Nitrogen immobilization in soil as a function of soil and plant traits in salt marsh and mangrove ecosystems: A test of the rapid N stabilization hypothesis

Background/Question/Methods

Human activities have accelerated the deposition of reactive nitrogen (N) to the biosphere. Excess N supply has unfavorable consequences, generating interest in identifying ecosystem structures that contribute to N retention. This interest is keen in productive mangrove forests and salt marshes that buffer coastal waters from land-derived runoff. We test the hypotheses that soil and vegetation properties of intertidal ecosystems affect the immobilization rate of reactive N in stable soil organic matter (SOM). The emerging model of terrestrial N cycling emphasizes rapid N immobilization, but we poorly understand whether N is shunted directly into stable forms, or first into labile (bioavailable) forms that are subject to remineralization and N loss. Along 10 km of the subtropical west coast of peninsular Florida, we established 9 pairs of 100-m² plots (18 plots total). Each pair comprised a mangrove forest and a Juncus salt marsh plot. At each plot replicate soil cores were injected with 99 at% ¹⁵NH₄Cl and reburied for 8 hours, 2 days, or 21 days; unlabeled reference cores were also taken. After in situ incubation, soils were extracted with potassium sulfate; residual soil was analyzed for isotope ratio and %N content. We define immobilized ¹⁵N as any ¹⁵N detected in this extracted soil, and we assume it is component in insoluble organic molecules.

Results/Conclusions

Immobilization of reactive N in SOM increased through time. Approximately 15% of added ¹⁵N was recovered in the immobilized phase 8 hours after addition. Immobilization rose asymptotically to 28% of added ¹⁵N by 21 days. Immobilization of reactive N was positively correlated with the size of the particulate organic matter fraction (combustible >53-μm material). Immobilization did not differ between mangrove and Juncus ecosystem types. In the mangrove (but not the Juncus) habitat, immobilization was greater where soil molar C:N was lower. These results indicate that in organic rich intertidal soils (SOM = 38% of soil mass), soil organic matter can retain a large fraction of deposited reactive N within mere weeks of deposition. This outcome is true in both forested and unforested ecosystem types, and soil particle size distribution appears to be an important ecosystem structural feature for the critical service of N retention. This conclusion takes on grave importance for coastal water quality and marine food security when the dramatic loss of these intertidal buffer habitats to urban development is considered.