Pollen Species Research Articles

Airborne pollens in Wuhan urban area and its correlation with hospital visits of allergic rhinitis patients from 2017 to 2018

Objective: To investigate the species and concentrations of airborne pollens in Wuhan urban area and their correlation with the number of visits of allergic rhinitis patients. Methods: Retrospective analysis of pollen dispersal characteristics and the number of patients with allergic rhinitis presenting to Tongji Hospital of Huazhong University of Science and Technology in Wuhan city from October 2017 to September 2018, as well as pollen allergen testing results of patients with allergic rhinitis presenting to the Department of Allergy during the same period. Pollen data was collected by a 1-year air sampling conducted in Wuhan City during the same period using the volumetric method. The samples were examined microscopically to identify airborne pollen species and counted, and the concentrations of various pollens were calculated. Information on patients with allergic rhinitis who came to the hospital during the same period was collected, and the correlation between pollen concentration and the number of patient visits was statistically analyzed using Pearson correlation analysis. Results: A total of 35 types of airborne pollen were collected from October 2017 to September 2018. The dominant pollens in spring were Moraceae (68.46%, 1 042/1 522), Pendula (12.22%, 186/1 522) and Cupressaceae (2.30%, 35/1 522); in summer and autumn, the dominant pollens were Artemisia (3.81%, 58/1 522), Humulus (4.01%, 61/1 522) and Ambrosia (0.59%, 9/1 522). The peak number of visits for allergic rhinitis patients occurred in March-April and July-September, both exceeding 2 200 visits and reaching a maximum of 2 545 visits. There was a very weak correlation between the number of visits and the total pollen concentration (r=0.17, P=0.001). The average monthly pollen skin prick test positive rate of patients with allergic rhinitis was highest in March-May, exceeding 40% with a maximum of 45.73%, and there was a significant correlation between the positive pollen skin prick test positive rate and the average monthly pollen concentration (r=0.62, P=0.031). Conclusions: Pollen species and concentrations fluctuated continuously with time in Wuhan urban area, with peak pollen dispersal in spring from March to April and in autumn from August to September. The number of visits to patients with allergic rhinitis and the positive pollen skin prick test positive rate increased accordingly during the peak pollen concentration periods.

A new species of Eusphalerum Kraatz, 1857 from the Eocene Baltic amber (Coleoptera, Staphylinidae, Omaliinae).

A new species of the genus Eusphalerum Kraatz, 1857 (Omaliinae) from the Eocene Baltic amber is described and illustrated: Eu. bukejsi Shavrin, sp.n. The specimen was studied using X-ray micro-computed tomography (μCT), to obtain images of the habitus, body parts and aedeagus. Based on the shape of the aedeagus, it was compared with East Asian species of pollens and japonicum groups.

Semantic segmentation of pollen grain images generated from scattering patterns via deep learning

Pollen can lead to individuals suffering from allergic rhinitis, with a person’s vulnerability being dependent on the species and the amount of pollen. Therefore, the ability to precisely quantify both the number and species of pollen grains in a certain volume would be invaluable. Lensless sensing offers the ability to classify pollen grains from their scattering patterns, with the use of very few optical components. However, since there could be 1000 s of species of pollen one may wish to identify, in order to avoid having to collect scattering patterns from all species (and mixtures of species) we propose using two separate neural networks. The first neural network generates a microscope equivalent image from the scattering pattern, having been trained on a limited number of experimentally collected pollen scattering data. The second neural network segments the generated image into its components, having been trained on microscope images, allowing pollen species identification (potentially allowing the use of existing databases of microscope images to expand range of species identified by the segmentation network). In addition to classification, segmentation also provides richer information, such as the number of pixels and therefore the potential size of particular pollen grains. Specifically, we demonstrate the identification and projected area of pollen grain species, via semantic image segmentation, in generated microscope images of pollen grains, containing mixtures and species that were previously unseen by the image generation network. The microscope images of mixtures of pollen grains, used for training the segmentation neural network, were created by fusing microscope images of isolated pollen grains together while the trained neural network was tested on microscope images of actual mixtures. The ability to carry out pollen species identification from reconstructed images without needing to train the identification network on the scattering patterns is useful for the real-world implementation of such technology.

