The Primate Finger Maze was designed to overcome the time demands and the limited application of the Wisconsin General Test Apparatus to specific cognitive abilities of nonhuman primates. It was first described in an experiment by Tsuchida et al. (2003).

The Primate Finger Maze Test involves a problem-solving task wherein the subjects manipulate the food reward placed within the apparatus to the feeding box. The apparatus consists of four levels or steps with one end opening to an error box and the other to the next step (feeding box in the final step).

Mazeengineers offers the Primate Finger Maze.

The Primate Finger Maze is a 4-step non-correction-method-type finger maze originally described by Tsuchida et al. (2003) in their paper assessing the global cognitive abilities of long-tailed macaques. The puzzle apparatus was designed to overcome the time demands and the limited application of the Wisconsin General Test Apparatus to specific cognitive abilities of nonhuman primates. The Primate Finger Maze Test is based on puzzle feeders, which are popularly used as enrichment for primates (De Rosa, Vitale & Puopolo, 2003; Roberts, Roytburd, & Newman, 1999).

The Primate Finger Maze Test involves a problem-solving task wherein the subjects manipulate the food reward placed within the apparatus to the feeding box. The apparatus consists of four levels or steps with one end opening to an error box and the other to the next step (feeding box in the final step). Small slits in the apparatus face-aligned approximately with the spot the reward falls on the step, permit a single digit to manipulate the reward direction. The combination of small slits and error boxes require higher problem-solving capabilities such as pre-planning of the move, learning from, and recalling previously made errors. Additionally, the assay can be extended to evaluate fine motor skills as well.

The Primate Finger Maze can be applied in investigations involving pharmacological manipulations, brain lesions, and neurodegenerative and neuropsychiatric disorders to obtain an objective assessment of cognitive abilities. Other primate mazes include the Human Vertical Maze and the Virtual Star Maze. For human assays, click here.

The Primate Finger Maze is constructed using clear acrylic. The maze consists of four steps measuring 180 mm long and 280 mm wide with a height of 32 mm. Each end of the step has a 25 mm wide opening. One end of the step ends in a 32 × 55 mm error box that prevents the animal from accessing the reward. The position of the error box is alternated for every step. The end opposite the error box opens to the next step or to the feeding box in case of the last step. The feeding box allows the subject to pick up the food reward without any restriction. Both faces of the maze box are covered with clear plastic to allow visual access to the puzzle and the reward. On the side facing the subject, small slits for each step are present. The slits ensure that the subject can only use a finger to manipulate the food reward on the step. The apparatus can easily be mounted onto the mesh of the animal cage.

Cleaning the apparatus may not be necessary during a session; however, it is recommended that any distracting cues that may be left from previous trials are removed. An automated tracking and recording system such as the Noldus EthoVison XT may be used to assist with the observations.

Following is a sample protocol for Primate Finger Maze Test. The duration and number of trials and sessions may vary depending on factors such as age.

Acclimation

Perform acclimation trials to familiarize the subject to the apparatus and the experimenter and to associate the feeding box with reward. Place a food reward in the feeding box of the apparatus with the experimenter present. Allow the subject 120 s to collect the reward. Repeat the trials until the subject meets the success criterion in consecutive trials or until a maximum of 10 trials have been performed. If the subject fails to meet the criterion, repeat the acclimation session the following day.

Finger Maze Training

Once the subjects have been acclimated to the apparatus and have learned to retrieve rewards from the feeding box, begin the training sessions. First, place the food reward on the last step that opens to the feeding box. Repeat trials until the subject meets the learning criterion or until 10 trials have been completed. Use an intertrial interval of 10 s. Repeat sessions until the subject meets the learning criteria. Once the learning criterion is met, place the food reward on the step above and repeat the training process. Repeat the training process for each step in the fashion described above. Perform at least 2 sessions per day.

Finger Maze Learning Test (Random Test)

On successful completion of training, begin the learning test. Place the food reward in a pseudo-random manner on one of the steps and observe the performances. Perform 20 trials per day with an intertrial interval of 10 s for 2 days.

Finger Maze Retention Test

Following a retention period, evaluate the retention memory of the subjects. Perform the retention test in the same manner as the learning test.

Investigation of the effects of perinatal exposure to thiamazole on learning and memory

Inoue, Arima, Kato, and Ebihara (2014) investigated learning and memory performances of infant cynomolgus monkeys exposed perinatally to thiamazole in the Primate Finger Maze Test. Pregnant cynomolgus monkeys were divided into thiamazole and control groups. Thiamazole groups were orally administered either 2.0 or 3.5 mg/kg during the late gestational period (days 120 to 150 of gestation) while the controls did not receive any treatment. All animals were allowed for a natural birth. Infant cynomolgus monkeys began the Primate Finger Maze Test protocol starting postnatal day (PND) 253 using a piece of apple as the reward and training was performed in 2 daily, 10 trial sessions. No groups displayed any statistical differences in the number of sessions required to meet the training criteria. The learning test was performed between PND 279 to PND 334. Infants of the thiamazole groups displayed comparable performances in the learning test to that of the control groups. Success rates of the 2.0 and 3.5 mg/kg group were 98.3% and 97.0% on day one and 97.5% and 97.0% on day two, respectively, while the controls maintained a success rate of 94%. Retention memory was evaluated 2 months following the learning test and revealed a deterioration of memory in the thiamazole groups. While all groups showed lowered success rates on day 1 of the test, the success rate of controls jumped to 91% on day 2, which was comparable to the learning test performances. While the 2.0 mg/kg group showed an increase to 81.7% from 65.0%, the 3.5 mg/kg group’s success rate remained relatively similar on both days (65.0% and 68.0% on days 1 and 2, respectively).

Investigation of the effects of gestational and lactational exposure to TCDD in rhesus monkeys

Negishi et al. (2006) evaluated the learning abilities of the offspring of rhesus monkeys that were administered 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) during gestation and post- parturition. Pregnant rhesus monkeys were subcutaneously injected with 30 and 300 ng/kg of TCDD on gestational day 20, which was then followed by either 1.5 ng/kg or 15 ng/kg TCDD injections for every 30 days till postnatal day 90. The control group was administered vehicle (a mixture of toluene and DMSO) following the same regime. Following birth, offspring were weaned at the age of 12 to 14 months and subjected to the behavioral assay. Primate Finger Maze Training was performed using a piece of apple as a reward in 2 daily sessions consisting of 15 trials each for 5 days per week. Though the 300 ng/kg group displayed quicker learning than the controls, no significant effect of TCDD exposure on learning could be observed. The learning test was performed on two consecutive days in 16 trial sessions per day. As with the training performance observation, no effect of TCDD exposure could be observed.

Investigation of cognitive ability of long-tailed macaques

Tsuchida et al. (2003) evaluated the learning capabilities of experimentally naïve long-tailed macaques (Macaca fascicularis) in the Primate Finger Maze Test. The 7-year-old monkeys were trained in two, 15 trial sessions per day performed 5 days per week using a piece of apple as the food reward. Subjects were observed to make significantly more errors when being trained on the second last step in comparison to the other three steps. However, performances improved with each step, suggesting that the subjects were able to quickly acquire the maze learning. Learning tests were performed in two sessions consisting of a total of 32 trials following the final training session. A success rate of 91.07% was observed in the learning phase, which was significantly better than the chance performance of the animals.