Modifying the maximal light-use efficiency for enhancing predictions of vegetation net primary productivity on the Mongolian Plateau

Maximal light-use efficiency (LUE), e(max), is a measure of the conversion efficiency of the photosynthetically active radiation absorbed by plants to net primary productivity (NPP), based on the principle that an optimal plant productive capability exists when the LUE is at its maximum. e(max) is an important parameter for modelling regional NPP and is conventionally applied at the biome level using a constant value. In this study, we estimated type-specific e(max) values for three dominant land cover types on the Mongolian Plateau: 0.621 g C MJ(-1) (MJ is the mega Joules and 1 MJ = 10(6) J) for meadow steppe, 0.534 g C MJ(-1) for typical steppe, and 0.520 g C MJ(-1) for desert steppe. With these steppe-specific modified e(max) values, we were able to examine changes in NPP for 2001-2015 on the plateau, as well as their likely responses to regional climate change. The use of different e(max) values for each steppe type improved the accuracy of the Carnegie Ames Stanford Approach (CASA) Biosphere model predictions of grassland NPP by 18.8% (R-2 = 0.48 to R-2 = 0.57; R-2 is the coefficient of determination) over the observation period. Previous studies based on a constant e(max) (0.541 g C MJ(-1)) appear to underestimate e(max) and NPP in meadow steppe, highlighting the importance of setting type-specific e(max) values for different land cover types in the remote-sensing modelling of grassland NPP. However, more detailed maps of biome sub-classes, with species composition, would be valuable for future attempts to determine appropriate e(max) values. The growing season (April-October) NPP on the plateau increased significantly from 2001 to 2015, with an annual increment of 4.44 g C m(-2) y(-1). This trend was strongly governed by the change in summer NPP across the plateau. In comparison, NPP in the spring and autumn did not influence the change in total NPP, which was likely due to their relatively small values. Summer precipitation and the related drought stress were the chief factors responsible for plateau-scale NPP changes, due to the high proportion of summer precipitation and NPP in the annual totals. This may induce important environmental features, such as dzud (a Mongolian term for a severe winter) and desertification on the plateau.