Search PMN  


Peer Reviewed

© 2001 Plant Health Progress. 
Accepted for publication 9 March 2001. Published 13 March 2001.

Development of Management Strategies for Hop Powdery Mildew in the Pacific Northwest

William W. Turechek,
Assistant Professor, Cornell University, NYSAES, Geneva, NY 14456; Walter F. Mahaffee, Research Scientist, USDA-ARS, Corvallis, OR 97330; and Cynthia M. Ocamb, Assistant Professor, Oregon State University, Botany and Plant Pathology, Corvallis, OR 97331

Corresponding author: William W. Turechek.

Turechek, W. W., Mahaffee, W. F., and Ocamb, C. M. 2001. Development of management strategies for hop powdery mildew in the Pacific Northwest. Online. Plant Health Progress doi:10.1094/PHP-2001-0313-01-RS.


Hop powdery mildew, caused by Sphaerotheca macularis, was first discovered in the Yakima Valley of Washington in 1997 and has since become the most serious disease of hop (Humulus lupus) in the Pacific Northwest. Lack of understanding of the epidemiology of S. macularis has made it difficult to develop sound management practices. Results from our field and laboratory studies suggest that control measures applied early in the growing season are probably the most important in shaping the epidemic in a particular field and that late season control measures may not need to be applied at the same intensity as in early to mid-season.


Fig. 1. Symptoms of powdery mildew on: A, an emerging shoot completely colonized by S. macularis, i.e., a flagshoot; B, a close-up of a colony on a leaf; C, a female flower; D, undeveloped; and E, developed cones (click image or larger view).


In June of 1997, despite years of quarantine efforts, hop powdery mildew (Sphaerotheca macularis [Wallr.:Fr] Lind. (synonym S. humuli [DC.] Burrill)) was reported for the first time in the Pacific Northwest (Fig. 1)(9). The disease was first observed on greenhouse-grown plants and subsequently found in nearby fields in the lower Yakima Valley. Initially, it appeared limited to a few locations but soon spread throughout the Washington hop-growing region by the end of July of the same year. Washington growers removed more than 1200 hectares of highly susceptible hop varieties resulting in an estimated loss of $9.5 million in combined production and establishment costs in 1997. Moreover, the movement of hop plant material across the three Pacific Northwest states was restricted to prevent spread of this disease. Despite these efforts, powdery mildew was detected in both Oregon and Idaho in 1998. Powdery mildew and its control is estimated to have cost Pacific Northwest hop growers over $30 million in 1999 and 2000, or about 15% of their total crop revenue (personal communication, Ann George, Administrator, U.S. Hop Industry Plant Protection Committee).

In 1998, powdery mildew was managed in Washington using an intensive preventative fungicide program. Prevention typically required application of demethylation-inhibiting (DMI) fungicides and horticultural oils (3,4) on a 10-14 day spray schedule. Using this schedule, growers made up to 15 pesticide applications over the course of the season and achieved reasonable control of powdery mildew. In 1999, many were reluctant to engage in the same preventative fungicide program due to the expenses incurred in the previous year. Thus, growers relied on scouting information to alert them to the presence of the disease before taking action and/or many growers waited until flower formation before applying pesticides. Unfortunately, flowers and developing cones were exposed to high levels of inoculum from uncontrolled leaf infections, which resulted in high disease incidence of the cones even though pesticides were being applied once the disease was detected.

Limited information has been published on the biology of the pathogen and disease epidemiology (8,11). This has hampered the development of IPM-type approaches to disease management, leaving growers with very few options other than cultural strategies and routine calendar sprays of pesticides. The objectives of this study were to: 1) determine the optimal range of temperature for hop powdery mildew development on a susceptible cultivar; 2) determine factors that influence epidemic development in commercial fields; and 3) develop testable hypotheses for disease management based on the results of this study.

Effects of Temperature on Infection Frequency and Lesion Size

A study was conducted to determine the range of temperatures conducive for disease development. Hop plants of the highly susceptible cultivar ‘Symphony’ were propagated from greenwood cuttings (6), planted in Sunshine Mix #1 (SunGro Horticulture, Bellevue, WA) in 15-cm pots, and grown under greenhouse conditions. Leaves were tagged as they unfolded to monitor leaf age. A single-chain isolate of S. macularis was obtained from an Oregon hop field, propagated on greenhouse-grown hop plants, and maintained at 13°C in growth chambers. Inoculum was prepared by washing conidia from infected hop leaves with a 0.005% (vol/vol) solution of Tween 20 and adjusting the suspension to a concentration of 50,000 conidia/ml. Plants were inoculated with a handheld atomizer and placed into growth chambers with a 15h day length (~300 mmol) set at 12, 15, 18, 21, 24, 27, or 30°C. Although it is generally recognized that conidia of many powdery mildews lyse when suspended in water, some powdery mildews, including S. macularis, are capable of surviving for short periods when suspended in water (1,10). When possible, inoculation via water suspension is preferred to settling tower inoculations because it permits a quantitative measurement of inoculum concentration to be made which allows for precision and repeatability in controlled experiments.

