A COMPARISON OF ON-ROAD AND OFF-ROAD BIRD COUNTS - DO YOU NEED TO GO OFF-ROAD TO COUNT BIRDS ACCURATELY

On- and off-road point counts were established in two National Forests in northern Minnesota to determine whether breeding bird parameters d erived from two different types of counts conducted on- or off-road we re comparable. The first design compared single-year counts randomly p laced along a road (like the U.S. Fish and Wildlife Service Breeding B ird Survey) and counts conducted at least 200 m from the road that wer e placed in a specific habitat. In the second approach, differences an d similarities were examined in data gathered over 3 yr from points pl aced near roads to points in the same habitat (stand) >200 m from the road. Data from the first approach indicated that on average, two more species and four more individuals were observed on roadside counts th an on off-road counts. Twenty-four individual species were more abunda nt on road than off road. Many of these species were ones associated w ith openings or shrubs that develop along roads. Of the five species t hat were more abundant on the off-road counts, three had specific asso ciations with lowland conifer habitat, which was not as commonly sampl ed with the on-road counts. Data from the second approach indicated th at number of species, individuals, and individual species-abundance pa tterns were similar between the paired within-stand points. Greater st atistical power was achieved for data gathered with habitat specific c ounts off-road primarily because standard errors were lower for bird p arameters in this data set than in the data collected on road. It is s uggested that points can be placed on roads with the restriction that points be selected randomly and placed within distinct habitat types, and that roads selected for sampling have a closed canopy. Evidence pr ovided here also suggest that some points be placed off road in habita ts not sampled with on-road counts.