Dynamics of development of late blight [Phytophthora infestans] in ...

84

Epidemiology / Épidémiologie

Dynamics of development of late blight [Phytophthora infestans] in potato, and comparative resistance of cultivars in the highland tropics O.M. Olanya, P.S. Ojiambo, and R.O. Nyankanga

Abstract: Host resistance is an important component for the management of late blight [Phytophthora infestans] on potato, in the highland tropics, where effective fungicide use is limited because of the cost of application. Potato cultivars with major resistance genes to late blight (population A) and minor or quantitative resistance genes to the disease (population B) were evaluated in field studies at two locations in Kenya during 2000, 2001, and 2002 cropping seasons. Disease severity, area under disease progress curve (AUDPC), infection rates, tuber blight, and tuber yields were assessed to determine the effectiveness of cultivar resistance to potato late blight. Significant differences (P < 0.05) in AUDPC were detected among cultivars. Progress of late blight on potato cultivars was best described by the nonlinear form of the logistic model. Infection rates ranged from 0.0047 to 0.3105 logits per day and were generally higher on susceptible than resistant cultivars. Rates of disease progress were highest on the susceptible control ‘Kerr’s Pink’ (0.3015 logits per day) and lowest on ‘Rutuku’ (0.0047 logits per day), a cultivar derived from population A. The rates of disease progress on population B cultivars were significantly (P < 0.001) higher than those on population A in seasons of severe late-blight epidemics. Significant (P < 0.05) differences in tuber blight development were also detected among the cultivars. Significant negative correlations were observed between AUDPC and tuber yield for cultivars of population A, but no correlations were observed for cultivars of population B. This suggests that cultivars of population B are less influenced in their yield by late blight than those of population A and may be more suited for use where late blight is a recurring problem. Key words: infection rates, Phytophthora infestans, population A, population B, resistance, potato, highland tropics, Kenya. 94

Résumé : La résistance de l’hôte est un élément important de la lutte contre le mildiou [Phytophthora infestans] de la pomme de terre dans les régions montagneuses des tropiques, là où un usage efficace des fongicides est limité par le coût de l’application. Des cultivars de pomme de terre avec des gènes majeurs de résistance contre le mildiou (population A) et avec des gènes mineurs ou de résistance quantitative (population B) conter la maladie ont été évalués lors d’études en champ menées à deux sites au Kenya au cours des saisons de végétation de 2000, 2001 et 2002. La gravité de la maladie, la surface sous la courbe de progression de la maladie (SCPM), les taux d’infection, le mildiou du tubercule ainsi que le rendement en tubercules ont été étudiés afin de connaître l’efficacité de la résistance des cultivars de pomme de terre au mildiou. Des différences significatives (P < 0,05) ont été trouvées parmi les cultivars pour la SCPM. La forme non linéaire du modèle logistique a le mieux expliqué la progression du mildiou dans les cultivars de pomme de terre. Les taux d’infection allaient de 0,0047 à 0,3105 logits par jour et étaient habituellement plus élevés pour les cultivars sensibles que pour les résistants. La vitesse de progression de la maladie était plus rapide pour le témoin sensible ‘Kerr’s Pink’ (0,3015 logits par jour) et plus lente pour ‘Rutuku’ (0,0047 logits par jour), un cultivar issu de la population A. La vitesse de progression de la maladie des cultivars de la population B était

Accepted 24 November 2005. O.M. Olanya.1 Agricultural Research Service, US Department of Agriculture, New England Plant, Soil and Water Laboratory, University of Maine, Orono, ME 04469, USA. P.S. Ojiambo. Department of Plant Pathology, University of Georgia, Athens, GA 30602, USA. R.O. Nyankanga. Department of Horticulture, Cornell University, Ithaca, NY 14853, USA. 1

Corresponding author (e-mail: [email protected]). Note: Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture, University of Georgia, or Cornell University.

Can. J. Plant Pathol. 28: 84–94 (2006)

Olanya et al.: late blight on potato / Phytophthora infestans / disease progress and severity / resistant cultivars

85

significativement (P < 0,001) plus rapide que celle des cultivars de la population A lors de saisons d’épidémie grave de mildiou. Des différences significatives (P < 0,05) ont aussi été trouvées parmi les cultivars pour le développement du mildiou des tubercules. Des corrélations négatives significatives ont été observées entre la SCPM et le rendement en tubercules pour les cultivars de la population A, mais aucune corrélation n’a été observée pour les cultivars de la population B. Cette observation laisse croire que les cultivars de la population B ont un rendement moins affecté par le mildiou que ceux de la population A et qu’ils peuvent être mieux adaptés aux endroits où le mildiou est un problème chronique. Mots clés : taux d’infection, Phytophthora infestans, population A, population B, résistance, pomme de terre, régions montagneuses des tropiques, Kenya. Olanya et al.: late blight on potato / Phytophthora infestans / disease progress and severity / resistant cultivars

