Przebieg i prognozowanie sezonów pyłkowych brzozy (Betula sp.) i dębu (Quercus sp.) w Poznaniu w latach 1996-2010
Loading...
Date
2012-06-25T10:46:12Z
Authors
Editor
Journal Title
Journal ISSN
Volume Title
Publisher
Title alternative
The course and forecasting of the pollen seasons of birch (Betula sp.) and oak (Quercus sp.) in Poznań, 1996-2010.
Abstract
Brzoza (Betula sp.) i dąb (Quercus sp.) wytwarzają ziarna pyłku o działaniu alergennym. Ich sezony pyłkowe (SP) charakteryzują się wyraźną zmiennością, co uzasadnia potrzebę opracowania modeli prognostycznych.
Głównym celem pracy jest przygotowanie modeli prognozujących początek i intensywność SP brzozy i dębu oraz określenie relacji między atmosferycznym SP a fazami kwitnienia brzozy brodawkowatej (B. pendula) i dębu szypułkowego (Q. robur).
Pomiar stężenia ziaren pyłku (1996-2010) dokonano metodą wolumetryczną. Do prognozowania przebiegu SP wykorzystano dwie grupy modeli: (i) wieloraką regresję liniową; (ii) modele termiczne. Obserwacje fenologiczne prowadzono w Poznaniu w latach 2008-2010.
Modele termiczne prognozujące początek SP wykazywały wyższy stopień dopasowania do danych niż modele regresyjne. Różnice między wartościami prognozowanymi a rzeczywistymi wynosiły od 1 do 5 dni. Modele prognozujące intensywność SP brzozy uzyskiwały wysoką wartość współczynnika determinacji (R2>0,90), a obliczone błędy predykcji były niskie. Słabsze dopasowanie do danych otrzymano w przypadku modeli prognozujących SPI dębu, a postawione prognozy były silnie niedoszacowane.
Wykorzystując przygotowane modele można określić początek SP z 2-3 tygodniowym, a jego intensywność nawet z kilkumiesięcznym wyprzedzeniem, co czyni je niezwykle pomocnym narzędziem w profilaktyce alergologicznej.
Birch (Betula sp.) and oak (Quercus sp.) pollen grains are an important cause of allergy. In Poznań, pollen seasons of these taxa show high year-to-year variations, which justifies the construction of forecast models. The main aim of this study was to forecast the intensity (SPI) and start of birch and oak pollen seasons and to establish the relationship between flowering phases of B. pendula and Q. robur and pollen data. Daily average pollen counts (1996-2010) were collected by volumetric spore trap. Two different methods for constructing forecasting models were applied: (i) multiple regression analysis; (ii) thermal time models. Phenological observations were conducted in Poznan during 2008-2010. The goodness of fit of the thermal time models was higher than those based on multiple regression analysis. The differences between estimated and observed start dates varied from 1 to 5 days. Models predicting the birch SPI had very high R2 values (>0.90) and the prediction errors were less than 20%. Conversely, in the case of oak the forecasts were not accurate (much lower than observed values). Based on the presented models it is possible to predict the start of pollen season in 2-3 weeks and the SPI within a few months of the season starting. Therefore these results may be a helpful tool for allergy prophylaxis.
Birch (Betula sp.) and oak (Quercus sp.) pollen grains are an important cause of allergy. In Poznań, pollen seasons of these taxa show high year-to-year variations, which justifies the construction of forecast models. The main aim of this study was to forecast the intensity (SPI) and start of birch and oak pollen seasons and to establish the relationship between flowering phases of B. pendula and Q. robur and pollen data. Daily average pollen counts (1996-2010) were collected by volumetric spore trap. Two different methods for constructing forecasting models were applied: (i) multiple regression analysis; (ii) thermal time models. Phenological observations were conducted in Poznan during 2008-2010. The goodness of fit of the thermal time models was higher than those based on multiple regression analysis. The differences between estimated and observed start dates varied from 1 to 5 days. Models predicting the birch SPI had very high R2 values (>0.90) and the prediction errors were less than 20%. Conversely, in the case of oak the forecasts were not accurate (much lower than observed values). Based on the presented models it is possible to predict the start of pollen season in 2-3 weeks and the SPI within a few months of the season starting. Therefore these results may be a helpful tool for allergy prophylaxis.
Description
Wydział Biologii: Instytut Biologii Środowiska
Sponsor
Keywords
ziarna pyłku, pollen grains, modele prognostyczne, forecasting models, alergia pyłkowa, pollen allergy, fenologia pylenia, pollination phenology