Eth.J.Sci & Technol.

5(2):90-98, 2008 ISSN 1816-3378

©Bahir Dar University, April 2008

Performance Stability Analysis of Potato Varieties under Rainfed and

Irrigated Potato Production Systems in Northwestern Ethiopia

Yigzaw Dessalegn1*, Fentahun Mengistu1 and Tesfaye Abebe2

1
Amhara Region Agricultural Research Institute, PO Box 527, Tel. 058-220-6400, Bahir Dar
2
Adet Agricultural Research Center, PO Box 8, Tel. 058-338-0237, Adet
*
Corresponding author: [email protected]

Abstract for small scale farmers, occupies the

In Ethiopia, farmers cultivate potato both in the largest area compared to other vegetable
rainy season under rainfed condition and the dry crops and produces more food per unit area
season using irrigation. Despite the variation in
climatic condition and production constraints
and time compared to cereal crops. In 2001
between the two systems, farmers grow the same E.C., 0.94 million tons of potato tuber was
variety both in rainfed and irrigated potato produced nationally from 164 thousand ha
production systems. Moreover, researchers release of land (CSA, 2003). The Amhara
improved varieties only based on performance National Regional State contributed 36.1%
under rainfed production system. Therefore, the
objective of this study was to assess the stability of
of the annual national potato production.
improved potato varieties across the two potato However, the regional average potato tuber
production systems. Nine potato varieties were yield (4.8 t/ha) is less than the national
evaluated using randomized complete block design average yield (5.7 t/ha) (CSA, 2003).
with three replications under irrigation and rainfed Factors such as late blight (Phytophtora
conditions for two consecutive years at Adet
Agricultural Research Center, northwestern
infestans) and bacterial wilt (Pseudomonas
Ethiopia. Results revealed that improved varieties solanacearum) infections, poor crop
performed better in the rainfed production system management and shortage of adaptable and
than the irrigated production system. In addition, high yielding varieties contributed to the
there was no stable variety in both production low productivity of potato in the region
systems. Zengena, CIP384321.3, Genet, Wochecha,
Guasa, Tolcha and Menagesha had regression
(Tesfaye and Yigzaw, 2008).
coefficients higher than unity indicating that these
varieties prefer favorable condition (rainfed potato In Ethiopia potato is produced in the rainy
production system). The local variety and Awash season under rainfed condition and dry
had regression coefficient less than unity indicating season using irrigation. In 2002, the
their response to the unfavorable condition
(irrigated potato production system). This depicts
irrigated potato production system
the relevance of fostering independent variety contributed 58.7% of the annual potato
selection program for each potato production tuber produced and 76.8% of the total area
system. However, currently varieties such as of land planted with potato in the country
Zengena and Guasa can be recommended for both (CSA, 2003). Likewise, in the Amhara
production systems.
National Regional State, irrigated potato
Key words: Potato; variety; yield; rainfed; production system contributed 84.2% of
irrigation; stability; Ethiopia the area and 65.5% of the annual potato
production. Although irrigated potato
production system contributed the lion’s
1. Introduction
share both in the country and the region, its
Potato (Solanum tuberosum L.) is one of
productivity (3.7 t/ha) is lower than the
the most important vegetable crops grown
rainfed (10.5 t/ha) system (CSA, 2003).
in the high and mid altitude areas of
This could be due to differences in climatic
Ethiopia. It serves as food and cash crop

