Monitoring marine biodiversity is a challenge in some vulnerable and difficult-to-access habitats, such as underwater caves. Underwater caves are a great focus of biodiversity, concentrating a large number of species in their environment. However, most of the sessile species that live on the rocky walls are very vulnerable, and they are often threatened by different pressures. The use of these spaces as a destination for recreational divers can cause different impacts on the benthic habitat. In this work, we propose a methodology based on video recordings of cave walls and image analysis with deep learning algorithms to estimate the spatial density of structuring species in a study area. We propose a combination of automatic frame overlap detection, estimation of the actual extent of surface cover, and semantic segmentation of the main 10 species of corals and sponges to obtain species density maps. These maps can be the data source for monitoring biodiversity over time. In this paper, we analyzed the performance of three different semantic segmentation algorithms and backbones for this task and found that the Mask R-CNN model with the Xception101 backbone achieves the best accuracy, with an average segmentation accuracy of 82%.

Taking the west slope of Cangshan Mountain in Yangbi County, Dali as the research site, on the basis of investigating the local natural geographical conditions, topography and biodiversity status of Cangshan Mountain, the CAP protection action planning method was adopted, and the priority protection objects were determined to be native forest vegetation, rare and endangered flora and fauna, alpine vertical ecosystems, hard-leaf evergreen broad-leaved forests and cold-tempered coniferous forests; The main threat factors were commercial collection, tourism development and overgrazing. Biodiversity conservation on the western slope of Cangshan Mountain should take species as “point”, regional boundary as “line”, ecosystem and landscape system as “plane”, so as to realize the overall planning structure system combining “point—line—plane”, which can be divided into conservation core area, buffer zone and experimental area. The results can provide a reference for biodiversity conservation on the western slope of Cangshan Mountain.

This paper explores the integration of Large Language Models (LLMs) and Software-Defined Resources (SDR) as innovative tools for enhancing cloud computing education in university curricula. The study emphasizes the importance of practical knowledge in cloud technologies such as Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS), DevOps, and cloud-native environments. It introduces Lean principles to optimize the teaching framework, promoting efficiency and effectiveness in learning. By examining a comprehensive educational reform project, the research demonstrates that incorporating SDR and LLMs can significantly enhance student engagement and learning outcomes, while also providing essential hands-on skills required in today’s dynamic cloud computing landscape. A key innovation of this study is the development and application of the Entropy-Based Diversity Efficiency Analysis (EDEA) framework, a novel method to measure and optimize the diversity and efficiency of educational content. The EDEA analysis yielded surprising results, showing that applying SDR (i.e., using cloud technologies) and LLMs can each improve a course’s Diversity Efficiency Index (DEI) by approximately one-fifth. The integrated approach presented in this paper provides a structured tool for continuous improvement in education and demonstrates the potential for modernizing educational strategies to better align with the evolving needs of the cloud computing industry.

In order to scientifically evaluate the germplasm resources of Momordica charantia in southern China, the diversity, correlation and cluster analysis were carried out on the main botanical characters of 56 Momordica charantia varieties, such as melon length, melon transverse diameter, single melon weight, internode length, stem diameter, leaf length and leaf width. The results showed that the variation coefficients of 7 agronomic characters of 56 Momordica charantia varieties ranged from 8.81% to 19.44%, the average variation coefficient was 14.21%, the maximum variation coefficient of single melon weight was 19.44%, and the minimum variation coefficient of melon cross diameter was 8.81%. The correlation analysis showed that there were correlations among the agronomic traits. The positive correlation coefficient between leaf length and leaf width was up to 0.978, and the negative correlation coefficient between single melon weight and internode length was up to 0.451. The 56 varieties were divided into 3 groups by cluster analysis, of which 92.86% of the materials were concentrated in the first and second groups, and there were only 4 materials in the third group. The results can provide a reference for the cultivation, utilization and genetic improvement of Momordica charantia resources in southern China.

Tropical dry forests are complex and fragile ecosystems with high anthropogenic intervention and restricted reproductive cycles. They harbor unique richness, structural, physiological and phenological diversity. This research was carried out in the upper Magdalena valley, in four forest fragments with different successional stages. In each fragment, four permanent plots of 0.25 ha were established and the light habitat associated with species richness, relative abundance and rarity was evaluated, as well as the forest dynamics that included mortality, recruitment and diameter growth for a period of 5.25 years. In mature riparian forest, species richness was found to be higher than that reported in other studies for similar areas in the Cauca Valley and the Atlantic coast. Values of species richness, heterogeneity and rarity are higher than those found in drier areas of Tolima. Forest structure, diversity and dynamics were correlated with light habitat, showing differences in canopy architecture and its role in the capture and absorption of radiation. The utilization rate of photosynthetic effective radiation in the forest underlayer with high canopy density is low, which is related to the low species richness, while the underlayer under light is more abundant and heterogeneous.

In light of swift urbanization and the lack of precise land use maps in urban regions, comprehending land use patterns becomes vital for efficient planning and promoting sustainable development. The objective of this study is to assess the land use pattern in order to catalyze sustainable township development in the study area. The procedure adopted involved acquiring the cadastral layout plan of the study area, scanning, and digitizing it. Additionally, satellite imagery of the area was obtained, and both the cadastral plan and satellite imagery were geo-referenced and digitized using ArcGIS 9.2 software. These processes resulted in reasonable accuracy, with a root mean square (RMS) error of 0.002 inches, surpassing the standard of 0.004 inches. The digitized cadastral plan and satellite imagery were overlaid to produce a layered digital map of the area. A social survey of the area was conducted to identify the specific use of individual plots. Furthermore, a relational database system was created in ArcCatalog to facilitate data management and querying. The research findings demonstrated the approach's effectiveness in enabling queries for the use of any particular plot, making it adaptable to a wide range of inquiries. Notably, the study revealed the diverse purposes for which different plots were utilized, including residential, commercial, educational, and lodging. An essential aspect of land use mapping is identifying areas prone to risks and hazards, such as rising sea levels, flooding, drought, and fire. The research contributes to sustainable township development by pinpointing these vulnerable zones and providing valuable insights for urban planning and risk mitigation strategies. This is a valuable resource for urban planners, policymakers, and stakeholders, enabling them to make informed decisions to optimize land use and promote sustainable development in the study area.