Biology of a southern mallard: Florida's mottled duck

THE BIOLOGY OF A SOUTHERN MALLARD:
FLORIDA'S MOTTLED DUCK

By

PAUL NEIL GRAY

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1993

ACKNOWLEDGEMENTS

My parents have supported me above and beyond the call

of duty, and to them I dedicate this dissertation. And to

my father, who was trained as an engineer but has worked as

a businessman to support his family, I extend a special

tribute. I hope in some measure it was gratifying to him to

provide me with an opportunity he did not have.

My advisor, H. Franklin Percival, has been a tremendous

help throughout. He nurtured me with great care and treated

me as a person, not a commodity. I will miss arguing with

him about topics important and otherwise.

Fred Johnson, of the Florida Game and Fresh Water Fish

Commission (GFC) funded much of this project, showed me

study sites and taught me Florida Duck biology. Also of

GFC, the crew of the Okeechobee Biological Field Station was

most helpful in a variety of ways. Don Fox, in particular,

was most cooperative and loaned a lot of equipment to me.

Don Brown could fix things as well as I could destroy them.

John Wulschlager, Gary Warren, Steve Miller and Rich

Turnbull were helpful in many ways.

Special acknowledgement must go to John Conomy and

Spencer Simon who gave me superlative help in the field and

lab. They both gracefully braved heat, fire ants, snakes,

air boats, and the town of Okeechobee. Barbara Fesler, of

the Coop Unit office, took me into her brood and helped me

in myriad ways. She does not deserve an acknowledgement,

she deserves an award.

My committee, Drs. G. Ronnie Best, Katherine C. Ewel,

Ronald F. Labisky, James D. Nichols, and George W. Tanner,

was very helpful and provided constructive criticism

throughout. I owe special thanks to Kathy Ewel, who through

my badgering and her patience, has virtually served as my

second advisor.

Many landowners in the Okeechobee area graciously

allowed me access to their properties. They include

Clarence Lofton of the Flying B Ranch, Nancy Goolsby of the

Flying G Ranch, Woody Larson and Bill Bradley of Larson

Dairies, Kent Bowen of McArthur Farms, Jeff Clemons of the

Okeechobee Livestock Market, John Pearce and Skip Croncich,

of Pearce Farms, and Frank Teale of Remilu Ranch. Almost

every one of them towed us out of a mudhole or helped fix

our truck at some time or another.

Additionally, Ross and Liz have been great friends and

roommates. I have been amused by their critters, especially

Rollie, the little orange weirdo. Thank you to Laurie whose

marvelous company helped me extend completion of this by at

least a year. I also extend a tribute to the late Dennis

Raveling; his ideas and approaches to waterfowl biology have

greatly influenced my thinking.

iii

TABLE OF CONTENTS

page
ACKNOWLEDGEMENTS...................................... ii

ABSTRACT.............................................. vi

CHAPTERS

1 GENERAL INTRODUCTION............................ 1
Study Area ...................................... 4

2 SURVIVAL, HABITAT USE, AND ECOLOGY OF FLORIDA
DUCK HENS AND THEIR BROODS................. 7

Introduction. ................................... 7
Materials and Methods............... ............. 9
Results ..................................... 14
Discussion...................................... 21

3 PHYSIOLOGY OF THE ANNUAL CYCLE OF FLORIDA'S
MOTTLED DUCK ............................... 29

Introduction...................................... 29
Materials and Methods............................ 34
Results .......................................... 43
Discussion...................................... 61

4 MOLTS AND PLUMAGES OF THE FLORIDA DUCK.......... 84

Introduction.................................... 84
Materials and Methods........................... 88
Results .......................................... 95
Discussion...................................... 139

5 SYNTHESIS: ADAPTATION IN FLORIDA'S
MOTTLED DUCK. ............................... 149

APPENDICES

A PEARSON CORRELATION COEFFICIENTS, PROBABILITY
LEVELS, AND SAMPLE SIZE FOR PHYSIOLOGICAL
VARIABLES OF MALE FLORIDA DUCKS.............. 153

B PEARSON CORRELATION COEFFICIENTS, PROBABILITY
LEVELS, AND SAMPLE SIZE FOR PHYSIOLOGICAL
VARIABLES OF FEMALE FLORIDA DUCKS............ 156

LITERATURE CITED.......................................... 159

BIOGRAPHICAL SKETCH ............. ..................... 172

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

THE BIOLOGY OF A SOUTHERN MALLARD:
FLORIDA'S MOTTLED DUCK

By

Paul Neil Gray

December 1993

Chairman: Dr. H. Franklin Percival
Major Department: Wildlife and Range Sciences
(Forest Resources and Conservation)

Florida's Mottled Duck (called the Florida Duck herein)

is endemic to the Florida peninsula and one of six morphs of

mallards in North America. Brood-leading hens were radio-

monitored to test the hypotheses that brood mortality limits

population growth and that brood-leading hens prefer to be

nocturnally active on mudflat-type habitats. Internal

physiology was examined to test the hypothesis that other

mallards in North America should have respectively greater

annual energy demands due to their migration and colder

environment than the non-migratory Florida Duck. Also,

competing hypotheses of whether members of the mallard

complex have two or three body molts annually, were

evaluated.

The survivorship of hens and their ducklings during the

brood-rearing period was estimated at 1.0 and 0.5,

vi

respectively, which, in itself, should not limit population

growth. Hens and broods used dense vegetation diurnally and

open areas (most frequently mudflats) nocturnally. At all

times, broods and hens favored water < 10 cm deep.

Florida Ducks acquired large amounts of body reserves

before breeding, lost reserves during breeding, partially

recovered reserves before wing molt, and gained large

amounts of reserves from autumn through winter. This

pattern is similar to that in migrant mallards; however,

Florida Ducks had a lower amplitude of change in body

reserves than the migrants. Lipids were stored and utilized

more readily than proteinaceous reserves.

The mass of internal organs changed seasonally in

Florida Ducks. Heart mass was positively correlated with

body mass. The liver and kidneys enlarged during egg laying

in hens, and during wing molt in both sexes. The gut

enlarged during autumn, and also enlarged during the egg-

laying period in females. These trends were similar to

those in all mallards and reflect similar requirements.

Florida Ducks underwent a head-body-tail molt three

times per year: in spring, before wing molt, and after wing

molt. Both sexes changed plumage aspect during the summer

and autumn molts.

vii

CHAPTER 1
INTRODUCTION

Florida's Mottled Duck (Anas fulvigula fulvigula) is a

non-migratory mallard that is endemic to the Florida

peninsula and does not migrate from the state. This duck,

hereafter referred to as the Florida Duck (Bellrose 1980),

is one of more than 20 kinds of mallards worldwide (Scott

1972, Palmer 1976, Weller 1980), but is reproductively

isolated from other populations because it is the only

mallard that breeds in Florida. When confined with other

North American mallards, such as the Mottled Duck (A. f.

maculosa), the Mexican Duck (A. f. diazi), the Mallard (A.

platyrhynchos), or the Black Duck (A. rubripes), the Florida

Duck can interbreed and bear fertile offspring (Johnsgard

1961). Hence, although Florida Ducks reportedly have been

extant since the Illinoian stage of the Pleistocene

(Brodkorb 1957), they still are closely related to other

mallards.

The Florida Duck merits close attention from wildlife

managers because it appears to have a declining population

(Johnson et al. 1984, 1991). Considering the phenomenal

rate of human population growth (Terhune and Floyd 1982) and

the massive wetland losses in Florida (Dahl 1990, Frayer and

Hefner 1991), a declining population is not surprising.

1

2

However, the main reasons for the population decline must be

determined before efficient corrective measures can be

enacted. Hence, I tested the hypothesis that brood-rearing

success is not adequate to sustain the population. If

survivorship of either brood-leading hens or ducklings is

unduly low, managers can target management activities toward

increasing survivorship during this period.

Florida Ducks are of interest ecologically because they

are a mallard species that lives in a unique different

environment (Scott 1972, Palmer 1976). Much attention has

been given to the bio-energetics of the annual cycle of

ducks in general, and the mallards in particular (Krapu

1981, Reinecke et al. 1982, Whyte and Bolen 1984, Heitmeyer

1988a, 1988b, 1988c, Cowardin et al. 1985, Whyte et al.

1986). Researchers have proposed cause and effect

relationships between changes in body reserves and organs

during the annual cycle and factors such as migration, cold

temperatures, breeding stress, molting stress, and food

availability. Many of the assumptions underlying these

hypotheses can be tested by observing Florida Ducks who live

in a sub-tropical (warm) environment, do not migrate,

maintain similar breeding and molting requirements, and have

somewhat different food habits than other mallards.

There is debate about whether mallards have two

(Witherby et al. 1939, Palmer 1972, Weller 1980) or three

molts (Heitmeyer 1987, Gray et al. 1989) per year, and

3

confusion on when the molts occur. Much of the confusion

results when researchers are unable to observe molts that

occur during migration. Florida Ducks do not migrate and

can be monitored continuously. Further, with the

concomitant physiological data I collected on each duck, I

tested hypotheses on the energetic and nutritional costs of

molt on ducks.

Florida's Mottled Duck is one example of a natural

ecological experiment in which change in isolated mallard

populations can be documented. Around the world, there are

more than 20 species mallards living in environments ranging

from Siberia to the tropical islands of the south Pacific

(Scott 1972, Weller 1975). Mallards are classic generalists

in that they are able to feed in an amazing variety of

habitats, nest almost anywhere, and synchronize their

reproductive cycle to either day length or water

availability (Witherby et al. 1939, Frith 1967, Weller 1975,

Palmer 1976, Soothill and Whitehead 1978, Bellrose 1980).

Through all these different habits and habitats, these birds

retain a remarkable similarity to each other. One must

wonder if differences among mallard species are due to

adaptation, which implies a distinct advantage from the

change, or if differences arise simply from non-selective

changes in the population. In this dissertation, I will

document differences (or similarities) between Florida Ducks

4

and other mallards, and discuss whether these can be

interpreted as true adaptations, or just change.

I capitalized the common names of all birds throughout

this dissertation to differentiate when I refer to a species

or a group. For example, when I refer to "Mallards," I am

referring only to A. platyrhynchos, rather than all the

ducks called mallards (Black Duck, Mottled Duck, Mexican

Duck and so on). Similarly, I do not capitalize "mottled

ducks" when I refer to the three North American species, but

do capitalize "Florida Duck" when referring to A. f.

fulvigula, and "Mottled Duck" when referring to A. f.

maculosa. Only bird names are capitalized.

Study Area

The population size of Florida Ducks probably varies

from about 7,000 to 20,000 breeding birds (Sincock 1957,

Johnson et al. 1984). In this paper, I refer to Florida's

Mottled Duck as the Florida Duck because of uncertainty that

trends observed will apply to the other mottled ducks.

Florida Ducks occur only in the peninsula of Florida,

southward of 29 N lat.

Brood rearing was studied principally on Lake

Okeechobee, in Okeechobee and Glades Counties, Florida.

Some unmarked broods were observed on the prairies north and

west of Lake Okeechobee. Ducks were collected principally

from the prairies north and west of Lake Okeechobee.

5

Southern Florida receives about 125 cm of precipitation

yearly with 25-30 cm falling during the dry season (November

through April) and 90-100 cm in the wet season (May through

October)(MacVicar 1983). Summers are hot and humid with

average temperatures of 28C (average daily maximum is 35C).

Winters are cooler and drier with average daily temperatures

of 16C (average daily maximum is 18C)(McCollum and Pendleton

1971, MacVicar 1983).

Rainfall was below normal during the study and regional

water levels were low (Marban et al. 1989). Most areas of

the prairies were dry. Lake Okeechobee was about 1 m below

normal water levels.

The prairies of southern Florida are flat grasslands

with saw-palmetto (Serenoa repens), cabbage palm (Sabal

palmetto) hammocks, and freshwater wetlands regularly

interspersed (VanArman et al. 1984). Soils are generally

sandy throughout with scattered pockets of organic peats and

mucks (McCollum and Pendleton 1971). Lake Okeechobee has

38,000 ha of freshwater marshes on its western border,

including submergent bed, emergent and moist soil

communities (Pesnell and Brown 1977, Milleson 1987).

Common plants in the lake marshes and prairies include

maidencane (Panicum hemitomon), torpedo grass (Panicum

repens), spikerush (Eleocharis spp.), sawgrass (Cladium

jamiacensis), pennywort (Hydrocotyle spp.), smartweed

(Polygonum spp.), cattail (Typha spp.), bullrush (Scirpus

6

spp.), pickerelweed (Pontederia cordata), duck potato

(Sagittaria lanciolata), willow (Salix spp.), hyacinth

(Eichhornia crassipes), and hydrilla (Hydrilla

verticillata), beakrush (Rhylncispora spp.), lotus (Lotus

spp.), and spikerush (Eleocharis spp.)(Sincock and Powell

1957, Pesnell and Brown 1977, Milleson et al. 1980,

Winchester et al. 1985).

CHAPTER 2
SURVIVAL, HABITAT USE, AND ECOLOGY OF
FLORIDA DUCK HENS AND THEIR BROODS

Introduction

The Florida Duck population may be declining (Johnson

et al. 1984, 1991). In 1984, the United States Fish and

Wildlife Service (FWS) classified the Florida Duck as a

Species of Special Emphasis in the Regional Resource Plan

and cited inadequate population data as an impediment to

knowledgeable management. Specific information about hen

and duckling survivorship during the brood rearing period is

needed, but is not available (LaHart and Cornwell 1971,

Johnson 1974).

Florida Duck broods and hens apparently use open,

shallow areas at night, but reports are anecdotal (LaHart

and Cornwell 1971, Johnson 1974). Banding crews report

frequently finding broods at night in very shallow water or

on mudflats (Fred Johnson, GFC, pers. comm.), but it is

unknown whether that is because broods simply are more

visible or if there is a real preference for these habitats.