90
Yigzaw Dessalegn et al., 2008

conditions and production constraints of select and release improved potato

the two production systems. Furthermore, varieties solely based on performance
researchers have never released improved under rainfed condition.
varieties for the irrigated potato production
system. Yield is a complex trait in potato and is
generally considered to have low
The prevailing average monthly maximum heritability (Lynch and Kozub, 1991).
temperature is higher in the irrigated potato Hence, indirect selection could be useful
production system than in the rainfed strategy to bring considerable genetic
system. The average monthly minimum improvement on potato tuber yield.
temperature is low and causes frost injury Therefore, knowledge on the
to the plant during the irrigated potato interrelationships of characters with tuber
production system. Therefore, irrigated yield seems very important for high
potato production is affected both by the yielding potato variety development. Thus,
prevailing higher maximum and lower this experiment was conducted to assess
minimum temperature compared to the the stability of potato varieties and to
rainfed potato production system (Fig.1). evaluate the association of different
On the other hand, in contrast to the characters with tuber yield under the two
irrigated system, the rainfed potato potato production systems.
production system is more affected by late
blight. 2. Materials and Methods
The performance of nine potato varieties
A temperature of about 20oC is ideal for was assessed in this study. The experiment
potato tuber development (Mondal and was conducted at Adet Agricultural
Chatterjee, 1993). At higher temperatures Research Center using a Randomized
the plant fails to initiate tuber formation Complete Block Design with three
and at low temperatures vegetative growth replications both under rainfed and
is restricted by frost (Horton, 1987). irrigated potato production systems for two
Similarly, Mondal and Chatterjee (1993) consecutive years (2001 & 2002). Plot size
reported that tuber initiation starts early at was 9 m2 (3 m × 3 m) with a spacing of 75
lower than at higher temperatures. Besides, cm between rows and 30 cm between
the number of tubers produced per plant is plants. The trial was planted in the first
higher at lower than at higher temperature. week of June for the rainfed production
The seed tubers produced at higher system and in the second week of
temperatures (34oC) are low yielding November for the irrigated production
compared to seed tubers produced at cooler system. In both systems nitrogen and P2O5
temperatures (7.7oC) (Mondal and were applied at rates of 81 and 69 kg/ha,
Chatterjee, 1993). Therefore, temperature respectively. The whole P2O5 was applied
affects not only the yield of treated plants at planting. The nitrogen split equally into
but also the productivity of its progeny. three and was applied at planting, at the
first ridging and at flowering. All other
Bhagsari et al. (1988) reported significant crop management practices were applied
variations among potato genotypes in tuber as recommended.
setting and harvest index in different
subtropical environments. Despite Number of main stems per plant, plant
significant climatic variation between the height, number of tubers harvested per
two systems, potato breeders in Ethiopia

91
Eth.J.Sci & Technol. 5(2):90-98, 2008 ISSN 1816-3378
©Bahir Dar University, April 2008

plot, marketable tuber yield, unmarketable environments. Elobu and Osiru (1994) also
tuber yield and total tuber yield were observed significant marketable tuber yield
recorded and subjected to analysis of difference among three heat tolerant potato
variance using MSTAT-C statistical genotypes under a hot lowland tropical
software. Genotype x production system condition of Uganda.
and genotype x year interaction were
assessed using combined analysis of In the irrigated potato production system,
variance. Duncan Multiple Range Test the maximum temperature was very high
(DMRT) was employed to separate means. and the minimum temperature was very
Stability analysis was performed following low. High yielding genotypes had
Eberhart and Russell (1966) method using maximum number of main stems per plant,
the model Yij = μI + βiIj +δij. Regression high number of tubers per plant and long
coefficient (bi) was considered as a plant height. Therefore, genotypes
parameter of response and deviation from difference in marketable tuber yield under
regression (Sdi2) as a parameter of stability irrigated potato production system could
(Singh and Chaudhary, 1977). Simple be ascribed to their difference in capacity
linear correlation was computed to to initiate tuber at higher temperature and
evaluate the association of different traits frost tolerance. Likewise, Thornton et al.
at different potato production systems. (1996) reported that high temperature
retarded vegetative and tuber growth rate
3. Results and Discussion in potato. Similarly, Horton (1987)
In the irrigated potato production system, reported that at high temperatures, the
genotypes were significantly different in potato plant fails to initiate tuber and at
number of main stems per plant, plant low temperature its vegetative growth is
height, number of tubers harvested per retarded by frost. In the rainfed potato
plot, marketable tuber yield, unmarketable production system, genotypes significantly
tuber yield and total tuber yield (Table 1). varied in plant height, number of tubers
Zengena was the tallest in plant height harvested per plot, marketable tuber yield,
among evaluated genotypes. It also had unmarketable tuber yield and total tuber
maximum number of main stems per plant yield (Table 1). Unlike in the irrigated
and maximum number of tubers harvested potato production system, genotypes were
per plot among evaluated genotypes. not statistically different in number of
Marketable tuber yield ranged from 6.7 main stems per plant. Marketable tuber
t/ha to 22.2 t/ha. The lowest and the yield ranged from 1.44 t/ha to 33.1 t/ha.
highest marketable tuber yield were The local variety was the shortest and had
recorded from Wochecha and the lowest marketable tuber yield
CIP384321.3, respectively. Zengena and compared to all evaluated varieties. The
CIP384321.3 were not statistically low productivity of the local variety is
different in marketable tuber yield. ascribed to its susceptibility to late blight.
Therefore, Zengena and CIP384321.3 On the other hand, Zengena was the tallest
seem promising varieties for irrigated variety and ranked first in marketable tuber
potato production system at Adet. yield, but it was not significantly different
Similarly, Bhagsari et al. (1988) reported from CIP384321.3 and Guasa in
significant variation among potato marketable tuber yield.
genotypes in tuber setting and harvest Therefore, Zengena, CIP384321.3 and
index while grown in different subtropical Guasa appeared to be promising varieties