Effects of landscape structure and climate seasonality on pollen intake by honeybees in Neotropical highland agroecosystems

AbstractPollen is the main protein source for honeybee brood and so colony development relies heavily on the availability of pollen in the environment. Intensification of agriculture and climate seasonality are known to alter honeybee pollen intake in temperate regions through changes in resource availability; however, little is known about how honeybees respond to such environmental factors in tropical regions. Pollen collected by honeybees was sampled from apiaries in a Neotropical highland region of Colombia. Pollen species were identified and the effects of landscape diversity, forest area and mean monthly precipitation on the pollen intake by honeybees were evaluated for all pollen species together and pollen species segregated according to forest and anthropic areas. Honeybees were found to be much more associated with anthropic than forest pollen species regardless of landscape structure or precipitation. However, pollen intake from all species and forest species responded positively to landscape diversity and forest area. Precipitation was found not to be related to the overall amount and overall richness of pollen collected by honeybees. Nonetheless, overall pollen diversity was negatively affected by precipitation in less diverse landscapes, whereas anthropic pollen diversity was negatively affected by precipitation in more forested landscapes.

Identifying the spatial footprint of pollen distributions using the Geoforensic Interdiction (GOFIND) model

Geoforensics leverages various spatial analytical, biological, ecological, and geological techniques to improve criminal investigations, including efforts to thwart terrorism, evaluate humanitarian crises or identify the origin of fraudulent goods. In addition to investigating what took place, geoforensic efforts often focus on determining where and when the event occurred. Forensic palynology is a promising subfield of geoforensics that uses pollen to link persons or objects to particular places and times. This promise is driven by pollen's strong utility as a biomarker. It is ubiquitous, durable and demonstrates a relatively predictable distribution in time and space. As a result, pollen often provides important clues for determining the provenance of hard-to-trace items, such as computers, fraudulent goods, digging tools, clothing, and undetonated explosives during geoforensic investigations. One major limitation that has reduced the implementation of forensic palynology is the lack of computational tools that relate pollen species to geographic locations. The purpose of this paper is to introduce a robust geocomputational framework that uses pollen sample composition to identify the relative likelihood of potential origin locations for improving geoforensic efforts. Using USDA CropScape data for the state of Texas, our results suggest that this framework allowed for multiple potential origin sites to be identified simultaneously, with solution properties that were better than a random process and offer a possible alternative to the single site joint probability approaches commonly used in the field of forensic palynology.

Species of Fungi and Pollen in the PM1 and the Inhalable Fraction of Indoor Air in Homes

Airborne microbial fragments in the PM1 fraction (particles with aerodynamic diameters less than 1 µm) are a cause for concern as they may potentially deposit in the alveoli of the human airways. This study aimed to use qPCR to identify and quantify 24 different species or groups of genera in the PM1 and the inhalation fraction (particles that may enter the mouth or nose during breathing) of indoor air and to relate this to what has previously been found for each species. Results showed that eight fungal species, and Aspergillus/Penicillium/Paecilomyces variotii, as well as Alnus/Corylus and actinobacteria belonging to the Streptomyces genus were detected both in the PM1 and the inhalable fraction. Five fungal species were only detected in the inhalable fraction. A significant effect of season was found on the fungal composition in the PM1 (p = 0.001) and the inhalable (p = 0.017) fraction. This study demonstrated that it is possible to use qPCR to identify and quantify different microbes in the PM1 fraction, and it has improved our understanding of the qualitative and quantitative relationship between the PM1 and the inhalable microbial particles in indoor air. Combined with the literature review it also shows a large variation within and between species in the share of fungi which is present as fragments.

