Historical forests of the Black Hills, South Dakota, USA, determined using General Land Office surveys

Abstract

Forests in the western United States generally have increased in tree density since Euro-American settlement, particularly through increases in fire-sensitive species, such as spruces, firs, and junipers. Like most areas, the Black Hills region in western South Dakota and eastern Wyoming was logged for forest products and underwent agricultural conversion before historical forests were documented. To supplement historical reconstructions and accounts, we compared tree composition and densities (diameters ≥12.7 cm at 1.37 m above ground height) from historical General Land Office (GLO) records (years 1878 to 1915) and current Forest Inventory and Analysis (FIA) tree surveys (years 2011 to 2016) in the Black Hills Highlands of South Dakota. For composition, ponderosa pine (Pinus ponderosa P. Lawson & C. Lawson) decreased from 95% to 86% of all trees, with a consequent increase specifically of white spruce (Picea glauca (Moench) Voss) from 1.5% to 6.7% of all trees. Ponderosa pine currently is smaller in mean diameter by 7.4 cm, while white spruce is larger in mean diameter by 2.4 cm than historically. When the 35% of historical survey points without recorded trees were excluded, historical tree densities indicated an overall forested structure of savannas and open woodlands with tree densities ranging from 66 trees ha–1 to 162 trees ha–1. However, historical forests of the Black Hills incorporated dense stands. Tree densities have increased two- to more than four-fold, to 311 trees ha–1 currently. These comparisons provide another source of information, paralleling changes documented in surface fire-dependent pine and oak forests throughout the United States, of transitions in forest composition and structure since Euro-American settlement.

Surveyor and analyst biases in forest density estimation from United States Public Land Surveys

“…The most extensive assessment and correction for surveyor bias in PLS density estimation is the methodology of Hanberry, Yang, et al (2012) applied in 11 publications covering Missouri, Minnesota, Wisconsin, northern Michigan, and small areas of Mississippi, Oregon, Washington, and South Dakota (Hanberry, 2020; Hanberry & Dey, 2019; Hanberry & He, 2015; Hanberry et al, 2014, 2015, 2016, 2018; Hanberry, Kabrick, et al, 2012; Hanberry, Palik, et al, 2012; Hanberry, Justice, et al, 2020; Tatina & Hanberry, 2022). This methodology combines a rank‐based approach (essentially assuming a modeled average rank of witness tree distance) and a bias‐based approach (using quadrants occupied, Maltese Cross azimuthal bias, and empirical “line” tree composition and size).…”