How weather and landscape interact over time to influence Colorado potato beetle population dynamics

Population dynamics are controlled by the environment and moderated by life history. These factors determine the timing and magnitude of population abundance and are used to predict crop pest populations. Forecasts of pest populations typically focus on heat accumulation during the growing season when the pests are active or emerging from the soil if they have a diapause period. Weather conditions during diapause can impact population dynamics as well but tend to be understudied. This study uses 14 years of Colorado Potato Beetle (CPB) abundance data combined with landscape and daily weather data to understand the importance of seasonal weather and landscape on predicting population abundance. We used boosted-tree multi-class models to predict abundance at each life stage and found that across all models, cumulative degree days (CDD), a measure of developmental time, was overwhelmingly important. To understand how the top non-temporal variables could interact with time, we used general additive models to examine interactions between CDD and the top six ranked variables for each life stage, then generated predicted abundances across the growing season for the 10th, 50th, and 90th percentiles of each variable, keeping all other variables constant. This revealed that while CPB abundance is most strongly affected by heat accumulation, other climate factors like precipitation and humidity, as well as winter temperatures can also influence abundance trends. Landscape variables demonstrated less variation over time and smaller differences in predicted abundance but indicated that some non-potato crop cover can influence abundance.