Modeling spatial and temporal changes of ponderosa pine forests in northern Arizona since Euro-American settlement

This dissertation research was conducted on 13 historically stem-mapped, permanent plots, representing the pure ponderosa pine ( Pinus ponderosa Laws. var. scopulorum) and the ponderosa pine-Gambel oak (Quercus gambelii Nutt.) in north-central Arizona. These plots were established in 1909 and 1913, respectively, sampled until the 1950s, and resampled between 2002 and 2006. For this research, I developed structural reference conditions and examined the long-term vegetation dynamics, focusing on changes in tree size structure, tree spatial patterns and contemporary pine recruitment.;I used spatial analyses to examine the short- and long-term changes exhibited at one site, a 2.59-ha historically stem-mapped ponderosa pine plot that was initially harvested in the late 1800s, and showed that tree spatial and temporal recruitment patterns were not random. The harvest also changed spatial patterns by homogenizing tree size within forest patches. Post-harvest recruitment patterns suggest that pine seedlings initially established in natural grass openings and then proceeded to colonize stump patches created by harvesting until tree growing space was fully occupied in the 1940s.;I used six of the historically stem-mapped, permanent plots, containing detailed information on stumps, snags, and logs, to develop a reconstruction model and provide structural reference conditions for ponderosa pine forests of north-central Arizona. Furthermore, I proposed a new method for delineating groups of trees or patches, and quantified tree and patch spatial pattern for the above reconstructed reference conditions. The method I proposed standardizes between- and within-patch analyses and, when coupled with stand-level analysis of spatial patterns, provides a more complete picture of forest structure prior to Euro-American settlement.;Reference conditions for these plots in 1873-1874 included the following reconstructed stand attributes: tree densities ranged from 45 to 127 trees ha-1, quadratic mean diameter ranged from 41.5 to 51.3 cm, and basal area ranged from 9.2 to 18.0 m2 ha-1. The reconstructed diameter distributions (for live ponderosa pine trees ≥9.14 cm; diameter at breast height (1.37 m); DBH) prior to fire exclusion varied in shape, but generally displayed an irregular, uneven-aged size distribution with one or two dominant size cohorts. Presettlement patch densities ranged from 10 to 27 patches ha-1, which accounted for 62% to 75% of the stand basal area (m2 ha-1).;Lastly, I analyzed forest structural characteristics and tree recruitment patterns on a ponderosa pine-Gambel oak site that received harvests from three different systems in the early 1900s (Seed Tree, Group Selection, and Light Selection). Using Ripley's K(t) univariate and K12(t) bivariate analysis, I examined the short- and long-term consequences of harvest and livestock grazing, as well as provided insight into stand dynamics for trees ≥9.14 cm. Intensive harvests (i.e., Seed Tree) effectively converted the spatial patterns to random, while the less intensive (i.e., Group Selection and Light Selection) harvesting systems had varying effects, but generally accentuated the patchiness of pre-existing spatial patterns.;The long-term impact of timber harvesting and livestock grazing, which occurred over 90 years ago, has complex and lasting effects that are reflected in the contemporary ponderosa pine forest structure of north-central Arizona. Quantifying the influence of land-use legacies on forest structure provides insight into stand dynamics and development, which have been shown to affect wildlife habitat, understory production, and fire behavior; therefore, it should be explicitly incorporated into management and restoration activities.