Spatio-temporal patterns of precipitation-use efficiency of vegetation and their controlling factors in Inner Mongolia

Abstract

Aims Precipitation-use efficiency(PUE) is an important indicator for understanding how net primary productivity(NPP) in arid and semi-arid ecosystems responds to variations in precipitation. The objective of this study was to determine the spatio-temporal patterns and responses to climatic and biotic factors of PUE at a regional scale. Methods CASA(Carnegie-Ames-Stanford Approach) model was used to simulate NPP in Inner Mongolia during 2001–2010 based on the MOD13A1 data and spatially interpolated meteorological data. PUE was calculated as the ratio of NPP to annual precipitation. The effects of fraction of vegetation cover(FVC) and leaf area index(LAI) on PUE were also investigated. The FVC was calculated with the dimidiate pixel model based on the MOD13A1 data. LAI data were acquired as the MODIS LAI products. Important findings The multi-year average PUE of Inner Mongolia was 0.94 g C·m–2·mm–1, exhibiting apparent increasing trend at an average rate of 0.55 g C·m–2·mm–1 per 10° with changes in longitude from 105° E to 120° E. The spatial patterns of PUE showed significant differences among vegetation types. The PUE was highest in shrubs and lowest in desert. The spatial distribution of PUE responded differentially to climatic factors in different precipitation ranges. Where precipitation was less than 75 mm, PUE showed a significant negative correlation with temperature and precipitation(R2 = 0.226, p 0.05). In the area with precipitation of 175–300 mm, PUE exhibited a significant positive correlation with temperature and precipitation(R2 = 0.878, p 0.001), and increased significantly(R2 = 0.94, p 0.001) with precipitation at a rate of 0.57 g C·m–2·mm–1 per 100 mm. In the area where precipitation was higher than 475 mm, PUE increased spatially with increasing temperature and decreasing precipitation. In this precipitation range, the effect of temperature on spatial variance of PUE was 8.61 times of that of precipitation. The inter-annual variation of PUE also had different responses to climatic factors in different precipitation ranges. Where precipitation was less than 220 mm, PUE displayed a positive correlation with precipitation and a negative correlation with temperature. In the area with precipitation of 220–310 mm, precipitation had a much greater effect than temperature on the inter-annual variations of PUE. Where precipitation was higher than 310 mm, PUE showed positive correlations with both temperature and precipitation. In relatively humid parts of this precipitation range, however, the PUE showed a poor correlation with precipitation, but a stronger correlation with temperature. FVC was linearly related with the spatial distribution(R2 = 0.73, p 0.001) and inter-annual variations of PUE(R2 = 0.11, p 0.001). LAI showed a linear relationship with the inter-annual variations of PUE(R2 = 0.42, p 0.001). In the area where LAI was less than 3.15 and with non-forest vegetation, LAI was linearly related with the spatial distribution of the PUE.