Predicting Pollen mast years using Agent-Based Modelling

The volume of pollen produced by trees differs from year to year. Especially beach and oak trees have mast years. Every few years, trees produce more than the average amount of fruit to increase their chances of survival. The crop of fruits and nuts is also called mast, which is why we call these years with large crops mast years. During these years, there are abundant fruits, more than animals can eat, increasing the chances of new tree development. However, trees do not want to produce massive fruit yearly, as many fruits will weaken the tree.

Mast years only work when trees in the same neighbourhood synchronize their masting. Interestingly, trees in the same areas are synchronized in mast years. It is often called a mystery how trees know that they have to produce more fruit in some years. There are several hypotheses on how mast years work and why they are synchronized, including the predator satiation hypothesis, the pollination efficiency hypothesis and the environmental predication hypothesis (Pears et al. 2016). Similar weather conditions can cause synchronization.

Due to climate change, weather conditions change, and when mast years are driven by weather, these mast years will also change. This can cause large impacts, not only on the trees but on the complete ecosystem. It can, e.g. influence migration patterns of birds and, ultimately, humans. Large numbers of fruits can only be produced when there are large volumes of pollen. High volumes of pollen cause years of severe pollen allergies.

Agent-Based models are very suitable for testing hypotheses and reproducing patterns. A model can be created that runs a single hypothesis but can also combine different hypotheses. This way, the best possible hypothesis can be selected. The test can be repeated for different areas to see if the model produces good results for all test sites.

Pattern-Oriented Modelling is based on the fact that complex systems produce patterns and that good models should be able to reproduce these patterns (Grimm et al., 2005). Mast years have temporal patterns; they happen every 3, 4 or 5 years. In this research, you will check to what extent the hypothesis of masting can reproduce patterns produced via your ABMs.

In this topic, you will:

• Collect and Analyse data on mast years for a few test sites
• Design an ABM to test the masting hypothesis
• Implement the ABM and test this model for the selected sites
• Make a comparison between your model results and the spatial-temporal patterns of your test sites to select the best or combination of hypotheses.