African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 6, Num. 2, 1998, pp. 137-142

C. LUNG'AHO, P.S. OJIAMBO¹ and H.M. KIDANEMARIAM¹

National Potato Research Centre, P.O. Box 338, Limuru, Kenya
¹International Potato Centre, SSA, P.O. Box 25171, Nairobi, Kenya

(Received 20 November, 1997; accepted 6 March, 1998)

Code Number:CS98015
Sizes of Files:
      Text: 37K
      Graphics: No associated graphics files

ABSTRACT

Eleven promising potato clones possessing late blight tolerance were evaluated for tuber yield performance, stability and adaptation across ten environments of medium and high potential potato growing regions of Kenya. Results of combined analysis of variance for tuber yield showed significant effects of genotypes, environments, and genotype by environment interaction. The mean tuber yield for individual clones ranged from 22.80 to 37.63 t ha^-1. When the genotypic tuber yields were subjected to stability analysis against an environmental index, the regression coefficients for individual clones ranged between 0.427 and 1.687. Among the clones tested, KP90188.3 was the most stable genotype across all the ten environments, while four of the highest yielding clones; 387792.5, 378699.2, 381381.20 and 381381.13 appeared to be specifically adapted to favourable growing conditions. Two clones, 381381.13 and 381381.20 exhibited mean superior performance in yield across all environments, and were pre-released in 1996 pending multiplication of sufficient quantities of seed.

Key Words: Multilocation trials, potato clones, Solanum tuberosum

Onze clones prometteurs de pomme de terre ayant une tolérance au mildiou ont été évalués pour performance en rendement des tubercules, la stabilité et l'adaptation dans dix environnements à potentiel moyen et élevé pour la production de la pomme de terre au kenya. Les résultats de l'analyse combinée de la variance du rendement en tubercules ont mis en évidence des effets significatifs des génotypes, des environnements, et de l'interaction entre génotype et environnement. Le rendement moyen en tubercules pour les clones individuels varie de 22, 8 à 37, 6 t ha^-1. Lorsque les rendements génotypiques de tubercules étaient soumis à l'analyse de la stabilité contre un indice environnemental, les coefficients de régression pour les clones individuels variaient de 0, 427 à 1,687. Parmi les clones testés, KP 90188.3 était le génotype le plus stable dans tous les environnements, alors que quatre des clones 387792.5, 378699.2, 381381.20 et 381381.13 semblaient spécialement adaptés aux conditions favorables de croissance. Deux clones, 381381.13 et 381381.20, ont montré une performance moyenne supérieure à travers tous les environnements. Ils étaient pré-diffusés en 1996 en attendant la multiplication des quantités suffisantes de semences.

Mots Clés: Essais multilocaux, clones de pomme de terre, Solanum tuberosum

Developing crop cultivars that perform well across a wide range of environmental conditions has long been a major challenge to plant breeders. In practice, genotype by environment interaction complicates the identification of superior genotypes (Allard and Bradshaw, 1964). Although it is important to detect such interactions by conducting yield trials over a series of environments, this alone gives no measure of the stability of individual genotypes. Hence, stability measurements are important since they give an indication of the adaptability of genotypes to general or specific conditions (Getinet, 1988).

Potatoes (Solanum tuberosum L.) are an important food crop in Kenya. Potato late blight, a disease of potato foliage and tubers is caused by Phytophthora infestans (Mont.) de Bary and is considered the most critical disease that limits the expansion of potato production in Kenya. The loss in tuber yield of potatoes infected by the disease varies with the host cultivar and growing environment and total destruction of the crop has been reported (Fry, 1978). All the fourteen varieties that were released prior to 1996 and the >60 informally released varieties are susceptible to late blight (Maingi et al., 1991; Kinyae et al., 1994). Although fungicides are available to control the disease, they are expensive, environmentally unfriendly and pose health hazards for small holders who grow most of the crop (Schuster and Schroeder, 1990; Pilings and Jepson, 1993). Therefore, development of plant host resistance as the foundation of integrated late blight management is considered an appropriate option in keeping losses due to the disease at a minimum.

