Final report for FNE23-066
Project Information
The purpose of this project was to apply replicated experimental evolution to wild yeast collected from Rogers Orchards to develop strains capable of producing a unique, terroir-driven cider from dessert apples without the faults commonly associated with wild fermentation. We collected wild yeast communities using the pied de cuve method, placing fresh apple juice beneath orchard trees to capture locally adapted strains, and transferred these communities to UConn's Department of Ecology and Evolutionary Biology for selection experiments. Using 96-well plates with 88 replicates across two treatments, control and high alcohol tolerance, we selected for yeast that performed well in sterilized evercrisp juice over 14 days, then isolated and evaluated 24 strains through successive rounds of taste testing across multiple juice types and volumes, ultimately identifying four top-performing strains.
These four strains, when used in combination, produced ciders with distinctive tropical flavor profiles, with notable banana and coconut notes, and successfully fermented dessert apple varieties including evercrisp, honeycrisp, empire, mcintosh, and pink lady. Ethanol-evolved strains produced higher average alcohol content (3.08% ABV) compared to control strains (2.12% ABV) at the small-scale testing phase. At commercial scale, approximately 250 gallons of cider were produced, fermented at 60°F, aged for two months, and kegged at a final alcohol content of around 6% ABV. The resulting cider, named Kanavu, sold over 400 servings and generated approximately $2,000 in revenue during roughly one month of sales at Long View Ciderhouse. We consider the project a success in demonstrating that evolutionary biology tools can produce predictable, flavorful wild-fermented ciders, though further work is needed to preserve the early-fermentation tropical flavors as the cider reaches full dryness. We shared our results at CiderCon 2026, where we received strong interest from fellow cidermakers and farmers, including praise from renowned cidermaker Tom Oliver, and we plan to continue educating both producers and consumers about this approach through educational videos.
This project seeks to:
- Establish a proper methodology for the collection, selection and propagation of wild yeast populations from both agricultural and non-agricultural environments using replicated experimental evolution.
- Discover a number of previously unknown wild yeast strains with unique and appealing sensory attributes and selected for resistance from common faults in cider.
- Determine if wild yeast strains can successfully ferment abundant dessert apples, rather than the more traditional tannic cider apples.
- Select for wild yeast strains from Rogers Orchard, creating a unique cider that is true to the terroir- made with apples and yeast from the same orchard
Questions:
- What range of diversity among wild yeast populations can we expect from the various geographical locations and farming practices from which we will be collecting samples?
- How much variability in perceived sensory attributes is available to cider makers through utilizing wild yeast in their ciders?
- Using the tools of replicated experimental evolution, to what extent will we be able to select for advantageous traits, such as desired aromatics and flavors? Conversely, to what extent will we be able to deprive future generations of yeast of disadvantageous traits?
- What best practices in wild fermentation methods can be learned from the evolutionary tools in this study that can then be replicated by cider makers in the future without access to such tools?
Problem
Cider makers have long been pitching commercial yeasts as the safer route when making cider, especially on a commercial scale. However, cider makers also strive to distinguish themselves and create signature flavors unique to their region that are enjoyed for their complexity. Many commercial cider makers “wild ferment” their ciders using the native yeast in their orchards, processing equipment or fermentation facilities, however the process is also well known to be fraught with challenges, prone to faults due to yeast nutrient deficiencies causing hydrogen sulfide, and spoiled batches from acetobacter and brettanomyces infections.
Proposed Solution
With Swapna’s expertise in replicated experimental evolution and tools available to her in the Department of Ecology and Evolutionary Biology at UConn we propose to broaden what is currently known about using wild fermentations in cider making through a rigorously scientific process. We will start by collecting diverse starting colonies of yeast through PDC fermentations and select among them strains with advantageous traits. By repeating this process over time, with test trials and experimental evolution laboratory methods, our aim is to select for yeast strains capable of fermenting cider in ways measurably beneficial both with respect to cellar practices and (of course) flavor.
