JIA-2018-09

2099 WANG Di et al. Journal of Integrative Agriculture 2018, 17(9): 2096–2106 related). For maize, the global Moran’s I as calculated by Cliff and Ord (1981) is: ∑ n i =1 I = ( y i –y ) 2 ∑ n i =1 ∑ n j =1 w ij ∑ n i =1 ∑ n j =1 w ij ( y i –y )( y j –y ) n (1) Where, I is the global spatial autocorrelation index; n is the number of sampling units; y i is the maize planting area in the i th sampling unit; y j is the maize planting area in the j th sampling unit; y is the average value of maize planting area in all sampling units; and W ij is the spatial adjacency between sampling units i and j . The spatial adjacency matrix was calculated using ArcGIS Software, and its type was Contiguity-Edges-Corners, which means the sampling units that share a boundary or node will influence the computation of the target unit. The same equation was applied to rice area and the other analyzed variables. The significance of Moran’s I is tested based on the Z -score as given by eqs. (2)–(6). The Z -score was also calculated using ArcGIS Software. Z -score= I – E ( I ) var ( I ) (2) E ( I )= – 1 n– 1 (3) var ( I )= ×[ n 2 S 1 –nS 2 +3( ) 2 ] – ( n– 1)( n+ 1)( ) 2 ∑ n i =1 ∑ n j =1 w ij ∑ n i =1 ∑ n j =1 w ij 1 ( n– 1) 2 1 (4) S 1 = ( w ij + w ji ) 2 1 2 ∑ n i =1 ∑ n j =1 (5) Administrative boundary data Spatial autocorrelation analysis Data Spatial autocorrelation index Sampling unit size Moran’s I calculation Spatial sampling frame Stratified sampling design Stratification criterion Cultivated land fragmentation Moran’s I calculation Moran’s I calculation Moran’s I calculation Spatial autocorrelation analysis Crops mapping data DEM data Cultivated land data Rational sampling unit size Spatial sampling scheme for crop acreage estimation Optimized stratification criterion Crop planting intensity Ground slope Fig. 1 The overall flowchart of this study.