The use of discriminant analysis and neural networks to forecast the severity of the Poaceae pollen season in a region with a typical Mediterranean climate

Biological particles in the air such as pollen grains can cause environmental problems in the allergic population. Medical studies report that a prior knowledge of pollen season severity can be useful in the management of pollen-related diseases. The aim of this work was to forecast the severity of the Poaceae pollen season by using weather parameters prior to the pollen season. To carry out the study a historical database of 21 years of pollen and meteorological data was used. First, the years were grouped into classes by using cluster analysis. As a result of the grouping, the 21 years were divided into 3 classes according to their potential allergenic load. Pre-season meteorological variables were used, as well as a series of characteristics related to the pollen season. When considering pre-season meteorological variables, winter variables were separated from early spring variables due to the nature of the Mediterranean climate. Second, a neural network model as well as a discriminant linear analysis were built to forecast Poaceae pollen season severity, according to the three classes previously defined. The neural network yielded better results than linear models. In conclusion, neural network models could have a high applicability in the area of prevention, as the allergenic potential of a year can be determined with a high degree of reliability, based on a series of meteorological values accumulated prior to the pollen season.     

Pollen grains ofAmbrosia in the air of Sofia,Bulgaria

Up to now,Ambrosia pollen has not been considered a cause of allergy symptoms in Bulgaria. However, the results of pollen monitoring suggest that such allergies will become inevitable in the near future. This conclusion is supported by allergometric data on 120 patients.     

Pollen grains ofAmbrosia in the air of Sofia,Bulgaria

Up to now,Ambrosia pollen has not been considered a cause of allergy symptoms in Bulgaria. However, the results of pollen monitoring suggest that such allergies will become inevitable in the near future. This conclusion is supported by allergometric data on 120 patients.     

Predicting the intensity of the birch pollen season

We present a model for the prediction of the magnitude ofBetula flowering and pollen dispersal which may be used in the management of birch pollinosis and in the planning of clinical trials. The pollen sum during the flowering season is regressed on the temperature sum from May 1st to July 20th during the initiation year, the pollen sum of the initiation year, and the temperature sum during the main pollen season in the flowering year. We suggest that the fluctuating flowering pattern inBetula alba-species is primarily determined by the availability of assimilation products during inflorescence initiation and development during the spring one year before anthesis. When inflorescences, which are initiated during the previous year, elongate in the beginning of anthesis, they act as strong sinks to stored carbohydrates, and thus compete with developing leaves and shoots. The result is an initially reduced photosynthetic capacity in years with intense flowering, and a limited potential for the initiation of new inflorescences for the following year. The ambient temperature during catkin initiation affects assimilation efficiency and is a determinant of about equal importance to flowering intensity as is the magnitude of the flowering in the initiation year. The amount of pollen dispersed is also dependent on the weather during anthesis, which is not possible to predict until about one month in advance. The two other independent variables are available during the previous summer, making it possible to give a sufficiently valid prediction to allergologists about the magnitude of the next birch pollen season, according to its botanical determinants. We suggest that the varying reproductive output inBetula alba should not be described as true masting. A more parsimonious explanation to the flowering pattern is that an individual continually maximizes reproductive effort, according to what is possible, but that reproduction is often constrained by the environment.     

Predicting the intensity of the birch pollen season

We present a model for the prediction of the magnitude ofBetula flowering and pollen dispersal which may be used in the management of birch pollinosis and in the planning of clinical trials. The pollen sum during the flowering season is regressed on the temperature sum from May 1st to July 20th during the initiation year, the pollen sum of the initiation year, and the temperature sum during the main pollen season in the flowering year. We suggest that the fluctuating flowering pattern inBetula alba-species is primarily determined by the availability of assimilation products during inflorescence initiation and development during the spring one year before anthesis. When inflorescences, which are initiated during the previous year, elongate in the beginning of anthesis, they act as strong sinks to stored carbohydrates, and thus compete with developing leaves and shoots. The result is an initially reduced photosynthetic capacity in years with intense flowering, and a limited potential for the initiation of new inflorescences for the following year. The ambient temperature during catkin initiation affects assimilation efficiency and is a determinant of about equal importance to flowering intensity as is the magnitude of the flowering in the initiation year. The amount of pollen dispersed is also dependent on the weather during anthesis, which is not possible to predict until about one month in advance. The two other independent variables are available during the previous summer, making it possible to give a sufficiently valid prediction to allergologists about the magnitude of the next birch pollen season, according to its botanical determinants. We suggest that the varying reproductive output inBetula alba should not be described as true masting. A more parsimonious explanation to the flowering pattern is that an individual continually maximizes reproductive effort, according to what is possible, but that reproduction is often constrained by the environment.     

