pH and nutrients disparities


There are dramatic differences in N levels and pH in the four beds, presumably because of what's been grown there before. If these tests are to be carried out only after sowing and planting, how are results of this experiment expected to be valid and credible?

Seems like this is the only way to acquire reliable data:


Hello Victoria . As much as I remember the test had to be done before sowing and after harvest . Personally like others ( I presume ) done the first test after the sowing , because my test kit arrived after . I haven’t finished the growing season yet so after that should get the second test . For me my best result and surprise was the monoculture plot of Cobra beans . Even with the soil depleted , had satisfying results ( for me ) . At least in the case of beans the ipothesis is there should be more Nitrogen after the harvest . We shall find out soon :) .

Best wishes to all .


Hi Victoria, Thanks for the great question and the link to an interesting paper.

Yes, as Ionut suggests, what we are looking for here is the change in your plots during the growing season in relation to what we are growing in them. Because we will compare each plot with itself (e.g. monoculture beans in start and after end of experiment), then we can account for the impact of what you've grown before in any particular bed.

It is interesting to see what your starting conditions are - and that your plots are very different. As you suggest, this can reflect what you've grown in the past, as well as any differences in management. For some people, part of their plot might have had manure added 2 years ago, for example.

For the GROW Experiment, however, the really interesting results, will come when we compare the start with the end. We will ask you to do the second set of observations after your last harvest. This is because for nitrogen-fixers, like our beans, they fix nitrogen during growth for their own use. It is only afterwards, as they break down, that this becomes released into the soil for other plants to use.

I hope that helps answer your question? I'm very happy to continue the discussion if you have more details that you'd like to know. Just ask here.

Finally, it's worth noting that this is the first citizen science experiment that we know of looking at small scale regenerative practices in food growing across many countries. Our methods and approaches can't be the same as if we were doing this in more controlled conditions, such as in a farmer field with access to laboratory equipment and resources. More on this, and much, much more in our up-coming free online course "Citizen Research: From Data to Action" which will soon be open on FutureLearn for registration. I hope you will join us there.

Best wishes, Naomi

Thank you Ionuț and Naomi for getting back to me. I was just wondering whether the soil tests should have been done well before sowing, so one could allocate beds for the experiment with soils in similar conditions, in order to have like for like comparisons between poly and mono-cultures yields.

My poly-culture beans are yielding heavier than mono-culture ones. Beans are known to favour slightly acidic soils. My poly-culture pH is between 6 and 6.5, whilst the mono-culture one is neutral. So is the better performance caused by the presence of spinach and radishes or simply by the lower pH? Comparing with other growers' data can give an inclination, but not a definite answer. And we are supposed to minimize variability between the control and treated plot.

I understand that academic research on inter-cropping is done on large scale farms, but maybe some of it's theories can apply to our experiments?

Best regards, Vic

Hi Vic,

Ah, yes, in an ideal world we would have loved to have ensured that the conditions were relatively similar right from the start. With the nutrient levels, you could potentially have had to do quite a lot of tests on various spaces before you found several plots the same. This would need more tests, time and space. We are aware of how tight time and space is for people and we have learned firstly that some people had already planned their space by the time they knew about the experiment (we will be starting a bit earlier next year!) and, secondly, that some people didn't have enough space to do it at all. So, finding plots that were likely to be similar is balanced against, well, finding plots. :)

Even experiments in a single large plot, or farm, expect to have variation between individual plots which is why replication is so important. Where you have just one plot of each type you can draw some insights, but it is harder to be sure -why- you see any differences - as you rightly point out. In farm-scale experiments, there will be several plots of each type (replicates). In our citizen-science experiments, there is also replication and this is spread across many different people.

Measuring the nutrients as we have done can help us to determine possible reasons for variations in crops like those differences that you have found. For example, if all the higher-yielding bean plots are associated with a particular pH range more so than whether they are polyculture/monoculture then we would understand that the planting arrangement is less important than the soil pH. Within the replications, it is likely that some people have a lower pH in the monoculture and some in the polyculture so this should balance out when we look at all the results together.

I love that growing trials that take place in real growing environments ("in the field" - be it a farm, garden, or allotment) are so different to lab trials. In laboratory trials everything is controlled - the temperature, light, humidity, pH, nutrients, even pollinators and pests. But the real world is more messy and complex than that. I'm an ecologist, though, and I know that this complexity is important, even if it's not always convenient.

