Energetics and Water Turnover of Tippler Pigeons

By: GESSAMAN
Date: 1991
Category: General

FLIGHT PERFORMANCE, ENERGETICS AND WATER TURNOVER OF TIPPLER PIGEONS WITH A
HARNESS AND DORSAL LOAD

(I Received 9 August 1990. Final acceptance March 1991.)

JAMES A. GESSAMAN
Department 0f Biology and Ecology Center, Utah State University, Logan, UT
84322

GAR W. WORKMAN

Department of Fisheries a d Wildlife, Utah State University, Logan, UT 84322

MARK R. FULLER

U.S. Fish and Wildlife Service Patuxent Wildlife Research Center, Laurel, MD
20708

Abstract. We measured carbon dioxide production and water efflux of 12
tippler pigeons (Columba spp.) during seven experimental flights using the
doubly labeled water (DLW) method. Prior to the experiment birds were
randomly assigned to one of two groups. One group flew as controls (no load
or harness) on all seven flights. The other group wore a harness on two
flights, a dorsal load/harness package (weighing about 5% of a birds mass)
on two flights, and they were without a load in three flights.

Plight duration of pigeons with only a harness and with a dorsal
load/harness package was 21 and 26% less, respectively, than the controls.
Pigeons wearing a harness, or wearing a dorsal load/harness package lost
water 50-90%, and 57-100% faster, respectively, than control pigeons.

The mean CO, production of pigeons wearing a harness or a load/harness
package was not significantly different than pigeons without a harness or
load. The small sample sizes and large variability in DLW measurements
precluded a good test of the energetic cost of flying with a harness and
dorsal load.

INTRODUCTION

Homing pigeons (Columba livia) fitted with a harness and dorsal load were
slowed 27% and >3 1% on 90 and 320 km flights, respectively, and their
hourly CO, production was 41-52% higher during 320 km flights (Gessaman and
Nagy 1988). Those pigeons flew at 74 km/hr, which is faster than their most
efficient speeds. Pennycuick (1968) used aerodynamic equations to compute a
minimum-power air speed of 3 1 km/ hr and maximum-range speed of 58 km/hr
for pigeons, whereas a minimum-power air speed of 40 km/hr was measured by
Rothe et al. (1987) on pigeons flying in a wind tunnel. Gessaman and Nagy
(1988) concluded that high performance homing pigeons work longer and harder
during a long distance flight when wearing a dorsal load (e.g., a radio
transmitter), but that the effects of a dorsal load on flight performance
and metabolism of avian species that normally fly at more efficient flight
speeds would be less dramatic. We designed the present study using tippler
pigeons to overcome three limitations ofthe Gessaman and Nagy (1988) study:
(1) the inefficient flight of homing pigeons (i.e., the high cost of
transport), (2) the inability to track the flight of the pigeons, visually
or by telemetry, so flight distances and durations of both controls and
experimental birds could not accurately be measured, and (3) the lack of a
measurement of the effect of the harness alone (not including a load) on
flight metabolism.

Tippler pigeons (Culumba livia) typically fly in a flock of 3-12 birds at a
slower air speed than homing pigeons (pers. observation), and remain in view
throughout flights lasting several hours. Therefore, the flight duration of
each bird can be measured accurately and flight distance flown per hour by
control and experimental birds is nearly identical (although not
measurable), because they remain in a flock throughout most of the flight.

METHODS BIRDS

Juvenile tippler pigeons were purchased from suppliers in Arizona and
Wisconsin in August 1987 and May 1988. Their sexes were unknown. Following
the methods of Curley (196 l), birds were trained to fly for several hours
and then return to a loft in Smithfield, Utah. Training flights began in
June 1988, using 26 birds. After five weeks, 14 birds that did not fly
continuously for 22 hr or did not fly in a flock were removed from the
flock, leaving 12 pigeons for the experiment. Flocks of L 12 birds tended to
separate into ~2 smaller flocks after several minutes of flight, making it
difficult to visually track the groups, and the distances flown by the
different groups could differ. Before the experiment we assigned each bird
to either a control or an experimental group. The experimental birds flew
without any load in flights 1, 2, and 7; wore a harness only in flights 3
and 4; and wore a harness and load in flights 5 and 6. The control group
flew without a harness or load in all seven flights in which Vco, and water
flux were measured with doubly-labeled water (DLW), subsequently referred to
in this paper as the DLW flights. We released all 12 pigeons together. We
released all birds every other day for exercise flights between the seven
DLW flights.

