As permafrost degrades, the amount of organic soil carbon (C) that thaws during the growing season will increase, but decomposition may be limited by saturated soil conditions common in high-latitude ecosystems. However, in some areas, soil drying is expected to accompany permafrost thaw as a result of increased water drainage, which may enhance C release to the atmosphere. We examined the effects of ecosystem warming, permafrost thaw, and soil moisture changes on C balance in an upland tundra ecosystem. This study was conducted at a water table drawdown experiment, established in 2011 and located within the Carbon in Permafrost Experimental Heating Research project, an ecosystem warming and permafrost thawing experiment in Alaska. Warming and drying increased cumulative growing season ecosystem respiration by~20% over 3 years of this experiment. Warming caused an almost twofold increase in decomposition of a common substrate in surface soil (0-10 cm) across all years, and drying caused a twofold increase in decomposition (0-20 cm) relative to control after 3 years of drying. Decomposition of older C increased in the dried and in the combined warmed + dried plots based on soil pore space 14 CO 2 . Although upland tundra systems have been considered CH 4 sinks, warming and ground thaw significantly increased CH 4 emission rates. Water table depth was positively correlated with monthly respiration and negatively correlated with CH 4 emission rates. These results demonstrate that warming and drying may increase loss of old permafrost C from tundra ecosystems, but the form and magnitude of C released to the atmosphere will be driven by changes in soil moisture.

Permafrost soils currently store approximately 1672 Pg of carbon (C), but as high latitudes warm, this temperature‐protected C reservoir will become vulnerable to higher rates of decomposition. In recent decades, air temperatures in the high latitudes have warmed more than any other region globally, particularly during the winter. Over the coming century, the arctic winter is also expected to experience the most warming of any region or season, yet it is notably understudied. Here we present nonsummer season (NSS) CO2 flux data from the Carbon in Permafrost Experimental Heating Research project, an ecosystem warming experiment of moist acidic tussock tundra in interior Alaska. Our goals were to quantify the relationship between environmental variables and winter CO2 production, account for subnivean photosynthesis and late fall plant C uptake in our estimate of NSS CO2 exchange, constrain NSS CO2 loss estimates using multiple methods of measuring winter CO2 flux, and quantify the effect of winter soil warming on total NSS CO2 balance. We measured CO2 flux using four methods: two chamber techniques (the snow pit method and one where a chamber is left under the snow for the entire season), eddy covariance, and soda lime adsorption, and found that NSS CO2 loss varied up to fourfold, depending on the method used. CO2 production was dependent on soil temperature and day of season but atmospheric pressure and air temperature were also important in explaining CO2 diffusion out of the soil. Warming stimulated both ecosystem respiration and productivity during the NSS and increased overall CO2 loss during this period by 14% (this effect varied by year, ranging from 7 to 24%). When combined with the summertime CO2 fluxes from the same site, our results suggest that this subarctic tundra ecosystem is shifting away from its historical function as a C sink to a C source.

Climate change is causing rapid changes to forest disturbance regimes worldwide. While the consequences of climate change for existing disturbance processes, like fires, are relatively well studied, emerging drivers of disturbance such as snow loss and subsequent mortality are much less documented. As the climate warms, a transition from winter snow to rain in high latitudes will cause significant changes in environmental conditions such as soil temperatures, historically buffered by snow cover. The Pacific coast of North America is an excellent test case, as mean winter temperatures are currently at the snow-rain threshold and have been warming for approximately 100 years post-Little Ice Age. Increased mortality in a widespread tree species in the region has been linked to warmer winters and snow loss. Here, we present the first high-resolution range map of this climate-sensitive species, Callitropsis nootkatensis (yellow-cedar), and document the magnitude and location of observed mortality across Canada and the United States. Snow cover loss related mortality spans approximately 10° latitude (half the native range of the species) and 7% of the overall species range and appears linked to this snow-rain transition across its range. Mortality is commonly >70% of basal area in affected areas, and more common where mean winter temperatures is at or above the snow-rain threshold (>0 °C mean winter temperature). Approximately 50% of areas with a currently suitable climate for the species (<-2 °C) are expected to warm beyond that threshold by the late 21st century. Regardless of climate change scenario, little of the range which is expected to remain suitable in the future (e.g., a climatic refugia) is in currently protected landscapes (<1-9%). These results are the first documentation of this type of emerging climate disturbance and highlight the difficulties of anticipating novel disturbance processes when planning for conservation and management.

Understanding plant community succession is one of the original pursuits of ecology, forming some of the earliest theoretical frameworks in the field. Much of this was built on the long-term research of William S. Cooper, who established a permanent plot network in Glacier Bay, Alaska, in 1916. This study now represents the longest-running primary succession plot network in the world. Permanent plots are useful for their ability to follow mechanistic change through time without assumptions inherent in space-for-time (chronosequence) designs. After 100-yr, these plots show surprising variety in species composition, soil characteristics (carbon, nitrogen, depth), and percent cover, attributable to variation in initial vegetation establishment first noted by Cooper in the 1916-1923 time period, partially driven by dispersal limitations. There has been almost a complete community composition replacement over the century and general species richness increase, but the effective number of species has declined significantly due to dominance of Salix species which established 100-yr prior (the only remaining species from the original cohort). Where Salix dominates, there is no establishment of "later" successional species like Picea. Plots nearer the entrance to Glacier Bay, and thus closer to potential seed sources after the most recent glaciation, have had consistently higher species richness for 100 yr. Age of plots is the best predictor of soil N content and C:N ratio, though plots still dominated by Salix had lower overall N; soil accumulation was more associated with dominant species. This highlights the importance of contingency and dispersal in community development. The 100-yr record of these plots, including species composition, spatial relationships, cover, and observed interactions between species provides a powerful view of long-term primary succession.