Nitrogen dynamics in conifer-dominated forests with and without hardwoods

Abstract

Nitrogen and carbon in the surface 12 cm of mineral soil, N in leaf litterfall, anaerobic N mineralization rates in the soil and forest floor, and root and N accretion to sand traps placed in surface soil layers were compared in forests with hardwoods either completely or partially removed during a conifer thinning 3 years before. An adjacent unthinned conifer–hardwood stand was also included. Conifer stocking did not differ between thinned stands with and without hardwoods. Stands without hardwoods averaged 520 kg/ha more N in mineral soil (p < 0.001), 20% more N mineralized from soil during 7-day incubations (p < 0.001), and lower soil C:N ratio (p = 0.02) than stands with hardwoods. These variables did not differ between thinned and unthinned mixed stands. Soil N did not correlate with the number of hardwoods removed. Weight of forest floor and rate of N mineralization from the forest floor did not differ between mixed and pure stands. However, stands with hardwoods returned about 10 kg•ha−1•year−1 more N in leaf litter (due to higher N concentration in conifer litter as well as the presence of high-N hardwood litter); stands without hardwoods accreted about 10 kg•ha−1•year−1 more N in sand traps. Soil N mineralization in mixed stands correlated positively with N mineralization in the forest floor but not with N accretion to sand traps, while the opposite was true in pure conifer stands. Although pretreatment variability among stands cannot be ruled out, the replicated treatments within a relatively uniform area make it appear likely that differences were related to the presence or absence of hardwoods. This was not a simple additive effect, however, but a community-level phenomenon, that is, conifers cycled N differently when mixed with hardwoods than when in pure stands.