Reconstructing Pre-EuroAmerican Settlement Forest Composition in the Ouachita Mountain Ecoregion, Arkansas, USA
Topics:
Keywords: species distribution modeling, forest composition, presettlement land survey records, historical ecology, Arkansas (USA), Ouachita Mountain ecoregion
Abstract Type: Poster Abstract
Authors:
Willow Harper, University of Central Arkansas
William Flatley, University of Central Arkansas
Yaqian He, University of Central Arkansas
,
,
,
,
,
,
,
Abstract
Prior to EuroAmerican settlement the Ouachita Mountains were likely dominated by shortleaf pine and oak woodlands. These woodlands provide habitat for a number of Arkansas’ threatened plant, insect, bird, and reptile species. Today these woodland habitats are in decline due to the removal of wildfire that maintained their unique composition and structure. It is for these reasons that shortleaf pine-oak woodlands are a focus of conservation and restoration in Arkansas. However, little is known of the historical distribution of these fire-adapted habitats in the Ouachita Mountain region. Early 19th century General Land Office (GLO) surveys provide data on the composition and structure of vegetation during early EuroAmerican settlement in the Ouachitas. The goal of this research was to reconstruct pre-EuroAmerican forest composition and structure for the Arkansas’ Ouachita Mountain ecoregion using GLO survey data and species distribution modeling. We hypothesized that shortleaf pine and fire-adapted oak species would be the dominant pre-settlement tree species, and they would be primarily found on south-facing slopes, along ridgelines, and other dry sites conducive to frequent low-intensity wildfire events. Location and species information for line trees were extracted from GLO survey notes using optical character recognition software and mapped using R scripts. We then related the presence of each species to topographic and edaphic site conditions, including elevation, slope, aspect, mean solar radiation, topographic position index, topographic wetness index, dominant soil order, soil pH, available water storage, and distance from streams. This was used to create a species distribution model using random forest classification.
Reconstructing Pre-EuroAmerican Settlement Forest Composition in the Ouachita Mountain Ecoregion, Arkansas, USA
Category
Poster Abstract