WATER VAPOUR FLUX IN RAIL EGGS 
69 
Parameter 
Observed 
Swamphen 
Ratio observed: 
mean predicted means 
Observed 
Moorhen 
Ratio observed: 
mean predicted means 
Length (mm) 
52.12 ± 2.19 
(53; 16) 
50.55 ± 2.49 
(20;6) 
Max. breadth (mm) 
36.41 ± 0.91 
(53; 16) 
35.03 ± 1.52 
(23;8) 
Initial mass (g) 
35.14 ± 3.25 
(31;13) 
33.42 ± 2.89 
(21 ;7) 
Estimated initial mass (g) 
37.45 ± 2.80 
(45; 10) 
33.82 ± 4.39 
(14;4) 
I (days) 
24 
24 
M Hi0 (mg-day- 1 ) 
185 ± 38 
(61;13) 
0.83 
188 ± 35 
(136:16) 
0.91 
G h ,o (mg.[day.kPa]->) 
74.36 ± 12.03 (12; 11) 
1.30 
71.43 ± 16.09 (15; 12) 
1.35 
L(gm) 
235 
(20:15) 
0.85 
270 
(12; 10) 
1.02 
N 
6087 
(20; 15) 
1.18 
5132 
(12; 10) 
1.07 
A p (mm 2 ) 3 
1.04 
1.33 
1.15 
1.38 
a “observed” value calculated from measured G HlQ and L 
Table 1. Size and water vapour flux parameters of rail eggs. Numbers in parentheses are sample sizes of (eggs; 
nests). I = incubation period; M H2 q = daily rate of water loss; G HlQ = water vapour conductance; L = shell 
thickness; N = pores per egg; A p = total effective pore area. Mean K w used to obtain initial egg mass estimate 
was 0.541 (swamphen) and 0.539 (moorhen). Estimates of I based on personal observations and on Falla et al. 
(1985), Garnett (1978), O’Grady & Lindsey (1979), National Photographic Index (1985) and Williams (1966). 
Sources for predictions: M H , 0 , Ar & Rahn (1980); G H , Q , L and N, Ar & Rahn (1985); A p , Ar & Rahn (1978); 
predictions based on egg mass/incubation period. 
for 6 nests); only three values were less than 3.5 
kPa. Mean initial egg mass, derived as for the 
swamphen, was 33,6 g. At the mean M H?0 
recorded, an egg of this mass would have a frac¬ 
tional mass loss of about 13% over the 24 day 
incubation period. G h , g corrected to 25°C 
ranged from 46.24 to 107202 mg.(day.kPa) -1 ; the 
mean conductances of three older eggs with an 
average initial mass of 32.3 g and nine fresh eggs 
with an average mass of 34.9 g did not differ sig¬ 
nificantly (77.67 ± 9.02 versus 69.25 ± 20.08 
mg.(day.kPa)" 1 ; t (10) = 0.686, p >0.05). AP H2 o 
was calculated to be 2.6 kPa and the mean satu¬ 
rated vapour pressure of the egg contents was 
therefore estimated to be 6.2 kPa, which trans¬ 
lates into a mean incubation temperature of 36- 
37°C. Pore density of the shell averaged 26 ± 8, 
25 ± 6 and 31 ±8 pores per 25 mm 2 at the 
pointed pole, equator and blunt pole respect¬ 
ively. Pores were unbranched, had open orifices 
and were mostly posthorn-shaped. 
DISCUSSION 
Given the prevailing ambient temperatures and 
relative humidities in the study region (see 
Methods), mean water vapour pressure in the 
nests of both rails must have been substantially 
above environmental levels. The calculated 
mean saturated vapour pressure of the egg con¬ 
tents of the two species was similar to that 
reported for many other bird species (Walsberg 
1980, Rahn 1984) and translates into a mean 
incubation temperature which is fairly typical of 
many birds that maintain a high incubation con¬ 
stancy (Drent 1972). Mean P n in both species 
was substantially higher than the averages 
reported for a large sample of bird species breed¬ 
ing in a variety of environments (2.0-2.8 kPa; 
Rahn et al. 1977a, Ar & Rahn 1978, Walsberg 
1980, Rahn 1984). Mean AP H; , 0 was substan¬ 
tially smaller than the mean values calculated 
for many other bird species nesting in various 
environments (3.6-4.7 kPa; Rahn & Ar 1974, 
Rahnetal. 1977b, Ar& Rahn 1978, 1980, Wals¬ 
berg 1980, Rahn 1984). 
However, P n and AP H?0 were clearly not so 
unusual as to necessitate any significant 
compensator}' changes in G H ,o or the shell para¬ 
meters that influence it. Mean values obtained 
for these factors all lie within the 95% confi¬ 
dence limits of the values predicted by the 
allometric equations referred to in Table 1. In 
acknowledged wet-nesters, such as grebes and 
brush turkeys, where P n is much higher (4.3-6.4 
kPa) and AP H2 q much smaller than in the rails, 
Gh 2 o and N usually exceed allometrically pre¬ 
dicted values by more than two standard errors 
of estimate (Seymour & Rahn 1978, Davis et al. 
1984, Ar & Rahn 1985). This contrast clearly 
occurs because the eggs of grebes and brush tur¬ 
keys, unlike those of swamphens and moorhens, 
are permanently or intermittently covered by 
warm, moist vegetation, and the eggs of grebes 
are also often in contact with liquid water. 
Comparative data on egg water vapour flux 
for other rail species are few. M H ,o and N are as 
expected from egg mass in the king rail, Rallus 
