Click on author name to view affiliation information
Expansion of the beer brewing industry in the United States has encouraged hop production outside of the traditional Pacific Northwestern region. In the Midwest, producers grow hop cultivars developed for the Pacific Northwest, but differences in climate cause inconsistent yields and flavor profiles. The objective of this study was to characterize the potential genetic value of 12 hop genotypes to the development of progeny adapted to Nebraska. Progeny of the genotypes were given a visual performance score as observed in the field in June, July, and August from 2019 to 2021. Estimated breeding values for the performance scores (EBVPSs) of the genotypes and their progeny were calculated using a linear mixed model analysis of the unbalanced data collected. Variation among EBVPSs of the genotypes and progeny indicates that there is potential to breed for progeny that are adapted to the Midwest. Four parental genotypes had positive EBVPS for the June, July, and August visual performance scores indicating their progeny performed above average throughout the season. However, some progeny performed below average across monthly ratings, demonstrating the importance of selecting appropriate parents for developing Midwest-adapted cultivars.
The United States is the largest producer of hops (Humulus lupulus L.), with 104.9 million pounds being produced in 2023 (Hop Growers of America 2024). More than 99% of hops produced in the United States are from the Pacific Northwest states of Idaho, Oregon, and Washington (Hop Growers of America 2024) and production in these states has nearly doubled over the past decade because of the expansion of the beer brewing industry (Brewers Association 2021).
In response to national trends, growth of the Midwest craft beer industry, and increased demand by brewers for regionally produced hops, the University of Nebraska-Lincoln initiated a regional hop breeding program to develop locally adapted cultivars using conventional hops breeding. However, because conventional breeding requires evaluating large numbers of progeny at the cost of labor, space, and other resources, it is essential to select parental genotypes that produce progeny containing the desired profile of traits. Typically, selection in conventional hop breeding is based on phenotypic observation of progeny derived from selected parental genotypes. However, selection based on phenotypic observations can be difficult when progeny are evaluated across multiple years and locations, and when parental genotypes, breeding populations, and progeny are represented unequally across the years and locations in which the evaluations were conducted (Henderson 1953). Moreover, phenotypic measurements are a function of both genetic and environmental effects, and selection decisions must be based on relative magnitudes of genetic effects (Bertan et al. 2007). When data are unbalanced, linear mixed model (LMM) analysis based on Henderson’s best linear unbiased predictors (BLUPs) (Henderson 1953, 1975) can be used to calculate estimated breeding values, with these values then being used to make more accurate selection decisions. This method adjusts for unequal number of progeny and/or replicates in the phenotypic data and separates genetic effects from environmental effects (Henderson 1953).
Researchers previously characterized parental genotypes and progeny in the Pacific Northwest for numerous important characteristics such as brewing qualities, storage stability, and yield (Henning and Townsend 2005; Henning et al. 1997a, 1997b). Henning et al. (1997a, 1997b) used historical data of accessions to determine their general combining ability for yield, brewing qualities, and hop storage index. In both studies, the authors identified superior genotypes for use in creating progeny with multiple traits of interest. Henning and Townsend (2005) conducted a field-based study that assessed the brewing qualities and yield of five female and five male parents; they found that the female commercial cultivar Magnum produced progeny with the highest alpha acid content. Henning and Townsend (2005) also observed that progeny of the male accession 21267M exhibited high alpha acid content and yields. Accession 21267M was later released for use in public breeding programs because of its potential value in producing progeny with high yields and alpha acid content (Henning et al. 2018).
Still, little is known about Midwest performance or adaptation of Pacific Northwest cultivars and their progeny. Producers in the Midwest have made various attempts to grow cultivars adapted to the Pacific Northwest in the Midwestern climate. However, environmental differences between the Midwest and the Pacific Northwest often resulted in low yields and adversely altered secondary metabolite profiles that impart flavor, aroma, and antimicrobial properties (Pavlovič et al. 2010).
The objective of this study was to characterize the potential value of 12 parental genotypes and their progeny using EBVPS as an essential first step in breeding for Midwest-adapted hops. Progeny resulting from crosses between six maternal and six paternal genotypes were evaluated for adaptability and performance in eastern Nebraska. Evaluations were conducted in June, July, and August from 2019 to 2021 using a single numeric scale that represented an aggregate value of several agronomic traits.