Introduction Potato (Solanum tuberosum L.) is an important cash and food crop in many parts of the world (Haverkort 1990). In the tropical highlands of East Africa, normal temperature conditions (15–22 °C) and average rainfall (>1200 mm) during the cropping season are often suitable for potato cultivation. Potato cultivars are usually grown in solid plantings or mixed cropping systems and concentrated in the densely populated tropical highlands in Kenya, where altitude ranges from 1500 to 2500 m (Jaetzold and Schmidt 1982). The crop is grown with minimum fungicide and under dense-planting and high-rainfall conditions, which favor the development of late blight [Phytophthora infestans (Mont) de Bary] (Olanya et al. 2001). Late blight is a significant constraint for potato production in the tropical highlands of East Africa. Yield losses attributed to the disease are estimated to be in the range of 35% to 75% (Olanya et al. 2001). Although control of potato late blight in tropical Africa can be achieved through the use of fungicides, the fungicide costs are often prohibitive for the small-scale potato growers, who are the major producers of potato within the region. Moreover, periodical fungicide applications, timing, and dosage are often made more difficult by frequent tropical rainfall and limited knowledge of small-scale growers on fungicide application (Kankwatsa et al. 2002; Nyankanga et al. 2004). In addition, the development of resistance to commonly used fungicides such as metalaxyl (Ciba Corporation, Basel, Switzerland) has also made control of late blight very difficult (Mukalazi et al. 2001). The International Potato Center (Centro Internacional de la Papa; CIP) and national potato programs in sub-Saharan Africa have made major efforts, for many years, to develop host resistance or tolerance in potato, one of the most economic methods for late-blight control in the tropical highlands (El-Bedewy et al. 2001). Utilization of two populations of potato genotypes (populations A and B) developed at CIP has been documented for late-blight control (Landeo et al. 1997). Population A consists of genotypes with major genes for resistance to late blight, which are effective on some races or isolates of the pathogen. Population A was developed by introgression of resistance derived from Solanum demissum Lindley into S. tuberosum subsp. tuberosum L., S. tuberosum subsp. andigena Hawkes, and neotuberosum clones (Landeo and Turkensteen 1989; Landeo et al. 1995). Population B consists of genotypes with quantitative resistance to late blight, which is effective against a broad range of pathogen races or isolates. From a four-way hybrid cross between Solanum acaule Bitter, Solanum bulbocastanum

Dunal, Solanum phureja Juz & Buk, and S. tuberosum, genotypes with horizontal resistance free of R genes have been developed (Landeo et al. 1995, 1997; Trognitz 1995, 2001). R genes were removed by crossing the resulting clones of population B to S. tuberosum cultivars free of R genes (Landeo et al. 1995). The absence of race-specific R genes in population B has been tested and confirmed (Landeo et al. 1995, 1997, 2000, 2001). Many of the potato cultivars grown in Kenya have very low to moderate levels of resistance to late blight. Quantification of late blight in these cultivars with different levels of resistance is important for prediction of yield performance under various scenarios of late-blight pressure. Because the population of P. infestans in East Africa is of the US-1 genotype (old clonal lineage) and A1 mating type (Ochwo et al. 2002; Vega-Sanchez et al. 2000), it is anticipated that cultivar resistance or tolerance (Browning et al. 1977) would be a suitable strategy for disease control. The US-1 clone of P. infestans is not as aggressive as the US-8 genotype (Lambert and Currier 1997), therefore, it requires less fungicide applications, which is more cost effective for resource-poor growers. However, cultivar resistance can be deployed and managed by potato farmers to control late blight effectively under tropical highlands of East Africa. The objectives of the present study were to quantify levels of resistance to late blight, disease progress, tuber blight, and the relationships of disease severity to tuber yield in potato cultivars with different levels of resistance to late blight in Kenya. This would provide valuable information for the utilization of resistant cultivars as part of integrated management of potato late blight in tropical highlands of East Africa.

Materials and methods Description of study sites and data collection Experiments were established at two locations, Loreto (2100 m) and Kabete (1800 m), during the 2000, 2001, and 2002 cropping seasons. At Loreto, experiments were conducted during the long-rain season (LR; April–August) in all 3 years, while at Kabete, the experiments were conducted during the short-rain season (SR; September–December), in 2000, and during the LR, in 2001 and 2002. Soils at the two experimental sites are typically Humic Nitosols (Jaetzold and Schmidt 1982). Potato is often severely infected by P. infestans throughout the year at these two sites, and artificial inoculation was not necessary. Eight potato cultivars: ‘Asante’, ‘Tigoni’, ‘Kerr’s Pink’, ‘Nyayo’, ‘Rutuku’, ‘Awash’, ‘Genet’, and ‘Cruza 148’ were evaluated in this study (Table 1). The cultivars were bred

86

Can. J. Plant Pathol. Vol. 28, 2006 Table 1. General description of potato cultivars used for quantifying various parameters of epidemics of late blight [Phytophthora infestans] at two locations in Kenya. Cultivar

Clonea

Maturityb

Skin color

Genotype groupc

‘Asante’ ‘Awash’ ‘Cruza 148’ ‘Genet’ ‘Kerr’s Pink’ ‘Nyayo’ ‘Rutuku’ ‘Tigoni’

381381.2 378501.3 720118 800984 — — 720097.1 381381.13

Medium-late Early Late Medium Early Early Late Medium-late

Red White Purple white White Pink White Deep red White

Population Population Population Population — — Population Population

B B A B

A B

a

Clone from which the cultivar was derived. Maturity (in days after planting): early < 75; 75 ≤ medium < 90; 90 ≤ medium-late < 110; late ≥ 110. c Population A, cultivars with qualitative resistance and one or two major genes; population B, cultivars with quantitative resistance and minor resistance genes. b

