From 2011-2016 citizen scientist observers conducted repeated point counts for montane birds through the Mountain Birdwatch 2.0 program, administered by the Vermont Center for Ecostudies (www.vtecostudies.org). Jason Hill and John Lloyd used N-mixture models, in a Bayesian statistical framework, to estimate abundance of Bicknell's Thrush (Catharus bicknelli) within their estimated U.S. range. The model that produced these estimates included, as covariates, elevation, latitude, and percent forest cover, and this model accounted for imperfect detection by observers. The model included a temporal component, and we used the model values from 2016 for the purpose of projecting this population estimate across the northeastern U.S. We cross referenced the USGS Protected Areas Database to determine what portion of the Bicknell's Thrush's range and population occurred on lands identified as conserved within the Protected Areas Database.
The variable displayed is Nhat and the values are the mean expected abundance for Bicknell's Thrush given the elevation, latitude and forest cover. For example, if you counted all of the thrushes on parcels with the same elevation, same latitude, and same forest canopy coverage and on average you had 0.3 thrushes per parcel, you'd have to put 100 of those parcels together to have habitat for 30 Bicknell's Thrushes.
http://vtecostudies.org/about-us/staff/kent-mcfarland/