If you haven't been following Tomatogate—my name for the minor controversy I caused with two previous articles that challenged the commonly held belief that refrigeration is never an acceptable way to store tomatoes—then you may not know just how strongly people feel about the topic.
Let's just say I ticked a lot of people off. Some suggested that my conclusion was proof that I had no idea what a good tomato tastes like. Some said that my methodology (a series of multiple casual blind taste tests that I conducted throughout the summer) was invalid. Others didn't like my tone, accusing me of being both sensationalist with my headline (kinda true), and arrogant (also kinda true—in my defense, I was just trying to have some fun throwing down the gauntlet on a long-held and oft-repeated piece of dogma; clearly I rubbed some folks the wrong way).
A small group of commenters celebrated my findings: They were closet refrigerated-tomato-eaters who felt validated by evidence that what they'd been doing really wasn't as terrible as everyone else insisted.
A lot of people asked legitimate questions, and pointed out shortcomings in the tests I had done. Many wanted more data.
More data meant eating more tomatoes, so it didn't take much arm-twisting to convince me and Kenji that we should load up on the late-season crop and run more rigorous tests. After all, more data is always a good thing, and reproducibility is the foundation of any reliable experimental result. Either my initial observations would hold, or I'd have to revise them or even fully retract them.
Kenji and I split up the tasks. Here on the East Coast, I went to the farmers market and bought a large load of both regular red tomatoes and a variety of heirlooms.* My plan was to do one giant blind tasting in the office, with as many people as I could wrangle, and more quantitative measures (as opposed to the strictly qualitative assessments I had done earlier in the summer). Then I'd follow that with some triangle tests to see just how well tasters could really differentiate between refrigerated and room-temp samples.
* For those who theorized that I had bought crappy tomatoes from a wholesaler at the market, it's worth pointing out that NYC Greenmarket only allows vendors to sell produce they've grown themselves, and that I had talked to the farmers to confirm the tomatoes had not been previously refrigerated.
Meanwhile, over yonder in the Bay Area, Kenji went and picked his own tomatoes straight from the vine, just to remove any lingering question about handling practices of the middle men; he then did his own blind tastings with those tomatoes. He also examined the refrigerated tomatoes for signs of mealiness.
So, what were our results? Would you believe it if I told you that they undermine the never-refrigerate rule even more than I had previously argued? Well, that's what happened.
My East Coast Tests
When I ran my initial series of tests, it was at the height of summer in New York City with temperatures well above 80°F. Without air-conditioning in my apartment, I found that more often than not, refrigerated ripe tomatoes tasted better than ones that continued to sit out on the counter, suggesting that in instances where room-temperature is above roughly 80°F, refrigeration was often preferable.*
* In researching the topic, I found several scientific studies that compared refrigeration to "room-temperature" conditions, in which the room-temp conditions were all below 70°F (colder than many summertime rooms in real life); I didn't find a single study that compared refrigeration to high-temperature storage conditions. I also found no studies that examined truly ripe tomatoes—they seemed to have all been designed with the concerns of large-scale tomato growers, who pick their tomatoes while still green, in mind, and not the concerns of those of us at home.
But when I went out to run this latest round of tests, temperatures in New York had fallen considerably, down to the 60s and 70s. I'll admit, this made me really nervous: My whole argument revolved around very hot summertime conditions, and I had made no claim that the refrigerator was equal to or better than temps in the 70s and below. Frankly, I was still fairly certain that the refrigerator was worse in these lower-temp instances. How could I provide better data if I suddenly didn't have the conditions to replicate my earlier tests?
I went ahead with the tests anyway. I'm glad I did, because they suggest that I was too conservative in my initial analysis.
The Blind Tasting
As I mentioned above, I bought a variety of tomatoes at the farmers market. Most of them were regular red slicing tomatoes, the rest were an assortment of heirlooms. Their quality was variable. I put half of each type of tomato in the refrigerator, the other half out on the counter. The next day, I took the refrigerated ones out and let them come back up to room temperature.
I then cut up each tomato and assigned it a number. I had ten tasters work their way through the samples, each in a different order, to ensure that no single tomato was disadvantaged due to tasters' palate fatigue. Tasters evaluated the tomatoes on a scale of 1 to 10 on four criteria: overall preference, flavor, aroma, and texture. None of the averaged scores diverged significantly from the overall preference score, so we'll look at the overall-preference scores as a proxy for the other criteria scores:
Before going into the specifics, I want to reiterate that this test compared tomatoes stored at low 70-degree temperatures to refrigerated ones, essentially pitting refrigerated tomatoes against much more ideal conditions than my previous tests.
