Abstract: Understanding the potential distribution of Spartina alterniflora under climate change is necessary to formulate effective management measures and mitigate invasion danger. In this study, a multi-source data integration technology was developed to collect occurrence records of Spartina alterniflora in the Chinese coast. The MaxEnt model was used to predict the potential distribution of Spartina alterniflora under the current and future(2050s) climate change scenarios, and the effect of climate factors on species distribution was analyzed. Results show that sufficient occurrence records of Spartina alterniflora were obtained using the data integration technology, which can overcome spatial sampling bias and solve the difficulty of species information collection in complex environment. The prediction results of MaxEnt model were satisfactory, and the average values of AUC(area under curve) and standard deviation were 0.943 and 0.005, respectively. Under the current climatic condition, suitable areas for the distribution of Spartina alterniflora account for 9.35% of the total coastal areas, with the highly suitable areas accounting for 1.35%, which are mainly located in salt marshes of Jiangsu, Zhejiang and the north of Fujian. The spatial distribution of Spartina alterniflora are greatly affected by the climate factors, including isothermality, mean diurnal range, mean temperature of warmest quarter, etc., and these factors contribute more than 70% of the variation. Under two climate scenarios of RCP4.5 and RCP8.5, the suitable areas of Spartina alterniflora in 2050s will increase by 22.82% and 24.19%, respectively, compared with the current climate model, and show an expansive trend from coastline to inland. The high suitable area will increase by 7 470.04 and 13 208.47 km², respectively, and the expansion of medium and low suitable areas will be more obvious.

Key words: Spartina alterniflora, multi-source data integration, maximum entropy, potential distribution, climate change