The Kenyan potato programme depends on multilocational trials referred to as National Performance Trials (NPT), to evaluate promising genotypes in varying potato growing environments before best performing genotypes can be released as varieties. Information on the adaptation and stability of the genotypes over seasons and also over sites is useful as it is the basis for recommending the type of varieties that should be grown under particular production environments as well as the yield expectations of the test clones.

The identification of superior genotypes is, however, complicated by genotype by environment interactions (Allard and Bradshaw, 1964). A method to estimate adaptation of cultivars to different environments based on analysis of variance and regression coefficients was developed by Finlay and Wilkson (1963). They interpreted regression coefficients approximating 1 as being indicative of average stability and, when associated with high mean yield, genotypes are considered to have acceptable adaptation and stability. Eberhart and Russell (1966) used the same method of regression analysis to classify the yield stability of genotypes, but they defined other parameters of stability thus, a genotype with desirable adaptation and stability was defined as one having a high mean yield, a regression coefficient equal to unity, and a deviation from regression equal to zero (s2d=0). In alternative approach to stability analysis, Lin et al. (1986), defined complete similarity in three different ways; i) equality of genotype's response across locations, ii) equality of all within location differences, and iii) equality of all within location ratios.

For quantitative traits such as tuber yield, the relative contribution of genotype by environment interaction influences the rate of progress in selection. Hence, under Kenyan conditions where potatoes are grown across a varied range of agroecological zones, identification of suitable genotypes adapted both to specific and general environments is necessary so that appropriate recommendations regarding growing environments can be made (Assefa et al.,1995) and is indeed an important consideration in cultivar release.

The objective of the present study was therefore to determine the yield performance and stability of eleven potato clones across a range of environments and to recommend the best performing clones that are adapted to the major potato growing areas in Kenya.

A yield trial, consisting of eleven promising clones was conducted at five different locations: Tigoni, Embu, Njoro, Njabini, and Kisii starting from 1994 to 1996 as part of the national performance trials. Each location in a given season was considered an individual environment. Sites were grouped according to annual rainfall and production regions. The locations differed in altitude, soil characteristics and other climatic factors, representing a range of growing conditions (Table 1). The source of these clones was the International Potato Centre (CIP). The clones had previously been evaluated for a number of years in replicated and unreplicated trials during on-station trials at National Potato Research Centre, Tigoni, Kenya and selected for their late blight tolerance and high yields before being tested in multilocational trials. Tolerant clones had been defined as those having a maximum score of 5 on a scale of 1-9 where 1 = no foliar late blight symptoms and 9 = all foliage wiped out by late blight. Each experiment at each site was a randomised complete block design with three replications. The eleven potato clones were randomised independently at each experimental site. All seeds for the clones were produced at one site (Tigoni) and sent to each collaborator at the same time. Detailed instructions for conducting the experiment including crop husbandry and data collection were provided to each collaborator. Planting dates and seasons were selected by the individual collaborators based on environmental conditions at each site. Plot sizes consisted of three rows of ten tubers each. Sprouted tubers were planted at a spacing of 75 x 30cm giving a plant density of 44,444 plants per hectare. The crops were grown under natural late blight epidemic with no fungicides applied to control the disease and the recommended normal agronomic practices for potato production were followed at each experimental site.