An HSI model constructed for mottled duck breeding habitat

(Rorabaugh and Zwank 1983) relies mostly on data from the

Mottled Duck (A. f. maculosa) in Louisiana and Texas, thus

managers in Florida do not know if these data are suitable

CHAPTER 2
SURVIVAL, HABITAT USE, AND ECOLOGY OF
FLORIDA DUCK HENS AND THEIR BROODS

Introduction

The Florida Duck population may be declining (Johnson

et al. 1984, 1991). In 1984, the United States Fish and

Wildlife Service (FWS) classified the Florida Duck as a

Species of Special Emphasis in the Regional Resource Plan

and cited inadequate population data as an impediment to

knowledgeable management. Specific information about hen

and duckling survivorship during the brood rearing period is

needed, but is not available (LaHart and Cornwell 1971,

Johnson 1974).

Florida Duck broods and hens apparently use open,

shallow areas at night, but reports are anecdotal (LaHart

and Cornwell 1971, Johnson 1974). Banding crews report

frequently finding broods at night in very shallow water or

on mudflats (Fred Johnson, GFC, pers. comm.), but it is

unknown whether that is because broods simply are more

visible or if there is a real preference for these habitats.

An HSI model constructed for mottled duck breeding habitat

(Rorabaugh and Zwank 1983) relies mostly on data from the

Mottled Duck (A. f. maculosa) in Louisiana and Texas, thus

managers in Florida do not know if these data are suitable

8

for the Florida Duck (A. f. fulvigula). Mottled Ducks breed

largely in estuarine areas (Stutzenbaker 1988) and Florida

Ducks breed mostly in freshwater areas (Johnson et al. 1984,

1991), making comparisons problematic.

Florida Duck breeding also is of interest because it

occurs at the end of Florida's dry season. Nest hatch in

other North American Anatidae (including mallards) coincides

remarkably well with water availability and peak abundance

of foods (Bookhout 1979, Murkin et al. 1982), but Florida

Ducks hatch when wetland water levels are at their lowest

(Steiglitz and Wilson 1968, Thomas 1982). Florida Ducks

nest in greatest concentrations on the ephemeral ponds of

the prairies in south-central Florida (Sincock 1957, Johnson

et al. 1991). During some years, most wetlands on the

prairies dry completely, possibly leaving broods stranded.

Mallard and Black Duck broods apparently do not select

mudflat-type habitats (Reed 1975, Courcelles and Bedard

1979, Ringleman and Longcore 1982, Talent et al. 1982,

Mulhern et al. 1985, Monda and Ratti 1988). Presently,

there is no ecological explanation for Florida Ducks nesting

during the dry season.

I used radio telemetry to follow brood-leading hens on

Lake Okeechobee, Florida. I tested the hypotheses that low

brood-rearing success is limiting population growth in

Florida Ducks, that broods are most active during nocturnal

periods and that they preferentially use mudflat-type

9

habitats. The last hypothesis has a corollary that Florida

Ducks have similar brood-rearing habitats to Mottled Ducks,

Mallards and Black Ducks. Florida Ducks are expected to

differ because they live in a different geographical region

and because of their habit of nesting during the dry season.

I also compared these data with those in the HSI model for

the mottled duck (Rorabaugh and Zwank 1983) to test that

model for applicability to the Florida Duck.

Materials and Methods

Hen Capture

Twenty-one brood-leading female Florida Ducks were

captured from April through August in 1989 by nightlighting

from an airboat on Lake Okeechobee. Broods were captured

with the hen, and all were transported to the laboratory to

be processed. Each hen was fitted with a 25 g radio

transmitter (using a harness similar to that of Dwyer

(1972)), and ducks were released at the capture site at dawn

(within six hours). Transmitter packages apparently do not

impede normal activities in either Mallards (Cowardin et al.

1985) or Mottled Ducks (Baker 1983). No adverse effects

from the radio package were detected in this study.

I estimated age of broods from the relationship between

duckling mass and age for Mottled Ducks (Stutzenbaker 1988).

I was not able to determine the age of brood-leading hens by

plumage characteristics (Carney 1981, Stutzenbaker 1988) or

examination of the bursa. The use of plumage

10

characteristics to determe the age of Florida Ducks is not

considered reliable (Carney 1981), and bursas may not be

present in some first-year birds.

Survivorship

I located hens using a triple-beam yagi antenna and

attempted to count ducklings each time the brood was

located. I ceased locating hens during daylight periods

after 9 attempts because broods always were hidden in dense

cover, precluding a count of the ducklings. I subsequently

concentrated on nocturnal periods to insure collecting data

for survivorship estimates and habitat use during this more

active period.

Brood success was defined as the probability that one

or more members of the brood survives to flight stage; and

duckling survival was the probability that an individual

duckling leaving the nest would reach flight stage. I used

the Mayfield (1961) method to estimate duckling survival.

This estimator is used to estimate brood, duckling and hen

survivorship (Johnson 1979, Klett and Johnson 1982,

Ringleman et al. 1982, Ringleman and Longcore 1983, Kirby

and Cowardin 1986) because it relies on a calculated daily

survival rate that can be extrapolated over the entire

period of interest. Confidence intervals for daily survival

rates were calculated using formulas from Hensler and

Nichols (1981).

11

An average daily survival estimate was computed for

individuals in each brood, and these were averaged to obtain

one estimate for the brood. Survivorship of individual

ducklings within broods cannot be considered independent

from brood mates. Individual brood averages were combined

to calculate a weighted (by number of ducklings) mean

survivorship based on all broods (with a corresponding

variance).

The exact fate of a brood can be uncertain when only

the hen is marked. If the hen cannot be located, she may

have abandoned a live or destroyed brood, taken the brood to

an undetected location (one radio-marked brood-hen evaded

location for six weeks), or there could have been radio

failure. Two hens in this study had young broods that

disappeared on subsequent observations and were considered

lost. These broods were included in the "best-estimate"

calculation of brood survival. Three other hens (hence,

also their broods) disappeared and were omitted from the

best-estimate, but were considered destroyed for the "worst-

case" calculation. These broods were older and could have

been alive and abandoned, hence the worst-case estimate is

probably too low.

Brood success was calculated using a Kaplan-Meier

(1958) estimator modified for a staggered entry design

(Pollock et al. 1989) to estimate the percentage of broods

with at least one duckling reaching fledging. Age was used

12

to stratify broods for survival estimation. Precision of

these estimates is low because sample sizes averaged only 11

marked broods for any given time period. At least 20 at any

time is recommended. However, the use of this method is

preferable to a direct calculation of the percentage of

broods that survived because the latter almost certainly

would be too high. I used formula 3 in Pollock et al.

(1989) to estimate the variance:

var([t])=[S(t)]2[l-S(t)]/r(t)

and formula 4 to estimate 95% confidence intervals:

S(to)+1.96[var S(to)]1/2

where S was the survival function, to was the initial time,

and r was the number of animals at risk.

Estimating survival by counting unmarked ducklings of a

radio-marked hen had several potential biases. Survivorship

estimates may be positively biased when hens lose ducklings

to predation but adopt other ducklings between observations.

Survivorship estimates may be negatively biased because of:

1) uncounted ducklings due to visibility problems; 2) hens

abandoning a live brood before it fledges, incorrectly

indicating the brood was lost; 3) ducklings experiencing

delayed mortality due to stress of capture instead of

natural causes; 4) sublethal stress from the radio package

that decreases the hen's ability to care for the brood; and,

5) presumed loss of a surviving brood resulting from radio

failure or the brood moving to an undetected location.

13

Inaccurate counts also could bias survival estimates. In

all analyses, I omitted individual observations that were

suspect.

Habitat Use

Exact microhabitat use was impossible to assess because

I was unable to locate broods before they detected me.

Further, broods are mobile and often use more than one

habitat type. Accordingly, I determined the area of a

wetland the brood was using and quantified habitat

accordingly. Measuring an "area of use," although

subjective, was probably the most accurate representation of

the needs of a brood, considering broods use several

habitats for different activities (e.g., feeding, sleeping)

or at different ages (Lallart and Cornwell 1971, Ringleman

and Flake 1980, Paulus 1984).

Water depth was measured to the nearest cm at the spot

the brood first was sighted. Within an approximate 10 m

radius of the brood, average height of vegetation, and

average foliar cover, in categories of 0-5, 6-35, 36-65, 66-

95 and 96-100% cover, were estimated. Distance to escape

cover and the size of the opening being used also were

estimated. Dominant and subdominant plant species, and

wetland type according to the Cowardin et al. (1979) system,

also were recorded.

Habitat use by broods was compared between two regions

of Lake Okeechobee to test whether birds in disjunct areas

14

utilized habitats with similar characteristics. Because

vegetation foliar cover variables were replicated (in time)

for many hens, the data were bootstrapped by performing 10

Chi-square tests on one randomly selected observation from

each hen. I used t-tests (treating each observation of a

hen as a nested replicate (Littell et al. 1991) to compare

water depth, vegetation height and distance to cover between

the two areas. To test whether birds used habitats with

different characteristics during diurnal and nocturnal

periods, I used a binomial test pairing diurnal observations

(N=6) of birds with nocturnal observations of those same

birds (Zar 1984).

Results

Survivorship

None of the radio-instrumented hens (N=21) died during

brood rearing. The survival estimate during the brood-

rearing period was 1.0; consequently confidence intervals

could not be calculated. All hens dispersed from brood-

rearing areas shortly after broods reached fledging age.

The best estimate of duckling survival for a 56-day

brood-rearing period was 0.50 (95% CI = 0.38-0.66). The

worst case estimate for the 56-day period was 0.24 (95% CI =

0.17-0.35)(Table 2-1). Broods appeared to fledge at about

eight weeks of age as evidenced by final observations of

four flightless broods at seven weeks of age, four at eight

weeks, and one at nine weeks. All ducklings were considered

15

destroyed in two of 19 broods, yielding a Kaplan-Meier

estimate of brood survivorship of 0.82 + 0.073 (95% CI). If

three broods for which I had incomplete information were

destroyed, then brood survivorship was 0.62 + 0.080 (95% CI,

16 of 21 survived).

Adoption of ducklings by brood-leading Florida Ducks

appeared to be common. Older broods had more ducklings

(Table 2-2) and two broods increased in size during the

study, one from eight to 14 and the other from six to seven

ducklings. Further, I captured a hen with 19 ducklings, a

number greater than clutch size.

Habitat Use

Florida Duck broods utilized areas with dense foliar

cover during the day and more open areas at night

(P<0.05)(Table 2-3). Eight of nine diurnal observations

occurred in emergent vegetation that was robust and non-

persistent. In contrast, 49% of all nocturnal observations

were in habitats with less than 35% emergent cover, and 70%

were in habitats with less than 65% emergent cover.

Overall, broods used emergent marshes, and aquatic beds, 50%

of the time, respectively. Aquatic beds were composed of

either unconsolidated sediments mudflatss) or rooted

vascular plants (Hydrilla and occasionally sparse Lotus

spp), and usually had a water depth of 0-5 cm. Dominant

plants in the areas used by broods were Typha and Scirpus,

with others at frequent but lower occurrences (Table 2-4).

Table 2-1. Survival (S)
Okeechobee, 1989.

of Florida Duck ducklings on Lake

Daily S 56-Day S 95% CI 63-Day S 95% CI

Best
estimate 0.98775 0.50 0.378-0.664 0.46 0.335-0.631
(n-19)

Worst
scenario 0.97501 0.24 0.166-0.352 0.20 0.133-0.309
(n=21)

Table 2-2. Mean number of ducklings in newly-captured
broods of Florida Ducks on Lake Okeechobee, 1989.

Age (weeks

0 1 2 3 4 5 6 7 8

Mean
brood size 7.0 4.0 6.3 5.7 8.3 10.3 7.8

Number of
broods 6 4 3 3 3 4 5 -

Table 2-3. Characteristics of habitats in which Florida
Duck broods were located during nocturnal and diurnal
periods. Numbers in each cell are the total number of times
a brood was located in that category. Thus, single broods
may be represented in more than one cell.

NOCTURNAL PERIOD (n=51)

Vegetation height (cm)
Foliar Row
cover 0 1-50 51-100 >100 total
class (%)

0-35 11 7 3 4 25

36-65 6 1 3 1 11

66-100 6 1 3 5 15

Totals 23 9 9 10

DIURNAL PERIOD (n=9)

Vegetation height (cm)
Foliar Row
cover 0 1-50 51-100 >100 total
class (%)

0-35 1 1

36-65 -1 1

66-100 7 7

Totals 0 0 1 8

18

Cover used during diurnal periods was taller than that

used during nocturnal periods (P<0.05); 89% of the diurnal

observations were in vegetation >250 cm in height whereas

only 20% of the nocturnal observations were in vegetation

>100 cm. I located birds in areas of 0 cm vegetation height

(either on mudflats or prostrate hydrilla) 45% of the time

during nocturnal periods. At night, broods were found

within 10 m of cover 66% of the time and within 50 m 90% of

the time (Table 2-5). The maximum distance to escape cover

was 100 m.

Brood-use areas were characterized by very shallow

water. Broods were never observed in water more than 30 cm

deep (Table 2-6), and 63% of all observations were in water

<5 cm deep. The depth of water at brood locations did not

differ between day and night (P=0.32).

Eight and 10 broods, respectively, were tracked on

Indian Prairie (IP) and near the Harney Pond Canal (HPC) of

Lake Okeechobee. These areas are about 10 km apart. Broods

on IP used deeper water (P<0.05) and taller (P<0.05) and

more dense (P<0.05 for five of 10 bootstrapped tests and

P<0.1 for seven of the 10 tests) vegetation than on HPC

(Table 2-7) during nocturnal periods. There was no

detectable difference (P>0.16 in all bootstrapped tests) in

distance from cover of broods in IP or HPC.

Hens from many different locations lead broods to

specific locations that presumably were the most favorable

Table 2-4. Percentage occurrence of plant species found at
brood site locations.