Each pigeon wore a numbered, colored leg band and was uniquely color marked
on the under wings with Testers fast drying paint. Individual birds could be
identified during flight and their flight duration could be measured to the
nearest minute.

HARNESS AND LOAD DESIGN

The harness was a 2.5 x 6 cm piece of leather held flat on the back of the
bird by a loop of flexible, vinyl-coated, multi-stranded wire in front of
the wings and another behind the wings. The front loop fit around the crop
and the rear loop fit around the chest and over the keel (Fig. 1, Kenward
1987).

A 1 cm length of monofilament line connected the two loops ventrally along
the midline. We adjusted the harness so that an index finger (about 1.75 cm
diameter) could be easily inserted beneath the anterior half of the leather
piece and a little finger (about 1.5 cm diameter) beneath the posterior
half.

We attached a harness weighing an average of 7.2 g (SD = 0.1) to each
experiment bird 11 days before flight 3; the harness loops were completely
covered by breast feathers within a few days. These birds were then flown on
four different days before flight 3 to allow them to habituate to the
harness. We fastened loads to the harnesses of the experimental birds six
days before flight 5. Birds were then flown on two separate days before
flight 5 to allow habituation to the additional load. Each load was designed
to simulate the size and mass of a radio transmitter. Each load (made from a
plastic vial cap) weighed 6.5 g (SD = 0.4) had a diameter of 1.8 cm and a
thickness of 0.3 cm. Each load had a whip antenna (guitar string) 18 cm in
length. The load was weighted with steel shot so that the total weight of
the harness and load was 5% of the body mass (Mb) of the pigeon (Fig. 1). A
load was fastened to the leather patch with Velcro and oriented so that the
whip antenna extended over the tail. We removed the load/harness from the
experimental birds one day before the seventh flight.

CO, PRODUCTION AND WATER FLUX

We measured rates of CO, production and water flux of each pigeon with
doubly-labeled water (Lifson and McClintock 1966, Nagy 1980) during seven
flights between 16 July and 2 September 1988. Before each DLW flight we
weighed each bird and injected each with 0.4 ml of isotopic water
(containing 95.9% I80 and 0.3 mCi3H) into the abdominal cavity along the
midline, midway between the cloaca1 opening and the posterior edge of the
keel. One to 1.5 hr later, we punctured the brachial wing vein of each bird
with a 2 1 -gauge hypodermic needle and collected the blood in 2-3
heparinized glass capillary tubes. We flame-sealed the tubes with a propane
torch within 15 min. Prior to flights 1 and 2, birds were injected the
previous evening, blood was sampled
1 hr later and they were held in carrying cages for 11 hr and 9.5 hr,
respectively, before releasing them the next morning. For the other flights,
we injected birds about 05:OO hr, sampled blood 1 hr later, and released
them as a group near sunrise about 15 min after collecting the last blood
sample.

Typically the birds flew in a flock for > 1.5 hr. During the next few hours
birds landed in small groups or alone on one of four perches (the loft roof,
a house TV antenna, power lines, or a barn roof). The birds usually sunned
for > 1 hr before entering the loft. After entering the loft, we weighed the
birds, collected a second blood sam548

