, four, and

five years later; one returned four and five years later; two returned five and six years after banding. All except 15 of 109 females were

taken when nesting during May, June, and July. Of the seven taken in

April, three were known to be nesting. Eight were taken early in the

season and may not have been nesting.

Immature females also have a tendency to return and nest at or near the place of nativity. Of 144 females returning during the nesting sea­ son in Massachusetts, 111 were adults when banded, 14 were immatures, and

19 were locally reared birds. McLaughlin and Grice (1952:255) reported 14-7

that year-old females migrate northward somewhat later than the old

females, occupy sites in their natal regions, and later spread into

the surrounding areas. Sowls (1955*37) considered this to be the

situation with ducks in general.

Females may not always return to their natal areas or earlier

nesting sites. A local female was banded in Maryland on September 6,

1952, and recovered in New York, about 300 miles northward from the

banding place, on October 15, 1955. This female overshot its natal

area by 300 miles. Of thirty-four Vermont-banded birds which were

recovered in subsequent years more than 100 miles north, east, or west of the banding place, eight were adult males, three were adult

females, seventeen were immature males, and six were immature females.

Three of the females were taken more than 100 miles north of the band­

ing place; two were taken more than 100 miles east; four were taken more

than 100 miles west. A young female banded in Maine on July 7, 1953, was recovered in Michigan 500 miles west-southwest from the banding

place on October 1, 1954-. This is the point most distant north, east, or west from the natal or breeding area where a female was later recover­ ed.

There were no females for which both the banding and the recovery dates occurred during the nesting or brood seasons, so it is impossible to state whether or not females always return to their natal or earlier breeding sites. Certainly, they sometimes nest at or near their hatch­ ing places. Besides non-homing there are two possibilities to explain the frequent wide separation of banding and recovery localities: (1) the birds may have been reared or they may have nested earlier near their U 8 banding places and moved in an early-fall radial dispersal to the recovery locality; or (2) they may have moved in an early-fall radial dispersal from their natal or earlier nesting area to the banding locality.

Both adult and immature males shoved little tendency to return to the natal areas, earlier breeding sites, or to earlier migration routes.

Lincoln (1939*75) reported that among ducks the female influences the movements of the male. Ducks from many parts of the United States mingle to some extent in the South. Birds from Vermont and Wisconsin may simultaneously occupy the same swamp during the pairing season.

Observations I made at the Savannah Refuge in January before the ducks were mated indicated that females followed males or males followed females with equal regularity before pairing. Pairing takes place before northward migration. After pairing the male follows the female in all local movements. The male probably follows the female in northward migration when she returns to the general or exact locality where she was reared or where she nested earlier. The probability of a male re­ turning to his place of nativity or earlier breeding area is propor­ tional, then, to the number of females from his home area occupying the same swamp at pairing time. The extent to which local populations remain together through the winter is not known, but wood ducks certain­ ly migrate southward in groups to some extent. Gigstead (1938:609) re­ ported that wood ducks leave Illinois in the fall in large concentrated flights. On the other hand, Bent (1923:167) reported that they migrate in small flocks, probably family parties. Wood ducks reared in the South mingle during the winter with birds

from the North; when pairing takes place, a southern male may pair with

a northern female. The male then follows the female northward to her

nesting site. A locally reared male, banded in Louisiana on August 6,

1938, and recovered in Minnesota on October 3, 1939, probably reached

the North in this manner. Similarly, males from the North may remain

in the South through the nesting season after pairing with southern

females. This is probably the explanation for the presence in Arkansas

on May 18, 1951, of a male banded in New York, on September 27, 194-9.

There is no ready explanation for the presence in South Carolina on

July 9, 1954-, of a female banded on September 30, 1953, in New York.

This female may have been crippled and unable to migrate or there may

have been an error in the date the bird was found.

Males from any given point in the northern part of the wood duck's

range may pair in the winter with females from any part of the United

States or southern Canada. This makes for an increased dispersal of

males from a given point throughout the east-west and north-south

spread of their range in successive years. Actually, the extent of the

dispersal of males is restricted by the fact that most females adhere to

a relatively narrow migration lane. Among New York recoveries, there

were more indirect recoveries (11.3 per cent) taken from west of the

Ohio and Mississippi Rivers as compared with the direct recoveries

(7.5 per cent) from that region. This results from the movement of males from the Atlantic Coast area to the more western area. Recovery of Pen-reared Birds

The migration of released pen-reared birds appears to parallel that of wild-trapped birds. The kill in the states where the birds were released may be slightly higher than for the wild birds, but the pen- reared birds apparently join the wild birds in their migrations.

McCabe (1947:108) reported that some wood ducks released near Madison,

Wisconsin, remained through the following winter after release. He reported that ducks of both sexes returned to nest the following two years. It seems probable that the males present in the following nesting seasons may have been continuously in the area.

Summary

Analysis was made of data from 230 wood ducks banded in Ohio during 1955, and of 3,843 recovery records obtained from the United

States Fish and Wildlife Service.

Fifty per cent or more of the birds shot the first hunting sea­ son after banding are normally shot in the state where they were banded.

Twelve to 30 per cent of birds shot in years subsequent to the first year after banding were shot in the state where they were banded. On the basis of the Ohio-banded sample and published information, it was calculated that hunters kill 68 per cent of the Ohio wood duck popula­ tion; Ohio hunterskill 39 per cent.

Radial dispersal occurs among wood ducks during late September and early October. Most of the ducks move only 10-15 miles, but some travel as much as 800 miles northward. This dispersal involves birds

150 151

of the southern as well as the northern part of the wood duck*s range.

Adults and young of both sexes are involved. Some birds reared in a

given state are thus lost by hunters in that state.

Some wood ducks remain in the northern part of their breeding range throughout the winter, but most of them migrate southward. South­ ward migration starts in late October. Most of the ducks from the northern part of the wood duck * s breeding range do not migrate south­ ward beyond 38° N. in the western part of the United States; the majority of the eastern population travels southward beyond 35° N. There is a fairly distinct Atlantic Coast migration lane which follows the swampy habitat within 125 miles of the Atlantic Coast. Southward movement appears to continue at least until January 1. Most of the ducks in the eastern United States do not move southward beyond 29° N. Most of the wood ducks reared or breeding in the New England states winter in the

Atlantic coastal states; some winter in the Gulf coastal states. Most of the birds reared or breeding in the northern mid-western states winter in Arkansas, Lousiana, Mississippi, and Texas; some winter in

Florida and Georgia. In southern Georgia, the Atlantic coastal lane extends westward into Alabama. Some birds from the New England states migrate through Kentucky and Tennessee to Louisiana and Mississippi.

There was an increase in the number of Ohio birds bagged in Florida in 1956, when there was no open season on wood ducks in Ohio, as compared with 1955 when there was an open season in Ohio.

The longest southward migratory flight was 1,600 miles. One bird was known to have traveled at least 65 miles per day for 5 days. 152

Northward migration starts in late February. Locally breeding

birds are the first to arrive; northbound transients arrive later.

Females have a strong tendency to return to the same nesting

place in successive years. First year birds usually return to their

natal area or spread into the surrounding area. They may sometimes

nest as much as 200 miles north, east, or west of their natal area.

Both immature and adult males seldom or never return to their

natal area or earlier breeding ground. "When in the South, males may pair with females from any part of the wood duck’s range. The female returns

to her natal or earlier breeding area, and the male goes with his mate.

Released pen-reared birds exhibited a migration pattern similar to wild birds, except that some males were taken near the point of release the year after being released. CHAPTER VIII

FOOD AND FEEDING

Since living birds were necessary in working toward nearly all of the major objectives of this study, the killing of birds for study of food items was largely avoided. Some incidental observations on food items and feeding habits were made, and other information was gleaned from the literature.

Food

Nhen swimming about on the surface of the water, ducklings were observed pecking at the surface of the water or low vegetation, and it was assumed that they were feeding on insects., Bent (1923:166) stated,

"A large part of its food consists of insects which it finds on the surface of the water or on the leaves and stems of aquatic plants, such as beetles, mayflies, locusts, and various creeping insects."

Audubon (1843:274.) stated that food of the young consists of aquatic insects, flies, mosquitoes, and seeds. Ducklings in captivity were seen catching insects resting on the grass in their pen. Hanson (1951:70) reported that a large portion of the food of captive-reared ducks was insects attracted to a light in the ducks* cage, and he attributed his rearing success to this fact.

After the ducklings are several weeks old, seed is an important part of their diet. A duckling 5-6 weeks old was collected on June 9,

1956, near a pond at the Olentangy Station; its crop contained 800 seeds of Eleocharis obtusa, 44 seeds of Polygonum muhlenbergia and 8 seeds of

153 154-

Scirpus gp. Twenty-six seeds of Geum avens were removed on June 1A,

1956, from the gullet of a wood duck 5-6 weeks of age taken in a wood­

land pond at the Olentangy Station. Captive ducklings were frequently

seen eating grass and leaves from small corn plants.

Food was present in the crops of only two females taken from their

nests in areas where no feeding stations were being maintained. The

crops of these two birds contained corn, and information furnished by

local farmers indicated that hog lots were the most likely source of

this corn. In some areas under observation, the young ducks started

feeding in farmers' hog lots soon after the ability to fly was attained,

and the females probably led the young to the feeding sites which they

had been using earlier. All observations on food and feeding of wood

ducks in Ohio indicated that corn obtained in farmers' hog lots is the

chief food through August and September. Reports from farmers agreed

that wood ducks frequently feed in hog lots at this season of the

year. Paul Gibson, a duck hunter and bird watcher, reported that wood

ducks were seen feeding on shelled corn with a flock of turkeys on a

farm near Plain City, Ohio.

Through August and September corn seemed to be a preferred food, or perhaps the birds do not readily change from eating one food to eat­

ing another. At Buckeye Lake in 1956, an abundance of elderberries, blackberries, and wild grapes hung low over the water's edge near a bait­

ing station, but these fruits were used little or not at all by the 125 ducks which fed at the station, even though the supply of corn was some­ times exhausted. In 1955, at Magee Marsh, dogwood was much used by wood ducks. All of the fruit on the bushes within about eighteen inches of the water was eaten. 155

Gigstead (1933:608) reported that wood ducks feed on chafua, wild millet, and smartweed in August in Illinoisj in September they feed on sago. Stollberg (1950:217) noted that seeds of the oaks and bur-reed were much used by wood ducks but not by other species of ducks.

Schoffman (194-6) noted that the wood duck does not compete with the coot at Reelfoot Lake in Tennessee where cow-lily makes up 68 per cent by volume of the food of the duck. Pirnie (1935:307) reported the results of examination of stomachs of three wood ducks collected in August.

These ducks had eaten pondweed seeds and tubers, unopened blossoms of wild rice, dogwood seeds, bulrush, sedge, and yellow water-lily seeds.

Animal matter consisted of a snail, a spider, a water beetle and frag­ ments of other beetles, and grasshopper fragments. Beyer, £t al.

(1907:320) reported that the water chinquapin is an important food of the wood duck in Louisiana, Martin and Uhler (1939:72) stated that southern red oak, willow oak, and pin oak are much used by mallards, black ducks, and wood ducks. During the period of January 1-9, 1957, at the Savannah Refuge, all evening feeding flights of wood ducks vent to oak groves, and the ducks were believed to be feeding largely on acorns from trees in these groves.

In Ohio on September 4, 1956, a number of ducks which had gathered at a baiting station were observed swimming rapidly about and seemingly capturing insects from the surface of the water. A large gathering of

Gyrinidae had been observed at the spot some 15 minutes earlier, and it is assumed that the wood ducks were feeding on these beetles.

Coutler (1955:264) examined the crop and gullet contents of 39 wood ducks taken in Maine during the early spring. The food of these ducks was 90.9 per cent vegetable matter and 9.2 per cent animal matter. The

vegetable matter consisted of Cyperaceae, 36.4- per cent; Sparganiaceae,

35.7 per cent; Rosaceae, 4-.3 per cent; Cornaceae, 2.0 per cent; Hydro-

charitaceae, 2.0 per cent; Alismaceae, 1.0 per cent; miscellaneous

plant food, 2.3 per cent; and unidentified plant food, 7.1 per cent.

The animal food consisted of mayflies, beetles, flies, treehoppers,

grasshoppers, stoneflies, whirligig beetles, water striders, and moths.

Also in Maine, Norton (1909:439) reported that wood ducks have shown a

fondness for bulbs of Sagittaria. In Iowa early in May, Dreis (1951:

U 6) reported observing wood ducks feed on two species of duckweed,

Lemna minor and Spirodela oolvrhiza. In Michigan during April, Pirnie

(1935:307) collected a bird which had fed on fragments of arrow arum and bur-reed seeds, and one mermaid weed seed.

From examination of 4-13 stomachs Mabbott (1920:38,4.6) reported that 90.19 per cent of the food of the wood duck consists of plant material: duckweeds, 10.35 per cent; cypress cones and galls, 9.25 per cent; sedge seeds and tubers, 9 .14- per cent; grasses and grass seeds, 8.17 per cent; pondweeds and their seeds, 6.53; acorns and beech nuts, 6,28; seeds of water lilies and leaves of water shield, 5.95 per cent; seeds of water elm and its allies, 4--75 per cent; seeds of smart- weeds and docks, 4-.74- per cent; coontail, 2.86 per cent; seeds of arrow arum and skunk cabbage, 2 .4-2 per cent; seeds of bur marigold and other composites, 2.38 per cent; seeds of buttonbush and allied plants, 2.25 per cent; seeds of bur-reed, 1.96 per cent, wild celery and frogbit,

1.31 per cent; nuts of bitter pecan, 0.91 per cent; grape seeds, 0.82 per cent; seeds of swamp privet and ash, 0.72 per cent. The remaining 9.4- per cent was made up of a number of minor items. The animal food

made up 9.31 per cent: dragonflies and damselflies and their nymphs,

2.54- per cent; bugs, 1.56 per cent; beetles, 1.02 per cent; grasshoppers

and crickets, 0.23 per cent; flies, ants, bees, and wasps, 0.07 per cent

miscellaneous insects, 0.97 per cent; spiders and mites, 0,.63 per cent;

crustaceans, 0.08 per cent; miscellaneous animal matter 2.71 per cent.

On January 13, 1957, digestive tracts were removed from three ducks

in bags of hunters on the Santee River in South Carolina. Two of these

digestive tracts contained nothing but a small amount of sand in the

gizzard; one contained corn. The hunters who killed the duck containing

corn suggested that no corn was available in the area unless it was

used for bait by hunters.

When acorns become available, wood ducks use them for food. A wood

duck collected on Big Darby Creek on October 16, contained only acorns

in its digestive tract; one entire acorn was in the crop, one entire

acorn was in the glandular stomach, and a pulverized mass of about one

acorn was in the gizzard. These were acorns of the white oak, 21 x 13 millimeters in size, and they seemed rather large for a wood duck to

swallow entire. Morris (1930:4-20) reported that acorns of white oak

swollen to three-fourths of an inch in diameter were eaten by wood

ducks.

Equally impressive is the ability to grind acorns through muscular

contractions of the gizzard. Mabbott (1920:44) reported that the wood duck feeds extensively on nuts of the bitter pecan, a nut with a hard­ ness equal to that of the northern hickory nut. 158

The acorns are normally picked from the ground, and I have never

seen a wood duck perch on the branches of a tree when obtaining food.

At Magee Marsh in 1955, fruiting dogwood lined the water’s edge at many

places. Only those fruit were eaten which could be reached by the birds

from the surface of the water, and all that could be thus reached were

eaten.

During the fall and winter months, difficulty was experienced in

getting wood ducks to feed at baiting stations when they were accustomed

to feeding at other locations. Ducks continued to make twice daily

visits to the feeding sites and were not attracted by food placed

elsewhere and seemingly more favorably located in relation to their

natural habitat. They continued to fly to accustomed feeding places

at the regular feeding time several days after the food supply was

exhausted. Food placed at fall and winter loafing stations was con­

sistently left without being utilized, although the ducks sometimes

spent hours loafing on the feed through the middle of the day, and food

was scarce at their regular feeding places.

Feeding

Pre-flight wood ducks were trapped at various hours, and they

appeared to feed at all times of the day. In addition to trapping re­

sults, all-day feeding was indicated by observations of ducklings at

feeding stations. After the ability to fly was attained, the ducks

apparently fed only in the morning and evening. In four observations,

the morning feeding flight started 30 minutes before sunrise when it was yet too dark to make a reading with a photographic exposure meter. 159

The ducks left the roosting places in groups of less than twenty birds

and flew to the feeding places. Small groups sometimes continued flying

to and from the feeding site for three or more hours, or until about

9:30 A.M. In five observations of flights to hog lots on the Scioto

River, feeding was completed and all ducks left the area within an hour

after their arrival. When the hogs were not feeding at the arrival of

the ducks, the ducks went directly to the corn pile and proceeded to

feed on corn shelled by the hogs. When the hogs were feeding at the

arrival of the ducks, the ducks circled overhead a great deal before

alighting. When finally alighting, they selected a spot which had been

earlier used as the hog feeding place and now contained less available

corn.

When the feeding place was near wood duck habitat, the birds

started gathering at the feeding place in mid-afternoon. Scattered

individuals and small groups continued to arrive through the remainder

of the afternoon. When feeding in farmers' hog lots, the ducks some­

times gathered during this afternoon period elsewhere than near the feeding place, and they then flew together in the evening to the feeding

site. This gathering movement of 100-150 ducks extended through several hours; consequently, there is believed to be no relationship between the hour of this movement and light intensity.

Although gathered around the feed, the birds did not eat immediately on their early arrival in the afternoon. Soon after 5:30 P.M. E.S.T., the wood ducks started to feed, and most of the ducks in the group started at the same time. The starting time of actual feeding varied 50 minutes in six observations within seven days of each other. The light intensity 160

varied from 400-2,500 foot candles at the times feeding started in

these six observations. There appears to be no relationship of feed­

ing time with light intensity.

When ample corn was available, the ducks filled their crops in a

very short period. The crop was filled to capacity, and the esophagus

was also filled to the bird*s mouth. As many as 190 grains of corn or

98.9 grams were removed from one bird {Figure 39). The corn sometimes

swelled in the crop after being eaten, and the capacity of the crop was

exceeded. In such cases the ducks disgorged some of the corn, and 12

ducks kept in captivity over night disgorged an averaged of 22 grains

each.

S u m m a r y

Newly hatched wood ducks feed largely on insects found on the sur­ face of the water or low vegetation. After the ducks are several weeks of age, seeds are their most important item of diet. During the nesting season and late summer in Ohio, corn is the most important food item of adults. Acorns become the chief item of food in October as they continue to be during the winter. Results of several published reports of studies of wood duck foods are presented.

Pre-flight wood ducks feed at all hours of the day; post-flight ducks feed chiefly in morning and evening. Morning feeding flights started about half an hour before sunrise; evening feeding started about one and a half hours before sunset. Gathering into pre-feeding flocks started 2-3 hours before the feeding time. No relationship was found between feeding times and light intensity. Figure 39

Corn and Sand Removed from Digestive Tract of Trapped Wood Duck

(on left from crop; upper right from glandular

stomach; lower right from gizzard) Feeding was completed in 10-30 minutes when abundant feed was available. The ducks filled their crops and gullets to the mouth.

After swelling of the corn in the crop, the capacity of the crop was exceeded, and excess corn was disgorged. CHAPTER IX

ROOSTING AND LOAFING

Roosting

Ducklings just out of the nest spend their nights being brooded by the female near water. The only ducklings actually found being brooded were on the flat top of a railroad tie among other drift on the

Scioto River. The brooding female was completely unconcealed, but, because brooding is so seldom observed, it is thought that a more con­ cealed situation is ordinarily chosen. Probably female wood ducks ordinarily brood their young hidden among vegetation adjacent to water.

Ducklings 5-6 weeks of age sometimes spend part of the night on the water among the brush or plant cover. When going up the Scioto

River nearly an hour after sunset on June 26, 1956, I found a brood of ducklings on the water near where a brood 5-6 weeks old had been seen about two hours earlier. The ducks found in the darkness were ob­ viously unable to fly and were thought to be the same ones earlier noted. When first noticed in the darkness these ducks were near the shore among overhanging tree branches.

In the late summer when post-brood concentrations were formed at

Buckeye Lake, all of the 160 ducks present in one general locality gathered at the same roosting site. This was in a shallow bay where approach from the land was prevented by a marginal strip of cat-tails several hundred feet wide, and approach from the water was prevented

163 164 by a similar strip of spatter-dock (Figure 40)• The birds roosted on the water among the spatter-dock at the edge of the cat-tails. Most of the ducks gathered into a small area, but a few scattered small groups alighted as much as several hundred feet from the main group and re­ mained where they alighted, at least until advancing darkness destroyed visibility.

I found no roosts of the fall concentrations on the Scioto River.

A farmer along the river informed me that in the late summer of 1954 there was a wood duck roost along the river, and that the roosting ducks perched on the limbs of trees which had fallen into the river.

He stated that the ducks gathered on a few large limbs which were covered with droppings.

In the extensive marsh of the Savannah Refuge, it was not feasible to follow the ducks by land or water to their roosting place. More than two thousand wood ducks flew to an inaccessible pond. Most of the ducks which participated in the roosting flight went to this pond.

Several small groups, however, were known to roost on small pools among the maiden cane. The roosting flight started half an hour before sun­ set and lasted until dark. The sharp decrease in numbers just before visibility was lost indicated that the flight was nearly completed before dark.

In an effort to determine whether correlation existed between light intensity and time of evening flight, light readings were taken with a photographic light meter before and during the period of evening flights. The meter was aimed directly overhead. In five observations of the times of the evening feeding flights, there was no close correla- Figure 4.0

Wood Ducks Roosted Among the Spatter-dock Near the Cat-tails 166

tion between times of arrival and light intensity. No roosting flight

occurred before the light intensity fell to 200 foot candles, but be­

low that level first flights occurred down to 30 foot candles. Peaks

of flights occurred at light intensities from 2-50 foot candles. There

appeared to be little or no relationship of time of roostiig flights

with light intensity.

Loafing

After the wood ducks fed in farmers' hog lots, they left the feed­

ing areas en masse and flew to the water areas. A few individual ducks

and small groups broke away from the flock now and then and went to

scattered spots where they spent the mid-day loafing and preening.

Reassembling for the evening feeding flight started in mid-afternoon.

When the ducks fed at baiting stations immediately within their habitat,

they left the feeding place in small groups and were more scattered

through the day than when they left their feeding place en masse.

On the Savannah Refuge, where several thousand wood ducks were present,

minor concentrations occurred at loafing places where the ducks flew in

small groups. This behavior suggests that concentrations of loafing

birds result from independent selection of the same loafing place.

In the fall and winter after feeding is largely or entirely com­

pleted in two short periods in the early morning and late evening, a

large part of the 24-hour period is left. Most of the life of the wood

duck is spent roosting, preening, and loafing, with the latter punctuat­

ed with occasional drinks of water. When loafing, the birds sit on the water, emergent wood, shore or low branches of trees overhanging the 167

water, with their necks drawn in toward their bodies, preening, or with

heads laid on their backs, seemingly sleeping. Occasionally some will

swim slowly about for short distances (Figure 4.1).

S u m m a r y

One female was observed brooding young in an exposed situation, but

brooding birds are seldom seen, and because captive ducklings spend much

time under the brooding female, I believe that wild birds are usually

brooded among protective cover. Ducklings 5-6 weeks of age may some­

times spend the night on the water among plant cover. All of the late

summer roosting sites located were on water among emergent plants.

Roosting flights started about one-half hour before sunset and con­

tinued until dark. There appeared to be no correlation of the time of the roosting flight with light intensity.

Wood ducks during the fall and winter spend most of the day in groups of various sizes preening or loafing. Figure 4-1

Occasionally Some of the Birds Swim Slowly about for Short Distances CHAPTER X

NESTING AND BROOD HABITATS

Water, trees, and brood cover are three requisites for wood duck habitat, but the nature and position of these may vary. If nesting places cannot be found nearer, wood ducks will nest one to two miles from water. Hawkins et al. (1941:393) reported good results with boxes as muoh as three-fourths of a mile from water, and the utilization of boxes in upland woods was equal to that in bottomland woods. Upon leaving the nest, the female soon leads the brood to water. This water may be only a small pool a few feet in diameter, a pond, a lake, a small brook, or a river. An ideal habitat for wood ducks is shown in Figure 42.

During 1956, 501 nesting boxes were checked in the four habitats: large river, small stream, lake, and farm pond (Table IX). Utiliza­ tion was best at Buckeye Lake, or the large lake habitat, where 15.8 per cent of 37 boxes were used by wood ducks. Buckeye Lake is particu­ larly attractive to wood ducks because of its relatively shallow water and abundance of emergent vegetation. The second highest rate of nest­ ing box utilization, 9.5 per cent, was found on farm ponds where 250 nesting boxes were checked. The rate of nesting box use on farm ponds might have been still higher except for the fact that 96 boxes of a little-used type were involved in the test. The utilization rate of

7.7 per cent on the Scioto River, or the large river habitat, might have been better except for the high density of the boxes: 104 boxes were erected in an 11-mile section of the Scioto River. There were

169 Figure l £ .

A Woodland Pond Affords Excellent Wood Duck Habitat 171

Table IX

Comparative Utilization of Nest Boxes in Different Habitats in 1956

Large Small Large Farm river streams lakes Ponds

No. nest boxes checked 10A 90 57 250

No. nest boxes used bv wood ducks 8 L _9... 23

Per cent utilization of nest boxes 7,7 L.L 15.8 9,5 .

No. successful nests 7 3 7 16

Per cent successful nests 87.5 75.0 77.7 76.2*

No. destroyed bv raccoons 1

Per cent destroyed bv raccoons 12.5

No. nests deserted 1 2

Per cent nests destroyed 11.1 8.3

No. dump nests 1 1 2

Per cent dump nests 25,0 11.1 8.3

No. nests disrupted bv starlings 3

Per cent disrupted by starlings 13.0

^Desertions caused by this study are discounted in this calculation. 172

numerous large sycamore trees adjacent to the river, and these trees

contained many natural cavities. The small stream habitat, with a

nesting box utilization of only 4*4 per cent, must be assumed to be a

marginal habitat.

In Massachusetts, McLaughlin and Grice (1952:245) found a higher

rate of usage of nesting boxes over permanent and temporary water than over dry land areas. Utilization of boxes was 51 per cent over perman­

ent water areas, 47 per cent over temporary water, and 20 per cent over land areas, Lee (1956:2) found that boxes erected over marsh and water habitat were used more than boxes over land or swamp areas, none of 10 boxes in a swamp being used. McLaughlin and Grice (1952:253) found larger average clutches in boxes over water (13.1) and marsh (14*2) areas as compared with boxes over dry land (10.6). A greater number of dump nests in the water and marsh habitats indicates greater pressure for nesting places in these two habitats as compared with land habitat.

Through use of nesting boxes, larger nesting populations were built up on marshes than in wooded areas and on small streams.

