Demographic parameter estimates are essential for understanding population ecology and developing management plans for species of concern. We inferred measures of breeding success using radiotelemetry in the marbled murrelet (Brachyramphus marmoratus), a secretive, forest‐nesting seabird, from 1998 to 2001 in Desolation Sound, British Columbia, Canada. Our estimates of mean annual nesting success and fecundity (0.19–0.23 female offspring/adult female/yr) are among the highest reported for the species. We suspect that our estimates are high compared to previous estimates in our study area (Manley 1999), primarily because of our higher success in areas inaccessible to ground‐based human observers. We detail how behavior‐based inferences of activity at different reproductive stages did not differ between confirmed nest sites and suspected nest sites that were physically inaccessible to us. We were able to accurately predict initiations of breeding and incubation success from the duration of adults' repeated daily shifts from the ocean to their inland nest sites. Chick‐rearing success was accurately predicted by visitation rates of adults during provisioning. We discuss the assumptions and potential biases of our methods and their effects on our results. Our method may overestimate early breeding failure, but it likely provided unbiased fecundity estimates for our population. Accurately inferring breeding success through radiotelemetry is costly and labor intensive. However, radiotelemetry could provide crucial demographic information once thought impossible to obtain for secretive breeding species.

Summary1. An indirect, physiological method to assess reproductive state in individuals of unknown status is described. The plasma levels of two main yolk precursors, vitellogenin (VTG) and very-low density lipoproteins (VLDL), are focused on as indices of egg production, for the characterization of fecund females. 2. Data for a species where breeding chronology could be directly assessed, at the population level (Cassin's Auklet, Ptychoramphus aleuticus ), confirmed the validity of this approach: plasma VTG levels were highest during the defined egg-laying period, and the highest proportion of females were defined as 'egg-producing' in this period. 3. Analysis of samples for Marbled Murrelets ( Brachyramphus marmoratus ) caught off-nest (i.e. where all individuals were of unknown status), clearly identified a putative egg-laying phase, with a single, protracted laying period (cf. multiple-broodiness). 4. Analysis of body mass confirmed our characterization of 'egg-producing' females: birds with elevated plasma VTG were on average 40 g heavier than other females, equivalent to the mass of the single egg (36-41 g). 5. Indirect, physiological assessment of reproductive state provided valuable information on the breeding biology of Marbled Murrelets which would have been difficult to obtain in any other way (e.g. proportion of fecund females, breeding phenology, single vs multiple-clutch breeding pattern). Despite some limitations, this technique should be applicable to any oviparous vertebrate population where essential information on breeding biology cannot be obtained by more traditional methods.

No abstract

Population size is a fundamental variable of interest in most conservation programs, as are the demographic processes underlying variation in numbers. The Marbled Murrelet (  Brachyramphus marmoratus ) is a seabird nesting in coastal old‐growth forest in North America and is at the heart of a controversy concerning forest management. Despite the urgent need for a careful assessment of the demographic state of populations, the species is so secretive that its life history is poorly known and reliable estimates of the required vital rates are rare or missing. Here we present the first survival estimates obtained through capture‐recapture data from a population in British Columbia, Canada. This population is in a site that has been heavily logged in the last 50 years and continues to be logged. Survival estimates ( 0.8289 and 0.9289 based on different samples corresponding to two capture techniques ) fell within the range specified for small alcids, indicating that neither mortality nor emigration out of the study area is particularly high in the population. We used capture‐recapture data to estimate the realized population growth rate (  λ ) from 1991 to 2000. λ^ was < 1 ( 0.9851 ), but the confidence interval was large and included 1 ( stationary population ). This estimate was obtained using part of the data ( the sample with the higher survival rate ) to meet the assumptions of the analytical approach. This lambda value might not reflect the growth rate of the overall population but may be specific to the segment of the population that successfully bred. The hypothetical value of λ corresponding to a survival probability of 0.8289 was 0.8593, which suggests that the population may be declining if the local survival rate based on the combined samples is closer to the true value. Because of our low recapture rates and poor precision of estimates, there is still uncertainty regarding the health of this population. Development of field techniques and achievement of higher detection probability should reduce uncertainty in future demographic studies of Marbled Murrelets and allow refined assessment of the conservation status of populations.