from different sources and included clones derived from CIP populations A and B (Table 1). ‘Tigoni’ was used as resistant control while ‘Kerr’s Pink’, a Dutch cultivar introduced and grown in many parts of Kenya, served as susceptible control. ‘Nyayo’ is of unknown origin but widely grown in Kenya and has moderate levels of resistance to late blight. Potato seed for the eight cultivars were derived from tissue culture and obtained from the plant quarantine station located at the Kenyan Plant Health Inspectorate Service facility in Muguga. These were planted in a randomized complete block design with three replications, using a two-row tractor planter. Field plots consisted of four rows each containing 20 tubers with spacings of 75 cm and 30 cm between rows and seed tubers, respectively. In all field plots, normal agronomic practices for potato production were followed. Fertilizers (175N and 175P) were applied at the rate of 500 kg of diammonium phosphate per hectare at planting. Insecticide (metasystox, Bayer East Africa Ltd., Nairobi, Kenya) was applied as needed for aphid control. No fungicides were applied to the field plots. Field plots were assessed for late-blight development by visual rating of foliage for percent leaf area blighted or defoliated (disease severity) beginning from the time when 2% to 5% leaf-area damage was noticed on the most severely infected cultivar. Subsequent evaluations of disease severity were recorded weekly, based on visual assessments on a scale of 0% to 100% (0%, no disease; 100%, total foliage damage) (James 1971), until severity on the most susceptible cultivar approached 100%. At harvest, tubers from the two inner rows of each experimental plot were harvested, weighed, and separated into seed size (tuber diameter < 6.4 cm) and table stock (tuber diameter > 6.4 cm). Tuber masses were converted to metric tons per hectare for subsequent analysis. Incidence of tuber blight in the field was assessed at harvest by recording the number of tubers with typical symptoms, for each cultivar and plot, and was expressed as a percentage of total number of tubers. Data analysis Data on percent leaf area blighted were fitted to nonlinear forms of the logistic and Gompertz models (Neher and Campbell 1997), using the Marquardt method in PROC NLIN procedure of SAS®, release 8.2 (SAS Institute Inc.

2005) to describe temporal late-blight progress and estimate model parameters (initial disease severity, rate of disease progress, and final disease severity). The most suitable model for parameter estimation was determined using standardized residual plots, adjusted coefficients of determination, and asymptotic standard errors of estimates of initial disease severity and rates of disease progress (Neher and Campbell 1997). Estimates of the rate parameter, r, derived from the model with the best fit, were compared among cultivars by ANOVA after fitting the best model to observed disease severity values of each cultivar for each replication (Reynolds and Neher 1997). Mean rates of disease progress among cultivars were then separated with Fisher’s protected least significant difference (LSD) test at α = 0.05. Within each cropping season, area under disease progress curve (AUDPC) was calculated for each cultivar, following the midpoint method (Campbell and Madden 1990), as: AUDPC = Σ{[(yi + yi+1)/2](ti+1 – ti)} where y is the percent disease severity, t is the time period, i and i + 1 represent weekly observations from 1 to n. Tuber blight, AUDPC, and yield for different cultivars were compared with ANOVA. Fisher’s protected LSD was employed to separate means of potato cultivars by location and year where the ANOVA indicated significant differences (P < 0.05). The ANOVA was performed using SAS/STAT, General linear model – VARCOM, version 6 (SAS Institute Inc. 2005). The relative resistance of potato cultivars was determined by ranking cultivars within years, on the basis of their AUDPC values. To determine the general trend of the effect of late-blight epidemics on potato yield and yield components, correlation analysis was conducted between AUDPC and yield (t/ha), unmarketable tuber yield, number of seed tubers, and total number of tubers. At each location, the relationship between AUDPC and yield and between AUDPC and yield components within each population were determined with PROC CORR procedure of SAS (SAS Institute Inc. 2005) based on combined data over seasons.

Results Disease severity and progress of potato late blight The disease progress curves on the cultivars had similar trends with observable differences between seasons (Figs. 1

Olanya et al.: late blight on potato / Phytophthora infestans / disease progress and severity / resistant cultivars Fig. 1. Temporal progress of late blight [Phytophthora infestans] on eight potato cultivars with different levels of resistance to the disease, during the long-rain (LR) cropping seasons at Loreto, Kenya, from 2000 to 2002. Data points are mean disease severity values, the vertical bars represent standard errors of the mean observed disease values, and the curves are predicted disease severity values obtained from fitting a nonlinear logistic model to observed late-blight severity values.

87

88

Can. J. Plant Pathol. Vol. 28, 2006

Fig. 2. Temporal progress of late blight [Phytophthora infestans] on eight potato cultivars with different levels of resistance to the disease, during the short-rain (SR) and long-rain (LR) cropping seasons at Kabete, Kenya, from 2000 to 2002. Data points are mean disease severity values, the vertical bars represent standard errors of the mean observed disease values, and the curves are predicted disease severity values obtained from fitting a nonlinear logistic model to observed late-blight severity values.


89

Table 2. Estimates of infection rates (r) of late blight caused by Phytophthora infestans on potato cultivars with different levels of resistance, during three cropping seasons at Loreto and Kabete, in Kenya, from 2000 to 2002. Loreto

Kabete

Cultivara

r ± SE

R2 (%)

r ± SE

2000 ‘Asante’ ‘Awash’ ‘Cruza 148’ (A) ‘Genet’ ‘Kerr’s Pink’ ‘Nyayo’ ‘Rutuku’ (A) ‘Tigoni’