Cooperators
- (Researcher)
- - Technical Advisor (Researcher)
Research
Cider making is a complex process that draws on centuries of tradition, artistry and science. In short, ripe apples are ground to a pomace and pressed to extract the juice. Fermentation can then be initiated spontaneously or through inoculating with commercial yeast. Once primary fermentation is complete the cider is racked to a second vessel to undergo a second fermentation and further age. Attention is paid to the sugar, acid and tannin levels of the apples and timing of harvesting. Measurements of brix, specific gravity and pH are among the most commonly tracked metrics throughout the fermentation process. Yeast nutrients, pectic enzymes, clarifying agents and other adjuncts may be used in the process but are not mandatory. The aging process is shaped to a large extent by the amount of oxygen contact with which the cider comes in contact; stainless steel prevents oxygen while plastic and oak barrels invite more contact.
The pied de cuve method will yield a starting culture of yeast that is highly diverse and contains a large number of local yeast strains- that may or may not yield successful hard ciders. This process requires outdoor temperatures in the range of 50 to 65 degrees and occur over a period of no more than 72 hours. Four liters of unfermented cider (must) was transferred to a sanitized HDPE bucket. We then covered the bucket in a fine cheese cloth adhered by tape. The bucket was then placed in the desired apple growing location, specifically in the understory of the tree. We used this method at Rogers Orchards, which uses EcoCertified spraying methods (including fungicides which inhibit native yeast colonies) and an unsprayed apple tree outside of Rogers Orchard. We used this method on different varieties of apples in the West Lot of the orchard, including honeycrisp, gala, and pear trees. In the field, we used a juicer to juice apples from a tree, and placed the juice under the same tree, to ensure that the yeast colonizing the juice will more likely be evolved, or locally adapted, to the same variety of apple. Between each tree, we sterilized the juicer using vodka to avoid cross contamination of yeast. We then allowed wild yeast to colonize the must for a week in consistent temperatures, before sampling.
To go from this highly diverse starting community to yeast strains that can reliably yield delicious ciders, we conducted selection experiments using sterile technique.
We took the starting cultures from pied de cuve and inoculate them into YPD media with penicillin-streptomycin. YPD media consists of 1% Yeast Extract, 2% Peptone from animal tissue, 10% dextrose solution (20% glucose), and 2% penicillin-streptomycin solution. This media is autoclaved at 250℉ at high pressure to ensure that it is sterile of all outside bacteria and yeast. The dextrose solution is vacuum-filter sterilized and then added to the media, and penicillin-streptomycin is added to kill any bacteria. This ensured that our strains do not produce acetic acid due to Acetobacter. After growing in 500 µL YPD media for 24 hours, we froze the initial pied de cuve communities by adding glycerol to make a 40% glycerol yeast solution in YPD and placing them in a laboratory grade -80℃ freezer. This freezing method will continue to be used in these methods as yeast can be stored indefinitely using this method and will stay alive in the freezer and can be revived at any time.
Based on the taste and smell of the pied de cuves, we decided to focus on the honeycrisp pied de cuve. We then took this pied de cuve community and put them through a series of replicated evolution experiments in 0.2 µm vacuum-filter sterilized evercrisp apple must, pressed 2 days before sterilization. We then inoculated pied de cuve yeast community into the sterile evercrisp must in 96 well plates. 88 wells will receive the yeast community, and 8 wells in each plate do not contain yeast as negative controls to detect any contamination. Replicating the experiment 88 times will ensure that we do not lose advantageous yeast strains due to random chance. We conducted two different experiments in this manner, one that simply selects for yeast that perform well in the evercrisp juice, and one that selects for yeast that perform well in evercrisp juice with 5% ethanol added, for higher alcohol tolerance. Both will also be subject to low temperature tolerance, as fermentations conducted at 50℉ yield better results. The experiments will continue for 2 weeks, the amount of time that Rogers Orchards keeps their yeast in fermented cider.
After the selection experiments, each community had a much smaller amount of yeast strains that have survived the selection process. We took each community and used streaking on YPD agar plates to isolate yeast strains. YPD agar contains the same elements as YPD media, with an additional 1.6% agar added to make the media solidify. This streaking method isolates single cells of yeast that will then replicate themselves to create a colony on the plate that is one yeast strain. In both of the treatments, we chose 12 random wells and used streaking to isolate the most abundant yeast strain, yielding 12 yeast strains with traits advantageous for cider making with dessert apples and 12 yeast strains with higher alcohol tolerance.
