Litter quality outweighs climate as a driver of decomposition across the tundra biome

Abstract

Considerable uncertainty exists regarding the strength, direction and relative importance of the drivers of decomposition in the tundra biome, partly due to a lack of coordinated decomposition field studies in this remote environment. Here, we analysed 3717 incubations of two uniform litter types, green and rooibos tea, buried at 330 circum-Arctic and alpine sites to quantify the effects of temperature, moisture and litter quality on decomposition. We found a surprisingly linear positive relationship between decomposition and soil temperature across all sites, counter to theory and previous model estimates. Litter mass loss was greater at wetter sites, even where soils reached almost full water saturation. However, litter quality was the strongest driver of litter mass loss across the tundra biome, explaining six times more variation in summer decomposition than soil temperature. Our results indicate that climate warming will directly increase decomposition across tundra environments. However, the indirect effects of climate change on vegetation communities, and thus plant litter inputs and quality, could have a more profound impact than direct effects on the balance of this globally important carbon store.

Microclimate explains little variation in year-round decomposition across an Arctic tundra landscape

“…To measure decomposition dynamics, we used commercially available Green Tea and Rooibos Tea (EAN 8714100770542 and 8711327514348; Lipton, Unilever Inc.), on the basis of the TBI protocol (Keuskamp et al 2013). The TBI has been used previously in numerous locations across the Arctic tundra (Sarneel et al 2020, Gallois et al 2022, Björnsdóttir et al 2022, Thomas et al 2023. The method should therefore be well suited to investigate the effects of multiple landscape-scale environmental gradients on litter decomposition.…”

“…To put decomposition rates from our study area into context, we obtained data on mass loss, k and S from the TundraTea initiative, which encompasses coordinated studies applying the TBI protocol across Arctic and alpine tundra regions (Thomas et al 2023). We only included data from year-round incubations in untreated plots comparable to our study, and from regions with at least three unique plots to ensure variation in environmental conditions.…”

Reasons to not correct for leaching in TBI; Reply to Lind et al. (2022)

“…Pre-leaching may often extract more material from the tea than the environmental settings would. For instance, Lind et al (2022) and Thomas et al (2023) measured 7%-12% of green tea mass to be leached in outdoor soils, whereas >35% of the initial tea mass was leached in indoor treatments that resembled pre-leaching treatments (Figure 1). Therefore, a soaking pre-treatment could make the tea less comparable to local litter (Boulton & Boon, 1991).…”

“…Data obtained from literature sources (Blume‐Werry et al, 2021 ; Djukic et al, 2018 ; Friesen et al, 2018 ; Harmon, 2016 ; Jiang et al, 2016 ; Keuskamp et al, 2013 ; Lind et al, 2022 ; Mori et al, 2021 ; Pouyat et al, 2017 ; Seelen et al, 2019 ; Thomas et al ( 2023 and data file nr. 179 in the TBI database); Xiong & Nilsson, 1997 ).…”

“… (a) variation in leaching estimates of rooibos and green tea in literature sorted from short to long incubation durations on a square root transformed scale to enhance visibility. Studies included Djukic et al ( 2018 ) leaching 3 min at 100°C, Pouyat et al ( 2017 ) leaching 80 min at 60°C (only green tea), Seelen et al ( 2019 ), leaching 3 h under outdoor conditions (9.5–14°C), Lind et al ( 2022 ) leaching 3 h outdoor measurements, and at 8, 19 and 60°C in the laboratory, Blume‐Werry et al ( 2021 ), leaching 12 h at 25°C, Mori et al ( 2021 ) leaching 24 h at 3, 15 and 25°C and Thomas et al, ( 2023 ) leaching 24 and 48 h under outdoor conditions and 24 h at room temperature (±20°C). Shaded areas represent the ranges of leaching of local litter reported in (A) Jiang et al ( 2016 ), (B) Friesen et al ( 2018 ), and (C) Xiong and Nilsson ( 1997 ).…”