Little difference in the survival of broods was found in the var­ ious Ohio habitats (Table X). Six unusually small broods 5-6 weeks of age of four and five ducklings were observed in a 10-mile section of

Alum Creek, and this small brood size was thought to be correlated with an unusual abundance of minks. Farmers interviewed along the creek re­ ported frequent trouble from minks destroying chickens. Brood sizes in most areas were larger than those found on Alum Creek. In U-tests

(Siegel, 1956:116) for differences of brood survival in the different habitats (Tables XII-XVII), significant differences were found in the Table X

Numbers of Ducklings in Broods at Various Stages of Development

in Different Habitats

On Farm Ponds - Hatched in May

No. leaving Age in weeks nest 0-1 1-2 2-3 3-4 4-'? 5-6 6-7 7-8 8 ulus weeks 12* 13 ? 10 9 7 8 7 8 1 2 12 11 9 13 7 3 6 7 _ 1 0 11 10 8 13 5 5 6 9 9 9 12 9 8 5 13 11 8 4 9 6 9 6 11 8

1 2

1 0

1 3

1 3

1 L

Average brood sizes 11.3 11.2 9.2 9 9.4 7 6.5 6.5 6

^Figures in horizontal lines are hot to be construed as representing

the same broods. The counts are from random observations of separate

broods. 174

T a b l e X (Continued)

On farm ponds - Hatched in June

No. leaving .Age in weeks nest 0-1 1-2 2-3 3-4 _ 4-5 5-6 6-7 7-8 8 olus 10 8 8 7 5 7 8 4 5 8 11 8 6 3 8 6 10 8 8 6 4 2 6 7 5 5 8 5 6 Average brood sizes 9.3 8 7.8 7.5 6.4 6.5 6.1 6.2 5

Large lake habitat - Hatched in May

No. leaving Age in weeks nest 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8 nlus 11 9 8 8 7 9 8 6 7 12 5 8 9 4 6 4 4 6 13 10 9 7 3 7 7 3 4 11 8 11 8 10 4 3 9 12 12 7 6 10 8 6 9 6 7 L . 5 Average brood sizes11.6 8.8 3.2 7.5 7.1 6.8 6.5 5.2 5.6

Large lake habitat - Hatched in June

No. leaving Age in weeks nest 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8 nlus 10 10 A v e r a g e brood sizes 10 175

T a b l e X (Continued)

Small stream habitat - Hatched in May-

No. leaving Age in weeks nest 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8 nlus 3 12^,13 11 8 6 5 6 5 s' 10 6 11 5 6 4 9 7 7 9 4 9 5 12 5 11

L 10 6

_ _ . 4 Average brood sizes 10 9 13 10 7-7 3.5 4 3 7 7

Small :stream habitat - Hatched ini June

No leaving A^e in weeks nest 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8 nlus 6 A v e r a g e b r o o d s i z e 6 Table XI

Brood Counts for May and June with U-Tests for Differences in Large River Habitat

Month Hatched

May June May June May June May June May June May June May June May June May June May June Leaving 0-1 1-2 2-3 3-4 4-5 5—6 6-7 7-8 more Nest week weeks weeks weeks weeks •weeks weeks weeks than 8 13 5 5 3 9 7 10 5 3 6 7 8 9 8 4 6 7 10 7 15 6 11 6 5 3 5 8 3 8 2 9 3 6 9 14 7 7 2 6 5 8 4 5 5 9 3 7 7 4 12 15 7 2 6 6 3 8 9 4 4 3 8 5 4 13 2 7 10 3 8 4 7 4 3 6 7 6 12 7 u 6 6 4 2 8 10 7 4 8 13 15 4 7 10 9 9 7 7 15 6 9 9 9 8 9 13 6 8 9 4 4 5 4 11 9 4 1 4 9 4 2 6 4 2 3 1 16.5 16.5 3.5 2.5 62.5 21.5 54-.0 43.0 35.5 4.5 525 7 .5 25.510.5 10 0 n 7 0 4 0 11 3 6 1 7 12 17 6 8 5 10 6 9 4 10 0 (4.-5 weeks) m n - o 51 o 14.28 x - .210 .500 .571 .050* .417 .015* .008* .165 Table XII

Brood Cotints for Small Stream (SS) and Large River (LR) Habitats with U-Tests for Differences

Hatched in May

SS LR SS LR SS LR SS LR SS LR SS LR SS LR SS LR SS LR Leaving nest 0-1 week 1-2 weeks 2-3 weeks 3-4 weeks A-5 weeks 5-6 weeks 6-7 weeks 7—8 weeks 8 13 12 5/ 13 5 11 9 8 10 6 3 5 7 6 9 5 4 13 10 8 7 15 10 11 6 6 11 3 5 8 6 8 4 9 9 9 7 14 7 7 2 9 6 8 4 5 9 9 5 7 12 15 7 12 2 6 3 5 9 4 11 8 13 2 7 10 8 4 7 3 10 7 12 7 14 6 4 6 8 10 4 13 15 4 10 4 9 9 7 15 9 9 9 8 13 8 9 4 5 11 9 1 9 2 6 4 2 3 1

U 6.0 15.0 5.5 6.5 7.5 3.5 16.0 8.0 13.0 8.0 26.0 8.0 2.5 53.5 10.0 20.0 25.5 19.5 m , n 3 7 3 4 1 11 4 6 3 7 2 17 7 8 3 10 5 9 P .192 .571 .417 .238 .333 .146 .001** .234 .399

<1 Table XIII

Brood Counts for Farm Ponds (FP) and Large Lake (LL) Habitats with U-Tests for Differences

Hatched in May

FP T.T, FP LL FP LL FP LL FP LL FP LL FP LL FP LL FP LL Leaving nest 0-1 week 1-2 weeks 2-3 weeks 3-4 weeks 4-5 weeks 5-6 weeks 6-7 weeks 7-8 weeks 12 11 13 9 5 8 10 8 9 7 7 9 8 8 7 6 8 7 12 12 12 5 11 8 9 9 13 4 7 6 5 4 6 4 7 6 10 13 11 10 10 9 8 7 13 8 5 7 7 5 3 6 4 9 11 9 8 9 11 8 12 10 9 4 8 8 9 5 13 12 12 11 7 6 8 10 8 6 4 9 9 6 7 6 4 5 9 6 11 8 12 10

1 3

1 3 14

U 23.5 36.5 19.5 4-520.5 9.5 15.5 2.5 33.0 15.0 10.5 9.5 6.5 5.5 10.5 5.5 8.5 6.5 m , n 13 5 4 6 5 6 3 6 8 6 4 5 2 6 4 4 5 3

P .387 .086 .214 .083 .141 .548 .571 .343 .500

t» Table XIV

Brood Counts for Large Lake (LL) and Small Stream (SS Habitat with U-Tests for Differences

Hatched In May

LL SS LL SS LL SS LL SS LL SS LL SS LL SS LL SS LL SS T.aaving Nest 0— 1 week 1—2 weeks 2-3 weeks 3—L weeks A—5 weeks 5—6 weeks 6-7 weeks 7—8 weeks 11 8 9 12 8 13 8 11 7 8 9 6 8 5 6 6 7 5 12 13 5 8 8 9 10 4 6 6 11 4 5 4 6 6 4 13 9 10 7 9 7 7 8 9 7 7 4 3 9 4 5 11 8 11 8 12 10 4 8 5 9 11 12 12 7 6 10 8 6 4 10 9 6 7 4 5 6 4

U 10.5 4.5 10.0 8.0 0.0 6.0 4.0 26.0 8.5 9.5. 3.5 6.5 33.5 8.5 3.5 8.5 6.5 8.5 m , n 5 36361 6463 52674335

P .286 .452 .143 .057* .548 . 429 .051* .314 . 500 Table XV

Brood Counts for Farm Ponds (FP) and Small Stream (SS Habitats with U-Tests for Differences

Hatched in May

FP SS FP ss FP SS FP SS FP SS FP SS FP SS FP SS FP SS Leaving nest 0-1 week 1-2 weeks 2-3 weeks 3-4 weeks 4-5 weeks 5-6 weeks 6-7 weeks 7-8 weeks

12 t 13 12 5 13 10 11 9 8 7 6 8 5 7 6 8 5 12 13 12 8 11 9 10 13 6 7 11 5 5 6 6 7 4 10 9 11 7 10 8 7 13 9 5 4 5 9 6 5 9 9 9 12 12 9 5 8 5 11 13 11 8 4 4 10 9 6 6 9 6 4 11 8 12 10 13 13 14

27.0 12.0 9.5 2.5 0 5 3.5 8.5 15.5 8.5 3 5 11.5 2.5 5.0 7.0 11.5 13.5

13 3 4 3 5 1 3 4 8 3 4 2 2 7 4 3 5 5

.182 .200 .167 .314 .315 .400 .167 .429 .500

co o Table XVI

Brood Counts for Farm Ponds (FP) and Large River (LR) Habitats with U-Tests for Differences

Hatched in May

FP LR FP LR FP LR FP LR FP LR FP LR FP LR FP LR FP LR T.flavincr n a s t 0-1 weak 1 - 2 weeks 2 - 3 weeks 3 - 4 weeks 4 - 5 weeks 5 - 6 weeks 6 - 7 weeks 7 - 8 weeks 12 13 13 5 5 5 10 9 9 10 7 3 8 7 7 9 8 4 12 10 12 7 11 15 9 11 13 6 7 3 5 8 6 8 7 9 10 9 11 14 10 7 8 2 13 6 5 8 5 5 9 6 7 9 12 9 15 9 7 2 12 6 9 3 9 8 4 5 8 13 13 11 2 7 8 10 8 7 3 4 7 9 12 7 14 6 6 4 8 10 4 9 13 15 6 4 10 9 9 7 11 15 8 9 9 9 8 12 13 8 9 4 10 5 11 9 13 1 9 13 2 14 6 4 2 3 1

U 39.0 52.0 8.0 8.031.0 24.0 10.5 7.5 42.0 14.0 43.0 25.0 4-5 11.5 8.5 31.5 19.0 26.0 m , n 13 7 4 4 5 11 3 6 8 7 4 17 2 8 4 10 5 9

U - 5 weeks) j u u = ^ o = „ ^ x = __go6

P .321 .557 . 371 .452 .060 . 209 . 267 .071 .350 Table XVII

Brood Counts for Large Lakes (LL) and Large River (LR) Habitat with U-Tests for Differences

Hatched in May

LL LR LL LR LL LR LL LR LL LR LL LR LL LR LL LR LL LR Leaving nest 0-1 week 1-2 weeks 2-3 weeks 3-4 weeks 4-5 weeks 5-6 weeks 6-7 weeks 7-8 weeks 11 13 9 5 8 5 8 9 7 10 9 3 8 7 6 9 7 4 12 10 5 7 8 15 9 11 4 6 6 3 4 8 4 8 6 9 13 9 10 14 9 7 7 2 8 6 7 8 7 5 3 9 4 7 11 12 8 15 11 7 8 2 10 6 4 3 8 9 9 4 8 12 13 12 7 2 6 7 10 10 8 8 6 7 3 7 12 9 6 7 7 14 4 6 4 5 8 10 4 13 15 4 10 9 9 7 15 9 9 9 8 13 8 9 4 5 11 9 9 2 6 4 2 3 1

U 15.5 19.5 10.5 13.537.5 28.5 16.5 19.5 23.0 19.0 53.0 32.0 11.5 36.5 10 30.0 9.0 18.0 5 76^-611 6667 5 T7684. 10 39 (£-5 weeks) = 42.5 0 = 12,7639 x = -.823

P .433 .457 .366 .469 .418 .206 .071 .094 .241 smaller brood sizes of the 5-6 weeks age group on small streams as com­

pared with large lakes and in smaller broods in the 2-3 weeks age group

on large lakes compared with small streams. The conflicting results

probably derive from the vagaries of sampling. Larger samples are need­

ed. Significantly smaller broods were observed among the June-hatched

birds in 3-4, 5-6, and 6-7 weeks age categories. A larger sample might

show smaller brood sizes in all June-hatched broods.

Working on 32 small man-made farm marshes in New York, Klein (1955:

17) found an increased rate of successful utilization of marshes with more cover. Only 2 out of 11 (18 per cent) of the marshes with 15 per cent or less of their area in cover produced ducks to maturity; 16 of

21 (76 per cent) of the marshes with 25 per cent or more of cover pro­ duced ducks. This difference was found to be statistically significant.

If there are creek habitats nearby, broods hatched on farm ponds with little cover frequently do not remain on the ponds for more than a short time. If the broods are not disturbed, they can frequently be held at the ponds by increasing hiding cover and by artificial feeding.

Boxes erected on farm ponds have a good chance of being used, and with the 10,257 farm ponds in Ohio (Handley: undated), there is an abundance of potential wood duck habitat. Integrated management of wood ducks and beavers may have important possibilities in some regions (Beard, 1953).

Brooks (1948:176) reported increasing use by wood ducks of beaver dams in West Virginia. S u m m a ry

A higher rate of usage has been found among boxes placed over water than over land. Wood ducks will nest one to two miles from water, and utilization of boxes in upland woods has been found equal to that in bottomland woods. Of 501 boxes erected in four habitats, lake, pond, river, and small stream, those along small streams were least used.

Little difference in survival of broods was found in different habitats. Slightly smaller brood sizes in one area seemed to be corre­ lated with an abundance of minks. The rate of successful utilization of ponds and marshes increases with increased cover.

184. CHAPTER XI

DECIMATING AND LIMITING FACTORS

Insects

Honey bees, bumble bees, wasps building cellular type paper nests,

and wasps building mud nests frequently use wood duck nesting boxes.

If the boxes are repaired in the spring, swarms of honey bees can

be removed from all of the boxes at that time; new swarms will not appear

until after the main part of the duck nesting season. Wasps pose a

hazard in checking nesting boxes, but McLaughlin and Grice (1952:254-)

found that wasps did not materially interfere with use of the boxes by

wood ducks. Klein (1955:74-) found no evidence that wasps interfere with wood duck nesting, and he quoted personal correspondence from F.C.

Bellrose reporting a similar conclusion.

Fish

Taverner and Swales (1907:52) considered the pike one of the worst

enemies of ducklings at Point Pelee, Ontario. Lagler (1956:120) reported

examining 1,213 stomachs of pike taken during the waterfowl brood season in Michigan, and ducklings were found in only three of these stomachs.

Raney (194-2:50) reported results of examination of six alligator gar taken in Texas, and four of these gar held large birds in their stomachs. Raney reported an observation of a gar swallowing a water turkey perched several inches above the water. Gar were seen taking live decoys used for duck hunting, and gar took several ducks before the

185 186

ducks could be retrieved after being shot.

Mumford (1952:4-0) stated that the largeraouth black bass is a known

destroyer of ducklings.

Bullfrog

On June 4j 1956, at one of the small ponds at the Olentangy Station,

five day-old ducklings were toe-clipped and released onto the water of

a pond. The five ducklings swam into a clump of cat-tails and remained hidden several minute's,*’" after which four swam out of the clump of cat­

tails. I immediately captured the four remaining ducklings and attach­

ed bands and large fish hooks to the legs of two which were returned to the pond near the spot where the one had disappeared. Within ten minutes bullfrogs had attacked both of the ducklings. One duckling was half swallowed, head first. The frog was not hooked with the fish hook, but it failed to disgorge the duckling before it was captured.

The duckling removed from the frog's mouth survived the ordeal. For an unknown reason, the second frog disgorged the other duckling after having it in his mouth. For a frog to capture a free duckling is a different matter than to capture one attached to a string and fish hook.

Some splashing in the water was observed before the frogs captured the ducklings; since the frogs captured the ducklings by the heads, it would appear that they could seldom capture a duckling free to escape from them as they were about to attack. However, Wayne Knisley, Delaware

County Game Protector, reported (personal conversation) that he saw a bullfrog capture a mallard duckling not handicapped by being tethered.

McAtee (1921) also reported a case of a large bullfrog capturing and eat- 137

ing a duckling.

Howard (1950:152) reported the capture of a brown towhee by a

bullfrog. The frog sprang at the bird from behind a lily pad, caught

the towhee by a foot, and submerged, thus drowning the bird.

Snapping Turtle

No direct evidence of predation by snapping turtles on ducklings

was obtained in central Ohio. Milton B. Trautman examined the stomachs

of two turtles taken during the 1956 brood season from wood duck habitat

near the Scioto River, and neither turtle contained duckling remains.

At Buckeye Lake, 15-20 large turtles are taken each year from the bay where many of the wood duck broods normally gather (conversation with local residents), and it is probably that some ducklings are lost to turtles.

At Buckeye Lake, two immature ducks with parts of feet missing were taken in 1956. The third toe of one foot of each of these two ducklings was cut squarely off to the first joint, and the remaining toes extended to the same length. The manner in which the different toes were removed to the same length suggested that the ducks had been caught between broad jaws, such as those of a turtle, but the escape alive from the jaws of a turtle seems improbable. Just how these toes were lost must remain a iqystery.

Taverner and Swales (1907:52) stated that the snapping turtle is one of the worst enemies of ducklings at Point Pelee, Ontario. Coulter

(1957) examined the digestive tracts of 157 snapping turtles taken in eastern Maine during the waterfowl brood season; forty-two specimens 188

(27 per cent) of the turtles contained evidence of a minimum of 52

birds of which 25 were ducks.

The capture of an adult hen ring-necked pheasant by a snapping

turtle was recorded by Edminister (1953). The turtle had caught the

pheasant by the head, presumably as the bird had lowered its head to

drink. Laskey (194-7) reported snapping turtles apparently preying on

yellow-billed cuckoos, grackles, blue jays, robins, and a mourning dove.

SnakBS

In an Illinois study during 1939-1945, Bellrose (1955b: 11) found that the bull snake was responsible for 10 per cent of the losses due to nest predation. He stated, "Although the ducklings are in the nest cavity only about 24 hours, a surprisingly large number of broods are found and eaten by bull snakes during that short interval of time."

The bull snake does not occur in Ohio, and no nesting loss attribut­ able to snakes of other species was noted.

Questionnaire replies from the various state game commissions in­ dicated that the gray rat snake is the number one nest predator in

Alabama, Florida, Georgia, and Louisiana, This snake is notorious throughout the South for visiting chicken houses and eating the eggs in the nests (Schmidt and Davis, 1941:151).

Gigstead (1938:607) reported observing a common water snake pur­ suing young wood ducksj the snake was killed, and capture of the duck­ lings by the snake was not observed. The Hooded Merganser as a Nest Competitor

This species utilizes wood duck nesting boxes. A single nest was found in the 977 nest boxes I checked in Ohio. Mixed clutches of hooded merganser and wood duck eggs have been reported by Bent (1923:

25, 162) and other authors. Bent (1923:25) quoted George A. Boardman as having witnessed dual occupancy of a nest by a hooded merganser and a wood duck. The two species were reported to have taken turns incu­ bating. Potter and Murray (1949:231) reported a joint clutch of a hooded merganser and a wood duck which was incubated by the merganser.

The mixed brood left the nest safely. More often the mixed clutches are incubated by wood ducks.

Nest parasitism is fairly common among waterfowl. Hochbaum (1944:

48) reported that redheads, ruddy ducks, and lesser scaup ducks fre­ quently drop their eggs in nests of the canvasback. The redhead is a common nest parasite.

Starlings as Nest Competitors

In Ohio, the starling is an important competitor of the wood duck for a nesting place; of the 977 boxes checked, 125 were being used by starlings. Most wood ducks do not start nesting as early as the star­ ling, and if the starling occupies a cavity before the wood duck, the cavity is unavailable to the duck. During the time the wood duck is egg-laying, its nest is likewise unsafe from starlings. The wood duck enters the cavity and lays its egg in the early morning, and through the remainder of the day the starling sometimes takes over the cavity and covers the duck eggs with nesting material. Mien the duck returns to

189 190

its nest and finds the cavity containing a starling*s nest, the nest is

deserted by the duck.

Starlings have been observed to fly at wood ducks which were in­

specting cavities containing occupied starling nests, and the starlings

seemingly prevented the ducks from entering the cavities. After the wood duck has commenced incubation, it appears to be safe from inter­

ference from the starling. Starlings were several times seen to alight on nesting boxes containing incubating wood ducks, but they no more than touched the box before they flew. Incubating female wood ducks gave a hiss in response to my scratching on the outside of the box, this hiss probably prevents starlings from entering nesting boxes occupied by wood ducks.

By several times removing the nest from a cavity the starling can finally be discouraged from use of a box. It is possible to erect enough boxes for both the wood ducks and starlings, and this may be cheaper than the persistent removal of starling nests needed to keep boxes available for wood ducks.

In Vermont, Miller (1952:5) found that starlings did not inter­ fere with nesting box utilization by wood ducks, as the use by ducks usually precedes the use by starlings. When conflict was observed, the wood duck won.

Miscellaneous Avain Nesting Box Competitors

During the three years of this study, nesting boxes were checked to the extent of 977 box-years, a box-year being one box available for a nesting season. A variety of birds were found using the boxes, but many 191 of these were unimportant competitors.

No flicker nests were found in the 977 wood duck nesting boxes checked in Ohio in 1955-1957. On April 27, 1956, I observed a flicker at a box containing a wood duck nest. The flicker did not enter the box, although the duck was not on the nest. Leopold (1951:210) noted that flickers broke some eggs in a wood duck nest. Bellrose (1955b:8) stated that flickers and red-headed woodpeckers may puncture the eggs of wood ducks.

One nest each was found of tufted titmouse, white-breased nuthatch, and English sparrow in the 977 boxes checked; two nests each of crested flycatchers and sparrow hawks were found. Four nests were found each of birds and grackles. Ten nests of tree swallows were found, all at

Magee Marsh except one at Knox Lake. Screech owls nested in nine boxes and roosted in nine additional boxes.

In any large-scale nesting box erection program, the use of nest­ ing boxes by these miscellaneous species is a relatively minor problem.

The possible exception is the tree swallow. At Magee Marsh in 1957,

9 of 116 boxes contained nests of tree swallows (Karl Bednarik, personal correspondence); however, there were enough boxes in this area to meet the needs of both species.

Miller (1952:5) found that tree swallows did not interfere with use of boxes by wood ducks, as use by ducks preceded use by tree swallows.

Where conflict occurred, the ducks won. McLaughlin and Grice (1952:254-) noted that wood ducks will kill or drive out tree swallows when active competition occurs. 192

Red-winged blackbirds have been reported (Jopson, 1956) using

nesting boxes, but none was found in the 977 boxes checked in Ohio.

Avian Predators

Morris (1892) reported observing a sharp-shinned hawk striking

a wood duck and fastening itself to the back of its prey, whereupon

both were shot. Rapp (194-1) reported flushing a red-shouldered hawk from the ground where it had been eating an adult wood duck. It was thought that the duck had been dead an hour or two, and the cause of the duck's death was not known. Saunders (1937) recorded observations

in which a red-shouldered hawk first attacked a brood of ducklings and then the female duck, but the hawk was unsuccessful in all attacks.

Avian predators probably exact a very small toll from the wood duck population.

Opossum

Of 820 nests dest"oyed by predators in the Illinois River Valley during 1939-194-5, Bellrose (1955b:4-7) attributed the destruction of 15 or 1.8 per cent to the opossum. Of 297 tree boxes checked in Ohio in

1957, two contained opossums. One contained a litter of young.

Mice

White-footed mice were frequently found in wood duck nesting boxes. One of these mice was seen to enter a wood duck nest when the female was off feeding; no injury was done to the nest. The egg membranes in boxes mounted on trees are regularly eaten by mice after the ducklings have left the nest. Raccoon

Bellrose (1955b:10) considered the raccoon the chief wood duck

nest predator in Illinois. Of 820 nests destroyed by predators in

Illinois in 1939-194-5, 304- or 37.1 per cent were destroyed by raccoons,

304 of 1,579 nests (19.2 per cent) were destroyed by this animal.

Actually more were destroyed by fox squirrels, but because the raccoon

destroyed the nests in a more advanced stage of incubation and killed

about one-third of the incubating females in the nests they raided,

Bellrose considered the raccoon the most destructive nest predator.

McLaughlin and Grice (1952:247) reported 26 of 30 (86.6 per cent)

wood duck nests destroyed by raccoons. Klein (1955:75) reported 6 of 30

(20 per cent) destroyed by raccoons. Questionnaire replies from the various state game commissions indicated that the raccoon, in most

areas, holds first place as a wood duck nest predator.

The raccoon poses a serious problem in wood duck management because

of nest predation and denning habits. In the Lake Erie marshes of

Ohio where there are few natural denning places, raccoons use many of the available wood duck nesting boxes, thus making them unavailable to wood ducks. At Magee Marsh, in Ottawa County, Ohio, Karl Bednarik

(data in files of Ohio Division of Wildlife) found 70 of 185 boxes occupied by raccoons in 1956. In Ottawa County, Ohio, Laurel Van Camp

(Nolan, 1956) examined 151 wood duck nesting boxes and found raccoons in

31. Van Camp (personal conversation) observed an increased use of nest­ ing boxes by raccoons at the season when the females had young; Brown and Bellrose (1943:303) observed a decreased use at this time. McLaugh­ lin and Grice (1952:254) found that raccoons seldom used the boxes for dens but merely visited them in search of eggs, 193 194

Bellrose (1955b:10) observed that nesting wood ducks in Illinois

developed a fear of raccoons, and the incubating birds in uplands left

nests readily on the approach of an observer after raccoons had been

driven to the uplands by flooded bottomlands.

Individual variation, which was not suspected of being associated with raccoon disturbance, was observed in the flushing of incubating female wood ducks in central Ohio. In two cases of two wood duck nests within 50 feet of each other, and with all four boxes equally accessible to raccoons, one female from each pair of nests flushed as I approached, while the other one of each pair remained on the nest until forcibly removed.

"Wood ducks, at least during the fall, do not ordinarily exhibit fear of raccoons. At Buckeye Lake, Ohio, on August 16, 1956 when some

50 wood ducks were feeding around a baiting station, a raccoon was seen to come out of the brush and to feed along with the wood ducks. On the approach of the raccoon the wood ducks moved back a few feet, but they soon returned and fed within A-5 feet of the raccoon.

Because the raccoon is such an important egg predator, considera­ tion must be given to the problem in any wood duck nesting box manage­ ment program. Perhaps the best solution would be the increased human use of raccoon fur. A market demand which would restore the price of pelts to the $12.00 to $15.00 of 35 years ago should do much toward re­ ducing the currently excessive raccoon population in much of the country.

In lieu of that hope becoming a reality, the best that can be done is to attempt to prevent these animals from entering wood duck nesting boxes. 195

Several methods of raccoon-proofing nesting boxes are discussed

in Chapter XII.

Squirrels

In Ohio, one duck egg was removed from each of two nests, but the

nests were otherwise successful. The egg removal was thought to have been

the result of red squirrel depredations, but definite proof was lacking.

Pieces of egg shell were found on the ground beneath the nesting boxes,

and some of the contents of reasonably fresh eggs were found on the

top and sides of boxes.

In studies on the Illinois River, Bellrose (1955b:10) found that fox squirrels destroyed more wood duck nests than any other one animal; of 820 nests destroyed by predators during 1939-1945* 419 or 51.1 per

cent were destroyed by this squirrel.

Bent (1923:160) reported a case of two gray squirrels living in a cavity only a few inches from a successful wood duck nest. M.B. Trautman

(personal conversation) told of a similar incident in Ohio. Leopold

(1951:210) noted that a considerable number of gray squirrels occupied the lawn near wood duck nests; the squirrels never destroyed duck eggs.

Klein (1955:75) also found gray squirrels innocent of damage to du'ck nests.

Competition for nesting boxes was the chief problem posed by squirrels in Ohio. Of 297 tree-mounted nesting boxes checked in 1956,

20 contained squirrels or their nests. Tdhen a nesting box contains an unoccupied squirrel nest it is somewhat less acceptable to ducks. Appar­ ently the fox squirrel destroys wood duck eggs but the gray squirrel does not. 196

Wood ducks appear to have no fear of squirrels. Breckenridge

(1947) described an encounter of a pair of wood ducks with a red squirrel.

The red squirrel rushed toward a duck perched in a tree but stopped short

a few feet from the duck. After this was repeated several times without

disturbing the duck, the squirrel ran along the bottom of the limb to

the duck, but when the squirrel's head appeared around the edge of the

limb, the squirrel was met by a peck from the duck.

Mink

It is difficult to obtain direct evidence of the relative importance

of the different predators responsible for brood decimation. Because

of nocturnal hunting activities the mink is not likely to be observed

capturing ducklings, and because of the fur value of the mink, it is

not generally feasible to destroy them through the duckling brood

season merely to examine their stomachs.

