Citizen science programs that record wildlife observations on and along roads can help reduce the underreporting of wildlife-vehicle collisions and identify and prioritize road sections where mitigation measures may be required. It is important to evaluate potential biases in opportunistic citizen science data. We investigated whether the opportunistic observations of live animals by volunteers along a 46-km section of Highway 3 in the Crowsnest Pass area ("Road Watch in the Pass" data collection program) in Alberta, Canada, had a similar spatial pattern as systematically collected data by the researchers along the same road section. A permutation modeling process that compared the number of observations between the two datasets for each 1-km segment, a randomization method that tested for and compared hotspot observation locations, and a bivariate Ripley's L1.2-function analysis along a continuum of spatial scales all showed spatial agreement between the two datasets. There was spatial agreement at a scale between 1 and 4km, and three clear hotspots of wildlife observation activity were identified for both processes. This suggests that the data collected by the volunteers are reliable and robust enough to be used to help identify road sections that may require mitigation measures. In addition, volunteers proved to be able to collect a sufficient number of observations relatively quickly. Within one year, 24 volunteers collected 640 wildlife observations, and we found that using only 150 or more of these observations always resulted in spatial similarity with the systematic observations collected by the researchers. We conclude with recommendations for other citizen science data collection programs and for further research.
- Citizen science