Assessment of foraging devices as a model for decision-making in nonhuman primate environmental enrichment

Abstract

Continued progress to move evidence-based best practices into community and regulatory animal welfare standards depends in part on developing common metrics to assess cost, benefit, and relative value. Here we describe a model approach to evidence-based evaluation and an example of comprehensive cost-benefit assessment for a common element of environmental enrichment plans for laboratory-housed nonhuman primates. Foraging devices encourage a species-typical activity that dominates the time budget of primates outside captivity and provide inherent cognitive challenges, physical activity demands, and multi-sensory stimulation. However, their implementation is not standard, and is challenged by perception of high costs and labor; nutritional and health concerns; and identification of best practices in implementation (that is, device types, food type, frequency of delivery and rotation). To address these issues, we directly compared monkeys' engagement with different foraging devices and the comprehensive cost of implementing foraging opportunities. We recorded 14 adult male cynomolgus monkeys' interactions with 7 types of devices filled with a range of enrichment foods. All devices elicited foraging behavior, but there were significant differences among them both initially and over subsequent observations. Devices that afforded opportunity for extraction of small food items and that posed manipulative challenge elicited greater manipulation. The cost of providing a foraging opportunity to a single monkey is roughly US$1, with approximately 80% attributable to labor. This study is the first to perform a rigorous cost-benefit analysis and comparison of common foraging devices included in environmental enrichment. Its broader significance lies in its contribution to the development of methods to facilitate improvement in evidence-based practices and common standards to enhance laboratory animal welfare.

Figure 1.

Illustration of model process for assessment of environmental enrichment strategies.

Figure 2.

Filled foraging devices used in study 1. Novel devices included a (A) pipe, (B) food feeder, (C) treat dispenser, and (D) combination panel. Familiar devices included a (E) challenger ball and F) paint roller.

Figure 3.

Food feeders with (A) circular hole and (B) rectangular hole covers used in study 2.

Figure 4.

Number of intervals (mean ± SEM) with manipulation for (A) pipe, (B) combination panel, and (C) challenger ball averaged across all observation periods for each of the 3 phases. Symbols refer to significant differences between devices. Post hoc tests showed that for the pipe, manipulation during Phase 1 was significantly higher than that during phase 2 (P < 0.0001) and phase 3 (P = 0.001). For the combination panel, manipulation in Phase 3 was significantly higher than during phase 2 (P < 0.0001) or phase 1 (P = 0.0004). For the challenger ball, manipulation in Phase 3 was significantly higher than Phase 2 (P = 0.0003), and Phase 1 (P = 0.006).

Figure 5.

Number of intervals (mean ± SEM) with contact for all devices during (A) the initial observation period and (B) averaged for the 1- and 2-h observation periods. The dashed horizontal lines indicate the maximum possible number of observation intervals. The dashed vertical lines separate the 2 devices that held small forage foods from those that held larger fruit or vegetable pieces. Symbols refer to significant differences between devices. Post hoc tests showed that for the initial observation, manipulation of the paint roller was significantly higher than that for the challenger ball (P = 0.006), food feeder (P = 0.002), and treat dispenser (P < 0.0001). Manipulation of the combination panel was significantly higher than that of the challenger ball (P = 0.04), food feeder (P = 0.01), and treat dispenser (P < 0.0001). Manipulation of the pipe was significantly higher than that for the food feeder (P = 0.04) and treat dispenser (P = 0.0003). Manipulation of the challenger ball was significantly higher than the treat dispenser (P = 0.04). For sustained contact across the 1- and 2-h observation periods, manipulation of the pipe was significantly greater than that of the paint roller (P = 0.0014), combination panel (P = 0.004), challenger ball (P = 0.001), food feeder (P = 0.002), and treat dispenser (P = 0.009).

Figure 6.

Number of intervals (mean ± SEM) with contact for food feeders having 2 types of covers during the initial observation period, the average of the 1- and 2-h observation periods, and the 24-h observation period. The dashed horizontal lines indicate the greatest number of observation intervals possible. Asterisks indicate significant (P < 0.05) differences between covers. Significant main effects for device cover and observation period were observed, as well as a significant interaction between them. Post hoc tests showed that manipulation of the food feeder was significantly (P = 0.0001) higher for the circular cover than for the rectangular cover for the sustained manipulation at the 1- and 2-h observations.