Disease development was monitored daily and the latent period was recorded. At two latent periods, three to five plants were evaluated for disease from each chamber (i.e., temperature). This period was used as the standard time for disease evaluation at a specific temperature because disease developed at different rates at the various temperatures and secondary spread was likely to occur within chambers if incubated longer. The infection frequency (number of lesions per unit leaf area) was recorded for each leaf on every plant. Leaf area was determined using a Li-cor LI-3000 leaf area meter (Li-cor Inc., Lincoln, NE). Average lesion area was determined by calculating mean lesion area from 3 arbitrarily selected lesions on each diseased leaf. Lesions were assumed to be circular, thus area was calculated using the formula for an area of a circle (Br2) using the average of perpendicular cross sections as a measurement of diameter (2r).

The experimental design consisted of three replications arranged in a randomized block design with temperature serving as the treatment and replication in time serving as the blocks. Temperatures were randomly assigned to growth chambers for each replication to minimize growth chamber effects. Infection frequency and lesion size were averaged over identical leaf ages in each replication for each temperature treatment and analyzed using the GLM procedure in SAS (version 6.12) with leaf age serving as a covariate. The logarithmic transformation was used for each response variable to stabilize the variance.

Disease developed at 12, 15, 18, 21, 24 and 27°C but not at 30°C (Fig. 2). Inoculation of water agar and glass slides indicated that conidia failed to germinate when incubated at 30°C (data not shown). The latent period was approximately 10 days at 12 and 15°C compared to 5 days at 18-27°C. Temperature had a significant effect on infection frequency (Fig. 2A) with 27°C having a significantly lower infection frequency than the other temperatures. Temperature also had a significant effect on lesion area (Fig. 2B), with temperatures above 21°C resulting in significantly smaller powdery mildew colonies. Leaf age had a significant effect on infection frequency (Fig. 3A) and lesion size (Fig. 3B) with young, expanding leaves being more susceptible than mature leaves (>12 days old). Leaves 15 days or older did not become infected.

Fig. 2. The effect of temperature on infection frequency (A) [number of lesions per cm2] and lesion area (B) of S. macularis on ‘Symphony’ hop leaves in a controlled environment study. Infection frequency and lesion area were averaged across leaf age. Error bars indicate ± standard error (click image or larger view).

Fig. 3. The effect of leaf age on infection frequency (A) [number of lesion per cm2] lesion area (B) of S. macularis on ‘Symphony’ hop leaves in a controlled environment study. Infection frequency and lesion area were averaged across temperatures. Error bars indicate ± standard error (click image or larger view).

Field Survey

In 1999 and 2000, hop fields in Washington and Oregon were surveyed from shoot emergence through harvest to gather information on epidemic development. In each field, two to nine transects (rows) were randomly selected. In each transect, 10 leaves (n) were arbitrarily selected from the first 75 to 100 plants or until the end of the transect was reached. If any one of the 10 leaves were infected then the plant was considered diseased. Disease incidence [p] for an individual transect was calculated using ∑x/N where x=1 if a plant was diseased and x=0 otherwise (12).

Similar sampling strategies were used to assess disease early in the season (i.e., incidence of flagshoots) and early secondary spread before there was enough leaf material to use the strategy defined above. Flagshoots are early spring shoots completely colonized by S. macularis which are believed to originate from mycelia overwintering in buds, presumed to be the initial inoculum (Fig. 1). For early-season disease assessments, 200 plants (N) in each of 2-9 randomly-selected transects (depending on the size of the field) were rated for the presence of flagshoots or secondary spread (leaf lesions) regardless of the number of flagshoots or leaf lesions on any single plant. Plant incidence was calculated as before. Cone incidence (final disease rating) was assessed within 3 weeks of harvest date using a modification of the leaf-sampling strategy. Due to the morphology of hop plants, groups of cones were collected. Two to three groups of cones (inflorescences) were collected from two to three heights along the bine from each plant, and the number of diseased cones (x) was rated out of the total number of cones collected (n). Final cone incidence was determined using ∑x/nN.

In total, 44 transects of leaf disease incidence were collected from 12 hop fields over the 1999 growing season in Oregon and Washington. Based on the 1999 data, 2000 sampling was intensified with more fields and more transects per field sampled. A total of 445 transects were sampled in Washington and 381 transects were sampled in Oregon from 42 different hop fields. The selected fields represented distinct growing regions, multiple varieties, and numerous planting and management strategies.

