Enhancement of whooping crane Grus americana recruitment by egg removal, Wood Buffalo National Park, Alberta and Northwest Territories, Canada
Boyce M.S., Lele S.R. & Johns B.W. (2005) Whooping crane recruitment enhanced by egg removal. Biological Conservation, 126, 395-401
Since an all time low of 16 individuals recorded in 1942, the population of the endangered whooping crane Grus americana (as of winter 2004) has risen to 217 birds in Wood Buffalo National Park (Canada), with reintroduced populations of 48 in Wisconsin and 79 in Florida (USA), plus 128 in captivity.
In order to obtain birds to facilitate recovery through captive breeding, foster-parenting and reintroduction, one of two eggs (90% of whooping crane clutches contain two eggs) was removed from 62% of nests in Wood Buffalo National Park, Canada during 1967–1996. Egg removals were justified because whooping cranes usually only rear one chick, the other often dying due to siblicide or predation. Concerns exist that the wild population may have recovered more quickly if nests had not been disturbed and no eggs taken. The consequence of the egg removals on recruitment of young whooping cranes, (i.e. the proportion of nesting pairs yielding a chick that survived to appear with the wintering flock at Aransas in Texas) was examined.
Study site: Wood Buffalo National Park (44,807 sq. km) is situated on the plains of north-central Canada, within Alberta and the Northwest Territories.
Egg collection: One of two eggs was removed from 62% of nests (496 eggs in total) in the National Park during 1967–1996. Eggs were removed by the end of May each year, with eggs being taken from nests throughout the nesting area. Selection of nests for egg removals was made during survey flights. Sampling was not influenced by access constraints because there were no sites that could not be accessed by helicopter. Only nests along the Nyarling River were less likely to be sampled as it is distant from the other nesting areas, thus requiring more travel time. Lay date was estimated by flying low over the nest until the incubating adult stood and the number of eggs could be counted. If no egg was present repeat visits were made until eggs were laid. At the time of egg collection, eggs were often tested for viability to ensure that the egg left was fertile. If neither appeared viable, an egg was sometimes substituted with one from a nest in which both were fertile.
Field sampling: During nest visits, pond depth was measured at each nest 1 m from the edge of the nest in three random directions. These measurements were averaged over all nests to obtain mean pond depth for each year.
Chick monitoring: Chicks were ringed, when still flightless, with colour leg bands during 1977–1988. In August, aerial surveys were conducted on the nesting grounds, attempting a complete census. In December aerial surveys of whooping crane wintering territories at Aransas National Wildlife Refuge (Texas), were conducted by US Fish and Wildlife Service staff. On rare occasions a ringed chick was missed in the August survey and appeared on the wintering grounds, the August survey results were corrected.
Effect of egg removal: As the egg removals were not designed as an experiment, statistical methods to control for variables that might influence results were employed. The Mantel–Haenszel estimator (see Agresti 1990 & original paper) was used to correct for year and nesting area effects to compare recruitment from unmanipulated clutches and those with an egg removed. After finding a significant main effect using this estimator, further analysis was undertaken to better understand recruitment success incorporating several ecological factors which might influence recruitment, including: precipitation, pond water depth, lay date, egg substitution, egg removal, and changes in lynx Lynx canadensis and mink Mustela vison (both potential predators) harvests. Mink and lynx harvest data were obtained from registered traplines in the nesting area.
Effects of egg removal on recruitment: Contrary to expectation, removing one of two eggs from a whooping crane nest increased the probability of nest success. Recruitment (excluding novice-pairs) from unmanipulated nests was 0.385 (±0.0013) whereas recruitment was 0.498 (±0.0012) for nests where an egg was removed. The egg removal effect was not influenced by excluding pairs nesting for the first time from the analysis, or the number of years that a site was used. Egg removal effect appeared not confounded by year or site effects.
The ratio of immatures to adults at wintering grounds in Texas varied considerably among years (with a high of c. 0.47 to a low of zero recruitment in two years) and recruitment varied among nesting sites (see Table 1, attached). Excluding 32 nests from the analysis at which egg substitutions were performed, survival of chicks to August and recruitment of young into the wintering flock was positively influenced by egg removal. On average recruitment was lower from nests where an egg substitution was made, even though some of these egg transfers resulted in surviving chicks that presumably were not possible otherwise. Low success from nests with substituted eggs might be expected if nests with infertile eggs are associated with poor nest sites or young, inexperienced. Of chicks surviving until August, 317 out of 440 (72%) survived the migration to Texas.
Effects of other factors: Declines in lynx and mink pelts taken from the nesting area are highly correlated with reductions in chick recruitment. Predation on whooping cranes as alternate prey is considered most severe at the time of predator declines induced by the collapse of snowshoe hare populations. Neither precipitation or pond water depth significantly influenced survival of chicks to August or December.
Conclusions: Removing one of two eggs from a whooping crane nest has enhanced recruitment to the Wood Buffalo/Aransas population. The joint US–Canada egg removal program has benefited conservation and population recovery, and allowed establishment of two new wild populations and several captive flocks.
Agresti A. (1990) Categorical Data Analysis, Wiley, New York, USA.
Note: If using or referring to this published study, please read and quote the original paper.