California Pepper Commission


Annual Research Report On Pepper Powdery Mildew


Mike Coffey, Professor
Dept. of Plant Pathology
UC Riverside
CA 92521

February 2007

Introduction and Background

Powdery mildew of peppers, caused by aggressive strains of the fungus Leveillula taurica, has emerged in recent years as a serious economic threat to processing and fresh market pepper producers in California. Its essentially unpredictable nature, high capacity for sporulation, very long latent period (18-21 days) and explosive epidemic potential can make it very difficult to control. Fungicides are frequently applied too late often resulting in severe sunscald of the pepper fruit, typically to significant and sometimes disastrous economic crop losses. In the summers of 2003 and 2004, there were several dramatic instances where crop losses in some cases approached 100 percent. In 2005, by way of contrast, powdery mildew was rarely seen. In 2006 the disease was very late in its appearance and caused significant losses. These events and their unpredictable nature underline the need to invest in research both on more effective control measures: new fungicides and the development of mildew-resistant varieties, as well as in evolving accurate early warning methods.

There is an urgent need to investigate new fungicides with good mildew activity, and in 2006 we focused on V-10118 and Quintec. Overall, there needs to be a closer examination of the efficacy of such fungicide chemistry towards the strains of powdery mildew currently attacking peppers in California. There has been no critical work on this and the outbreaks of mildew in 2003 and 2004 are a warning that current control strategies are probably not very effective when high disease pressure exists.

A second research project continued with season long evaluation of different selections of putative mildew-resistant pepper lines, including the C. annuum line 04A3 which is a red bell pepper breeding line developed in France, several dozen lines tested in prior years, as well as an F2 population (J207 X J214) derived from a selection of C. chinense, provided by Dr Kevin Crosby at the Texas Ag Exp Station.

The onset of powdery mildew occurrence has been impossible to predict in recent seasons in California. A critical factor in these outbreaks is the role of weather conditions. What are the key environmental conditions, especially temperature and relative humidity that trigger the onset of mildew epidemics? Some simple experiments in 2005 using growth chambers set with various temperature/humidity environments provided evidence that the ëHollisterí pepper strain used in our research behaves similarly to a tomato strain used in the development of the Guzman-Plazola Model for mildew prediction. Last year we looked at the effect of different temperatures and humidities using a constant environment. While this is an artificial situation it allowed us to determine that relatively high temperature (28 C) could shut down initial mildew development completely over a wide range of relative humidities. This year we focused on testing two alternative field based data logging technologies, Hobo and WatchDog. The Hobo weather station was set up for several months in a pepper field in Oasis in the Coachella Valley.

Questions still needing to be addressed include: what are the other hosts including weed hosts? We are still uncertain of the host range, including both weed hosts and other crops, and such knowledge might be very important for an understanding of how outbreaks get started. How does it survive from season to season? Again we do not know.

The 2006 research proposal embraced a broad range of aspects of the effective control and management of pepper powdery mildew. One important focus in 2006 was on the efficacy of the newer fungicide chemistries especially V-10118 and Quintec. The season long performance of 40 different pepper lines was monitored for mildew resistance through fruit production and ripening. This included 100 F2 progeny from a cross involving a resistance P. chinense line. Finally, we will establish a single site with a weather station to assess its performance and measure temperatures, humidity and leaf wetness.

Lath house evaluation of V-10118 and Quintec


Plants grown two per row in 2 gallon pots with automated irrigation. Fertigation was provided via drip using PetersÆ ExcelÆ 21–5–20 (Scotts-Sierra Horticultural Products Company, Marysville, OH) at the recommended rate. Osmocote 14–14–14 was also as a slow release fertilizer.

There were 8 treatments each of 5 replicates. Each replicate consisted of 4 plants. Transplants of a highly susceptible yellow bell hybrid LY1 (Seminis, supplied by Headstart Nursery, Oasis) were transplanted into 2 gallon pots with UC Mix on 4 May 2006. Each replicate of 4 treated plants was interspersed with an additional guard row of the susceptible cultivar (two plants) plus the highly susceptible variety ëFooled Youí was deployed in a single row between the two treatment rows once the transplants have grown to the flowering stage. These additional plants served as guard rows to minimize fungicide drift and as adjacent sources of mildew inoculum. The trial was set up in a randomized block design.