The maternal genotypes used in this study consisted of five commercial cultivars and one wild hop collected in Lancaster County, NE, USA (Table 1). The paternal genotypes consisted of four accessions retrieved from the US Department of Agriculture, Agricultural Research Service’s (USDA-ARS) National Clonal Germplasm Repository in Corvallis, OR, USA (USDA NRCS 2004), as well as two locally collected wild hops: one collected in Douglas County, NE, and the other with an unknown collection site (Table 1). The parental genotypes were grown in separate male and female nurseries at the University of Nebraska-Lincoln’s East Campus Research Farm in Lincoln, NE, USA (lat. 40.8136°N, long. 96.7026°W). The nurseries were located 320 m apart, with the male nursery sheltered by windbreaks to prevent cross-pollination and facilitate controlled crosses. Nebraska commercial hop growers provided the parental genotype starting material, prioritizing commercial cultivars based on agronomic performance rather than brewing qualities, selecting those that produced high yields and thrived in Nebraska’s unpredictable climate.
Crosses between parental genotypes were made in July and August of 2018 and 2019. In 2018, all possible cross-combinations of selected maternal and paternal genotypes were attempted but only 12 attempted crosses successfully produced progeny (Table 2). Crosses of parental genotypes that failed to produce progeny in 2018 were again made in 2019, with two crosses resulting in progeny (Table 2). The methods used to make crosses and seed treatment were described in full by Alas (2022) and briefly here. Pollen from each paternal genotype was collected in 4¼-inch × 4¾-inch × 2½-inch × 15½-inch water-repellant Canvasback tassel bags (part no. T415, Seedburo Equipment Company, Des Plaines, IL, USA). One bag with pollen originating from one paternal genotype was placed over a lateral branch containing receptive female flowers of one maternal genotype. This process was repeated, with each paternal genotype being crossed multiple times with the set of maternal genotypes. The number of pollinations made with each male parent varied and was dependent on the amount of pollen available. Each year the maternal genotypes were grown in the maternal nursery, with each female being replicated four to six times. Some planned crosses failed due to lack of nick between time of pollen shed of the male and receptiveness of the female flowers. At maturation, bags were collected and seeds from the cones were cleaned and stratified for 8 weeks. Following stratification, seeds were germinated and seedlings were grown and maintained in the greenhouse until transplanting in the progeny nursery in the following year.
Progeny for field evaluations were initially selected based on specific crosses, prioritizing those whose parents were likely to produce high-performing progeny in the Nebraska environment. Due to occasional germination challenges, progeny were subsequently selected based on parental genotypes producing seed with higher viability. Progeny in the greenhouse were not assessed for any specific traits. Field evaluations were based solely on cross-specific selection and germination viability.
Two progeny nurseries, located at the University of Nebraska-Lincoln’s East Campus Research Farm in Lincoln, NE, USA (lat. 40.8136°N, long. 96.7026°W), were used to evaluate progeny. The two nurseries were established parallel to each other, spaced 9 m apart, and arranged in 46-m long rows. The number of nursery rows varied from season to season, depending on the number of new progeny created each year. One of the two nurseries was used each year for evaluations, with nurseries alternating from season to season. New progeny were added to the two nurseries each year, and the progeny evaluated changed accordingly each year. For example, progeny resulting from 2018 crosses were evaluated in one of the progeny nurseries during 2019. Selected progeny were vegetatively propagated and maintained in a greenhouse until being planted in one of the progeny nurseries the following season. Progeny that were not selected were eradicated. The following season, progeny that were selected the previous season along with new progeny were transplanted from the greenhouse to the alternate progeny nursery. The use of alternate nurseries prevented contamination from nonselected genotypes from the previous year. Because hops have extensive rhizome systems, it is important to fallow the nursery to prevent regrowth and ensure the eradication of the nonselected lines before using the nursery again for evaluation. Within each nursery, progeny were arranged using a completely randomized design, with plants spaced 0.3 m apart and two to three bines of each genotype were trained to grow vertically using 2-m-tall bamboo stakes. Biodegradable plastic sheeting was placed on the soil surface to suppress weed growth and drip irrigation provided needed supplemental water.
