Automatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance the accuracy of the deep learning models in bird detection. Trained with virtual birds of sufficient variations in different environments, the model tends to focus on the fine-grained features of birds and achieves higher accuracies. Based on the 100 terabytes of 2-month continuous monitoring data of egrets, our results cover the findings using conventional manual observations, e.g., vertical stratification of egrets according to body size, and also open up opportunities of long-term bird surveys requiring intensive monitoring that is impractical using conventional methods, e.g., the weather influences on egrets, and the relationship of the migration schedules between the great egrets and little egrets.
Landslides are common in tropical and subtropical regions with hilly terrains and heavy rainstorms, and can cause significant economic, ecological, and social impacts.Natural forest succession is usually slow on landslides due to poor soil structure and lack of woody plant seeds, and often comes with a higher risk of repeated landslides.Active forest restoration has been increasingly suggested as an effective alternative to restore exposed landslide scars. Here, we evaluated the effectiveness of using different seed coatings and surface soil treatments on seed germination and seedling establishment of pioneer and late-successional tree species on landslide scars in Hong Kong. Our results show that biochar-dominant seed coating boosted seed germination rate by an additional 9.33% (Standard error = 0.04), while clay-dominant seed coating did not show significant effects. The effects on seedling survival were not significant in both seed coating and surface soil treatments. Our results also show that large seed and later successional species have significantly higher germination rates than pioneer species and are able to successfully establish on landslide sites.These results suggest that direct seeding using a biochar seed coat with large seeds or late-successional species is a useful method to enhance tree seed germination-an essential first step to restore the forests after landslide disturbances, with potential to be extended to other humid tropical and subtropical forests.
The cover image is based on the Research Article Vegetation regeneration on natural terrain landslides in Hong Kong: Direct seeding of native species as a restoration tool by Ying Ki Law et al., https://doi.org/10.1002/ldr.4492.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.