Abstract: Abstract: Production-living-ecological space is a typical spatial carrier to coordinate economic, social， and ecological benefits. Its coordinated development has been greatly significant to promote the ecological civilization, even to optimize the spatial planning in land use. However, the spatial function differentiation is still lacking on the different plots in the same land type, particularly in the most current evaluation on the production-living-ecological space. The clustering or Coordination Degree (CD) is also less sufficient in production-living-ecological space. The purpose of this study is to improve the evaluation system of production-living-ecological space, further to partition its aggregation and coordination degree for the land-use spatial optimization. An integration of multi-source spatial data was implemented to evaluate the functions of production-living-ecological space in different plots with the same land from the grid scale, with emphasis on the perspective of spatial heterogeneity. A systematic analysis was then made on the spatial autocorrelation and trigonometric coordinates. As such, the clustering and coordination characteristics were revealed on the production-living-ecological space on the township scale. Zoning schemes were finally proposed for the optimization of land space. The results are as follows. 1) The high score areas of production space in Hainan Island were concentrated in the coastal cities and counties, indicating the distribution characteristics of "higher around and lower in the middle, higher in the north and lower in the south". There were also differences between the urban and rural production spaces. The high score areas of living space were mainly concentrated in the cities and counties, with the spatial pattern of "small part gathering and most scattered". The high score areas of ecological space were concentrated in the middle, with the distribution characteristics of "higher in the middle and lower around". 2) The Moran's I indexes for the production, life, and ecological space function were 0.569, 0.221 and 0.716, respectively, in Hainan Island obtained from the township scale. It inferred that there was the strongest spatial correlation between the spatial autocorrelation characteristics and ecological space. Therefore, a zoning scheme was proposed for the whole island using the characteristics of spatial clustering. 3) There were great differences in the coordination degree of production-living-ecological space in different towns of Hainan Island. The high value areas were dominated by ecological space and coordinated type, accounting for 51.72% and 43.97% of the total number of towns, respectively. The low value area was dominated by ecology space, accounting for 72.41% of the total number of towns, where no living dominant town was observed. In the case of spatial features, a zoning scheme was proposed for the whole island using the characteristics of spatial coordination. The heterogeneity of spatial functions was effectively identified in the production-living-ecological space for the different plots with the same land type. A specific scheme of production-living-ecological spatial zoning was also proposed from two new perspectives of spatial clustering and coordination degree, suitable for the current practice of land space optimization. An optimization path of production-living-ecological spatial evaluation can provide the zoning schemes, according to the spatial characteristics. The finding can offer a potential decision making for the land space optimization, as well as the construction of free trade port and ecological civilization in Hainan Island of southern China.