Human activities affecting the species richness of urban spontaneous herbs under a three-scale factor framework

Abstract

Spontaneous herbs provide natural solutions for the sustainable development of urban environments. However, a wide range of species losses have occurred because of the lack of knowledge regarding the impacts of human activities. The results from field investigation and analyses based on the three-scale factor framework were as follows: 1) At the district scale, the top four contributing factors were all tax and revenue indicators related to governmental income according to the Hierarchical Partitioning results. Compared with Ridge regression, Lasso regression demonstrated substantially higher predictive performance. 2) At the patch scale, among the variables investigated, Cement Roads, One-Floor Buildings, Public Service Land, and Main Roads were the most important (adjusted R² > 0.2). Three different multiple linear regression models explained 24.9 %–56.1 % of the variance. 3) At the plot scale, the herbs’ growth condition indicators presented significant influence. This study provides a three-scale factor framework and valuable data for quantifying the influence of human activities, which is critical for the sustainable development of cities.