Diurnal periods % Nocturnal periods %

Scirpus 67 None (mudflat only) 24
Typha 11 Hydrilla verticillata 24
Lotus 11 Lotus 16
Salix 11 Scirpus 12
Typha 10
Mixed grass/sedge 8
Other 7

Table 2-5. Percentage of time broods were found at
different distances from cover during diurnal and nocturnal
periods.

Time of Distance (m)
day 0 1-10 11-50 50-100

Diurnal 100 0 0 0

Nocturnal 36 30 25 9

Table 2-6. Percentage of time Florida Ducks broods were
located using different water depths. There was no
difference (P=0.32) between depth of water used diurnally
or nocturnally.

Time of Depth of water (cm)
day 0 1-5 6-10 11-15 16-30

Diurnal 22 33 11 11 22

Nocturnal 24 40 18 8 10

Table 2-7. Comparison of habitat characteristics used
nocturnally by Florida Duck broods on Indian Prairie and
near Harney Pond Canal on Lake Okeechobee.

Variable Indian Harney
Prairie Pond Canal

Mean water depth

Mean distance to cover

Mean vegetation height

9.6 cm

10 m

109 cm

Percentage use (%)

Vegetation
cover class

Vegetation
height class

0-5%
6-35%
36-65%
66-95%
96-100%

0 cm
1-199 cm
>200 cm

2.0 cm

22 m

33 cm

21

brood-rearing habitat in an area. Four of eight broods near

Fish-eating Bay moved as much as 3 km, from various

locations, to use the same small area that other broods were

using. Concentration of broods in certain areas also has

been observed by duck-banding crews (Fred Johnson, GFC,

pers. comm.).

Discussion

Survivorship

The hypothesis that low brood rearing success limits

population growth in Florida Ducks is unsupported by my

data. The survival rates estimated for Florida Duck broods

are comparable to those of other mallards (Reed 1975, Ball

et al. 1975, Ringleman and Longcore 1982) and should sustain

Florida Ducks under normal circumstances. However, adult

Florida Ducks appear to have the lowest annual survivorship

of any North American duck (Fred Johnson, GFC, pers. comm.).

With present estimates of annual survival for adults (S=0.4,

Fred Johnson, GFC, pers. comm.) and estimates of duckling

survival to fledging (S=0.50, this study), nest success

(assuming nine ducklings hatch per successful nest) would

have to be about 60% to sustain the population. This rate

is much higher than typically recorded for mallards (Balser

et al. 1968, Steiglitz and Wilson 1968, Duebbert and

Lokemoen 1980, Baker 1983, Cowardin et al. 1985,

Stutzenbaker 1988).

22

This assessment of the status of the Florida Duck

population must be tempered by several factors. The annual

survivorship estimates are derived largely from banding

studies conducted in the 1980's, a period of drought in

Florida. Florida Ducks and other waterfowl usually suffer

population declines during droughts (Sincock 1957, Lynch

1984) Still, their ability to rebound when wetter

conditions return is uncertain in the rapidly changing

landscapes of peninsular Florida.

My estimate of brood survivorship was taken only from

Lake Okeechobee. In years of normal rainfall, the majority

of the population is found on the prairie areas north and

west of Lake Okeechobee and in the St. Johns River marshes

(Sincock 1957, Johnson et al. 1991). Habitat, water

conditions and predator suites probably are different in

these areas and brood survivorship also may differ.

Further, duckling adoption, which occurs on Lake Okeechobee,

appears rare in Mallards (Eadie et al. 1988, Boos et al.

1989), and may not occur in areas with low duck densities

where there is less potential for brood contacts. Duckling

survivorship may suffer as a result (Ball et al. 1975).

LaIIart and Cornwell (1971) recorded decreasing brood size

with age in areas away from Lake Okeechobee, rather than

increasing brood sizes as reported here. Hence, it is

unknown how well these estimates of brood survival apply

elsewhere in Florida.

Habitat Use

The hypothesis that broods are active nocturnally and

use mudflats preferentially was supported. Broods were

located in open mudflat or shallow water areas 70% of the

time during the night. Although they used open areas at

these times, "escape" cover was usually within 50 m.

Although I occasionally observed unmarked broods during

daylight, I never observed marked broods and there appears

to be little diurnal activity.

Although there were significant differences between the

IP and HPC areas (Table 2-7) in nocturnal habitat use, the

differences probably were related more to the different

structure of the habitat between the areas than to

differences in habitat use by the ducks. Mudflats on HPC

were more open and broods were found farther from cover than

on IP. However, the overall pattern of activity on somewhat

open, shallow areas during nocturnal periods was similar

between the two areas. Ducks stayed in vegetation during

the day.

A pronounced nocturnal activity period and the use of

shallow water and open habitat largely contradicts the

hypothesis Florida Duck broods have similar habits to other

North American mallards. Mottled Duck broods in Louisiana

and Texas are more active during the day than night and,

while they seldom use mudflats, they do use shallow water

extensively (Baker 1983, Paulus 1984, Stutzenbaker 1988).

24

Like the Florida Duck, these Mottled Ducks were associated

with emergent vegetation in all habitats. Broods of North

American Black Ducks also use wetlands with an interspersed

of emergents or other cover plants but avoid open areas and

do not use unconsolidated sediments (Reed 1975, Courcelles

and Bedard 1979, Ringleman and Longcore 1982). Mallards

follow a pattern similar to Black Ducks by using shallow

wetlands with robust vegetation interspersed with openings

(Talent et al. 1982, Mulhern et al. 1985, Monda and Ratti

1988). Unfortunately, few of these studies examined

nocturnal activity even though Ball et al. (1975) and

Swanson and Sargent (1972) observed nighttime activity by

Mallards on the breeding grounds.

Use of mudflats by brood-rearing hens appeared

selective. Rainfall was very limited in 1989 and Lake

Okeechobee was much lower than normal (Marban et al. 1989),

creating mudflats as the water levels dropped. Hens had

myriad habitats to select, including dense vegetation, open

vegetation, or vast areas of open water. Mudflats comprised

much less than 1% of the habitat on Lake Okeechobee (John

Richardson, Univ. Florida, pers. comm.), so by using

mudflats more than 45% of the time at night, hens displayed

a distinct "preference" for this habitat type.

The desirable characteristics of mudflats for broods

are unknown. Abundant food appears to be a major factor in

habitat use by waterfowl broods (Talent et al. 1982,

25

Pehrsson 1984) and, whereas I suspect this is true for

Florida Ducks, no data exist on seed banks or invertebrate

populations in this habitat type for south Florida. Open

habitats mudflatss) might increase exposure to predators but

duckling mortality was not excessive in this study.

Conversely, predator approach could be detected easily.

Most mortality to ducklings in Louisiana occurred at night

(Paulus 1984).

It is unclear why brood activity was mostly nocturnal.

Predator avoidance does not seem compelling because there

appear to be numerous nocturnal and diurnal predators in

Florida. Heat avoidance or prey availability are potential

factors. It is conceivable that invertebrate prey would

migrate downward in the substrate to avoid midday heat and

be unavailable to feeding ducklings. Until mudflats are

sampled for food availability, these relationships will

remain unclear.

A predilection for using mudflats may be a clue as to

why Florida Ducks nest near the end of the dry season. As

ponds dry, mudflats become abundant and if there is a

special reason to use mudflats to raise broods, then it

behooves hens to time nesting to coincide with the

occurrence of mudflats. The risk of this timing is that

lakes often dry completely. In the present human-altered

environment, partly drained wetlands, which are becoming

increasingly common in Florida (Johnson et al. 1991), could

26

dry too soon to allow broods to fledge and serve as a trap

for Florida Duck broods.

The prairies, where most Florida Duck production

probably occurs (Johnson et al. 1991), were unusually dry

during this study and most ponds were dry during the brood-

rearing period. I observed only eight broods in two years.

Of these, five were on small ponds (ca 0.5 to 5 ha) with

mudflats. In the other three ponds, the broods were

observed in open areas with sparse vegetation. These

limited observations indicated habitat use by broods

probably was similar on the prairies to that on Lake

Okeechobee.

Habitat Suitability Index Models: Mottled Duck

Rorabaugh and Zwank (1983) published a Habitat

Suitability Index Model for the Mottled Duck based mostly on

data from the Mottled Duck in Louisiana and Texas. The

model does not describe some of the habitat needs of Florida

Ducks during the brood-rearing period. Variables in the

model are directed specifically toward brood-rearing

habitat, and if applied to Florida Ducks, may seriously

mislead the user.

Variable V4 calls for 45-55% "continually submerged

substrate covered by woody or herbaceous emergent

vegetation" for optimal habitat. This may be too narrow a

requirement because as long as Florida Ducks have an

adequately large area of mudflat-type habitat and cover-type

27

habitat (I do not know how large this requirement is), the

percent of each is trivial. Variable V5 describes escape-

type habitat and rates it of high value, which is proper.

However, neither variable specifically calls for mudflats

(or open habitat of any kind) or gives open habitat any

value. Having the V4 and V5 variables place most emphasis

on cover that can provide protection from predators probably

is adequate to describe the diurnal requirements of Florida

Ducks (or at least the needs for loafing, resting and hiding

areas), but neglects the nocturnal (activity and feeding)

requirements.

Variable V7, water depth, calls for areas where water

is not deeper than 30 cm at low mean tide, but would be

deeper most of the time. Most Florida Ducks do not nest in

tidal areas, but the depth requirement probably still is not

shallow enough. Florida Ducks never were found in water

more than 30 cm deep and used water less than 5 cm about 60%

of the time.

Summary

Brood-leading female Florida Ducks on Lake Okeechobee

used marsh habitats with very shallow water, or mudflats.

During daylight hours, broods stayed in dense vegetation and

at night were found in open areas less than 100 m from

cover. These habitat observations support earlier

predictions. Survivorship of hens and their ducklings

during brood rearing was 1.0 and 0.50, respectively. These

28

rates should sustain the population, but if adult Florida

Ducks have annual survivorship as low as predicted,

population decline is expected (Johnson et al. 1984). The

HSI for mottled ducks (Rorabaugh and Zwank 1983) must be

modified before it will apply to Florida Ducks.

CHAPTER 3
PHYSIOLOGY OF THE ANNUAL CYCLE OF FLORIDA'S MOTTLED DUCK

Introduction

Waterfowl undergo physiological changes to meet

energetic and nutritional demands during the annual cycle.

Much of the ecological theory about the changes that occur

during the year is based on the assumption that the observed

changes truly are adaptive traits. In this model, the

environment places demands on ducks that compel them to

undergo certain physiological changes to complete their

annual cycle. If this model is realistic, biologists should

be able to use information about the various environments

that waterfowl occupy to predict how they would respond.

Florida Ducks live in an environment very different from the

Mallard and Black Duck, and therefore, should have

predictably different physiological changes.

Mallards accumulate lipid reserves that are used during

migration. Post-migratory mallards regain spent lipid

reserves before winter and lose of lipid reserves again in

late winter (Heitmeyer 1988a, Hanson et al. 1990). Indeed,

Mallards and Black Ducks appear to have an endogenous rhythm

of increasing body mass in early fall, followed by loss of

weight during the severely cold mid-winter and subsequent

replacement of reserves in the spring (Owen and Cook 1977,

29

CHAPTER 3
PHYSIOLOGY OF THE ANNUAL CYCLE OF FLORIDA'S MOTTLED DUCK

Introduction

Waterfowl undergo physiological changes to meet

energetic and nutritional demands during the annual cycle.

Much of the ecological theory about the changes that occur

during the year is based on the assumption that the observed

changes truly are adaptive traits. In this model, the

environment places demands on ducks that compel them to

undergo certain physiological changes to complete their

annual cycle. If this model is realistic, biologists should

be able to use information about the various environments

that waterfowl occupy to predict how they would respond.

Florida Ducks live in an environment very different from the

Mallard and Black Duck, and therefore, should have

predictably different physiological changes.

Mallards accumulate lipid reserves that are used during

migration. Post-migratory mallards regain spent lipid

reserves before winter and lose of lipid reserves again in

late winter (Heitmeyer 1988a, Hanson et al. 1990). Indeed,

Mallards and Black Ducks appear to have an endogenous rhythm

of increasing body mass in early fall, followed by loss of

weight during the severely cold mid-winter and subsequent

replacement of reserves in the spring (Owen and Cook 1977,

29

30

Reinecke et al. 1982, Hepp 1986, Delnicki and Reinecke 1986,

Whyte et al. 1986, Heitmeyer 1988a, LaGrange and Dinsmore

1988, Loesch and Kaminski 1989). These authors hypothesize

that this pattern has the adaptive advantage of insuring

that ducks are pre-programmed to gain maximal reserves to

sustain them through the coldest part of winter and allow

them to rest during especially stressful periods (cold

fronts). Morton et al. (1990) questioned the universality

of this pattern and noted that at lower latitudes (in

Virginia) Black Ducks did not exhibit this pattern. In the

above examples, migration causes weight loss and cold winter

weather supposedly stimulates an endogenous, mid-winter

weight loss.

I hypothesize that Florida Ducks should gain fewer

reserves in early winter and experience less weight loss

because of less thermoregulatory stress and lack of

migration. The endogenous mid-winter weight loss seen in

other mallards should be absent. Lipids are the most labile

component of the body reserves and account for the largest

part of the body reserve changes (Owen and Cook 1977,

Reinecke et al. 1982, Whyte and Bolen 1984, Hepp 1986,

Delnicki and Reinecke 1986, Whyte et al. 1986, Heitmeyer

1988a, Ankney and Afton 1988, Loesch and Kaminski 1989).

Florida Ducks should acquire and deplete less lipid reserves

than mallards during the mild Florida winters and the

periods the mallards are migrating.

