| Popis: |
Nonlinear models were used to estimate first emergence and peak abundance dates for Aphthona lacertosa Rosenhauer and A. nigriscutis Foudras, two flea beetles introduced to control leafy spurge, Euphorbia esula L., in North America. For model development, 26 field sites were sampled for flea beetle abundance at weekly intervals for eight weeks in three western Minnesota counties in 2000, 2001, and 2002. A three-parameter Weibull function, fit to observed cumulative probability distributions, were used to predict accumulated degree-days (ADD) to first emergence. Bias testing indicated the Weibull function provided a useful estimate of first emergence for A. lacertosa (304 ADD, lower developmental threshold 7.5 °C), but failed to produce a useful estimate for A. nigriscutis. A third-order polynomial was used to approximate seasonal abundance and predict peak abundance for each species. Estimated ADD to peak abundance of A. lacertosa was 594 ± 24 (DD > 7.5 °C) and 670 ± 15 (DD > 9.3 °C) for A. nigriscutis. Models were validated with additional data sets from Minnesota, Montana, and North Dakota. Estimated date of peak emergence provided useful predictions of peak emergence for Minnesota and North Dakota, but failed to predict peak emergence in Montana. We speculate that variation in climate and environmental conditions between Midwestern states and Montana were responsible for differing emergence patterns. We conclude that phenology models should be developed regionally to provide useful predictions of peak emergence for land managers. Maps were developed for Minnesota to spatially display predicted dates of peak abundance for A. lacertosa and A. nigriscutis. |
Full Text from ScienceDirect