two factors potentially controlling the population need to be con- 
sidered: (1) extremely heavy mortality suffered by the young birds 
when they leave the nest, and (2) delayed sexual maturity. The fol- 
lowing calculations will demonstrate the limitations of these postu- 
lations and the kind of field data still needed to make our picture 
of the population dynamics of this species a definite one. 
Population equilibrium when all birds breed at one year 
of age.--If for each pair of breeding adults alive at the start of 
the nesting season, 2.19 young are fledged (Hammond and Henry 199) 
the true age ratio in the population at fledging time will be 
modified by adult mortality during the nesting period. I take this 
period to be about 2 months. Since adult mortality is about 30 per 
cent per year, in 2 months roughly 10 per cent of the adults die, 
It follows then tnat the age ratio at fledging time is 2.19 young 
per 0.9 pair; that is, 2.43 young per pair or 1.2 young per adult. 
With a median fledging date of about July 10 (my estimate), the 
young would have to suffer a minimum 33 per cent mortality in the 
next 3 weeks in order to reach the life-table age ratio of 0.8 per 
adult. This postulated mortality seems to me to be too high. I 
therefore regard the hypothesis of breeding at one year as not 
sustained at this time. 
Population equilibrium when all birds begin to breed at 2 
years of age.—-Under this hypothesis, the [I birds alive at the start 
of age interval 1-2 in table 36 are nonbreeding. The 99 young alive 
at the start mist then be produced by 9 adults; this means a life- 
table age ratio of 1.05 young per adult on August 1. This may now 
be compared to the age ratio of 1.2 per adult (as of about July 10) 
inferred from Hammond and Henry's work. To drop the 1.2 value to 
1.05 would require a minimum juvenile mortality of 12.5 per cent 
over and above that sustained by the adults. I do not believe 
that this 12.5 per cent is a reliable estimate of what takes place 
in nature. All the samples (productivity and life table) are small 
enough to involve fairly large sampling errors. We can, however, 
conclude that initial breeding at 2 years of age offers the best 
available explanation of vital statistics for this species as we now 
know them. 
Some confirmation of this situation is found in the plumage 
descriptions of marsh hawks. According to Bent (1937, p. 84), the 
juvenal plumage is worn for about one year, and the adult plumage 
is usually acquired by a complete molt that takes from July to 
October or later. In my experience with the duck hawk or peregrine 
falcon (which has an identical type of molt), such immature birds 
are readily accepted as mates but the females either lay no eggs 
or have very small clutches. It seems to me entirely possible that 
this situation is likewise true of the marsh hawk. With so few band- 
ing recoveries available for the precise calculation of subadult and 
adult mortality rates, and with the probability that some band. loss 
may mar the results, we are not able to say with certainty that all 
one-year old marsh hawks raise no young. Quantitative studies of 
nesting success and productivity from various parts of this species! 
86 
