Land use change geo-spectrum and its topographic gradient effect in the core areas of Hani Terraces, a World Heritage Site

Topographic factors have been confined to the temporal and spatial changes of land use and resource allocation in mountainous heritage areas. Taking the Core Area of Hani rice Terraces as the study area, this study aims to explore the geo-spectrum of land use and topographic gradient, thereby enchancing the landscape integrity and ecological service function of terraced fields for sustainable development. The land-use types were first classified as grassland, road, construction, shrub, dry land, bare land, tree, plantation, paddy field, river, and garden using Google Earth 0.55m×0.55m high-precision historical images in 2005, 2010, 2015, and 2020. After that, the transfer matrix was calculated to determine the geo-spectrum, distribution index, topographic index (TI), hypsometric index (HI), slope length factor, slope steepness factor (LS), topographic wetness index (TWI), the temporal and spatial variation of land use and topographic gradient in the study area from a long time scale and a complete spatial range. A field test was finally conducted to verify the data using visual interpretation. The results showed that: 1) The main types of land use in the study area were the arbor woodland and paddy field, followed by the dry land and shrub land from 2005 to 2020. Specifically, there was an annual increase in the arbor forest, garden land, plantation land, road, and house building, where the comprehensive growth rates were 12.73%, 148.32%, 118.68%, 138.76%, and 70.58%, respectively. The reduction rates of the paddy field and dry land were -16.56% and -28.56%, respectively. The water areas fluctuated from 0.28 km2 in 2005 to 0.69 km2 in 2020, where the overall growth rate was 147.27%. By contrast, there was a fluctuated reducing trend in the shrub land and grassland with the comprehensive decreasing rates of -10.87% and -29.58%, respectively. 2) The information maps illustrated the stable, continuous, and repeated types of land use, accounting for 37.43%, 25.36%, and 18.41%, respectively. The stable type was the paddy field and arbor forest land, concentrating mainly on the low LS gradient and high TWI gradient. The continuous change type was from the grassland to the arbor forest, and from the paddy field to the plantation, focusing on the low TI gradient and high LS gradient. The recurrent variations were mainly in the paddy-dry land and arbor-shrub forest, concentrating on the low HI, high LS, and low TWI gradient. The land use map in the study area presented a significant effect of terrain gradient on the HI, LS, and TWI indices, indicating the dynamic characteristics of the terrain, erosion intensity, and potential spatial variation of soil moisture. Consequently, the dynamic mechanism of terrain gradient was effectively interpreted using the information maps of land use. 3) Some recommendations were addressed: (1) To strengthen the main connection of "point (village) - axis (water system, ditch) - surface (forest and terraced fields)", as the tourism developed; (2) To implement the differential development and complementarity of different terraced gradient areas; (3) To maintain the current characteristics of the stable areas in the geo-spectrum, while optimizing the spatial structure of the pre-change and intermediate change areas, and slow down the expansion of the late-change type; (4) To promote the continuously and repeatedly changing areas to the stable type in the heritage areas.