Overlap of Ecological Niche Breadth of Euglossa cordata and Eulaema nigrita (Hymenoptera, Apidae, Euglossini) Accessed by Pollen Loads and Species Distribution Modeling

Urban areas can serve as biodiversity refuges for pollinators because of the high diversity of available floral and nesting resources. However, it remains unclear what plant species commonly used for urban landscaping provide floral resources that pollinators actively use. Here, we integrate data from the pollen and species distribution models of two abundant euglossine bees-the large-bodied Eulaema nigrita (Lepeletier, 1841) and the small-bodied Euglossa cordata (Linnaeus, 1758)-in urban areas to investigate their overlap in diet breadth and distribution. We hypothesized that because bees with larger body sizes tend to have larger foraging areas, large-bodied bees would have a wider diet breath than small-bodied bees. Contrary to our hypothesis, we found that Eg. cordata has a wider diet breadth than El. nigrita with the former species showing higher diversity of pollen types collected (per pollen load and on average across pollen loads). Pollen grains from Solanum paniculatum and Tradescantia zebrina represented 63% of the diet of Eg. cordata, whereas pollen from S. paniculatum and Psidium guajava represented 87% of the diet of El. nigrita. After overlaying the distribution of both bee species and the three most important pollen resources, the distribution models revealed that these three plant species can co-occur with both euglossine bees throughout a large portion of eastern Brazil near the coast. Thus, we conclude S. paniculatum, T. zebrina, and P. guajava should be considered key plants for the maintenance of these two urban euglossine bee species. The results of this study provide important information for urban landscaping programs that aim to protect and preserve pollinators.

Characterization of pollen profile of Apis mellifera L. in arid region of Pakistan

Honeybees rely exclusively on pollen and nectar-producing plants for strengthening their colonies and manufacturing honey. Little is known about the indigenous melliferous flora of arid zones of Khyber Pakhtunkhwa (KPK) which is crucial for honey production and how different pollen assessment techniques effect the identification of indigenous melliferous pollen flora. Visual survey and loads ensnaring through pollen traps were used to identify the botanical profile of melliferous pollen flora of Dera Ismail Khan (DIKhan), Khyber Pakhtunkhwa. The test time extended for two consecutive years 2018 and 2019. The study revealed 56 plant species as pollen flora with 18 significant pollen producing species in visual survey technique while 8 species as predominant flora in pollen trapping technique. The major pollen species found common in both the techniques were Brassica napus L., Brassica campestris L., Trifolium alaxandrinum L., Zea mays L., Acacia modesta L., Citrus aurantium L., Euclyptus spp., and Morus alba L. Pollen interception and palynological analysis of pollen were found to be more reliable techniques as compared to focal observations. More than fifty % differences were found by comparing the results of the visual survey and pollen trapping technique in major flora of DIKhan. Based on the availability, utility status and flowering duration of apiphilic flora, mid-February to mid-May was found to be a significant pollen flow period in the study area. Maximum benefit can be taken in this period through trapping ample amount of pollen and stored for using in artificial diets, selling and feeding bees during dearth period.

Visual Machine Learning and Artificial Intelligence Application in Aeroallergen Identification During Spring, Summer, and Fall Pollen Season

Continued validation of automated pollen identification (API) is needed. We performed a side-by-side device comparison of pollen counts identified using a Burkard sampler (BS) and an Automate Pollen Sampler (APS) by Pollen SenseTM. The designated gold standard device for comparison, the BS, was co-located with the APS in Waterloo, Iowa. Pollen collection, processing, and identification were performed following NAB requirements. The APS was provided by Pollen SenseTM. The APS collects particulate matter volumetrically from ambient air, automatically images the particulates, and uses a convolutional neural network to identify the individual pollen species. For this first iteration, 57 days were compared in the fall 2019 and spring-summer 2020. The top 10 pollen reported to the NAB in the upper Midwest were chosen for reporting. Accuracy is determined as the number of days a species was identified as present or absent in the APS that matches the Burkard, divided by the number of counting days x 100. Accuracy: ragweed 96, mulberry 50, juniper 8, oak 46, nettle 0, grass 65, ash 58, elm 8, pine 65, birch 23. Current accuracy of APS needs continued improvement. Varying correctness in pollen species identification rely on several factors including whether the APS was trained for that pollen species, morphology complexity, and amount of debris in the background. Identification and enumeration using visual machine learning methods in artificial intelligence is a process that improves automatically through experience. It requires multiple iterations of repetitive learning to achieve desired accuracy.