Study onPlatanus hispanica Miller pollen content in the air of Seville, southern Spain

The work was carried out using a Cour trap that sampled the air of the city for 8 consecutive years (1987–1994). The pollen ofPlatanus hispanica is the fourth most abundant in the air of Seville (a mean of 11.05% of the total pollen collected). The variation throughout the years in the sum of weekly concentrations ofPlatanus hispanica pollen presents a certain biennial rhythm, in which years of high and low collection of pollen alternate. The starting day of the main pollination period (MPP) is negatively related with the mean of the mean temperatures for February (r=0.73,r ²=0.53,P=0.0398) and is earlier (at the beginning of March) when the mean temperature for February is high, and vice versa. The pattern of pollen variation inPlatanus hispanica remains constant through the years—pollen appears abruptly in high weekly concentrations (> 150 grains/m³) in March (sporadically at the beginning of April), with a week of maximum pollen emission (WMPE) in which more than 50% of the annual pollen is collected (in 6 of the 8 years), and a main pollination period (MPP) of 2 or 3 weeks (except in 1989 when it was 5 weeks). In every year (except 1989), weekly mean temperatures increased during the MPP, the duration of which depends on mean temperature and mean rainfall: mean temperatures > 16°C and absence of rainfall shorten the MPP, while lower temperatures and presence of rainfall lengthen it. The meteorological conditions most often found during the WMPE are mean temperatures > 15°C and rainfall absent or almost so.     

The relationship between the flowering phenophase and airborne pollen of Betula in galicia (N.W. spain)

The aim of this work was to investigate thephenological behaviour of Betula in Galicia, NW Spain, and to examine therelationship between the Betula pollencurves and the flowering phenophase. Threetrees were chosen from the each of ninepopulations of Betula located atdifferent altitudes and phytogeographicpositions. Phenological observations of theflowering periods of Betula were made ineach of them. Environmental factors such asfrequency of mist, latitudinal and topographicposition, proximity of the ocean, degree ofsolar exposure, and altitude result inphenological differences between theinvestigated populations. The correlationbetween the Pollinic Production Index of Betula pollen in Galicia and theaerobiological data of the seven monitoringstations showed that the period in which thehighest concentrations were registered wasalmost synchronous with the flowering times atmost of the phenological stations studied.Other factors such as transport and reflotationshould also be taken into account to provide anadequate interpretation of the aerobiologicaldata of Betula pollen in the atmosphere.     

Short-term prediction of Betula airborne pollen concentration in Vigo (NW Spain) using logistic additive models and partially linear models

Betula pollen is a common cause of pollinosis in localities in NW Spain and between 13% and 60% of individuals who are immunosensitive to pollen grains respond positively to its allergens. It is important in the case of all such people to be able to predict pollen concentrations in advance. We therefore undertook an aerobiological study in the city of Vigo (Pontevedra, Spain) from 1995 to 2001, using a Hirst active-impact pollen trap (VPPS 2000) situated in the city centre. Vigo presents a temperate maritime climate with a mean annual temperature of 14.9 °C and 1,412 mm annual total precipitation. This paper analyses two ways of quantifying the prediction of pollen concentration: first by means of a generalized additive regression model with the object of predicting whether the series of interest exceeds a certain threshold; second using a partially linear model to obtain specific prediction values for pollen grains. Both models use a self-explicative part and another formed by exogenous meteorological factors. The models were tested with data from 2001 (year in which the total precipitation registered was almost twice the climatological average overall during the flowering period), which were not used in formulating the models. A highly satisfactory classification and good forecasting results were achieved with the first and second approaches respectively. The estimated line taking into account temperature and a calm S–SW wind, corresponds to the real line recorded during 2001, which gives us an idea of the proposed models validity.     

Regional variations in the pollen season of Betula in Burgundy: two models for predicting the start of the pollination

Today, 10% of the French people suffer frompollen allergy, a world-wide rising ailmentsince the end of the last century.To prevent the appearance of the symptoms, physicians need precise and early information on the pollination date of the main allergenic species. This wascarried out in Burgundy on birch (Betula), which is responsible for frequent pollinoses inMarch–April. Two of the tested methods were both successful and complementary: the sum oftemperatures and the multiple regression. The first one is very precise but only allows forecastingfive days in advance: this is enough for people suffering from rhinitis or conjunctivitis, who muststart a preventive treatment one or two days before pollination is effective. But for asthmaticpatients, the treatment must one or two weeks in advance; the second method, although alittle less precise, is interesting becauseforecasting can be done at least two weeks inadvance. The results, from 1995 to 1998, were used prospectively in 1999.     

