Roe deer Capreolus capreolus behaviour affects density estimates from distance sampling surveys

1. Roe deer Capreolus capreolus were surveyed at night along tracks and roads in plantation forests in North Yorkshire, UK. Distance sampling was used to estimate their density. This method has been proposed for monitoring deer populations in state-owned forests throughout the UK. 2. Most deer were stationary on detection and the vast majority did not change their behaviour during observation. Few deer were observed on the transects, suggesting that they were avoiding tracks and roads at night as opposed to moving in response to the observer. 3. This has implications for data analysis and hence the results of surveys. Left-truncation or grouping of data close to the transect line to cope with road avoidance decreased the precision of the population estimate and may have impacted upon its accuracy. 4. Use of roads and tracks in forests is the only realistic option for transect surveys of deer at night but the influence of these features on deer distribution and hence density estimates must be taken into consideration when conducting such surveys.