Climatic effects on litter decomposition from arctic tundra to tropical rainforest

Abstract

Climatic effects on the decomposition rates of various litter types in different environments must be known to predict how climatic changes would affect key functions of terrestrial ecosystems, such as nutrient and carbon cycling and plant growth. We developed regression models of the climatic effects on the first‐year mass loss of Scots pine needle litter in boreal and temperate forests across Europe (34 sites), and tested the applicability of these models for other litter types in different ecosystems from arctic tundra to tropical rainforest in Canada (average three year mass loss of 11 litter types at 18 sites), the USA and Central America (four litter types at 26 sites). A temperature variable (annual mean temperature, effective temperature sum or its logarithm) combined with a summer drought indicator (precipitation minus potential evapotranspiration between May and September) explained the first‐year mass loss of the Scots pine needle litter across Europe with a higher R2 value than actual evapotranspiration (0.68–0.74 vs. 0.51) and with less systematic error for any sub‐region. The model with temperature sum and the summer drought indicator appeared best suited to the other litter types and environments. It predicted the climatic effects on the decomposition rates in North and Central America with least systematic error and highest R2 values (0.72–0.80). Compared with Europe, the decomposition rate was significantly less sensitive to annual mean temperature in Canada, and to changes in actual evapotranspiration in the USA and Central America. A simple model distinguishing temperature and drought effects was able to explain the majority of climatic effects on the decomposition rates of the various litter types tested in the varying environments over the large geographical areas. Actual evapotranspiration summarizing the temperature and drought effects was not as general climatic predictor of the decomposition rate.