Cupressaceae pollen in the atmosphere of Ascoli Piceno (Central Italy) and sensitization of allergic subjects

The aim of this work is to study the incidence of pollinosis in the Health District of Ascoli Piceno, Central Italy (U.S.L.24), this being an underestimated pathology from the clinical point of view and also as a result of the recent introduction of this taxa in the National Aeroallergen Network. Since 1990, 5055 patients of both sexes with respiratory symptomatology of suspected IgE mediated aetiology have been examined in our Centre with the Skin Prick Test (SPT) using allergen panels including Cypress; 171 (3.38%) patients were found to be positive to this allergen. These results show that the subjects with symptoms in the period January–March in most cases have a sensitization toCupressus pollen and new studies will evaluate the possibility of specific immunotherapy.     

Ambrosia pollen in the air of Lublin, Poland

Studies on Ambrosia pollen concentrations were carried out in Lublin in the period 1995–2004. The effects of a number of meteorological factors were analysed. In the first period of the study, the gravimetric method was used (1995–1999), while in the second period, the volumetric method was applied. The results show an increasing trend in the amount of airborne pollen. The Ambrosia pollen season in Lublin lasts from August to October. Over a period of 5 years, the highest number of pollen grains was recorded in September (53%), followed by August (44%) and October (3%). There were wide variations in annual totals. The annual total pollen counts was 167–1180 grains, with the peak value in 2002. Maximum daily pollen concentrations (56–312 pollen grains m⁻³) were recorded in the first half of August and in the first half of September. On the days when high Ambrosia pollen concentrations occurred, the temperature was above 21°C and the winds were mainly from the southeast, south and east. Maximum intradiurnal concentrations of pollen grains occurred in the afternoon hours. These results indicate, to some degree, that Ambrosia pollen is transported for long distances before descent.     

The emission of mugwort pollen (Artemisia vulgaris L.) and its flight in the air

Summary Mugwort pollen flight and pollen emission are characterized by a distinct biological rhythm with emission in the early morning hours (6:00–11:00 a.m. CEST). Pollen traps at roof level monitor the daily and yearly variation of mugwort pollen flight later and in lower quantities than those installed at ground level because of the different daily variation of convection at various air levels. The daily variation of mugwort pollen flight can be described by mathematical functions. Because of these significant differences between the catch at different levels, the supposed high allergic potential of mugwort has to be discussed.     

Heat requirement for the onset of the Olea europaea L. pollen season in several sites in Andalusia and the effect of the expected future climate change

Olives are one of the largest crops in the Mediterranean region, especially in Andalusia, in southern Spain. A thermal model has been developed for forecasting the start of the olive tree pollen season at five localities in Andalusia: Cordoba, Priego, Jaen, Granada and Malaga using airborne pollen and meteorological data from 1982 to 2001. Threshold temperatures varied between 5°C and 12.5°C depending on bio-geographical characteristics. The external validity of the results was tested using the data for the year 2002 as an independent variable and it confirmed the models accuracy with only a few days difference from predicted values. All the localities had increasingly earlier start dates during the study period. This could confirm that olive flower phenology can be considered as a sensitive indicator of the effects of climate fluctuations in the Mediterranean area. The theoretical impact of the predicted climatic warming on the olives flowering phenology at the end of the century is also proposed by applying Regional Climate Model data. A general advance, from 1 to 3 weeks could be expected, although this advance will be more pronounced in mid-altitude inland areas.     