In conducting our sampling for initial secondary spread in May, we noticed a considerable variation from field to field in the amount of green tissue left after the growers had done their pruning. With hops, pruning is the practice of removing early plant growth in order to establish the optimal training date. There are three primary methods growers use to prune: mechanical, chemical, or propane pruning. Quality of pruning was subjectively rated using the scale of 1 (excellent; no green tissue left), 2 (moderate), and 3 (poor; numerous leaves and shoots visible on each plant). The degree of secondary spread found at the first leaf rating was correlated with the incidence of flagshoots found in a field and the quality of pruning (Table 1). Generally, data indicated that fields with low incidence of flagshoots or that were pruned well entered into the season with relatively low leaf disease incidence, whereas those fields with a high incidence of flagshoots and were pruned poorly had the highest level of foliar disease. 

There were regional differences in disease development in 2000 (Fig. 4). In Washington, the Reservation and Moxee regions had the highest level of disease incidence, followed by the Mabton and the Prosser regions. Oregon had a much lower level of disease incidence than most Washington regions. Furthermore, there was a strong relationship (r2=0.68) between cone incidence and leaf incidence (average over the season) in a hop field (Fig. 5).

Table 1. Proportion of hop plants with secondary powdery mildew infections in relation to initial flagshoot density and the quality of pruning.

Quality of Pruninga
1 2 3 All
0 0.09 (6)b 2.33 (1) 6.500 (1) 1.17 (8)
>0.0-0.5 0.06 (2) 2.76 (1) 13.75 (3) 5.36 (6)
0.6-1.5 7.14 (5) 0.00 (1) 2.30 (1) 5.43 (7)
1.6-5.0 2.50 (1) 3.21 (2) 27.52 (3) 15.25 (6)
All 2.78 (14)  2.30 (5) 16.58 (8) 6.33 (27)

aQuality of pruning was subjectively rated using the scale of 1 (excellent; no green tissue left), 2 (moderate), and 3 (poor; numerous leaves and shoots visible on each plant).
bNumber of fields sampled is shown in parentheses.

Fig. 4. Regional disease progress curves of hop powdery mildew incidence for the year 2000. Data points represent mean incidence of powdery mildewed plants from transects surveyed in the region and month indicated on the graph (click image or larger view).

Fig. 5. Relationship between incidence of hop cones with powdery mildew and the seasonal average of powdery mildew incidence on leaves. Data was collected from commercial hop fields in Washington and Oregon in 1999 and 2000. Data are plotted on logarithmic axes to achieve better separation of the data at lower incidences (click image or larger view).


The data presented here suggest that control measures applied early in the growing season are probably the most important in shaping epidemic development and that later season control measures may not need to be applied at the same intensity as in early to mid season. This is based on two findings. First, the temperature study indicated that powdery mildew developed best at cool temperatures on relatively young leaves. These conditions are prevalent during the early part of the growing season. Second, field surveys indicated that incidence of powdery-mildewed cones is highly correlated with the level of foliar disease. Based on these observations, a prudent management strategy would appear to be one that is very aggressive before flowering and less aggressive during cone development. This strategy is also supported by epidemiological theory for a polycyclic disease that indicates that the rate of the epidemic is best affected by early season control measures affecting secondary spread (2).

The field survey revealed a strong regional influence on epidemic development. However, several factors may be interacting with the regional factor to influence powdery mildew development. Incidence of flagshoots (see Fig. 1) is perhaps the most significant of these interacting factors. Because of the polycyclic nature of this disease, a relatively high incidence of flagshoots could lead to increased levels of foliar disease due to a greater inoculum potential early in the season. Foliar and cone disease incidence were greatest in the Reservation and Moxee regions where flagshoot incidence and early secondary spread were the greatest. Lower disease incidence in Prosser may be due to the relatively uniform management practices of the few growers in that region where most growers did an excellent job of removing all of the green tissue at the time of pruning. Disease development in Oregon may differ because the long, wet springs may hamper disease development, planting of cultivars with varying levels of susceptibilities, and perhaps genetic variability among strains of S. macularis.

In Oregon, growers crown plants (pruning of plants 5-10 cm below the soil line so that the old bines and associated buds are removed) to help control downy mildew. This method is also used in England to manage powdery mildew (8). In Washington, growers have used scratching (removal of buds in the top 5-7 cm of soil using a device with spinning steel spring tangs), harrowing, and propane pruning to reduce flagshoots. However, the observed relationship of percent flagshoots and quality of pruning to initial secondary spread in Washington (Table 1) indicates that there may not be any need for cultural management of flagshoots in Washington if pruning is done very well. This hypothesis is further supported by the results from two fields where growers performed a second pruning after observing heavy initial secondary infections and started a protectant program after regrowth (data not shown). Despite the heavy secondary infections in May, these fields had low leaf and cone incidence at the end of the season. This practice could save growers up to $400/ha each year depending on their cultural practices.

