Most of the measures to be described in this article are in some way designed to test and challenge the capacity for non-human primates to perform skilled manipulation of a device or challenging retrieval of food rewards. These functional tests are to some extent all dependent on the capacity to make what Lawrence and Hopkins (15) called “relatively independent finger movements” (RIFM). That is, the capacity to execute “fractionated” digit movements, best exemplified by the control of index finger and thumb.

These authors were careful to use the term “relative”: in fact, even in humans, there are limits to the extent to which the different digits can be moved in a completely independent manner (16). The thumb is the exception to this rule. Studies have shown that there are both biomechanical and neural constraints on independent finger movements (17, 18). The biomechanical constraints arise because of the complex interactions between the many different muscles acting on the digits. Some muscles (such as the extensor digitorum communis and the flexor digitorum profundus, FDP) act on more than one digit. Their motor units are partitioned in such a way that when one digit is moved, it allows associated movements of neighbouring digits. Compared with humans, macaque monkeys have a somewhat restricted repertoire of grasps, partly due to the fact that the monkey has a relatively shorter thumb length compared with finger length (18). Macaques also have more interconnections between the tendons of the FDP and the origins of the lumbrical muscles (18), which increases the mechanical constraints on the independence of individual digits.

Although a small number of muscles actually control the forces exerted at the fingertip (such as FDP), many others are needed to stabilise the articulated bony chain that extends from the shoulder through to the elbow, wrist, hand and phalanges. Only with such stability can accurate tip-to-tip forces be developed. The hand has a very large number of degrees of freedom, and it is very difficult to monitor movements and forces at every single joint, let alone the activity in all the different muscles that underpins this movement.

The early studies of Lawrence and Kuypers (9) on the effects of pyramidotomy on fine finger movements in the macaque monkey adapted the original Klüver board (Fig. 2a). This board contains food wells of varying diameter and depth; the smallest wells required precise co-ordination of thumb and index finger to “winkle out” the reward. After section of the pyramidal tract, monkeys were unable to retrieve food, especially from the smallest wells. Very young monkeys, who lack the capacity for RIFM and in whom the CST is immature, also fail on these tests (15).

Monkeys use visual cues to direct their reach to the appropriate well and to orientate their hand and digits to grasp the reward; once contact with the well and reward is made, additional tactile/haptic cues help the monkey to retrieve the reward. Brinkman and Kuypers (19, 20) developed a board which forced monkeys to use visual rather than tactile cues to retrieve food pellets (Fig. 2b). In this board, protruding pegs, of the same diameter and shape as the food pellets made it difficult to use touch alone to find the real pellets; in addition, slots which gave access to the real pellets were cut into the board at a variety of different angles, again forcing the monkey to use vision to guide its approach to the reward. Only monkeys with intact eye–hand control could retrieve food pellets with normal speed and accuracy (Fig. 2b).

Monkeys show hand preference for different tasks, so it may be important to determine this for the particular task that is adopted.

A modified Brinkman board was used by Freund et al. (21) to test macaque monkeys with incomplete spinal lesions (Fig. 3d), in which some subjects were treated with the neurite growth inhibitor, anti-Nogo A, to promote recovery. In studies of this kind, there are inevitably large variations in the size and type of lesion, so the methods used have to be very sensitive to detect any consistent changes in control vs. treated animals. The board used by Freund et al. (21) had both vertical and horizontal slots (Fig. 3d). Intact, control monkeys find it more difficult to retrieve pellets from the horizontal slots compared with the vertical slots, probably because they have to adopt a rather unusual hand orientation to do so. Lesioned monkeys often showed much bigger impairments on the horizontal slots (as measured by the number of pellets retrieved in a fixed test time of 30 s) than on the vertical slots. In a further examination of these data, Freund et al. (22) measured the “contact time.” This was defined as the time interval between the first contact of the index finger with the food pellet and the final ­successful grasp of the pellet utilising pad-to-pad opposition of the index finger and thumb. This contact time is a useful additional measure because it focuses on the dextrous grasp and excludes other components of the retrieval score in the timed period, such as the reaching time and any delay in the monkey eating or chewing the reward before starting a new retrieval.