The interrelationship between the accumulation of lipids,protein and the level of acyl carrier protein during the development of Brassica napus L. pollen

Lipid accumulation during pollen and tapetal development was studied using cryostat sections of unfixed anthers from Brassica napus (rapeseed). Diamidino-2-henylindole (DAPI), a DNA fluorochrome, was used to stain the pollen nuclei in order to identify ten stages of pollen development in Brassica. Storage lipids (i.e. triacylglycerides) were stained using the fluorochrome Nile red. Pollen coat lipids are formed in tapetal plastids between the mid-vacuolate and early maturation pollen stages. The pollen coat components, including lipids and a proportion of the proteins, are derived from the remnants of the tapetum, after its rupture, during the second pollen mitosis. Quantitative microfluorometric analyses demonstrated four phases of lipid body accumulation or depletion in the developing pollen cytoplasm. The majority of storage lipids found in the cytoplasm of the mature pollen grain accumulated during the late vacuolate and early maturation stages when the pollen is bicellular. The level of acyl carrier protein, a protein integrally involved in lipid synthesis, was also found to be maximal in the developing pollen during the bicellular pollen stages of development. This coincided with the most active period of lipid accumulation. These data could indicate that the lipids of the pollen are synthesized in situ, by metabolic processes regulated by expression of genes in the haploid genome.To whom correspondence should be addressed     

Sources,impact and exchange of early-spring birch pollen in the Moscow region and Finland

The paper presents an assessment of birch pollen seasons in Finland and Russia. The re-analysis covered the period from 1994 to 2005 and was focused on suspected long-range transport events that were recorded both in Moscow and at several Finnish sites. In order to trace the origin of airborne pollen before the onset of local pollination, we used both aerobiological and phenological observations combined with forward and adjoint (inverse) dispersion model simulations. It is shown that, although the Moscow region is surrounded by extensive birch forests, it still receives substantial amounts of foreign pollen before local pollination. In the Moscow region, the sources of long-range-transported pollen are in the south and south-west, sometimes even in the east. In Finland, there are frequently cases, before the local flowering season, in which Finnish territory receives Russian pollen; however in the opposite direction, from Finland to the Moscow region, no transport episodes were unequivocally registered. Analysis of the end of the seasons was more problematic, due to contributions to pollen observations from local sources; this results in difficulties in the reliable identification of the long-range transport episodes. Apart from its short-term effects on the pollen seasons, long-range transport can have substantial impacts on the exchange of genetic material within Europe. A quick atmospheric pathway for gene transport can be important for adaptation of plants to a changing climate.     

Control of the developmental pathway of tobacco pollen in vitro

We developed a new method for culture of isolated pollen. Using highly homogeneous populations of immature pollen grains of Nicotiana tabacum L. prepared by means of Percoll density gradient centrifugation, we could direct their developmental pathway by regulating certain culture conditions. When the pollen population was cultured in basal medium with glutamine, most pollen grains underwent normal maturation. On the other hand, when first cultured in basal medium without glutamine, most pollen grains did not mature but after transfer to medium with glutamine and sucrose began to divide. This method for inducing pollen cell division was possible only with midbinucleate pollen grains which are characterized by having no central vacuole and no or only a few starch grains. Evidently, some essential changes necessary for the embryogenic response can be induced by glutamine starvation only in pollen grains at a specific stage.     

A long-term study of winter and early spring tree pollen in the Tulsa, Oklahoma atmosphere

Since 1986 the atmosphere in Tulsa, Oklahoma has been monitored for airborne pollen and spores with a Burkard 7-day spore trap situated on the roof of a building at The University of Tulsa. The present study specifically examined the early spring tree pollen season for several local taxa and the occurrence of pre-season pollen during December and January. Knowledge of the local pollen season will help identify the presence of out-of-season pollen and possible long distance transport (LDT) events. Average daily concentrations of airborne pollen for species ofBetula, Quercus, Ulmus, and Cupressaceae were determined for each year from 1987 to 1996. The data showed that during the early spring the precise pollination periods for these allergenic tree species are highly variable. There were considerable variations in start date, season length, peak concentration, date of peak, and cumulative season total. The start dates forUlmus, Betula, andQuercus varied by 30 days or more, while the early spring Cupressaceae pollen showed the least variation in start date (only 23 days). More research is needed to understand the mechanisms which govern the onset and magnitude of pollen release. Although several reports have documented episodes of long distance transport (LDT) of pollen, the actual contribution of out-of-season or out-of-region pollen to local air spora is poorly known. The current study also re-examined the LDT ofJuniperus ashei pollen in Oklahoma.Juniperus pollen appeared in the Tulsa atmosphere on 40% of the days in December and January with concentrations as high as 2400 pollen grains/m³ of air; however, no local populations ofJuniperus pollinate at this time of the year. High concentrations occurred on days with southerly winds suggesting thatJuniperus ashei populations in southern Oklahoma and Texas were the pollen source. Since no local pollen is present in the Tulsa atmosphere in December and January, this example of LDT has been easy to document.     