Fungicides were applied applied 3 times at 7 day intervals using a CO2 sprayer @ 30 psi to both leaf surfaces. Peppers, LY1, were inoculated on 19 June, 2006 with an aqueous suspension of powdery mildew spores. On June 21, 2006 fungicides were applied at weekly intervals for at total of 3 applications. Visual ratings (lesions per plant, disease incidence) were made 6 times (every 2–3 days) over a 3 week period and the plants were grown to harvest.

TreatmentDosageActive Ingredient*Rate
1V-101182xHigh80 ppm 0.08 lb a.i.
2V-10118High40 ppm 0.04 lb a.i
3V-10118Half 20 ppm0.02 lb a.i
4QuintecHigh88 ppm6 fl. oz. 0.6 gal
5RallyHigh100 ppm4 oz
6Cabrio/RallyHigh100 ppm/100 ppm8 oz/ 4 oz
7Cabrio100 ppm 8 oz

*Calculations based on an application rate of 120 gal per acre

Results and Conclusions

Three weekly applications Rally and Cabrio used alone at the manufacturerís recommended rates provided effective control of powdery mildew through the entire growing season. The alternate application of Cabrio/Rally/Cabrio was equally effective. Quintec gave similar results. In marked contrast, V-10118, even when used at very high rates, gave little or no control of mildew. The plants were observed through harvest and the control of all initially effective treatments held with no further increase in disease. The growing season in 2006 at UCR was very hot and dry, however.

Quintec proved to have similar efficacy to both Rally and Cabrio. No phytotoxicity was observed with any fungicide at the rates used.

Assessment of Resistance in Pepper Breeding Lines

Breeding lines from Kevin Crosby (Texas Ag Exp Station) were assessed in comparison with additional control lines whose high resistance level had been established in previous years. The lines will be sown in seedling trays with UC Mix (18-21 May 2006). The TAES lines included C. chinense Panama Red, C. chinense 315017, C. chinense 543184, C. chinense 543188, Fidel, Hidalgo, C. chinense TmH, Lv 2323, C. chinense 257046, J214 p1, J214 p2, J214 p3, 215, 216, 217, 467, and C. baccatum Panama. The ëUCRí resistant reference lines included: C. annuum PBC 167, C. baccatum PBC 151, C. chacoense 904750050, C. chacoense 904750103, C. frutescens 8884750097, C. annuum 04A3, C. chinense M 1796 -1 and C. chinense 904750105. For each of these lines 4 replicate plants were grown in gallon pots with UC Mix. In addition, 100 seedlings were grown of a TAES F2 cross of J207 X J214.

Pre-germinated were sown in seed trays on May 19, 2006 in greenhouse 13C. Later, they were transplanted to 1 gallon pots and moved to the lath house on August 7, 2006. Ratings were made monthly through December 24, 2006.

1C. baccatum PanamaHRHRHRHRHR
2C. chinense Panama RedHRHRHRHRHR
3C. chinense 257046HRHRHRHRHR
4C. chinense 315017IRIRIRIRIR
5C. chinense 543184HRHRHR -HR
6C. chinense 543188HRHRHRHRHR
9C. chinense TmHHRHRHRHRHR
10Lv 2323SSSSS
18TAES 467HR*HR*HR*- HR*
19C. annuum PBC 167 MRMRMRMRMR
20C.baccatum PBC 151 HRHRHRHRHR
21C. chinense 904750105HRHRHR -HR
22C. chinense M 1796-1, Aji RedHRHRHRHRHR
25C. annuum 04A3 MRMRMR -MR
26C. chinense PI 438619 Habanero RHRHRHRHRHR
27C. chinense PI 195301 01 SMRMRMR -MR
28C. chinense 984750047SSS -S
31C.baccatum PI 439376HRHRHR -HR
32C. chacoense 904750050HRHRHRHRHR
33C. frutescens 884750097IRHRHRHRHR
34C. pubescens 904750052HRHRHRHRHR
35C. annuum PBC 81IRIRIR -IR
38C. chacoense 944750103HRHRHRHRHR
39C. chacoense 904750173HRHRIRHRHR
40C. chacoense PI 273419MRMRMRSMR

