Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - mar3

Pages: [1]
1
Citrus General Discussion / Analyzing and comparing satsuma sweetness
« on: March 26, 2022, 06:52:39 PM »
The University of California Riverside maintains records of various features such as brix and acidity for a number of satsuma varieties. This is a great resource, and they've made freely available online. I'm sure most of you folks are already aware of it. I was curious about the sweetness of different varieties of satsumas, so I used it to compare listed varieties with trees over around 4 or 5 years old.

The ratio:  soluble solids concentration (SSC) / total acid (TA) is the most common formula that I've seen used, and some states that have a satsuma industry use it to grade their product quality. At least one study has shown that an alternate method called "BrimA" better predicted how much consumers would like a fruit. You calculate BrimA by: SSC - k * TA Here, "k" is some coefficient used to weigh acidity more than SSC, since the idea is that acidity will have a greater impact on flavor. To me, subtracting weighted acidity from SSC makes sense for satsumas. For example, if you have a very low acid satsuma that is not very sweet, it can still score relatively well with SSC / TA (e.g. when it has an TA of ~0.5). But some might say it lacks richness and has a watered down flavor. But by instead using subtraction, you must have a high SSC in order to get a high brimA. If the SSC is high enough, and even if you have acidity, you can still beat out lower acid fruits (getting a sweet + tart flavor). This article mentions California switching to this method in 2012 for grading navals. For this post, I used brimA with a k-value of 3.

UCR's entries have info such as dates, SSC, TA, exterior color, and % juice. Different varieties ripen at different times. I wanted a more objective way to find a "ripe enough" date than something like taking an average of when they happened to pick fruit. I've noticed that many varieties will decline in juiciness as time goes on and that UCR tends to stop picking them when values fell below 40%, roughly. My guess is that after this point, the fruit is perceived to be poorer quality and "dry," despite the fact that sweetness continues to increase and acidity decrease. I did a linear regression on date and juice % and solved for the latest date at which juiciness was still at least above 40%. A couple of varieties never really went above that point, and so I excluded them. A few high quality, late ripening varieties never really fell below 40%, so for those I used the last harvest date in the dataset. These ripeness dates are mostly from Riverside CA. Disclaimer: I'm just guessing that 40% is a good value to use here and don't have any actual firsthand knowledge of it. Please let me know if this needs changing.

I then did a linear regression for brimA against date, as well as for exterior color. I solved for those variables on the ripeness date at which juiciness is probably above 40%. I believe this is a more robust way to compare varieties than simply finding averages, since there's a good deal of variation between fruit. The regression takes all the information and gives a good estimate of what you're likely to see on a given date. I've seen it done on various papers that study fruit quality against other variables.

Unfortunately, I couldn't find a color chart reference on the UCR's website. I think it's a safe bet that probably 0 = green, 10+ = red/orange, and yellow is in the middle somewhere. I've provided this color information since not all varieties have a ripe-looking color at the time that they reach the minimum juiciness threshold.

Note: this only tries to find "sweetness." Taste is subjective, and obviously some people prefer milder satsumas, some more tart, etc. Also, there are many other important features when choosing a tree, such as yield, susceptibility to disease, cold-hardiness, etc. That said, here are the results:
Code: [Select]
BrimA Date Color Variety
(k=3) cutoff   at
cutoff

11.26 December 13 7.56 Okitsu Wase
10.05 December 13 9.23 Frost Owari (AKA "Owari")
9.85 December 13 11.79 Clausellina
9.22 November 27 9.95 Lange #3
9.12 November 14 8.21 Kuno Wase
8.94 November 17 8.83 Miho Wase
8.68 November 17 8.76 Miyagawa
8.39 November 27 9.21 Lange #2
8.25 November 18 6.27 Dart North
8.21 November 17 5.80 Aoshima
7.90 November 10 6.95 China S-9
7.89 November 17 7.42 Xie Shan
7.80 November 18 4.94 Road 164 Satsuma
7.78 November 18 5.50 Dart South
7.65 November 08 7.27 Armstrong
7.63 November 18 5.80 Lange #1
7.49 November 17 6.20 China S-2
7.46 November 17 6.63 Silverhill
7.45 November 17 5.72 Iveriya
7.29 November 12 4.32 Agudzera
7.16 November 17 5.32 China S-7
7.15 November 17 4.90 China S-6
7.10 November 18 6.72 Mc Ewen
7.06 November 10 4.90 Dungan
6.57 November 10 7.34 China S-1
6.24 November 10 7.46 China S-11
6.02 November 10 4.30 China S-3
5.95 October 26 5.56 China S-5
5.83 November 18 6.27 China S-15
5.24 October 07 4.40 China 6-22
4.92 October 16 5.21 China S-18
4.21 September 27 3.76 China S-20
3.89 October 18 4.18 China S-12

Pages: [1]