Northern Cardinal is one of the most familiar birds in the Lower 48 states, occurring from Maine to Florida and west to Arizona and Colorado. It is a resident species not known to have any significant migrations, although they do perform short distance movements and may gather in flocks in winter.
We used Northern Cardinal as our very first example of these occurrence models from our STEM analyses. This story, which you can review here, focused on the changes in detectability that occur in this species throughout the year. Understanding these seasonal changes in detectability is essential to understanding these occurrence models. Changes in detectability could be mistakenly interpreted as migratory movements, as species occurrence appears to decrease in some areas and increase in others. Interpretation of the biological patterns visible on these occurrence maps must be made with caution, drawing on an understanding of known biological patterns for each species.
Our earlier analysis, which drew on a single year’s data to make predictions for the year 2008, was both a more rudimentary model and used a far smaller subset of the overall eBird database. Research since then has improved the scope of the eBird data that can be used in these models, and this iteration now draws on eBird data from 2005-2009. With the inclusion of this longer data set, the model was able to work with almost ten times more eBird data overall, which has resulted in much more accurate models. One sign of this improvement is the reduced variability in the Northern Cardinal occurrence maps. Compare the animation above for 2008 with the previous version for the same year available here.