Nudo et al. (23) used a modified Klüver board to examine the prehensile performance of squirrel monkeys. These authors counted the number of attempts (“flexions per retrieval”) made by the monkey to retrieve the reward (a banana flavoured pellet) from the board. This approach was important in demonstrating the plasticity of the motor system after an experimental cortical lesion, and in demonstrating the power of intensive rehabilitation in promoting plasticity and functional recovery (24). Moore et al. (25) used timed tests on a Klüver-type board to assess changes in dexterity in aged macaques. Owl and squirrel presented with a board of varying well size rapidly adapt a particular two or three digit strategy to optimise rapid food retrieval, and Xerri et al (26) showed that this caused long-term plastic expansion of the representations of these digits in the somatosensory cortex. Timed tests on a Klüver board were used by Belhaj-Saif and Cheney (27) to assess hand function after unilateral pyramidal tract lesions in macaques, and to show the compensatory function of the rubrospinal tract.

In a further development of the Brinkman board, Haaxma and Kuypers (28) designed a “rosette” device that presented the food in a narrow slot, allowing the monkey to insert its index and thumb on either side of the food pellet (Fig. 2c). The slot was highlighted by a white line which cued the monkey to orientate its hand and digits in such a way as to retrieve the reward. Additional “blind” slots were present: these did not give access to the pellet, and were not highlighted. The device could be rotated between trials, so that the highlighted slot was presented at different orientations on every trial. Monkeys with normal eye–hand coordination and able to correctly use the visual cue to appropriately plan its grasp rarely make errors on this task; those rendered apraxic by lesions to the occipito-frontal white matter lesions were at chance as to which orientation they chose, often inserting their digits into the blind slots (28).

Rotating the Klüver board containing the rewards makes additional demands on the monkey’s ability to plan its reach-to-grasp movement precisely so that it is timed to coincide with a moving target. This test was first introduced by Asanuma and Arissian (29) and subsequently used by Kaeser et al. (30), who used a board rotating at 10 revolutions per minutes, and varying the direction of rotation from clockwise to anticlockwise. They used this test to look for a second phase of recovery in fine hand function after monkeys were treated with autologous stem cells derived from the prefrontal cortex.

Examples of both the static and rotating Brinkman board can be found at: http://www.unifr/neuro/rouiller/motorcontcadre.htm.

A potential drawback of the Klüver board is that it does not allow easy visualisation of individual digit movement, nor of the dynamics and kinematics involved. A pioneering study by Hepp-Reymond and Wiesendanger (31) used a small hand-held device that monitored the precision grip forces exerted by the monkey, and this was combined with simultaneous EMG recordings from hand muscles (32). The force required was indicated to the monkey by a visual display. This approach demonstrated a significant deterioration in the level of precision grip force that follows damage to the pyramidal tract, and also showed that the rate at which grip force increased was also much slower, reflecting the slowness and weakness of movement mentioned in the Introduction.

Another approach was used by Muir and Lemon (33): this required the monkey to exert independent control of the index finger and of the thumb to move two spring-loaded levers into a required target zone, and hold them both there for 1 s in order to earn a food reward (Fig. 3a–c; (34, 35)). The displacement of each digit was in the order of a few mm, and forces required were <1.0 N. A cover mounted above the manipulandum levers restricted access to the thumb and index and prevented other digits from contributing, thereby highlighting the fractionation of digit movement and hand muscle activity (36, 37). Using a robotic device it is possible to simulate spring loads with different properties and these can be changed in either a blocked manner or trial-by-trial.

Other investigators have used a small reward mounted in a device that constrains the monkey to use a precision grip to retrieve it. Darian-Smith (38) reviewed the use of high speed digital video to capture macaques grasping a reward held in a device that required the monkey to develop different precision grip force levels to release the reward; these ranged from 0.5 to 2.0 N (see Fig. 4). This approach allows the precise kinematics of the grasp to be documented in great detail. The time taken to collect a glucose tablet held in a spring clip was used by Galea and Darian-Smith (39) to document the maturation of precision grip in neonatal macaques; while such movements are absent at birth, the speed and fluency of the grip develop rapidly over the first 5–6 months of life, and then show a slow but prolonged improvement into the third year. Fine movements of the digits are essential for social grooming, and young macaques are reported to begin grooming at around 5–6 months old (40).

Digital video was also used by Sasaki et al. (41) to demonstrate that unilateral lesions to the lateral corticospinal tract at the C5 level, which would have interrupted descending corticospinal control of the spinal segment in the cervical enlargement supplying hand and digit muscles, caused an immediate impairment of precision grip. Over the following weeks, the slowness and clumsiness of the grip recovered significantly, and this was again documented by digital video. Recovery may have involved a number of different mechanisms (42), including ipsilateral propriospinal systems and uninjured but slowly conducting corticospinal fibres (43), and possibly a contribution due to collateral sprouting from intact corticospinal fibres descending on the unlesioned side of the cord (44).