Table 1.
Distribution of focal chicks with different numbers of heavier competitors at d15 in the two brood size treatments and in total.
Table 2.
Summary of models testing the effect of the brood size treatment on other potential continuous measures of early-life adversity and current state.
Table 3.
Correlation matrix showing how potential measures of early-life adversity and current state relate to one another.
Fig 1.
Effects of developmental competition on judgment bias.
(A) Latency to probe as a function of stimulus valence for birds in the two brood-size treatments. Data are mean ± 1 s.e. latency to probe in the cognitive bias test trials. (B) The same data shown in panel A standardised so that the latencies to probe POS and NEG are 0 and 1 respectively. This standardisation removes differences in speed to probe POS and NEG between treatments, and hence reveals the differences in the shapes of the generalisation gradients between treatments. Note that this standardisation is for visualisation purposes only and was not used in the data analysis (see text for details). (C) Latency to probe as a function of stimulus valence for birds at the top of the weight hierarchy in the nest (0 or 1 heavier competitors) and birds at the bottom of the weight hierarchy (2–6 heavier competitors). Data are mean ± 1 se latency to probe in the cognitive bias test trials. The dichotomization of the data into the groups 0–1 and 2–6, is for visualisation purposes only; all statistical analyses were conducted using the number of heavier competitors as a continuous predictor variable. (D) The same data shown in panel C standardised so that the latencies to probe POS and NEG are 0 and 1 respectively. Note that the standard errors shown on all of the plots in this figure give a false impression (underestimate) of the significant differences between the groups due to the fact that birds from the same genetic family are present in both treatment groups.
Fig 2.
Effects of developmental telomere attrition on judgment bias.
Latency to probe as a function of telomere attrition score. Telomere attrition score is the difference between telomere length at d4 and d55 adjusted for regression to the mean (see methods for details); negative values correspond to greater telomere loss. Data points are the mean ± 1s.d. of the ln latency to probe in the 24 judgment bias test trials with ambiguous stimuli (i.e. valences 2–4) for each bird (n = 22). The solid line shows the predicted regression line derived from the model described in the main text.
Fig 3.
Effects of genetic family on judgment bias.
Latency to probe as a function of stimulus valence for each of the 8 genetic families. Families FH6 and P11 contained 3 birds and the other families contained 4 birds. Data are mean ± 1 s.e. latency to probe in the judgment bias test trials. The dotted lines show the mean speed for each family (the mean of the mean latencies to probe POS and NEG). The top row of families were ‘pessimists’, meaning that their mean latencies to probe ambiguous stimuli were predominantly greater than the mean speed and more similar to NEG, whereas the bottom row were ‘optimists’, meaning that their mean latencies to probe ambiguous stimuli were predominantly less than the mean speed and more similar to POS.
Table 4.
Correlation matrix for pessimism index, mean speed and impulsivity.
Fig 4.
Effects of developmental telomere attrition on the ‘hunger factor’.
Data points are the scores extracted by the PCA (see text for details) for each of the birds for which we also had telomere attrition data (n = 19). Negative values of the hunger factor indicate birds with a ‘hungrier’ cognitive phenotype, whereas positive values indicate a more sated cognitive phenotype. The solid line shows the predicted regression line derived from a simple regression.