The yeast strains isolated from the selection experiments were then tasted, pruned, and scaled up in different types of juice, honeycrisp, evercrisp, and a blend. We then tasted and pruned again, until we determined which strains tasted the best, and scaled them up in isolation and combination. We then pressed a blend of apples to fill a tote (around 250 gallons), and inoculated the tote with the best performing strains in combination in YPD media which was then tested by the customers of Long View Ciderhouse.
In October 2023, we initiated an extensive study of wild yeast communities using the pied de cuve method across multiple orchard locations. Our approach involved collecting fresh juice samples, maintaining strict sterilization protocols between collections, and storing the samples in sterile buckets covered with sterile cheesecloth. The initial phase of the project focused on careful preparation, including gathering and sanitizing materials, and developing a robust methodology for field collection.
Field research was conducted on October 2nd and 3rd, 2023, spanning various orchards and apple-growing modalities, including wild and unsprayed conditions. We collected two sets of wild yeast samples: the first after 3-4 days of pied de cuve fermentation (October 6th), and the second after 8-9 days (October 11th). We identified pied de cuves as potentially interesting based on taste and smell, and honed in on the Honeycrisp pied de cuve from Rogers Orchards, which was sprayed. All the pied de cuve samples were transferred to UConn laboratories, where they were preserved in rich yeast media with glycerol for long-term storage at -80°C. We also stored one set of samples with penicillin streptomycin added, to remove any bacterium that were undesirable and preserve just the yeast. To provide a visual reference for our work, we streaked the samples on plates.
2024: Following collection, we conducted evolution experiments with the pied de cuve samples using sterilized Evercrisp juice. The experiment included two conditions: one with added ethanol and one without, running for 14 days while tracking fermentation progress. We successfully isolated 12 strains from each evolution experiment, resulting in 24 total evolved strains. These strains were then tested in 15 mL of sterile Evercrisp juice and fermented for two weeks. Through taste testing and alcohol content measurements, we found that ethanol-evolved strains produced higher average ethanol content (control average: 2.12% abv, ethanol evolved average: 3.08% abv), though this seemed to come at the expense of taste quality. From these tests, we identified six particularly interesting strains based on their flavor profiles, three from the control group and three from the ethanol evolution group. We had initially thought that testing for off characteristics such as H2S production, but determined that taste is the most important metric for cidermaking and focused on that.
We further evaluated these six evolved and identified strains using five different juice types: sterile Evercrisp, Honeycrisp, and cider blend, as well as unsterile Honeycrisp and cider blend. Using an optimized cell concentration in 50 mL volumes, we monitored fermentation and conducted taste tests across all 30 samples. A key finding was that unsterilized juices developed vinegar characteristics, providing valuable insight for future process optimization. Through this comprehensive testing, we identified four strains that performed consistently well across all juice varieties. We were surprised and excited to find that various strains, even when evolved to the same condition, had different flavor profiles, and when put in different juices resulted in different flavors also.
2025: After identifying these promising four strains, two evolved to higher ethanol and two evolved to the plain evercrisp juice control, we scaled up to gallon sized jugs of unsterilized fresh pressed juice in March of 2025. To conduct gallon sized experiments, Swapna incubated the four strains of yeast in 1L each of YPD media for seven days, to ensure that the cultured wild yeast we isolated and evolved were robust enough to outcompete any other wild yeast and bacteria already present in the juice.
After culturing for seven days, on March 7th Swapna used a centrifuge to separate the yeast from the YPD media, and we prepared 8 one gallon glass jugs with a juice blend of Empire, McIntosh, and Pink Lady apples that were pressed the previous day. We used these 8 gallon jugs to conduct 8 different experiments. One jug had no yeast, four had only one yeast isolate, one had the two isolates evolved to ethanol, one with the two isolates evolved to the control juice, and one with all of the isolates mixed together.
We tasted these gallons after 6 days (3/13/25) and found that fermentation was occurring in all the experiments except for the wild, and the gallons with ethanol evolved strains had more active fermentation.
After 13 days of fermentation (3/20/25, the color had visibly lightened and cleared up, and we detected really interesting tropical flavors at this stage- very reminiscent of bananas and coconut. The wild yeast control tasted foul at this stage.
