Search PMN  


PDF version
for printing

Peer Reviewed

© 2010 Plant Management Network.
Accepted for publication 7 July 2010. Published 23 August 2010.

Potassium Phosphite Mixed with Other Fungicides Reduces Yield Loss to Downy Mildew on Collard

Anthony P. Keinath, Professor of Plant Pathology, Clemson University, Coastal Research and Education Center, Charleston, SC 29414

Corresponding author: Anthony P. Keinath.

Keinath, A. P. 2010. Potassium phosphite mixed with other fungicides reduces yield loss to downy mildew on collard. Online. Plant Health Progress doi:10.1094/PHP-2010-0823-01-RS.


Downy mildew of collard occurs frequently in the southeastern United States, and fungicides have become an essential part of economical control strategies for the disease. Fungicides were evaluated in 2007 and 2008 either alone or combined with a reduced rate (2 pt/acre) of potassium phosphite. Two formulations of potassium phosphite also were tested alone at the full rate (4 pt/acre). Presidio, Presidio plus Pro-Phyt (2007) or K-Phite (2008), K-Phite, Amistar plus ProPhyt, and Aliette reduced downy mildew severity at the final rating and also reduced AUDPC over the two years. Presidio plus potassium phosphite and Amistar plus ProPhyt increased the weight of healthy leaves and stems over the water control and also increased economic return calculated after fungicide and other production costs were subtracted from the crop value. The yields with Ranman, Sonata, and Presidio applied in combination with potassium phosphite were 16% greater than with the fungicides applied alone. Potassium phosphite may be useful in a tank-mix with other fungicides applied to collard to prevent yield loss to downy mildew.


Downy mildew of vegetable crops in the Brassicaceae is caused by the oomycete Hyaloperonospora parasitica (Pers.: Fr.) Constant. (4). The pathogen is favored by low to moderate temperatures of 46 to 68°F (7). Sporangia failed to germinate at 82°F, although downy mildew developed at this temperature when plants were inoculated at 68°F. High relative humidity is conducive for the disease, but after infection downy mildew also will develop at low humidity levels (7).

Collard [Brassica oleracea L. subsp. oleracea convar. acephala (DC.) Alef. var. viridis L.] is traditionally planted in the fall for harvest in December for the New Year’s Day collard market, and some growers in South Carolina also produce collard year-round (5,6). When grown during the cool periods of the year, collard is particularly vulnerable to downy mildew, because periods of moist weather and moderate temperatures that are favorable for downy mildew occur frequently in the southeastern United States (7). All collard cultivars are susceptible to downy mildew.

Several fungicides with activity against oomycete pathogens are registered for application to leafy brassica vegetables (8). There is a considerable range in price of these fungicides, from $65/acre for aluminum tris O-ethyl phosphonate (Aliette 80WDG, Bayer Crop Science, Research Triangle Park, NC) applied at the maximum labeled rate of 5 lb/acre to $11/acre for copper hydroxide (Kocide 2000, Du Pont, Wilmington, DE) applied at 1.5 lb/acre. When environmental conditions favor downy mildew development, growers in South Carolina must spray collards weekly for periods lasting several weeks. Fungicide costs exceeding $106/acre/crop are considered excessive (2). Thus, two applications of Aliette at 4 lb/acre would account for the entire fungicide budget. Fungicides applied to control downy mildew must be effective and provide an economic return.

Mono- and di-potassium salts of phosphorous acid (potassium phosphite) are registered as a biopesticide by the United States Environmental Protection Agency (3). Potassium phosphite was first reported to be effective against brassica downy mildew when applied to cauliflower heads (11). The ProPhyt 4L formulation (Helena Chemical Co., Collierville, TN) of potassium phosphite was more effective than maneb at reducing severity of downy mildew on foliage of Chinese broccoli [B. oleracea L. var. alboglabra (L.H. Bailey) Musil] (13). In another field trial, potassium phosphite mixed with chlorothalonil was more effective than chlorothalonil applied alone against downy mildew on Chinese broccoli (14). Thus, potassium phosphite may be an effective alternative fungicide against downy mildew on collard.