Progeny were evaluated in 2019, 2020, and 2021. Each year, plants were transplanted to the respective nursery during the first 2 weeks of June. Collection of data for the evaluations began 2 weeks after transplantation. Progeny were not replicated in the nursery during the first year of evaluation. Progeny that had been selected for reevaluation following the first year were replicated from one to five times each year, depending on the space available. Downy mildew occurred naturally in the nursery and the magnitude of resistance was recorded as a percentage of lesions on the leaves. The height of the tallest bine of each plant was visually compared with the 2-m bamboo stake as a measure of vigor. The length of lateral branches and internodes was another measure of vigor, where shorter internodes and longer lateral branches are desirable. Yield was not measured because the tested plants were not grown to full maturity in this study. The selection index for progeny evaluations was based on a visual performance score, which represents a single numerical value for observed performance and agronomic productivity, including resistance to downy mildew, vigor, lateral branching, and internode length. Visual performance scores ranged from one to five, where 1 = poor performance and 5 = exceptional performance. Progeny visual performance was evaluated during the months of June, July, and August. Because the visual performance score resulted from the assessment of multiple traits, the underlying reasons for each genotype’s rating varied from genotype to genotype. For example, a score of 1 could have been assigned to a genotype because the genotype was not vigorous, or because the genotype was highly susceptible to downy mildew.
Data from 684 progeny genotypes, consisting of 572 nonreplicated and 112 replicated genotypes, were analyzed using LMM analysis. Data from progeny that failed to survive the growing season were not included in the analysis. The number of progeny included from each parental genotype ranged from 2 to 490. The number of progeny originating from each parental cross ranged from 1 to 227. The number of replicates for each genotype ranged from one to nine. LMM analyses were used to calculate BLUPs for this dataset having unequal numbers of progeny and progeny replicates across years of evaluations. The BLUPs represent the EBVPSs of both the parental genotypes and their progeny, where parental EBVPSs are based on progeny performance. The EBVPSs were calculated separately for each month to evaluate how progeny performed throughout the growing season. All analyses were executed using Echidna Linear Mixed Model Software (Gilmour 2020) using a statistical model with years specified as fixed effects, and genotypes specified as random effects. Estimated breeding values are calculated and commonly expressed as deviations around a mean of zero and all subsequent references to EBVPSs use this convention.
Twelve parental genotypes were characterized based on EBVPSs to determine their potential for developing progeny adapted to Nebraska. The EBVPS for the June rating of maternal genotypes ranged from −0.436 to 0.225 (Table 3). Paternal genotypes had a similar range of EBVPSs for ratings in June, ranging from −0.395 to 0.430 (Table 4). The maternal genotypes had a widening range in EBVPSs throughout the growing season, with EBVPSs ranging from −0.857 to 0.425 in July and −0.899 to 0.793 in August (Table 3). In contrast, the EBVPSs of paternal genotypes had a narrower range in EBVPSs later in the season, with EBVPSs ranging from −0.122 to 0.088 in July, and −0.212 to 0.323 in August (Table 4). The larger ranges of values for maternal genotypes compared with paternal genotypes might be explained by broader genetic diversity of the maternal genotypes as the females were of commercial and wild origin, while the males were only of wild origin.
Moreover, results showed that the maternal genotypes Galena, Glacier, and Sorachi Ace and paternal genotype PI 635403 all had positive EBVPSs for each of the June, July, and August ratings (Tables 3 and 4). Positive EBVPSs indicate that these four parental genotypes produced progeny that performed above average across all 3 months of the growing season. One maternal genotype, Columbus, and two paternal genotypes, 18NEHOPS026 and PI 635246, had negative EBVPSs for each of the 3 months, June, July, and August (Tables 3 and 4), indicating below average progeny performance across all months. The remaining five parents had EBVPSs that varied across June, July, and August (Tables 3 and 4). The parents with these mixed positive and negative EBVPSs have potential future use when creating breeding populations, but given their inconsistent pattern of performance, should be paired only with parents that together produce progeny with desired traits. In general, these results provide a basis for making decisions about which parents have potential for developing progeny adapted to the Midwest.
The EBVPSs provide evidence on which breeding crosses produced superior progeny for adaptation to Nebraska.
Compared with parental genotypes, progeny of breeding crosses had a wider range of EBVPSs, ranging from −0.499 to 0.357 in June, −0.874 to 0.458 in July, and −0.942 to 1.137 in August (Table 5). The wider range of variation demonstrates the diversity of progeny developed from the crosses. Overall, EBVPSs for August had the widest variation (Table 5), which was expected because by August, all progeny have become established in the field and display the greatest observed differences. In addition, by August, many progeny flowered, allowing for gender-based selection of genotypes to advance to the next stage of the breeding cycle.
