Soil C:N ratio is the main driver of soil δ15N in cold and N-limited eastern Canadian forests

Abstract

Significant relationships have been observed between soil N isotopic natural abundance (δ15N) and both climate and soil characteristics across a large range of ecosystems over the globe, suggesting strong and consistent effects of these variables on N cycling. However, whether the strength and the nature of these relationships vary at regional scales and with soil depth is less documented, especially in northern cold and N-limited forest ecosystems. In this study, we analyzed δ15N in soil horizons at 21 forest sites in eastern Quebec along a gradient of concomitant decreasing N deposition and temperature (MAAT) and increasing precipitation (MAP). We hypothesized that both soil δ15N and the magnitude of increase in soil δ15N would decrease along this gradient, in accordance with relationships reported at a global scale. The data show an increase in δ15N with soil depth, although it remained constant or sharply decreased between the B- and the C-horizon at most sites. The natural abundance of 15N in the forest floor (FF), in the B-horizon and in the C-horizon averaged 2.2 ± 0.9‰, 6.2 ± 1.2‰ and 4.9 ± 2.0‰, respectively while total soil profile δ15N ranged from 3.8‰ to 7.4‰. Contrary to our hypothesis, soil δ15N was poorly correlated with climate, vegetation and most soil metrics. As a consequence, there was no spatial gradient in soil δ15N values and in the magnitude of increase in δ15N with soil depth across the study area. Soil C:N ratio was the only variable significantly correlated with soil δ15N. Multivariate models including the C:N ratio explained 47%, 60% and 36% of the inter-sites δ15N variation in B-horizon, C-horizon and total soil, respectively. In contrast with global scale studies, which have reported higher soil δ15N at sites with low soil C:N ratio, the relationship between these two variables was positive across the study area. The possible influence of ecto-myccorhizal association on this pattern is discussed. Overall, our data show that soil δ15N is controlled by complex mechanisms influenced by several variables with potential antagonist effects. Climate and most soil metrics appear to have no direct influence in the cold and N-limited forest ecosystems studied here and soil C:N ratio can affect soil δ15N in an opposite manner to what has been commonly observed.