The objectives of this study were to (i) evaluate various fungicides alone and in combination with potassium phosphite for control of downy mildew on collard and (ii) determine which fungicide combinations decreased losses of collard affected by downy mildew and examine resulting economic returns to growers.

Field Plots and Fungicide Applications

The experiments were done in 2007 and 2008 in the same field at the Clemson University Coastal Research and Education Center in Charleston, SC. The soil type was Yonges loamy fine sand with a pH of 5.6. The field was cropped to collard in 2005 and 2006. After disking, raised beds were shaped 1.5 ft wide on 3-ft centers. Fertilizer (800 lb/acre of 10-10-10) was applied over the entire field, and the beds were cultivated lightly. In 2007, 1.5 pt/acre Treflan (Dow AgroSciences, Indianapolis, IN) and in 2008, 10 lb/acre Dacthal W75 (Amvac Chemical, Newport Beach, CA) was applied over the entire field as a pre-emergent weed control. Collard cv. Champion was direct seeded with a Stanhay belt seeder (Stanhay Webb Ltd., Grantham, Lincs, UK) on 9 February 2007 and 29 February 2008 so that the final plant density was approximately one per 5 inches. Plants were side-dressed with 150 lb/acre 10-10-10 on 18 March and 12 April 2007, with 150 lb/acre of 33.5-0-0 on 24 April 2007, and with 250 lb/acre 10-10-10 on 18 April 2008.

The experimental design was a randomized complete block with four replications. Plots were 15 ft long and two rows wide in 2007 and one row wide in 2008. Plots were separated by one nonsprayed row of collards and 5 ft of nonsprayed plants within rows. Fungicides were applied on 5, 13, 18, 23, and 27 April and 2 May 2007 and 24 April, 1, 8, 15, and 22 May 2008 after downy mildew was detected in the field. Fungicides were applied at 60 psi with a CO2-backpack sprayer that had a 36-inch boom with three TeeJet (Wheaton, IL) TX 26 nozzles, orifice size 0.094 inch, spaced 18 inches apart. When potassium phosphite was mixed with other fungicides, it was added at one-half the rate it was applied when tested alone. To promote downy mildew, leaf wetness was increased by using Senninger (Clermont, FL) low pressure mini-wobbler sprinklers to produce a mist for 30 min in the early morning and evening from 23 March through 7 May 2007 and 14 April through 27 May 2008, except on days when fungicides were applied (Fig. 1). Average daily maximum and minimum temperatures for April and May were 77/51°F and 83/60°F in 2007 and 75/47°F and 82/53°F in 2008, respectively.


Fig. 1. Initial symptoms of downy mildew on collard leaves with leaf wetness provided by mini-wobbler sprinklers.


Fig. 2. Chlorotic and necrotic symptoms of downy mildew on a mature collard leaf from the lowest third of the plant.


Plots were rated visually for percentage leaf surface area with symptoms of downy mildew with a 0 to 15 modified Horsfall-Barratt scale from 5 April through 7 May 2007 and 25 April through 27 May in 2008 (Fig. 2). In most plots, necrotic lesions were more common than chlorosis of leaf laminas. In 2007, each row was rated separately, and ratings were averaged. Downy mildew symptoms were confirmed by observing sporulation of H. parasitica on the lower leaf surfaces. On 15 May 2007 and 28 May 2008, plants were harvested from a 5-ft section of one row per plot. Leaves with any symptoms of downy mildew were removed from the harvested plants. Healthy plant tissue (stems and asymptomatic leaves) were weighed. Data were analyzed with SAS v. 9.1 PROC MIXED (SAS Institute Inc., Cary, NC); treatment and year were fixed variables and block (and block-by-year in datasets combined across years) was a random variable. Disease severity ratings were converted to the mid-point of the percentage range associated with each rating before analysis. Percentages were used to calculate area under the disease progress curve (AUDPC). Single-degree-of-freedom contrasts were written to compare the mean of fungicides mixed with potassium phosphite to the mean of these fungicides applied alone.

