EFFECT OF STAND BASAL AREA ON PONDEROSA PINE WOOD QUALITY: FINDINGS FROM A REPLICATED DENSITY EXPERIMENT IN ARIZONA, USA

Restoration treatments in the southwestern United States are currently generating large volumes of woody byproducts. These are typically sold into low-value markets such as pallet stock (Lucas & Kim 2016), if utilized at all. Meanwhile, contractors struggle with the high cost of operations relative to the low value of harvested material, therefore targeting higher-value products is essential. With this in mind, the aim of this study is to quantify and build models to predict the variation in important intrinsic wood quality characteristics of northern Arizona ponderosa pine (Pinus ponderosa) trees. Using destructive and non-destructive techniques, we assessed dynamic stiffness, static mechanical properties, and wood density of trees from six different stand densities. Our objectives were to 1) develop models for predicting mechanical properties of small clearwood specimens, and 2) develop models for the pith-to-bark variation in wood density. Our study site (Taylor Woods) is a replicated ‘levels-of-growing stock’ experiment near Flagstaff, AZ, in which plots have been maintained at specified basal areas (6.9, 13.8, 18.4, 23.0, 27.5, or 34.4 m 2 ha -1 ) by thinning every ten years (Bailey 2008). We selected 55 trees from 18 plots and measured their diameter at breast height, total tree height, canopy base height, and standing acoustic velocity. We then felled the trees and collected a bolt from just above breast height as well as 3-5 cm-thick cross sections at approximately 2.4-m intervals along the stem. To address Objective 1, we processed each bolt into small clear mechanical testing specimens in accordance with ASTM D143 (ASTM International 2014). Each tree yielded between two and seven specimens, depending on its diameter. We assessed modulus of elasticity (MOE) and modulus of rupture (MOR) for each specimen on a Tinius Olsen 5000 Universal Testing Machine (Tinius Olsen, Willow Grove, PA). For Objective 2, we processed tree cross-sections into 2x5-mm radial strips mounted on 2 mm-thick hardwood strips. We used a QTRS X-ray densitometer (Quintek Measurement Systems, Knoxville, TN) to determine the wood density of the radial strip at 25-micron intervals. We developed nonlinear mixed-effects models using the statistical program R (R Core Team 2018). Mixed-effects models were necessary due to the hierarchical structure of the data (i.e. samples nested within trees), non-constant variance, and unbalanced data (Pinheiro & Bates 2000). To satisfy Objective 1, we developed two models: one for MOE and one for MOR. Objective 2 analyses are currently being conducted and will be completed during Summer 2018. A modified Michaelis-Menten equation with Rings Per Inch, Ring Number, and basal area treatment provided the best fit to the data for both MOE and MOR. Moving from pith to bark, mechanical properties increased steeply at first before approaching an asymptote near the bark. Stands with higher levels of growing stock generally have higher mature wood values. Models fit the data well, with the MOE model performing slightly better than the one for MOR. In all trees, the same general trend was observed of rapidly increasing MOE and MOR near the pith that levelled off in the mature wood. Outside of the juvenile wood zone, wood properties compare favorably with those of other species that are used for structural lumber. The implications of this are that foresters and logging contractors in the southwestern US may be able to sell ponderosa pine restoration byproducts in higher-value markets.