Sr isotope and elemental variations in bald cypress tree rings as tracers of water composition through time in urban and rural streams

As urbanization in Central Texas rapidly expands, municipal water leaking from infrastructure poses an increasing threat to water quality and potentially enhances climate resilience in riparian zones through more consistent baseflow. Strontium isotopes (⁸⁷Sr/⁸⁶Sr) are effective tracers of sources of dissolved ions in Central Texas streams, indicating a significant fraction of municipal water contributing to baseflow in urban streams. In the extensively urbanized Waller Creek, a significant municipal water contribution to streamflow is reflected in the stream’s high ⁸⁷Sr/⁸⁶Sr values, while rural Onion Creek streamwater exhibits a relatively low ⁸⁷Sr/⁸⁶Sr signature. Tree ring chronologies from Waller and Onion Creeks reveal significant differences related to the input of municipal water, and suggest that trees from Waller Creek watershed may have recorded the changes that urbanization has induced in water quality. To track the history of infrastructure degradation and its impact on streamwater, we reconstruct a time series of water composition in two distinct watersheds using ⁸⁷Sr/⁸⁶Sr ratios in rings of riparian bald cypress trees. We identify contaminants introduced to tree cores through traditional methods of core preparation for dendrochronology and optimize new methods to avoid contamination. Wood digestion methods are also optimized to yield high reproducibility of ⁸⁷Sr/⁸⁶Sr. Bald cypress trees, dating back to the 1840s in Onion Creek and the 1940s in Waller Creek, record their environmental conditions, as reflected in the ⁸⁷Sr/⁸⁶Sr values of their rings. Results from this study demonstrate that bald cypress growth rings record ⁸⁷Sr/⁸⁶Sr values from the streamwater in which they grow. Based on the ⁸⁷Sr/⁸⁶Sr record in tree rings, we conclude that the water compositions of both Onion Creek and Waller Creek have remained largely unchanged over the lives of our sampled trees. This suggests that the extent to which urban infrastructure in Waller Creek has been degraded and has contributed to Waller Creek baseflow has been occurring since at least the 1940s. We constrain the effects of translocation to determine the precision of our dendrochemical record, concluding that translocation is not an important factor in the overall variability of the ⁸⁷Sr/⁸⁶Sr record through time.