Maria Raquel Kanieski

Exposure to allergenic pollen is a major public health concern, as it is a key trigger for respiratory allergies, including seasonal allergic rhinitis, which affects approximately 20 % of the global population. Monitoring airborne pollen is essential for prevention and clinical management, yet traditional identification methods, such as light microscopy, are time-consuming and often limited to genus- or family-level resolution. Here, we present a high-throughput approach combining flow cytometry with machine learning to identify pollen from urban environments. We collected a reference database of pollen from 97 species across 34 genera, representing the dominant allergenic trees and other common airborne taxa in Montreal, Canada. Using flow cytometry, we measured particle size, granularity, and fluorescence intensity across multiple excitation and emission channels, and applied a Random Forest classifier to distinguish pollen taxa. At the species level, the model achieved a mean F₁-score of 0.76, while genus-level classification reached 0.90, with misclassifications largely occurring among closely related species. Granularity and fluorescence parameters from the violet and blue lasers were the most distinctive features. Our results demonstrate that flow cytometry combined with machine learning provides an efficient, scalable alternative to microscopy, with potential for large-scale urban pollen monitoring.

Exposure to allergenic pollen is a major public health concern, as it is a key trigger for respiratory allergies, including seasonal allergic rhinitis, which affects approximately 20 % of the global population. Monitoring airborne pollen is essential for prevention and clinical management, yet traditional identification methods, such as light microscopy, are time-consuming and often limited to genus- or family-level resolution. Here, we present a high-throughput approach combining flow cytometry with machine learning to identify pollen from urban environments. We collected a reference database of pollen from 97 species across 34 genera, representing the dominant allergenic trees and other common airborne taxa in Montreal, Canada. Using flow cytometry, we measured particle size, granularity, and fluorescence intensity across multiple excitation and emission channels, and applied a Random Forest classifier to distinguish pollen taxa. At the species level, the model achieved a mean F1-score of 0.76, while genus-level classification reached 0.90, with misclassifications largely occurring among closely related species. Granularity and fluorescence parameters from the violet and blue lasers were the most distinctive features. Our results demonstrate that flow cytometry combined with machine learning provides an efficient, scalable alternative to microscopy, with potential for large-scale urban pollen monitoring.

Exposure to allergenic pollen is a major public health concern, as it is a key trigger for respiratory allergies, including seasonal allergic rhinitis, which affects approximately 20 % of the global population. Monitoring airborne pollen is essential for prevention and clinical management, yet traditional identification methods, such as light microscopy, are time-consuming and often limited to genus- or family-level resolution. Here, we present a high-throughput approach combining flow cytometry with machine learning to identify pollen from urban environments. We collected a reference database of pollen from 97 species across 34 genera, representing the dominant allergenic trees and other common airborne taxa in Montreal, Canada. Using flow cytometry, we measured particle size, granularity, and fluorescence intensity across multiple excitation and emission channels, and applied a Random Forest classifier to distinguish pollen taxa. At the species level, the model achieved a mean F1-score of 0.76, while genus-level classification reached 0.90, with misclassifications largely occurring among closely related species. Granularity and fluorescence parameters from the violet and blue lasers were the most distinctive features. Our results demonstrate that flow cytometry combined with machine learning provides an efficient, scalable alternative to microscopy, with potential for large-scale urban pollen monitoring.

Exposure to allergenic pollen is a major public health concern, as it is a key trigger for respiratory allergies, including seasonal allergic rhinitis, which affects approximately 20 % of the global population. Monitoring airborne pollen is essential for prevention and clinical management, yet traditional identification methods, such as light microscopy, are time-consuming and often limited to genus- or family-level resolution. Here, we present a high-throughput approach combining flow cytometry with machine learning to identify pollen from urban environments. We collected a reference database of pollen from 97 species across 34 genera, representing the dominant allergenic trees and other common airborne taxa in Montreal, Canada. Using flow cytometry, we measured particle size, granularity, and fluorescence intensity across multiple excitation and emission channels, and applied a Random Forest classifier to distinguish pollen taxa. At the species level, the model achieved a mean F1-score of 0.76, while genus-level classification reached 0.90, with misclassifications largely occurring among closely related species. Granularity and fluorescence parameters from the violet and blue lasers were the most distinctive features. Our results demonstrate that flow cytometry combined with machine learning provides an efficient, scalable alternative to microscopy, with potential for large-scale urban pollen monitoring.