31

Conversely, Florida Ducks and other mallards should

have similar physiological changes during similar events in

the annual cycle. All mallards lay eggs and defend

territories during breeding, and therefore, have similar

requirements. The Anatidae lay large clutches compared to

other birds (Murton and Westwood 1977) and female Mallards

and Black Ducks obtain and store all the needed lipids and

about half the protein necessary to lay eggs before

initiating egg-laying (Krapu 1981, Reinecke et al. 1982).

Males lose weight during territory defense. I hypothesize

that Florida Ducks will follow the same pattern and

magnitude of lipid and protein change as in other mallards

during the breeding season.

Competing hypotheses exist about whether lipid or

protein reserves ultimately limit clutch size in ducks.

These hypotheses are based on the observation that waterfowl

lose lipid reserves during breeding, but maintain protein.

The protein-limitation hypothesis states that, during egg-

laying, hens rely on easily acquired lipid reserves while

feeding exclusively and inefficiently on proteinaceous food

(invertebrates) used to produce the clutch of eggs (Drobney

1980, 1982, 1991, Krapu 1981, Reinecke et al. 1982 and

others). Conversely, the lipid-limitation hypothesis states

that protein is easily obtained (as evidenced by little net

loss of protein reserves during egg-laying), and that

cessation of laying is caused not by scarcity of protein,

32

but rather depletion of lipid reserves (Ankney and Afton

1988, Afton and Ankney 1991, Ankney et al. 1991, Alisauskas

et al. 1992). A third hypothesis, the "migrational-

uncertainty hypothesis" (Rohwer 1992), states that fat

storage is a hedge against an unpredictable migratory

environment and waterfowl "dump" the fat into the clutch of

eggs after arrival on breeding areas, when migratory

uncertainty is past. This last hypothesis can easily be

tested on Florida Ducks because they have no migrational

uncertainty and should not need to "dump" fat into a clutch

of eggs. The other hypotheses do not seem mutually

exclusive and simply seek different explanations of the same

phenomenon, that of a negative energy and nutritional

balance during breeding in wild waterfowl (Eldridge and

Krapu 1988, Arnold and Rowher 1991). If Florida Ducks do

not differentially deplete either protein or lipids, then

the respective limitation hypotheses will appear invalid.

Wing molt is another physiological demand that all

mallards must meet, hence Florida Ducks should not exhibit

different changes in body reserves as other mallards. Post-

breeding mallards regain some body mass before wing molt,

during which they either retain a constant body mass or lose

a slight amount of mass (Young and Boag 1982, Pehrsson 1987,

Panek and Majewski 1990). After wing molt, mallards again

accumulate reserves that are used during migration and

wintering periods. During the wing, molt the breast muscles

33

atrophy and leg muscles hypertrophy. Molt creates a large

demand for protein for feather growth (Young and Boag 1982,

Pehrsson 1987, Panek and Majewski 1990) and Florida Ducks

should accumulate protein reserves before wing molt.

Internal organs also change to meet the varying demands

of the annual cycle and most of these changes should be

similar in all mallards. One of the largest influences on

size of the digestive organs of birds is the amount of fiber

in the diet (Miller 1975, Barnes and Thomas 1987, Kehoe et

al. 1988, Brugger 1991). Gizzard, intestinal, and cecae

mass tend to increase in mallards during the fall and winter

because they generally increase the percentage of fiber in

their diets (Miller 1975, Whyte and Bolen 1985, Heitmeyer

1988b). During spring and summer, mallards consume more

animal matter, which is lower in fiber and digestive organs

reduce in size. Florida Ducks follow this pattern in food

consumption (Morton and Uhler 1939, Beckwith and Hosford

1957, Steiglitz 1972, Montalbano 1980, Montalbano et al.

1983, O'Meara et al. 1982), and if diet does indeed strongly

influence the size of the digestive organs, then Florida

Ducks will follow the same patterns as the other mallards.

The mass of the kidneys, spleen, pancreas and liver

should increase in females during the metabolically active

egg-laying period and during periods of rapid body reserve

acquisition (Lucas and Stettenheim 1972, Drobney 1982,

Hazelwood 1986, Heitmeyer 1988b). Heart mass should

34

increase to compensate for increasing body mass and should

correlate with body mass during breeding (Krapu 1981). If

the correlation between body mass and heart mass is real, it

should continue during all seasons.

There is a large body of data on physiological changes

during the annual cycle in Mallards and Black Ducks, and

abundant hypotheses to explain the changes. The influence

of some of the mechanisms thought responsible for these

changes (e.g., migration, thermoregulation) can be tested by

observing Florida Ducks. For events in the annual cycle

that appear similar (e.g., breeding, wing molting), Florida

Ducks can be observed to see if these patterns are as

generally applicable as claimed.

Materials and Methods

I collected ducks from January 1987 through December

1988 from Okeechobee, Glades, and Palm Beach counties.

Collections were made after observing the ducks to determine

their pairing and flocking status. To prevent collection

bias, no birds were trapped or shot over bait or decoys

(Walsh 1971, Weatherhead and Greenwood 1981, Burnham and

Nichols 1985, Reinecke and Shaiffer 1988, P. N. Gray,

unpubl. data). I attempted to collect ducks according to

their availability in the field: if I observed mostly

isolated pairs, I collected mostly isolated pairs. Age of

Florida Ducks was determined by presence of the bursa of

Fabricus (Johnson 1961, Heitmeyer 1985) because wing

35

characteristics are unreliable (Carney 1981, Stutzenbaker

1988).

I collected: 192 ducks by shooting; four and seven

wing-molting ducks by nightlighting from airboats in 1988

and 1989, respectively; and 12 birds from hunters on several

south Florida lakes in December 1988. The hunter-shot birds

were of unknown social status and were treated separately in

all analyses.

All collected ducks were placed on ice. Body mass was

recorded after the ducks were transported to the lab excess

water was removed from the plumage. External measurements

included body length (from tip of bill to tip of pygostyle),

wing cord (from proximal end of carpo-metacarus to the tip

of the flattened, ninth primary), tarsus length, and

relative color and brightness of the bill and legs. Ducks

were completely plucked and reweighed to estimate wet mass

of the air-dried plumage.

The skin is the major fat depot on ducks (Whyte and

Bolen 1984, Heitmeyer 1988a); thus, all ducks were skinned

(except for the pygostyle and areas distal to the wrist) to

obtain wet-skin weight. Omental fat, another indicator of

fat reserves (Woodall 1978), was collected during dissection

of internal organs.

For estimating actual changes in lipid reserves, I used

the formula:

Total body fat = -43.7 + 1.27*(wet skin weight).

36

Although Whyte and Bolen (1984) derived this formula for

wintering Mallards (r=0.95, N=634), Florida Ducks probably

are morphologically similar enough to Mallards to justify

its use. The formula is linear, so the calculated curve is

similar to the curve of wet skin weights and will provide an

estimate of total body fat to compare with other mallards.

No statistical tests were conducted on estimates of lipid

content derived with this formula; this formula was used

only to estimate fat reserves for comparison with other

studies.

After skinning, I excised and weighed one breast muscle

(including the pectoralis, supercoracoideus and

corabrachialis) and all muscles attached to the knee joint

(between the tibia and fiblula and femur). These two muscle

groups are the largest protein deposits in ducks and often

hypertrophy or atrophy in response to demands throughout the

year (Ankney 1979, Krapu 1981, Drobney 1982, Reinecke et al.

1982, Pehrsson 1987, Hietmeyer 1988a). Ducks then were

frozen for later dissection of internal organs.

Ducks were thawed and dissected within one year of

collection. I recorded the mass (g) of the heart, gizzard

(full and empty), liver, spleen, kidneys, large and small

intestines (full and empty), cecae (full and empty), bursa

(when present), testes, ovary, oviduct, three largest

follicles and oviducial egg, if present. Length

measurements of intestines are problematic because of

37

gunshot separations and non-uniform stretching (Freehling

and Moore 1987), thus, mass was used as the analysis

variable. Length and width were recorded for the testes,

three largest follicles, and eggs in the oviduct. Organs

damaged by gunshot were excluded from analyses.

Food was removed from the proventriculus, sorted into

plant or animal matter, and wet weight was recorded. Food

mass from the proventriculus was added to the mass of the

contents of the gizzard, intestines and cecae, were added to

estimate total food mass in the duck. I subtracted the food

mass from the total body mass to obtain the true mass of the

duck.

Protein content is highly correlated with water content

of the carcass (Whyte et al. 1986, Heitmeyer 1988a),

therefore, all birds were reweighed after drying at 70C for

14 days, at which time water loss had ceased. Carbohydrate

content was not measured because waterfowl accumulate only

minor amounts (Drobney 1982, Heitmeyer 1988a).

For analyses, ducks were grouped into 11 "social

status" categories that reflect distinct chronological

events in the annual cycle of the ducks. These categories

are listed in chronological order for ducks that follow a

"normal" reproductive cycle.

These categories are as follows:

Flocked,_paired. These ducks were in flocks composed

mostly of pairs. Single drakes and hens occurred

38

infrequently. These flocks were the continuation of the

fall, paired category listed below, but were observed after

January 1. The division between the fall, paired category

and this category coincides with increasing daylength, when

ducks begin preparing for nesting.

Territorial (paired). These birds had left the flocks

and were isolated from others, often occupying their own

pond. Drakes were engaged in territorial activities such as

3-bird flights where they chased other pairs from the

territory.

Egg-laying/incubating hens and their attendant drake.

This period represented the part of the territorial period

when nesting occurred. Drakes that were defending

territories in the absence of the hen (presumably laying

eggs or incubating) were included in this category.

Brood-leading. No drakes were included in this

category because they do not actively participate in brood-

rearing. Instead, drakes moved directly from the

territorial category to the post-breeding category.

Post-breeding. These birds were collected after the

peak of the breeding season and their previous history was

unknown. Ducks in this category appeared to have no

breeding intentions because they were not paired or

territorial, and upon dissection the gonads were found to be

regressed.

39

Wing molt. This category included only adults that had

molted their primary and secondary feathers and were

flightless. Flying birds with blood quills (newly grown

flight feathers) were placed in the fall, single category.

Fall, single. These birds included both adults that

had finished wing molt and hatch-year birds. All Florida

Ducks appear single when finishing wing molt; they pair

after reaching some level of physiological conditioning or

social status.

Fall, paired. This category included post-wing molt

birds, whether adult or hatch-year. Most of these pairs

were in flocks. This category includes only ducks collected

before the winter solstice.

Fall, unknown. This category consisted of a sample of

hunter-shot birds from December and early January; hence

social status (pairing or flocking status) was unknown and

this category was created. Most of these ducks were

probably aligned most closely with the fall, paired birds

although some may have been single.

Flightless duckling. This category includes all

ducklings that could not fly.

Flying duckling. This category included only recently

fledged (and flying) ducklings that still had blood in their

primary and secondary feathers. If the flight feathers had

hardened, these birds were placed in other categories.

40

When I compared my data on Florida Ducks with mallard

data from the literature, the time periods described and

methods of collecting data varied among studies. To make

comparisons meaningful, I placed their birds into categories

parallel to mine. When necessary, I recalculated some

quantities, or estimated values from figures. For example,

my fall, single and paired categories were not used by most

authors, so I classified their birds from early fall periods

as equivalent to my single birds, and I placed their late-

fall birds opposite my paired birds (because most Mallards

and Black Ducks are paired by late fall (Bellrose 1980)).

No comparisons were made with any birds from pen studies.

To address whether ducks maintained body mass during

the wing molt, I captured 28 and 46 wing-molting ducks by

nightlighting from airboats in 1988 and 1989, respectively.

I recorded length of blood quill of primary feathers, tarsus

length, and body mass (to the nearest 10 g). Length of the

blood quills was used as the indicator of stage of wing

molt, and an ANOVA (Proc GLM, SAS Institute, INC. 1988) was

run to determine if early-, mid- and late-molting birds had

different body masses (tarsus as a covariate for body size).

These ducks were captured at Garcia Ranch (27 47' N lat., 80

37' long., about 70 km north of Lake Okeechobee), and were

banded and released at the capture site.

Change in body mass of hens and ducklings during the

brood-rearing period was assessed using 34 brood-leading

41

hens captured by night-lighting. Body mass, tarsus length,

and average mass of ducklings were recorded. External

measures are useful for correcting for differences in body

size when comparing body mass, lipid content, organ size and

so on (Whyte and Bolen 1984, Ringelman and Szymczak 1985,

Johnson et al. 1985, Trippel and Hubert 1990). An ANOVA

(Proc GLM, SAS Institute, INC. 1988) was used to compare

body mass of hens during time periods of: early- (ducklings

< 300 g), mid- (300-600 g) and late-brood-rearing (ducklings

> 600 g), using tarsus length as a covariate for body size.

Hens were marked and released. Two brood-leading hens were

collected for dissection.

Analyses

I used 2-factor ANOVAs (social status and year as

classes)(Zar 1984, Proc GLM, SAS Institute, INC. 1988) to

compare body mass, organs and indices of protein (water

mass) and fat (wet skin weight) between social status

classes. Protected LSD's were used to test for differences

between means (Carmer and Swanson 1973). Sexes were tested

separately. Total body length, instead of tarsus length,

was used in these analyses as a covariate for structural

size because it was a more powerful indicator of body size.

Generally, body length was a significant covariate only for

measures of body mass, fat and protein indices, but not for

internal organs (Table 3-1).

Table 3-1. Significance levels of an ANOVA on variables
among status categories in Florida Ducks during the annual
cycle. Status category and year are main effects with body
length as a covariate.