Control of Downy Mildew on Collard with Fungicide Applications

When fungicide applications began in 2007, downy mildew was present on collard in all of the plots at ≤ 2% leaf surface area diseased. In 2008, downy mildew was detected on wild radish (Raphanus raphanistrum) in the field 1 week before the first fungicide application. Thereafter, downy mildew was found on collard in 4 of the 44 plots 1 day after the first fungicide application. The differences in the initiation times of downy mildew epidemics and fungicide applications are the most likely explanation for the higher severity in 2007 than in 2008. Nevertheless, there were no treatment-by-year interactions for end-of-the-season ratings (P = 0.44), area under the disease progress curve (AUDPC) (P = 0.33), or weight of healthy plants (P = 0.12).

At the next-to-the-last rating at 80 days after seeding in 2007, only plots treated with Sonata (AgraQuest, Davis, CA) mixed with ProPhyt had a lower severity of downy mildew than the control plots sprayed with water (Table 1). In contrast, at the next-to-the-last rating at 77 days after seeding in 2008, six treatments, Ranman (FMC Corp., Philadelphia, PA) mixed with ProPhyt, Presidio (Valent, Walnut Creek, CA) applied alone and in combination with K-Phite (Plant Food Systems, Zellwook, FL), K-Phite, Amistar (Syngenta, Wilmington, DE) mixed with ProPhyt, and Aliette, had lower severity than the control. The differences between the years may indicate that these fungicides were more effective when applied before infection of collard by H. parasitica than after infection. At the final rating, severity had decreased in 2007 and had plateaued in 2008 relative to the previous rating, most likely due to daily maximum temperatures that exceeded 82°F, which prevents spore germination in H. parasitica (7), on half (2007) or two-thirds (2008) of the intervening days. At the final ratings averaged across years, Presidio applied alone and in combination with potassium phosphite, both formulations of potassium phosphite, Amistar mixed with ProPhyt, and Aliette reduced severity of downy mildew compared to the control (Table 1).

Table 1. Severity of downy mildew on collard in 2007 and 2008.

Treatment, and
rate per acre
Severity (%),
30 April 2007
Severity (%),
16 May 2008

Area under
Potassium phosphite
(ProPhyt 4P), 4 pt
     10.3 abw       4.0 abc     4.6 cd      131 c
Potassium phosphite
(K-Phite 7LP), 4 pt
     13.5 ab       1.7 c     4.4 cd      138 c
Cyazofamid (Ranman 40SC), 2.75 oz      14.4 a       4.0 abc     7.0 abc      176 bc
Ranman, 2.75 oz +
ProPhyt, 2 pt
     15.3 a       2.5 bc     6.0 abcd      163 bc
Bacillus subtilis
(Sonata), 2 qt
     12.6 ab       6.3 ab     8.7 ab      194 ab
Sonata, 2 qt +
ProPhyt, 2 pt
      7.8 b       4.0 abc     5.9 abcd      126 c
(Presidio 4SC), 3 oz
     14.4 a       1.0 c     2.9 d      128 c
Presidio, 3 oz +
ProPhyt (2007), 2 pt or
K-Phite (2008), 2 pt
     12.6 ab       1.0 c     3.3 cd      127 c
Azoxystrobin (Amistar 80WDG), 3.5 oz +
ProPhyt, 2 pt
     10.3 ab       2.5 bc     4.8 cd      134 c
Aluminum tris O-ethyl
phosphonate (Aliette
80WDG), 2 lb
     11.9 ab       2.5 bc     5.1 bcd      139 c
Water      16.1 a       7.0 a     9.5 a      240 a
Mean without ProPhytx      13.8       3.8     6.2      166
Mean with ProPhyty      11.9       2.5     5.1      139
Probability of a
greater t valuez
0.16 0.18 0.21 0.0285

 u Rated 5 May 2007 and 22 May 2008, 87 and 88 days after seeding, respectively.

 v Mean over 2007 and 2008.

 w Means within a column with the same letters are not significantly different based on t tests of least-squares means, P ≤ 0.01.

 x Mean of Ranman, Sonata, and Presidio applied alone.

 y Mean of Ranman, Sonata, and Presidio applied in combination with ProPhyte. (K-Phite was used with Presidio in 2008.)

 z Preplanned, single-degree-of-freedom contrasts between Ranman, Sonata, and Presidio applied alone or in combination with potassium phosphite.

