Are greater nitrate losses following ammonium additions to a forest due to higher rates of nitrate production - or to something else?

Background/Question/Methods

There is a paradox in the response of a young forest to experimental additions of ammonium sulfate at the Fernow Experimental Forest near Parsons, West Virginia.  Additions of ammonium sulfate to a small watershed (WS3) since 1989 have increased stream-water nitrate losses to levels that are 64% greater than those measured in a nearby, but unfertilized, watershed (WS7).  This difference in nitrate losses, however, is maintained without any statistically detectable difference in the rates of soil nitrate production.

Since soil nitrate production was previously measured at only seven locations per watershed, a likely explanation is that large spatial variability, typical of in situ measurements of net nitrate production, masked the slight differences needed to account for the greater levels of stream-water nitrate that are lost from the fertilized watershed. 

To test this hypothesis, during the summer of 2011 we increased our sample size and measured the potential, net nitrification rates (30-day lab incubations) in the upper 5 cm of mineral soil collected from 100 locations in each watershed.  We also measured several soil properties that are often associated with differences in net nitrification rates, including soil pH and the soil C:N ratio.

Results/Conclusions

The more extensive measurements revealed that the average rate of net nitrate production in the soil of the fertilized watershed (39.8 mg N/m²*day) was 14% LOWER (p = 0.017) than the average rate measured in the unfertilized watershed (46.2 mg N/m²*day).  The lower rate of nitrate production in the fertilized watershed was also associated with a lower soil pH (4.12 vs. 4.52; p < 0.001) and a higher soil C:N ratio (17.6 vs. 14.6; p < 0.001).

Since our results indicate that greater nitrate losses, caused by the addition ammonium sulfate, are not attributable to elevated rates of nitrate production, other explanations should be considered.  We hypothesize (and will test) that the observed differences in nitrate loss are the result of a reduced biotic demand for nitrate in the fertilized watershed.  This could occur if trees switch to a greater uptake of ammonium in the fertilized watershed, or if chronic additions of ammonium sulfate reduce tree growth, and thus the overall uptake of all forms of nitrogen.  A better understanding of how this forest responded to chronic nitrogen additions should help us interpret past and future changes in stream-water chemistry in regions that have experienced high levels of acid deposition.