682 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 4, December 2011 
observations indicate that young grew to near the 
size of adults in a month and were apparently 
indistinguishable from adults in less than 
10 weeks. Birds were not yet breeding at the 
end of August or the beginning of September 
1923 and no birds showed signs of juvenile 
plumage. Thus, if they had bred earlier in the year, 
hatching would probably have been in April or 
May. This could fit with Spencer's impression 
that rails bred in April to May (supported by his 
nearly adult-size specimen from 7 Jun) and again 
in August to September when hatchlings were 
seen. Vaughn noted nests in April 1938. However, 
Lyons (1939) observed birds almost out of down 
on 13 June, just hatched on 25 June, downy on 25 
July, eggs on 16 September and 5 October, newly 
hatched on 23 October, more hatching beginning 
in November, and ~4 days of age on 19 
November. Eggs and downy young are not 
mentioned again until 20 July 1940. Thus, there 
seems to have been egg-laying throughout 1939 
and well beyond the time of year in which the 
rails did not breed at all in 1923 and were 
evidently not breeding in 1936-1937. Thus, the 
breeding season was probably influenced by 
environmental conditions such as drought, when 
breeding may not have occurred. 
The Wake Island Rail appears to have had a 
most unusual communal breeding system with up 
to three nests being placed together and nests 
apparently tended by more than two birds per 
nest. Young were attended, defended, and fed by 
groups of adults until well after hatching. This 
would presumably have been an effective way of 
deterring predation on eggs and young by hermit 
crabs and rats. Breeding seasons were at times 
irregular, or spread over 5-6 months. At least 
small groups of females would have to be capable 
of synchrony in laying that did not coincide with 
that of other groups of females, as apparently was 
the case in 1939, for communal breeding to be 
maintained. Unfortunately, it is not known 
whether males participated in this communal 
breeding pattern. 
We know less about the phenology of molt. 
Birds from the Tanager collections (Aug-Sep) are 
mostly extremely worn and faded. Lyons obser¬ 
vations of “straw-colored” primaries, which is 
their appearance when faded and abraded, includ¬ 
ed only one of 60 on 9 June 1939. but such 
individuals had become more numerous by 
August, in agreement with the Tanager speci¬ 
mens. The only available adult in fresh plumage 
was taken in November and the little available 
evidence suggests a full body molt in Seplember- 
October. 
POPULATION SIZE AND POTENTIAL 
SOURCES OF MORTALITY 
Abundance .—Wetmore (1970:4) noted that his 
collecting of specimens of the rail “was so 
distributed as not to bring undue pressure on the 
birds from any one area. Their abundance was 
shown when on my last day. in one limited space 
around a small opening. I counted 15 individu¬ 
als.” That was evidently 3 August 1923 when his 
journal mentions "about” 15 rails (Wetmore in 
Olson 1996:111), of which he collected eight, 
although the expedition did not leave until 5 
August. Wetmore collected eight rails on Wake 
Island itself on each of the following days: 28-29 
July. I and 3 August. Six were taken on Wake on 
2 August and eight were taken on Peale Island on 
30 July. 
The rat control team of 1936. when faced with 
the impossible task of capturing all the rails on 
Peale Island prior to poisoning operations, 
estimated there were 400 to 500 birds on Peale 
alone (Spencer 1936). Later. Spencer (1959) 
estimated the total population of the rail at a 
“few hundred” individuals. “The majority of the 
birds are found on the islet of Peale. Small 
scattered groups inhabit Wake and Wilkes Islets” 
(Spencer 1959). This is the only possible indica¬ 
tion that rails were ever found on Wilkes and, if 
true, they may have arrived there as a deliberate 
introduction. The fact that there were only “small 
scattered groups” on Wake Island proper in the 
late 1930s possibly indicates that the reduction in 
habitat caused by the construction of the airstrip 
may have had a negative impact on population 
size. 
Lyons (1939. 9 Jun) observed a flock of 60, by 
count, at the chicken feed storage shed on Peale 
and there was still a large flock by the greenhouse 
on 26 June. Many were seen on Wake in the 
meadow along the lagoon shore on 2 April— 
especially the meadow at the tip of the lagoon and 
many were heard calling (5 calls, 20 birds seen). 
(Lyons 1941). 
Direct Observations of Mortality .—Lyons men¬ 
tioned several instances of rail chicks dying 
shortly after hatching but the causes of their 
deaths were unknown. Perhaps hyperthermia 
caused by exposure of small, black chicks to the 
harsh heat and sun of Wake Atoll may have been 
