Beissinger 



Chapter 37 



Population Trends Projected from Demographic Analyses 



1.0 



0.8 



0.6 



0.4 



02 



I 0.0 

 3 150 



O 



y = 0.012x- 1.919 

 r = 0.993, P < 0.001 



/. \f v v 



III 



175 200 225 250 



Julian Date 



275 



Figure 2 The cumulative probability distribution function for fledging 

 dates of 74 Marbled Murrelet nests. Results of a linear regression of 

 Julian date (x) on the cumulative proportion of nests that fledged (y) was 

 fit to data and are given. No probability value can be calculated for the 

 regression because cumulative fledging values are not independent. 

 Data are from Hamer and Nelson (this volume a). Dates shown refer to 

 the end point of censuses used to adjust the juvenile ratio. 



rarely seen beyond 1 km offshore, whereas adults have 

 frequently been seen up 3 km off shore and were still 

 encountered up to 5 km (Ralph and Miller, this volume; 

 Strong, pers. comm.). A good example of this effect is from 

 studies in Clayoquot Sound, British Columbia (tables 1 and 

 2). Surveys through kelp beds where juveniles were known 

 to forage found juvenile ratios 3-4 times greater than total 

 area counts (surveys of all individuals in the sound). Thus, it 

 seems likely that onshore surveys will overestimate the 

 juvenile ratio, and at-sea surveys will underestimate them 

 unless the at-sea surveys include some transects close to 

 shore or through kelp beds. 



Second, the juvenile ratio increased during the breeding 

 season in every case at locations with repeated surveys (tables 

 1 and 2). This would be expected if nests in a population were 

 asynchronously fledging young (fig. 2), and juveniles, subadults 

 and adults remained in the general vicinity so that populations 

 were being surveyed. The universal increase in juvenile ratios 

 during the breeding season indicates that juvenile ratios may 

 be useful tools for tracking productivity of a population. Third, 

 sequential surveys often yielded similar juvenile ratios after 

 the percentage of juveniles observed was adjusted for different 

 survey dates using the linear model in figure 2. The closest 

 values generally occurred for surveys conducted in late July 

 and early August (tables 1 and 2). These adjusted ratios differed 

 by about 3 percent or less, in 6 out of 7 instances. Thus, 

 juvenile ratios appear to be sensitive to seasonal change, yet 

 provide repeatable measures for fecundity estimates. 



Table 1 Surveys of the ratios of juveniles to after-hatch-year birds (adults and subadults) for Marbled Murrelets 

 conducted during the breeding season along shorelines or from boats cruising only along kelp beds. The percentage 

 of juveniles (Pct.juv.) was adjusted for the timing of the survey (survey period) by using the cumulative frequency of 

 fledging dates (fig. 2) to estimate an adjusted percentage of juveniles (Adj. pct.juv.)for the end of the nesting season 



388 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