The mink is undoubtedly well able to capture ducklings, and it can

be expected that the importance of the mink as a duckling predator is

directly proportional to the local abundance of minks. This mammal is

probably more common in some areas than is generally suspected. A veteran mink trapper, Roy R. Walisa, who regularly works the Scioto

River from the Franklin-Pickaway County line to Waverly, a distance of

about 60 miles, annually takes 65-75 of these animals. Three other

trappers known to operate in parts of the same area each take J + ~ 5 mink per year. There are undoubtedly additional mink caught by trappers operating shorter trap lines. Hunting

Hunting is probably the most effective decimating factor of the wood duck after it is able to fly. A steady decline in the bag in the

Mississippi Valley between Wabasha, Minnesota, and Alton, Illinois, in five years of hunting was reported by Green (1951); in 194-6 wood ducks made up 14-,0 per cent of all ducks bagged, 12.0 per cent in 194-7, 7.3 per cent in 194-8, 3.3 per cent in 194-9, and 5.0 per cent in 1950. At the

Olentangy Station, the number of breeding pairs in nest boxes increased from 11 in 1956 following an open season to 17 in 1957 following a closed season.

The flesh of the wood duck is considered rather tasty by persons not relishing the so-called gamy flavor. Wilson (1854-s600) stated that as food, the flesh of the wood duck is inferior to that of the blue­ winged teal. Conversation with hunters indicated that the wood duck is generally considered inferior only to the canvasback. Because of the small size of the wood duck, half that of the mallard, hunters some­ times save their legal limit for the larger mallard.

Shooting the birds as they come to their feeding places or on their flight lanes is seemingly the most common method of hunting wood ducks in the South. These ducks can be readily shot from blinds placed at their feeding or loafing places. Jump shooting is a common method of hunting this species in the North. Wood ducks do not come to decoys as readily as do other species of ducks (Phillips, 1925:68).

197 Muskrat Traps

Included among the banding data are reports of wood ducks taken

in furbearing animal traps. Most of these were muskrat traps. Reports

came chiefly from New York and Ontario.

Loss of Nest-huntine Females in Chimneys

In the spring of 194-9, the C.R. Grabill family in Westerville, Ohio, heard a noise issuing from an unused chimney in the kitchen of their home. Thinking that a squirrel had fallen down the chimney and was un­ able to climb up its walls, they opened the bottom of the chimney, whereupon a female wood duck flew into the room. On April 25, 19-51,

I observed a pair of wood ducks on top of a chimney in Westerville,

Ohio, about a half mile from the Grabill home, but the female of this pair did not go down into the chimney, and the pair flew away as I watched.

Pettingill (1935) reported finding a dead wood duck in a chimney in the vicinity of Philadelphia, Pennsylvania. M.B. Trautman (personal conversation) advised me that he received reports of the finding of dead birds of this species in chimneys at Fut-in-Bay, Ohio, Hunt (1956:

11) reported having had so much trouble with ducks entering chimneys and dying that chimneys were screened in some areas in Michigan. Leopold

(1951:212) reported observing a pair repeatedly sitting on a chimney peer­ ing down the flue. He also reported the finding by his neighbor of a dead female above the damper in a fireplace flue. Sutton (1928:71) told of a female repeatedly entering a chimney in Hartstown, Pennsylvania, during the week of May 14., 1925. Phillips (1925:59) heard of several

198 199

instances of wood ducks entering chimneys and knew of one himself.

Mayfield (194-8:208) reported a nest eight feet from the top in a cottage

chimney near Quincy, Illinois,

A report of a wood duck in central Ohio successfully using a chim­ ney for nesting was received in 1957 from Edward S. Thomas. In mid-

June, Alma Raduege of Columbus, Ohio, observed some ducklings on the roof of her house. Thinking that the ducklings could only have come out of the chimney, she sent a boy up on the roof of the house to look down the chimney. The boy saw the nest and reported one duckling yet in it.

I examined the nest and chimney several days later. One dead duckling was in the nest. The hole of the chimney had been plugged 5*5 feet from its top, and this is where the nest of the wood duck laid. The opening of the chimney was about 8 inches square. Cement squeezed from between the bricks made rather rough sides on the inside of the chimney.

Eleven ducklings were reported to have come from the nest.

Among the recoveries of banded wood ducks in the files of the

U.S. Fish and Wildlife Service, .several were reported found dead in chimneys. The loss of wood ducks in chimneys can perhaps best be alle­ viated by providing ample nesting facilities.

Frost-cracking of Eggs

On April 8, 1956, an egg which had been frost-cracked was found in a nesting box at Buckeye Lake (Figure 43). This egg had not been covered with the loose saw dust in the nesting box, and it was presumably a dropped egg. If the egg had been covered, as is usual with nests to be incubated, it would, of course, have withstood a lower temperature with- Figure 43 200 Frost-cracked Egg 201 out being cracked. It may be that frost-cracking affects only dump nests, most of which would not produce young anyway. Sowls (1955:89) found no loss from frost-cracking of eggs of ground-nesting ducks.

Nest Desertion

Nests are sometimes found deserted for no apparent cause. The nest may be deserted after the eggs have been incubated several days, and the eggs are sometimes left neatly covered as though the female might have been killed when she left the nest to feed. One such nest observed in central Ohio was a second nesting attempt in late June, and it may have been deserted because the female had lost the incubation urge.

One incubated nest was found deserted in early May. McLaughlin and Grice (1952:251) found that friction occurs among the wood ducks themselves, thus causing nest abandonment. These workers reported that nest desertion increases with increased density of ducks. This friction occurs even when there are adequate nesting places. In a recent questionnaire three state game commissions, those of Connecticut, Cali­ fornia, and New York, listed nest desertion as the chief cause of nesting failure.

Loss of Ducklings in Water-current

On May 12, 1956, a brood of 15 ducklings, less than a week old, was observed near the shore of the Scioto River. As the brood was approached from the shore the ducklings quickly swam toward the center of the river, and the female flew up the river and disappeared from view. The current of the river was moving about 200 feet per minute, and the ducklings were 202

soon carried 3,000 feet downstream. They then swam toward shore and

seemed to start feeding. They continued feeding through more than an

hour as I watched. The female failed to return to her brood in that

time.

The extent broods become permamently separated from female ducks

is not known, but such separations, coupled with the habit of lost duck­

lings following any moving object, probably accounts for much of the

brood-mixing which can be observed. Loss of ducklings through separa­

tion of broods by the current of the river is thought to be a negligible

factor in wood duck decimation.

In June, 1957, Brice Metzger, Ohio Division of "Wildlife, observed

an entire brood of ducklings drawn by the current into the pen stock on a small pond at the Olentangy Station. All of these ducklings were lost.

Filamentous Algae

Since wood duck broods normally spend much time near or among emergent vegetation, they are also closely associated with alga beds which commonly grow in relatively shallow water. The emergent vegeta­ tion gives the ducklings a desirable combination of food and protective cover, but the filamentous algae also present a potential death trap

(Figure 44).

The algae do not appear to impose a real threat to ducklings unless the area is exposed to human or other molestation during the early brood period. When the ducklings become frightened on the water at some dis­ tance from emergent cover, they frequently dive and swim beneath the Figure 44-

Ponds Containing Abundant Surface Vegetation Are Potential Death 203 Traps Because of Filamentous Algae Beneath the Surface 204 water surface. Under such circumstances, the submerged ducklings some­ times become entangled among the filamentous algae and drown.

In 1955, at the Olentangy Station, one complete brood was believed to have been lost in this manner. On the approach of an automobile to ponds in connection with duck brood inventories, ducklings were fre­ quently seen diving and becoming entangled in algae. I once waded into the water and rescued five hopelessly entangled ducklings before they were drowned. On six occasions ducklings were observed to dive near algae without reappearing on the surface. The ducklings were being pur­ sued by the observer on four of these occasions. The drowned ducklings were not actually found, but since they had disappeared in large alga beds, inability to find the ducklings was to be expected, and entangle­ ment resulting in drowning was the assumed explanation for their failure to reappear on the surface.

Oedogonium and Spirogyra were the algae genera identified on the three ponds concerned. A heavy loss of diving ducks in this manner was observed in British Columbia by Munro (1941)•

Wood duck broods usually remained on their native ponds through their first two weeks out of the nest, and this is the age at which the ducklings are most vulnerable to alga injury. It is necessary, conse­ quently, to consider the relationship of algae and ducklings at the native ponds of the ducklings.

While various algacides are available, control of algae on every pond where a brood of wood ducks may occur is not generally feasible.

The loss of ducklings to algae on isolated farm ponds can be minimized by preventing disturbance at the ponds by humans or dogs through the 205 ducklings* first two weeks out of the nest or through the months of May and June if the former time is not known, When the pond is approached, effort should be made, through a conspicuous approach, to frighten the ducklings into cover before being too near the pond. The ducklings dive only when closely pressed.

Parasites and Disease

0*Roke (1934) bas found that infection by blood protozoa, Leuco- cvtozoon anatis. causes a malaria-like disease and may cause 10 to 100 per cent mortality among domestic and wild ducklings. Some samples from

Michigan contained 100 per cent infected birds. Similar to Plasmodium, which causes human malaria, the genus Leucocvtozoon has an asexual cycle called schizogony, which occurs in the vertebrate host, and a sexual cycle called sporogony which occurs in an invertebrate host.

The black fly Simulium venustum was found to carry the sexual stage of this protozoan. This fly is very common locally in Michigan, and

0'Roke found as many as 75 blood-gorged black flies on a single duckling at one time. . In the Lake Erie marshes near Sandusky, Ohio, 0'Roke found no Leucocvtozoon infections in July, 1932, and he reported that the black fly was not found in the area.

Mielcarek (1954!232) reported finding Plasmodium relictum and

Haemoproteus nettionsis in the blood of a wood duck taken in the fall of

1951. O'Meara (1956:207) reported the results of examining 422 blood smears collected from wood ducks in Maine. Two hundred and sixty birds contained Leucocvtozoon infections. 92 contained Microfilaria infections, and 284 contained Haemoproteus infections. An 80 per cent Leucocvtozoon 206

infection was noted in ducks from the Penobscot River, which has a

moderate to strong current and rather clear water; less than a 4-0 per

cent infection was found in ducks from Corinna Stream in Maine, a

sluggish, heavily polluted stream.

During September, 1956, I collected blood smears from 96 wood

ducks on the Scioto River and at Buckeye Lake. The technique used in

preparing the slides was suggested by A1 Dabag (personal conversation),

Veterinary College, The Ohio State University. The blood samples were

collected by pricking the tarsal veins of the ducks. A light film of

blood was spread over a microscope slide and the film was air dried.

The slide was later immersed in absolute methanol for three minutes and

then transferred to a dilute solution of Giemsa blood stain where it was

left for 4-5 minutes. The stain was mixed in distilled water at the rate

of one drop of stock solution of Giemsa stain per one cubic centimeter

of water. After removal from the stain, the slide was washed in dis­

tilled water and air dried. Examination was then made with a micro­

scope equipped with a mechanical stage. No blood protozoa were found

in these 96 samples. It appears that blood protozoa normally impose no

problem in wood duck management in Ohio.

The loss of ducklings to parathyphoid infections is rather prevalent among captive ducklings. Levine and Graham (194-2) reported the death of

4-00 of 500 wood ducks hatched in an incubator in Illinois. The birds were placed in pens which had been sprayed with a liquid disinfectant, and new electric hover brooders were used. The floors were covered with about an inch of sand. The symptoms which appeared about one week after hatching consisted of loss of appetite, sluggishness, and drooping tails. 207

The ducklings usually died within one day after the first symptoms

appeared. At autopsy, the ducklings showed pasted vents and white

diarrhea. Salmonella typhimurium was cultured from the livers of two of these birds. So far as is known, paratyphoid is confined to captive

ducks. It might become a problem in high-density wild populations.

Bellrose (194-5) reported the death of at least 100 wood ducks from a local population of 2,000 to 3,000 which had been feeding at a large crib of moldy corn during September, 1943, in Illinois. The ducks were found to be afflicted with aspergillosus and the causative organism was

Aspergillus fumigatus. The ducks apparently shattered the spores in feeding and breathed the spore-laden air. Aspergillosus has also been observed in the mallard (Neff, 1955), snowy owl (Meade and Stoner, 1942), and song sparrow (Manwell, 1954).

Six wood ducks taken in the late summer and fall of 1956 were care­ fully examined in a search for ecto parasites. Four species of Mallo- phaga were found: slender duck louse, Anaticola crassicornis. (from 4 ducks), common chicken louse, Menopon gallinae. (from 3 ducks), goose louse, Trinoton anseris. (from 2 ducks), and a species of Menacanthus

(from 2 ducks).

Two unidentified tapeworms (Hvmenolepis sp.) were taken from the small intestine of immature birds captured in September. An unidenti­ fied fluke was also taken from the caecum of an immature bird.

An exceptionally small immature female, weighing less than one pound and taken on September 13, 1956, was found to be unusually heavily parasitized. One tapeworm and six lice were collected from this bird.

This may have been a younger bird with its smaller size unrelated to 208

heavier parasitism.

Weather

Few observations were made on the effect of weather on hatching

success and survival of wood ducks. Brief comments are made in the

sections on "Incubation" and "Care of Young."

Summary

Bees and wasps occupy wood duck nesting boxes and pose a minor

problem when the boxes are being checked, but they need not interfere with utilization of the boxes by ducks.

Fish sometimes prey on ducks, but little information is available

on the extent of damage imposed. Two bullfrogs were observed catching

tethered ducklings. Little direct evidence of predation by snapping

turtles was observed, but this animal is known to be an important duckling predator in some areas. Snakes are important nest predators

in some areas but caused no trouble in Ohio.

The starling is an important competitor for nesting places in Ohio.

These birds can be discouraged from use of a box by destruction of their nests several times, or ample boxes can be erected for both the ducks and the starlings. The hooded merganser, sparrow hawk, screech owl, flicker, crested flycatcher, tufted titmouse, white-breasted nuthatch, tree swallow, bluebird, and grackle are minor nesting cavity competitors.

No loss of wood ducks to avian predators was observed.

Among mammals, opossums and mice are minor nesting cavity competi­ tors. Raccoons pose a more serious problem, using the boxes for denning 209

and thus preventing use of the boxes by ducks, as well as eating the

eggs, and sometimes killing the incubating ducks. Squirrels inter­

fere with use of boxes by ducks by filling the boxes with their stick

nests. Fox squirrels are serious nest predators; gray squirrels

apparently do not destroy the nests of wood ducks. Minks may be

important predators of ducklings where present in relatively large

numbers and are more numerous in some areas than is generally believed.

Hunting by man is probably the most important decimating factor

in the case of ducks after the ability to fly has been attained. Shoot­

ing the ducks from blinds at their feeding places is a common method of hunting in the South; jump shooting is most common in the North.

A few wood ducks are caught in muskrat traps, chiefly in New

York and Ontario.

Nest-hunting females sometimes enter chimneys of human dwellings and are unable to escape. Wood ducks sometimes nest successfully in chimneys.

Frost-cracking of eggs in dump nests has been observed, but it probably does not ordinarily affect eggs in nests to be incubated be­ cause such eggs are left covered with loose nesting material.

Nests are sometimes found deserted for no apparent cause. It may be that desertion results from friction among nesting wood ducks.

River ducklings are sometimes carried considerable distances by the current. If these ducklings are not later found by the female duck, they probably join other broods.

"When closely pressed, very young ducklings dive and swim under water. Ducklings occupying a shallow pond containing filamentous algae 210

sometimes become entangled in the alga strands and drown. Ducklings in

such areas should not be closely pressed but permitted time to make a

leisurely escape.

Ducklings are susceptible to a malaria-like disease caused by a

blood protozoan, Leucocvtozoon anatis. The sexual stage of this proto­

zoan is harbored by the black fly, Simulum venusturn. Leucocytozoon

anatis is sometimes a special problem in Michigan but does not appear

to pose a problem in Ohio. Other blood protozoa have been reported from

Maine ducks. Paratyphoid infections are prevalent among wood ducks

reared in captivity; the incidence of paratyphoid in wild birds is not

known. Salmonella typhimurium is the causative organism. Aspergillosus

is sometimes a problem among wood ducks where an abundance of moldy feed exists. Aspergillus fumigatus is the causative organism.

Four species of Mallophaga were collected from the plumage of wood ducks. Two unidentified tapeworms and one fluke were collected from

the digestive tracts of these birds. PART III

MANAGEMENT OF THE MOOD DUCK

211 CHAPTER XII

Natural Nesting Places

The cutting of the forest and consequent destruction of natural

nesting places was undoubtedly a contributing factor in the near­

extinction of the wood duck in the latter part of the nineteenth cen­

tury. With the silvicultural concepts of "ripe" and "wolf" trees yet unheard of, the primeval forest doubtless contained an abundance of

trees with suitable nesting cavities.

Cavities in trees are the usual natural nesting places for wood

ducks; however, Wilson (1854:599) reported that the nest is sometimes

constructed of a few sticks laid in a fork of tree branches. Nests have also been found in a recess among overhanging willows on a creek bank (Roberts, 1936:247), beneath upturned roots of a fallen tree

(M.B. Trautman, personal conversation), beneath the lower of two fallen trees where one had fallen across another (M.B. Trautman, personal con­ versation), on the ground under a heap of grape vines (Morris, 1930:

421), in a fissure of rock on the Kentucky River (Audubon, 1843:274) > on hay in barns in California and Connecticut (Bent, 1923:161), and in a vacant hen house (Forbush, 1925:229).

In the normal nesting place, the bottom of the cavity may be from a few inches to 8 feet below the entrance (Mayfield, 1948:208). Gig- stead (1938:604) examined 58 occupied natural cavities, and 52 inches was the greatest depth found. The diameters of the cavities were fairly uniform and averaged 10 inches. Entrances averaged 6 inches in diameter,

212 213

but the entrance to the cavity may be as small as 4 inches by 2.5

inches (Stuewer, 1945:236) or 3.25 inches by 3.50 inches (Dixon,

1924:46).

Dreis (1951:60) climbed to 52 natural cavities in a grid sample in Iowa, and on the basis of earlier observations of nests in natural

cavities judged each cavity suitable or unsuitable for wood ducks.

Of the 52, 41 were deemed unsuitable; 7 were considered suitable but were not occupied; 4 were occupied by wood ducks. In Massachusetts,

McLaughlin and Grice (1952:242) examined 69 natural cavities: 53 were considered unsuitable; 7 were deemed suitable but were not occupied;

6 were used by wood ducks; usage could not be determined in 3.

Although not all the cavities deemed suitable for wood ducks were used by this species, there were relatively few unused, and thoss which appeared suitable may have been unavailable for reasons not apparent to the investigators, such as earlier use by raccoons. Lack of suit­ able nesting places is undoubtedly a limiting factor in wood duck pro­ duction in some areas, and artificial nesting boxes have a distinct place in wood duck management.

Artificial Nesting Places

The use of boxes as compared with natural cavities can be expected to vary with the number of available natural cavities. On the Scioto

River in 1956, 101 nesting boxes were erected in an 11-mile section of the river. Eight of the 21 broods in this transect were known to have come from nesting boxes; 13 were supposedly from natural cavities.

McLaughlin and Grice (1952:242) reported that 45 of 51 known nests in an 214

area in Massachusetts were in boxes; 6 were in natural cavities.

Frank (1948:134.) reported only 12 wood duck nests in a total of

274 boxes during 6 years before 1948, and he concluded that use of

artificial nesting boxes would not be a feasible practice in wood duck management. Other workers have since found that wood ducks readily utilize nesting boxes, and a high rate of use has been obtained in reasonably large tests. Miller (1952:3) reported utilization of 87.5 per cent of 81 boxes in Vermont; Lee (1956:7) reported utilization of

83 per cent of 71 boxes in New Hampshire. Nichols (1955:367) reported that 61 (92.3 per cent) of 64 boxes were used in Connecticut; McLaugh­ lin and Grice (1952:247) reported that 1,427 (44*6 per cent) of 3,203 boxes were used in Massachusetts; Bellrose (1955b:46) reported 1,579

(49.1 per cent) of 3,218 boxes were occupied by wood ducks in Illinois.

History of Nesting Box Development

Several hundred slab boxes were erected on the Chautaugua National

Wildlife Refuge in 1936 by the United States Biological Survey. These boxes had a floor 9 inches square and a cavity 24 inches deep. The

5-inch entrance was 10 to 16 inches above the floor. Slightly less than 12 per cent of these boxes were used by wood ducks in four years

(Bellrose, 1955b:18). Because the slab boxes were heavy and difficult to mount, because the rate of use by wood ducks was low, and because of a high rate of nesting failure caused by predation, these boxes were con­ sidered unacceptable.

In 1939, A.S. Hawkins and F.C. Bellrose erected 450 boxes of rough- cut lumber in Illinois, chiefly in the Illinois River Valley. These 215

boxes were constructed of 1-inch cypress lumber; they had a floor 10

inches square and a cavity 23 inches deep. The bottom of the 4-inch

entrance was 16 inches from the floor. The rate of usage of these boxes during 1939 to 194-5 was about four times the rate of usage of the slab

boxes, but in some area predation was high and nesting success poor.

Through the use of entrances of different sizes and shapes, Bellrose

(1955b:21-27) then attempted to make these wooden boxes raccoon-proof.

The following were tested: a circular hole 3.5 inches in diameter, a diamond-shaped opening 5 inches wide and 3 inches high, a rectangular opening 5 inches wide and 3 inches high, a diamond-shaped entrance 4- inches wide and 3 inches high, and an elliptical opening 4 inches wide and 3 inches high. Raccoons readily passed through all of these open­ ings except the 3 by 4--inch diamond-shaped entrance and the 3 by 4- inch elliptical entrance. The diamond-shaped entrance was found to pre­ vent the entrance of wood ducks also, but the elliptical opening was found suitable for wood ducks. Four-inch circular holes were first cut into the boxes, and the circular holes were then masked with elliptical openings cut from sheet metal.

As squirrels were found to be important predators of wood duck eggs in Illinois, an effort was started in 1941 to make squirrel-proof boxes

(Bellrose, 1955b:24-31). Various types of metal guards were unsuccess­ fully used, but a wooden box covered with sheet metal and provided with a sheet metal pyrimidal roof proved to be squirrel-proof. In 1950, an all metal box was developed. This was made from a cold-air pipe of 26- gauge metal, 12 inches in diameter and 24 inches in length. The roof consisted of a 15-inch inverted metal cone. An elliptical opening 3 216

inches high by 4 inches wide was used. A circular piece of wood fitted

into and nailed to the bottom or a circular piece of sheet metal

soldered to the bottom made the floor. The box was lined on the inside with one-half inch insulating fiber board in order to enable the duck­ lings to escape from the box. Because the fiber board lining was not

sufficiently durable, it was later replaced with a thin layer of auto­ mobile undercoating. The females had difficulty entering boxes fitted with a thin layer of undercoating, and a thicker layer (one-half inch) was successfully used. The undercoating was mixed with vermiculite to make it possible to build up the thicker layer. This box is considered to be almost predator-proof.

Further Comments on Raccoon-proofing Nesting Boxes

In spite of the fact that they appear much larger than wood ducks, small raccoons can pass through any hole large enough for a wood duck.

Stullpen and Kirkpatrick (1953s126) reported that a 9.5 pound raccoon can pass through a circular hole 3.5 inches in diameter or a rectangu­ lar hole 3 by 3.5 inches in size. Raccoons are smaller in the South than in the North, and some weigh much less than 9.5 pounds during the nesting season. It is impossible to prevent young southern raccoons from enter­ ing wood duck nesting boxes through use of a small entrance hole.

Raccoons have an unbelievable ability to climb smooth metal pipes.

On May 15* 1955, the Columbus (Ohio) Dispatch carried a photograph of a raccoon perched high on the crossbars of a television aerial. In spite of the fact that wood duck boxes at Magee Marsh are mounted on steel fence posts, the boxes are much used for denning by these animals. Figure 45 Figure U 5 ro This Raccoon Readily Climbed a Smooth Pipe Two Inches in Diameter 218

shows a caged raccoon climbing a 2-inch pipe.

In an effort to find an effective method for raccoon-proofing

nesting boxes on posts over water, cones were placed around the posts

beneath 25 boxes at Magee Marsh in 1957. These cones were made of a

thin gauge aluminum, and were 12 inches in diameter at the top, 3 inches

in diameter at the bottom, and 20 inches high. The water level of the marsh dropped later, and many of the boxes were surrounded by soft mud

instead of water. Raccoon tracks all of the way up some of the cones mounted over soft mud indicated that these mammals were able to climb over

the cones.

Loss of Nests to Experimental Method

In connection with ground-nesting ducks Hammond and Forward (1956) objected that terrestrial visits to nests do not give reliable index to nesting success under natural conditions because the observer1s trail guides predators to the nests. Miller (1952:2) also suspected that raccoons followed the trail of the observer as he visited the boxes.

It is difficult to determine the role of the observer's trail in guiding predators to nests. Of the seven nests I knew to have been disturbed by predators, only one was disturbed in the interval between examinations; six were destroyed before the first examination of the nest­ ing box. It can be positively stated that not more than one nest was destroyed by predators as a result of their following the trail of the observer in the present study. Test of Different Types of Nesting Boxes

An experiment was conducted to compare the merits of several different types of wood duck nesting boxes. Fifty boxes each of five

different types were constructed. Nine of these were lost in high water on the Scioto River. Six hole-in-lid ammunition case type boxes were added. The 247 boxes were erected at random in regard to box

types. Two hundred and forty-seven cards were first marked with the

type of boxes. The cards were thoroughly mixed by protracted shuffling, and they were then taken into the field along with an assortment of nesting boxes. After the place for erecting a box was chosen, the next

succeeding card of the stack was examined and the type of box to be used was thus determined. The type boxes used in this test were the con­ ventional or basic types, tunnel-entrance, elliptical entrance, grease drum, and hole-in-lid ammunition case. All of these boxes were mounted on trees about 15 feet from the ground.

Tables XVIII and XIX show the rate of utilization and success in these different types of boxes. Statistical analyses of these data are given in Tables XX to XXIV. Results of further tests at Magee Marsh of raccoon- proofing devices are shown in Table XXV. Because raccoon predation was relatively unimportant in the unprotected boxes, the results are less con­ clusive than is to be desired. Likewise, use of boxes by ducks was not sufficiently extensive to give conclusive results.