When the five ratings in 2007 and four ratings in 2008 were used to calculate AUDPC, all fungicide treatments except Sonata applied alone reduced AUDPC calculated across both years (Table 1). The mean AUDPC across Ranman, Sonata, and Presidio applied in combination with potassium phosphite was lower than the mean AUDPC with the three fungicides applied alone (P = 0.03). However, the reduction in AUDPC by potassium phosphite was most pronounced when combined with Sonata, which was not effective when applied alone at any rating period. Thus, the effectiveness of Sonata plus ProPhyt likely was due to the activity of ProPhyt. Combinations of ProPhyte (or K-Phite) with Ranman or Presidio did not differ compared to effectiveness of Ranman or Presidio applied alone. In another study, Manex plus ProPhyt was more effective than Manex applied alone against downy mildew on cabbage (1).

No fungicide was effective across all four disease severity measurements (Table 1). Five treatments, Presidio, Presidio plus Pro-Phyt or K-Phite, K-Phite, Amistar plus ProPhyt, and Aliette, reduced downy mildew severity in mid-season 2008 and at the final rating and also reduced AUDPC. The number of effective treatments in these experiments likely was influenced by the low to moderate severity of downy mildew that resulted from natural inoculum in the two years. With a higher inoculum level or more conducive environmental conditions, it is likely that fewer treatments would show a significant reduction in downy mildew severity, as illustrated by the severity ratings on 30 April 2007.

In general, Presidio was more effective than Ranman or Sonata in these experiments. Presidio is not currently registered on leafy brassica vegetables, such as collard; however, it is registered on head and stem brassica vegetables and likely would be an effective management option for downy mildew on those crops. For example, Presidio also was more effective than Ranman against H. parasitica on broccoli (12).

Collard Yield and Economic Return Increase with Fungicides

Standards used to assess marketable yields in this study were slightly more rigorous than those used in commercial practice. For example, leaves with any downy mildew lesions were not considered marketable, whereas in commercial practice one leaf per bunch with one downy mildew lesion would likely be considered marketable. Nevertheless, yields were approximately 50% higher than the average yields used in the South Carolina collard production budget (2). There was no treatment-by-year interaction. The two treatments that increased yield over the control were Presidio plus potassium phosphite and Amistar plus ProPhyt (Table 2).

Table 2. Marketable yield and economic returns of fungicides applied alone and in combination with potassium phosphite across 2007 and 2008.

Treatment and
rate per acre
5-ft row
Net return/
ProPhyt 4P, 4 pt      6.86 abw 7546 144 3718
K-Phite 7LP, 4 pt      6.76 ab 7439 144 3611
Ranman 40SC, 2.75 oz      6.08 b 6690 126 2880
Ranman, 2.75 oz +
ProPhyt, 2 pt
     7.25 ab 7976 186 4106
Sonata, 2 qt      6.10 b 6707 102 2921
Sonata, 2 qt +
ProPhyt, 2 pt
     7.05 ab 7754 162 3908
Presidio 4SC, 3 oz      7.10 ab 7814 159 3971
Presidio, 3 oz +
ProPhyt (2007), 2 pt or
K-Phite (2008), 2 pt
     8.08 a 8886 219 4983
Amistar 80WDG 3.5 oz
+ ProPhyt, 2 pt
     8.00 a 8798 261 4853
Aliette 80WDG, 2 lb      7.43 ab 8178 180 4314
Water      6.17 b 6792    0 3108
Mean without ProPhytx      6.43 7070 129 3257
Mean with ProPhyty      7.46z 8206 189 4333

 s Weight of healthy leaves and stems after leaves with downy mildew symptoms were removed.

 t  Average price per pound for collard was $0.375 (2).

 u Fungicide costs were calculated for six applications at the rates used in these experiments with 2009 prices in Charleston, SC. Cost of labor and machinery was $4/application (2).

 v Total production cost was $3684/acre not including fungicides applied for downy mildew (2).

 w Means within a column with the same letters are not significantly different based on t tests of least-squares means, P ≤ 0.01.

 x Mean of Ranman, Sonata, and Presidio applied alone.

 y Mean of Ranman, Sonata, and Presidio applied in combination with ProPhyte. (K-Phite was used with Presidio in 2008.)

 z P = 0.0084 for single-degree-of-freedom contrast between Ranman, Sonata, and Presidio applied alone and in combination with potassium phosphite. Other means were not compared statistically, since they are based on yields.

