Robert F. Brzuszek

Recalling open old growth forests in the Southeastern Mixed Forest province of the United States

ABSTRACT Historical forests in the Southeastern Mixed Forest province of the United States have been less researched than other regions using historical tree surveys. We used 81,000 tree records from surveys during the 1800s to quantify composition of this ecological province. Upland oaks and pines comprised about 75% of all trees, with relatively equal composition. Oak composition may have comprised ≥ 45% to the northern and eastern sides of the province. Hickories were about 10% of composition and a few species were present at 1% to 2% composition. Currently, pine has increased to 49% composition; loblolly pine was 46% of all trees. Upland oaks decreased to 8% composition. Paralleling other historically oak- or pinedominated regions, fire-intolerant species increased to 40% of composition, particularly early-successional sweetgum. Historical oak-pine forests mostly have converted to loblolly pine plantations and broadleaf forests in this region. A large extent of the eastern United States historically was dominated by oak or pine forests, which likely were open old growth forests due to a frequent, low-to-moderate severity fire regime that reduced tree densities and infrequently disturbed overstory trees. Open old growth forests should be recognized as distinct ecosystems with unique characteristics, ecological functioning, and associated management practices.

Historical Forest Patterns and the Analysis of Site-Scale Forest Gradients for Natural Area Mangement

ABSTRACT: An investigation of General Land Office (GLO) records and township-scale environmental gradients were analyzed using multivariate statistical analysis to determine the usefulness of this methodology for determining historical forest patterns, correlating the historic data with modern environmental gradients, and as background information for the development of forest restoration and management plans. The study area encompassed one township within the Noxubee National Wildlife Refuge, Tombigbee National Forest, and John Starr Forest in east-central Mississippi. From the 1830s, GLO survey notes detailing witness tree locations and understory descriptions were transcribed into a Geographic Information System (GIS). Environmental variables describing soils, slope, roughness, elevation, and distance to streams were generated from contemporary data sets for each of the witness tree locations. Redundancy analysis was used to correlate witness tree species with environmental gradients. Dominant tree species were post oak (Quercus stellata), pine (Pinus spp.), willow oak (Quercus phellos L.), red oak (Quercus falcata), and sweet gum (Liquidambar styraciflua). Multivariate analysis revealed two significant non-random environmental patterns in relation to species composition. The lowland silty/ sandy loam soil association and witness tree distance to perennial stream were significant non-random associations described by the ordination. Three dominant associations are described: (1) post oak/pine woodland with open understory on sandy-loam uplands, (2) pine/white oak forest with thick understory on transitional slopes, and (3) willow oak/red oak/sweet gum forest on silty/sandy loam lowland terraces. The methods described have value to land managers in the development of forest management plans that address the establishment and maintenance of historical plant community diversity. At the site scale, plant association details can be delineated from GLO records and modern environmental data.