Exposure to allergenic pollen is a major public health concern, as it is a key trigger for respiratory allergies, including seasonal allergic rhinitis, which affects approximately 20 % of the global population. Monitoring airborne pollen is essential for prevention and clinical management, yet traditional identification methods, such as light microscopy, are time-consuming and often limited to genus- or family-level resolution. Here, we present a high-throughput approach combining flow cytometry with machine learning to identify pollen from urban environments. We collected a reference database of pollen from 97 species across 34 genera, representing the dominant allergenic trees and other common airborne taxa in Montreal, Canada. Using flow cytometry, we measured particle size, granularity, and fluorescence intensity across multiple excitation and emission channels, and applied a Random Forest classifier to distinguish pollen taxa. At the species level, the model achieved a mean F1-score of 0.76, while genus-level classification reached 0.90, with misclassifications largely occurring among closely related species. Granularity and fluorescence parameters from the violet and blue lasers were the most distinctive features. Our results demonstrate that flow cytometry combined with machine learning provides an efficient, scalable alternative to microscopy, with potential for large-scale urban pollen monitoring.

Every forest carries in its structure traces of exploitation and resilience, but how interpret this narrative? This study investigated whether different disturbance histories lead to distinct functional strategies in fragments of Mixed Ombrophilous Forest. We analyzed specific leaf area, leaf dry matter content, wood density, diameter at breast height, biomass, and carbon using community-weighted means, multivariate and statistical analyses. Conserved fragments showed significantly higher diameter, biomass, and carbon stocks (t-test, p < 0.05), indicating conservative strategies, whereas selectively logged fragments exhibited higher wood density and leaf dry matter content (t-test, p < 0.05), revealing unexpected patterns. Pearson correlations were stronger and more homogeneous in conserved fragments, while selectively logged forests displayed more variable association patterns among functional attributes. These findings suggest that selectively logged forests may follow distinct functional trajectories shaped by logging legacies and regeneration processes, highlighting the value of functional traits for understanding forest recovery and ecosystem dynamics.

Planting the proper trees can mitigate and adapt cities to the adverse impacts of climate change, in particular through increasing diversity. Despite being located in one of the most biodiverse countries in the world, low diversity and overabundance of exotic species in Brazilian urban forests are common. Our objective with this study is to group species based on their functional traits to provide a tool to help planners create more resistant urban forests in cities in the domain of the Dense Ombrophilous Forest (DOF), Atlantic Forest, in southern Brazil, as well as incentivize the use of DOF species based on their functional traits. We classified 77 native trees of the DOF in Santa Catarina state, southern Brazil, with potential use in urban environments into functional groups according to the following traits: seed mass, wood density, average height, and leaf persistence, using a hierarchical cluster analysis. The different groups formed represent species with different ecological strategies. Combining species from different groups will help create more resistant urban forests in cities of the DOF domain. We hope the method will also help to increase the use of native species from DOF. Furthermore, the method can be adapted to urban forests of different domains in Brazil and other countries.

Enhancing tree diversity is a key strategy for building climate-resilient cities. In Brazil, however, urban forests, except in urban forest remnants, are often dominated by a few species, mostly exotic, despite the country’s immense native biodiversity. This study presents a rapid and simple method to support planners in designing functionally diverse urban forests using native trees. We grouped 77 native species from the Dense Ombrophilous Forest (DOF) in southern Brazil into six functional groups based on four traits: seed mass, wood density, height, and leaf persistence. These groups represent distinct ecological strategies and can guide tree selection tailored to local conditions. By offering a low-cost, replicable tool grounded in functional ecology, this approach promotes the use of native species, enhances ecosystem resilience, and fosters a stronger local ecological identity. The method can be adapted for use in other formations and biomes and is particularly relevant for small and mid-sized cities that usually lack the technical capacity for ecological planning.

Passive restoration is widely adopted in restoration areas previously used by pine plantations. However, little is known about which environmental factors influence the initial natural regeneration in these areas. Our objective was to relate environmental variables to the floristic-structural patterns of early natural regeneration in areas under passive restoration post-harvest of Pinus taeda. We related natural regeneration data with vegetation and soil variables obtained in 19 plots (100 m²). The ordering of the floristic-structural data of natural regeneration was carried out by NMDS (Non-Metric Multidimensional Scaling) analysis, and the significant variables (p < 0.05) were plotted a posteriori. The results indicate that soil conditions are adequate for natural regeneration development in the study areas. The variables pH, clay, and canopy cover showed a relationship with the floristic-structural composition of the initial natural regeneration. These findings indicated a gradient of texture and acidity and a successional gradient among the sites. In general, tree species show an association with sites containing high canopy cover and soils of greater acidity and clay content. On the other hand, pioneer-shrub species predominate in sites with low canopy cover and soils that are less clayey and acidic.