Food preferences and demand in the brushtail possum (Trichosurus vulpecula)

The common brushtail possum (Trichosurus vulpecula) has been reported to eat vegetation, fruit, invertebrates, and occasionally fungi, eggs and meat in the wild. The relative preference between food types found in the wild, however, has not been investigated in a controlled laboratory study. In this series of experiments, single and paired preference assessments and demand procedures were conducted systematically under laboratory conditions to investigate the preference and demand for food types reportedly consumed by possums. In Experiment 1, 20 possums were used in a single stimulus assessment where the consumption of individually presented food items was measured. The foods presented were berries, raw chicken, egg, fivefinger leaves, locusts and mushrooms. More than 75% of possums consumed berries, locusts and mushrooms but fewer than 50% of possums consumed fivefinger, raw chicken and eggs. In Experiment 2, 12 possums were used in a paired stimulus assessment to establish relative preference for the same foods. The results showed that no single food was preferred by all possums. Overall locusts were the most preferred food, followed in order of preference by berries, egg, mushrooms, chicken and foliage. The single stimulus preference assessment confirmed the palatability of foods. The paired stimulus assessment provided a rank order of food preferences. In Experiments 3 and 4, the demand for these foods were measured under concurrent progressive-ratio (PR) and fixed-ratio (FR) schedules of reinforcement. In Experiment 3, the same six food types were used and every possible food pair (30 pairs in total) was presented to six possums in concurrent PR and FR 30 schedules. Exponential models of demand were applied to consumption rates and Pmax, break point and cross point values were generated. The rank orders for each parameter were compared. Overall, more responding was allocated to the PR schedule when a preferred food was available compared to the constant FR schedule. Cross points were larger for chicken, egg and locust, however, stable responding under the constant fixed-ratio schedule was not observed. The aim of Experiment 4 was to investigate possible reasons for the lack of stable responding. The experimental procedure of Experiment 3 was replicated with four food pairs and the constant FR schedule was alternated between 30 and 10 responses across sessions. Responding under the constant FR schedule was similar to that in Experiment 3 but the demand for foods under the PR schedule was similar. In Experiment 5, the same methodology in Experiments 3 and 4 was used except that each ratio requirement of the PR schedule was increased every five days, termed a PFR schedule. The same food pairs were used (berries and egg, and chicken and mushrooms). The same descriptions of demand were found in Experiment 5 as in Experiment 4, with higher demand for egg and chicken compared to berries and mushroom. A comparison of the linear (Hursh, Raslear, Shurtleff, Bauman, & Simmons, 1988) and exponential demand (Hursh & Silberberg, 2008) models showed that the exponential model provided better fits to the data. In addition, the cross price model (Hursh, Madden, Spiga, DeLeon, & Francisco, 2013) plotted with the exponential model of demand accounted for better cross points than the linear model. In Experiment 6, the aim was to confirm the similarity in performance under concurrent schedules where the incrementing schedules increased within or across sessions. Twelve possums were exposed to schedules that increased within a session using the same procedure as Experiments 3 and 4; and schedules that increased across sessions in a semi-replication of Experiment 5 where each ratio requirement was in place for one day. The progression of the incrementing schedule was also varied between a geometric sequence (basis 2), and an arithmetic sequence (step 5) to ascertain if progression type affected the demand for foods. The food pairs of berries and egg, and two new foods of a barley and coco-pop® mix and rolled oats were tested. The same response rate patterns were observed under the geometric and arithmetic progressions. The parameters of the exponential and cross price demand models predicted estimates that differed in their description of demand across PR and PFR schedules, progression and food type. The break points and cross points were larger under PFR FR schedules and geometric progressions. In conclusion, the preference assessments (Experiments 1 and 2) identified that locust was the most preferred food across possums but individual food choices were idiosyncratic. The demand procedures (Experiments 3 – 6) identified that possums are opportunistic in their food choice as they will respond for all foods at low ratio requirements and will respond at higher ratio requirements for more preferred foods even when another food is available for a lower cost. This series of experiments also highlighted that systematic studies are required to investigate what possums will eat and what they prefer to eat