Frank P. Day and Christine E. Conn

This report provides an overview of our completed work on microenvironmental patterns across the Hog Island chronosequence and their relation to subsurface decomposition rates. The results of the initial N- fertilization experiment and the effects on biomass allocation are also summarized. The work on aboveground NPP and root responses to N- fertilization observed by minirhizotrons is presented in reports by John Dilustro and Rett Weber.

A major objective of the study was to quantify patterns in nitrogen availability across the dune/swale chronosequence on Hog Island. The older sites had higher N levels in soil and soil water, especially in the swales. The older swales are more advanced successionally and are dominated by a nitrogen fixing associate, Myrica cerifera. Annual net nitrogen mineralization rates were also higher on the older sites. Generally swales had higher NH4+ levels than dunes, and dunes had higher NO2 + NO3- levels. The dune soils were aerobic, and the swale soils were generally anaerobic. The soil redox potential determines pathways of nitrogen transformation and regulates the rates. Net nitrogen mineralization rates were highest in the swales, possibly due to the dynamic nature of the hydrology of these sites. Moisture (flooding), redox conditions, and successional age of a site all impact nitrogen availability on Hog Island.

Litterbag studies on Hog Island indicate belowground decomposition of roots and nutrient immobilization dynamics are influenced by nitrogen availability, litter quality and hydrology. We hypothesized that roots buried in the most nitrogen poor dunes would immobilize nitrogen. However, roots buried in the oldest, most nitrogen rich dune decayed faster and, contrary to our hypothesis, exhibited net nitrogen immobilization. Nitrogen amendment stimulated decomposition and produced net nitrogen immobilization in the younger, more nitrogen limited dunes. In the saturated soil conditions of swales, woody tissues decayed slower than herbaceous tissues and conserved relatively more nitrogen and phosphorus. Differences in herbaceous litter quality in dunes resulted in different patterns of phosphorus release although decay rates did not differ. Herbaceous roots buried at all sites decayed slower in saturated swales than in the drier dunes. Phosphorus release was greater under saturated conditions. Measurement of tensile strength loss in cotton strips was used as an assay to remove substrate quality effects. Generally, tensile strength loss was more rapid in swales than dunes, and was faster in the summer followed by fall and winter/spring respectively. Even though environmental gradients existed across the chronosequence (nitrogen availability and water levels), there were no consistent, significant trends in decomposition rates across the chronosequence. During the winter/spring, decay rates decreased with increasing soil depth in the dunes. The inverse was observed in the swales. During the summer, these patterns reversed. In the summer, desiccation of the upper soil layer on the dunes apparently inhibits decomposition, while lower depths are more moist. At the same time in the swales, the upper soil layer is moist but not saturated like the deeper levels. Hydrology and temperature appear to be the predominant factors regulating decomposition on the Virginia coastal barrier islands. There were no clear associations between tensile strength loss and redox potential or nutrient availability.

In a one-year study, nitrogen additions resulted in biomass increases, principally in shoots, and apparently proportionally less allocated to the roots. Even though nitrogen availability increased across the chronosequence (with highest levels on the oldest dune), aboveground biomass decreased from the younger dunes to the oldest. Other factors appear to be more limiting than nitrogen over the range of nitrogen availability found across the chronosequence. Groundwater hydrographs suggest the oldest dune may be more prone to moisture stress. The dune plants sequestered more nitrogen but less phosphorus in nitrogen enriched plots. Generally there were fewer species in the fertilized plots.

Several long-term monitoring efforts and experiments are continuing along the Hog Island chronosequence. Groundwater monitoring is in its 4th year. Video recordings in the minirhizotrons are being obtained early and late in the growing season for the 2nd year following the initial year of frequent, more intensive observations. A long-term nitrogen fertilization experiment is in its 3rd year. Fertilization initially increased percent cover and density for A. breviligulata and S. patens. During the second year, continued density and percent cover increases for A. breviligulata coupled with decreases for S. patens in fertilized 24 and 36 year dune plots resulted in similar importance values between the two grasses. S. patens continued to be more important than A. breviligulata in fertilized 6 year dune plots. A. breviligulata became more important in fertilized 120 year dune plots and was probably influenced by greater pre-treatment density in fertilized plots. Community importance values for P. amarum remained constant throughout time in fertilized plots and were maintained by increases in percent cover but not in frequency or density. On the two younger dunes (6 and 24 year), P. amarum became more important in unfertilized plots due to increased frequency.