A day later on March 21, 2025, Jeff put the cider without the lees into high pressure bottles to hopefully preserve the flavors and for it to finish fermenting in the bottle.
On May 9th, Swapna prepared all four strains of yeast again in one liter bottles of YPD media to grow and inoculate a larger batch. On May 13th, we pressed a blend of evercrisp, macintosh, and pink lady apples to fill up around half a tote with cider.
We then opened the bottles to taste the finished cider- to find that there was massive pressure built up in the bottles, spraying a bunch of cider everywhere upon opening. Fortunately, we had just enough for Swapna, Daniel Bolnick (Swapna’s Ph.D. advisor), Emma Choi (a student in Swapna’s lab), and Jeff, to taste and decide on the combination of strains that tasted the best. The tropical flavors were preserved very well, and we received positive reviews about the taste and the uniqueness of the taste.
Based on this feedback, we inoculated the half tote (~250 gallons) with all four strains of yeast, we found that the flavors were the best when all four worked in concert.
Jeff then took the tote to a temperature controlled area and fermented it at 60 degrees fahrenheit. Once primary fermentation finished, he racked it (transferred to another tote and left behind the sediment). He continued to let the cider age for two months, occasionally topping off the tote with argon gas to remove the oxygen headspace and prevent oxidation of the cider. Once the cider was at a sufficient clarity Jeff transferred it to a bright tank and force carbonated the bright tank with carbon dioxide.
He then transferred the carbonated cider to kegs, again leaving behind any sediment, with a final yield of 250 gallons of cider. The kegs have since been kept in cold storage, and reached a final alcohol content of around 6%.
Before selling, we had to name the cider, and decided on Kanavu. This is a word from the language Tamil, and means dream, which is also what Swapna’s name means in Sanskrit, and also reflects that this was Jeff and Swapna’s dream to create a wild yeast cider!
Starting October, Jeff served the cider to customers. The ciderhouse closed in November, so Kanavu was on sale for around a month. During that time, Jeff sold over 400 servings of Kanavu and had about $2000 in sales. Jeff found that the concept was slightly difficult to explain, and that education about the process would be a worthy endeavor to explain to other cidermakers and customers.
The final product tasted slightly different from the bottled product because it was left to reach dryness rather than bottled after primary fermentation (due to the explosion factor). We believe that maybe this yeast takes longer to reach dryness, and we may need to treat the cider differently to preserve the flavors the wild yeast gives. Jeff plans on trying finer filtration, and potentially using a cuve, which is just the cider sediment that cidermakers replenish with juice. Jeff also plans on attempting canning of the cider in the early stages of fermentation, to preserve the early tropical flavors and sweetness that the wild yeast creates.
We took this cider during early fermentation to CiderCon 2026 so other cidermakers can taste the flavors brought by the wild yeast and can advise us on how best to preserve those flavors. We received a great amount of interest from other cidermakers and farmers, including the renowned Tom Oliver, who praised the cider and the research. Based on this feedback, Swapna plans on continuing researching wild yeast and cidermaking.
These results show very positive trends for using evolution to select for yeast optimized for cider production, and addresses both the issues of only using white wine yeasts on a commercial scale for cider as well as the unpredictability of wild yeast low intervention fermentation. Jeff said that this cider brought a unique new product to his profile of ciders, and was excited about the concept that this was a cider made with apples AND yeast from their orchard. We believe this could be a great way for cidermakers and orchard owners to create unique, but predictable, ciders without wasting product on failed attempts at natural ciders.
We sought to determine whether replicated experimental evolution, a tool from academic evolutionary biology, could be applied to wild yeast collection and selection to produce a commercially viable cider. Our objectives were to establish a replicable methodology for collecting and propagating wild yeast, discover strains with appealing sensory attributes and resistance to common faults, test whether wild yeast could ferment dessert apples (rather than traditional tannic cider varieties), and create a terroir-driven cider using apples and yeast from the same orchard at Rogers Orchards in Connecticut.
Beginning in October 2023, we collected wild yeast communities using the pied de cuve method across multiple orchard sites, placing fresh juice beneath trees to capture locally adapted strains. We focused on a honeycrisp pied de cuve from Rogers Orchards and subjected it to selection experiments in sterilized evercrisp apple juice using 96-well plates with 88 replicates per treatment. Two selection conditions were tested: a control environment and one supplemented with 5% ethanol to select for alcohol tolerance. After 14 days, we isolated 24 strains, which we evaluated through successive rounds of taste testing across five juice types, sterile and unsterile honeycrisp, evercrisp, and blends. We narrowed down to four high-performing strains, two from each treatment, that produced consistently appealing flavors.
