The use of non-traditional data sets can provide an historical perspective on bird communities and populations over time. I examined an extensive record of fieldtrip checklists maintained since 1941 by UMBS Field Ornithology class in northern Michigan. Most fieldtrip sites were clearcut and burned at the turn of the century but have since experienced forest regeneration and/or succession. I used checklists 1) to describe changes in bird diversity and assemblage composition in two northern Michigan counties over a 50-year period (regional changes); and 2) using historic aerial photographs and GIS, to determine the extent to which changes in the type and amount of cover found on a site correlate with changes in bird assemblages. I determined that while regional avian richness has not changed in northern lower Michigan over time, there has been significant variation in assemblage composition: the number of ground-nesting species declined, while the number of deciduous-tree nesters increased. Ground-nesting species have been reported to be at risk throughout the northeastern United States due to habitat fragmentation; my results corroborate declines and suggest that sites studied may not be immune to these effects. Additionally, the number of ground-gleaning omnivore species increased over time while ground-gleaning insectivore numbers dropped. These results conflict with earlier findings that specialist guilds dominate later successional stages. I also found that several feeding and nesting guilds significantly correlate with the amount and type of vegetative cover available in a given location. Ground-nesting species, for example, negatively correlate with deciduous cover on sites undergoing succession. Cover/guild relationships, however, are inconsistent and somewhat contradictory across sites. Thus, patterns revealed by this study are not likely to be generalizable. These findings underscore the importance of considering habitat variables other than vegetative cover in assessing habitat quality for species of concern.