Rheged in Galloway

[Jim Webster]

We can take the figures even further. The article comments that a lot of villa estates were around 300 acres plus. (I'm a bit wary of the figure to be honest, but let it stand)

The article also says "The village unit at Figheldean Down covered about 370 acres of arable, feeding some dozen families"

So assuming a yield of 400 kg per acre.
Assuming alternate years fallow which was common,
Then the land produced 74 tons of wheat
18.5 tons have to be kept as seed
If we take the Egyptian figure of a family needing 0.75 tons of wheat to be self sufficient, then somewhere between nine and twelve tons have to be kept back to feed the people who're growing it.

So of your 74 tons, 44 tons are 'surplus.'

Assuming 3lb of wheat per man (Engels, Alexander the Great and the logistics of the Macedonian army) this estate, selling its surplus to the army, will feed 90 men in the neighbouring garrison for a year

Jim

[Erpingham]

Fascinating to see this article again.  I used this a lot in my undergraduate dissertation (yes, it is that old).  It inspired me to have aquick scout around on the internet for more modern work and I found this useful set of comparators

https://core.ac.uk/download/pdf/77571.pdf

Go to appendix 5 on p.412

A key thing is standard roman sowing rate of 5 modii per iugerum which is apparently 135kg/ha.  On Columella's basis, this gives 540 kg per hectare, lower than Applebaum's figure. But Columella is rather lower than other Roman yield estimates.  Then again, compare the early medieval yields produced by small farmers rather than large estates, which are lower that Columella.  I think overall all I'm saying is there is a shed load of variability to contend with in trying to get post-Roman yield rates right.

[Jim Webster]

Fascinating to see this article again.  I used this a lot in my undergraduate dissertation (yes, it is that old).  It inspired me to have aquick scout around on the internet for more modern work and I found this useful set of comparators

https://core.ac.uk/download/pdf/77571.pdf

Go to appendix 5 on p.412

A key thing is standard roman sowing rate of 5 modii per iugerum which is apparently 135kg/ha.  On Columella's basis, this gives 540 kg per hectare, lower than Applebaum's figure. But Columella is rather lower than other Roman yield estimates.  Then again, compare the early medieval yields produced by small farmers rather than large estates, which are lower that Columella.  I think overall all I'm saying is there is a shed load of variability to contend with in trying to get post-Roman yield rates right.

there are a lot of variables, and average yields seemed to vary wildly around the Med. Wheat does best with sun and high temperatures, provided there is enough water, so I'd expect British yields to be at the low end

[Erpingham]

there are a lot of variables, and average yields seemed to vary wildly around the Med. Wheat does best with sun and high temperatures, provided there is enough water, so I'd expect British yields to be at the low end

I'm just working on what I've read rather than having real knowledge of farming but it looks that way to me too.  I also turned up a large meta-study of medieval English crop yields in which the authors didn't even look at yields per hectare because of the variables but just looked at the other yield measure of seed to crop.  Averaged out about 3.75 IIRC.  So our ballpark for our sub-Roman yields could plausibly be 270-540 kg per hectare normally, depending on where and scale of operation and have the British weather is an important wildcard variable.

[Jim Webster]

there are a lot of variables, and average yields seemed to vary wildly around the Med. Wheat does best with sun and high temperatures, provided there is enough water, so I'd expect British yields to be at the low end

I'm just working on what I've read rather than having real knowledge of farming but it looks that way to me too.  I also turned up a large meta-study of medieval English crop yields in which the authors didn't even look at yields per hectare because of the variables but just looked at the other yield measure of seed to crop.  Averaged out about 3.75 IIRC.  So our ballpark for our sub-Roman yields could plausibly be 270-540 kg per hectare normally, depending on where and scale of operation and have the British weather is an important wildcard variable.

Have you read Georges Duby's 'Rural Economy and Country Life in the Medieval West'? He's one who looks at yield to seed ratios for exactly the reason you mention.
But he sees an increase (for good agricultural reasons) fro 2:1 about 1000 AD to 5:1 or slightly better about 1300AD (with occasional good years or well run estates hitting as high as 11:1 on rare and much celebrated occasions.
He also comments that in 1812, the sous prefect of Marseilles, in reply to a query, wrote to say that the average harvest, over a ten year period, produced a yield of between 4.5 and 5 to one.

[Erpingham]

Only dipped into this area occassionally because of a residual interest as I say dating back to that article and others of the time (there was a great one about Iron Age Holland, I recall).  The thing the some of these studies did was to try to look at the wider economy of these farms - animal husbandry, other crops etc., which of course makes it more complicated.  For example, the Medieval English grew lots of peas and beans and various mixes of cereals to protect themselves against weather conditions working against one crop.  Did the Romans and any settlers do this and how might it affect calculations?  But maybe that's getting off too far off the track

One thing we might valuably do in which a good husbandman would have an insight is what impact does sustaining horsemen rather than footmen have on the equation?

[Andreas Johansson]

Assuming 3lb of wheat per man (Engels, Alexander the Great and the logistics of the Macedonian army) this estate, selling its surplus to the army, will feed 90 men in the neighbouring garrison for a year

Do we have any idea how many artisans, soldiers' wives, and assorted hangers-on a provincial Roman unit would have per fighting man?

[Jim Webster]

Assuming 3lb of wheat per man (Engels, Alexander the Great and the logistics of the Macedonian army) this estate, selling its surplus to the army, will feed 90 men in the neighbouring garrison for a year

Do we have any idea how many artisans, soldiers' wives, and assorted hangers-on a provincial Roman unit would have per fighting man?

In crude terms we could just assume the Egyptian 0.75 tons per head per year for fighting man and 'family' which would cut the number supported to 60 men (probably 50 because of double pay men etc)
Another way is to see what annona they would expect when still paid in kind

Jim

[Imperial Dave]

unless mentioned elsewhere, there's this article as well

https://www.academia.edu/1534036/The_Roman_Agricultural_Economy_Organization_Investment_and_Production