0.0588 0.0712 0.0514 0.0387 0.1012 0.0647 0.0487 0.0502

± ± ± ± ± ± ± ±

0.0118 0.0007 0.0009 0.0026 0.0018 0.0125 0.0011 0.0091


0.0892 0.1550 0.0778 0.0846 0.3105 0.1730 0.0921 0.0854

± ± ± ± ± ± ± ±


0.0951 0.1557 0.0700 0.0630 0.1584 0.0812 0.0368 0.0981

± ± ± ± ± ± ± ±

R2 (%)

bcd b d e a bc d cd

88.9 98.4 89.5 91.6 97.1 87.9 85.8 86.8

0.1936 0.2076 0.1312 0.1894 0.2889 0.2422 0.1605 0.1771

± ± ± ± ± ± ± ±

0.0233 0.0099 0.0196 0.0099 0.0575 0.0239 0.0179 0.0126

c c d c a b c c

94.1 98.9 96.1 98.0 92.5 97.8 91.8 97.4

0.0074 0.0111 0.0072 0.0117 0.0321 0.0147 0.0096 0.0101

c b c c a b c c

94.4 97.5 93.4 90.8 96.8 96.7 92.2 90.1

0.0833 0.1227 0.0690 0.0673 0.1267 0.1281 0.0788 0.1005

± ± ± ± ± ± ± ±

0.0142 0.0141 0.0119 0.0114 0.0110 0.0291 0.0189 0.0081

bc a c c a a c ab

91.2 95.1 93.4 90.1 91.4 86.4 90.9 96.9

0.0096 0.0101 0.0066 0.0088 0.0262 0.0078 0.0073 0.0115

cd ab cd cd a dc d bc

92.2 98.3 90.1 91.3 90.1 91.0 86.3 93.4

0.0724 0.0981 0.0488 0.0668 0.1024 0.0966 0.0583 0.0757

± ± ± ± ± ± ± ±

0.0087 0.0092 0.0047 0.0085 0.0079 0.0088 0.0101 0.0079

ab a d cd a a d bc

95.1 90.1 91.5 89.9 92.8 89.6 92.9 91.0

Note: Parameter estimates are from a nonlinear regression based on the logistic model. All infection rates ± standard error of the mean (r ± SE) were evaluated during the long-rain seasons at the two locations, except at Kabete in 2000, where the assessments were in the short-rain season. Rates followed by the same letter are not significantly different according to Fischer’s least significant difference at P = 0.05. R2, adjusted coefficient of determination for goodness of fit of the logistic model to disease severity over time. b Cultivars identified by (A) are from population A (‘Cruza 148’ and ‘Rutuku’); others are from population B.

and 2). At Loreto, final disease values were highest in the 2002 LR season and lowest in the 2000 LR season (Fig. 1). At Kabete, final disease levels were highest in 2000 SR season and lowest during the 2001 LR season (Fig. 2). ‘Kerr’s Pink’ had the highest final disease level, irrespective of the location and season. At both sites, the logistic model had a better fit to disease severity data for all the cultivars evaluated. At Loreto and Kabete, the adjusted coefficients of determination (R2) were high in all 3 years (Table 2). The estimated initial disease severity values were generally similar at both locations. At both locations, significant differences (P < 0.05) in infection rates were observed among the cultivars in all seasons of study (Table 2). At Kabete, ‘Kerr’s Pink’, a highly susceptible control, had significantly (P < 0.05) greater infection rates than ‘Rutuku’ and ‘Cruza 148’ from population A, in all the growing seasons. Differences in infection rates between ‘Asante’ and ‘Tigoni’ from population B and ‘Kerr’s Pink’ were not consistent. Cultivars derived from population A (‘Rutuku’ and ‘Cruza 148’) had significantly (P < 0.05) lower infection rates than those from population

B (‘Asante’, ‘Tigoni’, and ‘Genet’) in 2001 and 2002 LR season (Table 2). At Loreto, differences in infection rates were also recorded among cultivars. Infection rates were significantly (P < 0.05) higher on the susceptible control ‘Kerr’s Pink’ than on ‘Rutuku’ and ‘Cruza 148’ from population A, in all cropping seasons. In all seasons, except the 2000 LR, ‘Cruza 148’, a resistant cultivar from population A, had significantly (P < 0.05) lower rates of disease progress than those observed for ‘Tigoni’ and ‘Asante’, which are derived from population B (Table 2). Resistance of potato cultivars to late-blight infection There was a significant effect (P < 0.001) of population type on AUDPC (Table 3). Both the population type and cultivars used in the present study significantly affected tuber numbers (P < 0.0002 and P < 0.0001, respectively; Table 3), and yield (P < 0.0001 and P < 0.0001, respectively; Table 3). In general, AUDPC levels were lower on cultivars from population A than from population B in Loreto. Average AUDPC values were significantly (P < 0.05) greater for

90

Can. J. Plant Pathol. Vol. 28, 2006 Table 3. Analysis of variance on the effect of resistance of potato cultivars on tuber yield and yield components as impacted by late blight [Phytophthora infestans] at Loreto and Kabete, in Kenya, from 2000 to 2002. Source

DF

F value

P >F

AUDPC (% disease days) Replication (REP) 2 Population 1 Cultivar (CV) 7 REP*CV 14 Location (LOC) 1 CV*LOC 7 Season (SEAS) 2 SEAS*LOC 2

