Spatial variations in the relationships between road network and landscape ecological risks in the highest forest coverage region of China. (January 2019)

• Record Type: Journal Article
• Authors: Lin, Yuying; Hu, Xisheng; Zheng, Xiaoxue; Hou, Xiuying; Zhang, Zhengxiong; Zhou, Xinnian; Qiu, Rongzu; Lin, Jinguo
• Is Part Of: Ecological indicators. Volume 96(2019)Part 1
• Page Start: 392
• Page End: 403
• Publication Date: 2019-01
• Subjects: Landscape ecological risk -- Road network -- Kernel density estimation -- Geographically weighted regression -- The Minjiang River; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2018.09.016
• Languages: English
• ISSNs: 1470-160X

Abstract

Graphical abstract: Highlights: A comprehensive index was constructed to evaluate landscape ecological risk (LER). GWR was introduced to observe the spatial association between road network and LER. The variation in the associations was quantified and visualized. The different effect between high-level road and low-level road on LER was observed. The significant cumulative effects of road network on LER existed in the study area. Abstract: The road network is one of the most ubiquitous and significant long-term legacies of all types of human disturbances on the landscape. Taking the upper reaches of the Minjiang River in Fujian Province of southeast China as a case, the spatiotemporal dynamics of the landscape patterns and landscape ecological risk (LER) were explored, and based on the geographically weighted regression (GWR) model, the geographical heterogeneity in the correlations between the road network and the LER were identified. Our results showed that: (1) The distribution of the LER had a gradually decreasing trend from the middle to the periphery in 2007, with the high-risk area expanding to the western part of the study area in 2012 and 2016. The LER close to the road network was generally higher than those far from the road network. (2) The GWR model fit our case better than the ordinary least square (OLS) model, with both of the measurements of the road network (i.e., distance to the nearest road, DNR; and kernel density estimation, KDE) being significantly correlated with the LER at the 1% level. (3) According to the quantified coefficients estimated by the GWR model, we found that there were spatial variations in the associations between the two regressors and different level effects of roads on the LER. (4) The GWR analysis also indicated that the high-level roads mainly affected areas where human activities were more intensive, whereas the low-level roads infiltrated every corner of the region, mainly affecting areas that were far from the city. (5) The significant cumulative impacts of the road network on the LER were also observed in this study. Benefitting from the quantification and visualization of the spatial paradigm in regard to their trade-off and the synergistic associations between the LER and the road network at the grid level, our study provides suggestions for implementing more appropriate policies that will alleviate the impact of road construction on the landscape. This study also sheds light on further applications of the GWR model in future research on road ecology.