Introduction: Mountain peatlands are understudied globally, especially in tropical regions such as the Andes. Their high abundance across the landscape and thick carbon (C) rich soils establish them as nationally important C reservoirs. However, they are at high risk of degradation due to unsustainable land use and climate change. Mitigation of these threats requires detailed inventories of C stocks present and improved understanding of the major drivers of long-term C accumulation in these ecosystems.

Objectives: 1) What is the belowground C storage per unit area for Andean peatlands? 2) Are there differences in C storage between peatlands in the northern and southern climatic ecozones (páramo and puna) of the tropical Andes? 3) Which environmental variables (elevation, temperature, precipitation, solar radiation) best predict longer-term C accumulation rates in tropical Andean peatlands?

Methods: We cored 24 peatlands located between 3,000 and 4,800 m elevation across Colombia, Ecuador, Peru, and Bolivia, calculated C storage and long-term- and recent apparent rate of C accumulation (LARCA & RARCA, respectively), and tested their relationships to environmental variables (elevation, temperature, precipitation, and solar radiation).

Results: The peatlands had a mean thickness of 4.7 m (range, 0.7 m‒11.25 m). The mean age of peatland was 7,918 yrs B.P., with a range from 490‒20,000 yrs B.P. The mean C stock was 1743 Mg ha-1 and did not significantly vary by climatic region or basal age but did increase with elevation. LARCA was best predicted by age and elevation, while RARCA was negatively related to mean annual temperature.

Conclusions: These findings indicate that peatlands in the tropical Andes store thick deposits of soil C that are likely influenced by temperature, making them vulnerable to changes in climate. To inform climate policy, there is a need for science that will determine the potential for adaptation and mitigation treatments to increase the resilience of these C rich ecosystems to climate change.