0.03 10.95 64.54 1.85 26.99 4.23 1.96 1.30

0.9742 0.0010 0.0001 0.2214 0.0001 0.0010 0.1850 0.3535

Tuber blight (%) REP CV REP*CV LOC CV*LOC SEAS SEAS*LOC

2 7 14 1 7 2 2

2.38 15.24 0.80 61.99 1.12 2.49 0.47

0.1250 0.0562 0.6564 0.0001 0.3953 0.6877 0.2450

Tuber number REP Population CV REP*CV LOC CV*LOC SEAS SEAS*LOC CV*LOC*SEAS

2 1 7 14 1 7 2 2 14

3.28 15.17 61.82 0.72 14.33 2.63 52.72 2.81 4.81

0.0723 0.0002 0.0001 0.7651 0.0021 0.0171 0.0001 0.0001 0.0001

Seed mass REP Population CV REP*CV LOC CV*LOC SEAS SEAS*LOC CV*LOC*SEAS

2 1 7 14 1 7 2 2 14

2.53 3.12 18.19 1.46 141.15 2.73 127.66 42.81 7.25

0.0912 0.0843 0.0001 0.1594 0.0001 0.0200 0.0001 0.0001 0.0001

Yield (t/ha) REP Population CV REP*CV LOC CV*LOC SEAS SEAS*LOC CV*LOC*SEAS

2 1 7 14 1 7 2 2 14

0.15 22.80 21.79 0.64 14.01 5.39 227.69 119.46 6.39

0.8600 0.0001 0.0001 0.8390 0.0001 0.0001 0.0001 0.0001 0.0001

Note: Analysis from data collected on average tuber yield, mean seed number, tuber numbers, percent disease severity, and tuber blight assessed on potato cultivars with different levels of resistance to late blight during three cropping seasons and at two locations. AUDPC, area under disease progress curve; DF, degrees of freedom.

the susceptible control ‘Kerr’s Pink’ than for all other cultivars in all seasons at Loreto and in 2000 at Kabete (Table 4). At Loreto, for data averaged across seasons, AUDPC values were lowest for ‘Rutuku’ (180.3) and ‘Cruza 148’ (496.5) from population A; the average value for ‘Rutuku’ was significantly different (P < 0.05) from those of other cultivars while that of ‘Cruza 148’ was also significantly different from other cultivars except ‘Genet’ (Table 4). Cultivars from population B, such as ‘Tigoni’ and ‘Asante’, had significantly higher AUDPC values than cultivars from population A in 2001 and 2002 at Loreto (Table 4). In the 2001 LR season, AUDPC values averaged across cultivars were 840.3 at Loreto and 124.3 at Kabete. Tuber yield of potato cultivars Tuber yields were higher at Loreto compared with Kabete in the 2000 cropping season. However, tuber yield was higher at Kabete compared with Loreto in 2001 and 2002 LR seasons (Table 4). In the 2000 season, average tuber yield ranged from 30 to 60.7 t/ha at both locations (Table 4). No significant differences in the yield of ‘Tigoni’, ‘Asante’, and ‘Genet’, from population B, and ‘Rutuku’, from population A, ‘ were detected at Kabete. However, in the 2001 season at Kabete, yield ‘Rutuku’ was significantly lower than that of ‘Asante’ and ‘Genet’. The highly susceptible control, ‘Kerr’s Pink’, which had higher disease levels, also had significantly lower tuber yields at Kabete in 2001 and 2002 LR seasons. ‘Nyayo’ had higher tuber yield than cultivars from population A and B in 2002 at Loreto. In 2002, tuber yield was significantly (P < 0.05) lower than in the 2001 and 2000 cropping seasons at Kabete and 2000 cropping season at Loreto (Table 4). Tuber blight incidence on potato cultivars from field studies In the 2000 cropping season, no tuber blight was recorded at Loreto and Kabete locations. During 2001, the average tuber blight infection ranged from 0% to 3.3% at Kabete and from 0.8% to 13.8% at Loreto. Significant differences in tuber blight incidence were observed among cultivars at Loreto but not at Kabete in 2001. At Loreto in 2001, average tuber blight incidences were 13.8%, 0.8%, and 8.7% on ‘Tigoni’, ‘Asante’, and ‘Genet’, respectively. On ‘Rutuku’, ‘Cruza 148’, and the susceptible control ‘Kerr’s Pink’, average tuber blight incidences were 4%, 7.8%, and 10%, respectively. Tuber blight incidences on ‘Tigoni’, ‘Genet’, ‘Kerr’s Pink’, and ‘Nyayo’ were not significantly different from each other but differed significantly from those on ‘Asante’, ‘Rutuku’, ‘Cruza 148’, and ‘Awash’, which had average tuber blight incidences of 0.8%, 4%, 7%, and 1.3%, respectively. The tuber blight incidences on ‘Nyayo’, ‘Genet’, and ‘Cruza 148’ were not significantly different from each other but differed significantly from those on ‘Asante’, ‘Rutuku’, and ‘Awash’. Very low incidence of tuber blight was recorded at Kabete and Loreto in 2002, and differences in the incidence among the potato cultivars were not significantly different. Relationships of tuber yield, seed mass, and number of seed tubers with tuber blight and disease levels Significant variation in seed mass, number of seed tubers, and total number of tubers (P < 0.01) was observed among


91

Table 4. Area under disease progress curves (AUDPC) and yield of potato cultivars evaluated for development of late blight [Phytophthora infestans] at Loreto and Kabete, in Kenya, from 2000 to 2002. Loreto

Kabete

Cultivar

AUDPC (% disease days)

Yield (t/ha)a

AUDPC (% disease days)

Yield (t/ha)a

2000 ‘Tigoni’ ‘Asante’ ‘Nyayo’ ‘Rutuku’ ‘Genet’ ‘Kerr’s Pink’ ‘Cruza 148’ ‘Awash’ Mean