J. A. GJZSSAMAN, G. W. WORKMAN AND M. R. FULLER

FIGURE 1. Radiographs of the dorsal and lateral surfaces of a tippler pigeon
wearing a mock radio transmitter attached to a harness. (See text for
description of the harness). Two features of the mock transmitter show up in
the radiograph: (1) steel shot used to adjust the total weight of the
harness and transmitter to about 5% of the body mass of the pigeon, and (2)
the whip antenna (a piece of guitar string) that extended beyond the tail of
the pigeon. ple, and provided water and food. We micro-distilled pure water
(Wood et al. 1975) from each blood sample and assayed tritium activity of
each distilled water sample using Beckman Ready Safe scintillation cocktail
and a liquid scintillation counter (Beckman LS 5801). A separate portion of
each distillate was assayed for I*0 by K. A. Nagys laboratory at the
University of California at Los Angeles by converting I*0 to *F (by
cyclotron- generated proton activation of *O to 18F) and subsequently
counting the 18F in a Packard Autogamma system (Wood et al. 1975). We
calculated water flux and CO, production from the isotope measurements using
the appropriate equations (Nagy and Costa 1980, Nagy 1980). For flight 1, we
estimated total body water (TBW) volume from the dilution of tritium
injected into the bird about 1 hr before the flight. In the remaining
flights we did not measure the TBW of each bird; the mean value of TBW
before flight 1 was used in calculating water flux and CO, production for
these six flights. We assumed that fractional water content of a bird
remained constant through time, and we calculated TBW at recapture as the
product of body mass at recapture and fractional TBW of the animal at time
of injection. Also, we assumed any change in absolute TBW volumes was
linear, through time. The DLW method measured CO, production between the
times of the initial and final blood sample, i.e., a sum of CO, production
that occurred during rest and during flight. We computed CO, production
during the flight by subtracting an estimate of the birds CO, production
while at rest from the total CO, production between the initial and final
blood samples. We estimated an hourly rate of CO, production during rest,
i.e., resting metabolic rate (RMR), by measuring the resting rate of CO,
production of each pigeon in an open circuit respirometer at 2oc, and by
measuring the average daily rate of CO, production, i.e., average daily
metabolic rate (ADMR) by DLW of pigeons confined to the loft for 24 hr. ADMR
was measured on a day between experimental flights. RMR was measured for 2
hr about two months following the experimental flights. Mean ratio of ADMR
to RMR was 2.17; we selected 2 times RMR as the best estimate of the CO,
production of pigeons resting outdoors either in or out of the loft. Several
of the calculated values of CO, during flight were unexpectedly very high or
very low and we questioned their validity. In the literature, ratios of
flight metabolism (during continuous flapping flight) to RMR for pigeons
range from 8.0 (LeFebrve 1964) to 32.4 (when the birds carried a harness and
dorsal load that weighed 5% of the birds body mass, Gessaman and Nagy 1988).
We evaluated the probable validity of these high and low values by dividing
them by RMR. If a value of flight metabolism in our study was less than 7.5
times RMR or more than 34.0 times RMR we regarded it as an error and it was
eliminated from the data set.
In a few instances, the sample size was reduced more when the quantity of
water distilled from the blood was inadequate for isotopic analysis.

The DLW method also measures water influx (water gained from food, drink,
oxidative water and air breathed in) and water efflux (water lost by
evaporation and excretion) between the time of the two blood samples. We
measured water flux of pigeons on seven flight days and during one 24-hr
period when they were confined in the loft. During the 24-hr confinement in
the loft the pigeons had access to food and water, but not between the time
of initial and final blood samples on the seven flight days. We calculated
water flux of a pigeon during a specific flight as the difference between
water flux measured between initial and linal blood samples taken on that
llight day and an estimate of the water flux of that pigeon when resting
during the same period. We estimated water efflux occurring between the
initial and final blood samples when the birds were at rest by multiplying
the total duration of rest (hr) by the hourly rate of water efflux measured
on pigeons confined to the loft for 24 hr. We could not calculate water
influx during a flight, however, because we did not have an estimate of
water influx of resting pigeons without food and water (i.e., our only
measure of the water influx of resting pigeons was from pigeons that ate and
drank during confinement in the loft for 24 hr).

Flight time, CO, production and H,O efflux were each analyzed separately
with one-way ANOVA of 14 treatments; the data from the experimental and
control groups within a flight were each a data subset, resulting in 14
subsets (treatments) for the seven flights. Linear contrasts were computed
to compare: (1) the two subsets of birds wearing only a harness with the 10
subsets of birds without a harness or load, (2) the two subsets of birds
wearing a harness and a load with the 10 subsets of birds without a harness
or load, and (3) the two subsets of birds wearing only the harness with the
two subsets of birds wearing a harness and load.