92
Yigzaw Dessalegn et al., 2008

for the rainfed potato production system at Adet.

.
Table 1. Mean tuber yield and other agronomic characters of nine potato varieties under
irrigated and rainfed potato production systems at Adet, northwestern Ethiopia, in 2001
and 2002.

Variety Main Plant Tubers Marketable Unmarketable Total

stems height harvested tuber yield tuber yield tuber
per (cm) per plot (t/ha) (t/ha) yield
plant (t/ha)

Irrigated
Tolcha 3.9c 30.6d 97.0bc 9.66def 5.99abc 15.65cd
Awash 3.7c 35.3cd 141.3ab 12.90cde 7.50ab 19.50bcd
Genet 4.0bc 34.4cd 163.3a 14.56c 7.52ab 22.12abc
Guasa 5.8a 42.4bc 181.3a 16.02bc 8.90a 24.93ab
Local 4.6abc 34.2cd 153.8a 13.62cd 2.53c 16.15cd
Wochecha 4.2bc 32.0d 78.2c 6.72f 5.97abc 12.69d
Menagesha 4.9abc 46.7b 74.0c 8.25ef 5.72abc 13.97d
CIP384321.3 5.6ab 51.5ab 179.7a 22.19a 5.54abc 27.73a
Zengena 6.1a 59.6a 193.7a 19.65ab 4.19bc 23.85ab

Mean 4.8 40.7 140.3 13.63 5.99 19.62

CV (%) 19.9 14.5 22.2 19.5 47.2 22.4

Rainfed
Tolcha 4.8 45.8def 160.0abc 20.90bcd 4.95b 25.85cd
Awash 4.7 42.7ef 130.8c 14.25d 2.95bc 17.20e
Genet 5.2 53.3cd 207.3a 24.62b 2.66bc 27.27cd
Guasa 4.7 56.2bcd 183.0ab 26.88ab 5.57b 32.44bc
Local 5.0 38.2f 111.5c 1.44e 0.55c 1.99f
Wochecha 4.8 51.5cde 153.0bc 21.80bcd 2.39bc 24.19d
Menagesha 3.9 57.3bc 112.0c 17.19cd 3.55bc 20.73de
CIP384321.3 3.3 65.3ab 182.0ab 27.68ab 13.31a 40.99a
Zengena 4.2 74.8a 201.2ab 33.10a 2.11bc 35.20ab

Mean 4.5 53.9 160.1 20.87 4.23 25.09

CV (%) 23.2 12.4 19.5 21.6 65.4 16.1
Means followed by the same letter (s) within a column are not statistically different; each
value in the table is mean of 6 observations; * and ** stand for significance at p≤0.05
and p≤0.01, respectively; NS stands for not significant at p≤0.05

According to across systems and years yield and total tuber yield (Table 2).
average performance, varieties were Zengena and the local variety gave the
significantly different in plant height, highest and lowest marketable tuber yield,
number of tubers harvested per plot, respectively. In addition, results of
marketable tuber yield, unmarketable tuber combined analysis of variance showed that

