The concentrations of maternally derived androgens in the yolks of avian eg
gs vary within and among clutches, but a mechanistic basis for this variati
on has not been elucidated. We investigated in the American kestrel, Falco
sparverius, whether changes in plasma-prolactin concentrations induced by c
hanges in photoperiod and food supply affect yolk-androgen concentrations.
Over the course of a photoinduced breeding period in the laboratory, we mea
sured concentrations of plasma immunoreactive prolactin (ir-prolactin) in f
emale kestrels with ad libitum food availability (control) or food availabi
lity that was reduced during the early breeding period. In a second laborat
ory study, we administered via osmotic mini-pumps ovine prolactin (o-prolac
tin) to females beginning on the day they laid their first egg of a clutch
(egg-day 1) to determine the effects of high prolactin concentrations on yo
lk-androgen concentrations. In both this study and one on free-living kestr
els, we quantified changes in yolk-androgen concentrations with date of clu
tch initiation. Concentrations of ir-prolactin in nonlaying females rose wi
th date, irrespective of food treatment. Egg-day I ir-prolactin concentrati
ons were higher in control females laying late during the breeding phase th
an in those laying early. This increase was absent in food-reduced females.
Yolk-androgen concentrations in eggs 3 and 4 but not eggs I and 2 of the c
lutch were higher in clutches initiated late than in clutches initiated ear
ly in the breeding phase in both the field and laboratory. o-Prolactin trea
tment elevated yolk-testosterone but not androstenedione concentrations. Th
ese findings suggest that, in American kestrels, seasonal and laying-associ
ated increases in plasma-prolactin concentrations elevate yolk-testosterone
concentrations. Food availability and other factors may interact with date
to regulate the effects of prolactin on yolk-testosterone deposition. (C)
2001 Elsevier Science.