RESULTS CO, PRODUCTION AT REST

The average RMR of 10 pigeons weighing 3 17.8 g (SD = 25.9) was 0.8017 ml
CO, g-l hr-I, and the mean ADMR (n = 20) of 14 birds (six birds were
measured twice), weighing an average of 265.98 (SD = 23.1) was 1.7397 ml CO,
g-l hr-I. The ADMRRMR ratio averaged 2.17 (SD = 0.88) and ranged from 1.08
to 4.6 for individual birds.

CO, PRODUCTION DURING FLIGHT

Mean CO, production of control birds for the seven DLW flights ranged from
13.0 f 2.7 to 15.5 + 6.4 ml g- hr-I (Table 1). Mean CO, production of
experimental birds during flights 1, 2, and 7 was less than that of control
birds, but the differences were not significant (P > 0.05). The CO,
production of pigeons wearing either a harness or a load attached to a
harness did not differ significantly from the control pigeons (F though the experimental birds produced, on the average, more CO, when
wearing a harness or a load/harness package than control birds in flights 3
through 6. Average flight metabolism/RMR ratios of control birds for flights
1,2 and 7 combined and for all seven flights combined were 17.2 (SD = 3.0, n
= 9) and 17.8 (SD = 4.0, n = 22) respectively. For experimental birds the
ratio was 14.6 (SD = 2.8, n = 9) for three control flights (flights 1,2 and
7), 21.3 (SD = 5.8, it = 8) when carrying a harness (flights 3 and 4) and
21.2 (SD = 4.7, n = 7) when carrying a load/harness package (flights 5 and
6) (Table 1).

We believe that the present study is the first to measure flight metabolism
with DLW several times on the same individuals. We expected that the rate of
carbon dioxide production @co2 of a control bird would be similar from
flight to flight. This was the case with Pigeon 1 and Pigeon 2, the SD of
the mean Vco, was 7.4% (n = 4)

550 J. A. GESSAMAN, G. W. WORKMAN AND M. R. FULLER

TABLE 1. Mean rate of metabolism( + SD) [ml CO, hr-l g-l] of two
groups of tippler pigeons that flew together in the same f lock during seven
flights.

Metabolism was measured with doubly-labeled water. Experimental birds wore a
harness during flights 3 and 4. During flights 5 and 6 they wore a mock
transmitter attached to the harness( this load weighed about 2.5% of the
birds M,). In flights 1, 2, and 7, experimental birds were not wearing a
harness or transmitter. Control birds flew unencumbered in all seven
flights.

Plight Type of treatxnt to ID experimental birds

Plightm etabolism[ X0, hr g-l]

Experimentabl irds (n) Controlb ids (n)

Plight metablism/RMRb

Experimental birds (n) Control birds (n)

1 none 11.5 * 1.0 (2) 13.0 & 2.7 (3) 14.7 f 2.0 15.9 + 2.9
2 none 11.1 ?z 3.1 (5) 13.6 + 2.0 (4) 14.2 + 3.7 16.8 + 3.4
3 harness 17.4 -t 7.7 (3) 14.3 f 2.1 (4) 21.0 + 8.5 17.5 + 3.7
4 harness 16.4 -t 3.0 (5) 14.9 * 0.1 (2) 21.5 + 4.8 18.6 f 0.2
2 Tp + harness 148.64 ?+ 43.39 (43) 145.05 + 26.54 (34) 2139..72 +f 25.24
189.0 + 37.38
7 none 11.8 * 1.2 (2) 15.3 !z 0.2 (2) 15.7 + 1.4 20.1 + 0.4 p T = a mock
transmitter.

bR MR = resting metabolic rate, measured on these pigeons in the laboratory
and 10.1% (n = 5) respectively (Table 2). This was not the case with Pigeons
3, 4, and 5, where the SD was 36.5% (n = 4) 23.4% (n = 4) and 15.2% (n = 3),
respectively, of the mean Vco,. The large variance of bco2 of Pigeons 3, 4,
and 5 is undoubtedly a reflection of the low turnover of IsO during the
flights (discussed below).

