African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 9, Num. 1, 2001, pp. 41-48

African Crop Science Journal, Vol. 9, No. 1, March 2001, pp. 41-48

Code Number: CS01033

ABSTRACT

Sweetpotato production in western Kenya is constrained by widespread use of low yielding and late maturing traditional varieties. Studies conducted at several sites in Kenya between 1994 and 1996 identified some potential varieties for production in different sweetpotato growing agro-ecologies in western Kenya. The objective of this study was to assess the performance of the selected varieties under farmer management conditions. Four sweetpotato varieties Mugande, SPK013, SPK004 and Kemb 10 were evaluated by farmers at four clusters of trial sites between 1997 and 1999. Information on yield, taste and farmer acceptance was obtained at each trial site. Varieties Mugande, SPK013 and Kemb 10 performed better than the local check 'Mar Ooko' at all sites. Mugande and Kemb 10 had a wide adaptation while SPK013 yielded well in the Upper Midland high potential areas of western Kenya.

Key Words: Adaptability, farmer acceptance, Ipomea batatas, western Kenya

RÉSUMÉ

La production de la patate douce dans l'ouest du Kenya est limitée par l'usage généralisé de variétés traditionnelles tardives et à faibles rendements. Des études menées sur la production de la patate douce entre 1994 et 1996 dans différentes localités au Kenya identifièrent quelques variétés adaptées aux différentes agro-écologies trouvées à l'Ouest du pays. L'objectif de l'étude ici rapportée était d'évaluer, dans les conditions mêmes du fermier, la performance des variétées de patate douce précédemment sélectionnées. Ainsi, entre 1997 et 1999, les variétés Mugande, SPK013, SPK004 et Kemb 10 furent-elles évaluées par les fermiers de quatre agglomérations de chaque localité d'essai. L'information sur le rendement, le goût et l'acceptabilité des variétés a été obtenue pour chaque localité d'essai. Les variétés Muganda et Kemb 10 ont été largement adaptées alors que SPK013 a donné un bon rendement dans la partie du centre surélevé de l'ouest du Kenya.

Mots Clés: Adaptabilité,acceptabilité par les fermiers, Ipomea batatas, l'ouest du Kenya

INTRODUCTION

Sweetpotato (Ipomea batatas) is an important food security crop grown in Kenya mainly by women (Carey et al., 1997) for household consumption and as a source of family cash income. It is regarded as a 'poor man's crop because of its low input requirements, ease of production and ability to produce under adverse weather and soil conditions. In Western Kenya, the food security role of sweetpotato is currently critically important due to the failure of cassava as a result of epidemic cassava mosaic disease (CMD). In nutritional terms, sweetpotato is an excellent source of vitamin A, especially in the orange-fleshed varieties. Production statistics indicate that about 730,000 tons of storage roots were produced on 75,000 hectares of land (FAO, 1998), over 75% of which is found in the densely populated areas of western Kenya.

Sweetpotato production in western Kenya is constrained by widespread use of low yielding and late maturing landrace varieties, pests and diseases (Smit, 1997). Farmers in the region grow a wide range of varieties usually in mixtures to satisfy their preferences. Involving farmers in the selection process would therefore ensure that criteria of importance to sweetpotato farmers does not get overlooked or given inadequate weighting. Kenya Agricultural Research Institute (KARI) in collaboration with the International Potato Centre (CIP) have identified in a national-wide multilocational trials between 1994 and 1996 high yielding varieties that require testing in on-farm trials in western Kenya. The objective of this study was to evaluate these varieties under farmers' management practices in different sweetpotato growing agro-ecologies of western Kenya.

MATERIAL AND METHODS

Four sweetpotato varieties Mugande, SPK013, Kemb10 and SPK004 were evaluated between 1995 and 1999 at nine clusters of trial sites in Boro, Uranga, Malava, Butula, Sabatia, Teso, Ugunja, Nambale and Buhalalire in Western Kenya. These sites represent the major sweetpotato growing agro-ecological zones(AEZs) in western Kenya. Characteristics of the sites are shown in Table 1. Farmers participating in the trials at Boro, Uranga, Malava, Butula and Sabatia were selected in public 'barazas' (meetings) organised by researchers and the Ministry of Agriculture extension staff while those in Teso, Ugunja, Nambale and Buhalalire were selected by the community based organisation (CBO) partners collaborating in the project. These include Ateso Development Trust (ADT) in Teso district, Ugunja Resource Centre (URC) in Siaya district, REFSO and Appropriate Rural Development Agricultural Project (ARDAP) in Busia district.

The varieties were evaluated in 10 to 15 farms in each area along with one or two of the farmer's preferred local varieties. Each farmer received 60 cuttings of each variety to plant and maintain according to the normal practices in the area. The farmers received a further 30 cuttings to manage and harvest entirely at their discretion. Each farmer was used as a replicate.