In vitro pollen embryogenesis in Nicotiana tabacum L. and its relation to pollen sterility,sex balance,and floral induction of the pollen donor plants

Pollen sterility, sex balance, and floral induction of the pollen donor plants were tested for a possible relation to embryogenesis from in vitro cultured tobacco pollen (Nicotiana tabacum L. var. Badischer Burley). The pollen grains destined to become embryos in culture (P-grains) were sterile for the donor plants as judged by their staining reaction with acetocarmine and fluorescin-diacetate, and by an in vitro germination test. They were produced in high frequency in flowers which exhibited a shift in sex balance towards femaleness. Sex balance could be measured by the relative length of pistil to stamens. High P-grain frequency, high pollen sterility, and a shift in sex balance towards femaleness could be induced by raising the donor plants under short days and/or low temperature (18–15° C) as compared to long days at 24° C. Short days and/or low temperature also reinforced floral induction, revealing that the tobacco variety Badischer Burley is a quantitative short day and low temperature plant and that the variety follows the rule that conditions of strong floral induction shift sex balance towards femaleness. At 12° C and short days, contabescent flowers were formed with completely sterile anthers containing a few and mostly collapsed P-grains. Based on these results, it is now possible to predict conditions by which haploids via pollen embryogenesis might be produced in high frequency from low-yielding and recalcitrant species.Abbreviations DPF dead pollen grain frequency - LD24 long days at 24° C - PD pollen dimorphism - P:S ratio of pistil to stamen length - SD15 short days at 15° C     

Specific phosphoproteins in the initial period of tobacco pollen embryogenesis

Several phosphoproteins specifically correlated with the induction of embryogenic cells were detected in immature pollen grains of Nicotiana tabacum L. By regulating the concentration of glutamine in the medium the developmental pathways of immature pollen grains isolated at the mid-bicellular stage could be controlled, resulting in the formation of either mature pollen grains or embryogenic cells. Different phosphoproteins, designated as a-d and as e-i, respectively, were detected when the pollen grains either became embryogenic cells in glutamine-free medium, or when they were allowed to mature in glutamine-containing medium. The formation of embryogenic cells was suppressed by adding glutamine or cytokinin to the glutamine-free medium, nor did it occur with pollen grains at younger or older stages, and in these cases the phosphoproteins a-d were detectable only partially or faintly. The phosphoproteins a-d and e-i thus may be one of the factors necessary to direct the developmental pathway of immature tobacco pollen grains to embryogenic cells and to mature pollen grains, respectively.The authors thank Dr. V.S. Jaiswal (Botany Department, Banaras Hindu University, Varanasi, India) for his valuable suggestion in the preparation of the paper. This work was supported by a Grantin-Aid for special project research from the Ministry of Education, Science and Culture of Japan.     

Subcellular distribution of arabinogalactan proteins in pollen grains and tubes as revealed with a monoclonal antibody raised against stylar arabinogalactan proteins

Summary Monoclonal antibody PCBC3, raised against stylar extracts fromNicotians, alata flowers, was deduced from enzyme-linked immunosorbent assays and inhibition of immuno-gold labelling on tissue sections to bind specifically to carbohydrate epitopes on arabinogalactan proteins (AGPs) but not to other arabinose-containing cell wall polysaccharides. When pollen grains ofN. tabacum were hydrated in fixative, PCBC3 bound to vesicles in the vicinity of the endoplasmic reticulum but, when grains were hydrated for 20 min in culture medium before fixation, binding was restricted to the plasma membrane. The generative-cell plasma membrane was also labelled in grains ofLycopersicon peruvianum. In pollen tubes ofN. tabacum grown in liquid culture, the AGPs detected by PCBC3 were located in several regions, including the plasma membrane, tubular-vesicular structures (plasmalemmasomes) at and under the plasma membrane, and multilamellar bodies within vacuoles, features generally associated with endocytosis. Labelling was not evident in secretory vesicles or the plasma membrane at the pollen-tube tip. The AGPs detected with PCBC3 were also present in pollen-tube walls, near the interface between the inner, callosic layer and the outer, fibrillar, pectic layer. Pollen tubes ofN. tabacum grown in medium lacking added CuSO₄ produce a wall with an abnormally thickened fibrillar layer, and this layer was uniformly labelled with PCBC3. The disposition of wall AGPs thus changes in pollen tubes of different morphologies.Abbreviations AGP arabinogalactan protein - -L-Araf -L-arabinofuranose - ELISA enzyme-linked immunosorbent assay - MAb monoclonal antibody - PBS phosphate-buffered saline