The progeny of Galena × PI 635287, Galena × PI 635246, Sorachi Ace × PI 635287, Sorachi Ace × PI 635403, and Sorachi Ace × PI 635246 had positive EBVPSs across all 3 evaluation months (Table 5). Similarly, progeny from 18NEHOPS013 × PI 635246, Columbus × 18NEHOPS026, and Glacier × PI 635246 had negative EBVPSs across all 3 months (Table 5). Parental genotypes that were represented by only one cross produced progeny with EBVPSs like those of the parent. For example, maternal genotype Chinook had an EBVPS of 0.026 in June, −0.193 in July, and −0.277 in August (Table 3) and the progeny derived from Chinook had similar EBVPSs of 0.010 in June, −0.211 in July, and −0.308 in August (Table 5). Two maternal genotypes, Chinook and Columbus, and two paternal genotypes, PI 635403 and 18NEHOPS0031, were represented by single crosses in this study.
Progeny derived from a locally collected wild maternal genotype, 18NEHOPS013, performed poorly, especially during July and August (Table 5). Because 18NEHOPS013 was crossed with other locally sourced or regionally adapted wild hops, poor performance of progeny derived from 18NEHOPS013 is likely due to the presence of undesirable traits such as susceptibility to downy mildew and overly long internodes and lateral branches, traits commonly observed in unimproved wild hops collected by the program. Long internodes can lead to reduced lateral branching, ultimately resulting in fewer cones. Although excessively long lateral branches may indicate the potential for increased cone production, they tend to tangle and cause more shading, requiring plants to be spaced farther apart in the field, which reduces the number of producing bines and ultimately decreases overall yield. Although breeding populations derived from 18NEHOPS013 performed poorly, it may have value as a breeding parent as the genotype might contribute alleles that are favorable for Nebraska adaptation, which were not captured in this study. The choice of paternal genotypes with which to pair 18NEHOPS013 will require careful consideration and should include breeding crosses with complementary paternal genotypes that produce progeny with desired adaptability and agronomic performance. If complementary parents cannot be identified, 18NEHOPS013 will be discarded. Similarly, the choice of paternal genotype will need careful consideration when planning crosses to be made with the maternal genotypes Galena and Glacier, because their breeding populations had mixed performance depending on the paternal parent (Table 5). In general, the EBVPSs for the progeny produced in this study have utility in determining which parental genotypes and which crosses are expected to produce superior progeny adapted to Nebraska.
Although none of the parental lines evaluated were developed for the Midwest, several were identified that have agronomic potential for the region. Among the six maternal genotypes evaluated, Galena, Glacier, and Sorachi Ace all exhibited above average season-long performance based on their EBVPSs. Columbus performed below average, and Chinook and 18NEHOPS013 had mixed responses, performing above average on some rating dates and below average on others (Table 3). Similarly, PI 635403, a paternal parent, performed above average during the multiyear-long evaluations, whereas 18NEHOPS026 and PI 635246 performed below average, and 18NEHOPS031, PI 635242, and PI 635287 had mixed responses based on their EBVPSs (Table 4).
Based on EBVPSs for the three years, progeny adapted to Nebraska can be created using maternal genotypes Sorachi Ace, Glacier, and Galena and paternal genotypes PI 635403. These four genotypes produced progeny that performed above average throughout the season and have potential for use in future breeding efforts (Tables 3 and 4). Parental genotypes whose progeny performed inconsistently during June, July, and August should be used judiciously, pairing them only with other genotypes that show complementary responses to the Nebraska environment.
Some parental genotypes also showed strong potential for creating useful breeding crosses. The maternal parent Sorachi Ace consistently produced progeny with above average EBVPSs across all three rating dates in June, July, and August (Table 5). These consistent results indicate that Sorachi Ace has good potential for use as a parent when developing progeny adapted to Nebraska. In contrast, Galena and Glacier produced progeny that had inconsistent performance across the growing season, with some progeny performing above average and others performing below average (Table 5). The inconsistent performance of breeding populations originating from Galena and Glacier are indicative of the importance of pairing either of these maternal genotypes with paternal genotypes that would result in progeny having complementary responses that would provide desired adaptation and agronomic performance.
Likewise, the results underscore the importance of considering breeding values of progeny and individual parents when making breeding decisions. Paternal genotypes, such as PI 635403, which was represented by only one breeding cross, had EBVPSs similar to its progeny (Tables 4 and 5). These similar EBVPSs suggest that more data across multiple breeding crosses are needed for accurate assessment of the genotype’s full potential as a parent. Although PI 635403 was only represented by one breeding cross, the above average performance of PI 635403’s progeny suggests the potential value of the genotype for hops breeding. Conversely, the maternal hop 18NEHOPS013 was the only parental genotype with multiple breeding crosses for which most ratings were below average (Table 5). The poor performance of 18NEHOPS013’s progeny suggests 18NEHOPS013 lacks desirable alleles for adaptation and agronomic performance and has limited utility in developing progeny adapted to Nebraska. However, additional investigation is needed to determine which adaptive trait or traits led to the poor performance rating of these progeny so they can be avoided in future hops breeding efforts.
