The effect of Soil Freezing on N Cycling: Comparison of two Headwater Subcatchments With Varying Snowpack, Hokkaido, Japan

Over-wintering soil processes affect the leaching of NO3- to surface water especially during snowmelt. In Hokkaido, Japan the snowpack decreases from west (~2-3 m) to east (0.5 to 1 m) providyng a unique opportunity to test the effects of variable snowpack on soil freezing and N cycling. The Upper Nakagawa Catchment (UNC), eastern Hokkaido, had a mean annual stream water NO3- concentration of 2 mg L-1 with a mean maximum yearly snowpack of about 0.5 m. In contrast, in northwestern Hokkaido, the M3 catchment had a mean annual stream NO3- concentration and mean annual maximum snowpack of about 0.1 mg L-1 and 2.5 m, respectively. A buried bag experiment was conducted to elucidate whether the difference in stream water N between UNC and the M3 was due to differences in soil freezing between the 2 sites or litter quality. Soil samples from two landscape positions (upper and lower slope) in each catchment were mixed and buried at 0, 5, and 30 cm in five locations at each landscape position. Shibecha soil was buried in Uryu and Uryu soil was buried in Shibecha at each landscape position and was incubated between October 2004 and April 2005. Results suggested litter lignin content was different between Uryu and Shibecha with values of 0.31, 0.25, 0.02, and 0.18 g lignin g-1 Uryu lower slope, Uryu upper slope, Shibecha lower slope, and Shibecha upper slope, respectively. Under conditions without freezing (5 cm and 30 cm soil buried in Uryu), N cycling in Shibecha soil buried in Uryu was much higher than Uryu soil buried in Uryu. Any NH4+ that was produced was nitrified as evidence by the high final extractable NO3- in Shibecha soil buried in Uryu at 5 cm and 30 cm depths (~2.5 and 3.5 g N m-2 at the lower and upper slopes, respectively at both depths). There was a decoupling of nitrification and mineralization when soil freezing occurred. Sites experiencing severe soil freezing had higher net mineralization compared to sites having no or less freezing. This was most pronounced in Uryu soil buried in Shibecha and at 0 cm depth. Uryu lower slope soil buried in Shibecha lower slope at 0 cm had a net mineralization rate of 10.6 mg N m-2 d-1 compared to 5.7 mg N m-2 d-1 at Uryu soil buried in Uryu at the same depth. Since nitrifiers could not survive in severe conditions there was less subsequent nitrification and less NO3- release. For example, Uryu upper slope soil at 0 cm buried in Uryu had net nitrification of 2.7 mg N m-2 d-1 while Uryu upper slope soil at 0 cm buried in Shibecha had net nitrification of 1.7 mg N m-2 d-1. The marked differences in N cycling between Uryu and Shibecha soil were likely a function of differences in litter quality and soil freezing between the sites. This study was completed before the spring thaw. Increases in nitrification would likely occur in the spring in especially Shibecha due to an increase in labile sources of organic N and a rejuvenation of nitrifiers. This would also result in higher leachable NO3- in Shibecha. Freeze frequency should be considered when evaluating differences in N dynamics between catchments having a propensity for soil freezing.