Evaluation of sampling effort required to assess pollen species richness on pollinators using rarefaction.

PremiseUnderstanding the flower visitation history of individual pollinators is key in the study of pollination networks, but direct tracking is labor intensive and, more important, does not capture information about the previous interactions of an individual. Therefore, a protocol to detect most of the pollen species on the body surfaces of an individual pollinator could elucidate its flower visitation history.Methods and ResultsUnder a microscope, we observed 6.0‐µL droplets from a sample solution (1.0 or 3.0 mL) containing pollen grains collected from individuals of six major pollinator functional groups. To clarify how many droplets need to be observed to detect all pollen species within the solution, we examined up to 10 droplets collected from each individual insect. Sample‐based rarefaction curve analyses of the data showed that we could detect ~90% of the pollen species and the plant–pollinator links in the networks by observing six droplets.ConclusionsThe rarefaction curve analysis for pollen‐on‐pollinator studies is a useful preliminary step for minimizing the time and labor required while maximizing the data on the flower visitation history of each individual pollinator and revealing any hidden flower–pollinator interactions.

The Strong and the Stronger: The Effects of Increasing Ozone and Nitrogen Dioxide Concentrations in Pollen of Different Forest Species

The knowledge of pollen sensitivity and tolerance to stress factors such as air pollution is important for forest sustainability, ensuring the most efficient production with the highest benefits and lowest resource losses. This study intended to evaluate the influence of common air pollutants in four forest trees species, Betula pendula Roth, Corylus avellana L., Acer negundo L. and Quercus robur L., through a comparative analysis at the same experimental conditions. We aimed to investigate the effect that may occur in pollen fertility, protein content, oxidative stress and wall composition after exposure in vitro to ozone and nitrogen dioxide at concentration levels for vegetation protection in Europe. Our results suggest changes in pollen viability, protein content and differential sensitivity related to ROS synthesis, NADPH oxidase activity, as well as in wall composition. The results indicate that NO2 exposure affected more the pollen species studied mostly at the highest concentration exposure. As for ozone, there were less significant differences between samples; however, a different behavior occurs in O3 expositions, where the most influence happens at the legal limit for vegetation protection in Europe. Our study showed that significant pollen functions could be compromised even at common air pollutant’s concentrations.

Comparative proteomics of common allergenic tree pollens of birch, alder, and hazel.

BackgroundIn addition to known allergens, other proteins in pollen can aid the development of an immune response in allergic individuals. The contribution of the “unknown” protein allergens is apparent in phylogenetically related species where, despite of high homology of the lead allergens, the degree of allergenic potential can vary greatly. The aim of this study was to identify other potentially allergenic proteins in pollen of three common and highly related allergenic tree species: birch (Betula pendula), hazel (Corylus avellana) and alder (Alnus glutinosa).MethodsFor that purpose, we carried out a comprehensive, comparative proteomic screening of the pollen from the three species. In order to maximize protein recovery and coverage, different protein extraction and isolation strategies during sample preparation were employed.ResultsAs a result, we report 2500–3000 identified proteins per each of the pollen species. Identified proteins were further used for a number of annotation steps, providing insight into differential distribution of peptidases, peptidase inhibitors and other potential allergenic proteins across the three species. Moreover, we carried out functional enrichment analyses that, interestingly, corroborated high species similarity in spite of their relatively distinct protein profiles.ConclusionWe provide to our knowledge first insight into proteomes of two very important allergenic pollen types, hazel and alder, where not even transcriptomics data are available, and compared them to birch. Datasets from this study can be readily used as protein databases and as such serve as basis for further functional studies.

