Tree-ring δ 15N as an indicator of nitrogen dynamics in stands with N2-fixing Alnus rubra.

Tree-ring δ15N may depict site-specific, long-term patterns in nitrogen (N) dynamics under N2-fixing species, but field trials with N2-fixing tree species are lacking and the relationship of temporal patterns in tree-ring δ15N to soil N dynamics is controversial. We examined whether tree-ring δ15N of N2-fixing red alder (Alnus rubra) would mirror N accretion rates and δ15N of soils, and whether the influence of alder-fixed N could be observed in the wood of a neighbouring conifer. We sampled a 27-year-old replacement series trial on southeastern Vancouver Island, with red alder and coastal Douglas-fir (Pseudotsuga menziesii) planted in five proportions (0/100, 11/89, 25/75, 50/50, and 100/0, respectively) at a uniform stem density. An escalation in forest floor N content was evident with an increasing proportion of red alder, equivalent to a difference of approximately 750 kg N ha-1 between 100% Douglas-fir vs. 100% alder. The forest floor horizon also had high δ15N values in treatments with more red alder. Red alder had a consistent quadratic fit in tree-ring δ15N over time, with a net increase of $\sim$1.5‰, on average, from initial values, followed by a plateau or slight decline. Douglas-fir tree-ring δ15N, in contrast, was largely unchanged over time (in 3 of 4 plots) but was significantly higher in the 50/50 mix. The minor differences in current leaf litter N content and δ15N between alder and Douglas-fir, coupled with declining growth in red alder, suggests the plateau or declining trend in alder tree-ring δ15N could coincide with lower N2-fixation rates, potentially by loss in alder vigour at canopy closure, or down-regulation via nitrate availability.