Total production costs, including all variable, fixed, and other costs, were estimated to be $3684/acre (2). Fungicides were applied six times in 2007 and five times in 2008; the cost of six applications was used in calculations. The season-long cost of the fungicide applications tested ranged from a low of $102/acre for Sonata to a high of $261/acre for Amistar plus ProPhyt (Table 2). These costs included $4 per application for costs of labor and machinery (2). Variable costs as used in these calculations included the cost of three applications of copper hydroxide, which are made to protect plants from black rot, and excluded the cost of two applications of Aliette and one application of azoxystrobin for downy mildew, because the costs of downy mildew fungicides from this study were used. Net returns to growers, after costs of production were subtracted from the crop values, ranged between $2880/acre and $4980/acre (Table 2). The two treatments that had the highest yields, Presidio plus potassium phosphite and Amistar plus ProPhyt, also had the highest net returns, over $4800/acre. Although these two treatments were the most costly, fungicide costs were a relatively small portion of the overall production costs, and so net returns also were high.

The mean yield across Ranman, Sonata, and Presidio applied in combination with potassium phosphite was higher than the mean yield with the three fungicides applied alone (preplanned, single-degree-of-freedom contrast, P = 0.01). With each of these three fungicides, yield was approximately 1 pound per 5-ft row (16%) greater when ProPhyt (or K-Phite) was added than when the fungicide was applied alone (Table 2). Adding potassium phosphite increased the cost of the treatment by $60 per acre (six applications at 2 pints per acre costing $5 per pint) or 57% of fungicide costs (excluding application costs). Net returns were an average of 33% greater when these three treatments included potassium phosphite than when the corresponding fungicide (Ranman, Sonata, and Presidio) was applied alone (Table 2). In addition, Amistar plus ProPhyt had the second highest net return. Thus, adding potassium phosphite to other fungicides increased weight of healthy collard leaves and net returns to growers after fungicide costs and other production costs were considered.

Conclusions and Recommendations

Downy mildew severity and yield of collard treated with Aliette did not differ from that treated with two formulations of potassium phosphite (ProPhyt and K-Phite). Slightly more phosphorus acid equivalent was applied with ProPhyt (2.1 lb/acre) than with Aliette (1.6 lb/acre) at the rates tested. It is possible that, if Aliette had been applied at a higher labeled rate (3 to 5 lb/acre) rather than at the lowest labeled rate (2 lb/acre), differences between these three phosphorous acid products might have been apparent. However, in 2008, there were no significant differences between ProPhyt and K-Phite applied alone at 2 pt/acre and Aliette at 2 lb/acre (10). Because potassium phosphite products are $40/acre compared to $65/acre for Aliette at the highest labeled rates, growers could reduce fungicide costs by substituting ProPhyt or K-Phite for Aliette. Note that K-Phite is registered on all brassica vegetables, whereas ProPhyt is registered only on head and stem brassicas. Other potassium phosphite products are also registered for brassica vegetables and would be expected to perform similarly.

It could not be determined from the data collected why yields of healthy collard leaves and stems were increased with fungicides mixed with potassium phosphite over the yields of the corresponding fungicides applied alone. In 2007, weight of the diseased leaves removed was recorded (9). In that trial, six of the seven treatments that reduced diseased weight compared to the water control included potassium phosphite. Since every leaf with a lesion was removed from the plants, weight of diseased leaves likely was lower because fewer leaves had downy mildew lesions. (A reduction in the size of lesions or number of lesions per leaf would not have impacted weight of diseased leaves.) Downy mildew usually begins on the oldest leaves at the bottom of the plant, which are smaller than the leaves in the middle of the plant. If potassium phosphite prevented spread of downy mildew to progressively younger, larger leaves, this might have reduced the weight of diseased leaves by limiting disease to older, smaller leaves that abscise when disease is severe (7). Additional experiments and more detailed measurements of epidemic progress and yield components are needed to determine the mechanism of yield protection by potassium phosphite.

