Fig. 2.

Behavioural data showing the manual dexterity of six monkeys before the lesion (negative days on the abscissa) and post-lesion (positive days on the abscissa). The day of the lesion (day zero) is indicated by a vertical line. Each dot value corresponds to the total time (in seconds) needed by the monkey in a given session to empty the board containing 20 wells using the contralesional hand (see Fig. 1c). The median total time pre-lesion is given by the horizontal dashed line on the left part of each graph. Note the dramatic increase in total time immediately after the lesion, reflecting the deficit (increased difficulty to grasp the pellet using the precision grip). The values saturated at 200 correspond to sessions immediately post-lesion in which the monkey was unable to complete the task (empty the 20 wells). Then, there is a progressive recovery of manual dexterity. The recovery was considered as complete on the day indicated by the arrow, corresponding to the time point at which the monkey reached a plateau in the modified Brinkman board task (see Kaeser et al. 2010, 2011). The median total time at plateau of recovery is indicated by the horizontal dashed line on the right. In all monkeys, the post-lesion median total time was significantly longer than the pre-lesion median total time (p value given for each graph). The percentage of functional recovery was calculated by dividing the pre-lesion median total time by the post-lesion median total time at plateau, multiplied by 100. The percentage of functional recovery is given in each graph. For the interpretation of the data (see text), the volume of the cortical lesion in mm³ is indicated also in the graphs for each monkey. Note that Mk-BI exhibited a peculiar, bimodal behavioural pattern for this manual dexterity task, present, however, both pre-lesion and post-lesion: in some daily sessions, the monkey completed the task quickly (in about 25 s pre-lesion), whereas in other daily sessions, it took clearly more time (about 70–80 s pre-lesion as well). As the same bimodal distribution of total times was maintained post-lesion, it allows comparison of the median values. In line with a higher median value observed post-lesion for Mk-BI, note that for each of the two behavioural patterns, the variability across daily sessions was larger post-lesion as compared to pre-lesion, suggesting that the monkey was less regular and less comfortable with the task after the lesion. Moreover, pre-lesion Mk-BI exhibited the “quick” and the “slow” behaviours in equal proportions (50/50 %). After lesion (at plateau), the “slow” behaviour was more frequently present (73 % of the daily sessions) than the “quick” behaviour (27 %), also supporting the strong deficit induced by the lesion