Personality traits affect learning performance in dwarf goats (Capra hircus)

Abstract

A wide range of species exhibit time- and context-consistent interindividual variation in a number of specific behaviors related to an individual's personality. Several studies have shown that individual differences in personality-associated behavioral traits have an impact on cognitive abilities. The aim of this study was to investigate the relationship between personality traits and learning abilities in dwarf goats. The behavior of 95 goats during a repeated open field (OF) and novel object test (NO) was analyzed, and two main components were identified using principal component analysis: boldness and activity. In parallel, the goats learned a 4-choice visual initial discrimination task (ID) and three subsequent reversal learning (RL) tasks. The number of animals that reached the learning criterion and the number of trials needed (TTC) in each task were calculated. Our results show that goats with the lowest learning performance in ID needed more TTC in RL1 and reached the learning criterion less frequently in RL2 and RL3 compared to animals with better learning performance in ID. This suggests a close relationship between initial learning and flexibility in learning behavior. To study the link between personality and learning, we conducted two analyses, one using only data from the first OF- and NO-test (momentary personality traits), while the other included both tests integrating only animals that were stable for their specific trait (stable personality traits). No relationship between personality and learning was found using data from only the first OF- and NO-test. However, stability in the trait boldness was found to have an effect on learning. Unbold goats outperformed bold goats in RL1. This finding supports the general hypothesis that bold animals tend to develop routines and show less flexibility in the context of learning than unbold individuals. Understanding how individual personality traits can affect cognitive abilities will help us gain insight into mechanisms that can constrain cognitive processing and adaptive behavioral responses.

Keywords: cognition; discrimination learning; goats; novel object test; open field test; personality; serial reversal.

Figure 1

Timeline of the experimental procedure test sequence, housing condition and age of the dwarf goats (given in weeks).

Figure 2

Lateral view of a goat inside the compartment with learning device: 1 = entrance (there is only space for one goat at a time at the device to avoid observational learning by the pen mates); 2 = collar with responder for individual recognition at the device; 3 = light beam indicating when a goat leaves the device; 4 = yoke to put only the head through; 5 = computer screen; 6 = buttons to choose a stimulus; 7 = water bowl for reward dispense.

Figure 3

Symbol sets for training: training set 1 (A); training set 2 (B). Symbol sets during learning and reversal learning tasks: Initial discrimination learning task [ID, (C)] and the three reversal learning tasks: reversal learning 1 [RL1, (D)]; reversal learning 2 [RL2, (E)]; reversal learning 3 [RL3, (F)]. The rewarded symbols are marked with a square (for illustration purposes only). The position of the different stimuli changed after each individual choice.

Figure 4

Performance of different groups of goats in the reversal learning task: (A) proportion of goats that reached the learning criterion during reversal tasks RL1-RL3. (B) Number of trials to reach the learning criterion (TTC) in reversal tasks RL1-RL3. The groups Q1-Q4 indicate the learning performance during the initial discrimination task (ID): Q1: TTC in ID = 20, Q2: TTC in ID = 40–80, Q3: TTC in ID = 100–200, Q4: TTC in ID>220. Data are presented as least squared means and standard errors (LSM ± SE); *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5

Performance of goats with stable personality scores in activity (A) or boldness (B) in the reversal learning tasks: (A) number of trials to reach the learning criterion in the reversal tasks RL1-RL3 of goats assigned to the inactive (black bars) vs. active groups (gray bars). (B) Number of trials to reach the learning criterion in the reversal tasks RL1-RL3 of goats assigned to the groups unbold (black bars) vs. bold (gray bars). Data are presented as least squared means and standard errors (LSM ± SE); *p < 0.05.