93
Eth.J.Sci & Technol. 5(2):90-98, 2008 ISSN 1816-3378
©Bahir Dar University, April 2008

marketable tuber yield was significantly rate (Thornton et al., 1996). The average
different between production systems monthly minimum temperature was also
(Table 2). The rainfed potato production lower in the irrigated than in the rainfed
system gave the highest average potato production system (Fig. 1). Low
marketable tuber yield than the irrigated temperature retards vegetative growth and
system (Table 2). This could be due to the subsequently the productivity of the crop.
prevailing average monthly maximum Therefore, low productivity of potato in
temperature, which was higher in irrigated irrigated potato production system could
potato production system than in the also be attributed to the prevailing very
rainfed potato production system (Fig. 1). low minimum temperature. Combined
In agreement to this result, Mondal and analysis of variance showed that variety ×
Chatterjee (1993) reported lower potato production system, variety × year and
tuber yield in warm condition than in cool variety × system × year interaction effects
condition. In addition, tuber initiation were statistically significant (Table 3).
delays at higher than lower temperature That was an indication for different
(Mondal and Chatterjee, 1993). Moreover, varietal response to the changes in growing
higher temperature reduces tuber growth conditions or environments.

Table 2. Across years and systems mean agronomic and tuber yield performance of nine
potato varieties at Adet (2001 & 2002)

Variety Main Plant Tubers Marketable Unmarketable Total

stems height harvested tuber yield tuber yield tuber
per (cm) per plot (t/ha) (t/ha) yield
plant (t/ha)
Tolcha 4.4 38.2ef 128.5b 15.28d 5.47bc 20.75de
Awash 4.2 39.0ef 136.1b 13.13d 5.23bc 18.35e
Genet 4.6 43.9de 185.3a 19.59c 5.11bc 24.69cd
Guasa 5.3 49.3cd 182.2a 21.45bc 7.23ab 28.68bc
Local 4.8 36.2f 132.7b 7.53e 1.54d 9.07f
Wochecha 4.5 41.7ef 115.6bc 14.26d 4.18cd 18.44e
Menagesha 4.4 52.0bc 93.0c 12.72d 4.63bc 17.35e
CIP384321.3 4.5 58.4b 180.8a 24.93ab 9.43a 34.36a
Zengena 5.1 67.2a 197.4a 26.38a 3.15cd 29.53b

Mean 4.6 47.3 150.2 17.25 5.11 22.36

CV (%) 21.5 13.3 20.8 21.4 54.7 18.8
Means followed by the same letter(s) within a column are not statistically different
Each value in the table is mean of 12 observations; * and ** stand for significance at
p≤0.05 and p≤0.01, respectively; NS stands for non-significant at p≤0.05

Therefore, stability analysis was performed coefficient and minimum deviation from
to identify a relatively stable variety regression (S2di). Accordingly, the local
among evaluated genotypes.A variety is variety was the most unstable variety.
said to be stable and adaptive if it has high Zengena, CIP384321.3, Genet, Wochecha,
mean yield, about unity regression Guasa, Tolcha and Menagesha had

94
 Yigzaw Dessalegn et al., 2008

regression coefficients higher than unity production in both systems. Similarly,

indicating that these varieties prefer Abalo et al. (2003) reported instability in
favorable condition (Table 4). The local marketable tuber yield among potato
variety and Awash had regression genotypes grown at different localities of
coefficients less than unity indicating their Uganda. Hence, independent variety
response to the unfavorable condition. selection program is suggested for each
Therefore, there is no stable variety among potato production system.
evaluated genotypes to recommend for

35 a) Irrigated Average 35 b) Rainfed

30 minimum 30
Temperature (oC)

temperature 25
25
(oC)
20 20
Average
15 15
maximum
10 temperature 10
(oC) 5
5
0
0
1 2 3 4 5
1 2 3 4 5
Months
Months
Fig. 1. Average monthly minimum and maximum temperatures during the a) irrigated and b)
rainfed potato production systems at Adet.