Definition of resistance to powdery mildew used in this study

Highly resistant
(no visible symptoms)
Highly resistant *
(minute necrotic lesions or flecks)
Intermediate resistant
(few expanded lesions but no sporulation)
Moderately resistant
(few expanded sporulating lesions)
(large number of expanded sporulating lesions)

The TAES F2 cross of J207 X J214 produced mainly highly susceptible progeny. Of 100 seedlings tested, 92 were susceptible (S), 2 were moderately resistant (MR) and 6 were intermediate in resistance (IR).

Not all these more resistant progeny bore fruit. None of the progeny had a resistance level comparable to the top ranking lines with an HR rating with no visible symptoms (i.e., ëimmunityí to powdery mildew).

Outstanding sources of resistance (HR) included:
C.baccatum PBC 151, C. baccatum Panama
C. chacoense 904750050, C. chacoense 944750103,
C. chinense M 1796-1, C. chinense Panama Red, TAES 467,
C. frutescens 884750097,
C. pubescens 904750052.

#VarietyReplicates Resistance
1C. baccatum PanamaHRHRHRHRHR
5C. chinense 543184HRHR  HR
6C. chinense 543188HRHRHR HR
20C.baccatum PBC 151 HRHRHRHRHR
21C. chinense 904750105HRHR  HR
22C. chinense M 1796-1, Aji RedHRHR  HR
24C. annuum 04A3 (BB)*HRIRIR IR
25C. annuum 04A3 (CC)*IRIR  IR
26C. chinense PI 438619 Habanero RojoHRHRHRHRHR
27C. chinense PI 195301 01 SIR   IR
31C.baccatum PI 439376HRHR  HR
32C. chacoense 904750050HRHR  HR
33C. frutescens 884750097HRHRHRHRHR
35C. annuum PBC 81IRIRIR IR
38C. chacoense 944750103HRHR  HR
39C. chacoense 904750173HRHRIRHRHR
40C. chacoense PI 273419HRHR  HR

* BB and CC refer to seed collected from different plants.

A second lath house experiment largely confirmed the results of the first one with lines with immunity (HR) to mildew including:

C.baccatum PBC 151
C. baccatum Panama
C. chinense PI 438619 Habanero Rojo
C. frutescens 884750097

Plants of the resistant selection of C. annuum 04A3 were rated MR in the first experiment and either 1R or HR in the second experiment.

Mildew Forecasting using the Disease Prediction Model

A disease risk prediction model for Leveillula taurica has been developed by Dr. Remigio Guzman-Plazola, while working at UC Davis with Dr. Mike Davis. For pepper growers, field validation at sufficient locations and through several mildew years might encourage growers to use fungicides in a more effective manner, with economic and environmental benefits. In particular, such disease prediction technologies could be used to determine if and when to apply the first fungicide spray. Such field-based analysis and research would permit modifications and fine-tuning of prediction programs to make them more robust. One version of the Guzman-Plazola forecasting model is currently available to growers through the University of California Integrated Pest Management (UC IPM) website, and allows the grower to enter their own local data. Several other versions of this model are also available and they need to be compared for their abilities to forecast disease. In collaboration with Dr. Remigio Guzman-Plazola who is involved in developing a similar system for Mexican tomato and pepper growers we are in the process of analyzing and improving the available models.

