Cherine Jantzen

Masting, highly synchronised but temporally variable seed production, is initiated by weather cues and is thus highly sensitive to climate change. Changes in these cues can lead to a masting breakdown, reducing the fitness benefits of masting through decreasing pollination efficiency and increasing predation risk for seeds. Here, we use 50 years of individual tree data on annual seed production of European beech (Fagus sylvatica) in the Netherlands to assess temporal changes in masting patterns and their consequences for the selective benefits of masting. Additionally, we use a novel approach to identify which weather cues initiate reproduction, assess their temporal changes, and test whether they account for the observed changes in masting. We show that synchrony and inter-annual variation in beechnut production have declined, resulting in a masting breakdown in the late-2000s, since which there has been constant, but low, seed production each year. Consequently, predation risk increased almost three-fold, while pollination became less efficient, together reducing the fitness benefits of masting. Seed production was driven by precipitation and temperatures in the year of seed fall and the two preceding years, but the periods within the year in which trees respond to each climate variable differ in both timing and duration. Interestingly, only temperature, not precipitation, has changed over time, but this change only partly explained the observed changes in masting patterns. Masting breakdown is shown across the species range, but its fitness consequences remain understudied, because detailed, individual-level, long-term data are required but still rare. By using such a dataset, we here provide crucial evidence for the negative consequences of masting breakdown for beeches through reduced pollination efficiency and increasing predation risk. Using a new methodology, we further underline the strong effects of weather cues on reproduction, while showing that changing climate alone cannot be driving the masting breakdown and must interact with currently unidentified factors.

Ecological data is highly diverse due to the complex nature of the systems they describe. Proper documentation and management are often lacking or not designed for data reuse by others, making the data difficult to find, understand, and at risk to be lost. Adopting the FAIR (Findable, Accessible, Interoperable, Reusable) principles into data practices is a way to mitigate these problems. However, the FAIR principles are abstract and not easily understood by domain scientists. Despite a growing body of assessment tools and resources about FAIR, applying it in practice remains challenging as clear implementation guidelines are missing. We aim to fill this gap by translating the FAIR principles into four data components (metadata, storage, standard and structure) that can be successively worked on to enhance the FAIRness and structure of data and provide a general workflow together with a hands-on guide to give practical suggestions on how to improve the reusability of ecological data. For every workflow step, we introduce the rationale behind it and point towards implementation solutions tailored to ecology. Additionally, we introduce an evaluation tool that facilitates the entry to this workflow by guiding to only those steps that are necessary for the evaluated dataset. With the workflow, guide and tool introduced here, we lower the threshold for ecologists to start making ecological data FAIR, which will ensure long-term reusability of valuable data sources.