Male

Female

N Number of Status
categories category

Body
Year Length

N Number of Status
categories category

Uncorrected
body mass

Corrected
body mass

Dry mass

Water content

Wet-skin mass

Omental mass

Breast muscle

Leg muscle

Plumage mass

Testes

Ovary

Oviduct

Bursa

Heart

Liver

Kidney

Spleen

Pancreas

Gizzard

Small intestine

Large intestine

Cecae

114 9 0.01 0.01 0.01 101 11 0.01 0.02 0.01

0.01

0.01

0.05

0.01

0.03

0.01

0.01

0.01

0.01

0.13

0.01

0.01

0.02

0.40

0.03

0.01

0.01

0.02

0.01

0.01

0.24

0.01

0.07

0.69

0.57

0.91

0.48

0.94

0.96

0.01

0.72

0.99

0.66

0.04

0.32

0.01

0.06

0.01

0.01

0.09

0.01

0.41

0.63

0.02

0.01

0.01

0.48

0.32

0.05

0.27

0.70

0.82

0.03

0.42

0.85

0.18

0.59

0.01

0.01

0.01

0.01

0.18

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.65

0.24

0.01

0.01

0.81

0.06

0.10

0.29

0.01

0.59

0.06

0.20

0.01

0.77

0.72

0.92

0.02

0.02

0.07

0.02

0.08

0.81

0.03

0.01

0.79

0.06

0.02

0.32

0.01

0.92

0.98

0.01

0.01

0.45

0.15

0.56

0.26

0.48

0.01

0.03

0.05

0.96

0.14

0.09

0.03

0.11

101 9 0.07 0.92 0.31 96 11 0.01 0.34 0.47

Variable

Year

Body
length

Food content

43

Results

Overall Physiological Changes

All gross body measures (body mass, dried carcass

mass, water mass, wet-skin mass, plumage mass, breast muscle

mass and leg muscle mass) exhibited significant (P<0.05)

changes over time in both sexes except mental fat, for

which only drakes exhibited significant changes (Tables 3-2,

3-3 and 3-4). Between-year differences were detected in

both sexes only for water content and uncorrected body mass

(Table 3-1). Body length, was not a significant covariate

for dry mass, wet-skin mass or mental mass in either sex,

nor for plumage mass in females. All other covariate

effects on gross body measures were significant (Table 3-1).

The mass of the reproductive organs changed

significantly with status (P<0.01)(Table 3-5) and did not

differ between years or have a significant covariation with

body length. The mass of the heart, liver, kidneys,

gizzard, cecae, and small intestines changed (P<0.05) in

both sexes, but significant differences (P<0.05) occurred

only in the pancreas and large intestines of males (Tables

3-6 and 3-7). Spleen mass was the only measured variable

that exhibited no change over the entire annual cycle in

both sexes. Yearly differences in mass of organs were

variable; the heart, small intestine (P<0.05) and cecae

(P<0.06) showed differences for both sexes, and all others

differed for one sex but not the other. The covariate had

44

Table 3-2. Values of total body mass and food content in
the entire gut (g+SEM) of Florida Ducks during the annual
cycle. Means within each column followed by the same letter
do not differ (E<0.05, protected LSD test). Sample sizes
are in parentheses.

Uncorrected Corrected
Social body mass body mass* Food material
Status Male Female Male Female Male Female

Flocked, 1117+19a 989+17ab 1083+19a 952+18ab 36.3+4.5a 36.8+4.6a
paired (19) (22) (16) (22) (16) (22)

Territorial, 1071+14ab 949+18ab 1045+18abc 925+17ab 31.0+2.8a 28.2+3.0ab
paired (26) (19) (19) (18) (22) (18)

Egg-laying 992+18c 1016+22a 963+19c 994+25a 27.4+3.7a 25.7+3.0abc
(11) (10) (10) (9) (10) (10)

Brood- 814+104c 792+92d 22.0+11.1abc
leading (2) (2) (2)

Post- 1091+20ab 911+17b 1048+18ab 891+20bc 27.9+4.9a 20.1+4.6abc
breeding (13) (8) (11) (8) (11) (8)

Wing molt 1029+26bc 919133b 1003+24abc 910+32ab 26.7+3.5a 9.4+2.7c
(11) (8) (11) (8) (11) (8)

Fall, 1038+28bc 936+24ab 999+33bc 901+23b 46.9+8.0a 35.0+3.7a
single (15) (7) (13) (7) (13) (7)

Fall, 1101+28ab 982+25ab 1071+30ab 945+27ab 29.7+6.0a 25.9+5.5abc
paired (10) (10) (10) (9) (10) (9)

Fall, 1072+22ab 987+28ab 1057+28ab 960+27ab 20.8+4.0a 26.8+4.0abc
unknown (7) (4) (6) (4) (6) (4)

Duckling, 780+65c 806+49cd 29.9+6.8ab
non flying (4) (3) (3)

Duckling, 903+30d 787+25c 873+33d 740+19d 30.0+3.4a 16.5+2.9bc
flying (2) (7) (2) (5) (2) (5)

SCorrected body mass equals the uncorrected body mass minus the food content.

Table 3-3. Changes in fat indices (gSEM) in the carcass of
Florida Ducks during the annual cycle. Means within each
column followed by the same letter do not differ (P<0.05,
protected LSD test). Sample sizes are in parentheses.

Social Wet-skin mass Omental fat mass
status Male Female Male Female

Flocked,
paired

Territorial,
paired

Egg-laying

Brood-
leading

Post-
breeding

Wing molt

Fall,
single

Fall,
paired

Fall,
unknown

Duckling,
non flying

Duckling,
flying

114+7.8a 105+7.2abc
(16) (22)

92+6.1ab 91+4.3bcd
(25) (19)

66+4.0Ob
(11)

106+6.8a
(11)

74+6.7cde
(10)

47+12.0Oe
(2)

85+7.9cd
(8)

98+9.1a 106+15.5abc
(11) (8)

101+9.6a
(15)

117+10.9a
(10)

89+12.7bcd
(7)

121+12.9ab
(10)

113+8.3a 127+14.9a
(4) (4)

62+3.5b
(2)

77+18.2cde
(3)

57+5.7de
(5)

5.1+0.8a
(16)

3.4+0.6abc
(25)

1.0+0.4bc
(11)

4.0+0.9ab
(11)

2.8+1.0abc
(11)

3.6+1.Oabc
(15)

4.4+1.la
(10)

5.6+1.0a
(6)

0.6+0.01c
(2)

5.5+1.la
(22)

4.7+0.8a
(19)

2.1+0.5a
(10)

O. 10.Oa
(2)

2.9+1.0a
(8)

4.8+1.4a
(8)

2.6+1.la
(7)

6.4+2.0a
(10)

6.0+1.8a
(4)

1.5+1.2a
(3)

1.3+0.6a
(5)

CI h ^r X S~Ba I
-o m~ .8 x Q
s N A s 0 ca 0 ,
N .- .t- N a S > t +- n .- ct ) LA
+I1N +IA +50 IN +5i0 NO1 +LA
'0 *0u '0 LA LAv '0 1- I-

kma)z

0 0
u- U Q)

o 04

4I 9 e M

tn 0 4

0 Po (a
m 4. 41,

04-J O 4
o a
S4J -) 0

t 0 0i 0

-4 p
"- r.

(0 a)

Q) 4-4

al)10
(n *U1 O
o 4J
m :U 0

4 00I
m ag1

+1 0
C 10

E4 0 e )
J 4i U) 0

0 C0
U) 0t) 00

U 0

X! 4J tl
U 6 -P

10 ^C
EraB-r l-

0o I- r- t
- .- N MO a
+-'- +1- +Q0- +1-

u

U1
Ln
C0 I
+Irv
'0 '-
LA ^

a ma "0 t a o

+N+ N + +
to to t #0 +0 +0 lc # t #0
tMM M M MMM

4,
m

I

41
4i

u
U.-

0

U
(O
w 4
-3"

z

0 I
x

I

i>g
RI

o _
w Sr

*t' Na *a .^ "
-- 10 OLA -
+|- .-o +b +zo ++1o
N -t N-4 LyA N--
4t 4t .4.. '

+IN
LA
I'

.o -- ur
T T a 0 T O Ao G
N-' 1 N' Na .- +-' ?a N- a .-s N
+5N +50 +50 +5V +l5 0 l t+1 0- 4
ov- o- o '-' 'U 0o o o'>2 6 +5r si

-7 .0 .0 Q .0
. *'. .s Na .^ a ^ *1
N .'0 Ni, '-..- N,- +1.- M OiA. toj O -
+, -,- +5. 1, V5 5 '0
-0 NA _
tri tN ~=Vie.%= t' ~l +C,

U
0
a
+IN
0.

.0
+5' +5'0 +5 1 +IN +10 0c +1-4 +101 + fn + im +5LA
LAM in i0 gL LA -t

No ,0 s 0
'0> -0 '3

Jm n
-a -0
to (a a
- I-'^ .- a^ 5- a^ Wa
+t^ r- +5r +lco +5'.?
Nr r- .- s4r osv'
LA'' .. LA o
'0 '0 0o 10

(Q + U 0 0 X.
N0 0 3
X +UC.) + i+ I + ,.I

O ,.- .-- u u) C ... .. 0 N 0
N NNN N 1NN

#0 LU N Q LA 0 LA

+ +I L +- + ,,- N + A + 50 + ,0_ '

.0 .0 U
(U 0O -D
0. 'O,

+15- +)N +|--
10 '-' K0 LA -"
'0

- '0 o uLA '- -
+1'- +|fM *5.-
4' N d-
4 '4 10
~'G ctrN, t~u
+tt

Table 3-5. Changes in the mass of reproductive organs (gSEM)
and bursa in Florida Ducks during the annual cycle. Means within
each column followed by the same letter do not differ (P<0.05,
protected LSD test). Sample sizes are in parentheses.

Social Testes mass Ovary mass Oviduct mass Bursa mass
Status Male Female Female Male Female

Flocked,
paired

Territorial,
paired

Egg-laying

Brood-
leading

Post-
breeding

Wing molt

Fall,
single

Fall,
paired

Fall,
unknown

Duckling,
non flying

Duckling,
flying

2.0+0.5b
(16)

4.8+0.6a
(24)

6.3+0.8a
(10)

1.5+0.6bc
(11)

0.1+0.02bc
(11)

0.2+0.08bc
(15)

0.2+0.05bc
(10)

0.1+0.03bc
(6)

0.02+0.0c
(2)

0.8+0.2b
(22)

1.3+0.2b
(19)

22.1+2.9a
(10)

0.2+0.04b
(2)

0.6+0.2b
(8)

0.3+0.03b
(8)

0.2+0.05b
(7)

0.3+0.05b
(10)

0.3+0.09b
(4)

0.1+0.02b
(3)

0.2+0.04b
(5)

2.7+0.6bc
(22)

4.4+0.7b
(19)

29.5+2.1a
(10)

1.4+0.01bc
(2)

2.0+0.4bc
(8)

0.9+0.07bc
(8)

0.7+0.2c
(7)

0.7+0.1c
(10)

0.7+0.2c
(4)

0.3+0.08c
(3)

0.2+0.02c
(5)

0.09
(1)

0.5+0.08bc
(9)

0.5+0.2c
(3)

0.6+0.09 1.0+0.lab
(9) (4)

0.2
(1)

0.1
(1)

1.1+0.2a
(3)

0.6bc
(1)

0.9+0.03abc
(3)

1.1+0.03 0.8+0.2abc
(2) (5)

S0

0

100
0 C3

EO
mH 4
O-4 (
04 -

4-I

(I in

Q) ,I

10)
00 9
S 4

O

0 )
r-4

S4- r e )
en

0 N

mOmo
t 14

r-H *H

0 0 -4
0U (a

4
Q) 4)l
C--1

cn

tol

Q) .P

0 0

nr-I -A

o

fi ee
ty >-

(A

a

u-

IE

3E
.

c L
cu

A 0

4--.*

a) CD

.n .8 .8
'r 4 4 '0 N- LA
N-'. N-'- --- -- ^

4' N N N N
+!- +(-- +lin +10 +|C

rn C\1 (i V (M CM

'a m u 'a
XA w) LA N

a ah o in M 0 n a 4 (\ a
+IN +10 +10 +IN +1(0 +1C( +I- +10 +1.4
-(N Or- (7r- N '0 Nv 1- ') a rN -

u l
0' 0' 0' '0 c N c C6

0 a co -am c

+I0 +'Il +- : + L- +1- +A +10 +1 c0
lN- ON R-.- LA.- N-- 0'- 0- C-
0' 0' 00' N c- o0 0'

~-
0 0)
ih c > C
0 La 0) 0) C
Xi >0 i 4 0) 0 .5 -
IL u *- 6 TL 9 o

(a0
a 0 a a
CO m0 0 0 0 0 m 0'

+10 +1N + ++11 +I0 +1^ +I- +IM

NN N + N N
0 0 0 0 0 0" 0 0

0 0 0 0 0 0

+10 +114- +140 + +17' +I3 +i1 +10- +I'4
0 0- 0 0 0 0 0 0
0 N N 0 LA '0 .4

0-' 0-- 0 0-' 0 0" 0 0

+10 +)- +10 +1M +10 +10 +l. +10- +1i

0 0 0 0 0 O 0 (0

Ni 4') 0' M N Nn Ni n
m m t 0 m m
o N o i oi n o o o o
0' 0' 0-' 0-' O" 0C' 0-' 0-
+ln +10- +10 + +Ia0 +1o +14 +Io +|1

.+N i0. +10 0L +IGO + i'- -+10 +1 -
S -0 0 "! Z t *
u t a u 1 4a u a
Aj (\J M O n on o N

+ILA +1L0 +10 +1.- +I1- +iA +10 +17

S 1% 4 .4 4 C9 V!
NJ -' N -' r N n 0-
+in +li1 +10 +I* +100 +1CC +lin +1o +1T

.4 0X 0 LA M

No N N N IN M n N o
-N Nr Nr N yN M Nl Nr cv

Qu

+I'0 +11t + 1:

\j4 4(M

- o
**l +lin

0

-

S +IN

N

(a

rn
0 0O

a m

o

+ IN
S +IN
(0

vn M
0

0 -

+II+

0
-a

CO MC
+IM +U-1

in 0
a 0
o'
o -'

o0 '0

-C
oC 0-4-

4-4

0 49

rQ)

0) 0

0 0

+IH Q)

M0 $4

IC 0 9 0

U -Q I I
o 4 o 0

0 0 00 0" 0 C.Z
) 3 N +4-0 +410- +13 4-O w 0 0.

