[Untitled]

The experimental testing of the hypothesis of population reduction by mass trapping with the establishment of treatment/control areas is a logistic problem on landscape scale. We suggest a possible test based on time-series analysis of tree-mortality data. Our example is helped by: an island effect, a 20-year tree-mortality record, and a very competitive synthetic aggregation pheromone. The successful development of synthetic pheromone (ipsdienol and E-myrcenol) dispensers, and the dose-response of wick-dispensers is briefly described. In the 2000 ha of spruce (Picea mongolica) forest reserve at Baiyinaobao, surrounded by grassland, 80 traps were employed for 3 years, catching 0.5–1.7 million beetles/year. In spite of intensive conventional management since 1982 (17 years), the long-term tree mortality had been oscillating around 600. During the first three years of mass trapping, tree mortality was down to 118, 100, and 88, respectively. Neither natural population cycles nor weather conditions could explain the 3-year mortality drop. Cycles did not drive the drop, as auto-correlation function (ACF) and partial ACF plots do not show any clear periodicity, only a weak 5-year cycle is visible. Weather did not explain the drop, as the 3 years of mass trapping had average rainfall and higher than average temperatures. Climate data show negative correlations of tree mortality and temperatures on year basis and a 20-year trend of increased monthly averages, especially for July temperature. The increased temperature (global warming) will increase stress on trees, but mass trapping may have counterbalanced this new threat to the forest.