TABLE 1. Altitude, rainfall and soil characteristics of the experimental sites for national performance trials during the short rains 1994 and long rains 1996

Notes:
pH - using the KCI method
Soil Classification - FAO/UNESCO (1974) EC - Electro-conductivity
P - Precipitation
E0 - Potential evaporation a - Long rains season b - Short rains season

To determine the yield for each potato clone, tubers were harvested from all the hills in each plot to give yield per plot which was converted to tons ha-1 for subsequent data analysis. Analysis of yield stability of the potato clones followed four analytical procedures: 1) an F-test for significance for the tuber yield data combined over environments and clones according to Allard and Bradshaw (1964) and Hailu (1988); 2) all clone means for each trial were subjected to a stability test against an environmental index following the procedures described by Eberhart and Russell (1966) and Finlay and Wilkinson (1963); 3) the regression coefficients obtained were tested for significance using the Student's t-test; and 4) from the same analysis, deviations from regression and coefficients of determination were calculated.

The performance of the eleven potato genotypes was evaluated over ten environments which differed in altitude, soil type, and seasonal climatic factors (Table 1). Environmental mean yields ranged from 13.44 to 37.91 tons ha^-1. The combined analysis of variance for tuber yield exhibited a significant effect of the clone and environment components (Table 2), indicating a significant variation in the genetic yield potential of the potato germplasm evaluated, and a significant effect of seasons by locations. The clone by environment component was significant suggesting that the performance ranking of the clones was not constant.

TABLE 2. Combined analysis of variance for tuber yield (ton ha -1) of eleven potato clones evaluation across ten environments in the highlands of Kenya, 1994-1996

TABLE 3. Mean tuber yields and estimates of stability for tuber yield of eleven potato clones grown in ten environments in mid and high altitude regions of Kenya, 1994-1996

The highest yielding clone, 381381.20 had a yield of 37.63 ton ha^-1 while the lowest yielding clone, 382193.9, had a mean yield of 22.80 ton ha^-1 (Table 3). Because the genotype by environment component was significant (P=0.05), the individual clones were subjected to stability analysis following the model proposed by Eberhart and Russel (1966). The basic assumption for this analysis is that genotypic performance is positively and linearly related to favourable environmental conditions when other production factors are controlled.

Estimates of three stability parameters are shown in Table 3 and six representative regression equations and their corresponding coefficients of determination (R2) presented in Table 4. The slope for 382193.9 was significantly lower than 1.0 (P<0.05), while that of 381381.20 was significantly higher than 1.0. The slopes of the rest of the nine clones were not significantly different from unity. Of the eleven clones evaluated, six genotypes (387792.5, KP90188.3, 676070, 378699.2, 381381.13 and 381381.20) exhibited tuber yields above the grand mean, but only five clones had regression coefficients greater than or equal to unity. Thus, these five clones appeared to be superior in performance and were responsive to favourable growing conditions. Clone KP90188.3 which had a regression coefficient close to unity (b=0.964), a relatively minimal deviation from regression, and an intermediate yield, appeared to be stable across a wide range of environmental conditions. The lowest yielders, 382193.9, 382193.10 and 387419.3 had slopes below unity (b<1), exhibiting a poor tuber yield response to favourable environments. The three clones yielded below the mean across all environments (Table 3).

TABLE 4. Linear regression equations of yield (Y) against environmental index (x) of eleven potato clones tested in ten environments in mid and high altitude regions in Kenya

The values for the coefficients of determination (R²) (Table 4) suggested that 29.0 to 88.8% of the variation in tuber yield of the individual clones was accounted for by regression on the environmental index. Thus, regression analysis had considerable predictive value in estimating the stability of the test clones. Similar results were obtained by Assefa et al. (1995) and Sarian et al. (1990) working with wheat genotypes.

In accordance with the criteria set for stability analysis (Finlay and Wilkinson, 1963; Eberhart and Russell,1966) clone KP90188.3 was the most stable genotype across the diverse growing regions represented by the ten sites, having a regression coefficient close to unity, and a relatively minimal deviation from regression (Table 2). Although this clone exhibited general stability and acceptable yield levels, it cannot be recommended for release at present since it was highly susceptible to viral infections resulting in rapid degeneration. However, if and when potato growers begin to renovate their seed stocks on an annual basis then the clone will be recommended for production across diverse growing environments in potato growing regions of Kenya.

