Applying behavior‐analytic methodology to the science and practice of environmental enrichment in zoos and aquariums

Abstract

Environmental enrichment in zoos and aquariums is often evaluated at two overlapping levels: published research and day-to-day institutional record keeping. Several authors have discussed ongoing challenges with small sample sizes in between-groups zoological research and have cautioned against the inappropriate use of inferential statistics (Shepherdson, , International Zoo Yearbook, 38, 118-124; Shepherdson, Lewis, Carlstead, Bauman, & Perrin, Applied Animal Behaviour Science, 147, 298-277; Swaisgood, , Applied Animal Behaviour Science, 102, 139-162; Swaisgood & Shepherdson, , Zoo Biology, 24, 499-518). Multi-institutional studies are the typically-prescribed solution, but these are expensive and difficult to carry out. Kuhar ( Zoo Biology, 25, 339-352) provided a reminder that inferential statistics are only necessary when one wishes to draw general conclusions at the population level. Because welfare is assessed at the level of the individual animal, we argue that evaluations of enrichment efficacy are often instances in which inferential statistics may be neither necessary nor appropriate. In recent years, there have been calls for the application of behavior-analytic techniques to zoo animal behavior management, including environmental enrichment (e.g., Bloomsmith, Marr, & Maple, , Applied Animal Behaviour Science, 102, 205-222; Tarou & Bashaw, , Applied Animal Behaviour Science, 102, 189-204). Single-subject (also called single-case, or small-n) designs provide a means of designing evaluations of enrichment efficacy based on an individual's behavior. We discuss how these designs might apply to research and practice goals at zoos and aquariums, contrast them with standard practices in the field, and give examples of how each could be successfully applied in a zoo or aquarium setting.

Using Principles from Applied Behaviour Analysis to Address an Undesired Behaviour: Functional Analysis and Treatment of Jumping Up in Companion Dogs

“…ABA is the discipline in which tactics derived from the principles of learning and behaviour are applied to improve socially significant behaviour, and experimentation is used to identify the variables responsible for the improvement in behaviour [13]. Although ABA has proven highly useful in the human behaviour change arena for addressing a number of undesired behaviours (e.g., aggressive responses, anxiety, destructiveness, pica, self-injurious behaviour, stereotypic behaviour; [10,14]), the integration of ABA into the applied and/or clinical animal behaviour setting is still in its early stages [15][16][17]. Some of the unique features of ABA that are specifically valuable for animal behaviour professionals in facilitating behaviour change in animals and their caregivers ( [18,19]) include (a) a function-based approach to the identification of problematic behaviours and related development of interventions (e.g., functional analysis [FA]; [20]); (b) scientific methods that focus on continuous data collection of the individual's behaviour to implement and evaluate progress in an evidence-based approach (single-case research methods; [SCRMs]; [16,21]); (c) emphasis on working with owners and caregivers in the settings where the problematic behaviours occur; and (d) focus on humans' and animals' wellbeing in the decision-making process.…”

“…The first foundational pillar comprises a unique research methodology-single-case research methods (SCRMs) [29]. Research in the experimental and applied analysis of behaviour has extensively used inductive methods of investigation, in which many data points are collected on a few participants with the aim of assessing causal relations between environmental events and behaviour [16]. Behaviour-analytic research carefully identifies and defines target behaviours, assesses interobserver agreement and takes steps to minimize the influence of sources of bias, while maximising experimental control [21].…”

“…Behaviour-analytic research carefully identifies and defines target behaviours, assesses interobserver agreement and takes steps to minimize the influence of sources of bias, while maximising experimental control [21]. This is typically achieved by the use of SCRMs which focus on questions involving the application of behaviour-change interventions to individuals rather than questions about population-level patterns [16,30]. SCRMs should be contrasted from case studies which are often unsystematic, subjective, and qualitative in nature [16,21].…”

“…This is typically achieved by the use of SCRMs which focus on questions involving the application of behaviour-change interventions to individuals rather than questions about population-level patterns [16,30]. SCRMs should be contrasted from case studies which are often unsystematic, subjective, and qualitative in nature [16,21]. Key features of SCRMs include baseline assessment, continuous measurement, stability of performance, and use of different phases [31].…”

Patterns of animal–enrichment interaction in captive brown bears

“…The four PAIs described represents an easy methodology with simple statistical analysis that personnel involved in enrichment evaluation can perform. This simple methodology could facilitate the elaboration of a general database among the zoos to computerize, store and share information about enrichment programs (e.g., timetables, species, PAIs, types of enrichment devices or individual and specific variations about enrichment device; Alligood, Dorey, Mehkam, & Leighty, ; Shepherdson, ).…”