Our synthesis project used three long-term datasets to investigate the reliability of flowering communities from the perspective of pollinators in extreme environments. This study was possible because the authors were willing to share their long-term datasets within our research group. Making such datasets openly accessible to the public, however, is a priority for NSF. NSF recently funded the Environmental Data Initiative (EDI) to 1) archive of data from NSF-funded research programs such as LTREB and OBFS in the EDI Repository; 2) support and train researchers in those communities to archive high-quality data and metadata; and 3) develop best practices for particular themes of data, such as plant phenology, that will make data easier to discover and integrate. EDI will accelerate the pace of data syntheses such as ours in the future.
Sharing data from different locations allowed us to make more general inferences about the availability of floral resources in harsh environments than if we had examined each dataset independently. We calculated two indices (flowering frequency and turnover rates) to explore our hypotheses that communities with less reliable precipitation would have less reliable floral resources and that annual plants would show this pattern more strongly than perennials.
Our research yielded novel results. Reliable plant species that bloomed in the same month every year of each study were rare. Bloom frequency and turnover rates relative to precipitation reliability were consistent with our hypotheses. Looking at the entire community as a resource for foraging pollinators, we found that flowers are an unreliable resource, especially in unpredictable environments.
This analysis was conducted in 2010, prior to the advent of repositories such as the EDI Repository. Collaborators were identified through literature searches and personal contacts. Manuscript reviewers suggested we add more long-term datasets from extreme environments to the synthesis, but we could not locate any. Today, as EDI and other entities provide ecologists with basic information management skills, more data sets such as ours are being archived in open data repositories. Future researchers will more easily locate and re-use these valuable datasets. Data providers will also benefit as repositories assign DOIs to datasets, making data citations easy to track. We, as plant phenology researchers, can also collaborate with EDI to define best practices for formatting and documenting plant phenology datasets that will facilitate easy understanding and re-use in the future.