Table 3. Combined Analysis of variance (ANOVA) on stem number, plant height, days to
maturity, number of tubers harvested, marketable, total and unmarketable tuber yields

Source of No. of Plant Days to No. of Marketable Un- Total

variation main height maturity tubers/ tuber yield marketable tuber
stems/ (cm) plot (t/ha) tuber yield yield
plant (t/ha) (t/ha)
Year (Y) ** ** ** NS * ** *
Systems (s) * ** ** ** ** ** **
YxS NS ** ** ** ** ** NS
Variety (V) NS ** ** ** ** ** **
YxV NS ** ** ** ** ** NS
SxV * ** ** ** ** ** **
YSV ** ** ** ** ** ** NS
* and ** stand for significance at p≤0.05 and p≤0.01, respectively; NS stands for not
significant at p≤0.05

95
Eth.J.Sci & Technol. 5(2):90-98, 2008 ISSN 1816-3378
©Bahir Dar University, April 2008

Table 4. Stability parameters for nine potato varieties in two years and production seasons
(rainfed and irrigated) at Adet, northwestern Ethiopia, in 2001 and 2002.

Variety Regression Deviation from Coefficient of Mean marketable

coefficient (bi) regression (S2di) determination tuber yield (t/ha)
Tolcha 1.09 0.41 0.78 15.28
Awash 0.89 0.63 0.49 13.13
Genet 1.20 0.17 0.96 19.59
Guasa 1.12 0.57 0.66 21.45
Local -0.64 1.07 0.15 7.53
Wochecha 1.19 0.87 0.49 14.26
Menagesha 1.09 0.18 0.95 12.72
CIP384321.3 1.37 0.61 0.72 24.93
Zengena 1.69* 0.15 0.98 26.37
* stands for significant difference from unity at 5% probability

Knowledge of correlations among different reported a considerably high positive

characters is vital to design an effective correlation between tuber yield per plant
breeding program. Most economically and leaf area per plant at genotypic level.
important traits such as tuber yield are Therefore, these traits can be employed as
quantitatively inherited. The phenotypic indirect selection index for higher
expression of such traits is highly marketable tuber yield. Marketable tuber
influenced by genotype × environment yield had positive and statistically
interaction effect. Consequently, it is very significant association with plant height
difficult and time consuming trying to and number of tubers harvested per plot in
improve tuber yield through direct rainfed potato production system (Table
selection. Hence, it is very important to 5). This result revealed the importance of
estimate the degree of association of sink source ratio in both production
various highly heritable and easily systems. In the irrigated potato production
measurable agronomic characters with system frost retards vegetative growth
yield. Simple linear correlation analysis consequently tuber yield showed strong
was performed to assess the association of and positive association with number of
different characters under irrigated and main stems. On the other hand, in the
rainfed potato production systems. rainfed potato production system the
prevailing environment favors proliferation
In the irrigated potato production system, of main stem, subsequently showed
marketable tuber yield showed positive negative correlation with marketable tuber
and statistically significant association yield. However, plant height and number
with number of main stems per plant, plant of tubers harvested per plot showed
height and number of tubers harvested per positive correlation coefficient with
plot (Table 5). De la Morena et al. (1994) marketable tuber yield in both systems and
also indicated that variation in tuber yield can serve as indirect selection index for
among cultivars was associated with the higher marketable tuber yield in both
density of stems. In support of this production systems.
observation, Baye Berihun et al. (2005)

96
 Yigzaw Dessalegn et al., 2008

4. Conclusion which could be used as indirect selection

Varieties gave lower yield in the irrigated index for high marketable tuber yield.
system than in the rainfed, suggesting
variable responses in different systems. 5. Acknowledgements
Therefore, it is suggested to commence Authors are indebted to horticulture
independent variety development program research division of Adet Agricultural
for each system. Plant height and number Research Center technical assistants. We
of tubers harvested per plot showed are also grateful to the financial support of
positive and statistically significant the Amhara Region Agricultural Research
association with marketable tuber yield, Institute, Bahir Dar, Ethiopia.