Many small rather than few large sources identified in long-term bee pollen diets in agroecosystems

Bees provide essential ecosystem services such as crop pollination, but perennial colonies of social species require year-round access to floral resources, especially in resource-poor agricultural landscapes. We investigated pollen resources used by a social bee (Tetragonula carbonaria, Meliponini) in forests and orchards of subtropical Australia. Pollen DNA metabarcoding with the markers rbcL and ITS2 was used to identify hive pollens from 57 colonies collected at seven sites each season over two years. We identified 341 botanical sources of hive pollens from 37 orders, 72 families, 218 genera and 302 species. Interestingly, introduced species (e.g. Ageratum spp. and Raphanus spp.) and wind-pollinated plants (Poaceae, Cyperaceae) were common pollen sources in both orchards and forests. Orchard colonies used a subset of pollen species used by colonies in forests, with many Myrtaceae (Corymbia, Eucalyptus and Melaleuca spp.), Poaceae, Asteraceae, Fabaceae and Proteaceae species found in both land uses. We found T. carbonaria foraged on “many small” rather than a “few large” pollen sources each season, regardless of land use. This suggests stingless bees aim for diversity in pollen diets. As such, land managers and beekeepers should ensure colonies have access to a variety of floral resources year-round. This may be achieved through targeted planting of key families identified in this study (e.g. Proteaceae, Asteraceae, Myrtaceae, Poaceae, Brassicaceae, Araliaceae, Cannabaceae, Arecaceae, Rubiaceae and Sapindaceae) and / or maintaining weeds while they are flowering in the orchard. Land managers may consider planting in unproductive areas such as riparian zones, edges or between crop rows if space is limited.

Records of Social Wasps (Hymenoptera: Vespidae) from Mongolia

In Mongolia, there are 19 known social wasp species (12 spp. in Vespinae, 7 spp. in Polistinae), for some of which defi nite collection data are needed to confi rm their actual occurrence in this country. New distribution data are provided for seven vespine species and four Polistes species from Mongolia. Taxonomic comments are given to each species when necessary. New information about the nesting habits of some Dolichovespula and Polistes species is presented for the Mongolian populations. Keys to the Mongolian species of Vespinae and Polistinae are presented.

Manila grass (Zoysia matrella) Zoy m 1 allergen may contribute to allergic sensitization in tropical/subtropical regions due to extensive cross-reactivity with other group-1 grass pollen allergens.

Pollen of grasses in Chloridoideae and Panicoideae subfamilies is a major source of grass group-1 allergens in tropical/subtropical areas. Previously, most studies of subtropical grass pollen allergens have focused on Cynodon dactylon (Bermuda grass-Chloridoideae) and Sorghum halepense (Johnson grass-Panicoideae). However, little information is available about allergenicity of pollen from Zoysia matrella (Manila grass or Zoysia grass-Chloridoideae), which is among the most popular turfgrasses in tropical/subtropical areas. This study aimed to investigate the IgE reactivity and cross-reactivity of grass group-1 allergen from Z. matrella. In addition, the clinical relevance of Z. matrella in comparison with other species was assessed. IgE reactivity and cross-reactivity between recombinant proteins of group-1 allergen from Z. matrella (Zoy m 1) and C. dactylon (Cyn d 1) were determined by ELISA and immunoblot assays. Clinical relevance of Z. matrella pollen in Thai atopic patients was assessed using its pollen crude extract for skin-prick test, in comparison with extracts from four other pollen species. The Zoy m 1 had high IgE binding and could interfere with binding to C. dactylon crude extract. In addition, Z. matrella pollen extract elicited positive skin-prick test results comparable to previously reported allergenic species. Group-1 grass pollen allergen was confirmed to be a major allergen from Z. matrella among Thai atopic patients and was officially designated Zoy m 1.0101. Zoy m 1 allergen is a major allergen from Z. matrella that cross-reacts with other group-1 grass pollen allergens in the tropical/subtropical region.

Molecular allergy diagnosis using pollen marker allergens and pollen panallergens: Five patterns seen in multiple test reactions topollen extracts.

