Simulation of potential endangered species distribution in drylands with small sample size based on semi-supervised models

Abstract

Identifying suitable habitats for endangered species is critical to promote the recovery of endangered species. However, Conventional species distribution models (SDMs) need large amounts of labeled sample data to learn the relationship between species and environmental conditions, is difficult to fully detangle the role of the environment in the distribution of endangered species, which are very sparsely distributed and have environmental heterogeneity. This study’s first innovation used the semi-supervised model to accurately simulate the suitable habitats for endangered species with a small sample size. The model performance was compared with three conventional SDMs, namely Maxent, Generalized Linear Model (GLM), and Support Vector Machine (SVM). Applying the model to the endangered species Populus euphratica in the lower Tarim River basin, Northwest China. The results showed that the semi-supervised model exhibited better performance than conventional SDMs with an accuracy of 85% when only using 443 Populus euphratica samples. All models developed using smaller sample sizes exhibit worse performance in the prediction of habitat suitability areas for endangered species while the semi-supervised model is still excellent. The results showed that the suitable habitat for Populus euphratica is mainly near the river channel of the lower Tarim River basin, accounting for 13.49 % of the study area. The lower Tarim River still has enormous land potential for the restoration of endangered Populus euphratica. The model developed here can be used to evaluate the suitable habitat for endangered species with only a small sample size, and provide a basis for the conservation of endangered species.

Precise vegetation restoration in arid regions based on potential natural vegetation and potential normalized difference vegetation index

“…The potential spatial distribution of natural vegetation and higher vegetation coverage level in the LTRB was primarily near the river course, whereas the middle of the study area was unsuitable for vegetation growth (Liu et al 2014). This potential spatial distribution pattern was formed mainly because the groundwater level in the area near the water conveyance channel was shallow between À2 to À4 m, which was compatible with the optimal groundwater level for vegetation growth in the LTRB (Chen et al 2015). Compared with grass and shrubs, the potential distribution extent and potential vegetation coverage level of trees were significantly superior to the actual situation.…”

“…Fewer studies have been conducted on potential ecological restoration patterns in this region by comparing the most natural theoretical vegetation distribution patterns with the actual situation. Ci et al (2023) used PNV to simulate the potential suitable growth distribution of P. euphratica in the LTRB. However, accurate potential vegetation types, growth sites, suitable distribution areas, and coverage levels in LTRB are still unknown.…”