Beissinger 



Chapter 37 



Population Trends Projected from Demographic Analyses 



seabirds that nest and forage offshore (Ainley and Boekelheide 

 1990), there is no evidence that fish populations within 2 km 

 of shore, which murrelets mostly utilize, are affected. 



Some uncertainty in the measure of fecundity derived 

 from juvenile ratios is associated with the timing of censuses. 

 To convert juvenile ratios to a fecundity estimate, ratios had 

 to be increased to account for young fledging after the date of 

 census by using the cumulative frequency distribution for 

 fledged nests with known dates (fig. 2). This distribution was 

 comprised of nests from Alaska to California, because sample 

 size was not large enough to partition nests among portions 

 of the murrelet's range. Variation in the fledging dates exists 

 between Alaska, British Columbia, and the Pacific Northwest 

 (Hamer and Nelson, this volume a), although there is much 

 overlap. Future research might employ bootstrapping 

 techniques (Crowley 1992) to calculate an error estimate for 

 the cumulative frequency by date, as one way to determine 

 the inherent variability of the correction factor. 



Other approaches to estimating fecundity also yielded 

 low values, but are likely to have too many biases to be 

 useful yet. Juvenile ratios measured only using on-shore 

 counts tended to be higher than off-shore counts (table 2). 

 But fecundity will be overestimated by using only on-shore 

 counts because they undersample adults. Estimates of 

 fecundity from nesting success are likely to be less useful 

 than juvenile ratios because they must be corrected for many 

 factors that are difficult to measure (such as the proportion 

 of adults nesting, fledgling survival to the ocean, and renesting 

 frequencies). Furthermore, for the foreseeable future, fecundity 

 estimates based on nesting success are likely to depend on 

 small sample sizes because of the difficulty in finding nests. 



Predicted Rates of Decline of Murrelet Populations 



All scenarios of the demographic model predicted that 

 murrelet populations are likely to be declining (table 3). The 

 estimated rate of decline varied from 2-12 percent per year, 

 depending on the parameter estimates used. Based on the 

 discussion of the parameters above, the most likely rate of 

 decline would be based on fecundity values from juvenile 

 ratios intermediate between offshore juvenile ratios (which 

 may underestimate reproductive success) and nesting success 

 (which certainly overestimates fecundity), used with an estimate 

 of survival close to 0.90. These intermediate fecundity values 

 would suggest a rate of decline around 4 percent per year. 



A predicted decline of 4 percent per year is in close 

 agreement with population declines documented in two field 

 studies of murrelets. A 50 percent decline in murrelets detected 

 over 20 years of Christmas Bird Counts in Alaska (Piatt and 

 Naslund, this volume), despite an increase in observer effort 

 during this period, would represent a 3.4 percent average 

 annual decline. Similarly, the 40 percent decline in the 

 Clayoquot Sound murrelet population in British Columbia 

 over 10 years (Kelson and others, in press) would average to 

 a 5 percent annual decline. These studies are based on either 

 periodic but intensive sampling during few annual periods 

 (British Columbia), or low intensity but extensive sampling 



every year (Alaska). Despite, the sampling shortcomings 

 inherent in these two studies, the population trends that they 

 have documented are in good agreement with trends predicted 

 by the model in this paper. 



Model results suggest that murrelet populations may 

 even be declining at greater rates (table 3). A 7 percent 

 annual decline would be predicted from juvenile ratios based 

 on offshore counts in conjunction with high survival estimates. 

 This value is certainly a possibility for Pacific Northwest 

 populations of murrelets, which exhibit low offshore juvenile 

 ratios. It is even conceivable that murrelet populations could 

 be declining at 9- 1 2 percent per year (table 3). However, this 

 rate of decline is so high that it seems unlikely to go unnoticed 

 by field researchers. Furthermore, it is based on the most 

 pessimistic combinations of fecundity and survivorship. I 

 interpret the model predictions, in conjunction with the field 

 evidence, to suggest that murrelet populations are likely to 

 be declining at least 4 percent per year and perhaps as much 

 as 7 percent per year. 



Use of Juvenile Ratios for Murrelet Conservation 



Conservation efforts for Marbled Murrelets have been 

 hampered in part because of a lack of reliable biological 

 information. Demographic characteristics have been especially 

 difficult to measure because nests are very hard to find and 

 monitor, murrelets fly long distances both over the ocean 

 and across land, and the birds are difficult to capture, mark, 

 and telemeter (Quinlan and Hughes 1992). Juvenile ratios 

 provide one estimator of murrelet population health that 

 may be reasonably measured in the field. 



Juvenile ratios have great potential as estimators of 

 productivity. It is easy to obtain large sample sizes of juvenile 

 ratios compared to the difficulty of finding and monitoring 

 nests. It will be many years before enough nests are found to 

 yield sample sizes sufficient for accurate estimates of nesting 

 success. Additional information needed to convert nesting 

 success into annual fecundity (the proportion of birds that 

 nest and the number of attempts per year) will perhaps be 

 even more difficult to obtain. Juvenile ratios implicitly 

 incorporate these factors. Research will need to determine 

 optimal protocols for sampling juvenile ratios at-sea that 

 take into account apparent differences in habitat use by 

 juveniles and adults (tables 1 and 2) as well as other factors 

 that could bias these ratios. 



Changes in juvenile ratios could be a useful tool to 

 understand factors limiting murrelet population growth. 

 Juvenile ratios could be monitored in a regional areas (e.g., 

 over 30-50 kms of shoreline) and compared to landscape 

 characteristics to determine the effects of forest management 

 and other land use practices. Juvenile ratios may also be 

 useful for monitoring murrelet population trends. However, 

 changes in juvenile ratios can be caused either by changes in 

 recruitment (increased nesting success results in greater 

 proportions of juveniles) or changes in adult survivorship 

 (decreased survivorship results in greater proportions of 

 juveniles). Whether juvenile ratios change due to improved 



392 



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