WATER EFFLUX

The average rates of water efflux from control birds in flights 1, 2 and 7
(546.7 ml H,O kg- d-l, SD = 218.9, y1 = 9) and in flights 3 through 6 (577.6
ml H,O kg- d-l, SD = 212.6, n = 12) were similar (P > 0.05) (Table 3). For
experimental birds the average rate of water efflux was 404.9 ml H,O kg- d-l
(SD = 86.5, n = 9) in control flights 1,2, and 7, 780.2 ml H,O kg- d-l (SD =
246.5, n = 8) in flights 3 and 4, and 821.6 ml H,O kg-l d-l (SD = 231.5, n =
7) in flights 5 and 6. Therefore the rate of water efflux of experimental
birds increased about 90% (P = 0.0 14) with a harness and about 100% (P =
0.0 11) with a harness and a load. The average water efflux of all birds
flying without a load (5 16.4 ml H,O kg- d-l, SD = 102.3), also was 50% less
than that of birds wearing a harness (776.8 ml H,O kg-l d-l, SD = 19.1, F =
37.6) and 57% less than those wearing a load and harness (8 11.6 ml H,O kg-
d-l, SD = 101.6, F = 4.6).

FLIGHT PERFORMANCE

The flock flew in broad circles, ranging in width from 50 to 200 m.
Typically the flock flew 75 to 125 m above the loft area for several minutes
after release, and then moved to higher altitudes (altitudes were not
measured) for 15 to 30 min. On some flights the flock moved alternately
between low and higher altitudes, remaining at each for > 15 min. The birds
were usually in view throughout the entire flight, although at the high-

TABLE 2. CO, production (VcoJ [ml CO, hr-l g-l] during flight of five
control pigeons.d The ratio of flight CO, production to CO, production
during rest is in parentheses. Vco, during llight (tight icolresting vco,)
Plight ID Pigeon

1 Pigeon 2 Pigeon 3 Pigeon 4 Pigeon 5

1 HVa NM 10.1 (12.6) 13.7 (18.2) NM

2 12.5 (15.6) 16.7 (21.7) LV LV 12.9 (14.0)

3 13.3 (16.6) 17.1 (22.2) 14.5 (18.0) NMb 12.2 (13.2)

4 14.8 (18.4) NMb 14.9 (18.7) LVC LV

5 13.2 (16.5) 16.8 (21.8) NMb 11.9 (15.7) NM

NMb 13.3 (17.2) 23.9 (29.8) 8.6 (11.3) 16.2 (17.6)

7 NMb 15.2 (19.8) NMb 15.4 (20.4) NM

K (*SD) = 13.5 + 1.0 15.8 k 1.6 15.9 t 5.8 12.4 k 2.9 13.8 k 2.1

= HV = high value. The value divided by RMR was greater than 34.0 and was
removed from the data set. NMb = no measurement i.e., either the bird did
not fly or the quantity of isotopic distilled H,O was too small to analyze.
c LV = low value. The value divided by RMR was less than 7.5 and was removed
from the data set. d CO, production of individuals on some flights was
either not measured or was reasonably high or low and, therefore, was
removed from the data set(see footnotes and text for explanation). The two
measurements of CO, production obtained on the six the control pigeon are
not shown.

TABLE 3. Mean water efflux rates (? SD) of two groups of tippler
pigeons that flew together in the same flock during seven flights.
Experimental birds wore a harness during flights 3 and 4. During flights 5
and 6 they wore a mock transmitter attached to the harness (this load
weighed about 2.5% of the birds M,). In flights 1, 2, and 7, experimental
birds were not wearing a harness or a mock transmitter. Control birds flew
unencumbered in all seven flights.

Flight ID

Type of treatment to experimental birds

ml H,O OFF d-
Experimental birds Control birds
1 none 352.8 k 59.8 (2) 601.6 * 394.5 (3)
2 none 378.9 ? 69.8 (5) 474.5 + 49.3 (4)
3 harness 885.0 + 377.8 (4) 561.3 + 276.5 (4)
4 harness 790.3 * 208.1 (5) 514.3 + 171.6 (2)
5 p + harness 883.2 -t 216.4 (4) 469.5 + 339.5 (2)
6 Ta + harness 682.2 ? 248.2 (4) 679.9 + 128.0 (4)
7 none 522.0 + 26.0 (2) 608.9 + 164.3 (2)
* T = mock transmitter. est. altitudes they were barely visible.

