Fig 1.
Relationship between average winter temperature and March body condition of pink-footed geese Anser brachyrhynchus.
Average March body condition was assessed from the abdominal profile index (API) of (a) females and (b) males. Lines indicate least square fits from the linear mixed model (Table 1), and bars indicate standard errors. N indicates the number of geese included in the analysis.
Table 1.
Fixed effects output of the general linear mixed model to explain temporal persistence of the winter carry-over effect on pink-footed goose Anser brachyrhynchus body condition at three consecutive spring staging sites (Denmark, Trøndelag and Vesterålen).
Presented estimates are coefficients from a model with API (abdominal profile index) as response variable, “Year” and “Bird ID” as random effects and “Day of Month”, “Winter temperature” and the interaction between these as fixed effects. Winter temperature is the average December-February temperature (°C) in the preceding winter, and N the number of geese (females & males) with abdominal profile index (API) assessments for all combinations of month and staging sites. Day of month was fitted as a continuous variable (covariate).
Fig 2.
Relationship between early (March) and late (May) spring body condition of pink-footed geese Anser brachyrhynchus.
Average body condition was assessed from the abdominal profile index (API) of (a) females and (b) males. Black lines indicate the linear fit with 95% confidence limits, and grey dashed lines indicate hypothetical fits corresponding to no compensation (slope = 1) and full compensation (slope = 0). N indicates the number of geese included in the analysis.
Table 2.
Fixed effects model output of the general linear model to explain the effect of supplementary feeding on April API of pink-footed geese Anser brachyrhynchus.
Presented estimates are coefficients from a model with “Year” and “Bird ID” as random effects and “Day of month”, “Sex”, “Supplementary feeding”, “Sex * Supplementary feeding” and “Year * Supplementary feeding” as fixed effects. The supplementary feeding variable distinguishes between birds with API assessments inside and outside areas with supplementary feeding, and N indicates the number of geese included in the analysis. Day of month and Year was fitted as continuous variables (covariates).