Towards improved stratified random studies of tundra decomposition
Although our findings were overall inconclusive with regard to quantifying controls of tundra decomposition dynamics, they point to potential avenues for improving Arctic decomposition studies.
Incubating locally sampled litter alongside standardised substrate, in addition to including community traits as explanatory variables, would be an essential step to contextualise decomposition patterns in standard litter and enable more precise conclusions on feedbacks of vegetation change on decomposition (Joly et al. 2023). In a laboratory experiment, Duddigan et al. (2020) found that both tea types generally follow patterns of natural leaf litter decomposition, and studies in specific alpine (Didion et al. 2016, von Oppen 2017) and Arctic (Thomas et al. 2023) sites have shown that the two tea types well represent mass loss in local litter. However, a comparative study is still missing for the Arctic tundra biome at large, resulting in uncertainty about the representativeness of mass loss rates inferred from standard substrate. More generally though, the informative value of standard-litter experiments can be limited if nutritional differences between local litter and standard substrate are not accounted for explicitly (Joly et al. 2023). Given the importance of tundra carbon cycling, a systematic assessment of the representativeness of standard decomposition protocols (e.g., TBI) will be essential for adequate quantification of litter decomposition in Arctic environments.
Many regions across the Arctic tundra biome are experiencing notable vegetation changes in response to warming or degrading permafrost (Heijmans et al. 2022). However, a lack of coordinated systematic investigations of tundra decomposition dynamics across tundra zones and habitat types, is impeding biome-wide predictions of tundra carbon dynamics (Bonan et al. 2013); but see Thomas et al. 2023). Large-scale replication of stratified-random sampling could fill this gap, enabling representative sampling of environmental gradients across and within sites, as well as evaluation of the relative importance of macro- vs. microclimate for tundra litter decomposition (Joly et al. 2023). Our findings underline the importance of considering variables representing the complete decomposition process, particularly decomposer organisms as well as both local and standard litter, and both fine- and more broad-scale environmental data. Ultimately, such coordinated studies could be of high value to inform predictions of the future carbon balance of tundra soils with projected vegetation and environmental changes (van Gestel et al. 2018).

Conclusion

We incubated two types of standard litter of contrasting quality (Green Tea and Rooibos Tea) over 14 months across an Arctic tundra landscape, covering several environmental gradients of topography, vegetation, soil and microclimate. Although we found significant links between topography, vegetation, soils and microclimate, the micro-environmental variation itself explained little variation in decomposition responses. This suggests that other factors not accounted for in our study, like long-term moisture availability, snow patterns, or other seasonal effects might control activity and community composition of decomposers. Nonetheless, our systematic approach could serve as guidance for coordinated studies of tundra decomposition dynamics, particularly across large extents. Further research should explicitly include local litter material, decomposer communities, and temporal variation to improve model fit and explanatory power, and to support better predictions of Arctic tundra litter decomposition and carbon cycling dynamics.