In environments with favourable potato growing conditions clones 387792.5, 378699.2, 381381.20 and 381381.13 appeared to be highly productive. Clones 381381.20 and 381381.13 significantly outyielded the other nine clones and were ranked first and second respectively across all the environments included in this study. Clones 381381.20 and 381381.13 were pre-released in 1996 and are set for full release in 1997 for commercial production. Although clone 382193.10 had a relatively low yield, it is an early maturing clone with the characteristic of tuberising well under low moisture conditions. It is proposed for release in the medium and lower medium production environments.

Since the clones varied in response to diverse growing conditions of Kenya, future recommendations for cultivar release should take into account these findings. Stability analysis of tuber yield data collected across locations and seasons appears to be of value for recommending cultivars for specific or general environments provided representative sites are chosen for the study.

The authors wish to thank staff of co-operating stations for assistance in executing the study. The funds for this research were provided by the Agricultural Research Fund programme, the Regional Potato and Sweet Potato Improvement Programme for Eastern and Central Africa (PRAPACE) and the International Potato Centre (CIP).

Allard, R.W. and Bradshaw, A.D. 1964. Implications of genotype-environment interactions. Crop Science 4:503-507.

Assefa, S., Gelete, B. and Tanner, D.G. 1995. Yield stability analysis of nine spring Bread (Wheat) genotypes in the Central Highlands of Ethiopia. Africa Crop Science Journal 3:35-40.

Eberhart, S.A. and Russell, W.A. 1966. Stability parameters for comparing varieties. Crop Science 6:36-40.

Finaly, K.W. and Wilkinson, G.N. 1963. The analysis of adaptation in a plant breeding program. Australian Journal of Agricultural Research 14:742-754.

Fry, W.E. 1978. Quantification of general resistance of potato cultivars and fungicide effects for integrated control of potato late blight. Phytopathology 68:793-800.

Getinet, G. 1988. Grain yield stability of bread wheat cultivars in the highlands of Ethiopia. In: Fifth Regional Wheat Workshop for Eastern, Central, and Southern Africa and the Indian Ocean. Van Ginkel, M. and Tanner, D.G. (Eds.), pp. 61-65. CIMMYT, Mexico D.F.

James, C. W. 1981. Estimated losses of crops from plant pathogens. In: Handbook of Pest Management in Agriculture. Pimentel, D. (Ed.), pp. 80-94.

Kinyae, P.M., Lung'aho, C., Kanguha, E. and Njenga, D.N. 1993. The status of seed potatoes in Meru, Nyambene, Nyandarua and Laikipia districts. In: Second KARI/CIP Technical Workshop on Collaborative Research. Nairobi, Kenya. pp. 5-9.

Lin, C.S., Binns, M.R. and Lefkovitch, C.L. 1986. Stability analysis: Where do we stand? Crop Science 26:894-900.

Maingi, D.M., Ng'anga, N. and Kidanemariam, H.M. 1991. Collection and evaluation of farmer's informally released varieties in Kenya. In: KARI/CIP Technical Workshop Collaborative Research. Nairobi, Kenya. pp. 13-18.

Pilings, E.D. and Jepson, D.C. 1993. Synergism between EBI fungicides and a pyrethroid insecticides in the honey bee (Apis mellifern). Pesticide Science 39:293-297.

Sariah, M.A., Ndondi, R.V. and Mollel, M.J. 1990. Grain yield potential and adaptation of ten bread wheat varieties in Tanzania. In: Sixth Regional Wheat Workshop for Eastern, Central and Sourthen Africa. Van Ginkel, Tanner, D.G. and Mwangi, W. (Eds.), pp. 224-228. CIMMYT, Mexico, D.F.

Schuster, E. and Schroeder, D. 1990. Side effects of sequentially and simultaneously applied pesticides on non-target soil microorganisms in laboratory experiments. Soil Biology and Biochemistry 22:375-384.

Copyright 1998, African Crop Science Society