Conventional or Basic Type Box (Figure 46), - The box here designated as the conventional type box was widely used in early wood duck nesting box programs. It was developed by A.S. Hawkins and F.C. Bellrose in 1938 and was first used in Illinois (Bellrose, 1955b:19). This box was made

219 Table XVIII

Summary of Wood Duck Nest Box Survey in 1956

Hole-in­ Hole-in­ Ellip­ Grease Post vs. lid amm­ side amm­ Basic Tunnel tical drum tree unition unition Basic Handley (new) entrance entrance mount case case (old) design No. boxes checked AS AS A6 A9 92 56 77 69 16

Year Boxes erected 1956 1956 1956 1956 1956 1956 1956 195X 1955

In trees or on posts trees trees trees trees half each trees both posts both

No. boxes empty 20 18 7 30 35 A3 61 33 12 to

% boxes empty* XI .7 37.5 15.6 61.2 92.X • 79.2 X7.8 75.0

No. boxes used by bees 3

% boxes used by bees 13.7

No. boxes used by raccoons 1 1

% boxes used by raccoons 2.1 2.1

*Per cent figures are based on the numbers of boxes of each type. Table XVIII (Continued)

Hole-in- Hole-in­ Ellip­ Post vs. lid amm­ side amm­ Basic Tunnel tical Grease tree unition unition Basic Handley (new) entrance entrance drum mount case case (old) design No. boxes used by squirrels 6 5 5 1 2

% boxes used by squirrels 12.5 10.4 11.1 2.1 3.6 - No. boxes used by mice 1

% boxes used by mice 2.1

No. boxes used by starlings 9 17 29 11 1 2 1 23 1

% boxes used by starlings 18.7 35.4 64.4 22.4 1.1 3.6 1.3 33.3 6.2

No. boxes used by screech owls 4 2 1 3 1

% boxes used by screech owls 8.3 4.2 2.2 6.1 1.1

No. boxes used by grackles 4

% boxes used by grackles 4.3

No. boxes used by bluebirds 4

% boxes used by bluebirds 5.2 Table XVIII (Continued)

Hole-in- Hole-in­ Ellip­ Post vs. lid amm­ side amm­ Basic Tunnel tical Grease tree unition unition Basic Handley (new) entrance entrance drum mount case case (old) design No. boxes used by English sparrows 1

% boxes used by English sparrows 1.1

No. boxes used by tree swallows 1

% boxes used by tree swallows 6.2

No. boxes used by white­ breasted nuthatch 1

% boxes used by white­ breasted nuthatch 1.8

No. boxes used by tufted titmouse 1

% boxes used by tufted titmouse 1.5

No. boxes used by wood duck 8 U 3 A 2 11 12 222

% boxes used by wood duck 16.7 8.3 6.6 8.2 3.6 U.3 17.5 Table XVIII (Continued)

Hole-in- Hole-in­ Ellip­ Post vs. lid amm­ side amm­ Basic Tunnel tical Grease tree unition unition Basic Handley (new) entrance entrance drum mount case case (old) design No. wood duck nests

successful 6 A 2 A 1 9 7

% wood duck nests successful 75.0 100 66.6 100 50.0 77.3 70.0*

No. wood duck nests deserted 1

% wood duck nests deserted 33.3 No, dump nests 1 1 . 2

% dump nests 12.5 50.0 22.2

No. wood duck nests destroyed by predators 1

% wood duck nests destroyed by predators 12.5 No. deserted as result of study 2

% deserted as result of study 16.7

* Desertions caused by this study are discounted in this calculation. Table XVIII (Continued)

fHole-in- Hole-in­ Ellip­ Post vs. jlid amm- side amm­ Basic Tunnel tical Grease tree unition unition Basic Handley (new) entrance entrance drum mount 1 case case _(old) design No. wood duck nests disrupted by starlings 3

% wood duck nests disrupted by starlings 2.5 225

Table XIX

Summary of Wood Duck Nest Box Survey in 1957

Hole-in-lid, Tunnel I j Elliptical Grease j ammunition j Basic entrance • entrance drum case I

No. boxes checked AO 35 36 36 31 Year boxes erected 1956 1956 1956 1956 1956

No. boxes empty 21 17 U 24 19

Per cent boxes empty 52.5 4.8.6 38.9 66.7 61.3 No. boxes used by raccoons 1 1 1

Per cent boxes used by raccoons 2.5 2.8 3.2

No. boxes used by squirrels 2 6 11 4 5 Per cent boxes used by squirrels 5.0 17.1 30.6 11.1 16.1 No. boxes used by opossum 1 1

Per cent boxes used by opossum 2.5 3.2 No. boxes used by starlings 2 7 7 3 1 Per cent boxes used by starlings 5.0 20.0 19.4 8.3 3.2 No. boxes used by screech owls 1 2 2 1 1 Per cent boxes used by screech owls 2.5 5.7 5.6 2.8 3.2 No. boxes used by crested flycatchers 2 Per cent boxes used by crested flycatchers 6.5 No. boxes used by wood ducks 9 1 1 1 1 Per cent boxes used by wood ducks 22.5 2.9 2.8 2.8 3.2 226

Table XIX (Continued)

Hole-in-lid Tunnel Elliptical Grease ammunition Basic entrance entrance drum case No. wood duck nests successful 5 1 1 1 1 Per cent wood duck nests successful 55.5 100 100 100 100 No. boxes used by bees 1 1 Per cent boxes used by bees 2.5 2.9 No, wood duck nests deserted 1 Per cent wood duck nests deserted 2.9 No. wood duck nests destroyed by raccoons 3 Per cent wood duck nests destroyed by raccoons 7.5 No. wood duck dump nests 1 Per cent wood duck dump nests 2.5 No. sparrow hawk nests 1 1

Per cent sparrow hawk nests 2.8 3.2

No. boxes not in

repair A 1 1 2 Per cent boxes not in repair 10.0 2.9 2.8 5.6 Table XX

Comparative Use by Wood Ducks of Different Types of Nesting Boxes 1956

Tunnel Elliptical Grease Kole-in-lid Basic entrance entrance drum Ammunition case Total Numbers of boxes used by wood ducks 8 4 3 4 2 21

Expected 4.2 4.2 4.2 4.2 4.2 21

X2 = 4.95

X2^^ = 9.^88 (4 degrees of freedom)

1957

Numbers of boxes used by wood ducks 9 1 1 1 1 13

Ibcpected 2.6 2.6 2.6 2.6 2.6 13

X2 = 19.692 2 X = 9.488 (4 degrees of freedom) Table XXI

Comparative Success of Wood Ducks in Different Types Boxes 1956

Tunnel Elliptical Grease Hole-in-lid Basic entrance entrance drum ammunition case Total

Number of successful wood duck nests 6 4 2 4 1 17 Expected 6.5 3.2 2.4 3.2 1.6 17 Number of unsuccessful wood duck nests 2 0 1 0 1 4

Expected 1.5 .8 .6 .8 •U 4

Total 8 4 3 4 2 21 X2 - 3.66 X qj = 9.488 (4 degrees of freedom)

1957 Number of successful wood duck nests 5 0 1 1 1 8 Expected 5.54 .62 .62 .62 .62 8 Number of unsuccessful wood duck nests 4 1 0 0 0 5 Expected 3.46 .38 .38 .38 .38 5 Total 9 1 1 1 1 13 X2 = 6.663 X2o5 = 9 .^88 (4 degrees of freedom Table XXII

Relative Use of Different Types of Nesting Boxes 1956

Tunnel Elliptical Grease Hole-in-lid Basic | entrance entrance drum ammunition case Total

Number of boxes used by wood ducks 8 4 3 4 2 21

Expected -4.08 4.08 3.91 4.17 4.76 21

Empty 20 18 7 30 48 123

Expected 23.90 23.90 22.91 24.40 27.89 123

Other use 20 26 36 15 6 103

Expected 20.02 20.02 19.13 20.43 23.35 103

Total 48 43 46 49 56 247

X2 = 65.336 2 X n„ = 15*507 (3 degrees of freedom) Table XXII (Continued)

1957 Tunnel Elliptical Grease Hole-in-lid Basic entrance entrance drum ammunition case Total

Number of boxes used by wood ducks 9 1 1 1 1 13

Expected 2.92 2.56 2.63 2.63 2.26 13

Empty 21 17 U 24 19 95

Expected 21.35 18.68 19.21 19.21 16.54 95

Other use 10 17 21 11 11 70

Expected 15.73 13.76 14.16 14.16 12.19 70

Total 40 35 36 36 31 178

x2 = 26.4390

X205 = 15.507 (8 degrees of freedom) Table XXIII

Comparative Use by Starlings of Different Types of Nesting Boxes 1956

Tunnel Elliptical Grease Hole-in-lid Basic entrance entrance drum ammunition case I Total r Number of boxes used by starlings 9 17 29 11 2 68

Expected 13.6 13.6 13.6 13.6 13.6 68

X2 = 30.24.

X u?__ — 9.438 (4 degrees of freedom

1957

Number of boxes used by starlings 2 7 7 3 1 20 Expected 4 4 4 4 4 20

X2 = 8.00

X205 = 9-483 (4 degrees of freedom) Table XXIV Relationship of Nesting Box Usage by Wood Ducks and Starlings 1956

Tunnel Elliptical Grease Hole-in-lid Basic| entrance entrance drum ammunition case Total Number of boxes used by wood ducks 8 4 3 4 2 21 Expected 4.0 5.0 7.6 3.5 .9 Number of boxes used by starlings 9 17 29 11 2 68 Expected 13.0 16.0 24.4 11.5 3.1 Total 17 21 32 15 4 89 X2 = 10.97 X2 = 9.438 (4. degrees of freedom) 05

1957 Number of boxes used by wood ducks 9 1 1 1 1 13 Expected 4.33 3.15 3.15 1.58 .79 Number of boxes used by starlings 2 7 7 3 1 20 Expected 6.67 4-85 4.85 2.42 1.21 Total 11 8 8 4 2 33

X2 = 13.592

X205 = 9.488 (4 degrees of freedom) Table XXV

Utilization of Different Types of Nesting Boxes at Magee Marsh, 1957

Destroyed Successful Boxes Boxes with Type of box Used by Dump by wood duck with tree Empty wood ducks nests raccoons nests starlings swallows boxes Total

Unprotected 5 5 A 2 22 35

Elliptical entrance 1 1 1 1 5 8

Tunnel entrance 2 1 1 2 2 8 U

Post-funnel 5 2 1 5 A 2 17 3 A Total 13 2 2 12 11 7 52 91 Figure 46

Conventional or Basic Type Box 235 of 1-inch rough-cut cypress lumber. The box was 23 inches deep and 10 inches square inside and had a circular entrance hole 4- inches in diameter

k inches below the top of the box. The back of the box was extended two inches beyond the roof and the floor to provide a place for mounting.

The removable roof was held in position on the box with two strips of wood and was attached with a wire hook.

In this study, because of availability, uncured lumber of American elm was used for boxes of the basic, elliptical entrance, and tunnel entrance types. This lumber was cut to order and bought at $96.50 per thousand board feet. Three-fourth inch lumber was used. The lumber for each box cost about $1.00. The nails for assembling the boxes and the bolts and washers for mounting cost about $0,20.

Through the use of a band saw, the entrance holes were first cut into all of the box fronts. With two men working, these holes were cut in about one minute each. At $1.20 per hour for labor, each of these holes cost $0.04- each. In spite of the fact that the uncured lumber did not work nicely, 30-35 boxes were made in each 8-hour day through the aid of a power saw. At $1.20 per hour for labor, construction of the boxes cost $0.30 each. The total cost per box for material and con­ struction was $1.54- each. It is assumed for purpose of this calculation that necessary tools and equipment are available at no cost.

At $1.20 per hour for labor, mounting the boxes cost about $0.50 each. This makes a total cost of $2,04. per "box. There would be an additional variable cost for transportation when erecting the boxes.

It can be expected that the cost per box for construction would vary greatly with the speed and efficiency of the operator. 236

The conventional or basic type box was used by ducks more frequent­

ly than boxes of the other four types. In 1956, the rate of use of this

type box was double that of the other most used type, but this differ­

ence was not statistically significant. In comparison of 21 nests in the

5 types of boxes, a chi-square value of 4*95 was obtained when 9.488 is

significant at the 5 per cent level. In 1957, however, a statistically

significant difference was observed. Nine of 13 nests were in the basic

type box, and a chi-square value of 19.692 was obtained when 9.4-88 is

significant at the 5 per cent level (Table XX). Ducks showed a preference

for the basic type box in 1957 but not in 1956.

During 1956 and 1957 there were 34 wood duck nests in the boxes in­

volved in the test of 5 different box types. Seventeen of these nests were in basic type boxes. Four of the 17 were destroyed by raccoons.

Success in different types of boxes did not, however, show a statisti­

cally significant difference (Table XXI). In 1956, a chi-square value

of 3.66 was obtained, 9*488 being significant at the 5 per cent level;

in 1957 a chi-square value of 6.663 was obtained when 9*488 is signifi­

cant at the 5 per cent level.

Cronan (1957) observed that wood ducks prefer boxes without preda­ tor guards when given a choice, but they will readily use boxes with guards if others are not available.

Elliptical Entrance (Figure 47)* - For raccoon-proofing wood duck nesting boxes, Bellrose (1955b:23-24) recommended use of elliptical entrances 4 inches wide and 3 inches high. In the present experiment, elliptical ^ioles were cut in a strip of 24 gauge sheet metal 8 inches by 10 inches. This metal would cost $0.10 to $0.15 per box, and the Figure 47

Elliptical Entrance Nesting Box 238

charge of a tinsmith for making the holes would be about $0.15 each.

The metal sheets were nailed over the entrances of basic type boxes.

Wood duck use of this type box was less than of the basic type, and it was about equal to that of the tunnel entrance and grease drum

boxes. For some reason, starlings showed a statistically significant

preference for the elliptical entrance boxes. When utilisation by star­ lings of the five types of boxes is compared, a chi-aquare value of

30.24. is obtained when 9.438 is significant at the 5 per cent level.

Twenty-nine of 68 starling nests were in boxes of this type (Table

XXIII). No nests in this type box were lost to raccoons.

Tunnel Entrance (Figure 48). - Tunnel entrance boxes were first designed by the Massachusetts Division of Fisheries and Game (McLaugh­ lin and Grice, 1952). The U.S. Fish and Wildlife Service (Webster,

1954-J3) has also used tunnel entrances and reported limited success in raccoon-proofing nesting boxes with their use. The tunnel entrance is

10 inches long and 4 inches square inside. Actually the tunnels used in this experiment in Ohio were 6 instead of 10 inches long. The tunnels were nailed over the entrances of basic type boxes. This made an added cost of about $0.30 per box as compared with the conventional type box.

These boxes were used by wood ducks about equally to the grease drum and elliptical entrance boxes. There was no raccoon predation, except in one box where entrance was gained through the roof, but a raccoon was found in an additional box.

Grease Drum (Figure 49). - Chassis and gear lubricants are obtained by automobile service stations in steel drums holding 100 pounds of Figure I f i

Tunnel Entrance Nesting Box P Figure U 9

Grease Drum Nesting Box 241

lubricant. At some stations, drayage is paid for having the empty drums

removed; at some stations the drums are given to customers; at some

stations they are sold at $0.50 to $1.00 each. With some shopping

around, then, lubricant drums can be obtained without cost. There re­

mains the task of cleaning the drums, and when they had been used for

chassis lubricant, this is not an easy job. Gear lubricant rather than

chassis lubricant drums were found best because of greater ease in

cleaning.

After these drums were given some time to drain in an upside-down-

position, they were washed with turpentine and painted on the outside.

The front of the inside of the drum was covered with a layer of auto­ mobile undercoating mixed with sawdust. A 4“inch circular hole was

burned into a side near the top of the drum for entrance of the ducks;

a one-half inch hole was burned into the bottom of the drum to provide

for drainage; a one-half inch hole was burned near each end of the drum on the side opposite the entrance hole for use in mounting the drum.

The cost of cleaning the lubricant from the drums at $0.02 per drum for turpentine and $1.20 per hour for labor was $0.10 each. The cost of painting at $6.00 per gallon for paint and $1.20 per hour for labor was $0 .40. The automobile undercoating cost $0.03 per box. The holes were burned into the drum for a charge of about $1.00 each. The total cost of preparing the drum for a nesting place for wood ducks was $1.55»

Grease drum boxes were used about equally with tunnel entrance and elliptical entrance boxes. There was no raccoon predation in these boxes, but they were entered by raccoons. 242

Hole-in-lid Ammunition Case (Figure 50). - As surplus property

of the U.S. A r m y , The Ohio Division of Wildlife purchased several thou­

sand ammunition cases at $0.25 per case. The case is 18 inches by 14-

inches by 12 inches, with a hinged lid on the top. In preparing these

boxes for wood duck nesting places, a 4-inch circular hole was cut near

the center of one end of the lid; a piece of asphalt roofing material

was nailed on one end for a roof; the lid was cut in two pieces,

slightly to one side of the middle, and a board was attached to the

bottom of the box opposite to the entrance for use in mounting. The

bottom portion of the lid was nailed shut, and the hinged upper portion

was used for inspecting the inside of the box. The 4-inch entrance

holes were cut into the boxes with a hole saw. The total per unit cost

of these cases and of preparing them for use as wood duck nesting places was $0.70.

Only 2 of 56 (3.6 per cent) hole-in-lid ammunition cases were used

by wood ducks in 1956, and one of the nests was a dump nest. None of

an additional 92 boxes of this type used in a mounting test were used

by ducks in 1956. The rate of use of these boxes suggests that they are of a less desirable type for wood ducks than the other boxes used.

These boxes were more often empty than other types and were used less by starlings. Of 68 starling nests in the 5 types of boxes in 1956, only 2 nests were in boxes of this type.

Hole-in-aide Ammunition Case (Figure 51). - Not all of the ammunition cases put out for wood ducks by the Ohio Division of Wildlife were pre­ pared to exactly the same specifications. Some had the entrance hole in the side of the box rather than in the lid; these are here referred to as Figure 50

Hole-im-lid Ammunition Case Nesting Box Figure 51 ?o Hole-in-side Ammunition Case Nesting Box 245

the hole-in-side ammunition case type. The resulting difference was a

depth from the front to the back 2 inches greater and a width 2 inches

smaller than in the hole-in-lid type.

There were 11 wood duck nests in 89 (12.3 per cent) boxes of the hole-in-side ammunition case type. The two types of ammunition cases were not mounted in conjunction with each other, but there was seeming­ ly a comparable chance for utilization of both types so far as ducks and habitats were concerned.

Wood shavings had been placed in the hole-in-side boxes rather than sawdust as was placed in the hole-in-lid boxes. Sawdust was read­ ily used by wood ducks in other types of boxes, and the location of the hole seems to be the critical difference between these two boxes.

On the basis of this test it must be concluded that the hole-in- side is to be preferred to the hole-in-lid type of box. These boxes were used relatively little by starlings; consequently, they were more often available for wood ducks than boxes of most other types.

Galvanized Pipe Nest Box. - A metal box for wood ducks was first designed tjy Louis Ellebrecht, former manager of the Chatauqia National

Wildlife Refuge. Bellrose (1955b) found this box to be used by wood ducks and to be somewhat predator proof. The box used was of 26-gauge galvanized sheet metal, 12 inches in diameter and 24 inches high. An inverted galvanized cone, 15 inches in height, was used for a roof.

The bottom was made of a circular piece of wood. Two screws were used to attach the roof to the sides so the roof could be removed. The front of the inside of the box was coated with a layer of automobile under­ coating and vermiculite, one-half inch just below the entrance but 24.6

thinner elsewhere. Plans to test this type of box in Ohio during 1957 were thwarted because those ordered were received too late for the nesting season.

Using 18 of these boxes in Pennsylvania, 100 per cent utilization and 100 per cent success of nesting ducks was attained by periodically removing starling and tree swallow nests (R.E. Lattimer, personal communication, 1957). These boxes were painted green and mounted on posts over water. Comparisons were later made with boxes painted aluminum color with Galvinoleum paint. When in the hot sun the aluminum boxes average 5.5°C. cooler than the green ones. The alum­ inum colored boxes were used by the ducks in numbers equal to the green ones.

Metal boxes can now be purchased from H. Behrens Manufacturing

Company, Winona, Minnesota, for $4-8.00 per dozen.

Nail Keg Boxes. - Largely because of their cheapness and availabil­ ity, nail kegs have in the past been much used for nesting boxes for wood ducks. Bellrose (1955b:32) found that nail kegs are utilized only about half as much as board boxes when the two types are present in the same area. Klein (1955:70) inspected 77 board boxes and 58 of the nail keg type and found 26 per cent utilization of board boxes and 17 per cent use of nail keg boxes. The difference was found not to be statis­ tically significant. Boxes inspected by Klein were not erected in com­ bination with the board type boxes, but he thought the keg type was used less if there was an alternative choice available.

Nail keg boxes were not tested in central Ohio. These kegs are no longer readily available at hardware stores because cartons are being 2 47 increasingly used for shipment of nails.

Miscellaneous Types of Boxes. - Boxes of slightly different speci­ fications have been successfully used and recommended by various work­ ers. Handley (Cochran, 1955s206) has recommended a box with the floor

10 inches square, the front 18 inches high, the back 20 inches high, and with a circular entrance 4 inches in diameter. The front has a hinged door below the entrance hole for use in inspecting the box. Laurel

Van Camp has also recommended (personal conversation) a somewhat similar box; however, he prefers to make the entrance large enough so a hand can be inserted to remove unwanted material or to remove the female ducks for banding because of the increased hazards to the box imposed by a hinged door or removable roof.

New Versus Old Boxes

Miller (1952:5) found a higher rate of use the first year new boxes were erected and was led to suspect that boxes of new wood were more con­ spicuous and more easily found than old or weathered boxes. McLaughlin and Grice (1952:246) noted that when new boxes are added in an area con­ taining older boxes, the new boxes are used more than the old. In Ohio, a decline was noted in the use of tree-mounted boxes in the second year.

This decline was thought to have been caused by increased use of the boxes by raccoons. The boxes were erected only a few weeks before wood duck nesting started in 1956, and few were used by raccoons before the nest­ ing season was completed. Many of the boxes had signs of use by raccoons at the time of the 1957 inspection. Post Versus Tree Mounting

In Massachusetts in 194-6, McLaughlin and Grice (1952:244-) found

boxes mounted on posts over water preferred by wood ducks to boxes

mounted on trees. The boxes were mounted 3-5 feet above the highest

water level on cedar posts 14 feet long. In Vermont, Miller (1952:5)

found little difference in the rates of utilization of tree- and post­ mounted boxes.

In Ohio, 92 hole-in-lid ammunition case boxes were erected in

1956 in a post versus tree mounting test. These boxes were erected

throughout central Ohio in pairs on farm ponds. Where one box was

placed on a post over the water, another box was placed nearby on a

tree. None of these boxes was occupied during 1956. In 1957, however,

four of 32 boxes examined contained wood duck nests; two of these were

dump nests. All of these four nests were in boxes mounted on posts.

Most of the boxes erected by the Ohio Division of Wildlife have

been mounted on steel fence posts (Figure 52). The posts used are of the type used for mounting highway signs by the Ohio State Highway

Department. These posts sometimes bend in strong winds (Figure 53).

In the present study, used boiler flues were used, and these do not bend so readily. The boiler flues were purchased at $0.17 per foot.

Except at Magee Marsh where post-mounted boxes were heavily used by raccoons in 1956, the disturbance caused by raccoons in post-mounted boxes was negligible. Post-mounting apparently reduces the losses of wood ducks to raccoons.

24S 249

\

Figure 52

Wood Duck Nesting Box Mounted on Steel Fence Post Figure 53 w The Steel Fence Posts Are Sometimes Bent by Strong Winds o Density of Nesting Boxes

Wood ducks are to some degree social in their nesting habits. Of

44- nests examined in 1956, 20 were with a known neighboring nest or

nests within 50 feet. Five nests were so located in a small area that

the nests at the two extreme points were only 200 feet apart. On the

other hand, McLaughlin and Grice (1952:253) found that the number of

dump nests increased with the density of breeding ducks, and the number

of nest desertions increased with the number of dump nests. It is not

clear whether the increase in desertions was proportional to the popu­

lation increase or whether there were more desertions per breeding

duck. Bellrose (19556*4-0) reported best results with boxes spaced 50

to 100 feet apart. Every box was occupied in several units grouped

with 2 to 4 boxes per acre. McLaughlin and Grice (1952:24-5) had best

results when boxes were placed with 5 to 10, rather than more, in a

loose group. These workers suggested that there are too many boxes in

a given area unless 50 per cent or more of the boxes are used.

Sawdust or Other Loose Material in Nesting Box

Although Chapman (1924.: 198) stated that the nest consists of grasses, leaves, and twigs, the wood duck brings in no such material, and if no loose material is placed in the box in advance of the wood duck's nesting, its eggs can only be deposited on the bare floor.

The female does not pluck down from her breast until after several eggs have been laid, and before down is added the eggs are covered with loose material already in the nesting cavity. The chances of uti­ lization of a cavity are increased if the cavity contains loose material 252

I have seen only one occupied cavity with no loose material. Bellrose

(1955b:33) stated that a box lacking loose material will not be used

by ducks. Sawdust and wood shavings seem to be equally suitable for

nesting material.,

Height and Exposure of Boxes

McLaughlin and Grice (1952:24-5) noted that boxes were more readily

used when the entrances were visible from the duckx s usual travel lane;

these authors infer that such boxes are more easily found. Hawkins

and Bellrose (194-1:395) found it was not necessary to face boxes toward water which in the case of a stream would be the ducks1 travel lane.

Klein (1955:71) rated 135 boxes as bright or dim on the basis of

exposure to light and found no statistical difference in the rate of utilization in the two categories; 8 of 53 (15 per cent) bright boxes

and 22 of 82 (27 per cent) dim boxes were used by wood ducks.

Gigstead (1938:604-) and Klein (1955:71) found that utilization of boxes was not affected by the direction they faced. However, Klein

(1955:72) thought that 7 of 11 embryos were destroyed by exposure to weather in a clutch under his observation. The box containing this clutch was on an exposed, windy hilltop.

When boxes are placed on posts over the water of a pond it is not necessary that there be trees nearby. Boxes have been used on ponds as much as one-fourth of a mile from trees other than a few small trees in a fence row (Figure 54-) •

Bellrose (I955b:38) recommended that boxes be mounted at a height of 10 to 25 feet. Shreiner and Hendrickson (1951:7) found that the Figure 5U t o Boxes Were Used as Much as One-fourth Mile from Large Trees <5 254

height of boxes made little difference in the rate of utilization.

Klein (1955*72) also found no statistical difference in box utilization

correlated with height of boxes: 19 of 95 (20 per cent) boxes 0 to 8

feet high were used; 11 of 34 boxes (32 per cent) 8 to 15 feet high

were used. In this study in Ohio, boxes were successful when mounted

only one or two feet above water.

Vandalism (Figure 55)

In a nesting box program where the boxes are widely scattered,

vandalism is a problem to be considered. Hunters shot through the

sides of several boxes, seemingly to frighten squirrels which had taken

retreat in the boxes. At Knox Lake, a public fishing area, the boxes

can easily be reached from a boatj here the boxes were frequently

opened by fishermen and left without being closed.

Boxes mounted 15 feet from the ground on trees were not readily reached by boys and were little disturbed. In a woods adopted for a

shooting range, one tree-mounted box was used for a target. Actually,

little difficulty from vandalism was experienced in the three nesting

seasons of this study.

Checking Nesting Boxes

Faithful care of wood duck nesting boxes is an essential feature of any nesting box program. Because of competition from other species and because of the possible need for box repairs, it is desirable to check wood duck nesting boxes before the onset of each nesting season. This means that the boxes should be checked at least once annually, and in the latitude of central Ohio, this check might best be made in about the The Nesting Boxes Were Sometimes Lost to Vandals 256

last half of February. The time of nesting box checks should be

later in the season progressively northward and earlier southward.

In northern Ohio, the first half of March would be about the proper

time, and in Florida the check should be made as early as late January.

If all wood duck nesting boxes cannot be examined at least once annual­

ly, there is little reason for putting them out in the first place.

In any wood duck management program involving the use of nesting

boxes it is of primary importance to acquire some knowledge of the

extent of utilization of the boxes by the desired species as well as

the degree of nesting success. A second nesting box check of a pilot

sample is extremely desirable. This second check should be used chief­

ly for gathering information on the preferred types of boxes and mount­

ing situations, the proportion of boxes utilized by wood ducks, the

extent and nature of competition of wood ducks with other species of wildlife, and the overall effectiveness of the nesting box program.

This nesting box check should be used to guide further work with the species. If, for instance, 90 of 100 boxes examined contained success­ ful wood duck nests, the erection of more boxes might be indicated.

Summary

Wood ducks usually nest in tree cavities or nesting boxes, but nests have been reported from a variety of other situations. In Illinois,

58 occupied natural cavities were examined. The greatest depth found was

52 inches. The average diameter of the cavities was 10 inches. Entrances averaged 6 inches in diameter. The entrance may be as small as A inches by 2.5 inches or 3.25 inches by 3.50 inches. In Iowa and Massachusetts, 257

121 cavities were examined; 24 were considered suitable for wood ducks;

10 were used by these ducks. A nest was reported in a chimney 8 feet

from the top of the chimney.

Wood ducks readily use man-made nesting boxes. A rate of utiliza­ tion as high as 92.3 per cent of 64 boxes was attained in Connecticut.

