Geeking out on farming
There are many big claims in agricultural research. That GMOs are higher yielding than non-GMOs, that biochar is the next best thing, or that intercropping is the way to go. Many of these claims are supported by individual syntheses, each with thousands of individual experimental or quasi-experimental paired observations of the impact of treatments on crop yields or other outcomes measured against controls.
These data sets are scattered across the literature and the strength of the evidence for, or against, a given intervention across different locations in the world is often masked in existing scientific publications. This makes it hard to compare interventions side-by-side in a way that is relevant to a given context. This is a problem as much for researchers as it is for practitioners and donors.
We’ve been interested in trying to solve this problem for a while. We’re two data geeks who are interested in assessing big claims in agriculture. But being researchers, we also know there is massive heterogeneity in outcomes, and are interested in local solutions for global problems. This means knowing where the gaps in our knowledge lie, and how the evidence base differs across different places in the world. Over the last couple of years, off the side of our desks and with part-time help from students, we started to work on a solution to this issue, which we’ve called FarmGeek.
FarmGeek’s mission is to bring evidence from the peer-reviewed literature on the outcomes of agricultural management practices and food system interventions to anyone that is curious. We do this in a way that is geospatially explicit, so you can see first hand the scientific data underlying the big ideas and claims in the media and elsewhere on food system transformations.
Here’s the link: https://www.farmgeek.xyz/
FarmGeek currently includes data on over 144,000 observations of yields across many different farming interventions, such as GMOs, Organic, Biochar, Intercropping, Agro-forestry, Cultivar mixes, Biofertilizers and more. We’ve built the architecture to summarize this data and do it in a dynamic way that rescales the strength of evidence for any location you’d like to zoom into on the planet. You can even zoom all the way to see the experiments at that location, and get the DOI (s) for the publications from which we extracted the data.
We believe new methods of data visualization incorporating dynamic evidence synthesis and cross-scale analysis offer a new way to assess claims in food systems. FarmGeek shows this can be done for yields, and in the future we plan to extend this concept to incorporate additional outcomes such as biodiversity, and other aspects of environmental and human health.
Finally, FarmGeek was developed as a service for the community and we will continue to work on existing and new features based on your feedback and as time and resources allow. So come check it out, and if you have ideas or suggestions on ways to improve it, or just want to say you found it useful, we’d absolutely love to hear from you.
Dr. Zia Mehrabi & Dr. Navin Ramankutty