You made it!
This is where we identify, discuss, and refine the lab methods used in the Real Food Campaign to measure soil, food, and human health. You can learn more about the Real Food Campaign here.
We also try really hard to overlap our methods and process with that of like minded labs. We welcome contributors who help do this!
Why this matters
The methods identified in this working group will be used within the Bionutrient Food Association’s own testing lab and other Bionutrient Food Association partner labs who collect soil and food samples.
We also hope by building a common set of methods in a public + transparent way, a broad coalition of research partners can collect comparable data to answer complex questions about our food system.
Objectives of the Real Food Campaign
- Collect enough data to identify complex connections between soil, plant, and human health.
- Use that data to calibrate scalable methods/sensors
- Deliver real-time quality information to farmers and consumers using low-cost sensors.
What methods should we select?
These objectives put serious constraints on what lab methods we can use. While each working group will have their own unique constraints, all working groups have the following constraints.
Methods should be…
Low cost – Too expensive, and we can’t collect enough data to identify complex relationships. Ideally, <$75 per sample (food and soil is one sample) and no massive up-front equipment costs.
Representative – Low cost means each measurement should be as broad as possible. For example, measuring anti-oxidents generally is better than measuring a single anti-oxident.
Quality – It is more important to be comparable than to be accurate. If large amounts of sensor and metadata are collected, noise is acceptable as long as the noise is actually random.
Relatable – Methods which relate to measurable outcomes of interest, like our environment (soil runoff for example) or human health (prevalence of disease). Interesting measurements of soil for soil’s sake aren’t useful.
Scalable – While the lab methods themselves may not scale, the outputs of those methods would ideally correlate to a handheld sensor, image, and method that does scale. We should avoid lab measurements which cannot help us scale the delivery of results.
How can we improve over time?
To improve, we must have version control as not all users can upgrade to the most recent method immediately (for example, those running long-term experiments). Versioning should also be both human and machine readable, so that method updates propagate immediately.
In addition, we need a robust place between versions to discuss + deliberate.
To address these two related but unique requirements, I propose we use Discourse’s Wiki option for real-time discussions and updates, and Github for long-term version control + machine-readable access. This keeps the discussion focused here in the OpenAgDev forum rather than diluting it in google docs or other platforms.
We need to create the structure of the gitlab repository to make methods most widely accessible. For example, if there are multiple measurements for the same output (a 1 day soil respiration versus 4 day soil respiration test), we should identify the preferred method to maximize comparability whenever possible. We won’t get this right from day 1, and we’ll need to learn and iterate to improve how we store and deliver versions of the methods.
How is the final method decided?
There will be a smaller group of people within the working group selected to consent on the final method. We considered having broader consensus, but as there are so many opinions and perspectives, achieving a public consensus could just be too hard.
That being said, input from everyone is critically important to ensuring we have the best method possible, and moderators work hard to integrate comments and thoughts from everyone who is part of the discussion into the final product.
Where to start?
First - we need to start somewhere - so don’t let the perfect be the enemy of the good. We’re more interested in incremental improvement, not perfection on day 1. Here’s some good starting questions / points.
What are the rough categories of health, as it relates to the
Real Food Campaign (so for soil health, it may be “biological health”, “mineral availability”, and “structure” for example). An excellent starting point is the Sustainable Intensification Manuel, which categorizes outcomes of interest for development projects and some are relevant to the Real Food Campaign. FAO and UN may be other good places to look.
- Lit review of possible methods to capture information in those categories, and are those methods broadly accepted (if not, we have some work to do!).
- What validation / calibration work needs to be done? Add confidence to the methods we choose.
In addition, working groups will have monthly or bimonthly calls with notes that will also be posted on the forum.