EVI and NDVI as proxies for multifaceted avian diversity in urban areas

Improved random forest classification model combined with C5.0 algorithm for vegetation feature analysis in non-agricultural environments

In response to the challenges posed by the high computational complexity and suboptimal classification performance of traditional random forest algorithms when dealing with high-dimensional and noisy non-agricultural vegetation satellite data, this paper proposes an enhanced random forest algorithm based on the C5.0 algorithm. The paper focuses on the Liaohe Plain, selecting two distinct non-agricultural landscape patterns in Shenbei New District and Changtu County as research objects. High-resolution satellite data from GF-2 serves as the experimental dataset. This paper introduces an ensemble feature method based on the bagging concept to improve the original random forest classification model. This method enhances the likelihood of selecting features beneficial to classifying positive class samples, avoiding excessive removal of useful features from negative samples. This approach ensures feature importance and model diversity. The C5.0 algorithm is then employed for feature selection, and the enhanced vegetation index (EVI) is utilized for vegetation coverage estimation. Results indicate that employing a multi-scale parameter selection tool, combined with limited field-measured data, facilitates the identification and classification of plant species in forest landscapes. The C5.0 algorithm effectively selects classification features, minimizing information redundancy. The established object-oriented random forest classification model achieves an impressive accuracy of 94.02% on the aerial imagery for forest classification dataset, with EVI-based vegetation coverage estimation demonstrating high accuracy. In experiments on the same test set, the proposed algorithm attains an average accuracy of 90.20%, outperforming common model algorithms such as bidirectional encoder representation from transformer, FastText, and convolutional neural network, which achieve average accuracies ranging from 84.41 to 88.33% in identifying non-agricultural artificial habitat vegetation features. The proposed algorithm exhibits a competitive edge compared to other algorithms. These research findings contribute scientific evidence for protecting agricultural ecosystems and restoring agricultural ecosystem biodiversity.

Positive relationship between bird diversity and human mental health: an analysis of repeated cross-sectional data

BACKGROUND

An increasing body of research has examined the link between biodiversity of birds and human mental health, but most studies only use cross-sectional data. Few studies have used longitudinal or repeated cross-sectional data to investigate the mental health benefits of bird diversity. The aim of this study is to analyse the relationship between bird diversity and mental health at the national level using a unique repeated cross-sectional dataset.

METHODS

I used repeated cross-sectional health data from the German National Cohort health study, collected between March, 2014, and September, 2019, and annual bird citizen science data to investigate the effects of bird-diversity exposure on mental health. Mental health was measured using the summary score of the Patient Health Questionnaire depression module 9 (SumPHQ) and the Short Form Health Survey-12 Mental Health Component Scale. As a proxy for bird diversity, I created a unique indicator called reporting-rate richness and combined it with the health data. Reporting-rate richness measures the number of bird species within postcode areas across Germany in probabilities while accounting for variation in survey efforts. Alternative indicators of bird diversity, such as bird-species richness or abundance, were also calculated. Associations between bird diversity and mental health were estimated using linear regression with region and time fixed effects, adjusted for a range of sociodemographic and environmental confounders and spatial autocorrelation. Interaction terms between income levels and reporting-rate richness were also analysed to examine the moderating effect of socioeconomic status.

FINDINGS

I did the analyses for an unbalanced (n=176 362) and balanced (n=125 423) dataset, with the balanced dataset comprising only regions (postcode areas) in which health data were available for each year. The linear fixed-effects regression analysis indicated a significant negative association between reporting-rate richness and SumPHQ, as observed in both the unbalanced dataset (β -0·02, p=0·017) and the balanced dataset (β -0·03, p=0·0037). Similarly, regression results with both datasets showed a positive relationship between reporting-rate richness and Mental Health Component Scale (MCS; unbalanced β 0·02, p=0·0086; balanced β 0·03, p=0·0018). The moderator analyses revealed a significant influence of socioeconomic status on the relationship between reporting-rate richness and mental health. The robustness of these findings was confirmed through sensitivity analyses.

INTERPRETATION

The results suggest that a greater likelihood of having many different bird species in a person's area of residence might positively contribute to mental health, especially for people with lower socioeconomic status. These findings could have implications for biodiversity conservation and health policy decisions, as governments are facing challenges such as global biodiversity loss and growing public mental health problems.

FUNDING

None.

Distribution patterns and drivers of urban green space and plant diversity in Haikou, China

Investigating historical and ongoing changes in urban green space (UGS) and urban plant diversity (UPD) provides critical insights into urban ecology and urban planning development. The present study illuminates some of the transformations which can occur in rapidly developing urban landscapes. In this work, we used 30 m resolution images from the Landsat 5 satellite from 2015 to investigate UGS patterns in Haikou City, China. Metrics of UPD were obtained using field surveys, allowing the proportion of UGS and UPD to be determined in each urban functional unit (UFU) of Haikou. The results revealed that leisure and entertainment areas (such as parks) had the highest diversity, whereas roads and transportation hubs had the lowest. More frequent anthropogenic maintenance had a positive effect on the total number of species, including cultivated, tree, and herb species. Similarly, increased watering frequency had a positive impact on the diversity of cultivated and shrub species. By providing demonstrating a crucial link between UGS and UPD, the results provide valuable information for planning sustainable urban development in Haikou City and other tropical regions. They highlight the important role of UGS in maintaining biodiversity and providing a range of ecosystem services. This research will inform policymakers and urban planners about the need to consider UGS and UPD in urban planning and management process, in order to promote sustainability and conservation of biodiversity.

Effects of light and noise pollution on avian communities of European cities are correlated with the species' diet

Urbanization affects avian community composition in European cities, increasing biotic homogenization. Anthropic pollution (such as light at night and noise) is among the most important drivers shaping bird use in urban areas, where bird species are mainly attracted by urban greenery. In this study, we collected data on 127 breeding bird species at 1349 point counts distributed along a gradient of urbanization in fourteen different European cities. The main aim was to explore the effects of anthropic pollution and city characteristics, on shaping the avian communities, regarding species' diet composition. The green cover of urban areas increased the number of insectivorous and omnivorous bird species, while slightly decreasing the overall diet heterogeneity of the avian communities. The green heterogeneity-a measure of evenness considering the relative coverage of grass, shrubs and trees-was positively correlated with the richness of granivorous, insectivorous, and omnivorous species, increasing the level of diet heterogeneity in the assemblages. Additionally, the effects of light pollution on avian communities were associated with the species' diet. Overall, light pollution negatively affected insectivorous and omnivorous bird species while not affecting granivorous species. The noise pollution, in contrast, was not significantly associated with changes in species assemblages. Our results offer some tips to urban planners, managers, and ecologists, in the challenge of producing more eco-friendly cities for the future.