Spatial and temporal variations of C3/C4 relative abundance in global terrestrial ecosystem since the Last Glacial and its possible driving mechanisms

The primary factor controlling C-3/C-4 relative abundance in terrestrial ecosystem since the Last Glacial has been widely debated. Now more and more researchers recognize that climate, rather than atmospheric CO2 concentration, is the dominant factor. However, for a specific area, conflicting viewpoints regarding the more influential one between temperature and precipitation still exist. As temperature and precipitation in a specific area usually not only vary within limited ranges, but also covary with each other, it is difficult to get a clear understanding of the mechanism driving C-3/C-4 relative abundance. Therefore, systematic analysis on greater spatial scales may promote our understanding of the driving force. In this paper, records of C-3/C-4 relative abundance since the Last Glacial on a global scale have been reviewed, and we conclude that: except the Mediterranean climate zone, C-3 plants predominated the high latitudes during both the Last Glacial and the Holocene; from the Last Glacial to the Holocene, C-4 relative abundances increased in the middle latitudes, but decreased in the low latitudes. Combining with studies of modern process, we propose a simplified model to explain the variations of C-3/C-4 relative abundance in global ecosystem since the Last Glacial. On the background of atmospheric CO2 concentration since the Last Glacial, temperature is the primary factor controlling C-3/C-4 relative abundance; when temperature is high enough, precipitation then exerts more influence. In detail, in low latitudes, temperature was high enough for the growth of C-4 plants during both the Last Glacial and the Holocene; but increased precipitation in the Holocene inhibited the growth of C-4 plants. In middle latitudes, rising temperature in the Holocene promoted the C-4 expansion. In high latitudes, temperature was too low to favor the growth of C-4 plants and the biomass was predominated by C-3 plants since the Last Glacial. Our review would benefit interpretation of newly gained records of C-3/C-4 relative abundance from different areas and different periods, and has its significance in the understanding of the driving mechanisms of C-3/C-4 variations on longer timescales (e.g., since the late Miocene) with reliable records of temperature and atmospheric CO2 concentration.