Model results for cross pattern experiments. Summary of honeybee behaviour and simulated bees’ performance for the discrimination of simple cross patterns.

<p>In the behavioural experiments [<a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1005333#pcbi.1005333.ref011" target="_blank">11</a>] different groups of honeybees were differentially trained on a particular cross pattern pair, one rewarding (CS+) and one unrewarding (CS-). Blue: honeybee result, percentage of correct CS+ pattern selections after training. Red: performance accuracy of the DISTINCT simulated bee. Green: performance accuracy of the MERGED simulated bee. Behaviour Error bars for honeybee shows standard deviation. Error bars for models shows standard deviation of the Kenyon cell similarity ratios (as a percentage, and centred on the simulated bee performance value; which was equivalent to average Kenyon cell similarity ratio over the all simulation trials). <b>(a)</b> Discrimination of 90° cross and 45° rotation of this pattern. The DISTINCT simulated bee easily discriminates the patterns but honeybees cannot, the simulated bee based on the MERGED model achieved below 60% accuracy. <b>(b)</b> Discrimination of a 22.5° cross pattern and the same pattern rotated through 90°, both of the simulated bees and real honeybees can discriminate these cross patterns.