Planting forest on cutover peatlands may be regarded as a viable restoration technique in western Canada, where natural bogs are treed with a high density of Black Spruce, Picea mariana. Fertilizer is needed to promote P. mariana establishment on cutover peatlands; however, it also encourages spontaneous colonisation by non-peatland species such as Paper Birch, Betula papyrifera. This study aimed to assess the most appropriate fertilizer dose for P. mariana establishment and growth against the trade-off of birch invasion; consequently, we monitored the effect of B. papyrifera on P. mariana growth. Four levels of fertilizer dose were applied below-ground, but flooding of the site following planting allowed fertilizer to reach the surface and favoured the colonisation of B. papyrifera. Seven years after planting, fertilizer promoted P. mariana survival and the highest fertilizer dose improved both P. mariana and B. papyrifera growth, while the lowest fertilizer dose promoted spruce growth, to a lesser degree, without promoting birch growth as much as higher doses of fertilizer. Birch removal had a significant positive effect on the growth of P. mariana, possibly by allowing greater light penetration and higher near-surface soil moisture. Avoiding B. papyrifera colonisation on site is more effective than cutting due to the ability of birch to regenerate rapidly from stumps. In practice, if planting coniferous trees is the chosen restoration option, the risk of birch colonisation can be minimised by leaving a thicker remnant peat deposit, burying fertilizer near the planted seedlings, and planning planting to avoid flooding during the growing season post-planting whenever possible.

Tree restoration on degraded land has been identified as an effective and affordable capture carbon strategy but it is unclear if carbon sequestration rates are comparable to rates on non-industrially impacted silvicultural forests. To this end, we resampled a jack pine (Pinus banksiana) and red pine (P. resinosa) chronosequence 16 years after the initial measurement to quantify carbon pools following ca. 40 years of regreening on an acid and metal impacted landscape. Measured carbon pools were then compared to those reported in an unpublished study to determine how carbon sequestration rates have changed over time and if repeated sampling at the stand level validates temporal trends estimated by the chronosequence. Total ecosystem carbon (TEC) within the stands ranged from 55 to 136 Mg ha⁻¹ with a regional mean of 90.2 ± 9.6 Mg ha⁻¹. On average, tree and soil organic pools (SOC) were the two largest carbon pools, representing 47% and 42% of TEC, respectively. Compared with unplanted sites, tree restoration resulted in a significant increase in the mean TEC among all sites of 54.4 ± 10.2 Mg ha⁻¹ with a corresponding sequestration rate of 1.7 ± 0.3 Mg ha⁻¹ yr⁻¹. The chronosequence approach was only able to consistently detect a change in the tree carbon pool. In contrast, repeated sampling at the stand level identified changes in carbon sequestration rates within SOC, LFH and shrub carbon pools and showed that the chronosequence tree carbon sequestration rate was underestimated by a factor of 2.3. Chronosequence studies assume study sites have similar landscape history and environmental conditions, which may not be reasonable in highly degraded landscapes where past events (e.g., pollution, erosion, restoration) influence multiple landscape characteristics (e.g., local hydrology and topography). We conclude that tree restoration on impacted landscapes can sequester carbon at a rate comparable to silvicultural plantations in a similar climatic region and that reforestation of industrially damaged landscapes could be part of an effective carbon capture strategy.

Aims: This study assesses the relative effects of hydrology and colonization by vascular plants on belowground C and N mobilization, and emission of CO2 and CH4 in an extracted bog under restoration in Alberta (Canada). Methods: A wet (high water table) and dry (low water table) area were identified at the site and plots with cottongrass (Eriophorum vaginatum) or bare peat were established in each area. Plant growth, peat and porewater dissolved C (DOC) and N (TDN), microbial biomass and the emissions of CO2 and CH4 were monitored at the plots throughout the growing season. Results: The largest concentrations of DOC were measured in dry and bare sites. Lower E2:E3 ratios suggested a higher aromaticity of the DOC at these sites that were net sources of CO2 and CH4. The concentration of TDN was greater in plots with cottongrass and high water table, supporting a more abundant microbial biomass. Cottongrass dominated plots also had larger gas emissions as compared to bare plots even though they were net C sinks due to their high photosynthetic rates. Conclusion: Maintaining a high water table is key to reducing peatland C losses. While vascular plant presence seems to prime the release of N and greenhouse gases, the inputs of C exceeded the losses and recovered the C sink function of the peatland ecosystem in the short term. Carbon inputs are maximized under high water table and plant presence.