Effect of extraction pretreatment on radial variation of nitrogen concentration in tree rings

Past research in the paleoenvironmental subdiscipline of dendrochemistry has concluded that N concentration variation in tree rings cannot provide information on past conditions of environmental availability of N. The objective of this study was to test wood extraction pretreatments to remove wood extractives and sap, both of which may obscure the environmental signal of N availability in tree rings. Three increment cores were collected from each of six trees (three ponderosa pines and three Douglas-firs). Within each tree, the first core was left untreated (referred to as CONTROL), the second core was extracted for several hours in organic solvents and distilled water (referred to as EXTRACT), and the third core also was extracted but for a total time of 3 d (referred to as 3-DAY). A semimicro Kjeldahl method was used to determine total N on decadal groups of rings. Average N concentration of EXTRACT cores was significantly less than that of CONTROL, and the coefficient of variation of EXTRACT cores also was significantly less than that of CONTROL. Most CONTROL cores showed substantial temporal variation in N concentration related to heartwood and sapwood and/or recently formed rings, in contrast, most EXTRACT cores showed no substantial change in N concentration related to heartwood and sapwood and/or recently formed rings. The 3-DAY cores confirmed, but did not improve upon, results obtained with extraction using the shorter time duration. Thus, pretreating wood by extraction appears to substantially reduce the variation in N concentration of tree rings, which is a necessary first step toward interpreting ring N as an indicator of past environmental N availability.