Polysensitizations to tree, grass, and weed pollen are found in ~ 20% of pollen-allergic individuals. They are often based on broad IgE cross-reactivities to pollen panallergens belonging to highly conserved protein families: 1. profilins, 2. polcalcins (calcium-binding proteins in pollen), 3. cyclophilins. They represent highly conserved cross-reactive minor allergens present in all pollen species, but also in plant foods and other organisms. Despite being rarely clinically relevant they can hamper allergy diagnostic tests with extracts. In this situation, molecular allergy diagnosis is able to distinguish broad cross-reactivity due to allergen-specific IgE to pollen panallergens (i.e. profilins Bet v 2 or Phl p 12; polcalcins Bet v 4 or Phl p 7; and, in the future, cyclophilins Bet v 7 or Ole e 15) from primary IgE sensitizations to so-called marker allergens represented by important pollen major allergens: Bet v 1 for the birch and beech family (Fagales), Ole e 1 for olive and ash (Oleaceae), Phl p 1 for temperate climate grasses (Poaceae), Art v 1 for mugwort (Artemisia), Amb a 1 for Ambrosia species (Ambrosia). Five typical cases (A – E) with positive skin prick test results to tree, grass, and weed pollen extracts demonstrate typical patterns of IgE sensitization with a variable impact of pollen panallergens: A – profilins, B – polcalcins, C – profilins and polcalcins, D – presumably cyclophilins, E – primary polysensitization to tree, grass, and weed pollen without interference from profilins or polcalcins. Differences between pollen extract-based skin prick test diagnosis and molecular allergen-specific IgE testing are explained using the presented concept. This approach allows to reduce the number of allergen extracts – presuming they are also clinically relevant – for allergen immunotherapy (i.e., only tree and/or grass pollen extracts), particularly in pollen-polysensitized patients.

New pollen genera and species from the Oligocene of northern Spain and a systematic, biostratigraphic and biogeographic re-evaluation of coeval taxa

The deposits of the As Pontes Basin in north-west Spain yield extraordinarily well-preserved microfloral assemblages that have unmatched diversity within the European Oligocene. The comprehensive examination of 300 miospore taxa has led to the discovery of two new fossil pollen genera, Eucommiopsispollenites gen. nov. and Pistaciapollenites gen. nov., as well as seven new species: Eucommiopsispollenites compactus sp. nov., Eucommiopsispollenites gracilis sp. nov., Nymphaeapollenites asponticus sp. nov., Nymphaeapollenites perforatus sp. nov., Pistaciapollenites heteroporatus sp. nov., Punctilongisulcites cryptornatus sp. nov. and Symplocoipollenites diktyotos sp. nov. Moreover, the botanical affinity and geographic distribution of 13 additional species of particular interest are discussed in detail, because these are either recorded for the first time anywhere within the Paleogene (Lythraceaepollenites bavaricus, Myrtaceidites myrtiformis, Punctilongisulcites punctiechinatus, Rhuspollenites ornatus and Vitispollenites tener), in the Iberian Peninsula (Aglaoreidia pristina, Diospyrospollenites ovalis, Pseudospinaepollis pseudospinosus, Reevesiapollis reticulatus and Tricolporopollenites vancampoae), or are studied for the first time using scanning electron microscopy (SEM) (Pentapollenites regulatius, Polycolpites helmstedtensis and Triporopollenites robustus). SEM analysis has enabled the conclusive botanical assignment of two species: Polycolpites helmstedtensis, with affinity to the extant tropical genus Eriandra, and Triporopollenites robustus, related to Myricaceae. In view of these affinities, new combinations are proposed accordingly: Polygalacidites helmstedtensis comb. nov. and Myricipites robustus comb. nov. The results of this study carry significant biostratigraphic and palaeobiogeographic implications at a regional and wider, continental scale. Firstly, the similarities of the As Pontes microflora with other Oligocene assemblages located various thousands of kilometres apart indicates the spread of a relatively homogenous broadleaved evergreen flora in Europe and allows for the recognition of Myricipites robustus and the genus Pentapollenites as key biostratigraphic index taxa across the continent. Secondly, the last occurrence datums of thermophilous taxa that became extinct in the Eocene in more northern latitudes indicates that the Iberian Peninsula functioned as a refuge area for thermic vegetation at least since the early Oligocene. http://zoobank.org/urn:lsid:zoobank.org:pub:16AFE374-A8B4-4CE9-B5B4-31807E2A6962