87.3 64.7 121.6 25.0 2.5 471.3 25.7 — 114.3

bc bcd b cd d a cd

60.7 56.6 54.5 47.0 38.1 36.8 35.8 — 47.1

a a a a b b b

295.2 187.8 714.0 199.5 504.0 1777.7 204.2 1281.0 645.4

de e c e cd a e b

46.9 48.9 46.2 39.6 40.5 33.7 30.0 38.1 31.7

a a a ab a b b ab


703.4 834.2 902.1 323.2 334.8 2248.1 396.2 980.8 840.3

b b b c c a c b

17.8 13.1 11.8 15.2 11.8 2.8 14.5 9.5 15.5

a ab ab ab ab c ab bc

23.8 18.8 322.6 42.4 57.9 427.7 42.2 58.9 124.3

b b a b b a b b

41.9 45.1 32.5 28.7 43.1 16.9 35.4 30.6 34.3

a a b b a c ab b


2016.4 1799.1 1356.8 193.3 952.2 3086.0 1067.5 2612.5 1635.5

c cd de f e a e b

12.1 10.8 19.5 13.9 12.0 3.5 10.4 5.8 11.0

b b a b b c b bc

385.7 237.3 925.8 36.9 378.6 1078.8 305.8 863.1 526.5

b bc a c b a bc a

23.4 23.8 30.7 12.1 23.1 10.0 14.1 12.5 19.1

ab ab a bc ab c bc bc

Note: Cultivars were evaluated during the long-rain seasons (March to July) at the two locations, except at Kabete in 2000, where the assessments were in the short-rain season (September to December). The AUDPC and yield values followed by the same letters are not significantly different according to Fischer’s least significant difference at P = 0.05. a Average yield of potato tubers from three replications.

the cultivars with different levels of resistance and among cropping seasons or years. At Kabete, AUDPC of population B cultivars was negatively and significantly (P < 0.05) correlated with seed mass and number of seed tubers. Significant negative correlations were detected between AUDPC and tuber yield for cultivars from population A at Loreto and Kabete, but not on cultivars derived from population B (Table 5).

Discussion Resistance to late blight in the potato cultivars evaluated in this study was generally expressed as low AUDPC values and lower rates of disease progress. The highest infection rates were detected on the susceptible control ‘Kerr’s Pink’ compared with cultivars derived from population A or B. Except on ‘Asante’, infection rates observed on population A cultivars were significantly lower than rates observed on population B cultivars, particularly in seasons when disease level was high. Differences in infection rates observed in

this study may be explained by the inherent nature of resistance of cultivars to infection by P. infestans (Landeo and Turkensteen 1989; Ojiambo et al. 2000). ‘Cruza 148’ and ‘Rutuku’ with qualitative resistance showed the lowest disease and infection rates compared with population B cultivars, in the presence of severe late-blight epidemics. Lower disease and infection rates of the population A cultivars are indicative of their ability to limit infection (Ojiambo et al. 2000). ‘Tigoni’, ‘Genet’, and ‘Asante’ from population B have quantitative resistance that permits infection by the pathogen, but yield was not correlated with AUDPC (El-Bedewy et al. 2001). Thus, in this study, significantly higher infection rates were observed on the cultivars from population B than on those from population A. Several studies have documented the contribution of R genes to resistance to P. infestans and the resulting disease levels (Landeo et al. 1995; Trognitz et al. 1995, 2001). It has also been shown that R genes (major resistance genes) can be quite effective for late-blight control in the absence of compatible races or isolates of the pathogen (Landeo and

92

Can. J. Plant Pathol. Vol. 28, 2006 Table 5. Pearson’s correlation coefficients of area under disease progress curve (AUDPC), for development of late blight [Phytophthora infestans], and tuber yield and yield components for potato cultivars evaluated during three growing seasons at Loreto and Kabete, in Kenya, from 2000 to 2002. AUDPC (% disease days) Potato tuber yield and yield components

Loreto

Seed mass (t/ha) Unmarketable mass (t/ha) Yield (t/ha) Seed number Tuber number

–0.5290 0.1001 –0.7208* –0.2918 –0.3204

Population A

Kabete Population B 0.0549 –0.3000 –0.1165 0.1947 –0.5007

Population A

Population B

–0.2912 0.3151 –0.6947* –0.0253 –0.0723

–0.8760* 0.1983 –0.2617 0.8605* –0.0995

Note: Correlation coefficients followed by an asterisk indicate a significant correlation at Loreto (P = 0.0036, n = 14 and P = 0.0378, n = 9) for populations A and B, respectively, and at Kabete (P = 0.0221, n = 9; P = 0.0412, n = 6; and P = 0.0278, n = 6) for seed mass, yield, and seed number, respectively. At each location, the correlation analysis was performed based on combined data of AUDPC, yield, and yield components from the three growing seasons.

Turkensteen 1989). In this study, we attribute the low disease levels on ‘Rutuku’ and ‘Cruza 148’ (population A) to the presence of major resistance genes. Population B clones were developed from cultivars of S. tuberosum subsp. andigena with back-crosses to R-gene-free S. tuberosum to increase gene frequencies for quantitative resistance (Landeo et al. 1997). Even though previous studies have documented that the ability of host cultivars to limit development of late blight is an important criteria in reduction of infection rate or quantitative resistance (Landeo et al. 1997, 2001; Umaerus and Umaerus 1994), our data showed that rates of disease progress were more rapid in population B cultivars, which have quantitative resistance, and that disease effects on yield were minimal. Perhaps, cultivars from population B have differential expression of quantitative resistance (Landeo et al. 2000). Differences in AUDPC were detected among potato cultivars, but the relative ranking of cultivars for resistance to late blight were consistent across years and locations. The ranking of potato cultivars for resistance to late blight in 2000, 2001, and 2002 cropping seasons indicated that ‘Cruza 148’ and ‘Rutuku’ with major resistance genes consistently showed higher levels of resistance to late blight across years and locations than most population B cultivars. However, at Kabete in 2001, because of low disease levels, there were nonsignificant differences in AUDPC among cultivars from populations A and B. Differences in late-blight severity in response to cultivar resistance in tropical environments have been reported previously (Olanya et al. 2001). The variation in disease reaction between locations and years is attributed to environmental conditions during the cropping seasons. In 2001 and 2002 at Loreto and in 2000 at Kabete, environmental conditions such as low temperatures, higher relative humidity, and rainfall amounts were more conducive to disease development compared with other cropping seasons that were dry (data not presented). In tropical highlands where inoculum is present, environmental conditions consisting of lower ambient temperatures and high relative humidity and rainfall amounts are often conducive to late-blight development (Ojiambo et al. 2000; Olanya et al. 2001). In absence of favorable environmental conditions in 2000 and 2001 (data not shown), low late-blight epidemics were observed. No significant differences in tuber blight incidence under field conditions were detected on potato cultivars evaluated