In March 2025, we scaled from laboratory volumes to gallon-sized fermentations in unsterilized fresh-pressed juice. These trials revealed striking tropical flavor profiles including banana and coconut, and confirmed that our evolved strains could outcompete ambient microorganisms without sterilization. We then scaled to approximately 250 gallons of fresh-pressed cider inoculated with all four strains working in concert. The cider was fermented at 60°F, aged for two months, force-carbonated, and kegged at roughly 6% ABV.
We met each of our core objectives. We established a viable, replicable methodology combining pied de cuve collection with laboratory evolution and sensory-guided selection. We discovered strains with unique flavor attributes not achievable with commercial yeasts, and we confirmed that wild yeast can successfully ferment dessert apple varieties. Most importantly, we produced a terroir cider, named "Kanavu", made entirely with apples AND yeast from Rogers Orchards.
The commercial results were encouraging. In approximately one month of sales at Long View Ciderhouse, Kanavu sold over 400 servings generating about $2,000 in revenue. The cider also received strong praise at CiderCon 2026 from fellow producers, including renowned English cidermaker Tom Oliver. Jeff noted that Kanavu added a genuinely novel product to his lineup and reinforced the orchard's identity.
Going forward, we plan to experiment with finer filtration, cuve replenishment techniques, and early-stage canning to better preserve the distinctive mid-fermentation character. We also recognize that educating customers about the process is an important part of marketing this style of cider. Overall, this project demonstrates that evolutionary biology tools can bridge the gap between the unpredictability of wild fermentation and the uniformity of commercial yeast, offering cidermakers a path to unique, terroir-expressive products without the risk of spoiled batches.
Education & outreach activities and participation summary
Participation summary:
In Fall 2023 we lead a 20 person presentation and feedback session where we hosted other apple growers, cidermakers, winemakers and cider enthusiasts at Long View Ciderhouse.
In Winter 2024 we presented our initial research at the American Cider Association annual CiderCon event where our research poster was displayed to over 800 attendees comprising of apple growers, cider makers and academics specializing in apple growing, fermentation and oenology. Attendees traveled to the event, held in Portland OR, from around the country and around the world. We received helpful feedback and broad interest in our research.
In Fall 2025 Long View Ciderhouse served the results of our research, a cider we came to name "Kanavu" to the general public. The feedback was great and has encouraged us to develop the batch into larger commercial volumes.
In Winter 2026 we presented our findings through another CiderCon poster session on display to the over 800 attendees. We also presented the work in a 30 minute presentation and Q&A where we spoke in front of dozens of esteemed colleagues from the American Cider Association, the Cider Institute of North America and veterans in the cider industry who collectively hold decades of industry knowledge. Upon presenting the work, Tom Oliver --an award winning cidermaker of over 30 years who works strictly with wild fermentation in Herefordshire, England spoke to us at length about the project and asked if we could apply the research to his orchard. Gregory Peck, Professor at Cornell College of Agriculture and Life Sciences expressed his enthusiasm for the research and inquired if we intended to make a commercial product. Elizabeth Ryan from Hudson Valley Farmhouse called the work "very very exciting...and for a lot of us you're at the tip of the spear of a lot of questions."
Learning Outcomes
Following our presentation at CiderCon 2026, we were struck by the overwhelmingly positive response from the cider community, both for the quality of the cider (samples were distributed) and the new ideas raised by the research and methodology. Dozens of orchard-based cidermakers approached us asking whether this could be applied to their own operations.
One finding that generated the most surprise and curiosity was the degree of diversity we discovered within a single farm. Yeast populations varied not just across different orchard blocks, but between apple varieties growing in the same general location. The degree to which this diversity impacted flavor was unexpected even to us.
For apple grower-cidermakers in particular, this sparked a fundamental shift in perspective: the realization that every orchard already contains a reservoir of unique, untapped yeast cultivars — distinctive flavor engines that are native to that specific place and have never been identified, isolated, or used. These strains exist on every farm and always have but most producers simply didn't know to look for them.
