Testing the efficiency of structured and unstructured surveys for detecting a small population of Jack in the Pulpit (<i>Arisaema triphyllum</i>) plants

Detecting inconspicuous species with low abundance is challenging, yet failure to detect these species can have important conservation implications. Few experimental studies have been conducted to assess how survey protocols affect plant species detectability. We tested the efficiency of structured (transect-based) and unstructured (unplanned search) surveys for the detection of a small, simulated population of Jack in the Pulpit ( Arisaema triphyllum) plants. We found no significant difference in the mean percentage of plants detected between survey types (27% structured, 31% unstructured, 95% CI = −0.23, Inf). However, unstructured surveys detected slightly more plants per unit time, while structured surveys covered a slightly larger portion of the study area. Participants located 33% of plants across all 18 surveys. The probability of finding all plants in a single survey was very low (7.59 × 10 –4 to 8.89%). To ensure a 95% probability of detection of at least one plant in our study, a minimum of two to four surveys would be required. These results suggest that for population estimation or presence/absence surveys, particularly for species of conservation concern with low abundance or invasive species at the early stages of invasion, a single survey might not be enough.