at Kabete and Loreto. The lack of significant differences in tuber blight incidence incited by the US-1 genotype indicates that numerous factors, other than the amount of inoculum, such as rainfall, tuber maturity and susceptibility, and soil temperature and moisture, may have more influence on tuber blight development in tropical conditions. Differences in tuber resistance to late-blight infection in field conditions have been attributed to variation in the development of the periderm (Lapwood 1977), resistance in lenticels (Lacey 1967), maturity levels of tubers (Grinberg et al. 1995), or resistance of cultivars (Flier et al. 2001). Because of the low levels of tuber blight observed in the field studies, the effects of resistance levels on tuber blight infection could be ascertained. However, in other studies, resistance to late blight in potato foliage and tubers was noted to be determined by the same genes (Stewart et al. 1994). Differences in tuber resistance to blight under temperate conditions have been previously reported and attributed to differences in genotypes of P. infestans (Lambert and Currier 1997). The significant negative correlation between AUDPC and tuber yield observed at the two locations on genotypes from population A indicates that high late-blight epidemics can significantly affect tuber yield. Lower tuber yield on cultivars from population A, despite good resistance to late blight, have been previously reported (El-Bedewy et al. 2001). Quantitative relationships between late blight of potato and loss in tuber yield have been attributed to disease effects on foliage loss as well as cultivar effects (James et al. 1972). Therefore, high levels of late blight may have resulted in reduced tuber yields as a consequence of defoliation, which has been observed in previous studies (Nyankanga et al. 2004; Olanya et al. 2001). The lack of significant correlation between AUDPC and tuber yield on cultivars from population B suggests that those cultivars may be able to tolerate disease with little effect on tuber yield. Potato clones from population B have been observed to have better yield in the presence of late blight (El-Bedewy et al. 2001). However, the lack of significant correlations between AUDPC and seed mass, number of seed tubers, and total number of tubers imply that disease levels may not be major contributing factors to these yield components of potato, as opposed to what was observed previously by James et al. (1972). From the present study, it can be concluded that differential resistance of potato cultivars to late-blight infection is


indicated by differences in infection rates and late-blight severity. Cultivars derived from population A show high levels of resistance to foliar infection, have moderate tuber yield, and show a significant negative correlation between AUDPC and tuber yield. Cultivars of population B have better tuber yield, higher rates of disease progress, but no correlation between AUDPC and tuber yield. The ability of the cultivars derived from population B to maintain a high yield in the presence of the late-blight pathogen suggests that they could be a highly desirable component in the integrated management of potato for resource-limited potato growers in tropical highlands of East Africa.

Acknowledgements We would like to thank the International Potato Center, regional office for sub-Saharan Africa in Nairobi, for providing financial and logistic support that enabled this study to be conducted. We greatly acknowledge the support of the New England Plant Soil and Water Laboratory, Agricultural Research Service, US Department of Agriculture, Orono, Maine.

References Browning, J.A., Simmons, M.D., and Torres, E. 1977. Managing host genes: epidemiologic and genetic concepts. In Plant disease — an advance treatise. Vol. I. Edited by J.G. Horsfall and E.B. Cowling. Academic Press, New York. pp. 191–212. Campbell, C.L., and Madden, L.V. 1990. Introduction to plant disease epidemiology. John Wiley & Sons, New York. El-Bedewy, R., Olanya, O.M., Ewell, P.T., Lung’aho, C., Ojiambo, P.S., and Karinga, J. 2001. Evaluation of potato germplasm (populations A & B) for resistance to late blight in Kenya. Afr. Crop Sci. J. 9: 215–223. Flier, W.G., Turkensteen, L.J., Van den Bosch, G.B.M., Vereijken, P.F.G., and Mulder, A. 2001. Differential interaction of Phytophthora infestans on tubers and potato cultivars with different levels of blight resistance. Plant Pathol. (London), 50: 292–301. Grinberg, M., Kadish, D., and Cohen, Y. 1995. Infectivity of metalaxyl-sensitive and resistant isolates of Phytophthora infestans to whole potato tubers as affected by tuber aging and storage. Phytoparastica, 23: 165–175. Haverkort, A.J. 1990. Ecology of potato cropping systems in relation to latitude and altitude. Agric. Syst. 32: 251–272. Jaetzold, R., and Schmidt, H. 1982. Farm management handbook of Kenya. Kenya and German Agricultural Team of the German Technical Co-operation, Ministry of Agriculture, Rossdolf, Germany. James, C. 1971. A manual of assessment keys for plant diseases. Can. Dep. Agric. Publ. No. 1458. James, W.C., Shih, C.S., Hodgson, W.A., and Callbeck, L.C. 1972. The quantitative relationship between late blight of potato and loss in tuber yield. Phytopathology, 62: 92–96. Kankwatsa, P., Adipala, E., Hakiza, J.J., Olanya, O.M., and Kidamemariam, H.M. 2002. Effect of integrating planting time, fungicide application, and host resistance on potato late blight development in southwestern Uganda. J. Phytopathol. (Berlin), 150: 248–257. Lacey, J. 1967. Susceptibility of potato tubers to infection by Phytophthora infestans. Ann. Appl. Biol. 59: 257–264.

