Peat extraction alters the hydrophysical properties of the peat during the 15 to 40 plus years of extraction. Despite the importance of hydrologic conditions for driving carbon (C) emissions from this net C source, few studies in Canada have quantified the impact of peat extraction on energy partitioning and evaporation (E) rates. The removal of vegetation prior to extraction alters the controls and mechanisms of water loss, and drainage induced subsidence is expected to enhance vertical capillary connectivity through the peat profile. We thus conducted a multi-year study using eddy covariance to understand the energy balance and daytime E rates from actively extracted sites in Quebec and Alberta, Canada. Despite being a partially drained system, available energy was largely partitioned into latent heat. The relative importance of surface and atmospheric controls of E varied with hydrologic conditions; with greater water table depth (WTD), the relative importance of vapour pressure deficit decreased, and the relative importance of WTD increased. Our results highlight a need for continuous surface moisture measurements for accurate E prediction. During active extraction, site managers harrow (till) the top few cm of peat to create a dry, hydrologically isolated layer that is cheaper to extract and transport. A weighing bucket lysimeter experiment found that while harrowing initially elevated E rates, by ~ 4 hours post harrowing, the newly dry layer acted as a barrier to further water loss from the peat profile. These sites provide a unique opportunity to further our understanding of water availability and transport of water to the evaporating surface from bare peat, and will inform future modelling efforts to partition evapotranspiration from peatlands. An understanding of the impact of site management on E rates informs site water balance calculations and can aid in optimizing harvesting practices and effective restoration strategies post-extraction.

Peatlands are globally important long-term sinks of carbon, however there is concern that enhanced moss moisture stress due to climate change mediated drought will reduce moss productivity making these ecosystems vulnerable to carbon loss and associated long-term degradation. Peatlands are resilient to summer drought moss stress because of negative ecohydrological feedbacks that generally maintain a wet peat surface, but where feedbacks may be contingent on peat depth. We tested this ‘survival of the deepest’ hypothesis by examining water table position, near-surface moisture content, and soil water tension in peatlands that differ in size, peat depth, and catchment area during a summer drought. All shallow sites lost their WT (i.e. the groundwater well was dry) for considerable time during the drought period. Near-surface soil water tension increased dramatically at shallow sites following water table loss, increasing ~5–7.5× greater at shallow sites compared to deep sites. During a mid-summer drought intensive field survey we found that 60%–67% of plots at shallow sites exceeded a 100 mb tension threshold used to infer moss water stress. Unlike the shallow sites, tension typically did not exceed this 100 mb threshold at the deep sites. Using species dependent water content - chlorophyll fluorescence thresholds and relations between volumetric water content and water table depth, Monte Carlo simulations suggest that moss had nearly twice the likelihood of being stressed at shallow sites (0.38 ± 0.24) compared to deep sites (0.22 ± 0.18). This study provides evidence that mosses in shallow peatland may be particularly vulnerable to warmer and drier climates in the future, but where species composition may play an important role. We argue that a critical ‘threshold’ peat depth specific for different hydrogeological and hydroclimatic regions can be used to assess what peatlands are especially vulnerable to climate change mediated drought.