Top view of the three columns in high brightness mode.
Aggregation in the final aggregate was also faster in high brightness than in the other two brightness settings.
Nevertheless, the evolution of the number of woodlice per aggregate was less regular in high brightness than in other condition.
This absence of aggregation was observed in high brightness without shelter and could result from the increase in activity with light intensity [51], [52], [15].
After 20 minutes of experiments, in high brightness, the number of woodlice in the final aggregate slightly decreased to stabilize at around 20 woodlice.
Indeed, in the first 15 minutes, global aggregation was faster in high brightness than in medium or low brightness (Figure 3a; Kruskal-Wallis test, KW values>6.66, p<0.05 followed by Dunn's test: C≠B and C≠A, p < 0.05).
Indeed, at 10 minutes, 60% of woodlice were already aggregated in high brightness compared to 35% in medium and low brightness (Figure 3b; χ2 test, df = 2, χ2 = 6.79, p = 0.034).
However, the high variability of results did not permit any statistical differences to be observed (Figure 3b; comparison between 10, 30, and 45 minutes in high brightness; Friedman test, Fr = 3.937, p = 0.14).
Hence, in high brightness, after reaching a maximum very quickly, the number of aggregated woodlice progressively decreased during the experiments (Figure 3a; comparison between 10, 30, and 45 minutes; Friedman's test, Fr = 11.68, df = 2, p<0.01).
