Multi-decade patterns of gypsy moth fluctuations in the Carpathian Mountains and options for outbreak forecasting

摘要

The gypsy moth is the most important defoliator of broadleaved forests in the Northern Hemisphere. We explored the patterns in the moth's long-term dynamics and produced outbreak forecasts for seven countries located in central to southeastern Europe and extending into the Carpathian Mountains. We investigated how the outbreak periods and trends in the size of outbreak areas differ among the countries, the extent to which pest dynamics are synchronized, and how the observed patterns can be used to forecast outbreaks. We used long-term time series on outbreaks from about 1947 to 2013. The outbreak period ranged from 13 years in the northwest to 8 years in the southeast of the region; the periods were statistically significant in six of the seven countries (alpha = 0.05). Two distinct patterns in outbreak size were observed, i.e. while outbreak areas in the northwest were increasing, exceptionally large outbreaks occasionally occurred in the southeastern part of the region. Outbreak forecasts based on combined use of the Fourier Transform and ARFIMA approaches showed that outbreak predictability differs among the countries. An increase in outbreak areas, the control of which would require increased resources, was forecasted mainly in the central part of the region. Although the forecasts can support the forest management, there are limits to their use because of the complex relationships between the pest and the environment, which were not captured by our empirical forecasting models.