The utility of tree ring δ15N as an indicator of past nitrogen cycle dynamics

Background/Question/Methods

Due to the rarity of long-term records of nitrogen cycle dynamics, a growing number of recent studies use δ¹⁵N signatures in tree rings to make inferences about the past nitrogen cycle. Tree ring δ¹⁵N may be linked to nitrogen availability in terrestrial ecosystems due to the preference of nitrogen cycle biochemical processes for lighter nitrogen isotopes. However, the relationship between soil nitrogen processes and plant δ¹⁵N is not yet well understood. Long-term nitrogen cycle data in small watershed ecosystem studies provide a unique opportunity for validation of the relationship between tree ring δ¹⁵N and nitrogen availability. For this study, increment cores were taken from Quercus rubra L. and Liriodendron tulipifera L. trees in watersheds 4, 10, and 13 of the Fernow Experimental Forest, Parsons, WV. These watersheds have different stream nitrate concentrations, and watershed 4 has a 42-year stream nitrate record. Three time segments (1974-75, 1981-82, 1998-99) were selected to correspond with distinct phases of nitrate export in stream water from watershed 4. Cores were cross-dated and rings from these segments were analyzed for δ¹⁵N. In addition, δ¹⁵N was analyzed on a 5-year basis, and annually in the 1970s, in cores from 3 L. tulipifera trees in watershed 4.

Results/Conclusions

L. tulipifera tree ring δ¹⁵N in watershed 4, the watershed with the longest stream water nitrate record, was strongly associated with the temporal trend of increasing stream water nitrate (R² = 0.861, p < 0.001).  However, δ¹⁵N in the rings of Q. rubra did not reflect the temporal stream water nitrate trend; nor did it reflect a significant difference that was previously found in foliar δ¹⁵N of Q. rubra for the 1997-98 time period. L. tulipifera tree ring δ¹⁵N signals differed between watersheds and were consistent with known stream nitrate levels for watersheds 10 and 4, but did not reflect known differences between watersheds 4 and 13.  Thus, the utility of tree-ring δ¹⁵N as a measure of temporal nitrogen cycle dynamics may be species dependent, as the method was effective for L. tulipifera but not Q. rubra. Factors that may influence the utility of this method include nitrogen movement between rings, preferences in the form of nitrogen used, differing mycorrhizal relationships, and topographic locations that may promote denitrification. While there is potential for the use of tree-ring δ¹⁵N as a measure of nitrogen availability, a deeper understanding of the impact of these factors is needed.