A map first, since these letters will keep moving between two states. Everything with a heartbeat is at the Parrish place on the Gulf Coast of Florida — the pigs, the birds, the feed shed. Locke Holler is 4.75 acres in Copperhill, Tennessee, up in the Copper Basin on the edge of the Cherokee National Forest, and it has no stock on it yet: a fence line, a survey argument, and a truck being built to get us there. So when I say the gilt, she is in Florida. When I say the property line, that is Tennessee.
You can move a plan. You cannot move a due date. Once that gilt is bred, gestation runs 114 days whether or not the farrowing paddock is finished, so the litter lands around the 27th of January and weans about the 24th of February, and every job between here and there just got a deadline attached to it.
That is the whole character of September. Nothing dramatic happens this month. It is the last month where being a week late is free.
The unglamorous part comes first
Before the boar does anything, the gilt in Parrish gets fed harder. Starting around the 21st we raise her feed for the two weeks before service — flushing, it's called — because it lifts the number of eggs she drops on that first cycle. It is the least photogenic job on the farm and it is worth more than most of the fencing.
People ask why we wait until October at all. Because this is the Gulf Coast of Florida, where heat is the whole design constraint — and because heat works on a delay. A boar makes sperm about two months before he uses it, so a hot July shows up as a disappointing September — he looks fine, he acts fine, and the conception rate is quietly down. Farrowing and nursing carry their own heat risk: a sow sprawled out trying to cool off is a sow more likely to lie on a piglet.
So the rule we run on is simple
No services in June, July or August. No farrowing between June and September. That leaves October through May to place both events, which is plenty for two gilts if you stagger them — and staggering them is why the second one waits until November.
Fifty-four birds and a shed that emails me
The poultry side, also Parrish, is smaller news but it runs every single day. Fifty-four birds, a quarter pound of feed each, thirteen and a half pounds out the door daily. That number used to be a guess. In August we finally measured it against real bag data instead of estimating it, which is the difference between a feed budget and a feed hope.
The feed shed now tells on itself. There's an hourly check that emails me when any feed drops to its reorder point, or when it can't reach the tracker at all. It has been running since late August. The buttons that let whoever is doing chores log a feeding without a phone come in October, and the goal is to have the whole thing operable by the first of November.
What went wrong this month
Two things, both small, both worth writing down.
On the 31st we did a feed run. The Game Bird delivery went in the system and the quantity did not stick — ten forty-pound bags should have moved four hundred pounds of stock and it moved about fifty. The system noticed the gap and, having no better explanation, filed it as drift. Drift is supposed to be the measure of how wrong my model is. Now the first real reading I have on that feed is measuring a typo instead. I know what it is; I don't yet know what caused it, and I'd rather say that than pretend it's solved.
The other one is a single bag of hog grower. The tracker says eleven, the run sheet says twelve. Nobody bought a bag, so one of the two counts is wrong. Fifty pounds, about fifteen dollars. It gets settled at the count on the 5th, because the alternative is finding it in three months as a mystery.
The truck has to work first
Everything at the Tennessee place will arrive on a truck, and ours is a 1992 Chevy sitting in the Florida yard with its dash out on the bench. There is a long list of things that are free to do right now and miserable to do once that dash goes back in — the ground stud, the conduit runs, fifteen measurements you simply cannot take through a reassembled dashboard.
All of it is on hold until the 5th. That's the compression test, and those numbers decide whether this engine is worth the money we were about to spend on it. Doing the pretty work first and then finding out the motor is tired is how you end up with a beautiful dash on a truck that won't pull a trailer. So it sits, deliberately, for five more days.
The fence that would have been fifteen feet wrong
Up in Copperhill we've been walking the property line looking for the original monuments. Found three in July. All three sit fourteen to eighteen feet off where the mapping app puts the corner — and, more tellingly, all three are off in the same direction by about the same amount. That is not three mistakes. That is one shifted layer.
Had we built to the app, the fence would have been fifteen feet wrong down its entire length, and we would have found out from a neighbor rather than a survey pin. We're hunting a fourth monument to turn a strong pattern into something I'd stake a fence line on. Slow, and cheaper than the alternative.
Next month: whether she took, and what the compression numbers said.
— Dan