Apples to Apples

Establishing an apple orchard is a major financial investment. Growers decide which varieties to grow long before they have fruit ready for sale. While much is known about consumer preferences for varieties of apples for fresh use, this is not the case for hard cider apple varieties. Do consumers want sweet hard ciders? Floral notes? The complex flavors and aromas resulting from high levels of tannins? The goal of the Apples to Apples Project and the Comparing Apples to Apples report was to help hard cider growers—and makers—select apple varieties for a high-quality, artisanal product. To accomplish this, researchers compared the results of laboratory analysis and tastings of single-varietal ciders to identify desirable flavor characteristics.

Apples of the same variety from different trees and different orchards may look and taste different.

The full 48pp report Comparing Apples to Apples  profiles the various cider apples used, how the cider was made, laboratory analysis, data analysis and findings from several tasting sessions. Orchard owners and cider makers were on the research team from Albion Prairie Farm, Brix Cider, The Cider Farm and Cider House of Wisconsin. The report was authored by Dr. Julie Dawson, UW-Horticulture; Matt Raboin, Brix Cider; Nicholas Smith, UW-Food Science; and CIAS staff Michelle Miller, Eleanor Voigt, Cris Carusi and Ruth McNair.

Growers from four orchards near Madison, Wisconsin provided 41 varieties of hard cider apples in the fall of 2017 for testing. Staff at the UW-Madison Food Science Fermentation Sciences Program pressed each variety separately and produced about a gallon of single-varietal cider. Each single-varietal hard cider was tested in two ways to determine taste characteristics: first in the laboratory and then by at least one panel of tasters at four separate tasting events held during the winter of 2017-18.

Ciders made with a standard method but different apple varieties will look different and have unique flavor profiles.

Laboratory analysis included pH, titratable acidity (TA), degrees Brix (°Brix) and phenolics. Most of the varieties fell within ideal ranges for pH and °Brix; fewer fell within ideal ranges for TA and phenolics. Cider makers commonly blend apple juices to attain a product in the ideal range. Apples that are outside of the ideal range are useful in adjusting a blend. The initial and most comprehensive tasting provided baseline data on all varieties. The subsequent three tastings aimed to collect more specific data on flavor qualities, allowing comparison of flavors and preferences across different tastings. The tasting results were analyzed on four flavor components: acidity, sweetness, astringency and bitterness. The researchers also collected qualitative data on cider flavor, mouthfeel and strength.

It is easier and less expensive to analyze cider chemistry in the laboratory than to gather flavor and other sensory data. Furthermore, taste can vary from year to year, orchard to orchard, and taster to taster. To address this, project researchers used a multivariate analysis to explore how closely laboratory measurements of variables such as pH and phenolics reflected the tasters’ perceptions of characteristics such as acidity and bitterness.



Participants in tastings were assigned random individual ciders to taste.

As expected, phenolics were closely related to tasters’ perceptions of astringency and bitterness. °Brix was significantly correlated with bitterness and perceived strength (alcohol). Titratable acidity (TA) and pH were equally correlated with perceived acidity. The perception of acidity in the tasting trials traced closely to the laboratory results for all but one apple variety. The ciders rated highest overall by the tasting panels were predominantly perceived as sweet or balanced in flavor, with preferences leaning towards higher perceived sweetness and acidity. The highest-rated ciders were negatively correlated with phenolics, which were closely associated with bitter and astringent flavors. The lowest-rated ciders had high levels of TA, °Brix and/or phenolics.


Relationship of measured laboratory characteristics of apple juices and flavor components in the resulting hard ciders

Laboratory-measured variables are shown in blue, and sensory variables are shown in orange. The direction of each
arrow shows the direction in which that trait increases. The angles between the arrows show the correlation among the
variables. The length of the arrow corresponds to the importance of that factor in explaining differences among varieties,
with longer arrows representing a higher contribution to variation. The dashed horizontal and vertical lines represent the
two dimensions resulting from the Multiple Factor Analysis.

Principle Component Analysis (PCA) allows us to display data on multiple variables and varieties at once. It takes a dataset with multiple variables, finds the most important sources of variation, and creates a two-dimensional map to represent as much of the variation as possible, showing their relationships. Variety placement corresponds to the levels of each of the variables for that variety.

Relationship of single-varietal ciders in terms of the laboratory and flavor components

On the figure above, the position of each variety shows its relationship to other varieties and the variables on the first graph. For example, apple varieties found in the top right are more phenolic, and in the bottom right are more acidic. Varieties around the perimeter of the graph show more pronounced differences, while those clustered in the center show a more balanced contribution of all variables. The dashed horizontal and vertical lines represent the two dimensions resulting from the Multiple Factor Analysis.

Tasters with the Dawson Lab’s Seed to Kitchen Collaborative participated in the first and most comprehensive tasting.

To more accurately describe the chemical components and taste perception of hard ciders, multiple years of data collected from more locations are needed. Long-term trials in which the same apple varieties are gathered from the same farms, and brewed and tested over multiple years, would improve the accuracy of results. A larger, more segmented taste test is also necessary to understand consumer preferences. In a larger study, researchers may also gain a better understanding of how soil types, microclimates, weather or production practices might contribute to terroir, also known as “taste of place.”

This project was sponsored by the United States Department of Agriculture Sustainable Agriculture Research and Education (SARE) program in the North Central Region, project number ONC17-030, the David S. Bourne Foundation, and the People of the State of Wisconsin.