Abstract: Desert riparian forest is a unique forest community type found in extreme arid areas with a simple community structure.The minimum sampling area of Populus pruinosa communities in the desert riparian forest in the upper reaches of the Tarim River was studied using species-area,community coefficient-area,and importance value-area curves.Results showed the fitting effect was obviously different in the four saturated species-area curves.The S=aA/(1+bA) and S=a(1-e^-bA) models had higher fitting correlation coefficients (P<0.01) and the best fitting results compared with the two other models.Using these two models,the minimum sampling area for P.pruinosa communities corresponding to 60%-80% of the total number of species was 100 m².The minimum sampling area corresponding to 90% of the total number of species was 200 m².The minimum sampling area calculated by the community coefficient-area curves and importance value-area curves was 400 m²,with both methods overcoming the shortcomings of the species-area curves,which only focused on species appearance or absence,and the estimation problems of saturation species.The fitting effect better conformed to the actual situation,and the two methods were suitable for determining the minimum sampling area required in the desert riparian forest of the Tarim Basin.These results indicated that the minimum sampling area corresponding to 60%,80%,and 90% of total species number of the P.pruinosa community were 100 m²,200 m²,and 400 m²,respectively.The optimal sampling area of the P.pruinosa community in the desert riparian forest in the upper reaches of Tarim River was 400 m².