the breeding season, the propensity of females to attempt to breed again after the
conclusion of a previous attempt and the time required by the female to produce
a new clutch after the end of the previous attempt. In several cases it is known
that variation among areas or years in productivity due to these factors is consider-
able. The duration of the breeding season of the Common Snipe Gallinago
gallinagoand the number of breeding attempts per female were influenced
by flooding and soil moisture conditions. Annual and geographical differences
in the duration of the breeding season produced more variation in the number
of young produced per female than variation in success per nesting attempt
(Green 1988). Similarly, breeding productivity of Black-throated Blue Warblers
Dendroica caerulescens differed among forest habitats, but the differences
occurred because more breeding attempts were made per pair in the most
productive habitat (Holmes et al. 1996). The success rate of individual breeding
attempts did not differ. Variation in productivity among years in this species was
also mainly accounted for by differences in the mean number of broods that each
pair attempts (Holmes et al. 1992).
In spite of evidence from the field and from models (Ricklefs and Bloom
1977; Murray 2000) that annual productivity may not be accurately represented
by the average success of individual breeding attempts, a literature review by
Thompsonet al. (2001) showed that most published papers refer only to success
per attempt.
If there are adult birds that do not attempt to breed it is important to
realize that measures of the number of young fledged per breeding adult or pair
will overestimate productivity relative to the numbers of all adults. In such cases,
it is necessary to estimate the proportion of adult birds that attempt to breed,
usually separately for each age class, if the results are to be used in demographic
models. Estimating the proportion of birds of a given age that attempt to breed
is difficult. In some species, for example, many raptors and seabirds, some
non-breeding adults occupy breeding territories or nest sites and regular checking
from early in the season can indicate that they did not attempt to breed. It is
sometimes difficult to exclude the possibility that a pair did make a breeding
attempt, but quickly failed. However, in species in which egg-laying is always
preceded by conspicuous displays or nest refurbishment, it should be possible
to exclude this possibility providing that territories are visited early and often
enough. Other methods for measuring the proportion of non-breeders include
analysis of mark-recapture or mark-resighting data (Newton and Rothery 2001)
and radio-tracking of birds of known age (Kenward et al. 1999). In species in
which age classes can be recognized by plumage characteristics, postmortem
examination of the reproductive tract of females can be used to estimate the
Measuring annual productivity| 73