Nitrogen cycling in the upland boreal shield forest, response to an experimental addition of nitrate

Abstract

The emission of N gases by industrial and agricultural activity has increased the load of N by several-fold to many forested ecosystems during this century. Following a long-term elevated N input, the demand for N by plants and soil microorganisms may be satisfied and the terrestrial ecosystem may reach a state of "N saturation". One of the main features of N saturation is an increase in net nitrification rates in soils. Nitrification is a strong acidifying process because both acidity (2 moles of H^+ per mole of NH4^+ nitrified) and a mobile anion (NO3) replace the soil-bound NH4^+. Nitrogen saturation can cause freshwater acidification and forest decline. Presently, boreal and temperate Shield catchments efficiently retain mineral N inputs (NH4^+ and NO3^+) and buffer downstream acid-sensitive freshwaters from N-based acidification. However, the mechanisms responsible for N retention are not well understood. These retention mechanisms must be defined to evaluate the long-term potential of Shield catchments to mitigate the acidification of freshwaters.

The objectives of this study were 1) to describe the N cycle in small upland boreal Shield catchments at the Experimental Lakes Area (ELA), northwestern Ontario, and 2) to study the processes involved in N retention in this system using an experimental addition of NO3^+ to one catchment. The ELA Upland catchments are representative of sparsely vegetated rocky ridges common throughout acid-sensitive areas of the Canadian Precambrian Shield. As is typical of boreal forest, this landscape is a vegetation mosaic where thin forested soils (or "forest islands") are scattered among bedrock outcrops covered with discontinuous mats of lichens, mosses, and grasses (or "lichen patches"). In the first part of the study, the emphasis was on the comparison of the N cycle between forest islands and lichen patches. The information gathered was used to make prediction on the behavior of the landscape following an increased N input. During the second part of the study, 40 kg N ha^-1 yr^-1 as NaNO3 was added to one catchment (U3; 0.40 ha) for two years. This N input was similar to the highest level of N deposition presently observed in North America. The mechanisms of N retention and their efficiencies were evaluated using a combination of mass-balance budgets, soil N mineralization assays, the recovery of a ^15N label added with the NO3^+, plant growth, and plant nutrient content.

Under unmanipulated conditions, several aspects of the Upland catchments N cycle contradicted the traditional view for the boreal forest. As was expected for an unproductive conifer forest, mineral N inputs were efficiently retained. However, overall the catchments leaked more N than expected because of the export of dissolved organic N (DON). There was a striking contrast in internal N cycling in the different components of the landscape. Forest islands were N-limited, as indicated by an efficient mineral N retention, a low net N mineralization, the export of NH4^+ and NO3^+ in runoff was not related to precipitation inputs but to net mineralization and nitrification rates in lichen patches, demonstrating that this component of the boreal Shield ecosystem was unexpectedly close to N saturation.

The response of U3 to the NaNO3 treatment indicated a variable potential in time and space to retain the elevated N input. Nitrate retention was limited during the snowmelt period when biological retention mechanisms were less active. However, during the growing season N retention by U3 remained similar to reference catchments. Forest islands and lichen patches responded in opposite ways to the increased NO3^+ input. On bedrock surfaces, net nitrification rates doubled in lichen patches and by the second year of addition N was no longer retained. Although fast hydrological flushing and low biomass must have limited N retention on bedrock surfaces, the intrinsic N saturation of lichen patch soil microorganisms was determinant in preventing NO3 retention. In contrast, in forest islands N retention remained similar to the reference because soil microorganisms directly and indirectly contributed to N retention. Forest island soil microorganisms directly mediated NO3^+ retention by immobilizing N during the decomposition of litter with a high C:N. In addition, nitrogen retention was indirectly mediated by the tendency to convert NO3^+ inputs into NH4^+ during internal cycling. This indirect retention occurred because assimilatory NO3^+ reduction decreased the demand on the soil NH4^+ pool, allowing for NH4^+ to accumulate in the soil instead of NO3^+. Unlike NO3^+, NH4^+ can be retained in catchments by a variety of abiotic immobilization mechanisms and is less likely to be lost in runoff. Based on the recovery of the 15N label added with the NaNO3 treatment, N was retained by plant and lichen uptake in lichen patches and by both plant uptake and microbial immobilization in forest islands. Thus, in lichen patches the intrinsic N saturation of the soil microbial community left the onus of N retention on plants and lichens.

On the short-term, the upland boreal Shield landscape has a limited potential to prevent N-based acidification of downstream ecosystems because of a weak potential for N retention during a part of the year and the intrinsic N saturation of a portion of the landscape. Although forest islands were not N-saturated on the short term, the increased N load from upslope bedrock surfaces may accelerate the onset of N saturation following an increased N input. Because the different components of the boreal Shield landscape are hydrologically connected, N saturation may occur as a cascading effect in this system. The role of lichen patches at the head of the cascade had been hitherto unrecognized.