In 1936, slab boxes were first tested in Illinois for wood duck

nesting places. Various types of boxes have since been tested in Illinois, and an all-metal box is now considered best. This box is readily used by

ducks, and it is considered to be most predator proof of the various boros tested. Elliptical entrances, 4 inches by 3 inches, are reason­

ably effective in preventing entrance of raccoons in the northern part ofthe United States, but these are ineffective in the South where raccoons are smaller.

Predatory animals sometimes follow human trails and destroy birds* nests visited. In this study, not more than one nest could have been lost to this cause.

I tested 247 nesting boxes of five types in regard to utilization and nesting success. These boxes were randomly erected in regard to box types. Because of limited use of the boxes by ducks and little loss to raccoons, the test must be considered inconclusive. The conventional type box was most used by ducks, but nests in it suffered most predation.

Use of the tunnel and elliptical entrance and the grease drum boxes by ducks was about equal. Raccoons did not enter the elliptical entrance boxes. No nests were destroyed, but raccoons entered tunnel entrance and grease drum boxes. The hole-in-lid ammunition case was little used by ducks. 258

Some additional boxes, not erected in combination with the five

types listed above, were checked. Eighty-nine hole-in-side ammunition

case boxes contained 11 successful wood duck nests. This box was little used by starlings and was more often available to wood ducks. Galvan­

ized pipe nest boxes were not tested in this Ohio study. One hundred

per cent utilization and success in 18 boxes of this type was reported

in Pennsylvania.

Nail keg boxes were not tested in Ohio because of their decreasing availability. They have been tested in other states and found to be generally used less than most other types of boxes.

Various workers have found that boxes are sometimes used more the first year after erection than in subsequent years or new boxes are used more than old ones when new boxes are added among old boxes. This decline is thought to be caused by increasing use by raccoons as the boxes are in place a longer period.

There appears to be little difference in the rates of use of nesting boxes mounted on trees or on posts.

Best results with nesting boxes are attained when boxes are erected in loose groups, with 5 to 10 boxes per unit.

The inside of nesting boxes should be provided with some loose material such as sawdust or wood shavings.

There is conflicting opinion in the literature as to whether boxes should be conspicuously placed along the duck’s fly lane. The boxes can be on ponds as much as one-fourth mile from trees.

The boxes should be erected high enough from the ground or water that they are not readily accessible to boys. Boxes only a few inches 259 above water were used.

For maximum success in a nesting box program, boxes should be re­

paired each season before nesting starts. A nesting season or post­ nesting season check is desirable to guide further work in the use of nesting boxes. CHAPTER XIII

MARKING INDIVIDUAL WOOD DUCKS FOR SUBSEQUENT IDENTIFICATION

An effective means for insuring repeated identifications of indi­ vidual ducks or broods is an integral part of any detailed study of wood duck movements. Several methods of capturing wood ducks and of marking them for subsequent identification have been used. These methods have various degrees of usefulness.

Bait Trapping Wood Ducks

Because banding wood ducks in large numbers usually involves trapping in scattered locations, small portable traps are preferred.

The traps used were 6 feet long, 3 feet wide and 3 feet high. These traps were made collapsible so they were easily transported in the trunk of an automobile, on an automobile top rack, or in a station wagon. All of the six traps which were used could be transported at one time on an automobile rack (Figure 56). One of the traps could be removed from the rack and set ready for operation (Figure 57) in less than ten minutes.

One-inch-mesh fence 3 feet wide and made of 11-gauge wire was used. This was sufficiently rigid so that no reinforcement was necessary.

Twenty-six linear feet of wire were used and the following pieces were cut: two pieces 3 feet long, two pieces 2 feet long, and three pieces

6 feet long. These were fastened together with hog rings.

A piece 2 feet long was fastened to the end of a 6-foot piecej a piece 3 feet long was attached to the opposite end of the 6-foot piecej

260 Figure 56

The Six Wood Duck Traps Were Easily Moved from Place to Place Figure 57 to O' One of the Wood Duck Traps in Operation to 263

rings were placed to within 8 inches on one side of a second piece 3

feet long attaching it to the other end of the 3-foot piece; a 6-foot

piece was attached to the second 3-foot piece; and a 2-foot piece was

attached to the opposite end of the 6-foot piece. These pieces were

attached in a straight line and the sides of the trap were thus com­

pleted. The third 6-foot piece was then laid parallel with one of the

6-foot sides and attached with hog rings along the side opposite to that where rings were ommitted between the two three-foot pieces. This 6-

foot piece was the top of the trap. The 3-foot pieces were folded toward

the 6-foot pieces, thus forming a funnel. At the apex of the funnel at

the side opposite that where the 6-foot top was attached, an opening

for entrance of the ducks w a s made. Three cross wires were cut 8 inches

from the bottom. The 8-inch flaps of wire thus formed were bent to make an opening to admit the ducks. By folding the hinged funnel to­ gether, folding the two loose flaps at the opposite end of the sides

inward so that they laid against each other between the sides, and the top against the side on which it hinges, the trap was made ready for transportation.

In setting the trap, the sides were spread with the funnel opening against the ground. The hinged top was swung into place and fastened to the trap end, free side, and the funnel. The two loose flaps at the end of the trap opposite the funnel were swung into place, lapping over each other, and fastened together and to the top of the trap. Fine copper wire was used to hold the t o p and the end opposite the funnel in place.

The apex of the funnel where the birds enter was kept free of the ground so that the loose flaps could be freely spread by entering ducks. The 264

flaps of the entrance were separated about three inches. They must

not be too widely spread as ducks will then escape.

In general, wood ducks were rather easily trapped, especially

during the spring and summer months. During the nesting season, the

adults took corn placed for them and readily entered a trap later

operated over the corn. Likewise, the ducklings were easily trapped

with corn bait by the time they were 3-4 weeks of age when their legs were first large enough to carry a band. On the Scioto River, c o m was

spread over a smooth, bare spot of mud in the known range of one or more broods of wood ducks, and within a few days the corn usually was

being eaten by the ducks. After the ducks were accustomed to feeding

inside of the trap, operation of the trap was started.

In pond habitats at the Olentangy Station, four stakes were driven

into the mud in a rectangular position among vegetative cover, and a platform was laid on the stake tops slightly above the surface of the water. Of 13 separate attempts to trap wood duck broods in such a situation, failure resulted in only one attempt. Then the brood was believed to have moved from the proposed trapping site to new feeding range nearby at about the time the trapping effort was initiated. Two hundred and forty-three local wood ducks were banded in the two years of this study.

Trapping ducks in the autumn in Ohio and in the winter in South

Carolina proved to be a more difficult problem than trapping pre-flight ducks,. Wood ducks then had habitual feeding sites where it was impossible to operate a trap because of interference from feeding hogs as in Ohio, and because trapping stations could not be legally established on private 26$

land during the hunting season where the ducks were feeding in South

Carolina, ¥ood ducks in the autumn flocks could usually be trapped when they could be induced to feed at a definite location. It was, however, not always possible to induce the ducks to start feeding at a chosen location, and when corn was placed near where birds were seen through the middle of the day, the ducks were seen to sit on the c o m loafing and preening without showing the slightest apparent interest in the com. After starting to feed at a station, the ducks continued to re­ turn regularly and were trapped by alternating trapping periods with baiting periods.

The usual procedure for establishing a fall wood duck banding station consisted of locating a wide, bare, sand, mud, or gravel bar within the regular range of a local population of wood ducks. A bushel or so of corn was scattered over the bare area, and the area was left undisturbed through several dpys or a week. After the ducks had started to feed on the c o m the traps were put in place with the doors left open and with an abundance of corn inside the trap, ¥hen the ducks had become accustomed to feeding in the open trap, the doors were closed and trapping was started. A total of 554 wood ducks were banded from these traps in two years of trapping, and an additional 208 repeats or recov­ eries were recorded. As many as 28 wood ducks were taken in one trap at one catch.

Few casualities resulted from trapping. One duckling, 5-6 weeks old, died in a trap as a result of protracted exposure to the hot sun when my planned return was prevented because of outboard motor trouble.

At another time, following an evening when I had last attended a trap 266

about a half hour before sunset, four dead ducklings, killed by a mink,

were found in the trap the following morning, On one occasion trapping

operations were suspended for a week, and the funnel of a trap was wired

shut. During my absence, a raccoon had forced the entrance open, a

wood duck later entered the trap, and a raccoon killed the duck. All

of these casualities resulted from some detail of improper attention

to the traps.

Banding Incubating Female Wood Ducks

Wood duck nesting boxes were checked chiefly in early May in

central Ohio, and incubating females could usually be captured on their

nests. Most incubating birds remained on their nests (Figure 53) and

were readily removed for banding. The nests were approached as quiet­

ly as possible and a hand introduced into the box entrance, the partial­

ly opened lid, or the door, and the incubating female was quickly caught

and pressed into a corner of the box above the eggs where she was held

with her wings against the palm of the hand and fingers so that flutter­

ing of the wings was prevented. The lid or door of the box was then

removed or opened wider, the free hand introduced, and the bird removed

from the box.

After banding, the duck was replaced on her nest and a hand was held

over the entrance a short time to prevent the duck*s immediate departure.

When it was desired to minimize the interference imposed by banding, the

nesting box entrances were plugged with a burlap bag attached bo a rope.

Several hours after banding, the plug was pulled from the nesting box

entrance with the rope. With the observance of such precautions the Figure 58

Most Incubating Females Gould Readily Be Removed from the Nest for Banding ^ incubating birds continued to remain on the nests hours after the plug was removed.

No nest desertions resulted from my banding activities alone, although 4-0 incubating birds were removed from their nests, banded, and restored to the nest. As indicated by an automatic activity recording device, one bird remained off her nest through 49 hours after banding.

After 49 hours away from the nest she resumed incubation, and ducklings later hatched from all her eggs in that nest. While no desertions re­ sulted from banding activities alone, two resulted from more extended disturbance associated with research. Working in Vermont, Miller (1952:

4) reported nest desertion by 13 of 104 female wood ducks removed from their nests. The observer should not conclude that desertion has re­ sulted until the bird has remained off the nest more than two days.

No effort was made to avoid removing birds from the nests at early stages of incubation other than to delay the first nesting box check to a date when the early-nesting birds had been incubating several weeks.

The first general nesting box check was made in late April and early May, and the incubating females were removed for banding at the same time.

Most incubating wood ducks remained quietly on their nests until a hand was introduced into the cavity, but a small proportion left the nesting cavity somewhat sooner. In several instances a long stick was introduced from the ground into the nesting box opening, and the escape of the duck was thus prevented until it was possible to ascend a ladder and capture the duck. An automatic nesting box trap was used to capture two unusually wary females nesting in boxes on posts over water. U3Q of a Dog for Capturing Flightless Young

Wheaton (1882:528-529) reported that Kirtland was successful in

training a d o g to capture flightless young wood ducks and to bring them

to him without injury so that he maintained a caged flock for several

years at Poland, Ohio. Through the courtesy of Wm.B. Hendershot, Ohio

Division of Wildlife, an attempt was made to use a Labrador Retriever

for capturing ducklings at the Olentangy Station in 1956. Although the

dog was exceptionally well-trained and soft-mouthed it fatally injured

the only wood duck it brought to us. Young wood ducks normally make .a

concerted struggle when first captured, and this makes it very difficult

for a dog to retain hold of a duckling without injuring it. It is be­

lieved to be almost impossible to have a dog so well-trained that it

can capture and hold a worthwhile number of young wood ducks without

injury to some.

Gollop (1956) reported notable success in the use of retriever

dogs in capturing mallards for banding in Saskatchewan and Alberta.

A comparison of the number of recoveries from mallards captured through

the use of a dog with numbers captured by drive-trapping indicated no

difference in survival rates.

Through the cooperation of Marvin Mansfield, Ohio Division of Wild­

life, a German short-haired pointer was tested for use in finding young wood ducks, with the thought that the young wood ducks could be easily

picked up after being pointed by the dog. It was found that the young ducks usually continued running through the vegetation when they were being closely approached by the dog, and the dog held point such a short time that his point was easily overlooked. Undoubtedly, an occasional

269 270

duck could be picked up in this manner as can be done by an observer walking unaided through vegetation in wood duck territory, but the use

of a pointer dog does not seem to offer effective aid in capturing

young wood ducks in large numbers.

The young ducks will sometimes crouch in the concealing vegetation

and remain until they are caught by the bander. Usually, however, they

continue running through the vegetation when being pursued and cannot

be captured. Driving the ducklings into improvised corrals was not

tried, but it might yield limited success in habitats such as farm

ponds. By the time the ducklings have reached an age when they are

large enough to band, they scatter among the vegetation and move fifty or more feet from water. A relatively large number of drivers and/or a very large corral would be needed, and extensive success could not be expected because of the limitation and scattered distribution of occupied habitats susceptible to this method of coverage.

Imping Feathers

Wright (1939) has suggested a method of marking birds for sub­ sequent identification by attaching several conspicuously colored for­ eign feathers to the birds* natural feathers. A method somewhat similar to Wright*s was accidentally used in this study. Many females and duck­ lings were returned to the nests after being marked with airplane dope before the dope was thoroughly dry. Some duck down in the nest thus often became trapped in the wet dope. In several cases this down was held by the dried dope through as much as 3-5 weeks. In the case of two adult females and five ducklings, this down rather than the dope, called 271 attention to the identity of the birds. The protruding down was more readily detected than the color marks.

Painting and Dying of Young and Adult Wood Ducks

Yellow Testor*s airplane dope was applied to certain areas or com­ binations of areas of the plumage of young and adult ducks. The marked birds were most readily found in the field when a rather large surface of the bird, such as its entire tail or its entire head, was distinc­ tively colored. It was found hazardous to paint the flight feathers, as a male with painted primaries was unable to fly.

Four and five coats of airplane dope were applied, and time was allowed between coats for partial drying. Painted ducks could be identified at distances up to about 500 yards. Birds still wearing their distinctive marks were found four to five weeks after marking.

Color-marking proved to be an extremely useful method of marking in field studies. Sowls (1955:6-7) was successful in the use of color- marking of larger ducks, but he was less successful in using this method on smaller ducks such as blue-winged teals and shovellers.

In studies of brood movements, either the young or the adult fe­ male or both were marked at AS nests during 1955 and 1956. An addition­ al 20 birds, taken in bait traps, were color-marked as part of an effort to gather information on the distance of movements of birds flushed in river censuses.

Miller (1952:3) reported dying A3 unhatched young wood ducks by egg injection (Evans, 1951), but none of the birds were seen after leav­ ing the nest. This was not considered a suitable method for marking 272

ducklings of this species.

Color-banding

Sowls (1955*7-8) considered colored bands an indispensable aid in

local studies of waterfowl. Loafing places were created in order to

facilitate observation of the colored bands. In study of wood ducks

it was particularly desired to be able to identify females with broods,

and females with broods were seldom seen in such a way that leg bands

could be observed. Accordingly, no effort was made to use colored

bands.

Toe-clipping and Web-punching of Ducklings

A total of 102 ducklings in 8 broods were toe-clipped before they

left their nests. A razor blade was used to sever a claw at its base,

and the ducklings were thus marked as to their nests of origin. Only

19 of these toe-clipped ducklings were subsequently known to have been

recovered, although intensive trapping was conducted within the ranges

of most marked broods.

A disadvantage of toe-clipping derives from the fact that claws

and toes of ducks are sometimes lost as a result of other causes.

Twenty-four ducks of the 366 banded in 1956 had toe nails or parts of toes removed by means presumably other than the investigator1s razor blade. When pre-flight ducks were captured in a baited trap, they

sometimes scrambled about the trap and crawled up its sides, especially in the acute-angled corner beside the funnel. Some of these birds which had thus recently damaged their toes so that they were bleeding may have 273

been birds earlier toe-clipped, but because of the fresh injury they

could not be positively identified. Because the toes of ducklings are

sometimes damaged in banding traps, toe-clipped birds should be recog­

nized on the initial capture of the duck, rather than at times of re­

capture .

Toe-clipping is an easy method for marking nestling wood ducks

for subsequent identification, but it does not give a positive identifi­

cation .

Miller (1952:3) punched holes in the webs of 34-7 ducklings* feet,

and 21 of these were subsequently trapped and banded; in 1951 he

punched the webs of 527 ducklings of which 122 were later recovered.

Wing-taggjng of Ducklings

The use of wing tags presents some advantages when compared with

toe-clipping as a method of marking young wood ducks in the nest. Size

0 fingerling tags were placed on 25 captive ducklings in an effort to gain some information on the possible usefulness of such tags. Nine of these ducklings lost their tags within a week after tagging. The tags were placed on the fore edge of the patagium (Figure 59). The wing bones of the birds were held between the thumb and index finger of the left hand while the tag was pressed against these fingers and clamped into the skin. No wing bones were known to be damaged, but increased practice in application of the tags should reduce the number of lost tags. Because loss of tags cannot be entirely eliminated, the absence of tags in a marked population should not be taken to indicate foreign birds. Figure 59

Wing-tagged Duckling 275

The task of applying wing tags is somewhat more tedious than toe- clipping, and the tags are frequently lost by the birds, but the identification resulting from the tags is more certain than that re­ sulting from toe-clipping. The value of wing tags for marking nestling wood ducks was not thoroughly investigated in this study, but the work done indicates that such tags may be useful in certain local studies of the species. In Massachusetts, 387 ducklings were wing-tagged in 1952, and 123 were later captured (Anonymous, ion dated) .

Summary

Wood ducks, especially flightless local birds, were easily trapped for banding. A small portable trap was used. As many as 28 birds were taken at one catch in one trap, and a total of 554- birds were band­ ed from six of these traps in two years.

Forty incubating wood ducks were removed from their nests for band­ ing, and no desertions resulted from this alone.

The use of a dog in capturing flightless wood ducks for banding pur­ poses is not ordinarily feasible.

Either the adults or the young or both were color marked in 4-3 nests in 1955 and 1956. Twenty additional adults taken in bait traps were color marked in 1956, Down adhered and dried to airplane dope on two adult females and five ducklings and led to their identification in the field three to five weeks later. 276

Toe clipping of ducklings in the nest for individual identification when later trapped is not a reliable method, as too many toe nails are lost in other ways. Wing tagging is a satisfactory method for marking ducklings in the nest. CHAPTER XIV

CENSUS METHODS

A business-like administration of the wood duck resource is an

essential corollary with maximum yield on a long-time basis. During

a 10-year period, the total hunter kill of wood ducks will certainly

be largest if, within the limitations imposed by the breeding biology

of the species, the maximum number of birds are preserved to fly north

to nest each spring. A reliable method of measuring the size of the wood duck population available for hunters each year is paramount to

intelligent regulations. No known census method affords everything

to be desired, but some methods are better than others, and some of

their merits and demerits are considered in the following pages.

River Float Counts

The river float count is currently the most used method for measurement of trends in wood duck populations. Several states - notably Indiana, Ohio, and Tennessee - each annually cover more than a hundred miles of river transects. Data obtained by this method fur­ nish the chief basis for formulation of hunting regulations applicable to the wood duck. The method consists of two observers floating down rivers with a small boat and outboard motor and counting the ducks ob­ served. Effort is made to overlook no ducks and to avoid duplications.

Sexes are recorded when they can be determined, and during the nesting season the missing females are sometimes assumed when single or small groups of males are observed. The transects are usually covered three

277 278 times annually: in April for breeding pairs, in May for early broods, and in June for late broods. The trip is normally completed in the morning within four hours after sunrise.

In order to determine something of the reliability of these counts, forty separate cotints were made on an 11-mile transect of the Scioto

River in Ross County, Ohio, in 1956. Of the forty counts, ten were duplicate counts made by two observers at the same time; thirty were made by the same observer at different times. The thirty counts were made at different hours - morning, mid-day, and evening - between

April 11 and July 21, 1956. Based on a summary of all of these counts, plus nesting box checks, observations of marked broods, and trapping activities, twenty-one broods were estimated to have been produced in the sample transect in 1956.

A single count which exceeded the estimated breeding population was made on April 23. This may have resulted from counting errors, but there may also have been a segment of the population made up of birds later to nest along the river above and below the 11-mile transect or along tributary streams, or northward migration may still have been in progress. Two additional counts on this same date each totaled less than 50 per cent of the larger count. The totals of different counts from April 23 until mid-May varied from 4-5 to 61 per cent of the esti­ mated population of the transect. Most of the males moved out of the transect by the first of June. The proportion of the females observed after June 1 varied in different counts from about 10 to 48 per cent of the estimated population. The variation in results of this census method is such that the method does not indicate small changes in population 279

numbers. The least variation occurred in the counts of late April and

early May.

The largest proportion of the ducks present in an area are usually

found when the survey is made within four hours after sunrise. The

highest count in this study was made between 7:00 A.M. and 10:00 A.M.

E.S.T., but except for this single count all morning counts were close­

ly comparable with those made at other times of day. Disturbingly large variation was shown in the results of the morning counts. The least variation occurred in the evening counts, but the number of evening

counts is too small to support a convincing argument in favor of

evening counts.

In a study of this nature, where frequent trips are made with a motor boat through the study transect, there is a possibility that the behavior of the ducks is influenced by the noise of the motor and

presence of the boat. Flushing distances on the first trip through a transect were compared with flushing distances on the twentieth to twenty-fifth trips through another transect, and these distances were closely similar in the two cases. The flushing distances in 15 obser­ vations on the first trip through a transect varied from 10 to 150 yards, with an average of 61 yards, and in 20 observations on the twentieth to twenty-fifth trips the flushing distances varied from 20 to 175 yards, with an average of 54- yards. Prior use of the two transects by other persons in motor boats may not have been comparable.

These forty counts were made by one observer alone. River surveys are best made by two observers, one of which operates the outboard motor while the other watches for ducks. In order to test the efficiency of 280

one observer alone, five counts were made by two persons working to­

gether in the same boat, but ducks seen were not called to the attention

of the fellow observer until they had been passed. The counts of each

observer totaled approximately the same, but in each case 5 to 10 per

cent of the birds seen by one observer had not been seen by the

other.

Observations of marked adults revealed that the same birds some­

times continued down the river ahead of the observers as much as two

miles, and the same birds may, unknowingly, be counted repeatedly.

Some measurement was made of the extent to which different observers

agreed on whether observed birds were new to the count or duplicates.

Two observers together in a boat agreed on all birds being seen, but

recorded them secretly according to their individual judgment as to

whether or not they were duplicates. The independent decisions are

given in Table XXVI, Column 10, and the mutual counts agreed upon

through later conversation are given in Column 11. In 10 paired counts,

deviation amounted to as much as 7 birds or 2 U per cent, but the totals

for the 10 counts were the same. Greater variation could perhaps be

found between results of two observers if observers representing ex­

tremely liberal and extremely conservative characteristics were chosen

for comparison. No measurement was made of these personal characteristics of the observers whose data were above compared. The magnitude of the variation between different counts was nearly the same when two observers worked together as when one worked alone.

Adult wood ducks were sometimes seen sneaking through the brush along the shore as the boat passed, and they were so near to being over- Table XXVI Surveys of Vood Duck Population on an 11-mile Section of the Scioto River

Total Date No. Single Single Unsexed Broods No. Total adults Observer 1956 Time pairs males females ducks young young adult of two seen seen seen seen seen seen ducks observers A 4.-11 2:00 P.M.- 6:00 P.M. 14 1 1 5 15 A 4.-23 7:00 A.M.-10:00 A.M. 11 9 1 24 56 A 4.-23 10:00 A.M.-12:00 Noon 8 4 2 1 23 A 4-23 12:00 Noon- 3:00 P.M. 8 3 1 1 21

A 4-28 3:30 P.M.- 8:30 P.M. 7 4 2 9 29 B 4-28 3:30 P.M.— 8:30 P.M. 6 4 1 8 25 29 A 4-29 8:00 A.M.-10:00 A.M. 5 7 2 6 25 B 4-29 8:00 A.M.-10:00 A.M. 8 6 1 6 29 28

A 4-29 10:00 A.M.- 2:00 P.M. 5 3 1 6 20 B 4-29 10:00 A.M.- 2:00 P.M. 5 4 1 7 22 22 A 4-30 7:00 A.M.-10:30 A.M. 6 4 2 7 25 B 4-30 7:00 A.M.-10:30 A.M. 6 4 2 7 25 25

A 4-30 11:00 A.M.- 3:00 P.M. 5 3 0 6 19 B 4-30 11:00 A.M.- 3:00 P.M. 7 3 2 6 25 21 A 5-1 7:00 A.M.- 9:00 A.M. 2 10 2 12 28 B 5-1 7:00 A.M.- 9:00 A.M. 3 10 2 11 29 28

i A 5-11 3:00 P.M.- 6:00 P.M. ** 4 2 5 1 5 19 A 5-12 7:00 A.M.-10:00 A.M. 3 10 4 3 2 15 23 A 5-19 4:00 P.M.- 8:00 P.M. 4 3 2 1 2 2 14 B 5-19 4:00 P.M.-■ 8:00 P.M. 4 3 2 1 2 2 14 Table XXVI (Continued)

Total Date No. Single Single Unsexed Broods No. Total adults Observer 1956 Time pairs males females ducks young young adult of two seen seen seen seen seen seen ducks observers A 5-20 1:00 P.M.- 4:30 P.M. 6 2 7 3 1 7 24 B 5-20 1:00 P.M.- 4:30 P.M. 4 1 5 3 17 18 A 5-20 5:30 P.M.- 7:30 P.M. 9 6 1 1 1 25 B 5-20 5:30 P.M.— 7:30 P.M. 6 3 3 18 19

A 5-23 7:00 A.M.— 9:00 A.M. 1 10 2 1 2 14 A 5-23 9:00 A.M.—12:00 Noon 2 3 6 13 A 5-23 3:00 P.M.— 6:00 P.M. 2 5 9 A 6-1 5:30 A.M.— 7:30 A.M. 2 2 6

C* 6-1 7:00 A.M.— 9:00 A.M. 0 0 0 0 0 0 0 A 6—1 7:30 A.M.-11:30 A.M. 2 2 1 2 6 A 6-25 5:00 P.M.— 8:00 P.M. 4 6 37 4 A 6-26 6:00 A.M.— 8:00 A.M. 4 4 32 4

A 6-26 8:00 A.M.-12:00 Noon 3 3 2 12 6 A 6-30 5:00 A.M.— 8:00 A.M. 3 3 12 6 A 6-30 3:00 P.M.— 5:00 P.M. 9 6 28 9 A 7-9 6:00 A.M.- 8:00 A.M. 8 2 27 8

A 7-10 5:00 A.M.- 9:00 A.M. 3 6 24 3 A 7-10 9:00 A.M.—12:00 Noon 10 4 18 10 A 7-21 5:00 P.M.— 8:00 P.M. 4 2 13 4

# Count made by personnel of the Ohio Division of Wildlife. 283

looked that it does not seem far-fetched to assume that some actually

were thus overlooked. Also, there were doubtless some ducks which

flushed beyond curves in the river and flew behind hills or trees to

the river behind the observers and were thus not counted.

The practice of assuming the presence of a female on a nest when a

male is observed during the nesting season is probably justifiable, and,

if sexing is accurate, it increases the meaning of the count. The

practice, however, accentuates the importance of accurately sexing

the birds. If a single female is erroneously considered a male, the

error falsely increases the count by one bird. It was found to be

ordinarily impossible to satisfactorily sex all of the birds seen even when such experts as E.H. Dustman and M.B, Trautman occupied the ob­

server* s seat.

Opposing factors undoubtedly counterbalance each other to some

extent, and duplicate counts may on occasion closely compensate for birds which are overlooked. One count gave a total which was only seven ducks less than that revealed by intensive study of the area. If the different variables would operate in a somewhat constant fashion, river floats would probably yield reliable indications of population trends.