Although maternal and paternal genotypes showed a similar range of EBVPSs in June (Tables 3 and 4), the range of values for the maternal genotypes increased in July and August (Table 3). The range of values for the maternal genotypes and their respective progeny was widest in August, which was expected, as progenies become established and display differences in adaptive vigor and traits potentially related to yield (Table 5). Also, field conditions in August were conducive to secondary downy mildew infections, which impacted August rating scores. In contrast, Henning and Townsend (2005) found a lack of genetic variation for several brewing qualities between maternal parents. In the Henning and Townsend (2005) study, maternal parents were public elite cultivars, and the authors suggested that lack of variation was the result of ancestral relationships between tested genotypes. In the current study, limited information is known about the ancestral origins and pedigrees of the wild hops included. The USDA-ARS wild hop accessions were found in North Dakota or Canada and belong to the taxonomic variety lupuloides (USDA NRCS 2004). Small (1978) classified H. lupulus into five taxonomic varieties, three of which are native to North America: var. lupuloides, var. neomexicanus, and var. pubescens. Driskill et al. (2022) conducted a genetic marker analysis that included the three Nebraska wild hops, placing 18NEHOPS013 and 18NEHOPS026 in the wild North American and US-developed hop clade, with 18NEHOPS013 closely related to genotypes of var. pubescens and 18NEHOPS026 to var. lupuloides. In contrast, 18NEHOPS031 was placed in the English hop clade, showing close ties to var. lupulus, a European variety introduced to North America. In addition, plants growing in the wild are subject to direct selection pressure for traits important to commercial production, such as adaptation, disease resistance, pest resistance (Hampton et al. 2001; Haunold et al. 1993), and brewing qualities (Hampton et al. 2002; McCallum et al. 2019), which highlights the importance of wild hop genotypes in providing essential traits for the development of new cultivars tailored to regional production needs.
In contrast to responses of the maternal genotypes, the narrowest range of EBVPSs for the paternal genotypes was in July (Table 4). Quality traits, along with specific growth and disease resistance traits, will be assessed in subsequent multiyear evaluations extending beyond the duration of this study because of the importance of these traits to hop producers and beer brewers. Henning et al. (1997a) and Henning and Townsend (2005) found that progeny of paternal genotypes showed significant genetic variation for multiple traits, including yield, alpha acid, beta acid, and essential oil content. In the current study, progeny plants were immature, chemical quality traits were not assessed, and yield was subjectively assessed as a component of the rating score so subsequent longer duration studies will determine if the tested paternal genotypes contribute to progeny traits similar to what was observed by Henning et al. (1997a) and Henning and Townsend (2005). Hop genotype performance is influenced by their growing environment, and these genotype-by-environment interactions are important considerations when developing new regionally adapted hops; these interactions influence expression of key traits such as yield and brewing qualities (Beatson and Brewer 1994; Henning and Townsend 2005). However, genotype-by-environment interactions were not evaluated in the current study because of the unreplicated data and relatively small sample size. This study was designed as an initial evaluation of germplasm for adaptation to Nebraska, and hence data were recorded using a single numeric rating, that rating reflecting an aggregate assessment of various traits contributing to adaptation and agronomic performance.
In conclusion, characterization of the foundational genotypes and progeny from their breeding crosses indicate that these foundational genotypes have potential utility when breeding for Nebraska-adapted hops. The commercial cultivars included in this study were selected for exceptional regional performance based on input from local hop producers. The wild hops included in this study were either locally collected or found to persist following multiyear field evaluations before the start of this study. Future studies will incorporate additional genotypes not included in this study and any new parental genotypes added to the program. By focusing on these foundational genotypes and on their progeny that have high EBVPSs, the efficiency of breeding for Nebraska-adapted hops can be improved. The efficiency gains are due to reducing the number of progeny created during each crossing season, by targeting progeny from crosses known to contribute to local adaptation, and the time needed to develop, evaluate, and release cultivars.
Contributor Notes
The data supporting the findings in this study are available from the corresponding author, K.L.A., upon reasonable request.
We thank Matthew Sousek and Craig Ferguson for constructing the hop nurseries. This work was supported by the Nebraska Craft Brewery Board, Nebraska Department of Agriculture Specialty Crop Block Grant Program, and Midwest Hop Producers.