Syrphidae of Southern Illinois: Diversity, floral associations, and preliminary assessment of their efficacy as pollinators.

Syrphid flies (Diptera: Syrphidae) are a cosmopolitan group of flower-visiting insects, though their diversity and importance as pollinators is understudied and often unappreciated. Data on 1,477 Syrphid occurrences and floral associations from three years of pollinator collection (2017-2019) in the Southern Illinois region of Illinois, United States, are here compiled and analyzed. We collected 69 species in 36 genera off of the flowers of 157 plant species. While a richness of 69 species is greater than most other families of flower-visiting insects in our region, a species accumulation curve and regional species pool estimators suggest that at least 33 species are yet uncollected. In order to further the understanding of Syrphidae as pollinators in the Southern Illinois region, we produced a NMDS ordination of floral associations for the most common syrphid species. The NMDS did not sort syrphid species into discrete ecological guilds, and syrphid floral associations generally fit those predicted by traditional pollination syndromes. We also conducted a preliminary analysis of the pollen-carrying capacity of different syrphid taxa, which found several Eristalis species to carry pollen loads comparable to the European Honey Bee, Apismellifera, and showed significant differences in the pollen-carrying capacity of various syrphid species. Notably, the extremely common genus Toxomerus and other small Syrphinae species carried very little pollen, while large and pilose Eristalinae species carried large pollen loads.

Testing the Raman parameters of pollen spectra in automatic identification

Pollen identification and quantification are used in many fields of application and research has been conducted to attain accurate automatic pollen recognition aiming to reduce the laborious work and subjectivity in human identification. The aim of our study was to evaluate the capacity of Raman parameters of pollen spectra, calculated for only 7 common band intervals in a limited spectral range, to be used as future technique in pollen automatic identification. There were analyzed 15 different pollen species considered to induce allergic reactions. Raman spectra were acquired at an excitation wavelength of 785 nm in a spectral region from 1000 to 1800 cm−1, preprocessed and deconvoluted to determine the Raman parameters: wavenumber, full width at half maximum of the band and integrated intensity. Seven common band intervals of all Raman spectra, in the fingerprint areas 1000–1010, 1300–1460 and 1500–1700 cm−1, were chosen for the classification of the pollen species using SVM (support vector machine). Our results showed that the classification accuracy of all pollen species was 100% in the training step, while in the testing step 14 out of the 15 pollen species were correctly assigned (93.3%), including the discrimination between 5 Poaceae species and between Betula pendula and Corylus avellana. It was also observed that all Raman parameters are important in the classification as well as all wavenumber areas considered. So, our study indicates that the Raman parameters of pollen spectra can be a promising methodology for automatic pollen recognition.

On the Road of Automated Pollen Recognition

Identifying all the pollen species present on earth, and more particularly in a territory, is a major concern for palynologists. This is an arduous task that can be automated using artificial intelligence. Many studies have tried to solve this problem by using machine learning and deep learning. In this paper, we present three pollen recognition approaches: Classification with no examples, recognition with a sufficient number of examples, recognition with un-sufficient number of examples. For each of them, we propose respectively to use Visual Bag of Word and expectation-maximization clustering algorithms, Classification using Local Binary patterns and the Gabor Filter Feature, Local Binary Patterns and Prototypical Networks. We find 77,38% recognition for 10 pollen species rate for the first one, 90.80 % for training with a sufficient number of examples and 80 species, and 20 different pollen species and finally 84,30% for the third approach with one example for training and 20 species.