Moss and lichen decomposition in old-growth and harvested high-boreal forests estimated using the litterbag and minicontainer methods

•Forest mosses and foliose lichen are a more labile litter in forests than in wetlands.•Clearcut harvesting had no effect on decomposition rates.•Litter type was a better predictor of decomposition rates than litter chemistry.•The artefacts of the minicontainer method outweigh its statistical advantages.•In extreme climates monthly decomposition models may be preferable to annual models.

Bryophytes and lichen are important components of many boreal forest ecosystems, making the quantification of moss and lichen decomposition rates critical to understanding the C cycle of these forests. Cryptogam decomposition has been predominantly studied in wetlands, while few studies exist for forest-habitat mosses and even fewer for foliose ground lichen. We used a) the litterbag and b) the minicontainer (MC) method to quantify the decomposition rates of i) feathermoss, ii) forest peatmoss, iii) foliose ground lichen, and iv) alder leaves (reference litter) in cool, wet high-boreal Labrador black spruce forests. A total of 1560 litter samples (360 litterbags, 1200 MCs) were incubated for two years in six forest stands of different disturbance origin: three old-growth stands of wildfire origin, and three recently clearcut stands. Litter samples were retrieved after 6.5, 13, 47, 57 and 109 weeks of field incubation, and analysed for mass loss, C, N, nutrient, and fibre content.While clearcut harvesting had no significant effect on litter decomposition at all, decomposition rates significantly differed between litter types, with residual mass increasing in the order alder ≤ lichen < feathermoss ≤ peatmoss. Compared to wetlands, forest moss litter was more labile in the studied forest types, with lichen producing especially fast-decomposing litter. Litter type was a better predictor of decomposition rates than individual chemical parameters, indicating that, even in extreme climates like in Labrador, substrate quality is more important in determining decomposition rates than environmental factors. For all litter types, decomposition models accounting for the seasonality of decay dynamics performed better than models assuming constant or continuously decreasing decay rates. Compared to the litter bag method, which yielded decomposition rates comparable to previous studies, the MC method overestimated decomposition rates for alder and lichen due to fragmentation artefacts. The small sample size of the MC method therefore outweighs its statistical advantages arising from ease of replication. In order to derive reliable estimates of litter decomposition, both the field incubation method and the applied decomposition model must therefore be selected to suit the studied litter types.