Witness trees recorded in surveyors' notes have been extensively used to describe presettlement forest vegetation throughout eastern North America. Relatively few studies, however, have discussed surveying inconsistencies that could affect the quality of witness-tree data in Colonial metes and bounds surveys. To increase the accuracy of these data, we developed techniques to quantify and mitigate the effects of several irregularities in metes and bounds witness-tree data using Lancaster County, southeastern Pennsylvania, USA, as a model. Irregularities included unequal sampling intensities among regions of contrasting topographic relief, and on a finer scale, unequal sampling among landform classes. We mitigated the effects of these inconsistencies by subdividing the study area into physiographic (topographic) sections, and then subdividing each physiographic section into landform classes before witness-tree analysis. Pre-European settlement forests were dominated by Quercus velutina and Q. alba, with Carya spp. in regions of low topographic relief and Castanea dentata in regions of high topographic relief. We then assessed the effects of Native Americans using catchment analysis in which we tallied witness trees within 5 and 7 km of Susquehannock village sites and compared the forest composition to edaphically similar catchment areas of low Native American activity. Elevated frequencies of Carya spp., Juglans spp., and Robinia pseudoacacia, with depressed frequencies of Quercus alba, occurred on sites with a history of Native American occupation, possibly as a result of disturbance or active cultivation. Together, bias-correction techniques and catchment analysis greatly extend the usefulness of witness-tree data by improving methods for evaluating pre- and post-European settlement disturbance on forest change.
|Original language||English (US)|
|Number of pages||13|
|State||Published - Jan 1 2001|
All Science Journal Classification (ASJC) codes
- Ecology, Evolution, Behavior and Systematics