From genes to landscapes: Historical patterns and long-term effects of human land-use in Massachusetts forests

New England forests have been severely impacted by human land-use. Most forests in the region regenerated on abandoned agriculture fields, while most others were logged repeatedly. Old-growth stands constitute a tiny fraction of the region's forests. Although frequently located on marginal sites, these stands offer an opportunity to study the effects of land-use on biodiversity. This dissertation explores the patterns and effects of land-use on the genetic, community, and landscape scales. A comparison of genetic diversity and structure of five secondary Quercus rubra populations that regenerated on old pastures with that of five old-growth populations suggests that founder effects temporarily decreased genetic diversity in some secondary stands. Old-growth samples had significantly greater nuclear allelic richness than secondary stands at the four simple sequence repeats (SSRs) tested, but there was no significant difference at three chloroplast SSRs. Secondary stands had a tenfold greater F_ST at chloroplast loci than old-growth stands. On the community level, surveying the understory flora of five pairs of adjacent old-growth and secondary forest stands showed that old-growth was significantly lower in plant density than secondary forest. While most species encountered were found in both forest types, some species showed an association with old-growth or secondary. On the landscape scale, comparing the physical characteristics of land that remained forested in 1830 with cleared land demonstrated that 1830 forests were found on sites that are significantly more difficult to farm. Compared to small forest patches and open land, large forest patches were at significantly higher elevations, on significantly steeper slopes, more likely to have a western aspect, and on soils ranked as significantly less suitable for pasture. Compared to open land, small forest patches were located at significantly lower elevations and were ranked as significantly less suitable for pasture. Town-level analysis showed that towns with greater mean elevation and slopes were more forested in 1830. These studies illustrate an important point: sites with different land-use histories may differ in biological characteristics both because of the direct effects of land-use and because of the physical characteristics associated with a particular land-use.