A Method to Use Ranked Timber Observations to Perform Forest Composition Reconstructions from Land Survey Data

Abstract

The goal of this research was to develop a novel method to make use of ranked timber observation descriptions, which have been underutilized in survey reconstruction research. The new method involves assuming that ranking position reflects relative importance and thereby produces four separate measures that can be used to describe the geographic distribution, relative abundance and dominance of taxa recorded in historical surveys. We believe the method allows researchers to describe the composition of the forest from the line description data in a manner that would be impossible working only with witness tree data. The method is demonstrated using the Twenty Townships survey (central New York) as a case study, an area not previously discussed in the literature. We found that beech (Fagus grandifolia), maple (Acer spp.), basswood (Tilia americana) and hemlock (Tsuga canadensis) were the most widely distributed taxa and that these four taxa were more or less abundant in this same order. Where they were present, beech and hemlock were dominant, while maple and basswood were less so. Ash (Fraxinus spp.), birch (Betula spp.), elm (Ulmus spp.), cherry (Prunus spp.), pine (Pinus spp.) and cedar (Thuja spp.) were minor elements of the forest composition, but ash and cedar were both the primary constituents of swamps in the region, making them dominant taxa when they were present. Chestnut (Castanea dentata), oak (Quercus spp.), alder (Alnus spp.), ironwood (Ostrya virginiana), butternut (Juglans spp.) and walnut (modern day Carya spp.) were also present, but they were extremely rare.