Chapter 37 



Population Trends of the Marbled Murrelet Projected From 

 Demographic Analyses 



Steven R. Beissinger 



Abstract: A demographic model of the Marbled Murrelet is devel- 

 oped to explore likely population trends and factors influencing 

 them. The model was structured to use field data on juvenile ratios, 

 collected near the end of the breeding season and corrected for date 

 of census, to estimate fecundity. Survivorship was estimated for 

 the murrelet based on comparative analyses of allometric relation- 

 ships from 10 species of alcids. Juvenile ratios were generally low, 

 and were higher for counts made from shore or in kelp beds 

 (typically 10 percent) than conducted offshore (<5 percent). An- 

 nual survivorship was strongly related to body size in alcids. 

 Survival for the Marbled Murrelet was predicted to be 0.845 and 

 r?nge to 0.90. Lambda, the expected annual growth rate of the 

 population, was estimated for likely combinations of fecundity and 

 survival, and indicated that under all combinations murrelet popu- 

 lations are expected to be declining. Based on the best data, rates 

 of decline are predicted to be 4-6 percent per year, but the rate of 

 decline could conceivably be twice as large. Studies in Alaska and 

 British Columbia suggest population declines at 3-5 percent per 

 year, supporting model predictions. Results are discussed in rela- 

 tion to the factors affecting murrelet population growth, and the 

 use of juvenile ratios for monitoring murrelet populations. 



Recovering a threatened or endangered species depends 

 on determining its rate of population change and correcting 

 the factors that limit population growth. Despite the important 

 information on the biology and life history of the Marbled 

 Murrelet (Brachyramphus marmoratus) that has been brought 

 together in this volume, population trends for the murrelet 

 remain elusive because little long term data are available. 

 Christmas bird counts from five sites in Alaska found a 50 

 percent decline in the population over a 20 year period (Piatt 

 and Naslund, this volume). Murrelet censuses conducted in 

 Clayoquot Sound, British Columbia 10 years apart found a 

 40 percent decline in the population (Kelson and others, in 

 press). Comparison of historic and current data suggests that 

 the murrelet has disappeared or become very rare in large 

 portions of its nesting range in California, Oregon, and 

 Washington (Carter and Morrison 1992). But current 

 population trends in the Pacific Northwest remain unknown. 



Demographic modeling can give indications of likely 

 population trends and play an important role in the conservation 

 of the Marbled Murrelet. Simple demographic models based 

 on estimates of annual survival and fecundity can be used to 

 determine the rate of decline or increase of a species. They can 

 also help focus attention on critical demographic information 

 that needs to be gathered for future studies. Sensitivity analyses, 

 where demographic values are altered to see the effect on 



1 Associate Professor of Ecology and Conservation, School of Forestry 

 and Environmental Studies, Yale University, New Haven, CT 065 1 1 



population growth, can indicate which components of the life 

 history are most likely to affect population growth and where 

 the potential for management may be greatest. 



Unfortunately, only a little is known about the demography 

 of the murrelet. There are no estimates of survivorship for 

 birds of any age. Reproduction is slightly better understood. 

 Clutch size is known to be one egg, and a substantial proportion 

 of nests are known to fail (Nelson and Hamer, this volume b). 

 However, neither the age of first breeding nor the proportion 

 of adults that breed is known. The ratio of young-of-the-year 

 (hereafter juveniles) to after-hatch-year birds (subadults and 

 adults) has been monitored at-sea and is often very low (e.g., 

 Ralph and Long, this volume). 



This paper represents an initial attempt to model the 

 demography of the Marbled Murrelet to explore likely 

 population trends. Although few data are available, there is 

 enough reproductive information from murrelets to use, in 

 conjunction with predictions of survivorship derived from 

 analyses of past studies of alcids, to yield crude estimates of 

 the rate and direction of change of the murrelet population. 



Model Structure 



The model was structured to take advantage of the one 

 population parameter that could be best estimated from field 

 data - fecundity. In the absence of detailed data, the simplest 

 way to model the murrelet population is based on three life 

 stages: adults (birds that are breeding age or older), subadults 

 (birds that are greater than one year old but younger than the 

 age of first breeding) and juveniles (fledged young that have 

 reached the ocean but have not yet survived their first year of 

 life). The latter stage takes particular advantage of one of 

 two estimates of productivity available from field data - 

 namely the ratio of young to after-hatch-year birds surveyed 

 at sea. The virtue of this scheme - simplicity - is also its 

 weakness. Undoubtedly there may be age variation among 

 the demographic rates of murrelets, as there is with other 

 seabirds (Hudson 1985, Nur 1993, Wooller and others 1992). 

 But without any specific information on the age structure of 

 vital rates, assigning age structure to them would be arbitrary. 

 For the moment, simplicity has its virtue. 



The simplified population life cycle given in figure 1 is 

 based on post-breeding season censuses with a projection 

 interval of one year (Caswell 1989, Noon and Sauer 1992) 

 and is typical for long-lived monogamous birds (McDonald 

 and Caswell 1993). The flow of events is (1) censuses are 

 conducted at the end of the breeding season, (2) birds must 

 then survive to the next breeding season, (3) all individuals 

 are aged one year, (4) surviving adults then breed, and (5) 

 post-breeding censuses are conducted again. Circles or nodes 



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



385 



