Behavioral monitoring in zoos and aquariums have often relied on observational monitoring of animals, as this provides a non-invasive and low-cost approach to understand behavior.  Observational methods rely on a sampling protocol, a set of rules to decide how and when behaviors are recorded and for whom.

In ZooMonitor, you can use the Behavioral Observations module to record data using a combination of several standardized sampling methodologies commonly applied in behavior research.  Although different terms have described these methods, these common approaches include: 1) all-occurrence sampling (c.f., “all occurrences”, Altmann, 1974; “continuous”, Martin and Bateson, 2007), 2) interval sampling (c.f., “instantaneous sampling”, Altmann, 1974; Martin and Bateson, 2007), 3) continuous sampling (“c.f., “focal-animal”, Altmann, 1974; “continuous”, Martin and Bateson, 2007), and 4) group scan sampling.

At the most detailed level, continuous sampling records the full duration of behaviors for an individual.  Although this approach provides a perfect record of the behavior pattern, this sampling methodology is intensive and often limited to a single animal or specific behavior of interest. 

For many studies, a combination of interval and all occurrence sampling are often sufficient and allow more behaviors to be recorded for more focal individuals.  Interval sampling involves recording behavior at pre-determined time points (see Choosing the Best Sampling Methods for advice on choosing these intervals).  All occurrence sampling, on the other hand, involves counting each time a behavior occurs during an observation.  These approaches, when combined, allow you to track all occurrences of rare behaviors that would likely be missed on the interval time points, while also accommodating a general approximation of the behavior pattern (i.e., time budget) using interval sampling.

These two types of behavior patterns, that is brief, infrequent behaviors and long duration behaviors, are typically categorized as “events” and “states”, respectively.  Ultimately, it may be challenging to make a clear distinction whether a behavior should be categorized as an event or state and it may be necessary to track that behavior using multiple methods.

These observational sampling methods just described (continuous, interval, and all occurrence) are typically employed for focal animal sampling, where one or more individuals may be observed.  In scenarios with a large group of animals (e.g., 50+ flamingos), these individual-focused strategies are impractical.

When observing large groups of animals, a sampling strategy termed group scan sampling is often used.  In group scan sampling, you’ll count the number of animals are engaged in a given behavior.  For instance, you might score 30 flamingos standing, 10 walking, and 10 engaged in social behaviors like bickering.  With group scan sampling, your entire group of individuals is your focus.

Lastly, it’s important to consider how data from these sampling strategies are interpreted.  Data from continuous and interval sampling are often used to determine a time budget, and are expressed as percent of time.  All occurrence data provides a count (i.e., frequency) of the behavior that can be combined with the session duration to generate a rate value (occurrences per time unit).

Next Steps

• For more resources on behavioral research methods, we recommend the Methods for Animal Behavior DVD (https://www.aza.org/methods-for-animal-behavior-research-dvd) and Measuring Behaviour: An Introductory Guide by Martin and Bateson (2007).
• Continue onto Creating a Behavioral Observation Project to get started.