Vegetation control effects on untreated wood, crude cellulose and holocellulose δ13C of early and latewood in 3- to 5-year-old rings of Douglas-fir

Abstract

The stable carbon (C) composition of tree rings expressed as δ13C, is a measure of intrinsic water-use efficiency and can indicate the occurrence of past water shortages for tree growth. We examined δ13C in 3- to 5-year-old rings of Douglas-fir (Pseudotsuga menziesii (Mirb) Franco) trees to elucidate if decreased water supply or uptake was a critical factor in the observed growth reduction of trees competing with understory herb and shrub vegetation compared to those growing without competition. We hypothesized that there would be no differences in δ13C of earlywood in trees growing in plots with competing vegetation and those in plots receiving complete vegetation control during 5 years because earlywood formed early in the growing season when soil water was ample. We also hypothesized that δ13C in latewood which was formed during the later half of the growing season when precipitation was low, would be greater (less negative) in trees in plots without vegetation control. We then separated early and latewood from rings for three consecutive years and analyzed their δ13C composition. No significant differences in earlywood δ13C in years 3–5 were observed for trees in the two vegetation control treatments. δ13C of untreated latewood separated from wood cores was greater in 4- and 5-year-old rings of trees growing with competing vegetation compared to trees growing without vegetation competition (i.e., −25.5 vs. −26.3‰ for year 4, and −26.1 vs. −26.8‰ for year 5). Results suggest that water shortages occurred in Douglas-fir trees on this coastal Washington site in the latewood-forming portion of the growing season of years 4 and 5 in the no-vegetation control treatment. We also compared δ13C from untreated wood, crude cellulose extracted with the Diglyme–HCl method, and holocellulose extracted with toluene–ethanol to see if the extraction method would increase the sensitivity of the analysis. δ13C values from the two extraction methods were highly correlated with those from untreated samples (r2 = 0.97, 0.98, respectively). Therefore, using untreated wood would be as effective as using crude cellulose or holocellulose to investigate δ13C patterns in young Douglas-fir.