That the variables should operate in a constant fashion is, of course, an unrealistic hope, and river surveys, as now conducted, appear to be disturbingly insensitive measurements of population trends (Tables

XXVII to XXIX). Significant variation occurred at all hours of the day in April, May, and June. Least variation occurred in evening counts, but this is meaningless because of smallness of the samples. The money and effort now being spent on the three river float trips in April, May, 284

Table XXVII

Variation in Numbers of Wood Ducks Observed on an 11-mile Transect of the Scioto River in April

Time Number of Adults seen Date 2:00 P.M.- 6:00 P.M. 35 April 11 7:00 A.M.-10:00 A.M. 56 April 23 10:00 A.M.-12:00 Noon 23 April 23 12:00 Noon- 3:00 P.M. 21 April 23

3:30 P.M.- 8:30 P.M. 29 April 28 3:30 P.M.- 8:30 P.M. 25 April 28 8:00 A.M.-10:00 A.M. 25 April 29 8:00 A.M.-10:00 A.M. 29 April 29

10:00 A.M.- 2:00 P.M. 20 April 29 10:00 A.M.- 2:00 P.M. 22 April 29 7:00 A.M.-10:30 A.M. 25 April 30 7:00 A.M.-10:30 A.M. 25 April 30

11:00 A.M.- 3:00 P.M. 19 April 30 11:00 A.M.- 3:00 P.M. 25 April 30

7:00 A.M.-10:00 A.M. 10:00 A.M.-1 2:00 Noon 8:00 A.M.--10:00 A.M. 12:00 Noon- 3:00 P.M. 8:00 A.M.'-10:00 A.M. 10:00 A.M.- 2:00 P.M. 7:00 A.M.'-10:30 A.M. 10:00 A.M.— 2:00 P.M. 7:00 A.M.'-10:30 A.M. 11:00 A.M.- 3:00 P.M.

Observed 56 25 29 25 25 Observed 23 21 20 22 19 Expected 32 32 32 32 32 Expected 21.7 21.7

X2 = 22.875** X2 = 10.756**

3:30 P.M.*-8:30 P.M. 3:30 P.M.-8:30 P.M. Observed 29 25 Expected 27 27

X2 = 0.296

** Chi-square values significant at 5 per cent level are marked with two asterisks in this and two following tables. 285

Table XXVIII

Variation in Numbers of Wood Ducks Observed on an-11-mile Transect of the Scioto River in May-

Time Number of Young Seen Date

3:00 P.M.- 6:00 P.M. 5 May 11 7:00 A.M.—10:00 A.M. 15 May 12 11:00 A.M.- 3:00 P.M. 14 May 12 4:00 P.M.- 8:00 P.M. 2 May 19

4:00 P.M.- 8:00 P.M. 2 May 19 1:00 P.M.- 4:30 P.M. 7 May 20 5:30 P.M.- 7:30 P.M. 1 May 20 7:00 A.M.- 9:00 A.M. 2 May 23

7:00 A.M.-10:00 A.M. 7:00 A.M.- 9:00 A.M.

Observed 15 2 Expected 8.5 8.5

X2 = 9.94-1**

11:00 A.M.- 3:00 P.M. 3:00 P.M.- 6:00 P.M. 1:00 P.M.- 4:30 P.M. 4:00 P.M.- 8:00 P.M. 5:30 P.M.- 7:30 P.M.

Observed 14 7 Observed 5 4 - 1 Expected 10.5 10.5 Expected 3.3 3.3 3.3

X2 = 2.333 X2 = 2.627 2 86

Table XXIX

Variation in Numbers of Wood Ducks Observed on an 11-mile Transect of the Scioto River in June

Time Number of Young Seen Date

7:00 A.M.- 9:00 A.M. 0 June 1 7:30 A.M.-11:30 A.M. 2 June 1 5:00 P.M.- 8:00 P.M. 37 June 25 6:00 A.M.- 8:00 A.M. 32 June 26

8:00 A.M.-12:00 Noon 12 June 26 5:00 A.M.- 8:00 A.M. 12 June 30 3:00 P.M.- 5:00 P.M. 28 June 30

7:00 A.M.--9:00 A.M. 6:00 A.M.--8:00 A.M. 5:00 A.M.--8:00 A.M.

Observed O 32 12 Expected 14.7 14.7 14.7

X2 = 24.192**

7:30 A.M.-11:30 A.M. 5:00 F.M.-8:00 P.M. 8:00 A.M.-12:00 Noon 3:00 P.M.-5:00 P.M.

Observed 2 12 Observed 37 28 Expected 7 7 Expected 32.5 32.5

X2 = 7.142** X2 = 1.246 287 and June might be more wisely used if all three of the trips were made in late April or early May when there seems to be least variation in results. Some indication of the accuracy of each count would then be available, and little definite information would be lost by skipping the May and June trips. Mumford (1952:16) found that no correlation existed between pre-nesting and later brood counts and concluded that pre-nesting counts cannot be used to estimate production. Of the two methods - breeding pair counts and brood counts - I think the first is ordinarily much more representative of the actual population than the second.

Brood Counts

The variation in successive counts covering the same ponds or river transects suggests that brood surveys are not sufficiently sensitive to indicate true trends in wood duck populations. In an extreme case, variation amounting to 267 per cent occurred in two brood counts made

96 hours apart by separate pairs of observers under approximately simi­ lar conditions in regard to time of day, weather, water level stage, and proficiency of observers. One count contained only six broods, whereas the earlier count contained twenty-two broods.

The task of locating wood duck broods and of making brood counts is difficult and uncertain. The knowledge that a brood of wood ducks occupies a given area does not assure repeated and complete counts of ducklings in the brood. On repeated occasions color-marked wood duck broods have been observed leaving nesting boxes on ponds where they were not seen again in daily visits and hours of concealed observation through 288

the following two weeks. Then observations were made which suggested

that the broods had been on or near the ponds through the intervening

two weeks. During 1955, Kenneth Russell was employed by the Ohio Divi­

sion of Wildlife to make brood counts of ducks at the Olentangy Station.

I also made similar observations at the same time and place, though

independently. I made approximately 100 visits and Russell made

approximately 50 visits to all of the 43 ponds. There were 11 broods

seen by both observers; 11 broods seen by Russell were not seen by me;

15 broods I saw were not seen by Russell. It must be recognized that

these counts were subject to the many inaccuracies inherent in brood

counts.

In making brood inventories of ponds, the usual method was to

drive in an automobile to within approximately a hundred feet of the

pond and to walk the remaining distance. At some ponds where a com­

plete view of the pond was available from the road the car was driven

all of the way to the pond. The best brood counts were usually made

immediately on arrival at the ponds when the ducklings were observed before they escaped to cover. Walking around the ponds sometimes dis­ closed ducklings not otherwise detected, but it was frequently dubious as to whether or not the entire brood was thus found or whether the duck­ lings found represented only one brood. An observation blind was used to a limited degree to determine the brood composition of ponds, but trapping and banding proved to be a more feasible supplement to pond visits for use in brood inventories.

Brood inventories on Buckeye Lake were very difficult because of the presence of extensive beds of spatter dock which furnished excellent 289 feeding and retreat cover for wood duck broods. The best brood counts were usually made by observing the ducklings crossing open pools among the emergent vegetation while I waited or drove the small boat through the concealing cover.

Brood inventories of non-navigable creeks were made by walking in the creeks. Most of the flood plains adjacent to creeks contained an abundance of plant growth such as poison ivy and nettle thus making walking hazardous. Also, it was usually impossible to watch properly the creek for ducks when walking among the vegetation on the flood plains or creek banks. An effort was made to conduct the creek inven­ tories during periods of recently reduced flow of water, and the difficulty of walking the creek transects was thus held at a minimum.

The recently fallen water level was also responsible for leaving ex­ posed many soft, muddy areas where tracks of ducks could be readily found. In a 10-mile creek transect only one brood of ducks was seen, but the locations of five other broods were discovered by finding tracks in the mud. Counts were made, along with additional observations, in a subsequent return to these brood territories.

Brood inventories on the navigable Scioto River were by far the most easily and most effectively made of those in the various wood duck habitats. This was partly because of the ease in extensive coverage of above average wood duck habitat, and fifty miles or nearly five trips through an 11-mile transect were traveled during one day of most weeks in the brood season. This exposed the observer to the chance of making approximately one hundred brood counts in a day. Usually, however, less than one-fourth of the broods were observed in any one day. 290

Since as many as six broods sometimes gathered in the same area,

and broods mixed to some extent, the difficulty in making brood counts

was intensified by brood concentrations. Sometimes mixing of broods

was revealed by differences in duckling sizes and the one or two birds

foreign to the brood could be discounted. There was no assurance,

however, that a decline in the number in a brood was not the result of

loss to another brood occupying the same area, and mixing of broods

restricted the reliability of brood inventories where wood duck popu­

lations existed as brood concentrations, and not merely as isolated

broods.

Weather is a variable which may have an important effect on the

results of brood counts. In three trips over an 11-mile river transect

on June 1, 1956, no broods were seen at a time when ten broods less

than three weeks of age were known to have been present. This was a

cool day, and the minimum temperature was 8.3°C. The rather young

broods were presumably out of the water being brooded. A time should

be chosen for brood inventories when the temperature is not too cool, or not below about 13°C.

It was early found impossible to be sure entire broods were seen in every case, especially when the broods were first spotted near cover.

Small numbers of ducklings were sometimes seen under conditions suggest­ ing that the entire brood had been noted, but further observations re­ vealed additional ducklings. If the broods were detected before the fe­ male flew from them, she sometimes led them from concealment where they could be counted after waiting a half hour or less. 291

Wood duck broods frequently seek cover long before the observer is near enough to see the ducklings. In order to get some notion of the distance ahead of the observer at which ducklings seek cover, baiting stations visible at a considerable distance were maintained along the

Scioto River. These stations were watched with binoculars as they were being approached, and it was found that the ducklings disappeared from view before being readily visible to the unaided eye about one-third of the times approached. The disappearance distances in yards were as follows: 10, 15, 15, 25, 25, 30, 35, 35, 75, 200, 200, 275, 275. These observations were made with the three horsepower outboard motor running at medium speed.

A workable method for aging young in the field is essential in field studies of duck broods involving survival of young to different ages. Through observation of captive-reared ducks, Southwick (1953) developed a system of aging eight species of ducks, not including wood ducks, by classifications of plumage development. Through field studies,

Dreis (1954) developed a system of aging pre-flight wood ducks by obser­ vation of relative body size and behavior. A combination of both methods was used in this study.

As suggested by Dreis, comparative size is the best single criterion for aging pre-flight wood ducks in the field. After rearing wood ducks in captivity, handling marked wild birds of known age, and observing birds of known age in the field, one develops an ability for aging in the field which cannot be adequately conveyed to others and which is ac­ quired only by experience. This imparts an unfortunate element of sub­ jectiveness to the data. Nevertheless, aging of the young is an essential 292

part of brood counts.

The number of broods and the numbers of young in each brood

occupying limited areas cannot always be readily determined by field

observations alone. When field observations were supplemented with

trapping and banding activities, a much more complete analysis of

brood populations was possible. On repeated occasions one, two, and

three additional broods were found occupying limited areas thought, on

the basis of field observations, to contain only one or two broods.

Likewise, field counts of the numbers of young in given broods was

modified upward on numerous occasions by trapping and banding activi­

ties. Only one field count of a brood was modified downward, and this was a change of only one bird.

Bridge Stops

Chance observations from a bridge of a brood of wood ducks suggested that a systematic program of bridge stops might be a convenient method for making wood duck counts. In order to get some indication of the

potential value of bridge stops for making such counts, I made 88 different bridge stops in central Ohio during June and July, 1956.

These bridge stops were made between 5*00 A.M. and 8:00 A.M. E.S.T., and a wait at each stop lasted five minutes. Two separate broods were observed. Through August and September, 156 additional bridge visits were made' in order to determine the effectiveness of this method for following wood duck populations after the ability to fly has been attain­ ed. These visits were less systematic than the earlier bridge stops and were made incidental to other activities. There was no restriction as 293

to the time of day of the visits, and in many cases the streams were

examined for ducks when the automobile containing the observer or

observers was slowly moving. A flock of 20 ducks was seen during eight

of ten visits to one site, but no ducks were seen in any of the other

stops (Figure 60).

Obviously, the effectiveness of bridge stops for finding wood ducks

can be expected to be restricted by the distance of visibility up and

down the streams from the bridges. Most creeks and rivers of central

Ohio are so meandering that the views in each direction of the streams

are very limited. About 80 per cent of the views up or down streams

from bridges in central Ohio were restricted to 50 yards or less. As

a result of the restricted visibility, a relatively short section of

wood duck habitat can be scrutinized on each bridge stop. The effective­ ness of the method is correspondingly restricted.

When wood ducks occur, however, in a stream within the field of

vision open from the bridge, this method of observation presents some

special advantages for making observations as compared with the method

in which the observer floats down a stream. Because the ducks are al­

ready somewhat accustomed to human activity as it occurs on modern high­ ways, they can be readily observed from the bridges. An automobile is usually a more convenient means of visiting wood duck habitats than a

boat, as the automobile is the means with which the observer is most likely to leave his headquarters, while the boat requires special trans= portation to the water area. It is usually more convenient to make fre­ quent visits to wood duck habitat located near a highway bridge than to make equally frequent visits to a site accessible only by boat. When a Figure 60

A Flock of Twenty Wood Ducks Remained on This Section of the

Hocking River Through the Summer and Fall of 1956 295

departure date for wood ducks from a given habitat is desired, it can

sometimes be most easily obtained by regular visits to a bridge where

the ducks can be repeatedly observed.

Aerial Survey

Flights in a Cessna and a Cub airplane were made on October 13, 1956

over the Scioto River in Ross and Pickaway Counties and over Jonathan

Creek and the South Fork of the Licking River in Licking and Muskingum

Counties of Ohio. The flying height was between 150 and J + 0 0 feet. Several times when swimming flocks of ducks were observed on the Scioto River, the altitude was reduced to about 150 feet in an unsuccessful effort to flush the ducks. Only three ducks were seen in the 30-mile flight over the

Licking River and Jonathon Creek. These are relatively small, tree- lined streams, and it is thought, on the basis of experience on the

Scioto River, that wood ducks were overlooked because they were concealed by the trees and because they failed to flush. The Jonathon Creek-

Licking River area was not checked from the water or ground, but judging from banding recoveries in 1955, it can be assumed to be a fall concen­ tration area for wood ducks. The census flight over the Scioto River was made between 3:00 P.M. and 4-500 P.M. E.S.T., and the same area was covered twice within an hour. One hundred and twenty wood ducks were counted on the first survey and 112 were counted on the second. On Octo­

ber 1 4 , & boat trip was made over the same transect which had been cover­ ed on the previous day in the aerial survey, and 133 wood duck were counted. The difference was largely explained by the observation from the boat of 26 birds beneath trees along the edges of the river. Most 296

of the birds did not flush when the airplane passed over them.

When results of the two aerial counts on the Scioto River are

compared with results of one river float count, the aerial survey is

shown to yield more conservative results. The sample is too small to

justify a convincing conclusion, but the probability is indicated that

the fall population index yielded by the aerial survey, is as reliable

as that yielded by the river float method. Absolute density cannot be

derived by either method. The airplane is useful in wood duck surveys

only in situations where the ducks can be seen from the air without

frightening them into flight.

Feeding Flight Counts

The feeding sites of local late summer, fall, and winter populations of wood ducks were found by observing the directions of flights of small groups of ducks seen in the late afternoon. After a few small groups of ducks had been observed flying in the same general direction during late afternoon, continued movement in the direction of the flight usually brought the observer to their feeding site. The ducks usually went to their feeding sites as individuals or small groups, but sometimes all or most of the local population of wood ducks went to the feeding site in one group, and in such cases they went in the late evening a short time before sunset. When all of the wood ducks went together to the feeding site, the gathering place of the flock, rather than the feeding site, was found by noting the direction of flights of small groups of ducks in the late afternoon. The feeding flights were then observed as they left the gathering points. After the feeding sites were located, 297

positions were taken nearby where ducks could be seen coming froa all

possible directions. When it was possible to keep surveilance over all

compass directions, an accurate count was made of the numbers of ducks

coming to feed.

A baiting and trapping station was maintained for wood ducks

through August, September, and October 1955 and 1956 at Buckeye Lake,

Ohio. This was on an exposed mud bar in Honey Creek, and the birds

could be watched from a boat as they came to the area or as they swam

about after they were at the feeding station. Accurate counts were

made of the birds coming to the area on many different days. Here an

abundance of food was available, and it was believed that not all of the

birds feeding at the station visited it twice daily. Trapped birds

well-filled with corn were kept in captivity and after 12 hours their

crops still contained an abundance of corn. Twice daily feeding would

clearly by superfluous. The usual count for a single flight was only

about half of the total number of ducks shown by banding to be feeding

at the station. Counts of ducks in feeding flights at Buckeye Lake did

not yield a satisfactory estimate of the total population in the area.

It was not possible to see in all directions from some feeding

places because of obstructing woods, and the accuracy of final results

became somewhat questionable at such feeding sites. This was the situa­

tion at the Savannah Refuge January 2-9, 1957. Here more than two thou­

sand wood ducks were feeding in two areas at separate points in a long

strip of woods. Because of marshy and swampy conditions and an inter­ vening woods, no point could be found where the movements of all ducks

could be simultaneously observed as they approached the feeding site. 298

Neither of the two fire observation towers present on the refuge were

favorably situated for use in watching these flights, but observations

from favorably located towers should produce highly accurate counts of

the population.

In a vast area of wood duck habitat such as occurs in the Savannah

Refuge, it could not be definitely known that all of the ducks were

entering into each evening feeding flight. An observation blind was

erected near a spot where some 50 wood ducks regularly spent the day, and another strip of wood duck habitat occupied through the day by

some 150 ducks was checked by walking along and watching from a nearby dike. All of the wood ducks in this sample of some 200 birds were known to have participated in the feeding flight. At the time these observations were made, most of the birds were feeding on acorns, the available supply of which was rather depleted, and twice daily feedings by all ducks was probable.

Along the Scioto River in Ross and Pickaway Counties, Ohio, the fall feeding sites were found by first locating concentrations of wood ducks and then observing the evening flights from an exposed vantage point. The two feeding places found were in farmers* hog-feeding lots.

Here the ducks went en masse to the feeding sites.

Feeding flights of wood ducks occur in the morning as well as in the evening, but the morning flights cannot readily be observed because they occur too early for the ducks to be seen.

Counts of ducks in feeding flights appear to be useful in restrict­ ed populations such as on refuges, but each area has its peculiar problems.

The special problems in each situation must be analyzed, and counts of 299 feeding flights are not a suitable method for general application.

After several years of determining the constancy of use of feeding places, the construction of observation towers might prove feasible where needed on refuges.

Roosting Flight Counts

During September and October, 1955 and 1956, in Ohio, and during

January, 1957, at the Savannah Refuge, all of the wood ducks present within a radius of several miles gathered to roost in small areas. Such roosting places were found by noting the directions of the ducks' move­ ments as they left their feeding places within about a half hour before sunset. By advancing in the direction of observed flights, eventually a position was reached where flights merged at a central point. The exact roosting site was then easily found and the ducks counted as they came into the roosting place. I stationed myself in a place where I could look over the roosting area into the maximum of open sky, as some of the late-coming ducks would be missed if they entered against a tree background. Wood duck roosting flights observed in Ohio extend­ ed through a half hour or less and contained a total of less than 200 birds. Since the ducks came into the roosting places as single birds or groups of less than 20, the entire flight could be counted with a high degree of accuracy. The count was best made of the evening flight to the roosting place rather than of the morning flight from the roost­ ing place, as the birds left before light intensity had increased suffi­ ciently to permit their being seen. 300

Roosting counts are especially useful where a habitat island, such

as an isolated lake, is being worked for wood ducks. In such a habitat,

wood duck utilization and production in a well-defined sample can be

followed each year. In this situation excellent data can be obtained with a minimum of time and effort. Roosting flight counts are believed

to be the best means for obtaining estimates of wood duck populations

in restricted areas such as wildlife refuges. Late summer movements of wood ducks destroy the year to year comparative value of roosting

counts in non-restricted river habitats.

Nesting Box Check

Results of the present study suggest that counts of wood ducks using nesting boxes is the most accurate available method for reveal­ ing trends in wood duck populations. If each box is properly prepared for nesting wood ducks in advance of the nesting seasons, there should be a minimum of variables affecting the rate of nesting box utilization from one year to the next. The removal and restoration of natural nest­ ing places may be an important long-time variable. Certainly, in many past years of American history, decrease of natural wood duck nesting places proceeded more rapidly than their restoration. The time when these opposing trends are nearly in balance may be approaching, and the increase or decrease of natural nesting places should hold negligible importance in short-time comparisons.

The same wood duck sometimes returns through successive years to nest in the same cavity, and as the population is replaced with younger birds, there may be a gradual shift toward increased use of artificial 301 nesting places. It is not at present known whether there is any tendency for first-year breeding wood ducks to select a nesting cavity with bias toward or against their natal cavities. It is entirely possible that, with the decline of available nesting cavities and the increase of artificial nesting places, the nesting habits of wood ducks may be modified to the extent that eventually only nesting boxes will be utilized. This would closely parallel the situation with the purple martin in parts of its present range. This shift from natural to arti­ ficial nesting places should be a relatively unimportant variable in comparison of nesting box utilization in two successive years. Compari­ son of populations dn two periods separated by an interval of ten, fifty, or a hundred years might be an entirely different matter.

Preparation of the nesting boxes each year slightly in advance of the wood duck nesting season is an essential part of this method of population measurement. If the degree of utilization of nesting boxes is to be compared in successive years, it is imperative that an effort be made to make all boxes equally available to wood ducks. The boxes should be checked for needed repairs immediately before the onset of the wood duck nesting season. In the latitude of Ohio, effort should be made to make needed repairs during the last half of February.

There is some labor involved in keeping nesting boxes in a state of readiness for wood ducks, and because of nesting box competitors, the boxes sometimes retain their state of readiness for only a short period of time. Therefore, the boxes can be given best care if one person is responsible for only a small number of boxes. Also, the nesting box check will give the most comprehensive appraisal of the status of the 302

wood duck population if the boxes are extensively scattered through a

wide area and in all available duck habitats. For these two reasons,

it seems best to assign the care of a small number of boxes to each of

a large number of persons. If each county game protector in areas con­

taining wood duck habitat were made responsible for five or ten boxes,

individual boxes would be given excellent attention, and the resulting

sample should give a good picture of trends in wood duck populations.

With such a large number of participants involved, it can be expected

that some would lack interest in the project and that some administra­

tive checking and pressure would be needed to assure uniformity in

attention to the boxes.

Fluctuation in use of the boxes by nesting box competitors such as

starlings, screech owls, raccoons, and squirrels may be an important variable inherent in nesting box checks as a method of wood duck popu­

lation measurement. This problem may be best solved as at the Olentangy

Station where an ample number of boxes were provided for nesting box

competitors as well as for ducks.

Something should be said against the practice of counting known unsuccessful nests along with successful nests. If production is the ultimate information desired, seriously biased totals may result from failure to segregate unsuccessful nests. If, for instance, a pair of ducks has made three nesting attempts all of which were prematurely terminated, the one pair has contributed three nests to the count but no ducks to the fall population. This count is inversely related to nesting success and production. If the nesting boxes were checked during

June and only successful and active nests were counted, the resulting 303

data should ^ive a reasonably accurate appraisal of the wood duck

breeding population.

If areas of concentrated wood duck management, such as the Olentangy

Station or Magee Marsh, are used for sample units, some control of

variables can be maintained. The task of keeping the boxes in repair

and of checking them for utilization can then be handled 1j y one person

who is thoroughly familiar with the unique problems of the project. It

cannot now be definitely stated that wood ducks utilize such intensive

management units in exact proportion to the over-all population, and

disproportionate concentration of nesting birds is possible. McLaugh­

lin and Grice (1952:24-7) reported increased usage of boxes each of 3

successive years: 33 per cent usage was recorded the first year, 55

per cent the second year, and 66 per cent the third year. Results of

checking scattered nesting boxes should be available for comparisons with results from intensively managed units.

Boxes should be protected from raccoons as much as possible.

Boxes mounted on posts over water are usually somewhat protected.

Since nesting box checks are only a partial indication of a sea­

son's production, conclusions derived from nesting box checks should be reinforced by late summer roosting flight counts.

S um m a ry

The relative merits of seven methods of censusing wood ducks are considered. The river float count, the method now most used, is con­ sidered not sufficiently sensitive to show anything but major changes in population numbers, the way it is currently conducted with one trip 304

in each of April, May, and June. If the three trips were all made

within a short period in late April or early May, each count would be

accompanied with a measurement of variation.

The variation in brood counts is such that these counts, unless of

an intensive nature, give no indication of trends in wood duck popula­

tions .

Bridge stops are not useful for a general census because too few

ducks are ordinarily found within view of bridges. When flocks are

found within view of bridges, bridge stops are a convenient method for

making observations incidental to censusing.

Aerial surveys are useful only on large rivers where the fall

population of ducks can be seen on the water. Aerial surveys yield a more conservative estimate than river floats because scattered ducks hidden by trees are overlooked from the air.

Counts of ducks in feeding and roosting flights in combination

or in roosting flights alone are useful for following isolated popu­

lations. These counts are very accurate under ideal conditions, but

problems peculiar to individual areas limit their usefulness for general application.

Counts in June of successful or occupied nesting boxes provides an accurate method that can be generally applied for following trends of numbers in wood duck populations. CHAPTER XV

LEGAL PROTECTION FOR THE MOOD DUCK

It is a characteristic of most hunters to clamor for increased

liberalization of hunting privileges. With mpdern guns and transpor­ tation, the vast army of hunters who take to the field in the United

States could exterminate the continental duck population in a very short period of unrestricted hunting. Too often hunters fail to recognize that continued hunting is possible only through restricted hunting.

There is a record in Illinois (Gigstead, 1933:607) of 700 wood ducks having been illegally taken and sold for $3.00 each.

I was in the field a great deal in the course of this study, associating with hunters, and because nothing was done about law viola­ tions, gained hunters1 confidence and learned of their practices.

Shooting wood ducks before the season opened, shooting the protected wood duck, shooting ducks over baited areas, shooting beyond legal hours, exceeding the bag limit - all of these offenses were observed or reported to me from one to many times.

Benefit of a Selective Closed Season

During 1956 when wood ducks were not legal game in Ohio, Delmar

E. Handley (personal correspondence) reported that Ohio Division of Wild­ life personnel spent 37 hours observing 25 hunters who had 33 chances to shoot wood ducks, firing 25 times and knocking down 10 wood ducks. With the continued increase in the number of hunters and pressure on game, the inevitable solution may be increased license fees to support better

305 306 law enforcement.

The protection of the wood duck in Ohio during a selective closed season is to a large extent a matter of apprehending willful law vio­ lators. This was particularly true in most of Ohio during the closed season of 1956 because the wood duck was the only duck in wood duck habitat through most of the period when it was in Ohio during the shoot­ ing season. It was simply a matter of knowing the single species of duck which was being seen.

The inability of many hunters to identify wood ducks in the field before shooting and disrespect for the law by some hunters results in many wood ducks being illegally shot in a selective season. The re­ markable recovery that this species made from near extinction to its present state of abundance clearly attests to the worthwhileness of a selective closed season, although legal protection may be only one of the factors responsible for this increase.

What is really needed is increasingly appreciative and informed hunters. A 56-page Waterfowl Identification Guide with illustrations of the Peterson Field Guide type was first issued by the U.S. Fish and

Wildlife Service in 1956. The Atlantic Waterfowl Council and various state game commissions have had copies of this guide printed and distri­ buted. Other states have issued leaflets containing guides to waterfowl identification. Information on identification of waterfowl is readily accessible to hunters; only the stimulus to use this information by the rank and file of hunters needs to be provided by increased apprehension and newspaper reports of prosecutions in each state. 307

The case of the hunter's place in wood duck management would not be complete without a word of tribute to the many excellent sportsmen scattered over the country. There are those whose enjoyment in the hunt is measured not alone by the number of ducks bagged but by the number seen wild and free. Because many hunters love their sport beyond the mere bagging of a duck, they can and do learn to identify their game. Numerous hunters thoroughly able to identify wood ducks in the field were encountered. Some learn this ability because of interest; some prefer to claim that it cannot be learned.

