Evaluation of an automated recording device for monitoring forest birds

Monitoring of forest songbirds via auditory detections during point surveys can be enhanced by using preprogrammed recording devices. During May–July 2008, we compared boreal forest bird surveys conducted with SM-1 bird song recorders (Wildlife Acoustics, Inc.) with field surveys by observers and surveys recorded with the E3A Bio-Acoustic Monitor Kit (River Forks Research Corp.) in Ontario, Canada, to evaluate the utility of the SM-1 to generate reliable detections of forest birds. The SM-1 surveys identified, on average, 8.95 species, 0.76 fewer species per 10-min point count than field surveysx̄= 9.71 species) and 1.26 fewer species than the E3Ax̄= 10.21 species). SM-1 surveys also identified on average 11.6 individuals per 10-min count, 2.5 fewer than field surveysx̄= 14.1) and 2.3 fewer than E3A surveysx̄= 13.9), respectively. The lower number of SM-1 detections, however, was less than the reduction in detections made by field surveys later as compared to earlier in the breeding season. This suggests that SM-1 recorders set up early in the season would detect more birds than field surveys stretching late into the season. Moreover, lower detections with the SM-1 could be easily offset by collecting an additional 10-min sample on another day. Most species were detected equally well by all 3 methods with a few exceptions. Unattended recording devices are especially advantageous in situations where the number of experienced observers is limited, where access difficult, where multiple samples at the same site are desirable, and where it is desirable to eliminate inter-observer, time-of-day and time-of-season effects.