Restoration is moving towards a more mechanistic approach that emphasizes restoration of ecosystem services. Trait-based approaches provide links between species identity and ecosystem functions and have been suggested as a promising way to formally integrate ecosystem services in the design of restoration programs. While practitioners have been routinely using informal knowledge on plant traits in their practices, these approaches are underutilized as operationalization remains challenging. The goal of this paper is to provide guidance for applied scientists and restoration practitioners looking to apply a trait-based approach to restore forest ecosystems. We present a five-step framework: (1) selection of services to be restored, (2) trait selection, (3) data acquisition, (4) analytical planning, and (5) empirical testing and monitoring. We use three Canadian case studies to illustrate the applicability of our framework and the variety of ways trait-based approaches can inform restoration practices: (1) restoration of urban woodlots after an insect outbreak, (2) restoration of a smelter-damaged landscape surrounding an urban area, and (3) reclamation of remote upland forests after oil- and gas-related disturbances. We describe the major mechanisms and traits that determine vegetation effects on ecosystem services of importance in each case study. We then discuss data availability, methodological constraints, comparability issues, analytical methods, and the importance of empirical testing and monitoring to ensure realistic prediction of service restoration. By outlining issues and offering practical information, we aim to contribute to a more robust use of traits in ecological restoration.

Boreal peatlands are critical ecosystems globally because they house 30%-40% of terrestrial carbon (C), much of which is stored in permafrost soil vulnerable to climate warming-induced thaw. Permafrost thaw leads to thickening of the active (seasonally thawed) layer and alters nutrient and light availability. These physical changes may influence community-level plant functional traits through intraspecific trait variation and/or species turnover. As permafrost thaw is expected to cause an efflux of carbon dioxide (CO2) and methane (CH4) from the soil to the atmosphere, it is important to understand thaw-induced changes in plant community productivity to evaluate whether these changes may offset some of the anticipated increases in C emissions. To this end, we collected vascular plant community composition and foliar functional trait data along gradients in aboveground tree biomass and active layer thickness (ALT) in a rapidly thawing boreal peatland, with the expectation that changes in above- and belowground conditions are indicative of altered resource availability. We aimed to determine whether community-level traits vary across these gradients, and whether these changes are dominated by intraspecific trait variation, species turnover, or both. Our results highlight that variability in community-level traits was largely attributable to species turnover and that both community composition and traits were predominantly driven by ALT. Specifically, thicker active layers associated with permafrost-free peatlands (i.e., bogs and fens) shifted community composition from slower-growing evergreen shrubs to faster-growing graminoids and forbs with a corresponding shift toward more productive trait values. The results from this rapidly thawing peatland suggest that continued warming-induced permafrost thaw and thermokarst development alter plant community composition and community-level traits and thus ecosystem productivity. Increased productivity may help to mitigate anticipated CO2 efflux from thawing permafrost, at least in the short term, though this response may be swamped by increase CH4 release.