93

Lambert, D.H., and Currier, A.I. 1997. Differences in tuber rot development for North American clones of Phytophthora infestans. Am. Potato J. 74: 39–43. Landeo, J.A., and Turkensteen, L. 1989. Assessment of partial resistance to late blight (Phytophthora infestans) of major genes in potato. Am. Potato J. 66: 530. Landeo, J.A., Gatelo, M., Forbes, G., Zapata, J.L., and Flores, F.J. 1995. Breeding for horizontal resistance to late blight in potato free of R-genes. In Phytophthora infestans 150. Edited by L.J. Dowley, E. Bannon, L.R. Cooke, T. Keane, and E. O’Sullivan. European Association for Potato Research, Pathology Section Conference, September 1995, Trinity College, Dublin, Ireland. Boole Press, Dublin, Ireland. pp. 268–274. Landeo, J.A., Gastelo, M., Forbes, G., Zapata, J.L., and Flores, F.J. 1997. Developing horizontal resistance to late blight in potato. In International Potato Center (CIP), Program Report 1995–1996. CIP, Lima, Peru. pp. 122–131. Landeo, J.A., Gastelo, M., Roncal, E., and Mendoza, A. 2000. Phenotypic stability for horizontal resistance to potato late blight in population B. Am. J. Potato Res. 77: 406. Landeo, J.A., Gastelo, M., Beltran, G., and Diaz, L. 2001. Quantifying genetic variance for horizontal resistance to late blight in potato breeding population B3C1. In International Potato Center (CIP), Program Report 1999–2000. CIP, Lima, Peru. pp. 63–67. Lapwood, D.H. 1977. Factors affecting the field infection of potato tubers of different cultivars by blight (Phytophthora infestans). Ann. Appl. Biol. 85: 23–42. Mukalazi, J., Adipala, E., Sengooba, T., Hakiza, J.J., Olanya, O.M., and Kidanemariam, H.M. 2001. Metalaxyl resistance, mating type and pathogenicity of Phytophthora infestans in Uganda. Crop Prot. 20: 379–388. Neher, D.A., and Campbell, C.L. 1997. Analysis of disease progress curves using nonlinear regression. In Exercises in plant disease epidemiology. Edited by L.J. Francl and D.A. Neher. American Phytopathological Society (APS) Press, St. Paul, Minn. pp. 38–41. Nyankanga, R.O., Wien, H.C., Olanya, O.M., and Ojiambo, P.S. 2004. Farmers’ cultural practices and management of potato late blight in Kenyan highlands. Implications for disease management. Int. J. Pest Manag. 50: 135–144. Ochwo, M.K.N., Kamoun, S., Adipala, E., Rubaihayo, P.R., Lamour, K., and Olanya, O.M. 2002. Genetic diversity of Phytophthora infestans (Mont.) de Bary in the eastern and western highlands of Uganda. J. Phytopathol. (Berlin), 150: 541–542. Ojiambo, P.S., Nyanapah, J.O., Lung’aho, C., Karinga, J.K., and Kidanemariam, H.M. 2000. Comparing different epidemiological models in field evaluation of selected genotypes from Solanum tuberosum CIP population A for resistance to Phytophthora infestans (Mont.) de Bary in Kenya. Euphytica, 111: 211–219. Olanya, O.M., El-Bedewy, R., Ojiambo, P.S., Ewell, P.T., and Hakiza, J.J. 2001. Relationships of fungicide application to late blight development and potato growth parameters in the tropical highlands of Uganda and Kenya. In Scientist and farmer: partners in research for the 21st century. International Potato Center (CIP), Program Report 1999–2000. CIP, Lima, Peru. pp. 77–86. Reynolds, K.L., and Neher, D.A. 1997. Statistical comparison of epidemics. In Exercises in plant disease epidemiology. Edited by L.J. Francl and D.A. Neher. American Phytopathological Society (APS) Press, St. Paul, Minn. pp. 34–37. SAS Institute Inc. 2005. SAS®/STAT user’s guide. Release 8.2. SAS Institute Inc., Cary, N.C.

94 Stewart, H.E., Bradshaw, J.E., and Wastie, R.L. 1994. Correlation between resistance to late blight in foliage and tubers in potato clones from parents of contrasting resistance. Potato Res. 37: 429–434. Trognitz, B.R., Chacon, G., Pinedo, H., and Eslava, M. 1995. Screening for R-genes causing race-specific resistance to late blight in wild potato species. Am. Potato J. 72: 662. Trognitz, B.R., Orrillo, M., Portal, L., Román, C., Ramón, P., Perez, S., and Chacón, G. 2001. Evaluation and analysis of reduction of late blight disease in a diploid potato progeny. Plant Pathol. (London), 50: 281–291.

Can. J. Plant Pathol. Vol. 28, 2006 Umaerus, V., and Umaerus, M. 1994. Inheritance of resistance to late blight. Part 5. Inheritance of resistance to pests and diseases. In Potato genetics. Edited by J.E. Bradshaw and G.R. Mackay. Commonwealth Agricultural Bureaux (CAB) International, Wallingford, UK. pp. 365–402. Vega-Sanchez, M.E., Erselius, L.J., Rodiguez, A.M., Bastidas, O., Hohl, H.R., Ojiambo, P.S., Mukalazi, J., Vermuelen, T., Fry, W.E., and Forbes, G.A. 2000. Host adaptation to potato and tomato within the US-1 clonal lineage of Phytophthora infestans in Uganda and Kenya. Plant Pathol. 49: 531–539.