As a general approach to field trials, we try to control for as much as possible but we accept that we can't control everything; there will always be tiny microclimate changes, slope position, disturbance etc. Part of this variation is mitigated by using something like a randomised block design so that the same plot type (e.g. monoculture beans) is not always at the bottom of the slope / in the shady patch / in some other space with differences that you can't see), but instead these are scattered randomly amongst the other plots so any plot type has an equal chance of being affected (there was more on this - with some diagrams to clarify - in the "Soil to Food" MOOC). For things we can't control or know in advance, we can measure them to ascertain any differences that could affect our results - like the pH. For anything else that crops up (excuse the pun!) along the way, we take notes so that when we look back through our results we can see "oh yes, that particularly small crop coincided with the slug invasion" or similar.

So here you see that there is a balance between what you can do in your own growing space and what we can do collectively. To improve your results from your own growing, you could also add replication. If you have space, you can have two or more versions of each plot at the same time. Where space is limiting, and to understand the influences that the weather can have, then repeating the experiment for two or more years can help you get much clearer answers. I hope that you feel inspired to join us again next year for further exploration! If you do have more space to play with, let us know and we'll be happy to help you work out something a little more tailored that can give you further insights. :)

Likewise, if you have further suggestions for improvement, please keep them coming. We currently have our partners at the FAO (Food and Agriculture Organisation) analysing the nutrient test kits against laboratory results from James Hutton Institute so we can also gain further insights about the reliability and application of these tests.

Warm regards, Naomi

Reply — Last edited

Hi Naomi,

Thank you for clarifying this. I see that shortfalls in controlling variables are being compensated by Europe-wide replications and, hopefully, consistent patterns will emerge. I certainly wish to join next year, as I'd love to see this project evolve. I recently got interviewed about it by Radio Berkshire, visiting our community allotment, Southcote Growallot, now and then. Never been on air before! They thought it was really interesting :)

There's a new bed at Southcote Growallot, almost 50 square meters, currently occupied by pumpkins and a few tomato plants. It would be OK for me to use it, in principle, for the next year's experiments, maybe not all of it, but probably most of it. There's also a number of smaller raised beds. (Southcote Growallot is on Google maps, with quite a decent satellite image, but the new large bed isn't on it yet). If the next experiment will involve climbing beans, the polyculture plot will need to be a bit larger than the current one, if other crops will stand a chance growing amongst them.

Coincidentally, the University of Reading is undertaking this type of research, on a much larger scale and also on EU grant! Do you know anybody there, by chance? I wonder if some sort of collaboration could be possible. Of course, we are now at risk that Brexit will kill it all off :(

Excellent idea to analyze these soil test kits in a lab, as we only have the manufacturer's word.

Please do send me suggestions for the future experiments and I'll discuss them with other Southcote Growallot members. I'm sure, if we plan well enough ahead more space can be found :)

Best regards,


Hi Vic, Wow! Brilliant to hear you've been talking about it on local radio! Do you have any links to the programme/website that we might share?

Sounds like you've got an amazing spot for some detailed and robust experiments and I'll be glad to talk with you about this and what we can do there. What a brilliant opportunity! Could you email me on so we can discuss? Next week all the GROW partners are meeting in Dundee for some evaluation and direction-setting for next year, so I'll be sure to keep this in mind for our discussions and will get back to you in early October.

Yes, I know a few folks in Reading - more on the agroforestry side. There certainly seems to have been a (re)surgence in interest in agroecology at UK Universities in recent years which is very promising. Definitely many exciting opportunities for collaboration going forward.

I know that, whatever may come of Brexit, there are lots of amazing people on both sides of the channel who want to work together to make the world a great place, so I'm confident that we will continue to do so.

Warm regards, Naomi

Hi Naomi,

Here's the link to the iPlayer: Bill Buckley, Friday 14/09, 3pm (two thirds into the program). I don't think my first appearance on air was great haha :) However, they love visiting our site and, I guess, the talking will be easier once we've got some solid results that'll give us more insights.

I'll email you, as you asked, to discuss the next experiment at Southcote Growallot.

Best regards,