On some flights birds moved to a lower altitude after flying for >2 hr, and
remained there until some began to land. A tired bird that began to slow and
hover in preparation for a landing would usually lure a few other tipplers
in the flock to land with it; however, the lure was not effective before 1.5
hr into the flight. This behavior, which is well known among breeders and
flyers of tippler pigeons, undoubtedly lured some birds to land even though
they could have flown much longer. In addition, the availability of several
potential perches near the loft (e.g., TV antenna, barn roof, silo, etc.)
might have induced some birds to land prematurely, i.e., before they were
fatigued. Flight duration, therefore, was probably not a true measure of
flight duration capabilities of many of the pigeons; their capabilities
undoubtedly exceeded the measured duration. In spite of these confounding
circumstances, the flight time of the control birds (3.00 + 0.58 hr; range =
1.92 to 4.88 hr) was significantly longer than that of the birds wearing
only a harness (2 = 2.36 f 0.10 hr; range = 1.62 to 4.12 hr; F= 33.0, P 0.01) and of birds wearing a load attached to a harness (X = 2.22 + 0.002; F
= 19.87, P harness plus a load were not different (F -c 1 .O). This demonstrates that
tippler pigeons, wearing a harness alone or a load attached to a harness,
fatigue or tire of flying before control birds.

DISCUSSION CO, PRODUCTION AND POWER OF FLIGHT

Mean rate of CO, production of control tippler pigeons in this study was
very similar to that of homing pigeons flying without a harness or load
(control birds) in the study of Gessaman and Nagy (1988) (Table 4). We
anticipated that the rate of flight metabolism of tippler pigeons would be
less than that of homing pigeons, because tipplers fly slower than homers
and, therefore, their cost of transport should be less, i.e., less energy
should be used per km flown. However, we were unable to determine the
precise speed of flight of the tipplers, because the birds did not fly a
straight path between two points. The difference in mean Vco, among control
birds that were flying together throughout all or most of a flight was
larger than expected. For example, mean Vco, of Pigeon 2 was 9.6% higher
than that of Pigeon 1 (Table 2). The sex of these birds was unknown. It
appears that even within a flock of birds the energy expended by individuals
can be quite different. The somewhat similar increase in water efflux of
birds wearing only a harness, and of birds wearing a harness/transmitter
package, supports the idea that the effect of the transmitter is much less
than that of the harness alone.

Goslow et al. (1990), studying the flight of starlings (Sturnus vulgaris) in
a wind tunnel with cineradiography, proposed that as the wing is depressed
during the down stroke of flight the furcula spreads to inflate the
clavicular air sac and the sternum elevates to compress the posterior air
sacs. It is possible that the harness straps we used on our tipplers impeded
the movement of their furcula or sternum and reduced their depth of
breathing (tidal volume). They could have maintained adequate lung
ventilation (minute
volume) by increasing breathing frequency, a speculation that is consistent
with our observations of increased water efflux.

TABLE 4. A comparison of measured CO, production @co>) of homing
pigeons and tippler pigeons during

flight and values predicted from allometric equations.

Pigeon ID Mean body nnSS (9)

Mean CO, production [ml CO, hr g-1

Allometric Auometric peak- Measured flight VCOJ

Measured (n) prediction ion b resting vc.3,

California homing pigeons 412.5 13.9 (8) 10.1 8.2 17.5

Utah homing pigeons* 412.8 16.5 (4) 10.1 8.2 21.5

Tippler pigeon (this study) 281.2 13.5 k 3.1 (31) 11.1 9.1 16.9 f 3.9

Gessaman, J. A. and K. A. Nagy. 1988. Transmitter loads affect the flight
speed and metabolism of homing pigeons. Condor 90~662466.

2 Gessaman, J. A. and M. R. Fuller. 1989 (unpublished data).

a Predicted from F.q. 8 in Bernstein (1987).

b Predicted from the equation for minimum power output in Figure 2.8 of
Phillips et al. (1985).