(1)) 0 o +10 10 f c

OO
t-0 0 0 -

-4 I I I4 I I
+m f o0t o o oo +

(n 44-1

.-. .- N) -' NM N N (0
-.O) +I'C +4, +8 I0 +.-0 + n +1 +0 +IN
0 (d C) va 0

0.'-4 U -o U 0 -o
I n- Nz 1.4 a)s ID ID 2 () C, D 5
4-1 7 +uM +I- tl V +: !5 +1 T I G +

0 04- u -U -

+r o Ic + + t

I N5 'ON u I N-v v s N 0
a 4 tt '4- 4 5

(A 0 I4 0 9- r0 z 4'
r-q Ln In co in tn
I 0

M +1 0 +1 0
(a 0 .2 g z o>

-H ) v ( ID ID I I D
,0 l p .1 C N Q C 'o 0
4~
-H 44 +\ *, +''-

to r0 0N n u A
ID 0 00 ID I IDC
#4 r'. +I- 0' +1'0 +IN ( LA N

0 00 4) 0- +.0 +Lo 0 4-
( a-- r-4 ) u0a 0 N 0 w

0 M-9 00L-4to
UIDO 0-- L. 4 QD lI C -- NU
04'M 04 LD O- D'.

50

significant effects (P<0.05) in six of 18 possible tests of

internal organs, but never differed simultaneously in both

sexes (Table 3-1).

Reproductive Organs

Reproductive organs of Florida Ducks exhibited a marked

seasonal pattern (Table 3-5). Testes were slightly enlarged

in spring males that were paired and still in flocks. Most

hens in these flocks still had inactive ovaries, but some

exhibited slight follicle development and slight enlargement

of the oviduct. Territorial males had larger testes

(P<0.05) than flocked birds, and females had larger oviducts

and ovaries (P<0.05) than fall birds. Territorial females

had follicles that measured 3-6 mm in diameter. At the

onset of egg-laying, ovaries and oviducts of hens were

larger than at any time of year (P<0.05), and the testes

also were at the annual maximum. Hen reproductive tissues

at this time totaled more than 50 g (ovary=22 g and

oviduct=29 g). Reproductive organs in females regressed

rapidly after egg-laying and were not significantly larger

during the brood-rearing period than the quiescent

reproductive organs throughout the fall (Table 3-5).

Reproductive organs of post-breeding ducks in both sexes

were not significantly larger than those of fall ducks.

Overall, males had a longer period of enlarged reproductive

organs than females.

Body Mass, Lipids and Protein

Florida Ducks began the year in flocks of pairs with

maximal body mass and reserves (Tables 3-2, 3-3 and 3-4).

Upon formation of the territory, drakes began a significant

(125 g) weight loss culminating during the egg-laying

period. This change included a significant loss of about 60

g of body fat and 39 g of water mass, the latter indicating

substantial protein losses. Females weight varied little

from the flocked period to the egg-laying phase, but carcass

composition changed through losing about 40 g of lipids

while gaining about 56 g of water (therefore, protein)

reserves. During the egg-laying period, females weighed

more than males because females were near their annual

maximum mass and males were near the minimum.

Females lost significant amounts of body mass during

incubation, reaching their lowest annual body mass at the

onset of brood-rearing (Table 3-8). I estimate that,

between the egg-laying period and the initiation of brood-

rearing, hens lost an average of 42 g of lipids (leaving an

estimated 8 g), 104 g of water mass, and 224 g total body

mass. Hens were able to regain a significant amount (70 g)

of body mass by the end of brood rearing (P<0.02)(Table 3-

8).

Drakes in the post-breeding period were significantly

heavier than in the egg-laying period, and hens in the post-

breeding hens were significantly heavier than in the brood-

52

Table 3-8. Changes in body mass (g) of hens during the
brood-rearing period. Means followed by the same letter do
not differ (P<0.05). Periods are based on the average size
(age) of the ducklings.

Brooding Range in
Period N Hen mass+SEM duckling mass

Early 18 792+12a 34-264

Mid 9 786+15a 355-558

Late 6 861+24b 688-970

53

rearing period. Hens weighed 51 g more at initiation of

wing molt than when fledging broods, and 58 g heavier than

post-breeding, single hens. Drakes initiated wing molt 56 g

heavier than post-breeding, single birds, and 95 g heavier

than territorial males. Both sexes maintained their body

mass through the first part of wing molt, but exhibited a

significant loss of body mass before completion (P<0.05)

(Table 3-9).

The post-breeding increase in body mass and subsequent

loss of body mass during wing molt was due partly to changes

in lipid reserves. Drakes and hens gained about 50 and 48 g

of lipid reserves, respectively, while adding only about 32

and 36 g water mass (during the post-breeding period when

ducks prepare for wing molt). Hens gained about 75 g of

lipids before wing molt and lost about 22 g of lipids by the

post-molting period, an average loss of about 0.8 g per day

(28 days). During wing molt, the growing remiges and

coverts on the wings had a maximum (measured) wet-biomass of

42 g for 50-60 mm primary quills (a sample of four birds

with longer quills had less feather mass).

After wing molt, both sexes gained body mass throughout

the fall. Lipid reserves and protein reserves also

increased. Paired birds tended to be heavier than single

birds in both sexes during the fall, but this difference was

linked to the propensity of hatch-year birds to be single,

and adults to be paired (Table 3-10). There was no

54

Table 3-9. Body mass (g) of Florida Ducks during wing molt.
Means within each column followed by the same letter do not
differ (P<0.05, protected LSD test). Periods are based on
length of the blood quill of the 10th primary feather.

Mate Female
Molt
period N Quilt length Mass+SE N Quill length Mass+SE

Early 20 0-35 1059+17a 16 0-31 931+19a
Mid 27 40-95 1016+13b 6 45-85 960+20a
Late 19 100-160 982+18b 2 100-115 813+18b

Table 3-10. Numbers of ducks with, and without, the bursa
of Fabricus in different social status categories. Presence
of a bursa confirms ducks are in their first year. Absence
of a bursa usually identifies adults, but could occur in
first-year ducks.

Male Female
Number Number Bursa Number Number Bursa
Social with without mass with without mass
status bursa bursa (g+SE) bursa bursa (g+SE)

Flightless -3 0 0.9+0.03
duckling

Flying 2 0 1.1+0.03 5 0 0.8+0.2
duckling

Fall, 9 6 0.6+0.09 4 3 1.0+0.1
single

Fall, 1 9 0.2 3 7 1.1+0.2
paired

Fall, 1 5 0.1 1 3 0.6
unknown

Flocked, 0 16 9 13 0.5+0.08
paired

Territorial 1 20 0.1 3 13 0.5+0.2
pairs

56

detectable mid-winter loss in body reserves and Florida

Ducks entered the breeding season with high levels of body

reserves.

Breast and Leg Muscles

Male breast muscles weighed less (P<0.05) during the

wing molt than at any other time (Table 3-4). The recovery

of this muscle mass continued throughout fall. Although

territorial males exhibited a significant loss of body mass,

their breast muscle mass did not decline. Females exhibited

lowest breast muscle mass (P<0.05) during brood-rearing,

wing molt and the fall, single period.

Leg muscles followed a cycle opposite that of breast

muscles, being largest (P<0.05) during wing molt in both

males and females (and paired fall males)(Table 3-4).

Females also had an enlarged leg muscle mass during the

brood-rearing period, followed by a significant loss and

regrowth of mass between the post-breeding and wing molting

periods. Leg muscle mass declined (P<0.05) in territorial

males when overall body mass also was declining.

Internal Organs

Heart. Heart muscle mass was correlated positively

with body mass (r=0.54, P<0.01, N=103 in males and r=0.66,

P<0.01, N=99 in females)(Table 3-6). The heart was

smallerduring brood rearing than during other periods in

females and during wing molt in both sexes (P<0.05). Heart

mass also correlated strongly with breast muscle mass

57

(r=0.62, P<0.01, N=109 in males and r=0.67, P<0.01, N=101 in

females)(Appendix 1).

Liver. Liver mass was greatest during the

metabolically active egg-laying and wing-molting periods in

hens and lowest during the brood-rearing period (P<0.05)

(Fig. 3-1). Liver mass in males also was high before,

during and after wing molt, but was near the annual low

during the egg-laying period (P<0.05). Liver mass was

highly correlated with kidney mass (r=0.57, P<0.001, N=107

in males and r=0.61, P<0.001, N=101 in females) (Appendix

1). Liver mass was not significantly correlated with any

other gross-body variable in males except for a weak

correlation with leg muscle mass (r=0.29, P<0.002,

N=108)(Appendix 1). Liver mass was correlated significantly

with most gross-body variables in females, and highly

correlated with leg muscle mass (r=0.54, P<0.01, N=101).

The correlation between the liver and the leg muscles

probably is not a cause/effect relationship and results from

both enlarging during wing molt.

Kidneys. Female kidneys reached their greatest mass

during egg-laying and wing molt and exhibited significantly

lower levels during other periods of the year. Kidneys in

males were heaviest during wing molt and uniformly low

during other periods (Table 3-6).

Spleen. Spleen mass did not differ in either sex

during the year (Table 3-6).

Figure 3-1. Mean mass (g) of livers (Y1 axis) and kidneys
(Y2 axis) during the annual cycle of Florida Ducks.
Categories on the X axis include: FLOCK Springtime birds
before territory formation; TERR Territorial birds; EGGS -
Egg-laying hens and accompanying drake; BROOD brood-
leading hens only; POSTB post breeding; WMOLT wing
molting birds; FALLS single birds in fall; and, FALLP -
paired birds in fall.

Liver mass

t--

A/

Kidney mass

I~ 7t

/

L........ LL... .. I
TERR EGGS BROOD POSTB WMOLT FALLS

TIME PERIOD

FEMALE LIVER
MALE LIVER

TISSUES
--- FEMALE KIDNEY
-- MALE KIDNEY

35

30

25

5

--4

i-4

Li -~

FLOCK
FLOCK

J-- 2
FALLP

60

Pancreas. Pancreas mass did not differ throughout the

annual cycle in females. However, in males, pancreas mass

was greatest during the wing molt and lowest during the egg-

laying period.

Digestive Tract. The total mass of the digestive tract

(gizzard, intestines and cecae) declined in both sexes from

territorial formation through brood-rearing, increased to

highest levels during wing molt, and declined during the

fall. Gizzard mass in both sexes decreased (P<0.05) during

the territorial period, reaching a low during the egg-laying

period, and increased (P<0.05) during wing molt and fall.

Mass of the small intestine of both sexes increased (P<0.05)

during wing molt and declined (P<0.05) during fall. Mass of

the large intestine of females showed no change (P=0.81) in

females during the year but that of males declined (P<0.05)

during the territorial period. Cecae mass in both sexes was

lowest during the egg-laying and brood-rearing periods, and

highest in fall.

Development of Young Florida Ducks

Flying ducklings older hatch-year birds were comparable

in structural size to adult birds and few differences in

body components existed (Tables 3-2 through 3-7). Tests for

differences between adult and hatch-year birds within status

classes were non-significant for virtually every variable

although almost every parameter measured was less in young

birds. Young birds probably were smaller and weighed less

61

than adults, but my sample size was inadequate to detect

these differences.

The significant difference in many variables between

paired and single birds in the fall is partly attributable

to the tendencies of hatch-year birds to be single and

adults to be paired. Nine of ten hatch-year males and four

of seven hatch-year females collected during the fall were

single (Table 3-10). Conversely, only one of 10 paired

drakes and three of 11 paired hens were hatch-year birds.

Hatch-year birds tended to weigh less than adults, and this

helped contribute to the lower recorded body mass of the

single birds.

Discussion

Changes in Body Mass

Changes in body mass during the year in Florida Ducks

followed a similar phenology, but not a similar magnitude,

as other mallards. The annual magnitude of change in body

mass of females was 26, 15 and 20%, in Mallards, Black Ducks

and Florida Ducks, respectively, of maximum body mass (Fig.

3-2). All females gained some body mass from early spring

until egg-laying andthen lost 26%, 9%, and 20%,

respectively, between laying and brood-rearing. Mallard

hens lose about 18% of their body mass during incubation

(Gatti 1983); Florida Ducks lost about 20% of their body

mass during egg-laying and incubation. Hence, the

hypothesis that breeding should place similar demands on

Figure 3-2. Mean body mass of three species of the mallard
complex throughout the year. Categories on the X axis
include: FLOCK Springtime birds before territory
formation; TERR Territorial birds; EGGS Egg-laying hens
and accompanying drake; BROOD brood-leading hens only;
WMOLT wing molting birds; FALLS single birds in the
fall; and, FALLP paired birds in the fall. Values taken
from the literature include: Owen and Reinecke 1979 All
male Black Duck values and FLOCK in females; Krapu 1981 -
TERR and EGGS in both sexes of Mallard and BROOD in females;
Reinecke et al. 1982 All female Black Ducks except FLOCK;
Whyte and Bolen 1984 Mallard FLOCK, FALLS, FALLP in both
sexes; Panek and Majewski 1990 WMOLT in both sexes.

B
1400

1300-

1200

1100

1000

900-

800B

1400

1300

1200

1100

1000

900

800

63

MALES

ody mass

'4-

FLOCK TERR EGGS WMOLT FALLS FALLP

TIME PERIOD

SPECIES
BLACK DUCK -- MALLARD
SFLORIDA DUCK

FEMALES

Body mass

--\4

FLOCK TERR EGGS BROOD WMOLT FALLS FALLP

TIME PERIOD

SPECIES
BLACK DUCK MALLARD
4 FLORIDA DUCK

Yr

64

and create similar physiological changes in Florida Ducks

and the other mallards was not rejected.

