Fig. 6

Survival probabilities corresponding to the events “leaving the aggregation” and “joining the aggregation”. The survival probabilities have been estimated by the Kaplan–Meier method for the events “leaving the aggregation” (a–c) and “joining the aggregation” (d–f) during the period of winter aggregations for different water temperatures (a, d), different fish sizes (b, e) and the different times of day (c, f). These curves, which represent the survival function as a function of time, show the probability that the event of interest has not yet occurred by this time point. For example, in a, the probability for an individual of not leaving the aggregation after time = 100 (25 h), in other words, the probability of staying in the aggregation after 25 h spent inside, is 0.03 at [9; 14[ °C, 0.10 at [7; 9[ °C, 0.20 at [5; 7[ °C and 0.31 at [3; 5[ °C. The curves have been computed on the four studied winters together. The shading around the curves represents the 95% confidence interval of the mean survival curve. It is sometimes barely discernible because it is very narrow