Understanding patterns and potential drivers of forest diversity in northeastern China using machine-learning algorithms

Prepared by Weixue Luo, Chunyu Zhang, Xiuhai Zhao & Jingjing Liang

Mixed temperate coniferous and broad-leaved forest in northeastern China. Photo credit: Xiuhai Zhao.

Large-scale biodiversity patterns have been a focus of ecological studies over the past decades, but it remains unclear which anthropogenic and environmental factors have the strongest impact on tree diversity patterns, especially in temperate forest ecosystems of East Asia.

We studied this question based on geospatial covariates and ground-source forest inventory data collected from across the mixed temperate forests in northeastern China. We used these data and trained different machine learning and statistical models to study spatial patterns of tree species diversity and their underpinning drivers.

We found that the spatial distribution of tree species diversity, in terms of species richness and evenness, is strongly associated with climatic (annual mean precipitation and annual mean temperature), topographic (elevation and slope), and anthropogenic factors. More specifically, tree species richness is positively associated with mean annual precipitation, seasonality in precipitation, elevation, terrain slope, and human footprint. In contrast, tree species evenness is negatively associated with mean annual precipitation, mean annual temperature, maximum temperature of the warmest month, human footprint and terrain slope. Anthropogenic factors play a greater role in tree species evenness more than in tree species richness. These findings shed light on the processes behind community assembly and biodiversity patterns.

Geographic distribution of estimated tree species richness (S) on a 0.1-ha basis in 2017 across northeastern China’s mixed temperate forests. Figure 4 from the original paper (Luo et al. 2021).

Based on these relationships, we developed the first wall-to-wall map of tree species diversity for northeastern China at a 1-km resolution. This high-resolution map provides a benchmark for future biodiversity assessment and can help landowners and land management agencies in their decision-making processes about sustainable forest management. This timely map also facilitates biodiversity conservation and forest restoration, a priority task outlined by the recently implemented 2050 China National Forest Management Plan.

Geographic distribution of estimated tree species evenness (E) on a 0.1-ha basis in 2017 across northeastern China’s mixed temperate forests. Figure 5 from the original paper (Luo et al. 2021).

This is a plain language summary for the paper of Luo et al. published in the Journal of Vegetation Science (https://doi.org/10.1111/jvs.13022).