WATER EEELUX

Bernstein (1987) reports that the rate of respiratory evaporation (M,) from
flying birds is related to body mass (MB) by the allometric equation, M, =
259 M,O.*O where M, is in mg/min and M, is in kg. This equation predicts a
respiratory evaporation rate of 480.7 ml H,O kg- d-l for a tippler pigeon
with a M, of 28 1.2 g, 18.5% higher than the average water efflux rate that
we measured on experimental birds flying without a load (404.9 ml H,O.kg-I
d-l) and 22.2% less than that measured on the control pigeons (577.6 ml H,O
kg-l d-l). Water efflux is a sum of water lost by respiratory evaporation,
cutaneous evaporation and excretion. The results suggest that 20% or less of
the water lost by our control birds was by cutaneous evaporation and
excretion. We anticipated that excretory water loss would be small, because
the birds were without food for about 18 hr prior to the flight. Hudson and
Bemstein (198 1) reported that about 20% of the total evaporative water loss
(respiratory plus cutaneous) from Chihuahuan Ravens (Corvus cryptoleucus)
flying in a wind tunnel was from cutaneous evaporation.

Our data indicate that birds flying with a load attached to a harness lose
water at a much faster rate than controls (50 to 100% faster). Dehydration
could limit flight duration of a bird carrying a harness/load well before
the depletion of fat (energy) reserves. This needs further study.

SOURCES OF ERROR

The mean error in DLW measurements of CO, production are in the range of &7%
when 50% of the IsO injected into the animal is lost from the body during
the measurement period (i.e., one-half life) (Nagy 1989). The error
increases exponentially as the turnover drops below 50%. For example, when
I80 declines by only 20% of its initial value (regardless of the absolute
level of IsO), a 1% error in any one of the 4 isotope measurements yields an
error in calculated CO, production of about f 37% (Nagy 1980). The I80
turnover was about 24% in flights 1, 2 and 3 which lasted about 2.5 to 3.8
hr, about 19% in flights 4 and 5 which lasted 2.2 to 2.5 hr, and about 22%
in flights 6 and 7 which lasted 2.2 to 3.2 hr (Table 5). These low values of
I*0 turnover can result in a very large error in CO, production and are
undoubtedly the reason for the aberrant values of CO, production which we
eliminated from the data set (see Methods section). Flight durations of 6-7
hr would be required to turnover about 50% of the isotopic oxygen.

A f 1% error in any one isotope measurement will yield a very large error in
calculated CO, production when tritiated water (HTO) turnover is a large
fraction (e.g., >90%) of I80 turnover. In all flights HTO turnover was 40 to
60% of IsO turnover, or about half of the I80 loss was due to the loss of
CO, from the body. Turnover rates of HTO, therefore, should not have been a
source of error in this study.

The study has not shown that a dorsal load affects the flight metabolism of
tippler pigeons. The sample size is too small, and its variability too
large, to statistically elucidate any real effect. In DLW studies that
compare two subgroups within a species a sample size of nine (per subgroup)
is recommended for statistical tests (Nagy 1989). However, it is highly
unlikely that a flock of 18 tippler pigeons (9 experimental and 9 control
birds) could be trained to fly non-stop for 6- 7 hours, because flocks > 12
birds tend to separate, and pigeons carrying a harness or load seem to land
before flying 6-7 hours. An inaccurate estimate of TBW at the start of a DLW
measurement period can result in a small error in the estimate of CO,
production. For example, a 5% error in TBW causes a 5% error in the estimate
of CO, production (Nagy 1980). We measured TBW by I80 dilution before the
first flight, but not before the subsequent flights, since in flights 2-7
two blood samples would have been needed before each flight to measure TBW
by IsO dilution, one to measure background level of 180 (before I80
injection) and the other to measure IsO dilution (1 hr after injection).
(The background level of isotope preceding flights 2- 7 probably contained
residual isotope from the previous flight, since some DLW flights were only
5-7 days apart.) We were concerned that the handling associated with taking
two blood samples from a bird before the flight would decrease the birds
willingness or ability to fly for several hours, so we took only one
pre-flight blood sample (1 hr after injection) in flights 2-7. Since the
background of I80 was unknown before flights 2-7, the mean TBW measured
before flight 1 was used in computing CO, production for all seven flights.