But instead of seeing a clear and decisive difference between refrigerated tomatoes and counter tomatoes, the differences were very small. On average, the counter tomatoes just barely edged out the refrigerated ones, except for the small yellow heirloom tomatoes, in which case the refrigerated ones received the highest average score. That was also the highest average score of all the tomatoes, which means that even when compared to much more ideal conditions, refrigerated tomatoes are capable of coming out on top: absolutely not what we'd expect if refrigeration were really as bad as the common wisdom claims.
Even more interesting is that of all the tomatoes in the tasting, all of us (the 10 tasters plus myself) were unanimous in our agreement that the small yellow tomatoes—the ones that scored highest—were the best.
Meanwhile, the basic red tomatoes were the worst in terms of overall quality, and they also happen to be the ones where the refrigerated tomatoes scored the lowest. This lends further support to my theory that the higher-quality and riper the tomato, the less harm the refrigerator will have on it. Simply put, really good, ripe tomatoes tend to do well in the refrigerator, while bad tomatoes remain bad or get worse in the fridge: under-ripe tomatoes continue to be under-ripe and mealy tomatoes become mealier.
In fact, based on these and my prior tests, I'm increasingly convinced that those who have experienced significant negative effects of refrigeration on their tomatoes are probably buying or growing sub-par tomatoes.
One more very important detail: What the chart above shows are average scores. But within each group, the variance was significant. For the red tomatoes, for instance, I had individual refrigerated samples that scored as high as 5.5, and countertop tomatoes that scored as low as 3.6. So while the countertop tomatoes slightly edged out the refrigerated ones when averaged together, the distribution of individual tomato scores was all over the place, regardless of storage method.
The Triangle Tests
Some commenters on my earlier articles asked whether I had performed a triangle test. I hadn't, and truthfully, I wasn't totally clear on what the advantage of a triangle test would be: I had never claimed that refrigerated tomatoes were going to always be indistinguishable from room-temp ones. In my earlier tests, we were unable to differentiate between refrigerated and not-refrigerated about half the time. But in the other instances, the differences were apparent; it's just that in those cases, we tended to like the refrigerated ones more.
Still, I figured there was no harm in trying a triangle test out. I did a test run on Max one night using more of those not-so-great red tomatoes that I had set aside. Max has a very good palate, so I was curious to see how he would do. In each tasting round, I presented Max with three slices of tomato in random order (either 2 refrigerated and one countertop, or 2 countertop and 1 refrigerated), and his task was to see if he could figure out which of the three was the odd one out. After 12 rounds, Max had correctly identified the odd tomato 6 times, which is slightly better than chance (in a triangle test, random guessing should yield correct answers 1/3 of the time, which in this case would be 4 out of the 12 rounds). When he was able to correctly identify the tomato samples, he also picked the counter sample(s) as his preference.
But when he got it wrong, he sometimes picked refrigerated slices as his preference. This is consistent with the blind tasting results above: even though the red counter tomatoes edged out the refrigerated ones overall, there were individual refrigerated samples within the mix that scored higher than some of the counter samples.
But 12 rounds isn't enough, so the next day I bought even more tomatoes, refrigerated half overnight, and then had five different tasters for a new session of triangle testing, with 24 rounds total. Out of 24 rounds, we'd expect random guessing to be correct 8 times (1/3 the total number of rounds). By the end of my session, my tasters had been correct 9 out of 24 times, performing just a hair above the random-guessing rate.
I'll be honest: As I was slicing the tomatoes for these tests, I felt that the differences were more apparent, but then again, I knew which was which (in some cases, I thought the counter tomatoes were better, in others I thought the refrigerated ones were). What this test shows is that once that knowledge is removed, the differences can be subtle enough that tasters have a very hard time telling refrigerated and countertop tomatoes apart. So while I don't believe that room-temp and refrigerated tomatoes are totally indistinguishable, these tests indicate that the claims of horrible effects of refrigeration on ripe tomatoes are exaggerated to an extreme.
Out in the Bay Area, Kenji also ran his tests with the tomatoes he picked directly off the vine himself. Just as with my most recent tests, Kenji's house is in the low 70s and mid-to-high 60s—theoretically ideal tomato-storage conditions. I'm going to let Kenji tell you in his own words:
The tomatoes that I picked were fully ripe. I held them for two days, half in the fridge, half on the counter. I let the refrigerated tomatoes come back to room temp before tasting, then I did a blind taste test with six people. Of those six, two did simple side-by-side preference tests: They both picked the fridge tomatoes as superior. The other four did triangle tests: Of those, three correctly picked the odd tomato out, and all picked the refrigerated tomatoes as superior. The fourth person who did the triangle test selected the odd one out incorrectly, but she still picked the refrigerated tomato as her favorite.
