More reduction in plant productivity than respiration under extreme drought across grassland types along a rainfall gradient: A modeling analysis

Background/Question/Methods

Ecosystem processes of grasslands in the central US Great Plains are strongly governed by both rainfall amount and its variability. An increase in the frequency and intensity of extreme weather events, including severe drought has been projected for this region. Different grassland types, from xeric to mesic, may respond differently to severe drought and its variability. A Terrestrial ECOsystem (TECO) model was used to assess such differential sensitivity to long term drought, as well as its variability in grasslands along a rainfall gradient.

Results/Conclusions

In response to drought, early-year NPP decreased more than Rh across all grassland types and therefore caused NEE to decrease generally from C sink to C source, whereas in later years reduction in Rh surpassed reduction in NPP, causing NEE to increase.  Patterns of responses of C variables to drought types were contingent on grassland types. Across grassland types, the least mean annual sensitivity and variability of NPP occurred in Cheyenne mixed prairie, which was likely resulted from lowest annual mean temperature and lowest variability of annual rainfall. Interannual variability of the ecosystem C components and relative reduction in NPP to drought were generally controlled by annual rainfall amount. Relative reduction in Rh progressively increased with years in each of the grasslands, uncoupled with rainfall. In conclusion, grassland ecosystems release CO₂ under drought condition, but the intensity became less with time.