Hongyan Jia

It is crucial to accurately identify precious tree pests in a real, complex natural environment in order to monitor the growth of precious trees and provide growers with the information they need to make effective decisions. However, pest identification in real complex natural environments is confronted with several obstacles, including a lack of contrast between the pests and the background, the overlapping and occlusion of leaves, numerous variations in pest size and complexity, and a great deal of image noise. The purpose of the study was to construct a segmentation method for identifying precious tree pests in a complex natural environment. The backbone of an existing Mask region-based convolutional neural network was replaced with a Swin Transformer to improve its feature extraction capability. The experimental findings demonstrated that the suggested method successfully segmented pests in a variety of situations, including shaded, overlapped, and foliage- and branch-obscured pests. The proposed method outperformed the two competing methods, indicating that it is capable of accurately segmenting pests in a complex natural environment and provides a solution for achieving accurate segmentation of precious tree pests and long-term automatic growth monitoring.

Objective: To investigate the impact of different slope directions on the quantity and quality of the soil seed bank and seedling germination process of Castanopsis hystrix plantations. Method: Fixed sample plots in forest stands of Castanopsis hystrix were established on different slope directions (sunny slope, semi-sunny slope, semi-shady slope, and shady slope). The characteristics of the forest stand were investigated, and per-wood scaling was carried out. The temporal dynamics of the seed rain and seed bank were quantified using seed rain collectors and by collecting soil samples from different depths. The quantity and quality of the seeds were determined, and the vigor of mature seeds was measured throughout the study. Results: (1) The diffusion of Castanopsis hystrix seed rain started in mid-September, reached its peak from late October to early November, and ended in mid-December. (2) The dissemination process, occurrence time, and composition of the seed rain varied between the different slope directions. The seed rain intensity on the semi-sunny slope was the highest (572.75 ± 9.50 grains∙m⁻²), followed by the sunny slope (515.60 ± 10.28 grains∙m⁻²), the semi-shady slope (382.13 ± 12.11 grains∙m⁻²), and finally the shady slope (208.00 ± 11.35 grains∙m⁻²). The seed rain on the sunny slope diffused earliest and lasted the longest, while the seed rain on the shady slope diffused latest and lasted the shortest time. Seed vigor and the proportion of mature seeds within the seed rain were greatest on the semi-sunny slope, followed by the sunny slope, semi-shady slope, and the shady slope. (3) From the end of the seed rain to August of the following year, the amount of total reserves of the soil seed banks was highest on the semi-sunny slope, followed by the sunny slope then the semi-shady slope, and it was the lowest on the shady slope. The amount of mature, immature, gnawed seeds and seed vigor of the soil seed bank in various slope directions showed a decreasing trend with time. The seeds of the seed bank in all slope directions were mainly distributed in the litter layer, followed by the 0–2 cm humus layer, and only a few seeds were present in the 2–5 cm soil layer. (4) The seedling density of Castanopsis hystrix differed significantly on the different slope directions. The semi-sunny slope had the most seedlings, followed by the sunny slope, semi-shady slope, and the shady slope. Conclusions: The environmental conditions of the semi-sunny slope were found to be most suitable for the seed germination and seedling growth of Castanopsis hystrix, and more conducive to the regeneration and restoration of its population.