Data for Nature manuscript titled “Environmental drivers of increased ecosystem respiration in a warming tundra”Corresponding author Dr. Sybryn Maes – [email protected] Github contains all R scripts on https://github.com/mjalava/tundrafluxPart A. Meta-analysisThe bold names refer to scripts (see the Github repository https://github.com/mjalava/tundraflux) and names in italics refer to files in this repositorydf_0-Study design Figure 1 and Extended Fig. 1 from main textdf_1a-Effect size calculations of response (ER)-Links to df_1.csv file with raw flux and environmental data-Only the experiments that state ‘Open Access’ in the excel file Authors_Datasets (sheet 2). For experiments stating ‘Available Upon Request’, you need to contact the authors for the -raw flux data.df_1b-Effect size calculations of environmental drivers-Links to df_1.csv file with raw flux data data (see above) and Dataset_ID.csv (this file includes all dataset IDs to merge the drivers into one dataframe)df_2a-f-Meta-analysis (2a) and meta-regression models (2b-f) (ER, N=136)-Links to df_2.csv file with effect size data and context-dependencies and Forestplot_horiz_weights_fig.csv (this file includes the mean pooled Hedges SMD as well as the individual dataset Hedges SMD to plot figure 2)-Contains code for Figs. 2-4 and Extended Figs 2-3df_3-Meta-regression for experimental warming duration-Contains code for Fig. 5df_4a            -Effect size calculations of autotrophic-heterotrophic respiration partitioning (Ra, Rh, N=9)-Links to df_3.csv file with raw partitioning data of subset experiments (output file df_4.csv)df_4b            -Sub-meta-analysis models (ER, Ra, Rh)-Links to df_4.csv (input file)NOTES·       All additional input files for the meta-analysis R-scripts are included within the folders. ·       ER, Ra, Rh = ecosystem, autotrophic, and heterotrophic respiration·       N = sample size (number of datasets) Part B. Upscaling resultsFor upscaling, the input data is described in the code files (see the Github repository) and the accompanying Readme.txt.percentageChangeResp_tundraAlpine.tif: modelled change in respirationbaseResp_tundraAlpine.tif: baseline respiration (calculated from the data from literature)modResp_tundraAlpine.tif: modelled respiration after warming (our calculations: (percentageChangeResp_tundraAlpine+1) * baseResp_tundraAlpine)changeResp_tundraAlpine.tif: modResp-baseRespstandError_tundraAlpine.tif: standard error of modelled respiration (standError_tundraAlpine_onlyDataUncertainty.tif: standard error of modelled respiration where only data uncertainty is taken into account

Data for Nature manuscript titled “Environmental drivers of increased ecosystem respiration in a warming tundra”Corresponding author Dr. Sybryn Maes – [email protected] Github contains all R scripts on https://github.com/mjalava/tundrafluxPart A. Meta-analysisThe bold names refer to scripts (see the Github repository https://github.com/mjalava/tundraflux) and names in italics refer to files in this repositorydf_0-Study design Figure 1 and Extended Fig. 1 from main textdf_1a-Effect size calculations of response (ER)-Links to df_1.csv file with raw flux and environmental data-Only the experiments that state ‘Open Access’ in the excel file Authors_Datasets (sheet 2). For experiments stating ‘Available Upon Request’, you need to contact the authors for the -raw flux data.df_1b-Effect size calculations of environmental drivers-Links to df_1.csv file with raw flux data data (see above) and Dataset_ID.csv (this file includes all dataset IDs to merge the drivers into one dataframe)df_2a-f-Meta-analysis (2a) and meta-regression models (2b-f) (ER, N=136)-Links to df_2.csv file with effect size data and context-dependencies and Forestplot_horiz_weights_fig.csv (this file includes the mean pooled Hedges SMD as well as the individual dataset Hedges SMD to plot figure 2)-Contains code for Figs. 2-4 and Extended Figs 2-3df_3-Meta-regression for experimental warming duration-Contains code for Fig. 5df_4a            -Effect size calculations of autotrophic-heterotrophic respiration partitioning (Ra, Rh, N=9)-Links to df_3.csv file with raw partitioning data of subset experiments (output file df_4.csv)df_4b            -Sub-meta-analysis models (ER, Ra, Rh)-Links to df_4.csv (input file)NOTES·       All additional input files for the meta-analysis R-scripts are included within the folders. ·       ER, Ra, Rh = ecosystem, autotrophic, and heterotrophic respiration·       N = sample size (number of datasets) Part B. Upscaling resultsFor upscaling, the input data is described in the code files (see the Github repository) and the accompanying Readme.txt.percentageChangeResp_tundraAlpine.tif: modelled change in respirationbaseResp_tundraAlpine.tif: baseline respiration (calculated from the data from literature)modResp_tundraAlpine.tif: modelled respiration after warming (our calculations: (percentageChangeResp_tundraAlpine+1) * baseResp_tundraAlpine)changeResp_tundraAlpine.tif: modResp-baseRespstandError_tundraAlpine.tif: standard error of modelled respiration (standError_tundraAlpine_onlyDataUncertainty.tif: standard error of modelled respiration where only data uncertainty is taken into account