Practical problems for the California pepper grower include the collection, collation and conversion of the weather data to a format compatible with the UC IPM site. This is currently an extremely time-consuming process and it is highly desirable to develop the software to make data analysis straightforward and rapid. In fact the ideal system would be one where telemetry data is ëgrabbedí automatically and is then ëfedí into the prediction model and a disease forecast again automatically sent to the grower. This year we developed PERL-based software that can be used as an interface on a Linux-driven (Max OS X) server in my UCR office to instantly convert field data into a format compatible with the disease prediction program.

Another issue for growers will be the cost of the equipment for monitoring weather. Ultimately, it might be possible to obtain regional data from a network of telemetry stations. The validity of using regional weather data to forecast powdery mildew also needs to be assessed. In the short-term the monitoring of weather data over several ëmildewí years from a small number of well-maintained telemetry stations will provide the needed databases to validate or modify different powdery mildew forecasting models including those of Guzman-Plazola and Gubler/Thomas.

The Guzman-Plazola model is premised upon infinite inoculum availability. The disease triangle helps us understand the three interrelated factors required for disease onset: susceptible host, conducive environmental conditions, and the presence of the pathogen. The latter factor is presumed present in most disease modeling. But this assumption is often unfounded. One desirable objective will be to determine the key factors in triggering mildew epidemics and, if possible simplify the database requirements.

Our lab has been working with this pathogen for some time now, both in pepper and tomato. Although we have not determined the racial structure within the pathogen population, we have found that pepper and nightshades are far more susceptible to this disease than is tomato. Placement of highly susceptible pepper varieties such as ëFooled Youí in growersí plantings might serve as a bellwether of inoculum presence. Combined with usage of the forecasting model, our objective would be to demonstrate the utility of our and othersí research to reduce grower costs while assuring harvest.

We have numerous software analytical packages, but the intention in this project would be to feed the data collected to the UC IPM free interface to two versions of the disease risk model that use different leaf wetness scales. A third and original version, which is SAS-based and ëformattedí in an Excel-based program will also be tested with the assistance of Dr. Remigio Guzman-Plazola. From this, conducive weather would alert us to the disease risk. Combined with the inoculum bellwether, we would decide whether or not to make the initial application.

UC IPM Collaboration (Joyce Strand):

UC IPM will develop and make available a second version of the model for our project, so that he can compare model results based on the two different ways of determining wet hours. In addition, we will submit data files to Guzman so that he can run them against the original SAS model to make sure that UC IPM 's model generates the same results as the original.

A critical factor in this epidemiological research will be the accurate monitoring and recording of powdery mildew occurrence and increase during a growing season at the experimental sites. Without this careful evaluation and recording of disease progress the research will be meaningless.

We purchased a field-tested a HOBO Weather Station which offers research-grade measurement quality and up to 10 channels that accept plug-in smart sensors that eliminate complicated programming, wiring, and calibration. The entire HOBO Weather Station system, including sensors, is powered for one year using four AA batteries and can log over 500,000 measurements into the non-volatile memory.

Leaf Wetness Sensor

Data logger

We also purchased 2 WatchDog 450 Series Data Loggers featuring four channels with a capacity of 15,000 measurements (3,750 per channel). Data can be downloaded to a PC for analysis.

Select measurement intervals are those 1, 10, 15, 30, 60 and 120 minutes. A 30-minute measurement interval will record for 85 days before the logger's memory is full. EEPROM memory keeps data safe in case of power loss.

WatchDog Datalogger

Leaf Wetness sensor

Work Plan 2007:


I thank the California Pepper Commission for their generous financial support for this research. The University of California at Riverside Plant Pathology Greenouse staff were vital to the success of these projects. Researchers included myself, Alexey Kravtsov, Grazyna Szkuta, Tatiana Roubstova, Laura Gaggero, Avneet Brar and numerous student helpers, but especially Cale Carter who was reasponsible for most of the planting in the two large lath house experiments.

Valuable discourses were had with Remigio Guzman-Plazola, Joyce Strand, Joe Aguiar and Bob Heisey through the course of a year.

Additional support was provided by Headstart Nursery, Valent BioSciences and Dow AgroSciences.

Design and hosting provided by Spinhead Web Design