Factors impacting ecological network in Changsha-Zhuzhou-Xiangtan urban agglomeration, China—Based on the perspective of functional performance

Abstract

Research on ecological network of highly urbanised areas still suffers from a lack of species specificity in the identification of sources and corridors, solidified identification range with errors, and neglect of stepping stones. This study examined the evolution of ecological network components (habitats and corridors) of the active urban birds, with factors that impact them, from the perspective of functional performance, with the aim of evaluating ecological networks using a flexible identification approach. Habitat function performance (level of ecological source function performance, SFP; level of ecological stepping-stone function performance, SsFP) and corridor function performance (CFP) per grid were constructed using a flexible identification strategy to ensure that the ecological network function of each grid was not overlooked. Additionally, nonlinear relationships between bird habitats in ecological network components and three driving factor types such as topography, resistance, and landscape pattern were explored at the global scale using the Moran’s I, geographical detector, and regression equations. Local regressions were used to determine the impact of regional landscape patterns on CFP for management recommendations. The following results were observed: (1) Of the 238 grids with significant continuous changes in SFP and SsFP, scenario 16 (S16: SFP continuously decreasing with SsFP continuously increasing) accounted for 33.40%, indicating more grids had a transfer in primary function performance from source to stepping stones, mainly in the central part of Changsha-Zhuzhou-Xiangtan urban agglomeration (CZXUA). The amount of change in the continuous change scenario was concentrated within the range 0.2–0.6. (2) Spatial correlation analysis showed that the positive indicators, percentage of construction land (CLP) and nighttime lighting data (NLD), and the negative indicator, normalised difference vegetation index (NDVI), in resistance had consistently increasing negative and positive correlations with SFP, whereas the correlations with the landscape pattern indices remained stable. Topography and resistance factors contributed more to the spatial heterogeneity of SFP, whereas their contribution to SsFP demonstrated the opposite trend. The regression relationship between habitat function performance and explanatory variables showed a high R2 value for nonlinear fitting. (3) Multiscale CFP was identified based on habitat function and the contribution of resistance factors. Geographically weighted regression analysis showed significant positive and negative impacts on CFP at the local scale, with landscape diversity and patchy irregularities providing more spaces for potential ecological corridor functions, and that regional landscape regulation strategies need to be weighed from multiple perspectives. Overall, this study proposes a flexible method for identifying ecological networks and multi-scale urban ecological management strategies.