Piecewise‐Continuous Sampling: A Method for Minimizing Bias and Sampling Effort for Estimated Metrics of Animal Behavior

ABSTRACT Capturing qualitative features of animal behavior requires sampling occurrences of behavior over time. A continuously sampled dataset can capture many qualitative features of animal behavior; however, it can be very time‐consuming and sometimes infeasible. Instantaneous sampling can reduce the amount of labor required but will miss the fine structure present in continuously recorded datasets. We therefore explored a synthesis of these techniques which we call piecewise‐continuous sampling, where there are multiple continuous samples during randomly dispersed time intervals. To test the efficacy of this technique, we collected a continuous behavioral dataset of harvester‐ant workers and then randomly sampled from this dataset using continuous sampling, instantaneous sampling, and piecewise‐continuous sampling, each with approximately the same total amount of observation time. We then measured errors associated with measuring various attributes of animal behavior for each type of sampling technique. The different sampling techniques had dissimilar strengths and weaknesses. Using a multi‐objective optimization technique (desirability functions), we show that piecewise‐continuous sampling can be used to explore the gradient between continuous sampling and instantaneous sampling to find the strategy which simultaneously minimizes effort and error.