Data were taken on virus score two months after planting and at harvest using a scale of 1 to 5, where, 1 represent no virus symptoms and 5, heavy virus attack. Evaluation of roots was done using a scale of 1 to 5 where, 1 represented very poor roots and 5, very good roots. At harvest, CBO staff and researchers recorded yield data on marketable and non-marketable roots. Storage roots destroyed by weevils were separated and counted. Percent weevil infestation was determined by dividing the number of roots infected by weevils by the number of roots harvested from one plot and multiplying by 100.

Root samples of each variety were boiled in separate pots and evaluated by untrained taste panels who included farmers and field staff. Each panellist was asked to place a stick next to the plate containing the best tasting variety. The sticks were then counted and varieties ranked such that the variety with the highest number of sticks was ranked first. Farmers conducting the trials and other interested farmers did group assessment of variety performance. They also listed and ranked the most important selection criteria used in selecting varieties for production. Farmers also provided qualitative information on the bad and good qualities of the varieties being tested. Statistical analyses were done using the MSTAT-C (1993) statistical package.

RESULTS AND DISCUSSION

The combined analysis of variance (Table 2) indicated that mean marketable and total root yield had significant mean squares for year, site and variety. The absence of significant interaction between variety and site indicated that sites did not influence the performance of the four varieties. This was not expected since most sites had different soil types, mean temperature and annual rainfall (Table 1). Ngeve (1993) and Ndolo et al. (1995) reported significant interactions between sweetpotato varieties and environments. Mugande, SPK013 and Kemb10 produced higher mean root yield and number of roots per plant than the local check (Table 3). The number of roots is important in sweetpotato production since it is positively correlated to the root yield (Lowe and Wilson, 1974). It is also an important selection criteria for farmers practising piecemeal harvesting, as the presence of many small roots at the time of harvest may indicate that there is continued potential for production (Ndolo et al., 1995). Low root yield observed in local varieties was probably because of their late maturity characteristics and inherent low yield potential. The percentage of roots infected with sweetpotato weevils (Cylas spp.) was greater in Kemb10, SPK004 and the local variety than in Mugande and SPK013. This may be attributed to differences in rooting depths among the two varieties.

Table 4 gives the performance of varieties at different sites. Introduced varieties Mugande, SPK013 and Kemb10 consistently outyielded the farmer's variety in all sites except at Nambale where the widely grown local variety Mar Ooko performed better than Kemb10 and SPK004. The lowest mean root yield of the varieties was recorded at Bukhalalire while the highest yields were observed at Butula (Fig. 1). Although the two sites have similar soil types, soils in Bukhalalire are relatively shallow compared to those in Butula. The performance of varieties in 1995, 1996 and 1998 was higher than that of 1997 and 1999 (Fig. 1). This may be attributed to differences in the amount of rainfall received during the trials. It was not possible to record the amount of rainfall received during this period since there were no adequate rain gauges.

Stability analysis based on total root yield (Table 5) indicated that SPK013 and Kemb10 were adapted to all environments while Mugande performed well under favourable environments. Variety SPK004 was also widely adaptable but had lower root yields than the general mean of varieties under evaluation. SPK004 is being popularised because of its high b-carotene content (K'osambo et al., 1998). All varieties had taste scores above 3.0 meaning that the cooking quality of the roots was acceptable to farmers. Although there were significant differences in storage root yield among varieties in each agro-ecological zone (Table 6 and Fig. 1), no significant differences were observed among zones. The high yields observed at the relatively low potential zone (LM3) could probably be attributed to the effect of the El-nino rains.

Table 7 gives selection criteria used by farmers in selecting sweetpotato varieties. Time of maturity and root yields were the most important selection criteria used by farmers. Most farmers prefer early maturing varieties, which provide food fast during hunger periods in the early months of the year and release land for other crops the following season. Taste and resistance to drought were also important. In Malava where sweetpotato production is commercialised, farmers preferred yellow -fleshed varieties with high market demand (Obiero et al., 1998). The ranking of varieties by farmers in terms of varietal preference and taste of cooked roots was not consistent among sites. Mugande, SPK013 and Kemb10 were considered the best varieties by farmers (Tables 8 and 9). The qualitative consideration given by the farmers on these varieties are listed in Table 10.

CONCLUSION

The results of this study indicated that introduced sweetpotato varieties were superior to the local varieties grown by farmers in western Kenya. Mugande, SPK013 and Kemb10 were preferred by farmers because of their high root yields, early maturity and good taste. SPK013 and Kemb10 were adaptable in all the sweetpotato growing areas. Mugande gave high yields in all sites although the yield was affected by changes in weather and soil conditions. Farmers were encouraged to grow the orange- fleshed variety SPK 004 because of its high β-carotene content. Involving farmers in the study, has improved interaction between researchers, extension staff, community-based organisations and farmers. The four CBO partners alone or in collaboration with the farmer groups, have established on-farm seed multiplication units in the wet river beds. This should enhance faster dissemination of the new varieties.

