Title: Changes in carbon-stock and soil properties following afforestation in SW Iceland

Abstract

Afforestation does not only establish new forests on treeless lands, but also changes many other aspects of the ecosystem, including the fauna, ground vegetation and soil properties. One of the most important ecosystem changes is the influence on the ecosystem carbon (C) stocks in different aboveground and belowground C pools. If afforestation is to be used as a method to sequester atmospheric carbon dioxide (CO2), to mitigate climate warming, it is important not only to consider changes in aboveground tree biomass, but also in the other four major ecosystem pools (dead wood, ground vegetation, litter layer and soil organic carbon (SOC)). The true CO2 mitigating potential of afforestation is the net change in all those five pools, but especially the last two pools have often been neglected in prior studies.

Here I present a study of three afforestation sites in SW Iceland named Heiðmörk, Nesjavellir and Ölfusvatn forests. Heiðmörk is the largest site, planted with different coniferous tree species since ca. 1950 and also contains large naturally regenerated (self-seeded) areas of the native downy birch (Betula pubesecens). The Nesjavellir and Ölfusvatn forests are younger, since ca. 1997, and the former has planted stands of both birch or conifers and also some naturally regenerated areas self-seeded from local birch forest remnant, while the latter only has planted stands of either birch or conifers.

To examine the influences of afforestation: i) between sites, ii) between different forest types and iii) with respect to increasing forest age, the present study compared different ecosystem properties of adjacent treeless control sites with afforested areas. The ecosystem properties included: i) ground vegetation cover, composition and biomass, ii) soil physical properties (bulk density stoniness, soil and litter depths as well as soil and litter dry mass), iii) soil chemical properties (pH, SOC and N concentration in different soil layers, C/N ratio in both soils and litter) and iv) ecosystem C stocks (soils, litter, fine roots, ground vegetation and standing trees biomass). Another aim of the present study was to test if ecosystem C-stocks could be validated using minimum number of measurement plots in individual forests.

On average across all sites, forest types and forest ages, the soils of the forest sites had 12% larger SOC stocks compared to the treeless sites in 2017. Significant differences in the SOC stocks appeared mainly in the upper top soils (0-10 cm) depth. Litter C, necromass and thickness were also found to be significantly higher in the afforested sites, while ground vegetation was significantly reduced, but these properties also differed between forest types and with age of the forest. Soil bulk density, pH and C/N ratio were found to remain similar across all sites and species. Soils under conifer tree species were not found to become acidic contrary to what was hypothesized.

On average, pure coniferous plots contained somewhat higher SOC stock (10,991 g C m-2) than birch plots (10,340 g C m-2) and the difference was even more pronounced for the litter layer, where the pure conifer stands had on average 92% larger litter C stocks than the native birch stands. Ground vegetation, on the other hand, was significantly reduced under conifers (-77%) while it remained under downy birch forests (+23%), but its C stock was far the smallest of the three and had only minor effect on the ecosystem C-balance. The annual rates of litter C accumulation were 22.0 and 4.9 g C m-2 yr-1 in the pure conifer and the native birch, respectively, and the observed average annual sequestration rates of SOC were 84.2 and 64.2 g C m-2 yr-1, respectively. The changes in ground vegetation C-stock amounted to –3.5 and 1.3 g C m-2 yr-1, respectively.

My results indicate that “general C-sequestration rate constants” should be avoided when forest owners are estimating the mitigation potential of their afforested lands. At least their sites should be classed into coniferous stands and birch stands and their C-sequestration should be estimated separately. Another finding is that the C-sequestration in the litter and topsoil layers may be of the same magnitude as the aboveground forest biomass during the first 20-40 years after afforestation. The present work provides valuable equations that enable forest owners to estimate the different C-pools in planted and self-seeded forest stands for different forest types in SW Iceland.