Spring ephemeral herbs and nitrogen cycling in a northern hardwood forest: an experimental test of the vernal dam hypothesis

In the late 1970s R.N. Muller and F.H. Bormann posited their ”vernal dam” hypothesis, stating that spring-ephemeral herbs in deciduous forests serve as a temporary sink for N when overstory trees are dormant, and then release this N later, in the summer, when the trees are active. This hypothesis has gained wide acceptance, yet two of its critical assumptions have never been experimentally tested: (1) that N taken up by spring ephemerals would otherwise be lost from the ecosystem, and (2) that N from senesced ephemeral tissues contributes to increased rates of summertime N mineralization. To test these assumptions, I quantified patterns of N cycling and loss from a set of paired plots, half of which served as controls and from half of which all spring-ephemeral plants were removed. There were no significant differences in NO₃ ^– leaching between plots with and without spring ephemeral vegetation. These results are consistent with the relatively low rates of N uptake by the dominant spring ephemeral, Allium tricoccum, and its apparent preference for NH₄ ⁺, which is far less mobile in soil than NO₃ ^–. In addition, based on sequential sampling, I found that soil microorganisms took up 8 times as much N during the spring than did spring-ephemeral herbs (microbial uptake=3.19 vs. plant uptake=0.41 g N m^–2), suggesting that microbial immobilization of N is the dominant sink for N during this season. Removal of spring ephemeral vegetation also had no effect on summertime rates of net N mineralization. Furthermore, the addition of spring ephemeral litter to soil+forest floor microcosms did not significantly increase rates of N mineralization in a laboratory incubation. Instead, this experiment demonstrated the overwhelming influence of forest floor litter in controlling the release of mineral N from these soils. Overall, neither assumption of the vernal dam hypothesis holds true in this ecosystem, where patterns of N cycling and loss appear to be dominated by microbial decomposition of forest floor material and soil organic matter. Received: 24 August 1999 / Accepted: 23 March 2000