ACKNOWLEDGEMENT

The authors wish to thank the Government of the Netherlands, KARI-Agricultural Research fund,CIP and PRAPACE for their support.

REFERENCES

Anonymous, 1993. MSTAT-C. A microcomputer program for the design, management and analysis of agronomic research experiments. Michigan State University.

Carey, E.E., Gichuki, S.T., Ndolo, P.J., Turyamureeba, G. and Teri, J.M. 1997. In: Collaborative sweetpotato breeding in eastern, central and southern Africa. CIP Program Report 1995-1996. Lima, Peru. pp. 49-57.

Food and Agricultural Organisation (FAO), 1993. Agrostat PC database, Rome, Italy.

Food and Agricultural Organisation (FAO), 1995. Agrostat PC database. FAO, Rome, Italy.

K'oambo, L.M., Carey, E.E., Misra, A.K., Wilkes, J. and Hagenimana, V. 1998. Influence of age, farming site and boiling on pro-vitamin A content in sweetpotato (Ipomoea batatus (L.) Lam) storage roots. Journal of Food Competition and Analysis 11:305-321.

Lowe, S.B. and Wilson, L.A. 1994. Comparative analysis of tuber development in six sweetpotato (Ipomoea batatas Lam.) cultivars. 1. Tuber initiation, tuber growth and partitioning of assimilates. Annals of Botany 38:308-317.

Ngeve, J.M. 1993. Regression analysis genotype x environment interaction in sweetpotato. Euphtica 71:231-238.

Ndolo, P.J., Carey, E.E., Kamau, J.W., Maisiba, G., Lusweti, C., Gichuki, S.T., Ngun'gi, J., Irungu, J.W. and Maina, D.K. 1995. Multilocational testing of sweetpotato clones in Kenya. In: Proceedings of Symposium of the International Society for Tropical Root Crops-Africa branch. Malawi 22-28 October, 1995.

Smit, N.E.J.M. 1997. Integrated pest management in Eastern Africa. Ph.D. Thesis, Agricultural University, Wageningen, The Netherlands.

TABLE 1. Characteristics of clusters used in the sweetpotato adaptablity study

Site	Cluster	Altitude (m)	Annual rainfall (mm)	Temperature (°C)	Soils
Boro	LM2	1500-2000	1400-1600	21.4-23.0	Orthic to rhodic ferralsols
Sabatia	UM2	1500-2000	1000-2000	18.1-20.4	Orthic to chromic/orthis acrisols
Butula	LM1	1300-1400	1800-2000	20.5-21.7	Orthic ferralsols
Malava	LM2	1300-1500	1600-1800	20.9-22.0	Ferralo-orthic acrisols
Uranga	LM3	1200-1300	1000-12000	21.4-22.3	Pellic vertisols
Uranja	LM2	1300-1400	1400-1600	21.5-22.6	Orthic to rhodic ferralsols
Bukhalalire	LM2	1300-1400	1800-2000	20.9-22.0	Ferralo-orthic acrisols
Nambale	LM1	300-500	1700-2000	21.0-22.0	Orthic to rhodic ferralsols
Teso	LM1	1300-1500	1800-2000	21.0-22.0	Orthic ferralsols and orthic acrisols

TABLE 2. Combined analysis of variance for storage root yield of sweetpotato varieties planted in on-farm trials at nine sites in western Kenya.

Source	df	Mean sum of squares
		Marketable root yield	Total root yield
Year	3	209.0**	193.0**
Site	7	128.3**	146.5**
Variety	8	232.7**	263.0**
Year x Variety	12	20.4	23.6
Site x Variety	29	20.9	23.2

TABLE 3. Combined root yield, number of roots/plant, taste score, virus score and percent weevil infestation of four sweetpotato varieties planted at nine sites in western Kenya.

Variety	Marketable root yield (t ha-1)	Total root yield (t ha-1)	# of roots/plant	Virys score (1-5)	% weevil infestation	Tatse score (1-5)
Mugande	13.6a1	14.8a	4.2	..3	15.3	3.5
SPK013	11.3ab	12.4ab	3.1	1.0	18.7	3.2
Kemb10	9.8bc	11.0bc	3.5	1.2	40.5	4.7
SPK004	8.3bcd	9.6bcd	2.8	1.0	33.0	4.0
Local*	6.5cde	7.5cd	2.9	3.4	29.4	3.0
Mean	9.9	11.1	2.3	1.6	27.4	3.7
CV%	47.8	51.0	25.6	17.3	48.1

TABLE 4. Mean storage root yield of sweetpotato varieties planted at nine sites in western Kenya

