Context: The management of the European rabbit is of strong interest to land managers. However, rabbits can be difficult to detect due to being nocturnal and living in dense vegetation or burrows. Consequently, rabbit spotlight counts are frequently used as an index of their density, but they are generally not expected to accurately reflect their true density. Indices place paramount importance on the precision of the estimator, which is directly driven by variability in individual estimates.

Aims: To investigate how the variability of rabbit spotlight counts changes with the number of rabbits counted, repeated counts, and environmental variables.

Key results: We identified a significant negative association between the number of rabbits counted and count variability; spotlight counts as an index of rabbit density have greater statistical power and are more likely to detect similar proportional differences in density when density is high compared to when density is low. We did not find any effects of rainfall, temperature, cloud cover, wind strength, season or additional consecutive spotlight count nights on count variability.

Conclusions: Despite our comparatively large dataset, our results contrasted those of several previous studies; this suggests that many of the environmental factors that have previously been shown to impact rabbit activity or spotlight counts likely have small effects in reality.

Implications: Appreciating and recognising that spotlight counts are less likely to accurately detect or reflect changes in population size when rabbit numbers are low is critical to their effective use.

We defined variability as the relative deviation of a single spotlight count from the mean of a set of three consecutive spotlight counts on the same population. We modelled spotlight count variability across 22 Australian sites over the period 2006-2020 using generalised linear mixed models.