"Joe Giller" > wrote in message
...
> Tom S, et al.,
> You might be right Tom,
> My sampling procedure was this: Pick a berry from random vines across
> the plot (about 15 total), crush them all in a sandwich bag, take a
> reading with a refractometer.

I suspected as much. Individual berry sampling _can_ give good results, but
you have to have a much larger sample (~200 or more) and be sure to pick
berries that are randomly representative - not just the ones that look nice.
IOW, pick both sides of the vines, each vine, high and low on the trellis,
big and small, and high and low on the clusters.

It's far easier to just pick whole clusters, using similar parameters. Of
course this will _not_ impact your final yield, since you'll ferment the
samples in advance of the actual picking and add them into the fermenters.

Tom S