We carried out
in situ monitoring of methane (CH
4) fluxes of four ecosystems, alpine marsh, alpine meadow, alpine steppe, and alpine desert, in the Yushu Tibetan Autonomous Prefecture and Maduo County, with a portable greenhouse gas analyzer. Biomass, microorganism abundance, nutrients, soil moisture and temperature in each ecosystem were measured. These parameters were investigated to describe spatial-temporal variations in CH
4 flux and quantify the key influencing factors in all the ecosystems. The results showed that alpine marsh and alpine meadow were CH
4 sources, with the maximum flux in August. Alpine steppe and alpine desert were CH
4 sinks, with a minimum in August. During the growing season, there were significant differences in CH
4 flux among the four ecosystems (
P<0.05). The abundance of
mcrA gene in alpine marsh was highest, followed by alpine meadow, while the abundance of
pmoA gene was highest in alpine meadow, followed by alpine marsh. The lowest abundances of
mcrA and
pmoA genes were found in alpine desert. There were significant differences in
mcrA and
pmoA among the four ecosystems (
P<0.05). CH
4 flux in alpine marsh and alpine meadow was significantly positively correlated with soil temperature and
mcrA (
P<0.05), whereas CH
4 flux in alpine steppe and alpine desert was significantly negatively correlated with soil temperature and
pmoA (
P<0.05). Across all the ecosystems, CH
4 flux was significantly correlated with soil moisture, organic carbon, total nitrogen, biomass,
mcrA and
pmoA (
P<0.05). Results of path analysis showed that soil organic carbon and the abundance of
mcrA and
pmoA had significant direct effects on CH
4 emissions, while soil temperature and water content indirectly affected it by changing soil microorganism abundances. Among all these key factors influencing CH
4 flux, the relative contribution of
mcrA abundance was the highest (up to 30.53%), followed by organic carbon and biomass. The results indicated that variations in CH
4 flux among the different ecosystems were caused by differences in microorganism abundance, organic carbon, and biomass. It is therefore important to consider the heterogeneity of CH
4 emissions among different ecosystems when modeling and estimating CH
4 fluxes in alpine areas.