Finally, there remains the question of how host resistance or tolerance plays a role in epidemic progress. Differences in varietal susceptibility could be one reason for the observed difference between Oregon and Washington hop growing regions. Most of the Oregon acreage is planted in varieties that appear to be less susceptible to powdery mildew, while in Washington a high percentage of the acreage is planted to highly susceptible varieties. Despite the perceived benefits of less susceptible varieties, the implementation of varietal resistance is limited due to the perennial nature of the crop, the lack of suitable rhizomes (planting stock), the substantial replant costs, and market factors. In addition, the long-term utility of varietal resistance is also suspect due to the rapid breakdown of resistant lines in other countries (7, Peter Darby [England] and Bernhard Englhard [Germany], personal communications) and the apparent breakdown in the Pacific Northwest. In 1997, the cultivar CTZ was thought to be resistant to hop powdery mildew, however, in 1998 it was devastated by powdery mildew infections. Most fields were not harvestable due to the severe cone infections.

While there is still much to be learned about the epidemiology of hop powdery mildew, we are rapidly gaining the knowledge base for development of an integrated pest management program that will incorporate cultural practices, disease forecasting, and economics into management decisions. Understanding the influence of temperature on disease development and how the epidemic progresses in a field are the first steps to building sound management recommendations. The data presented here indicate that control measures early in the growing season are probably the most important in shaping the epidemic in a particular field and that later season control measures may not need be applied at the same intensity as in the early to mid season. Thus, there is potential that the relationship between temperature and infection rates and disease development can be used to construct a disease-forecasting model similar to the conidial phase of the Gubler/Thomas model for grape powdery mildew (5). Such a model would be useful in helping growers time pesticide applications and possibly reduce the number of applications late in the season.


The authors acknowledge the efforts of Isabella Cantrell, Julie Dileone, Kelly Donahue, Joesef Heffner, Sean McMullen, Matthew Scott, Tara Sechler, Chris Soskis, and Ryan Whitmore for long days collecting data. The authors also thank Mark Nelson, Ann George, Michelle Palacios, and the many hop growers of Washington and Oregon.

Literature Cited

1. Aldwinckle, H.S., Watson, J.P., and Gustafason, H.L. 1975. Relationship between greenhouse and field resistance of grape seedlings to powdery mildew. Plant Dis. Rep. 59:185-188.

2. Campbell, C.L., and Madden, L.V. 1990. Introduction to Plant Disease Epidemiology. John Wiley & Sons.

3. Davidson, N.A., Dibble, J.A., Flint, M.L., Marer, P.J., and Guye, A. 1991. Managing insects and mites with spray oils. IPM Education and Publications, Statewide Integrated Pest Management Project, University of California, Pub. 3347.

4. Dell, K.J., Gubler, W.D., Krueger, R., Sanger, M., and Bettiga, L.J. 1998. The efficacy of JMS Stylet-Oil on grape powdery mildew and Botrytis bunch rot and effects on fermentation. American Journal of Enology and Viticulture 49:11-16.

5. Gubler, W.D., Rademacher, M.R., Vasquez, S.J., and Thomas, C.S. January 1999. Control of powdery mildew using the UC Davis powdery mildew risk index. Online. APSnet Features.

6. Haunold, A., and Zimmermann, C.E. 1974. Pollen collection, crossing and seed germination of hop. Crop Science 14:774-776.

7. Liyanage, A. de S. 1973. Studies on resistance and overwintering in hop powdery mildew. PhD thesis, University of London.

8. Neve, R.A. 1991. Hops. Chapman & Hall, London.

9. Ocamb, C., Klein, R., Barbour, J., Griesbach, J., and Mahaffee, W. 1999. First report of hop powdery mildew in the Pacific Northwest. Plant Disease 83:1072.

10. Reeser, P., Hagedorn, D.J., and Rouse, D.I. 1981. Quantitative inoculation with Eriysiphe pisi to assess variation of infection efficiency on peas. Phytopathology 73(9):1238-1240.

11. Royle, D.J. 1978. Powdery mildew of the hop. Pages 381-409 in: The Powdery Mildews. D.M. Spencer ed. Academic Press, London.

12. Turechek, W.W., and Madden, L.V. 1999. Spatial pattern analysis of strawberry leaf blight in perennial production systems. Phytopathology 89:421-433.