Does Laying Order Influence Growth Rates of Common Eider Ducklings? A Skeptical View

Abstract

Erikstad et al. (1998:8) concluded that ...growth and presumably the fitness of [Common Eider Somateria mollissima] are related not only to egg-size, but also to the laying position of the eggs from which they hatch. They argued that this was consistent with an adaptive hypothesis that states that females allocate resources to eggs in a clutch in relation to the likelihood that they will hatch and produce offspring that survive. Erikstad et al. (1998) analysed the influence of egg size and laying order on hatching body mass of ducklings. After controlling for egg size, they found that laying order was negatively related to hatching body mass as hatching from second eggs were heavier than those from final eggs. To examine the relationship between egg size and growth, Erikstad et. al (1998) used a repeated analysis of variance and found that laying order and duckling age significantly influenced body mass of ducklings. There was not, however, a significant interaction between age and laying order in this analysis. So, Erikstad et al. (1998) performed a second analysis on data from hatched from second (largest) and last (smallest) eggs from clutches of four and five eggs. This analysis not only showed that laying order and age affected body mass, but also a significant interaction between these variables. Erikstad et al. (1998:6) interpreted this interaction to mean that ducklings from small final eggs gain less mass than from large second eggs...not only because of their smaller size, but also because of their position in the laying sequence. This interpretation may be erroneous because of two potential problems. First, Erikstad et al. (1998) did not control for egg size in their analysis of effects of egg size and laying order on growth of ducklings. Thus, their finding that laying order significantly affected growth might simply reflect the fact that second eggs are larger than final eggs. We note that Erikstad et al. (1998) did control for egg size when analysing effects of egg size and laying order on hatching body mass of ducklings. It was inappropriate, however, to assume that because laying order influenced hatching mass of ducklings, independent of egg size, that this would also be true of duckling growth rates. Second, Erikstad et al. (1998) apparently did not determine sex of ducklings. This is unfortunate because if were not all the same sex or if the sex-ratios were not 50:50 in each group (highly unlikely, given the small sample size for each group) then some of the variance in their growth data may have been due to different growth rates of males and females. We are unaware of data on sex-specific growth rates in Common Eiders, but in Lesser Snow Geese Chen c. caerulescens males are heavier than are females by 12 d of age (Ankney, upubl. data; see also Fig. 1 in Ankney 1980). More importantly, Erikstad et al. (1998:7) misinterpreted the larger mass gain of from second eggs compared to that of from last eggs to conclude that Ducklings from small final eggs in a clutch grew more slowly than those from large second eggs even up to 12 days. The larger mass gain of from second eggs does not show that from last eggs grew more slowly. By way of example, consider two groups of ducklings, one that weighed, on average, 100 g at hatch and 1000 g at 20 d of age, and the other that weighed, on average, 90 g at hatch and 900 g at 20 d of age. An analysis of variance of these data would show an interaction between group and age, even though the had grown at the same rate. To examine differences in duckling growth rates, Erikstad et al. (1998) should have log-transformed their body mass data before analysis. To approximate such an analysis, we calculated that mean body mass of from last eggs was 88% of that of from second eggs at hatch (data in Erik-