Climate warming is modifying the movement of air masses over Northern latitudes, producing warming and cooling events across the boreal regions. These new conditions changes may mismatch plant phenology from weather conditions, and affect the growing period of trees. Understanding the processes of local adaptation in bud phenology can help to predict the response of plants to these rapid and unexpected environmental changes. Our study monitored bud burst and bud set weekly during four growing seasons in black spruce [Picea mariana (Mill.) B.S.P.] saplings planted in a common garden and originating from five provenances representing the whole latitudinal distribution of the closed boreal forest in Quebec, Canada. We compared the variance in bud phenology among populations and years, and analyzed the relationships with the temperatures at the origin sites. Bud burst and bud set occurred in mid-May and mid-July, respectively, with a large variability among provenances and between the study years. A delayed bud phenology was observed in the provenances from warmer sites, with bud burst and bud set being 1.1 and 1.4 days later for every additional degree in mean annual temperature at the origin site, respectively. Populations with earlier bud bursts also showed earlier bud sets, thus the growing season was similar among provenances. The heritability of bud set was higher than that of bud burst, with estimates of 0.26 and 0.21, respectively. On average, variance in bud phenology among provenances reached 5.3%, which was higher than that within provenances (2.6%). The factor year explained 37.7-69.7% of the variance in bud phenology. Synthesis. Results demonstrate the plastic response of bud burst to changing temperatures and suggest the effects of endogenous factors on bud set. The earlier growth reactivation due to global warming occurring under higher frost risks in spring are expected to produce damage to the developing buds. Meanwhile, the ability of bud phenology to match the inter-annual variability in weather could help to cope with the changing environmental conditions expected in the future.

The Arctic is warming at twice the global average speed, and the warming-induced increases in biogenic volatile organic compounds (BVOCs) emissions from Arctic plants are expected to be drastic. The current global models' estimations of minimal BVOC emissions from the Arctic are based on very few observations and have been challenged increasingly by field data. This study applied a dynamic ecosystem model, LPJ-GUESS, as a platform to investigate short-term and long-term BVOC emission responses to Arctic climate warming. Field observations in a subarctic tundra heath with long-term (13-year) warming treatments were extensively used for parameterizing and evaluating BVOC-related processes (photosynthesis, emission responses to temperature and vegetation composition). We propose an adjusted temperature (T) response curve for Arctic plants with much stronger T sensitivity than the commonly used algorithms for large-scale modelling. The simulated emission responses to 2 °C warming between the adjusted and original T response curves were evaluated against the observed warming responses (WRs) at short-term scales. Moreover, the model responses to warming by 4 and 8 °C were also investigated as a sensitivity test. The model showed reasonable agreement to the observed vegetation CO2 fluxes in the main growing season as well as day-to-day variability of isoprene and monoterpene emissions. The observed relatively high WRs were better captured by the adjusted T response curve than by the common one. During 1999?2012, the modelled annual mean isoprene and monoterpene emissions were 20 and 8 mg C/m**2/yr with an increase by 55 and 57 % for 2 °C summertime warming, respectively. Warming by 4 and 8 °C for the same period further elevated isoprene emission for all years, but the impacts on monoterpene emissions levelled off during the last few years.