Among the treatments that included fungicides currently registered for use on collard, Amistar plus ProPhyt was the only one that reduced downy mildew severity and increased yield and net return. Because azoxystrobin was not tested alone in this study, it was not clear how much of the control and yield increase were due to the activity of ProPhyt. However, in 2008, Amistar alternated with ProPhyt was less effective than Amistar mixed with ProPhyt (10). Based on the performance of three other fungicides with and without potassium phosphite, Amistar plus ProPhyt also may be more effective than Amistar applied alone.


Ginny DuBose and Andrew Lassiter provided technical support. This material is based upon work supported by the Southern Region IR-4 Program and NIFA/USDA under project number SC-1700294.

Technical contribution no. 5814 of the Clemson University Experiment Station.

Literature Cited

1. Adams, M. L., Holmes, G. J., and Thornton, A. C. 2008. Evaluation of fungicides for control of downy mildew of cabbage, Sampson County, 2007. Plant Disease Management Reports 2:V119. Online. DOI:10.1094/PDMR02.

2. Anonymous. 2008. Enterprise Budgets: Collards-Irrigated (Hand Harvest). Online. Coop. Ext., Dept. of Applied Economics and Statistics, Clemson Univ., Clemson, SC.

3. Anonymous. 2009. Phosphorous acid and its ammonium, sodium, and potassium salts (076002), Mono- and di-potassium salts of phosphorous acid (076416) Fact Sheet. Online.Biopesticide Active Ingredient Fact Sheets, US EPA, Washington, DC.

4. Constantinescu, O., and Fatehi, J. 2002. Peronospora-like fungi (Chromista, Peronosporales) parasitic on Brassicaceae and related hosts. Nova Hedwigia 74:291-338.

5. Farnham, M. W., Davis, E., H., Morgan, J. T., and Smith, J. P. 2008. Neglected landraces of collard (Brassica oleracea L. var. viridis) from the Carolinas (USA). Genet. Resour. Crop Evol. 55:797-801.

6. Farnham, M. W., Keinath, A. P., and Smith, J. P. 2001. Characterization of Fusarium yellows resistance in collard. Plant Dis. 85:890-894.

7. Felton, M. W., and Walker, J. C. 1946. Environmental factors affecting downy mildew of cabbage. J. Agric. Res. 72:69-81.

8. Holmes, G. J., and Kemble, J. M., ed. 2010. Southeastern U.S. Vegetable Crop Handbook. Vance Publishing Corp., Lincolnshire IL.

9. Keinath, A. P., DuBose, V. B., and Lassiter, A. W. 2008. Evaluation of fungicides applied after infection for control of downy mildew on collard, 2007. Plant Disease Management Reports 2:V025. Online. DOI:10.1094/PDMR02.

10. Keinath, A. P., Lassiter, A. W., and DuBose, V. B. 2009. Evaluation of fungicides applied alone or in combination with phosphonates to control downy mildew on collard, 2008. Plant Disease Management Reports 3:V032. Online. DOI:10.1094/PDMR03.

11. McKay, A. G., Floyd, R. M., and Boyd, C. J. 1992. Phosphonic acid controls downy mildew (Peronospora parasitica) in cauliflower curds. Aust. J. Exp. Agric. 32:127-129.

12. Raid, R. N. 2009. Evaluation of fungicides for control of downy mildew on broccoli, Spring 2009. Plant Disease Management Reports 4:V060. Online. DOI:10.1094/PDMR04.

13. Raid, R. N., and Meister, C. 2006. Evaluation of a phosphonic fungicide, alone and in combination, for downy mildew control on Chinese broccoli, Spring 2005. Fungic. Nematic. Tests 61:V087. Online. DOI:10.1094/FN61.

14. Raid, R. N., and Meister, C. 2006. Influence of spray initiation on downy mildew of Chinese broccoli, Spring 2005. Fungic. Nematic. Tests 61:V089. Online. DOI:10.1094/FN61.