#93 We Estimate the Size of the Harvest

A mid-season estimate that rarely comes close to the final number.

Last week’s Substack #92 described how grapes progress in size from “bloom” in late May to harvest in mid-to-late October. By mid-July, we hope the berries have progressed enough in size to predict the size of the upcoming harvest.

At this stage we can see how many clusters of grapes each vine has produced. So, it would seem like a simple exercise to count berries; count clusters; count vines and then do the arithmetic to come up with calculate a harvest estimate.

Unfortunately, there are just too many vines, too many clusters and too many berries to count them all. This year I suggested that two of us—that our Vineyard Manager Kees Stapel (pictured below) and I—do our own independent estimates and then compare our results.

Kees (pronounced “case”) and I often confront problems like this. What should we do about this issue or that? And it has become apparent that our minds to not work the same way. We often arrive at the same answer, but via completely different routes.

And this is just how it went with the harvest estimate.

I did it my way and he did it his.

Over the 15 years that Kees has been the Vineyard Manager, he has looked at vines and grapes closely while leaf-pulling, hedging, spraying, grape-hoeing. He has spent literally several thousands of hours in a tractor and on the ground working with vines and grapes.

So, when doing the fruit estimate he relies on a more intuitive sense of the amount fruit out there. He also has 20 years of experience looking at vines and then experiencing the actual harvest.

My approach, on the other hand, is different. I have spent nowhere near the time that Kees has driving a tractor through row-after-row of vines. So, I cannot use intuition and experience to estimate the amount of fruit.

I do it the more mechanical way. I count the number of clusters on a random selection of ten different vines for each varietal. While counting them, I note the variation in size and cut off ten clusters that intuitively seem representative of the clusters I have counted.

I bring each group of ten clusters into the office, weigh them to get an average weight per cluster. I then multiply by the average number of clusters per vine, and then by the number of vines per acre for that variety. That gives me a total weight for all the berries in a section of the vineyard.

Before doing this exercise our business manager, Chelsea Stamp, has projected that we will need 105 tons of Riesling to meet our sales forecast for Riesling wine in the next, 12-month period. If we look like we are not going to get 105 tons, then we will have to find another grower to sell us the additional Riesling fruit.

So what were our estimates?

Kees’s estimate came in at 101 tons
Mine came in at 100 tons.

We did not expect the estimates to be so close. Either both of us will be accurate or both way off when the final numbers are counted in October.

This is also a little short of the 105 tons that Chelsea says we need. But our estimates definitely come with a “fudge factor;” plus or minus 10 percent. We all feel like we are in good shape for the volume of Riesling this year. We have also had a very good season so far, with regular periods of moderate rain and also temperatures that have been warmer than usual.

---

For those of you doing the math at home for this fruit estimate, there is a little more to this.

Both Kees and I have to estimate what fraction of the final berry size and weight we are evaluating at this stage of the season. (See Substack #92, All the Vineyard’s a Stage.)

I can weigh the sample berries and count the total clusters and vines. If the berries are 50 percent of their final weight, I will get one estimate. But if they are only at 40 percent of their final weight I will get different final estimate. This percentage-of-final-size is not something that it is possible to know precisely. Thus, the fudge-factor of ten percent.

We also know what our harvests have been in previous years. There can be dramatic variations, so that can be only so helpful. Often the current year’s harvest is affected by the size of the previous year’s. A very large harvest in one year may indicate a smaller harvest the following year.

The reverse can be true. We had a very small harvest in 2016 and then a large harvest in 2017. We had a small harvest in 2020, and then a large one ins 2021.

One other factor is the age of the vines. We have been planting new vines every two-three years. Newly-planted vines do not produce a full yield until Year 6 or 7. So, as the vines mature, they will gradually produce more and more fruit. When estimating the yield of Block A or B, we need to adjust for the estimate based on the age of the vines.

Also, the quality of the soil can vary from place to place. We have generally uniform soils, but there are specific spots that we know where the vines are more likely to struggle and yield less fruit than the rest of the vineyard. We have to take that into consideration.

So we take all this into consideration as we arrive at a final number. There is likely some underlying science to this, but it is still mostly “art and intuition.”