Statistical modelling of wildwood mushroom abundance

• Published in: Scandinavian journal of forest research 2008-06, Vol.23 (3), p.244-249
• Main Authors: Dahl, Fredrik A, Galteland, Torborg, Gjelsvik, Runhild
• Format: Article
• Language: English
• Subjects: air; animal feeding; autumn; cloud cover; domestic animals; Ecology; humidity; Internet; linear models; meteorological data; meteorology; mushrooms; radionuclides; rain; regression analysis; statistical models; summer; temperature; variance; wind speed
• ISSN: 0282-7581; 1651-1891
• DOI: 10.1080/02827580802036184

This study analysed the relationship between weather conditions and the amount of wildwood mushrooms. A model of this relationship may be used to predict the amount of mushrooms with an elevated level of radionuclides that are available to domestic animals feeding outside during the summer and autumn months in Norway. Data on mushroom abundance were collected through a web application run by the Norwegian Mushroom Association. The data were coupled with meteorological data. Through linear regression analysis, it was found that average local cloud cover 5–3 weeks before the observation was the most important variable for explaining mushroom abundance at a given time and place. When controlling for cloud cover, weather variables covering rainfall, temperature, air humidity and wind speed had only marginal explanatory value. Of these, only rainfall and wind speed led to a significant (but small) improvement in the model. There was substantial unexplained residual variance in the model (R2=0.23), but data analysis from identical locations and periods indicated that this was due mainly to variability in the response variable.