Perhaps most significantly, this research challenges a deeply held ethos within the wine and cider industries. The "low intervention" approach to wild fermentation ("hands off" and "let nature decide") has long been treated as a kind of philosophical virtue. What our work suggests is that rigor and wildness are not opposites. You can pursue the unique, place-specific character of your orchard with scientific intentionality. Curiosity, methodology, and a willingness to intervene at the right moments may actually get you closer to what wild fermentation promises than simply stepping back and hoping for the best.
Project Outcomes
As a result of this grant I am, I would say, less afraid of conducting scientific research in my cidermaking practice. As Edward O. Wilson put it, "The love of complexity without reductionism makes art; the love of complexity with reductionism makes science.” We somehow found a middle ground. The variables inherent to cider and winemaking are so numerous and interconnected (harvest variation, soil nitrogen, apple variety, fermentation kinetics, pH and total acidity of the must --to name a few) that a demand for perfect experimental control would have stopped us before we started. Instead of paralysis, Swapna and I talked through the complexities in our initial design phase, respected them, and then set them aside just enough to act. The result was a type of applied science that may have at times appeared imperfect by conventional academic standards but all-in-all actionable by farming ones.
The thing that excites me most about the research is that engagement with this idea doesn't require a laboratory. So even though I can only count myself as someone who changed a practice as a result of the findings, I know that Swapna and I have sparked a greater curiosity and interest among cidermakers based on our sustained conversations and connections. Producers at every level of technical sophistication can find an entry point in this research: whether that means trying a simple pied de cuve to capture local yeast from their own fruit, running informal home breeding trials to select for preferred flavors, or pursuing full laboratory collaboration like my work with Dr. Subramanian. In other words the methodology is scalable and flexible and hopefully broadens other cidermakers interest in exploring and cultivating the incredibly wide range of native yeasts in their orchards.
Our methodology was deliberately accessible: a pied de cuve approach that any orchard-based cidermaker can replicate and a selection process based on personal taste. That accessibility turned out to be a feature, not a limitation. It kept us moving forward rather than waiting for perfect conditions that would never arrive.
The key to our success was the partnership itself. I brought sensory evaluation, orchard knowledge, and cidermaking intuition. Swapna brought evolutionary biology, laboratory rigor, and the ability to isolate and stabilize what to my palate struck me as promising. Neither of us could have done this alone.
The central challenge was one we anticipated but couldn't fully resolve: the sheer complexity of wild fermentation. Our results actually deepened that challenge in a fascinating way. Did we answer the question we set out to study? Yes. We demonstrated that wild yeast can be captured, selected, evolved, and stabilized into reliable, replicable strains with distinctive orchard-specific flavor profiles. But in answering that question, we uncovered a more interesting one. The final cider fermentation we landed on, Kanavu, is not the product of a single dominant yeast strain. It was a blend of four distinct strains working in concert, which points to how complex wild fermentation actually is. A redesigned experiment would instead try to understand and intentionally choreograph the relationships between strains (how they interact, compete, support one another, and collectively shape flavor). To me that feels like the truer picture of what is happening in a wild fermentation, and it opens a much richer set of creative possibilities.
And that complexity is only the yeast side of the story. In order to isolate and study the yeast populations, Swapna used penicillin-streptomycin to kill off all bacterial activity, which means an entire dimension of wild fermentation, the bacterial ecosystem, was effectively off the table for this research. Lactic acid bacteria, acetic acid bacteria --the full microbiome of the orchard, that's a separate and equally fascinating area of study that this project deliberately set aside and that future research could meaningfully explore.
I'd also want to go wider and deeper into the orchard itself. For example, does the farm's grafted cider-specific apple varieties house meaningfully different yeast populations than those in our culinary apple blocks? I'd absolutely love to continue and expand this work and I believe the producers who stand to benefit most are small and mid-size orchard-based cidermakers, particularly those for whom a spontaneous fermentation program has historically felt out of reach. This approach puts it within reach.
It's my hope that any cidermaker who believes not only that place should express itself in the glass, but that the diversity of our cider and wine offerings should take their cues from the microbial diversity of the orchard itself, will find something useful here.