Table 5. Simple linear correlation coefficients among different traits of potato under different
potato production systems, across systems and years at Adet

Traits Plant Tubers Marketable Unmarketable Total

height harvested tuber tuber yield tuber
per plot yield yield

Rainfed
Main stems per plant -0.64 -0.03 -0.37 -0.73* -0.54
Plant height 0.64 0.85** 0.39 0.82**
Tubers harvested per plot 0.85** 0.31 0.79**
Marketable tuber yield 0.43 0.96**
Unmarketable tuber yield 0.68*
Irrigated
Main stems per plant 0.87** 0.58 0.69* -0.09 0.61
Plant height 0.50 0.68* -0.13 0.59
Tubers harvested per plot 0.92** 0.07 0.90**
Marketable tuber yield -0.05 0.94**
Unmarketable tuber yield 0.31
Across years and seasons
Main stems per plant 0.43 0.61 0.45 -0.09 0.34
Plant height 0.54 0.82** 0.27 0.74*
Tubers harvested per plot 0.83** 0.33 0.76*
Marketable tuber yield 0.56 0.97**
Unmarketable tuber yield 0.74*
* and ** stand for significant at p≤0.05 and p≤0.01, respectively

97
Eth.J.Sci & Technol. 5(2):90-98, 2008 ISSN 1816-3378
©Bahir Dar University, April 2008

6. References the International Society for Tropical Root

Abalo G, Hakiza JJ, El Bedewy R and Adipala E Crops, Kampala, Uganda 22-28 Nov. 1992. pp
(2003). Genotype × environment interaction 181-184.
studies on yields of selected potato genotypes Horton D (1987). Potatoes: production, Marketing
in Uganda. African Crop Science Journal. 11 and Programs for Developing Countries. West
(1): 9-15. view press, Colorado. pp 243.
Baye Berihun, Ravishankar R and Harijt S Lynch DR and Kozub GC (1991). The association
(2005). Variability and associations among between potato tuber yield and the components
tuber yield and related traits in potato. of yield in irrigated and dry land environments
Ethiopian Journal of Agricultural Sciences. in the Prairies. Canadian Journal of Plant
18:103-121. Sciences. 71: 279-287.
Bhagsari AS, Webb RE, Phatak SC and Mondal SS and Chatterjee BN (1993). Potao. In:
Jororski CA (1988). Canopy photosynthesis, Bose TK, Som MG, Kabir J (eds). Vegetable
stomatal conductance and yield of Solanum crops. Naya Prokash, Calcutta, India. pp. 382 -
tuberosum grown in warm climate. American 469.
Potato Journal. 65: 393-406. Singh RK and Chaudhary BD (1977).
CSA (2003). Ethiopian Agricultural Sample Biometrical methods in quantitative genetic
Enumeration, 2001/2002. Statistical report on analysis: Stability models. Kalyani Publishers,
socio economic characteristics of the Ludhiana, India. pp. 239 – 266.
population in agricultural households, land Tesfaye Abebe and Yigzaw Dessalegn (2008).
use, area and production of crops, Part I. Review of crop improvement research
Addis Ababa, Ethiopia. achievements and future focus in parts of
De La Morena I, Guillen A and Garcia Del Moral Western Amhara Region: the case of Adet. In:
LF (1994). Yield development in potatoes as Tesfaye Abebe (ed.) proceedings of the 1st
influenced by cultivar and the timing and level Amhara Region Regional Workshop on potato
of Nitrogen fertilization. American Potato research and development achievements and
Journal. 71: 165-173. transfer experiences and future directions
Eberhart SA and Russel WA (1966). Stability Bahir Dar, Ethiopia, 20-21 Dec. 2007. pp 85-
parameters for comparing varieties. Crop 101.
Science. 6: 36-40. Thornton MK, Malik NJ and Dwellee RB
Elobu P and Osiru DSO (1994). Performance of (1996). Relationship between leaf gas
three potato (Solanum tuberosum L.) genotypes exchange characteristics and productivity of
in relation to plant population and applied potato clones grown at different temperatures.
nitrogen in a lowland area of Uganda. In the American Potato Journal. 73 (2): 63-77.
proceedings of the 5th Triennial Symposium of

98