Clutch size in Florida Ducks averages 9.6 eggs

(Steiglitz and Wilson 1968). Eggs I weighed averaged 57 g

(n=6 nests, 32 eggs). Hence clutch mass probably is about

547 g, or about 55% of the hen's body mass. Mallard hens

lay an average of 9.7 50-g eggs that total about 46% of hen

body mass (Rowher 1988). Black Duck hens lay 9.5 62-g eggs

that total about 54% of hen body mass (Rohwer 1988). In

this comparison, energetic and nutritional demands of egg-

laying for these three species appear similar. Weller

(1975) noted that island-breeding Anatini lay larger and

fewer eggs than the parent stocks from which they were

derived, but this tendency was not apparent in Florida

Ducks.

Female Mallards lost about 12% of their body mass

during wing molt (Pehrsson 1987, Panek and Majewski 1990)

whereas Florida Ducks lost 13% (Table 3-9). Both Black

Ducks and Florida Ducks gained body mass before wing molt,

and all three species gained body mass during fall.

Physiological changes were similar in all mallards

undergoing wing molt.

Males of the three species also exhibited similar

changes in body mass (Fig. 3-2), losing mass during the

territorial period and gaining mass before wing molt and in

fall. The overall magnitude of change from maximal body

65

mass was similar in the three species: 9%, 12%, and 11% in

Mallards, Black Ducks and Florida Ducks, respectively. Male

Mallards lose approximately 12 to 18% of their body mass

during wing molt (Young and Boag 1982, Pehrsson 1987, Panek

and Majewski 1990), whereas Florida Ducks lost about 7%

(Table 3-9).

All three species increased their body mass during

fall, when migration occurs. Penned male and female Black

Ducks gained 21% and 16% body mass, respectively (Hepp

1986). Male Mallards gained about 8% of body mass

immediately after wing molt (Young and Boag 1982). Mallards

of both sexes in England gained mass during fall toward a

maximum in December; males increased 10% and females

increased 11% (Owen and Cook 1977). Female Black Ducks

gained 7% of body mass during fall (Reinecke et al. 1982).

Black Ducks in other geographical areas experience similar

increases (Hanson et al. 1990, Morton et al. 1990) as did

Mallards (Whyte et al. 1986, Heitmeyer 1988a, Hanson et al.

1990). Both male and female Florida Ducks were about 6%

heavier (non-significant) during the fall, paired period

than during wing molt. In all these cases, the Mallard and

Black Duck gained as much as, or more, body mass than

Florida Ducks during fall, supporting the hypothesis that

the migrants undergo greater physiological changes to match

their needs.

66

Florida Ducks appear to have no endogenous weight loss

during mid-winter, as is expected in migrant mallards (Owen

and Cook 1977, Reinecke et al. 1982, Hepp 1986, Delnicki and

Reinecke 1986, Whyte et al. 1986, Pattenden and Boag 1987,

Heitmeyer 1988a, Loesch and Kaminski 1989). The annual

cycle shown in Figure 3-3 groups Florida Ducks by month

instead of by social status to test the hypothesis that

there is an endogenous mid-winter weight loss. When my data

are grouped by month, females exhibit a non-significant

decrease in body mass during January and February. No trace

of this pattern was seen in males. Hence, Florida Ducks

follow the prediction of Morton et al. (1990) that ducks at

low latitudes should not experience a mid-winter, endogenous

weight loss.

Changes in Protein and Lipid

Female Florida Ducks, like Black Ducks, gained water

mass (protein) before egg laying and lost large amounts by

brood rearing (Fig. 3-4). The magnitudes of these changes

were similar, and the hypothesis that protein changes would

be similar during breeding cannot be rejected. However,

females gained little water mass before wing molt, casting

doubt on the hypothesis that ducks need to acquire protein

reserves before wing molt (Young and Boag 1982, Pehrsson

1987, Panek and Majewski 1990). Both Florida Ducks and

Black Ducks gained similar amounts of water in the fall but

the dramatic 22% loss of water reserves in the Black Ducks

Figure 3-3. Mean body mass of Florida Ducks grouped by time
of year. Although this method of classification groups
ducks from different social status categories, it shows
trends related to julian date, such as the large body mass
of females in December and subsequent lower mass in January
and February. Sample sizes for males and females,
respectively, by month are: January- 9, 9; February- 12, 9;
March- 5, 5; April- 5, 6; May- 11, 11; June- 15, 11; July-
11, 10; August- 15, 10; September-3, 3; October- 10, 7;
November- 8, 4; and December- 10, 9.

BODY MASS

1200

1 10O,.~

1000

41---+

900

800

700

600 --- 1
JAN FEB MAR APR MAY

JUN

JUL AUG SEP OCT NOV DEC

MONTH

i- MALES FEMALES

'A

-/

Figure 3-4. Lipid reserves and water content of females of
three species of the mallard complex throughout the annual
cycle. Categories on the X axis include: FLOCK -
Springtime birds before territory formation; TERR -
Territorial birds; EGGS Egg-laying hens and accompanying
drake; BROOD brood-leading hens only; WMOLT wing molting
birds; FALLS single birds in the fall; and, FALLP paired
birds in the fall. Data from the literature include: Krapu
1981 TERR, EGGS and BROOD for lipids; Reinecke et al. 1982
- all Black Duck lipid values and all water values except
FLOCK; Whyte and Bolen 1984 FLOCK, FALLS and FALLP for
lipids; Whyte et al. 1986 all Mallard water values;
Morton et al. 1990 FLOCK for Black Duck water.

70

FEMALES

ipid reserves

*

L
220

200

180

160

140-

120

100-

80-

60

40-

20

01

Water content
700-

650

600

-4-

400 '
FLOCK TERR S008 BROOD WMOLT FALLS FALLP
TIME PERIOD

SPECIES
* BLACK DUCK I MALLARD FLORIDA DUCK

i. I. i 1- L 1
FLOCK TERR EGGS BROOD WMOLT FALLS FALLP
TIME PERIOD

SPECIES
BLACK DUCK 4 MALLARD FLORIDA DUCK

FEMALES

Z

Q

71

during fall was attributed to unusually cold weather and ice

that prevented feeding and caused starvation in many birds

(Reinecke et al. 1982). Black Ducks in Virginia experienced

only a 2% (non-significant) loss of water mass during winter

(Morton et al. 1990). The 22% loss of body water by Black

Ducks was larger than the greatest annual loss (17%) in

Florida Ducks and reflects not a greater acquisition by

Black Ducks, but rather a greater loss during severe winter

stress that Florida Ducks do not encounter.

As hypothesized, changes in lipid reserves were

virtually identical in all three species during the

reproductive period (Fig. 3-4). All three species increased

lipid reserves in the fall, but with a marked difference in

magnitude; Mallards and Black Ducks acquired almost twice

the lipid reserves than Florida Ducks (210, 195 and 110 g,

respectively). These lipid reserves reached a maximum of

18%, 16% and 11% of total body mass in Mallards, Black Ducks

and Florida Ducks, respectively. Hence, migrants acquire

more reserves during the migratory period than Florida

Ducks. However, lipid reserves of Florida Duck hens in

December averaged about 148 g (14% of body mass), indicating

that Florida Ducks have the potential to acquire lipid

reserves of the magnitude observed in other mallards.

The hypothesis that Florida ducks would mimic the

classic pattern of Mallard hens by acquiring all the lipid

reserves and half the protein reserves needed for egg laying

72

before initiation of laying (Krapu 1981) was accepted.

Female Florida Ducks lost a significant amount of fat, and

gained a significant amount of protein prior to egg-laying

(Tables 3-3 and 3-4). However, these changes in Florida

Ducks were not accompanied by the overall weight loss

observed in Mallards (Krapu 1981).

Florida Ducks are flightless for about 4 weeks during

wing molt (Johnson 1973) and must rely on a single wetland

for all their needs. Before wing molt, both sexes gained

significant amounts of lipid reserves but did not gain

significant protein reserves (Tables 3-3 and 3-4). Lipid

acquisition occurred despite the fact that the growing

flight feathers are about 86% protein (Heitmeyer 1988c).

The hypothesis that ducks need to stockpile proteinaceous

reserves before wing molt appears unfounded.

Although these three species had overlapping values in

total changes in body mass, there appear to be differences

in the individual body components of protein and lipids

among species during certain parts of the year. Florida

Ducks maintained protein levels comparable to those of Black

Ducks during most seasons but did not lose as much protein

as Black Ducks during the stressful cold periods.

Concomitantly, Florida Ducks did not accumulate the same

magnitude of lipids in fall as the two migratory species.

During the breeding, brood-rearing and wing- molting

73

periods, all three ducks seem to be under similar stresses

and to respond in identical patterns.

It is unclear whether the differences between Florida

Ducks and the migrant mallards reflect (evolutionary)

genetic changes. Florida Ducks may have been extant since

the Pleistocene (Brodkorb 1957), giving them ample time for

genetic change. Winter temperatures often exceed 33C and

large lipid reserves might provide too much insulation.

Florida Ducks follow the same pattern of lipid loss and

protein maintenance during egg-laying as other mallards, so

I can neither reject or confirm the lipid-, or protein-

limitation hypotheses with my data. Lipid reserves

approached low levels, which may support the lipid-

limitation hypothesis, but protein levels in hens also

decreased dramatically, supporting the protein-limitation

hypothesis. However, both sexes lose body mass during

breeding, and whether they lose lipids because they are

feeding for protein, or vice versa, it seems that waterfowl

are unable to meet the short-term nutritional and energetic

needs of breeding in the wild.

The strategy of carrying large lipid reserves rather

than large protein reserves probably is related more to

physical than ecological constraints. Proteinaceous tissues

are more than 75% water by weight (Scanlon 1982, Afton and

Ankney 1991) and may be too heavy to stockpile. Lipids,

however, are calorically dense and have little free water

74

(Welty 1979, Heitmeyer 1985, Griminger 1986). Thus, lipids

are a more efficient body reserve. Large lipid reserves are

a logical solution to meet the negative energy and

nutritional balance not just of breeding, but also of wing

molting, migrating and other demands on waterfowl throughout

the year. Therefore, ducks build lipid reserves, instead of

protein, before wing molt. Emphasis on lipids rather than

protein also would explain why migrants obtain greater lipid

reserves, but not greater protein reserves, than Florida

Ducks during migration.

Breast and Leg Muscles

The prediction that breast and leg muscles of Florida

Ducks would change most during wing molt, and that the

changes would be similar in all mallards, was supported.

Both breast and leg muscles seemed to follow a use/disuse

type of pattern, growing largest when needed (e.g., legs

during wing molt and brood rearing) and reducing when least

used (e.g., breast muscle atrophy during wing molt). Breast

and leg muscles constituted 22% and 8% of total body mass,

respectively, in both sexes. As expected, both correlated

(P<0.03) with total body mass and water content (Appendix

1). All three species had large pectoral muscles at the

onset of breeding and maintained them in spite of large

losses in body mass during breeding. Females lost breast

muscle mass during brood rearing. Breast muscles in both

sexes then reached low levels of mass during wing molt and

75

subsequently regained mass in the fall (Fig. 3-5). These

trends were similar to the migrant mallards (Fig. 3-6).

Internal Organs

Gizzard mass in female Florida Ducks followed a pattern

very similar to that shown by Mallards and Black Ducks (Fig.

3-7), supporting the hypothesis that gizzard mass is large

when birds consume a high-fiber diet. All three species

showed a marked decrease in gizzard mass during egg-laying

and brood-rearing when hens consume larger amounts of

invertebrates (Steiglitz 1972, Krapu 1981, O'Meara et al.

1982, Reinecke et al. 1982, Gray, unpub. data). Gizzard

mass rebounded to large size during fall. The sharp drop in

gizzard mass in Black Ducks in fall (Reinecke et al. 1982)

was probably a reflection of the severe winter in Maine,

because Black Ducks in Virginia did not exhibit this loss,

maintaining a gizzard mass of about 44 g during winter

(Morton et al. 1990).

Small intestines in female Florida Ducks actually were

larger during egg-laying and wing molt than at other times

of year, contradicting the hypothesis that a low fiber diet

would make intestines shorter (Table 3-7). Female Wood

Ducks also enlarged their small intestines during laying

(Drobney 1982). Although a high fiber diet tends to enlarge

the small intestine (Kehoe et al. 1988, Brugger 1991), my

results supported the speculation by Ankney and Afton (1988)

that ducks eating a diet low in energy (mostly

Figure 3-5. Mean mass of one lateral breast muscle of three
species of the mallard complex. Categories on the X axis
include: FLOCK Springtime birds before territory
formation; TERR Territorial birds; EGGS Egg-laying hens
and accompanying drake; BROOD brood-leading hens only;
POSTB post-breeding; WMOLT wing molting birds; FALLS -
single birds in the fall; and, FALLP paired birds in the
fall. Values from the literature include: Young and Boag
1982 all male Mallard values; LaGrange and Dinsmore 1988 -
FLOCK for female Mallards; Krapu 1981 TERR, EGGS and BROOD
for female Mallards (these data calculated from dry mass
using formula from Scanlon 1982); Reinecke et al. 1982 all
female Black Duck values except FLOCK.

MALES

Pectoral muscle mass
140

A--

120

-- .-... ,

*

80 1 1 -
FLOCK TERR EGGS POST WMOLT FALLS FALLP

TIME PERIOD

SPECIES
-- MALLARD -A- FLORIDA DUCK

FEMALES

Pectoral muscle mass
150

i1-----

S-- -- --*

80

70 L-- ------- L ---- -
FLOCK TERR EGGS BROOD WMOLT FALLS FALLP
TIME PERIOD

SPECIES
BLACK DUCK I MALLARD
FLORIDA DUCK

140

Figure 3-6. Mean mass of the muscles from one leg of male
Florida Ducks and Mallards. Categories on the X axis
include: FLOCK Springtime birds before territory
formation; TERR Territorial birds; EGGS Egg-laying hens
and accompanying drake; POSTB post-breeding period; WMOLT
- wing molting birds; FALLS single birds in the fall; and,
FALLP paired birds in the fall. Mallard values from Young
and Boag (1982).