Protecting Local Northern Populations From Over Shooting

Bv a Later Opening Season

Especially on river habitats, disproportionately heavy shooting of local populations may sometimes occur before dispersal of the local population has taken place. This condition was observed among prairie ducks by Hochbaum (194-7:55) and was termed "burning out” duck popula­ tions. The local problem can be solved with a later opening season.

A season opening as late as November 1 would be unfair to northern wood duck hunters as they would have little or no opportunity to harvest the wood ducks their states had produced. Many of the wood ducks in central Ohio migrate southward soon after the first of November. In central Ohio wood duck populations where trapping and banding were con­ ducted, except on the Scioto River, shuffling of the local populations had taken place by the end of the first week of October. On the Scioto

River it occurred a few days after the opening of the hunting season. 308

A later opening season is very effective in safeguarding a local

northern population of wood ducks. Hawkins .gt al. (1939: 130, 186)

pointed out that in Illinois the illegal kill of wood ducks, which was

greater in 1938 than the legal kill of blue-winged teals, might have been greatly reduced by delaying the opening date of hunting season by one week.

The Place of Refuges in Wood Duck Management

Duck refuges at present operate chiefly in providing places of re­ treat during the winter months. The Savannah National Wildlife Refuge in southeastern South Carolina annually supports one of the largest concentrations of wintering wood ducks in the United States.

The writer spent the period of January 2-9, 1957, making observa­ tions on wood ducks in the Savannah Refuge. On November 30, 1956, in a complete inventory of all ducks in the refuge tjy U.S. Fish and Wild­ life Service personnel, the wood duck population had been estimated at approximately eight thousand, but a sharp decline in numbers had been noted before the writer*s arrival on January 2, 1957. The writer estimated a maximum wood duck populatinn of twenty-five hundred birds on January 8, Some of this discrepancy in estimates of the total wood duck population in the refuge possibly derived from differences in personal opinions of the observers, but there had certainly been a marked decline in the wood duck population of the refuge during the month of December,

During the period of January 2-9, 1957, the wood duck feeding and roosting flights were observed several evenings, and it was thought that 309 nearly all of the wood ducks of the refuge were feeding on acorns outside

its boundaries. Feeding flights of wood ducks were made well within

legal shooting hours, and a great deal of shooting occurred almost every

evening at the feeding site. It can be expected that shooting was effec­

tive and the population was thus significantly reduced. It is unthink­ able, of course, that the total reduction taking place during the month of December could have been caused by killing ducks at this one feeding

site, and it is postulated that some of the birds moved elsewhere, per­ haps as a result of shooting at the feeding site. Wood ducks in Ohio commonly move from areas because of shooting.

The chief function of duck refuges should be to preserve some breeding stock for the following nesting season. Because of the habitat requirements of the wood duck, a smaller proportion of the total popula­ tion of wood ducks than of other species of ducks may find retreat in refuges. .Since many of the wintering wood ducks are scattered through the southern United States as relatively small groups and are available to hunters, a refuge would best serve its purpose by safeguarding the wood ducks within its boundaries, rather than by feeding them into the hunters* bags through the shooting season. The refuge for wood ducks should contain feeding, roosting, and loafing facilities equal to the needs of the number to be conserved by the refuge.

Legal Shooting Hours and Baiting

Shooting ducks over baited areas has been illegal since 1930, but it is such an effective method for bagging the legal limit or more that hunters continue to clamor for a modification of the law or to shoot 310

over baited areas in spite of the law. The evening feeding flights

of wood ducks occur well within the legal shooting time, and this

species is particularly vulnerable to baiting. The law leaves an

important loophole in that shooting over feeding areas is legal if feed

was not placed for the purpose of attracting ducks to be shot. Shooting

wood ducks as they come to feed under an oak tree is quite as destruc­

tive and unsportsman-like as shooting them over an area baited with

corn.

Hunters are sometimes successful in concealing bait stations where

ducks have learned to feed in advance of the hunting season. By scatter­

ing shelled corn in shallow water each evening only to the amount con­

sumed at one feeding period, regular feeding flights can be maintained

with little physical evidence of baiting. Observation of feeding

flights of wood ducks provides an excel].ent tool for locating such bait­

ing stations. Movement toward the baiting station by the observer can

be commenced as soon as several small groups of wood ducks have been

observed flying in one direction or toward a central point in the late

afternoon.

Bag Limit

Because of the relatively precarious status of the wood duck popu­ lation in recent years, shooting of this species has been permitted only on a very limited basis. A daily bag limit of one or two wood ducks per hunter is the chief restriction that goes beyond the restrictions applied to other species of game ducks. Hunters frequently object to this re­ striction with the argument that they occasionally kill more than pne or 311

two birds with a single shot, and if hunting alone, the ducks shot in

addition to the legal limit must be left to waste. There is no reason­

able limit to the extent to which this argument could be used, as one

hunter reported shooting into a flock and killing four and crippling

two wood ducks at one shot. The logical method to avoid such waste

seems to be for the hunter to shoot more discriminately.

Summary

Many wood ducks are ignorantly shot during a selective closed sea­

son, and mai^r more are shot by flagrant law violators. The solution

is apprehension of violators and increased publicity to prosecutions.

An open season on wood ducks in the latitude of Ohio starting in

the last few days of October or the first of November would be effective

in protecting northern breeding populations, but it would be unfair to

hunters in that latitude because most of the ducks migrate southward

about November 1.

Feeding flights of -wood ducks occur well within legal shooting hours, and shooting over natural feeding places can be as destructive as shooting over baited areas. The feeding flight normally starts about one and a half hours before sunset.

To best serve its objective of conserving some breeding stock through the shooting season, a winter refuge should contain roosting, feeding, and loafing cover.

Though hunters object to a bag limit of one wood duck per day because they sometimes shoot more than one duck at a shot, the logical solution

seems to be more discriminate shooting. CHAPTER XVI

ESTABLISHING A LOCAL WOOD DUCK POPULATION

Rearing and Releasing

Artificial restocking of wood ducks was initiated as early as 1924

as a means of restoring this species to unoccupied habitat. Three thou­

sand wood ducks were reared and released in Connecticut during the period

1924.-1936 (Frank, 1948! 128). Artificial propagation of wood ducks ap­

pears to be a feasible method for obtaining birds to stock unoccupied

habitat. In 1944* 97 wood ducks were reared to 5-6 weeks of age in

Illinois by the Illinois Natural History Surveyj they were then shipped

to Madison, Wisconsin. After their arrival at Madison, they were held

in captivity about 10 days before being released into an area which was

for 50 years devoid of breeding wood ducks. In the same year, a ship­ ment of 95 Illinois-reared wood ducks was sent to the Jasper-Pulaski

State Game Preserve in Indiana. At least five females returned and nested the following year at Madison, and four females returned and nested in 1945 in Jasper-Pulaski Preserve.

The ducklings used in this experiment were hatched in an incubator from eggs gathered from nests of wild birds (Hanson, 1951). A tempera­ ture of 99°F. was maintained throughout the incubation period. The relative humidity was maintained at 80 per cent or above. A hatching success of 63.6 per cent was obtained from 467 eggs. The ducklings were brooded at a temperature of 90°F. and housed in a pen containing a one- half inch mesh wire floor. Of the 297 ducklings hatched, 242 were reared

312 313

to at least five to six weeks of age.

The ducklings were first fed on chicken eggs which had been

boiled 15 minutes. A 150-watt electric light bulb inside a reflector was kept near the floor to attract insects. The ducklings ate many

insects, and this was thought to have contributed to the health of the ducklings. Commercial turkey starting mash in pellet form was the only food given. Approximately 3.1 pounds of mash were consumed in attaining a body weight of one pound.

Some workers have experienced difficulty in inducing the ducklings to start eating. C.E. Knoder (personal conversation) advised placing the newly hatched wood ducks with mallard ducklings, as the mallards start eating more readily, and the wood ducks will start eating when placed with other ducklings which are eating. Hunt (1956:10) advised placing the newly hatched wood ducks with older ducklings which have already started feeding.

Hanson (1951) reported that the paratyphoid organism, Salmonella tynhimurium. was found to be the chief cause of mortality of wood ducks

reared in captivity. In a rearing experiment in 194-1 f 400 of 500 duck­ lings died, and in 1952 377 of 445 died. Most of the ducklings died within the first five days after hatching.

Wilson (1854-i 600) stated that the wood duck is often kept in

European menageries and breeds readily in captivity, but numerous workers have experienced difficulty in rearing this species. Yeagley

(1953:160) considered it necessary to drop the ducklings into water at about 24- hours of age, thus simulating natural conditions. According to

Laidlay (1940:159) grit is important for all young waterfowl. Hunt (1956: 3 H

8) considered sand necessary for ducklings and mixed 10 per cent by

weight of clean white sand with the feed which consisted of poultry

starting mash.

Extensive Versus Intensive Management in Wood Duck Production

At the Olentangy Station there are more than 4 0 ponds scattered

over a land area of about 6,000 acres. All of these ponds are now

provided with one to three wood duck nesting boxes. During 1956, only

two of thirteen nests were alone on ponds where another nesting box was available. Extremely high densities of nesting wood ducks can be

attained, as has been demonstrated in Massachusetts where 94- nesting

boxes in a 3-mile section of river produced 956 ducklings. The non­

territorial trait of the wood duck renders this species peculiarly

susceptible to intensive management. It apparently nests by preference near another nest of the same species.

At the Olentangy Station all of the wood duck broods from nesting boxes, as well as six additional broods, gathered on five ponds. If

these five ponds were given over to wood duck production exclusively,, production of optimum conditions for wood ducks would be a feasible goal. With a large number of nesting boxes in a small area, the boxes can be given maximum attention at minimum cost. In addition, the var­ ious factors of brood decimination can be controlled and maximum survival of ducklings attained. If an area, such as a pond or small lake, is managed solely or chiefly for wood duck production, control of locally injurious plants, bacteria, insects, amphibians, fish, reptiles, birds, and mammals may be feasible. Brood survival could be greatly improved as compared with an unmanaged habitat where the ducklings run a formidable 315

gamut of hungry mouths and decimating factors.

Past and current programs of wood duck management in most states

contain phases of both extensive and intensive management, with chief

stress on the extensive method. From work that has already been done

in several states, it appears that intensive methods should receive in­

creased emphasis. New and unforeseen problems would undoubtedly arise

in such a program of intensive management. It can be expected, for

instance, that providing food artifically would eventually become a

necessity. Maintaining an adequate breeding stock in the face of local hunting pressure might also be a problem. If depletion of the breeding

stock resulted from local shooting before the birds had become mixed with the general population, the problem might be solved by making

the production area a refuge containing roosting, feeding, and loafing habitat. Special problems would need to be detected and solved as they would appear.

Human Interference in Attracting Wood Ducks

The exact amount of human interference wood ducks will tolerate and yet utilize areas for nesting, brood rearing, feeding, and roosting may be variable. The limit of tolerance doubtless varies to some extent among individuals, but in general the wood duck is not averse to human propinquity.

In seeking nesting places, female wood ducks frequently enter chimneys of occupied human dwellings. Successful nesting frequently occurs over places of concentrated human activity such as busy fishing docks and city streets. If a human dwelling stands adjacent to 316

occupied wood duck habitat, the chances are good for occupancy of wood

duck nesting boxes properly placed in trees on the lawn. Dixon (1924:

42) reported that 10 of 12 nests observed were near inhabited farm

houses or along roads. Gigstead (1938:605) found four occupied nests

in the courthouse yard of a village.

Musselman (1948) suggested that the current tendency of wood ducks

to nest in parks, cemeteries, and other situations near human habita­

tions is a recent adaptation associated with an increase in the wood duck

population. Wood ducks nested near human habitations before the wood

duck population decline of the late nineteenth and early twentieth

centuries. Authors writing about the wood duck from the time of Audubon

have commented on the fact that these birds nest near human habitations.

Even with few nesting wood ducks near human dwellings such nesting may

formerly have occurred commensurate with the degree that there was a

breeding population present. With the increase of human dwellings, there

is an increased probability of nests being near dwellings.

Feeding flights often bring wood ducks into close proximity with

man. Wood ducks in central Ohio regularly feed in farmers* hog lots

during the summer and early fall, and such feeding places usually occur

within relatively short distances of the farm buildings. The ducks

usually feed near but not actually among the hogs.

At the Olentangy Station in Ohio during 1955, nesting boxes were

present on all of the 43 ponds, but fishing was prevalent on only part

of the ponds. No correlation was found between use of ponds by fisher­ men and wood ducks. A lower brood survival rate could be expected due to frequent dispersal of the broods and loss of young by their becoming 317

entangled in filamentous algae. Unusually heavy mortality should not

occur on ponds containing a large area of emergent vegetation.

It was noted that persons most successful in attracting wood ducks were persons radical about not frightening the birds, but wood ducks will often tolerate human disturbance. It has been noted, also, that

fall populations of wood ducks move to new locations after hunters

have shot at them a few times.

The Use of Loafing Logs for Wood Ducks

Since wood ducks regularly use stumps and logs as loafing places under various conditions, it was thought that the presence of a log on a pond might aid in attracting ducks to chosen ponds for nesting or

it might facilitate brood counts on small ponds. Through the coopera­ tion of Delmar E. Handley, Ohio Division of Wildlife, logs were placed on 20 randomly chosen ponds at the Olentangy Station. These logs did not alter the utilization of ponds for nesting, and they did not aid

in finding broods. The logs were never observed being used b y wood ducks until mid-Julyj young ducks already able to fly were noted sitting on the logs on one occasion.

Summw-Yy

In Connecticut and Illinois, wood ducks have been captive-reared and released for restocking purposes. Some captive-reared females returned and nested near the release point in the following year. The eggs were gathered from the nests of wild birds and incubated at 99°F. and 80 per cent or more relative humidity. The ducklings were first fed on hard-boiled chicken eggs and later on pellets of commercial turkey 318

starting feed. Approximately 3.1 pounds of feed were consumed in attain­

ing a weight of one pound. An electric light kept in the cage attracted

insects which the ducklings ate. Difficulty inducing the ducklings to

start eating can be overcome by placing the newly hatched ducklings in

cages with other ducklings which have already started eating. Fine

grit should be provided.

Intensive management of the wood duck appears to be feasible. In

Massachusetts, 956 ducklings were hatched-in 94- boxes in a 3-mile section of river.

Wood ducks will tolerate considerable human association, particular­

ly in the selection of a. nesting site, but fall populations sometimes move to new areas because of human disturbance.

Specially placed loafing logs were little used by wood ducks. CHAPTER XVII

MISCELLANEOUS TECHNIQUES IN WOOD DUCK MANAGEMENT

Finding Nests of Wood Ducks

Wood ducks have been known to nest on repeated occasions in towns

or elsewhere near human dwellings without their presence being recog­

nized until the young left the nest. This failure to observe the nest­

ing bird is partly because the incubating female often leaves the nest

to feed and returns in the early morning. When the female leaves the

nest or when she enters the nest, she moves so rapidly that she is often

overlooked. The difficulty in finding nests of wood ducks is increased

by the fact that the incubating bird usually cannot be flushed from the

nesting cavity by rapping on the trunk of the nesting tree. Because

the young usually leave the nest in mid-forenoon and because they travel

on the ground, they are frequently observed soon after they leave the

nest.

The nests of wood ducks using artificial nesting places can be

found by checking these nesting places. When the ducks are nesting in

natural cavities, their nests cannot be so easily found. Slowly float­

ing down a river in a boat or cruising in an automobile over country

roads in the early morning near wood duck habitat are useful methods for locating wood duck nests. Before egg laying commences, the pair of wood

ducks often spends much time in the morning before 8:00 A.M. perched in the tree containing their nest or in a tree nearby. By watching such birds, they can usually be observed to enter their nesting cavity. During

319 320

the incubation period, wood ducks can be seen flying toward their nests

as they return from feeding. When they return from feeding the male

usually accompanies the female to a point near the nest, and the female

enters the cavity while the male continues flying or alights nearby.

When a cavity entrance can be seen from the ground, the presence

of nesting down around the edge of the entrance can sometimes be detected

with binoculars, and the presence of down usually indicates a cavity is

occupied. Two cavities lacked nests when their entrances contained

duck down.

Use of Nest Eggs for Inducing Nesting

Since wood ducks frequently lay eggs in nests of others of the

same species, it was thought that the use of nest eggs might induce

them to occupy given boxes. In late April of 1957, a wood duck egg was

placed in each of 20 boxes at the Olentangy Station, and a second box

on each pond was left without a nest egg. Two boxes with and two with­

out nest eggs were occupied. The use of nest eggs did not induce the birds to utilize given boxes.

Aging the Eggs

Various phases of the present study made it necessary to be able to predict the hatching times for wood duck clutches. Weller (1956) described a simple candler which consisted of a tube small enough for an egg to lay in one end. The observer then looks through the tube as he points it toward the sun or sky. Weller presents illustrations show­ ing the appearance of the eggs at different stages of incubation. In 321 my study, Weller's method was adopted as a simple and at the same time

satisfactory method. Actually, the hollow cardboard cylinders on which toilet tissue is wrapped were used. These were about the proper size and were readily replaceable.

Several other methods for aging eggs are available. Hanson (1954) described a portable egg candler making use of batteries. Illustrations of the appearance of wood duck eggs at various stages of incubation were presented. Westerskov (1950) described a flotation method for determining stages of incubation in eggs of the ring-necked pheasant.

Gigstead (1938:605) removed an egg from each duck's nest, marked the eggs to identify them with the nests, and incubated them under domestic hens. When pipping of the sample egg started, the nest from which the egg came was visited for observations on hatching. Sowls (1955:82) removed an egg from duck nests, broke the egg, and compared the embryo with embryos of a known age which had been artifically incubated.

Egg Membrane Counts for Gathering Information on

Hatching Success and Production of Ducklings

In connection with a post-hatching nesting box check, it is possible to gather some information on the total number of young hatch­ ed per nest and the proportion per clutch of eggs producing young. When the young hatch from the eggs, they normally leave the empty shell in two main pieces: a cup with the general form of the original egg and consisting of the more pointed end of the egg, and a cap consisting of the shell and membrane of the large end of the egg (Figure 61). After the young hatch, the dry brittle shells are soon broken into small Figure 61 322 Egg Shell and Membrane After Hatching of Duckling 323

pieces by trampling of the young and hen; the shells and egg membranes

become mixed among the duck dovn and shavings or sawdust in the nest.

Most of the shell becomes separated from the membrane, but the membrane

remains in its earlier two pieces. These membranes can be sorted from

the other material with which they become mixed and the total number

of egg ends of both types counted (Figure 62).

According to Leopold (1951:216) the female never removes egg shells

or membranes from the nest, and the number of ducklings hatching can be

determined by counting the membranes. If it were certain that the female never removes any of these egg membranes from the nest and the counts were made soon after the young have left the nest, counting pieces of membrane should be a perfect method for gathering information on hatchl­

ing success. It cannot, however, be insisted that the female never re­ moves egg membranes from the nest. On May 16, 1955, I observed with

7 x 35 binoculars a female wood duck flying at a range of about 75 feet with something in her bill which almost certainly was an egg or perhaps shell from a hatched egg.

Gigstead (1938:606) reported that the female sometimes removes shells of hatched eggs from the nest. Breckenridge (1956:19) noted that the egg membranes had disappeared from a nest equipped with an automatic recording apparatus which indicated that the bird had not been off the nest during the time that the membranes disappeared. This observation prompted the hypothesis that the egg membranes are sometimes eaten by the female or ducklings.

On two occasions egg membranes have been found on the water or on the nearby land beneath nesting boxes after ducks had left boxes over Figure 62

U) Egg Membranes After Ducklings Left the Nest £ 325

water. The membranes may have been thrown from the nests by the female

duck or by another bird.

A case where a female was observed throwing egg shells from her

nest was reported to me by M.B. Trautman. Jacob Verduin reported to

Dr. Trautman that he (Verduin) had seen a female toss the shells out

of the nesting cavity to the ground below, but two eggs which contained

live young nearly ready to hatch were also thrown from the nest. This nest was near a house and near a road where the incubating female was

frequently disturbed by humans. Early in the nesting cycle a large

limb has broken from the tree containing this nest, so that the bird was constantly exposed to passersby as she sat on the nest, and it can

be doubted if her behavior in throwing the unhatched eggs from the nest should be considered normal behavior. Consequently, the same doubt centers around her disposition of the egg shells.

It has frequently been impossible to account for known numbers of eggs by use of membrane counts. In one box showing every other evidence of a successful nesting, not a single egg membrane could be found. The usual broken shells were in the nest. This box was on a post over water, and an explanation for the absence of egg membranes was not avail­ able.

When boxes are mounted on trees, white-footed mice frequently enter the boxes and feed on egg membranes soon after hatching of the wood ducks. On one occasion a mouse was observed entering a wood duck box before the eggs had hatched and when the female was off the nest. In ten boxes on trees where repeated membrane counts were made bi-weekly, progressively lower counts resulted in the majority of cases. Mice were 326 found in the boxes, and gnawing of the membranes was in evidence.

Membrane counts in connection with nesting box checks may have some value in determining nesting efficiency. Data thus obtained, however, cannot be accepted at face value. In 20 nests where the number of hatched young had been determined by counting the eggs in the nests within one week before hatching, and the membranes after hatching, or by counting the numbers of young actually in the nests, nine nests were found where the membrane count did not agree with the count made in another more reliable manner. The total discrepancy in the nine nests amounted to 8.44 per cent.

Membrane counts yielded a total below the actual hatch, and when information on hatching success is sought it is preferable to use a more reliable method, such as counting the number of eggs in the nest before and after hatching. Such dual counts necessitate an additional visit to the boxes. Embryos in unhatched eggs should then be aged to determine if any of the eggs left in the nest were laid by a foreign female after the observer's earlier visit. Foreign eggs are sometimes added near the end of the incubation period.

Obtaining Data on Sex Ratios

Before the characteristic plumage markings of the sexes developed on the ducklings, sex determination was made by cloacal examination.

Because of the difficulty of finding the occluded oviduct, the absence of a penis was taken to indicate a female. In most cases, the sex was readily ascertained. Subsequent recapture, however, of 22 birds banded at 3-4 weeks of age disclosed tiro wrongly sexed birds. All of the sexing 327 data for ducklings which had been sexed at 3-4 weeks of age, conse­ quently, were discarded. Sex determination by use of white cheek patches of the male can be made at 5-6 weeks of age.

Of 172 pre-flight wood ducks trapped and sexed in 1956, 85 were males and 87 were females, and of the 79 late summer immatures 38 were males and 41 were females. Of 42 pre-flight birds trapped and sexed in 1955, 23 were males and 19 were females, and of 144 late summer immatures sexed in 1955, 80 were males and 64 were females. It is be­ lieved that males and females are randomly distributed through the popu­ lations of pre-flight ducks and late summer immatures so that sexing trapped birds gives a reliable index to the true sex ratio.

Adult males and females,however, are not believed to be randomly distributed through the late summer population and should not be con­ sidered in calculating sex ratios of trapped late summer populations.

Of 38 adults trapped in central Ohio in late August and September, 14 were males and 24 were females. It is doubtful if this is the true sex ratio of wood ducks at this season. Benson and Cummings (1956:212) trapped samples indicating that from 10 to 100 per cent of the popula­ tion were males. Trapping and banding data obviously cannot be used for determining sex ratios of wood ducks unless more random samples can be taken.

In observations from a blind and general field observations at the

Savannah Refuge, a record was made of the sex of all wood ducks seen when the sex could be determined. The same birds were not knowingly counted twice, but some duplication was probable. Totals of 116 males and 115 females were thus observed. The sexes appeared to be approximately equal 328 in this population. In field observations of 251 wood ducks in Idaho,

Oregon, and California, Evenden (1952) found 149 males to 102 females.

Of 211 banded by Mcllhenny (1937:118) 145 were males and 55 were females.

Uneven sei ratios sometimes occur among wood ducks. Females on the nest are susceptible to predation by raccoons. A rather large number of females may be lost each year as a result of their entering chimneys from which they are unable to escape. These two factors could be ex­ pected to cause a reduced proportion of females in the population as the males are not exposed to these hazards.

Obtaining Data on Age Ratios

After duckling down was replaced with feathers, the young ducks were aged by examination of the tail feathers (Kortright, 1942:31).

Nearly all of the immature ducks contained at- least one V-pointed tail feather until trapping was discontinued at the end of September. Bursal probing was used in a few cases, but this was considered too time con­ suming for extensive use on living birds.

I found a higher proportion of immatures to adults at Buckeye Lake in 1955 than could be expected to be the true age ratio. Of 103 wood ducks trapped during August 9-25, only seven were adults. Trapping and banding in New York during the late summer and early fall, Benson and

Cummings (1956:211) also found a higher proportion of immatures to adults trapped than was indicated by other observations. Benson and

Cummings (1956:207) found no difference in the age groups in regard to susceptibility to trapping, and there appears to be some segregation of the sexes so that trapping in any given area does not provide a typical 329

sample of the entire population. Sutton (1928:71) reported segregation of adult and immature wood ducks in the late summer at Pyraatuning Lake.

Hunter bag checks in Vermont (Miller, 1954:4) showed 39 adults and 118 immatures among 157 wood ducks examined. Benson and Cummings

(1956:209) reported that on the average adults comprise 35 to 40 per cent of the early fall population in New York. Hunters* bags may contain a more random sample than that taken in banding traps earlier in the season.

Collecting Cron Contents From Living Wood Ducks

It was desired to collect information on the food of wood ducks, but because a study of natural wood duck production was the chief objective of the Ohio project, it was preferred that the birds should not be sacrificed. Consequently, an effort was made to secure samples from alimentary tracts of living wood ducks without bodily injury to the birds. If a bird had fed recently before capture so that the food was still in the crop, the food was readily worked out through the mouth by separating portions and following them up the esophagus with a squeezing and pushing motion of the thumb and index finger of one hand.

When the crop was crammed full of com, as usually happened when the ducks were in the trap, it was a little difficult to separate a small amount of corn from the remainder so it could be started up the esophagus.

A well-filled crop could be emptied in about twenty minutes.

Of five complete alimentary tracts from hunters* bags, three were empty except for a small amount of sand in the gizzard, and the two con­ taining food had food in the crops as well as in the proventriculi and t

330 gizzards. In these cases the small sample from the crop told the same story as the larger sample from the entire alimentary tract.

Samples from incubating wood ducks should be most useful for indi­ cating the natural food, but incubating birds with food in their crops were seldom found. Removal of food from crops of incubating birds consequently yielded only a small number of samples useful in deter­ mining the food of the wood duck under natural conditions. Except for the difficulty of finding birds with natural food in their crops,this method seemed entirely satisfactory for collecting food samples from living wood ducks. The wood ducks taken in baited areas fed on corn exclusively, and samples from trapped birds yielded no further clue to food items.

Alimentary Tracts From ¥ood Ducks in Hunters* Bags

The wood duck was not legal game in Ohio during 1956; thus collec­ tion of alimentary tracts from hunters* bags was not possible at that time and place. Several days were spent in January, 1957, among hunters in South Carolina in an effort to collect some wood duck alimentary tracts. Only three were collected. The hunters were generally very cooperative and did not appear to resent having the desired viscera re­ moved from their birds, but in several instances they voiced misgivings that the honest goal of the alleged food habits study was disclosure of illegal baiting. Fishermen Interviews

Because pursuit of their sport sometimes results in their spending much time in wood duck habitat, fishermen are in a favorable position to make incidental observations on wood ducks. Fishermen usually have sufficient general interest in nature that they notice the activity of wood ducks which chance to occur near them. Fishermen can thus provide the researcher with information leading to the location of nests, roosting and loafing places, or on the general behavior of the species.