Variety	Boro n=10	Bukhalalire n=9	Butula n=11	Malava n=9	Nambale n=5	Sabatia n=6	Teso n=12
Mugande	14.4a	9.9a	20.2a	12.1a	14.6ab	12.6ab	13.8a
SPK013	13.5a	7.1ab	15.9b	12.1a	10.8bc	15.4a	12.6a
Kemb10	12.2a	4.7bc	13.4bc	11.4a	8.0c	12.6ab	10.7a
SPK004	7.6b	5.3bc	12.3bc	9.0ab	9.5bc	8.5b	12.5a
*Marooko	5.7b	8.0ab	9.8c	6.3b	16.4a	-	12.9a
*Bungoma	-	5.6bc	-	-	-	8.4b	-
*Jayalo	-	2.6c	-	-	-	-	-
Pumpkin	-	2.6c	-	-	-	-	5.8b
Mean	10.6	6.1	14.3	10.2	10.7	11.5	11.8
CV%	40.7	40.7	34.4	36.3	37.1h	26.4	33.8

TABLE 5. Mean1 storage root yield of four experimental varieties in LM1, LM2, LM3 and UM2 agro-ecological zones in western Kenya

Variety	LM1	LM2	LM3	UM2
Mugande	14.9a	13.0a	15.9a	12.6a
SPK013	12.2ab	11.6ab	12.3ab	15.4a
Kemb10	10.0abc	10.2ab	11.9b	12.6a
SPK004	10.5abc	9.1bc	9.2b	8.5b
Mean	11.9	11.0	12.3	12.3
CV%	52.8	50.6	45.0	26.4

TABLE 6. Stability parameters for total storage root yield of four experimental varieties planted in nine cluster trial sites in western Kenya

Variety	Mean total root yield (t ha-1)	β1	δ2	R2
Mugande	13.6	1.14	2.64	0.75
SPK013	11.3	1.05	1.38	0.85
Kemb10	9.8	1.04	1.62	0.81
SPK004	6.5	0.80	1.90	0.68

TABLE 7. Criteria used by farmers in selecting sweetpotato varieties in four clusters in western Kenya

Creterion	Boro	Teso	Uranga	Bukhalalire	Butula	Nambale	Malava	Ugunja	Sabatia
Maturity	2	2	2	3	1	3	1	3	2
Yield	1	1	1	1	2	1	2	1	1
Ability of vines to withstand drought	3	4	4	2	4	2	4	2	2
Taste	4	3	3	4	3	6	3	5	4
In-ground storability	5	5	5	5	5	4	5	4	6
Resistance to pests	6	6	6	6	7	5	6	6	8
Market demand	7	7	7	7	6	7	2	7	5

TABLE 8. Overall ranking of four sweetpotato experimental varieties and the most popular local variety at each site

Variety	Uranga	Malava	Butula	Sabatia	Bukhalalire	Boro	Teso	Nambale	Uganja	Rank sum
Mugande	2	1	1	1	2	2	3	3	1	16
SPK013	1	2	2	2	4	1	4	5	-	21
Kemb10	3	3	3	3	1	4	1	1	2	21
SPK004	4	4	3	3	2	5	4	3	3	27
Local	4	5	5	4	5	3	2	4	3	35

TABLE 9. Relative ranking of four sweetpotato varieties based on taste tests

Variety	Uranga	Malava	Butula	Sabatia	Bukhalalire	Boro	Teso	Nambale	Uganja	Rank sum
Mugande	3	5	2	1	4	2	3	3	1	24
SPK013	1	2	4	2	3	4	5	5	-	26
Kemb10	1	1	1	3	1	1	2	2	2	12
SPK004	2	2	3	3	1	2	1	1	3	18
Local	4	4	5	4	2	5	4	4	3	35

TABLE 10. Qualitative information obtained from participating farmers about the four sweetpotato varieties

Variety	Good points	Bad points
Mugande	Early Maturing Big roots Withstands drought Performs well in fertile soils Very vigorous	Average taste Poor in-ground storability
SPK 013	Early maturing Big roots Good taste Good in-ground storability	Performs poorly in infertile soils Does not withstand drought
Kemb 10	Very early maturing Very good taste Attractive fresh colour when cooked	Very poor in-ground storability Very susceptible to weevils Does not withstand drought
SPK 004	Very attractive fresh colour Cooks fast Liked by children	Susceptible to weevils Poor yield Long and narrow roots Poor in-ground storability Vines do not withstand drought

Figure 1. Total root yield (t ha^-1) of sweetpotato varieties at different experimental sites in western Kenya.

Figure 2. Performance of sweetpotato varieties in different years in western Kenya

Figure 3. Performance of four sweetpotato varieties in LM1, LM2, LM3 and UM2 AEZs in western Kenya.

The following images related to this document are available:

Line drawing images