65-

60

55-

50-

45-

40

35

30'

TIME PERIOD

SPECIES

MALLARD --- FLORIDA DUCK

MALES

Leg muscle mass

TERRY EGGS POSTBWMOLT FALLS FALL
TERR EGGS POSTBWMOLT FALLS FALLP

FLOCK

Figure 3-7. Gizzard and liver mass of three species of the
mallard complex throughout the annual cycle. Categories on
the X axis include: FLOCK Springtime birds before
territory formation; TERR Territorial birds; EGGS Egg-
laying hens and accompanying drake; BROOD brood-leading
hens only; WMOLT wing molting birds; FALLS single birds
in the fall; and, FALLP paired birds in the fall. Values
from the literature include: Krapu 1981 TERR, EGGS and
BROOD for female Mallard gizzards and livers (these data
calculated from dry mass using formula from Scanlon 1982);
Reinecke et al. 1982 all female Black Duck gizzard values
except FLOCK; Morton et al. 1990 FLOCK values for Black
Duck gizzards and all liver values; Heitmeyer 1988b FLOCK,
FALLS and FALLP Mallard gizzards and livers.

81

GIZZARD MASS
FEMALES

GRAMS

A-

4\

lj
!/
7

20 -- - J
FLOCK TERR EGGS BROOD WMOLT FALLS FALLP
TIME PERIOD

SPECIES
BLACK DUCK I MALLARD FLORIDA DUCK

LIVER MASS
FEMALES

GRAMS

30-

25-
4'

4'

/ N

10 L J L-- J l
FLOCK TERR EGGS BROOD WMOLT FALLS FALLP
TIME PERIOD

SPECIES
BLACK DUCK i MALLARD h FLORIDA DUCK

82

invertebrates) also must lengthen their small intestine

toobtain as much from their food as possible. Presumably

that is why Northern Shovelers have a longer small intestine

than Mallards (and Florida Ducks) despite being smaller

bodied and eating a low-fiber diet consisting almost

exclusively of invertebrates (Ankney and Afton 1988).

Extending this hypothesis, Florida Ducks may have heavy

intestines in both sexes during wing molt because they are

eating many invertebrates in compensation for previously

acquiring large lipid reserves and relatively few protein

reserves.

Liver mass in female Mallards and Florida Ducks follow

a similar cycle (Fig. 3-7). No data were available to

determine whether the liver enlarges in Mallards or Black

Ducks during wing molt, although liver mass in wing molting

Canada geese increased slightly (Raveling 1979). As

predicted, the liver enlarged during metabolically active

periods, such as wing molting in both sexes, and egg laying

in females.

Heart mass is positively correlated with body mass in

Mallards (Krapu 1981) and Florida Ducks, supporting the

hypothesis that increased tissue mass requires greater blood

flow. As body mass decreased, heart mass also declined.

The spleen is used for blood production that may be useful

for feather development and growth (Lucas and Stettenheim

83

1972, Heitmeyer 1988b), but I detected no change in spleen

mass during the study.

Summary

Physiological changes during the annual cycle in

Florida Ducks were quite similar to those described in the

literature about Black Ducks and Mallards. The hypothesis

that the phenology of changes were nearly identical for all

parameters was accepted, but, migrants had larger changes in

lipid reserves during migratory and wintering periods,

rejecting the hypothesis that Florida Ducks would mimic the

migrants. Proteinaceous reserves in Florida Ducks changed

in a manner similar to migrant mallards except when the

migrants lost large amounts of reserves during severe

weather. Lipids probably showed the greater differences

because they are a more efficient reserve to store.

The hypothesis that internal organ changes would be

similar in all mallards was accepted because internal organs

changed in a virtually identical manner for all mallards.

Internal organs that did not change in mass during the year

included the spleen in both sexes, and the pancreas and

large intestine in only one sex. Internal organ changes

were highly correlated with changes in body composition,

that were correlated with events in the annual cycle.

CHAPTER 4
MOLTS AND PLUMAGES OF FLORIDA'S MOTTLED DUCK

Introduction

Molt patterns vary within the Anatidae. These patterns

presumably have adaptive importance and often are used to

assist taxonomic analysis of the Anatidae (Palmer 1976,

Livezy 1991). The Anserinae have one molt annually, most

Anatinae have two molts annually and the Mergini (also

Anatinae) have three molts annually (Humphrey and Parkes

1959, Salomonsen 1968, Weller 1980). The third molt in the

Mergini was first deduced from the striking patterns of the

Oldsquaw Duck (Clangula hymenalis)(Salomonsen 1949).

Although three distinct patterns are not readily noticeable

in the Anatini and the prevailing model of molt sequence

(Weller 1980) assigns Anatini 2 molts per year, Heitmeyer

(1985, 1988c) claims there are three body molts per year in

Mallards. I used the Florida Duck to test the hypothesis

that there are two annual molts in members of the mallard

complex.

Mallards and Black Ducks reportedly begin life in a

natal down, begin molting into juvenal plumage immediately,

and then molt into the first basic plumage near the time of

fledging. Females molt into the first alternate plumage the

first fall and into second basic plumage the following

84

__1_1_1_~ __1___ ^ ___IT____~_ _~__ ___1_________ _O~_lil_~_l_____i__-i~?ll-l-___~~__s~

CHAPTER 4
MOLTS AND PLUMAGES OF FLORIDA'S MOTTLED DUCK

Introduction

Molt patterns vary within the Anatidae. These patterns

presumably have adaptive importance and often are used to

assist taxonomic analysis of the Anatidae (Palmer 1976,

Livezy 1991). The Anserinae have one molt annually, most

Anatinae have two molts annually and the Mergini (also

Anatinae) have three molts annually (Humphrey and Parkes

1959, Salomonsen 1968, Weller 1980). The third molt in the

Mergini was first deduced from the striking patterns of the

Oldsquaw Duck (Clangula hymenalis)(Salomonsen 1949).

Although three distinct patterns are not readily noticeable

in the Anatini and the prevailing model of molt sequence

(Weller 1980) assigns Anatini 2 molts per year, Heitmeyer

(1985, 1988c) claims there are three body molts per year in

Mallards. I used the Florida Duck to test the hypothesis

that there are two annual molts in members of the mallard

complex.

Mallards and Black Ducks reportedly begin life in a

natal down, begin molting into juvenal plumage immediately,

and then molt into the first basic plumage near the time of

fledging. Females molt into the first alternate plumage the

first fall and into second basic plumage the following

84

__1_1_1_~ __1___ ^ ___IT____~_ _~__ ___1_________ _O~_lil_~_l_____i__-i~?ll-l-___~~__s~

85

spring. Males also molt into the first alternate in fall

but maintain it until their second basic is acquired before

wing molt. Adult females then continue a pattern of

acquiring basic in the spring and alternate in summer to

fall. Adult males gain basic before wing molt and alternate

in early fall (Witherby et al. 1939, Humphrey and Parkes

1959, Palmer 1976, Weller 1980). Florida's Mottled Duck is

a member of North America's mallard complex (Johnsgard 1961)

and reportedly has a molt sequence similar to the Mallard

(Palmer 1976).

Identification and enumeration of molts can be quite

difficult. Molt generations can overlap, and new molts may

start before the old molt has finished (Payne 1972).

Sequential plumages often are similar, and therefore,

difficult to discern (Heitmeyer 1985). Time frames for

individual molts can be confused when birds delay or suspend

molts that are in progress (Payne 1972, Richardson and

Kaminski 1992). Further, different feather tracts molt a

different number of times per year (Oring 1968, Humphrey and

Parkes 1963, Wishart 1985). Because waterfowl migrate,

observers in any one area are not able to trace birds

through the annual cycle. Florida Ducks do not migrate,

thus are useful for studying annual patterns of molt.

The discrepancies in the enumeration of molt sequences

are many. Stresemann (1963) contradicted Humphrey and

Parkes (1959) by stating that the juvenal plumage of

86

Mallards was replaced by the alternate. Humphrey and Parkes

(1963) reaffirmed their position by citing earlier authors

(Schioler 1921, Witherby et al. 1939). In male Mallards, "A

few odd body-feathers may occasionally be renewed in

spring," a time when the traditional model does not call for

molt (Witherby et al. 1939). Likewise, female Mallards

replace a few feathers in spring while replacing nest-down

(Witherby et al. 1939). Heitmeyer (1987) details the spring

molt in female Mallards that appears to be a third, separate

molt from the two described by Witherby et al. (1939).

Palmer (1976) and Weller (1980) outline two molts in female

Mallards, one in the spring and one in the fall, but do not

mention the late summer molt of females recorded by Witherby

et al. (1939).

Demonstrating the existence of three molts in the

mallard group would resolve some of these discrepancies.

First, it would explain why authors have outlined two annual

molts, but placed them at separate times of the year--

suggesting three molts total. The molt scheme outlined by

Weller (1980) has both sexes undergoing the pre-alternate

molt simultaneously in the fall. For pre-basic molt,

females and males purportedly molt at different times, in

spring and summer, respectively. In this scenario, females

breed in basic plumage whereas males breed in alternate

plumage. Then, males undergo wing molt while in basic and

females undergo wing molt in alternate. If there are three

87

molts, one in the spring, one before wing molt and the last

in fall after wing molt, both sexes could be labeled as in

the same plumage at the same time.

Another source of interest in molting is the energetic

or nutritional stress it places on ducks. Chapter 3

discussed loss in body mass and reserves during wing molt,

but the relationship between body reserves and the molt of

the body feathers is less understood. During the pre-basic

molt, Mallards need to allocate as much as 3 g of protein

per day toward developing feathers (Heitmeyer 1988c).

Considering the low concentration of protein in foods, and

loss of protein during assimilation of food, mallards may

need to consume more than 25 g of invertebrates daily for

this demand alone. Heitmeyer (1988a) detected lower body

weights during the middle of the pre-basic molt in hen

Mallards and concluded that the stress of molt caused this

weight loss. Heitmeyer (1985, 1988b, 1988c) further

speculated that waterfowl synchronize their body molts with

specific events in the annual cycle to avoid molting during

stressful times of the year, e.g., Mallard hens seemed to

delay pre-basic molt until after they were paired, but

completed it before migration. First, I will test the

hypothesis that molt in the Florida Duck is not conducted

during periods of large energetic or nutritional demands

(e.g., breeding, wing molting), but rather occurs between

them. Second, I will test the prediction that molting is

88

stressful enough to cause an overall loss of body mass,

lipid or protein reserves.

Materials and Methods

Molts and plumages were examined on the same ducks that

were examined in Chapter 3. Molt was detected by examining

feather replacement as identified by the presence of blood

quills. These new quills are not noticed immediately, but

become readily visible when they protrude about three mm

from the skin. Blood quills grow in the same follicle as

the existing feather and the old feather remains attached to

the new quill until the quill is about five mm long. Newly

grown feathers also may retain some living tissue near the

base that may not be observed when examining feathers,

especially in downy areas. Counting blood quills is a

conservative estimate of the total number of molting

feathers.

Molt was scored by counting transects of 50 feathers at

14 locations on the body and all feathers present on seven

areas that had less than 50 feathers, such as the primaries

and tail coverts. I calculated a percentage rate of molt

for each tract. I did not record stage of molt in the first

year of the study, hence I could not determine whether the

bird was beginning or ending the molt. Examining feathers

along transects is a more quantitative measure than the

"grab samples" recommended by Titman et al. (1990).

89

I saved a sample of feathers from 76 ducks that

represented all status and age classes, and included 24

adult females, 26 adult males, and ten and six first-year

females and males, respectively. These feathers were used

to identify the aspect (patterns) of the various plumages

and to help detect the actual time that each molt occurred.

Colors described were from Palmer (1962).

Terminology used in this paper is the same as outlined

by Humphrey and Parkes (1959, 1963) and further defined for

waterfowl by Palmer (1972); plumage refers "to a single

generation of feathers." Juvenal plumage is worn after the

natal down and before the first basic plumage. Basic

plumage formerly was known as the "eclipse plumage" and is

that plumage worn by the hen during nesting, and the drake

during wing molt. Alternate plumage (formerly nuptial

plumage) is worn by drakes during fall through breeding.

Alternate in hens is worn from after the wing molt until the

onset of breeding (Palmer 1972). All other molts are termed

supplemental, and include the spring molt of the drakes and

the molt preceding wing molt in the hens. All molts

apparently involve the head, body and tail, but only one

molt per year involves the wings. I label all molts as

"body molts" except the "wing molt," which will refer only

to the molt of the wing feathers themselves.

I used this terminology to remain consistent with the

literature. Unfortunately, this terminology creates

90

confusion because previous authors have recognized only two

plumages in the mallard group and I identified three molts.

These three molts are simultaneous in both sexes but must be

labeled with different names in the above scheme. Using the

above terms, in the spring, both male and female undergo

their pre-basic and supplemental molts, respectively. In

the summer, the molts before wing molt are labeled

supplemental and pre-basic in hens and drakes, respectively.

The fall plumage change (after wing molt) is called pre-

alternate in both sexes. The implications of opposite names

for simultaneous molts in spring and summer will be

discussed later.

Feather Tracts

The feather areas sampled were (Lucas and Stettenheim

(1972)) as follows:

Crown is the part of the capital feather tract on top

of the head and between the eyes.

Cheek is the part of the capital tract below the eye,

anterior to the ear and above the submalar tract.

Chin is the submalar tract.

Neck is the area along the border of the ventral and

dorsal cervical tracts, excluding the larger-sized feathers

of the body.

Upper back is the interscapular tract and the posterior

area of the dorsal cervical tract.

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
97	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file