I also did a test slicing tomatoes open and checking them (subjectively) for mealiness as they warmed up. I didn't notice any mealiness from a fully ripe tomato that had been put in the refrigerator."
Kenji's results support what I suggested above: refrigerated and countertop tomatoes won't always be indistinguishable, but even when they are, the refrigerator isn't by any means guaranteed to be worse.
Our results from these latest tests are frankly as surprising to me as I imagine they are to many of you reading this: Before these tests, I never, ever would have argued that tomatoes kept at a mild 70 degrees could be beaten by or mistaken for refrigerated tomatoes.
Yet here we have multiple tests, performed on two different coasts by two different people with many different varieties of tomato, and that's exactly what we're seeing. It's starting to look like I was overly conservative in my original articles: The refrigerator isn't just preferable to a very hot kitchen, it's often just as good and sometimes even better than a cool kitchen.
On the Value of Science and the Danger of Misusing It
Throughout this tomato-tasting experience, I've reflected quite a bit on the role of science in all of this.
Science itself has done nothing wrong: It's a beautiful system—the best one we've got—for answering questions about how pretty much everything in the observable universe works. But it's easy for us to misuse it, and I think it's just such a misuse that created the conflict over tomato storage in the first place. If you'll bear with me, I'll explain:
As I've written here and in the prior articles, all of the academic studies I found on tomato storage were based on a narrow set of conditions: namely, tomatoes picked when much less ripe, and stored in temperatures below 70 degrees. Those studies concluded—and I'm willing to believe that they're correct—that those tomatoes are harmed by refrigeration and are better stored at slightly higher temperatures in the 50s and 60s. The studies I found didn't examine tomatoes that are picked when fully ripe, and they didn't consider warmer storage temperatures, certainly not above 80°F.
So what happens next? This is just a scenario I've made up, but it's plausible to me and it shows how the farther we get from the original data, the more likely it is that the data is misinterpreted: The big-ag tomato growers follow the research and begin storing their under-ripe tomatoes in cool temperatures (but not as cold as a refrigerator), and wholesalers do the same. Word gets to the produce vendor: Don't refrigerate these tomatoes, it's bad for them. Then the produce vendor tells the customer: Don't refrigerate your tomatoes, it's bad for them.
But that last step is the problem: The don't-refrigerate wisdom was good throughout the supply chain and kept the tomatoes in the best possible condition, but it doesn't necessarily apply to the retail customer with different storage conditions at home, or to tomatoes that are picked when ripe—conditions that had not been tested in any scientific study I've found.
Another example of the misuse of science is in this Alton Brown Facebook directive: "Do me a favor: Never put tomatoes in the refrigerator," he implores. "If they drop below 50 degrees F, a flavor compound called (Z)-3-hexenal is just going to flip itself off like a chemical switch ... permanently." Uh oh, (Z)-3-hexenal gets switched off permanently? That sounds really bad.
Now, I have no idea if that thing about (Z)-3-hexenal is even true, but let's give him the benefit of the doubt that it is. My question is: So what? In a complex biological structure like a tomato, I'm supposed to believe that because one single aroma molecule goes dormant, that's therefore a good enough reason to never refrigerate a tomato? What about all the thousands and thousands of other complex processes that are taking place in a tomato as it ages? How can we possibly draw an actionable conclusion from one single data point?
I might as well tell everyone to stop breathing because air pollution is bad for our lungs.
This is a problem in a lot of science-based journalism: Scientists perform a study and publish their results. Lay publications, looking to make that information relevant to the lay person, try to find some kind of practical advice buried in the findings. Scientists have found that vitamins are important to the human body? Take them as a pill! Scientists have found that fat is bad? Stop eating fat! Scientists have found that fat isn't as bad as they thought, and carbs are the problem? Stop eating carbs!
The problem isn't necessarily with the scientists, it's with the people trying to give concrete advice about how to live and act based on the work of the scientists. It puts us in hot water far too often.
So Where Do We Stand on This?
None of my tests, nor Kenji's, are rigorous enough for publication in any kind of scientific journal, but I think we've each had significant enough results that no one should at this point continue to believe that the no-refrigeration rule is always true. At the very least, the rule exaggerates the harm that a refrigerator does to ripe tomatoes while not considering the sometimes greater harm that can befall that same tomato if left at room temperature—even, it seems, when room-temp is down in the low 70s.
As for me, next time tomato season rolls around, I'll stick them in the fridge when it seems like they're ripe enough. And I won't feel a smidgen of shame about it.