LIMITATIONS OF TIPPLERS IN FLIGHT STUDIES

We learned in this study that a flock of tippler pigeons which flies in
circles for hours is easy prey for raptors that move into northern Utah in
November for five months of residency. We lost three tipplers in one
November day to raptors. This restricts metabolic studies of tippler pigeons
in northern Utah to April through October. Properly trained tipplers can
regularly fly nonstop for 8-10 hours. This is usually accomplished with a
flock of 2-3 good flyers. Record tippler flights of more than 17 hr have
been observed (Curley 1961, Smith 1983). On 18 October 1988, the mean flight
duration of our control birds was 8 hr, 5 min, the longest flight observed
in this study. Our tipplers did not fly as long in the summer months as in
the cooler days of October. The longest mean flight duration of control
birds in July and August was 3 hr, 54 min. The negative effect of heat on
flight time was evident in another way; birds released at 07:OO hr (in the
cool of the day, just before sunrise) flew longer than birds released at
10:00 hr.

ACKNOWLEDGMENTS

We thank W. S. Seegar for providing support for this study, and D. G.
McAuley and E. Klaas for useful comments on a draft.

A. GESSAMAN, G. W. WORKMAN AND M. R. FULLER LITERATURE CITED

BERNSTEIN, M. H. 1987. Respiration in flying birds, p. 43-73. In T. J.
Seller [ed.], Bird respiration.
CRC Press, Boca Raton, FL.
CURLEY, J. T. 1961. The time-flying tippler pigeon sport. Howell Book House,
New York, NY.
GESSAMAN, J. A., AND K. A. NAGY. 1988. Transmitter loads affect the flight
speed and metabolism of homing pigeons. Condor 90:662-668.
GOSLOWG, . E., JR., K. P. DIAL, AND F. A. JENKINS, JR. 1990. Bird flight:
insights and complications.
Bioscience 40: 108-l 15.
HUDSON, D. M., AND M. H. BERNSTEIN. 198 1. Temperature regulation and heat
balance in flying White-necked Ravens, Contusc ryptoleucusJ.. Exp.
Biol. 90:267-28 1.
KENWARD, R. 1987. Wildlife radio tagging. Academic Press, New York, NY.
LEFEBVREE, . A. 1964. The use of D,O* for measuring energy metabolism in
Columba livia at rest and in flight. Auk 81:403-416.
LIFSON, N., AND R. MCCLINTCXK. 1966. Theory and use of the turnover rates of
body water for measuring energy and material balance. J. Theor. Biol.
12~46-74.
NAGY, K. A. 1980. CO, production in animals: analysis of potential errors in
the doubly labeled water method. Am. J. Physiol. 238:R466-R473.
NAGY, K. A. 1989. Doubly-labeled water studies of vertebrate physiological
ecology, p. 270-287. In P. W. Rundel, J. R. Ehleringer, and K. A. Nagy
[eds.], Stable isotopes in ecological research.
Springer-Verlag, New York.
NAGY, K. A., AND D. P. COSTA. 1980. Water flux in animals: analysis of
potential errors in the tritiated water method. Am. J. Phvsiol.
238:R454-R465.
PENNYCUICKC,. J. 1968. A wind-tunnel study of gliding tight in the pigeon
Columba livia. J. Exp. Biol.
49:509-526.
PHILLIPSJ, . G., P. J. BUTLERA, ND P. J. SHARP. 1985.
Physiological strategies in avian biology. Blackie, London.
ROTHE,H . J., W. BIESELA, NDW . NACHTIGALL.1 987.
Pigeon flight in a wind tunnel. II. Gas exchange and power requirements. J.
Comp. Physiol B 157:
99-109.
SMITH, J. F. 1983. Flying tippler endurance records.
Pigeon Review 6:53-54.
WOOD, R. A., K. A. NAGY, N. S. MACDONALD, S. T.
WAKAKU~A,R . J. BECKMANA, NDH. KAAZ. 1975.
Determination of oxygen- 18 in water contained in biological samples by
charged particle activation.
Anal. Chem.

//