Responses of Bradford-reactive soil protein to experimental nitrogen addition in three forest communities in northern lower Michigan

Abstract

Increased use of anthropogenically fixed N and the release of N in combustion products have led to concerns about possible long-term impacts on terrestrial ecosystems. Previous studies demonstrating the potential of atmospheric N deposition to influence forest soil carbon have focused on decomposition processes with much less known about potential impacts on mycorrhiza-derived carbon. Glomalin is a unique glycoprotein produced by arbuscular mycorrhizal (AM) fungi that has been implicated in the formation of soil aggregates and potentially a significant store of soil carbon. To determine the possible impact of experimental N deposition of such stores we examined the operationally defined glomalin-related soil protein (GRSP) levels over two growing seasons in three forest types receiving background N deposition (control) or treated with 80 kg N ha−1 year−1 as NaNO3. Three sites of each of three forest types, sugar maple-basswood (SMBW), sugar maple-red oak (SMRO), and black oak-white oak (BOWO), in northern Lower Michigan were studied during the 2001 and 2002 growing seasons. GRSP were extracted from air-dried soils with citric acid and measured by the Bradford method. Analysis of variance revealed significant differences related to forest type and sample date in easily extractable Bradford reactive (EE-BRSP) and Bradford-reactive soil protein (BRSP), but failed to detect significant effects of experimental N amendment. EE-BRSP and BRSP varied in a pattern that was consistent with an AM fungal origin; a pattern that reflected the mycorrhizal types of the dominant over and understory plants of each forest ecosystem. SMBW forests dominated by AM plants had the highest levels of protein. BOWO forests with low AM plant cover had the lowest protein levels and SMRO forests were intermediate. Both Bradford-reactive fractions and their ratio varied seasonally, generally being highest in fall samples. Significant correlations observed between BRSP fractions, phosphorous, and soil organic matter were likely related to covariation of soil properties across forest types. While not statistically significant, response patterns of BRSP to N deposition were ecosystem-specific and reflected mycorrhizal types of dominant species. Abundance of these proteins reflected previously observed changes in SOC in the two forest types examined with abundant AM hosts. Specifically, nitrate addition led to BRSP decreases in SMBW and increases in SMRO forests. Changes in BRSP accounted for a small fraction of the changes in SOC; appearing to increase as a fraction of residual SOC consistent with the idea that GRSP are recalcitrant. BRSP remained unchanged at BOWO sites despite a significant increase in SOC at these sites. Our results point to the potential of proteins as contributors to differential, mycorrhizal type-specific responses to changes in soil carbon following N amendment.