Needless to say, duplication by the researcher will be desirable of observations reported on behavior. One fisherman reported that when very young ducklings get in a fast-flowing ripple on the river, the mother bird goes to their rescue, they climb on her back, and she swims to safety! On the other hand, some seemingly unusual behavior described by fishermen was later observed the way it had been described.

Summary

Wood duck nests can be found by observing the direction of flight of females in the morning during egg-laying or incubation periods. The male can sometimes be seen in the tree near the nest before and during egg-laying. The presence of down on the entrance to a cavity usually indicates an occupied cavity.

Nesting eggs for inducing wood duck use of a given cavity were tested and found to be ineffective.

Eggs were aged with a simple candling device consisting of a cardboard tube directed toward the sun or sky.

331 332

Egg membrane counts give some information on hatching success;

because membranes are sometimes removed from the nest, presumably by

the female, such counts are not entirely reliable. Two trips to the

nest are necessary for an accurate count of the number of young hatch­

ing in a nest.

Before the plumage markings characteristic of the sexes appeared

at five to six weeks of age, the young were aged by cloacal examination.

Immature males and females may be randomly distributed in the population

during late summer, and trapped samples show the true sex ratio among

immatures at this season. Male and female adults, however, do not appear to be randomly distributed among the population, and trapped

samples do not show their ratio in nature. During late fall and winter, males and females may be randomly distributed in the population. Uneven

sex ratios have been reported among wood ducks, with males much in pre­ ponderance. Loss of incubating females to raccoons and loss of nest- hunting females in chimneys could cause a heavier mortality of females.

Until the end of September most of the ducks were aged by the presence of the V-pointed tail feathers of immatures. Bursal probing was used in a few dubious cases. The proportion of immatures in the trapped sample was higher than would be expected. Workers in New York also found a higher proportion of immatures in trapped samples than was indicated by other observations of the wood duck population. There appears to be segregation of adults and immatures during the early fall months. Hunters* bags may yield a more random sample.

Samples of crop contents were obtained from living birds by working these contents up the esophagus with a squeezing and pushing motion of 33 3

the thumb and index finger. Birds taken in corn-baited traps were always filled with corn exclusively. Incubating birds usually had no food in their crops, but two were found with corn thought to have been obtained in farmers* hog lots. Several digestive tracts were obtained from birds in hunters* bags. -

Interviews with fishermen encountered in wood duck habitats some­ times yielded information helpful in duck research. APPENDIX

LIST OF SCIENTIFIC NAMES

Mammals

Opossum Didelnhis virginiana

Raccoon Pro cvon lotoir

Mink Mustela vison

Chipmunk Tamias striatus

Red Squirrel Tamiasciurus hudsonicus

Gray Squirrel Sciurus carolinensis

Fox Squirrel Sciurus niger

Beaver Castor canadensis

White-footed Mouse Permomvscus sp.

Muskrat Ondatra Zibethica

Birds

Water-Turkey Anhinga anhinga

Mallard Anas platvrhvnchos

Black Duck Anas rubripes

Gadwall Anas strepera

Pintail Anas acuta

Green-winged Teal Anas carolinensis

Blue-winged Teal Anas discors

Cinnamon Teal Anas cvanoptera

American Widgeon Mareca americana

Spatula

33U 335

Redhead Avthva americana

Canvasback Avtha valisineria

Common Golden-eye Glaudionetta claneuta americana

Hooded Merganser LoDhodvtes cucullatus

Common Merganser Mereus merganser iSharp-shinned Hawk Accioiter striatus

Red-shouldered Hawk Buteo Uneatus

Sparrow Hawk Falco soarverius

Ruffed Grouse Bonasa umbellus

Bobwhite Colinus vireinianus

Ring-necked Pheasant Phasianus colchicus

Virginia Rail Rallus limicola

American Coot Fulica americana

Killdeer Chardrius vociferus

Passenger Pigeon Ectopistes mieratorius

Mourning Dove Zenaidura macroura

Yellow-billed Cuckoo Coccvzus americanus

Road-runner Geococcvx californianus

Snowy Owl Nvctea scandiaca

Screech Owl Otus asio

Yellow-shafted Flicker Colaptes auratus

Red-headed Woodpecker Melaneroes ervthroceohalus

Great Crested Flycatcher Mviarchus crinitus

Tree Swallow Iridoorocne bicolor

Bank Swallow Riparia riparia

Purple Martin Proene subis Blue Jay- Cvanocitta cristata

Tufted Titmouse Parus bicolor

Itfhite-brearted Nuthatch Sitta carolinensis

Robin Turdus migratorius

Eastern Bluebird Sialia sialia

Starling Sturnus vulgaris

Common Grackle Quiscalus guiscula

Red-winged Blackbird Agelaius phoeniceus

Brown Towhee Pipilo fuscus

Song Sparrow Melospiza melodia

Reptiles

Rat Snake Elaphe obsoleta

Bull Snake Pituophis savi

"Water Snake Natrix sp,

Snapping Turtle Chelvdra serpentina

Amphibians

Bullfrog Rana catesbiana

Green Frog Rana clamitans

Fish

Alligator Gar Lepisosteus spatula

Northern Pike Esox lucius

Largemouth Black Bass Micropterus salmoides

Insects

Black Fly Si mill ium venustum Bumble Bee Bombus sp.

Honey Bee Apis mellifera

Plants

Cat-tail Tvoha so.

Maiden Cane Panisum hemitomum

Willow Salix so.

White Oak Quercus alba

Post Oak Quercus stellata

Over-cup Oak Quercus lvrata

Bur Oak Quercus macrocarpa

Basket Oak Quercus michauxii

Live Oak Quercus vireiniana

Southern Red Oak Quercus rubra

Pin Oak Quercus palustris

Southern Red Oak Quercus texana

Black Oak Quercus velutina

Spanish Oak Quercus falcata

Swamp Spanish Oak Quercus paeodoefolia

Water Oak Quercus nigra

Black Jack Oak Quercus marilandica

Laurel Oak Quercus imbricaria

Willow Oak Quercus phellos

Nettle Urtica so.

Spatter-dock Nvmphae advena

Water Chinquapin Nelumbium luteum

Sycamore Platanusi occidentalis Blackberry Rubus sp.

Grape Vitis sp.

Willow-herb Epilobium. sp.

Dogwood Cornus sp.

Buttonbush Cephalanthus occidentalis

Elderberry Sambucus canadensis LITERATURE CITED

Allen, C.S. 1892. The nesting of the black duck on Plum Island. Auk, 10:53-59.

American Ornithologists Union. 1957. Check list of North American birds. 5th ed.

Andrews, Ralph. 1952. A study of waterfowl nesting on a Lake Erie marsh. Unpublished M.Sc. thesis, The Ohio State University, pp. 153.

Anonymous. Un- The wood duck - an opportunity for basic research, dated. Westboro, Mass.:Mass. Div. of Fisheries and Game.

Audubon, J.J. 1843. Birds of America. N.Y. and Philadelphia: Published by author. Vol. VI.

Aymars, G.C. 1935. Bird Flight. Garden City, N.Y.: Garden City Pub. Co.

Bagg, A.M. 1948. Wood duck courting a mallard. Auk, 65:451.

Bailie, J.L. 1951. Ontario-western New York region. Audubon Field Notes, 5:201-203.

Baldwin, S.P. and S.C. Kendeigh. 1932. Physiology of the temperature of birds. Sci. Pub. Cleveland Mus. Nat. Hist. 3:1-196.

1938. Variations in the weight of birds. Auk, 55:416-467.

Barnes, W.B. 1948. Unusual nesting behavior of a wood duck. Auk, 65.:449.

Baynard, O.E. 1913. Breeding birds of Alachua County, Florida. Auk, 30:240-247.

339 340 Beard, E.B. 1953. Importance of beaver in waterfowl management at the Seney National Wildlife Refuge. Jour. Wildl. Mgmt., 17:398-436.

Bellrose, F.C. 1943. Two wood ducks occupy same nest. Auk, 60:446-447.

1944* Sparrow hawk incubating wood duck eggs. Auk, 61:299.

1945. Aspergillosus in wood ducks. Jour. Wildl. Mgmt., 9:325-326.

1955a. Comparison of recoveries from reward and standard bands. Jour. Wildl. Mgmt., 19:71-75-

1955b. Housing for wood ducks. Illinois Nat. Hist. Surv. Circ. 45.

Bennett, L.J. 1938. Blue-winged teal, its ecology and management. Ames, Iowa: Collegiate Press.

Benson, Dirck and G.E. Cummings. 1956. Some implications of waterfowl trapping records from the Oak Orchard Game Management Area. N.Y. Fish and Game Jour., 3:205-216.

Bent, A.C. 1923. Life histories of North American wild fowl. U.S. Natl. Mus. Bull. 126.

Beyer, G.E., Andrew Allison, and H.H. Kopman. 1907. List of the birds of Louisiana. Auk, 24:314~321,

Breckenridge, W.J. 1947. Wood duck versus squirrels. Auk, 64:621.

1956. Nesting study of wood ducks. Join*. Wildl. Mgmt., 20 :16-21 .

Brooks, Maurice. 1948. Appalachian region. Audubon Field Notes, 2:175-176.

Brown, L.G. and F.C. Bellrose. 1943. Use of nesting boxes for wood ducks by other wildlife. Jour. Wildl. Mgmt., 7:298-306. 341

Bump, Gardiner, R.W. Darrow, F.C. Edminster, and W.F. Crissey. 1947. The ruffed grouse, life histojry, propagation, management. Albany, N.Y. State Conserv, Dept.

Catesby, Mark. 1731. The natural history of Carolina, Florida, and the Bahama Islands. Vol. IV.

Chapman, F.C. 1924. Handbook of birds of eastern North America. N.Y.: Appleton and Co.

Cochran, Rod. 1955• Ohio*s Wildlife resources. Columbus, Ohio: Division of Wildl.

Collias, N.E. and E.C. Collias. 1956. Some mechanisms of family integration in duck. Auk, 73:378-400.

Cooke, W.W. 1906. Distribution and migration of North American ducks, geese, and swans. Washington, D.C.: U.S. Dept, of Agri., Biol. Surv, Bull. 26.

Cooke, M.T. 1937. Flight speeds of birds, Washington, D.C.: U.S. Dept, of Agri. Circ. No. 428.

Cottam, Clarence, C.S. Williams, and C.A. Sooter. 1942. Flight and running speeds of birds. Wils. Bull,, 54:121-131.

Coues, E. 1903. Key to North American birds. 5th ed. Vol. II, Boston: Dana Estes and Co.

Coulter, M.W. 1955. Spring food habits of surface-feeding ducks in Maine. Jour. Wildl. Mgmt., 19:264-267.

1957. Predation by snapping turtles upon aquatic birds in Maine marshes. Journ. Wildl. Mgmt., 21:17-21.

Cronan, J.M. 1957. Effects of predator guards on wood duck box usage. Jour. Wildl. Mgmt., 21:468. 34 2

Dawson, W.L, 1903. The birds of Ohio. Vol. II. Columbus, Ohio: Wheaton Pub. Co.

Dixon, Joseph. 1924., Nesting of the wood duck in California. Condor, 26 :41-6 6.

Dreis, R.E. 1951• Productivity of the wood duck Aix snonsa (L.), in eastern Louisa County, Iowa, 1951. Unpublished

M.Sc. Thesis, I o w a . State College.

1954* A field observation method of aging wood ducks. Jour. Wildl. Mgmt., 18:280-281.

Dreis, R.E. and G.O, Hendrickson. 1952. Wood duck production from nest-boxes and natural cavities on the Lake Odessa Area, Iowa, in 1951• Iowa Bird Life, 22:18-22.

Dutcher, William. 1907. The wood duck. Bird-Lore, 9:189-192.

Edminster, F.C. 1953. Snapping turtle catches pheasant. Jour. Wildl, Mgmt., 17:383.

Errington, P.L. 1937. A wood duck marsh in northwestern Iowa. Auk, 54:533-534*

Evans, C.D. 1951. Method of color marking young waterfowl. Jour. Wildl. Mgmt., 15:101-103.

Evenden, F.G. 1952. Waterfowl sex ratios observed in the western United States. Jour. Wildl. Mgmt., 16:391-393.

Fenneman, N.M. 1938. Physiography of eastern United States. New York: McGraw-Hill Inc.

Fisher, A.K. 1901. Two vanishing game birds - the woodcock and the wood duck. Washington D.C.: U.S. Dept, of Agri,. Yearbook, 1901:447-458. I 343

Forbush, E.H. 1905. Decrease of birds in New England. Auk, 22:25«,31.

1912. History of the game birds, wild-fowl and shore birds of Massachusetts and adjacent states. Boston, Mass., State Bd. of Agri.

1925. Birds of Massachusetts and other New England states. Vol. I. Boston: Mass. Dept, of Agri.

Frank, W.J. 194-8. Wood duck nesting box usage in Connecticut. Jour. Wildl. Mgmt., 12:128-136.

Gabrielson, I.N. 1943. Wildlife refuges. New York: MacMillan Co.

Gigstead, G. 1938. Wood ducks in the Illinois river bottoms. Trans. N. Am. Wildl. Conf., 3:603-609.

Girard, G.L. 1941• The mallard: its management in western Montana. Jour. Wildl. Mgmt., 5:233-259.

Gollop, J.B. 1956. The use of retrievers in banding flightless mallards. Trans. N. Am. Wildl. Conf., 21:239-243.

Goslin, Robert. 1945. Bird remains from an Indian village site in Ohio. Wils. Bull., 57:131.

Green, W.E. 1951. An analysis of duck kill and hunter success. Upper Miss. River Cons. Comm. Proc., 7:90-93.

Grice, David. 1956. Detailed investigation of wood duck production^ survival and annual return to a study area. Mimeo­ graphed annual report, Massachusetts, proj. No. PR-19-R.

Hamilton, W.J. 1943. Nesting of the eastern bluebird. Auk, 60:91-94.

Hammond, M.C. and W.R. Forward. 1956. Experiments on causes of duck nest-predation. Jour. Wildl. Mgmt., 20:243-247. Handley, D.E. 1955. Wood duck nest box investigations. Columbus: Ohio Div. Wildl. Release No. 5*

Un- Trends in the construction of farm ponds in Ohio and dated, their importance in waterfowl management. Columbus: Ohio Div. Wildl.

Hanson, H.C, 1951. Notes on the artificial propagation of wood duck. Jour. Wildl. Mgmt. 15:68-72.

1954* Apparatus for the study of incubated birds eggs. Jour. Wildl. Mgmt., 18:191-198.

Hawkins, A.S., F.C. Bellrose and H.G. Anderson. 1939. The waterfowl research program in Illinois. 111. Nat. Hist. Surv. Biol. Notes 12.

Hickey, J.J. 1949. Survival studies of banded birds. Unpublished Ph.D. dissertation, University of Michigan.

Hochbaum, H.A, 1944* Canvasback on a prairie marsh. Washington, D.C.: Amer. Wildl. Inst.

1947. Effect of concentrated hunting pressure on waterfowl breeding stock. Washington, D.C.: Trans. N. Am. Wildl. Conf., 12:53-62.

Hough, R.B. 1907. Handbook of trees of the northern States and Canada east of the Rocky Mountains. Lowville, N.Y.: Pub. by the author.

Howard, W.E. 1950. Birds as bullfrog food. Copeia. 2:152.

Howell, A.H. 1932. Florida bird life. Fla. Dept. Fresh Water Fish co-op. with U.S. Fish and Wildlife Serv.

Hoyt, J.S.Y. 1948. Observations on nesting associates. Auk, 65:188-196. 345

Hunt, R.M. 1956. My experiences during 38 years of breeding and rearing game birds and wild waterfowl at the Mason State Game Farm. Lansing: Mich. Dept. Cons.

Jopson, H.G.M. 1956. Red-wing nesting in bird house. Auk, 73:284..

Kendeigh, S.C. 1952. Parental care and its evolution in birds. 111. Biol, Monog., 22:1-356.

Kiel, W.H., Jr. 1955. Nesting studies of the coot in southwestern Manitoba. Jour. Wildl. Mgmt., 19:189-198.

Kingsford, E.G. 1917,. Wood duck removing young from nest. Auk, 34s335-336.

Klein, H.G. 1955. Wood duck production and use of nest boxes on some small marshes in New York. N.Y. Fish and Game Jour., 2:68-83.

Kortright, F.H. 1942. The ducks, geese and swans of North America. Washington, D.C.: Amer. Wildl. Inst.

Lagler, K.F. 1956. The pike, Esox lucius Linnaeus, in relation to waterfowl on the Seney National Wildlife Refuge, Michigan. Jour. Wildl. Mgmt., 20:114-124.

Laidley, J.C. 1940. Important points in rearing of ducks. Avicul. Mag., 5:158-160.

Laskey, A.R. 1947. Snapping turtles apparently preying upon passerine birds. Auk, 64:315-316.

Layne, J.N. and A.H, Benton. 1954- Some speeds of small mammals. Jour. Mamm., 35:103-104.

Lee, J .A« 1956. Wood duck breeding studies. Mimeographed job completion rpt., New Hampshire, proj. No. W-7-R-12, job No. I-D.

Leopold, Frederic. 1951. A study of nesting wood ducks in Iowa. Condor, 53:209-220. 346

Levine, N.D. and R. Graham. 1942. Paratyphoid in baby wood ducks. Jour. Am. Vet, Assn., 100:240-24.1.

Lincoln, F.C. 1935. Migration of North American birds. Washington, D.C.: U.S. Dept, of Agri. Circ. 362.

1939. Migration of American birds. N.I.: Doubleday, Doran.

Low, J.B. 1945. Ecology and management of the redhead, Nvroca americana. in Iowa Ecol. Monog., 15:35-39.

1952. Great Basin, central Rocky Mountain region. Audubon Field Notes, 6:207-208.

Mabbott, D.C. 1920. Food habits of seven species of American shoal-water ducks, U.S. Dept, of Agri. Bull., 862.

Manwell, R.D. 1954- A case of aspergillosis in a song sparrow. Jour. Parasit. 40:231.

Martin, A.C. and F.M. Uhler. 1939. Food of game ducks in United States and Canada. Washington D.C.: U.S. Dept, of Agri. Tech. Bull. No. 634.

Mayfield, H.F. 1948. Middle western prairie region. Audubon Field Notes, 2:208-209.

Mayhew, W.W. 1955. Spring rainfall in relation to mallard production in the Sacramento Valley, California. Jour. Wildl. Mgmt., 19:36-47.

McAtee, W.L. 1921. Homing and other habits of the bullfrog. Copeia, 96:39-40.

1923. Local names of migratory game birds. Washington, D.C.: U.S. Dept, of Agri. Misc. Circ. 13.

McCabe, R.A. 1947. Homing of transplanted young wood ducks. Wils. Bull., 59:104-109. 4

347

Mcllhenny, E.A. 1937. Results of 1936 bird banding operations at Avery Island, Louisiana, with special reference to sex ratios and hybrids. Bird-Banding, 3:117-121.

McLaughlin, C.L. and David Grice. -* > 1952. The effectiveness of large-scale erection of wood duck boxes as a management procedure. Trans. N. Am. Wildl. Conf., 17:242-259.

McLean, D.D. 1930. Speed of flight in certain birds. Gull, 12:1-2.

Meade, G.M., and Dayton Stoner. 194-2. Aspergillosis in a snowy owl. Auk, 59:577-578.

Mielcarek, J.E. 1954. The occurrence of Plasmodium relictum in the wood duck (Aix sponsa). Jour. Parasit., 40:232.

Miller, W. DeW. 1925. The secondary remiges and coverts in the mandarin and wood ducks. Auk, 42:41-50.

Miller, W.R. 1950. Wood duck nesting studies. Vermont Fish and Game Ser., Mimeographed P-R Project 8-R-4 .

1951. Wood duck nesting studies. Vermont Fish and Game Ser., Mimeographed P-R Project W-8-R-5.

1952. Aspects of wood duck nesting box management. Vermont Fish and Game Ser., Mimeographed P-R Project 8-R.

1954. Sex and age data collection. Vermont Fish and Game Ser., Mimeographed job completion report, project No. W-8-R-7, job No. 6,

1955. Sex and age data collection. Vermont Fish and Game Ser., Mimeographed job completion report, proj. W-8-R-9, job. No. 6.

Morgan, Allen, and R.P. Emery. 1956. Northeastern maritime region. Audubon Field Notes, 10:234-236. 348

Morris, R.O. 1892. Notes from Springfield, Massachusetts. Auk, 9:74..

Morris, R.T. 1930. Concerning wood ducks. Bird-Lore, 32:420-421.

Mousley, Henry. 1924.. Further notes and observations on the birds of Hatley, Stanstead County, Quebec, 1919-1923. Auk, 4-1:572-539.

Mumford, R.E. 1952. Study of wood duck populations on Indiana streams. Unpublished M.Sc. thesis. Purdue Univ. Pp. 4-9.

1954-. Waterfowl management in Indiana. Indianapolis, Indiana: Indiana Dept, of Cons.

Munro, J.A. 194-1. Studies of waterfowl in British Columbia. Greater scaup duck, lesser scaup duck. Canadian Jour, of Res., 19:113-138.

Musselman, T.E. 194-3. A changing nesting habit of the wood duck. Auk, 65:197-203.

Neff, J.A. 1955. Outbreak of aspergillosis in mallards. Jour. Wildl. Mgmt., 19:415-416.

Nelson, A.L. and A.C. Martin. 1953. Gamebird weights. Jour. Wildl. Mgmt., 17:36-42.

Nichols, C.K. 1937. Early nesting of the wood duck. Auk, 54:533.

1955. Hudson-St. Lawrence region. Audubon Field Notes, 9:366-370.

Nolan, Val. Jr. 1956. Midwestern prairie region. Audubon Field Notes, 10:386-387.

Norton, A.H. 1909. The food of several Maine water-birds. Auk, 26:438-440.

0*Meara, D.C. 1956. Blood parasites of some Maine waterfowl. Jour. Wildl. Mgt. 20:207-209. 349

0*Roke, E.C. 1934. A malarie-like disease of ducks caused "by Leucocvtozoon anatis Wickware. Univ. Mich. Sc. of For. and Conser.

Bull. 4 •

Peterson, R.T. 194-7. A field guide to the birds. Boston: Houghton Mifflin Co.

Pettingill, O.S., Jr. 1935. Ducks in chimneys. Auk, 52:303.

Phillips, J C. 1925. A natural history of the ducks. V°l. 3. Boston and N.Y.: Houghton Mifflin Co.

Pirnie, M.D. 1935. Michigan waterfowl management. Lansing, Mich.j Mich, Dept. Conser., Game Div.

Potter, J.K, and J.H. Murray. 194-9. Middle Atlantic coast region. Audubon Field Notes, 5:231-232.

Raney, E.C. 1942. Alligator gar feeds upon birds in Texas, Copeia 24:50.

Rapp, W.F. 1941- Red-shouldered hawk eating a wood duck. Auk, 53:572-573.

Richdale, L.E. 1942. Supplementary notes on the royal albatross. Emu, 41:169-134.

1949. The effect of age on laying dates, size of eggs, and size of clutch in the yellow-eyed penguin. Wils. Bull., 61:91-9 8.

Roberts, T.S, 1936. The birds of Minnesota. Minneapolis, Minn.: University of Minnesota Press.

Rollin, Noble. 1957. Incubation by drake wood duck in eclipse plumage. Condor, 59:263-265.

Romanoff, A.T. 1943. A study of artificial incubation of runner duck eggs: requirements. Poultry Sci. 22:148-154* 350

Saunders, A. A. 1937. Injury-feigning by a wood duck. Auk, 54:202-203.

Schmidt, K.P. and D.D. Davis. 1941. Field book of snakes of the United States and Canada. N.Y.: G.P. Putnam*s Sons.

Schoffman, R.J. 1946. Preliminary report on the food of ducks at Reelfoot Lake, Tennessee. Jour. Tenn. Acad. Sci., 21:10-13.

Schreiner, K.M. and G.0. Hendrickson. 1951. Wood duck production aided by nesting boxes, Lake Odessa, Iowa in 1950. Iowa Bird Life, 21:6-10,

Seubert, J.L. 1952. Observations on the renesting behavior of the ring-necked pheasant. Trans. N. Am. Wildl. Conf., 17:305-329.

Shelley, L.0. 1933. Song sparrow in the stomach of a frog. Auk, 50:123.

Sibley, C.G. 1955. Behavioral mimicry in the titmice (Paridae) and certain other birds. Wils. Bull., 67:128-132.

Siegel, Sidney. 1956. Nonparametric statistics for the behavioral sciences. N.Y.: McGraw-Hill Book Co.

Southwick, Charles. 1953. A system of age classification for field studies of waterfowl broods. Jour. Wildl. Mjgmt., 17:1-8,

Sowls, L.K. 1949. A preliminary report on renesting in waterfowl. Trans. N. Am. Wildl. Conf., 14:260-272.

1955. Prairie ducks. Washington, D.C.: Wildl. Mgmt. Inst.

Stewart, P.A. 1937. A preliminary list of bird weights. Auk, 54:324-332.

Stollberg, B.P. 1950. Food habits of shoal-water ducks on Horicon Marshs Wisconsin. Jour. Wildl. Mgmt., 14:214-217.

Stoner,. Dayton. 1936. Studies on the bank swallow Riparia riparia (Linnaeus). Syracuse N/Y.: Roosevelt Wildl. Forest Experiment Station. 351

Stuewer, F.W. 19-45. Small entrance to a wood duck nest. Jour. Wildl. Mgmt., 7:236.

Stullpen, D.E. and C.M. Kirkpatrick. 1953. A determination of hole sizes which exclude certain predatory mammals. Jour. Wildl. Mgmt., 17:124.-131.

Sutton, G.M. 1928. Birds of Pymatuning Swamp and Conneaut Lake, Crawford County, Pennsylvania. Ann. Carnegie Mus., 18:18-239.

Taverner, P.A. and B.H. Swales. 1907. The birds of Point Pelee. Wils. Bull., 14.:37-54-.

Townsend, C.W. 1916. The courtship of the merganser, mallard, black duck, baldpate, wood duck, and bufflehead. Auk, 33:9-17.

Trainer, J.E. 1946. The auditory acuity of certain birds. Unpublished Ph.;D. dissertation, Cornell University, Pp. 97.

Webster, C.G. 1954-. Nest boxes for wood ducks. Washington, D.C.: U.S. Dept, of Int., Fish and Wildlife ser. Wildlife leaflet 351.

1955. Hatching of wood ducks after abandonment. Wils. Bull., 67:306.

Weller, M.W. 1956. A simple field candler for waterfowl eggs. Jour. Wildl. Mgmt., 20:111-113.

Westerskov, K, 1950. Methods for determining the age of game bird eggs. Jour. Wildl. Mgmt., 14.:56-67.

Wilson, Alexander. 1854. American ornithology. N.I., Charles L. Cornish,

Wing, Leonard. 194.7. Christmas census summary 1900-1939. Mimeograph. Pullman, Wash.: State College of Washington.

Wright, A.H. and Francis Harper. 1913. Biological reconnaissance of Okefinokee: the birds. Auk, 30:477-505. Wright, E.G. 1939. Marking birds by imping feathers. Jour. Wildl. Mgmt. 3:238-239.

Yeagley, H.L. 1953. Some surprises in research. Audubon Mag., 55:158-161 AUTOBIOGRAPHY

I, Paul Alva Stewart, was born near Leetonia, Ohio, on June 24*

1909* After obtaining an elementary education in a rural school of

Columbiana County, Ohio, my formal education was interrupted. I entered The Ohio State University as a special student in 1948. I became a regular student by earning a B average in all of my under­ graduate work and received the Bachelor of Science degree in 1952.

When the undergraduate work was completed, I received a research fellowship from the Ohio Cooperative Wildlife Research Unit and commenced graduate work at The Ohio State University. The Master of Science degree was received in 1953. Continuing with the fellowship, I worked during the following four years on the re­ quirements for the degree Doctor of Philosophy.

353